Inactivation of a Two-Component Signal Transduction System ... · Using microarray and real-time reverse transcriptase PCR analyses, we ... cell cultures were incubated at 37°C with

INFECTION AND IMMUNITY, Aug. 2006, p. 4655–4665 Vol. 74, No. 80019-9567/06/$08.00�0 doi:10.1128/IAI.00322-06Copyright © 2006, American Society for Microbiology. All Rights Reserved.

Inactivation of a Two-Component Signal Transduction System,SaeRS, Eliminates Adherence and Attenuates

Virulence of Staphylococcus aureusXudong Liang,1 Chuanxin Yu,1 Junsong Sun,1 Hong Liu,1 Christina Landwehr,1

David Holmes,2 and Yinduo Ji1*Department of Veterinary and Biomedical Sciences, College of Veterinary Medicine, University of Minnesota,

1971 Commonwealth Ave., St. Paul, Minnesota 55108,1 and Anti-Infective Research,GlaxoSmithKline Research and Development, 1250 S. Collegeville Rd.,

Collegeville, Pennsylvania 194262

Received 27 February 2006/Returned for modification 28 April 2006/Accepted 9 May 2006

Staphylococcus aureus is a major human and animal pathogen. During infection, this organism not only isable to attach to and enter host cells by using its cell surface-associated factors but also exports toxins toinduce apoptosis and kill invaded cells. In this study, we identified the regulon of a two-component signaltransduction system, SaeRS, and demonstrated that the SaeRS system is required for S. aureus to causeinfection both in vitro and in vivo. Using microarray and real-time reverse transcriptase PCR analyses, wefound that SaeRS regulates the expression of genes involved in adhesion and invasion (such as those encodingfibronectin-binding proteins and fibrinogen-binding proteins) and genes encoding �-, �-, and �-hemolysins.Surprisingly, we found that SaeRS represses the Agr regulatory system since the mutation of saeS up-regulatesagrA expression, which was confirmed by using an agr promoter-reporter fusion system. More importantly, wedemonstrated that inactivation of the SaeRS system significantly decreases the bacterium-induced apoptosisand/or death of lung epithelial cells (A549) and attenuates virulence in a murine infection model. Moreover,we found that inactivation of the SaeRS system eliminates staphylococcal adhesion and internalization of lungepithelial cells. We also found that both a novel hypothetical protein (the SA1000 protein) and a bifunctionalprotein (Efb), which binds to extracellular fibrinogen and complement factor C3, might partially contribute tobacterial adhesion to and invasion of epithelial cells. Our results indicate that activation of the SaeRS systemmay be required for S. aureus to adhere to and invade epithelial cells.

Staphylococcus aureus is an important community- and hos-pital-acquired pathogen that can cause serious disease, includ-ing skin and soft tissue lesions, as well as life-threatening in-fections such as pneumonia, endocarditis, and toxic shocksyndrome (38, 49). This organism’s ability to cause such avariety of diseases partially depends on the expression of itsmany virulence factors, such as surface-associated adhesins(17), a polysaccharide capsule, and a range of extracellularcytotoxins, proteases, DNases, and enterotoxins (27). Most ofthe virulence factors have been found to be controlled differ-entially by different two-component signal transduction regu-latory systems (TCS), such as Agr (45), ArlRS (18, 37), andSaeRS (20), and global regulators, such as SarA (9), Rot (41,53), and Mgr (39). Therefore, TCSs have been implicatedtogether with other regulators to play an important role inbacterial pathogenesis (45).

The well-studied regulatory system in S. aureus is the Agrquorum sensing system, which is composed of the AgrBDCAstructural genes and RNAIII, the effector molecule of the agrlocus (45). The agrB gene encodes a membrane-associatedprotease required for modifying the prepropeptide of AgrDand generating small peptide signaling molecules. The peptide

signal molecules can be recognized by the membrane-associ-ated sensor kinase (AgrC), which subsequently activates theresponse regulator (AgrA), which up-regulates RNAIII pro-duction (28, 62). The expression of RNAIII is temporal, withmaximal expression occurring in the transition from the post-exponential to the stationary phase. RNAIII is a dual regula-tor of staphylococcal virulence factors, including exoproteinsand surface proteins (e.g., protein A, coagulase, and someadhesins) (45). Recent studies suggested that Agr positivelyregulates cap5 expression both in vitro and in vivo (57). OtherTCS loci, such as arlRS (18), also positively regulate the ex-pression of virulence factors, such as Ser-Asp-rich bone sialo-protein-binding proteins, and repress some exported proteins,including cysteine protease, serine protease, the hla gene prod-uct, �-hemolysin, and leukotoxins (37). The srrAB system isinvolved in the adaptation to anaerobic growth of S. aureus andin the regulation of virulence factors such as toxic shock syn-drome toxin 1 and protein A (61).

In this paper, we investigate the SaeRS system, anotherimportant two-component signal transduction system involvedin the control of virulence gene expression (20, 23, 55). Theresponse regulator SaeR is similar to other regulatory proteins,such as DrrA from Thermotoga maritima and ResD and PhoPfrom Bacillus subtilis (20). The N-terminal region of SaeRcontains a highly conserved aspartate phosphorylation sitecommonly found in response regulators (6). The C terminus ofthe histidine protein kinase SaeS has an autophosphorylated

* Corresponding author. Mailing address: Department of Veterinaryand Biomedical Sciences, College of Veterinary Medicine, University ofMinnesota, 1971 Commonwealth Ave., St. Paul, MN 55108. Phone: (612)624-2757. Fax: (612) 625-5203. E-mail: [email protected].

4655

on October 22, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

on October 22, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

on October 22, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

on October 22, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

histidine residue that is similar to other sensors of histidinekinase, such as PhoR and YkoH from B. subtilis and VanSfrom Enterococcus faecium (20). The N-terminal region of theSaeS protein possesses two transmembrane domains that arealso found in other sensor proteins (6). It has been reportedthat the expression of sae is repressed in the presence of glu-cose as a consequence of changes in the pH (46). The SaeRSsystem up-regulates the transcription of hla, hlb, and coa (20)but has no effect on the expression of agr or sarA in vitro (21).Moreover, the SaeRS system also controls hla expression invivo, as the level of hla transcripts is significantly decreased insae mutant strains during infection (22, 23). The SaeRS systemmight function independently as a regulator since SaeRS-de-pendent and Agr/SarA-independent activation of hla wasfound in exudates accumulated in a guinea pig model of infec-tion (22, 23). It was reported that mutation of sae in S. aureusstrain Newman eliminates the transcription and expression offnbA and increases the expression of CP5; it also leads to asignificant decrease of the internalization of S. aureus by en-dothelial cells (56). Moreover, the role of the SaeRS systemas a virulence regulator has been demonstrated in severalanimal models of infection (3, 23, 51). Therefore, the SaeRSregulatory system appears to play an important role in themodulation of virulence gene expression during certaintypes of infection.

In this study, we aimed to examine more specifically theroles of the SaeRS system in bacterial pathogenesis. We cre-ated a saeS gene replacement mutant in an S. aureus isolatefrom a human and identified the regulon of saeRS using Af-fymetrix S. aureus oligonucleotide arrays. We also examinedthe impact of the mutation of saeS on bacterial adherence andinternalization in epithelial cells, S. aureus-induced apoptosisand death, and survival and/or virulence by using a murineinfection model. Our microarray data indicate that SaeRS mayaffect the expression of a bifunctional protein (Efb) whichbinds to both extracellular fibrinogen (48) and complementfactor C3 (36) and a gene encoding a hypothetical fibrinogen-binding protein (the SA1000 protein). In order to understandwhether these proteins are involved in pathogenesis, we con-structed allelic gene replacement mutants and examined theimpact of mutation on bacterial adhesion and invasion.

MATERIALS AND METHODS

Bacterial strains, media, and growth conditions. S. aureus strain RN4220 wasutilized as the primary recipient for allelic exchange constructs, together withstrain WCUH29 (NCIMB40771), a virulent clinical isolate, as a secondary re-cipient for phage transduction. S. aureus strain 15981 and the 15981 �saeRSstrain were kindly provided by I. Lasa (56). Escherichia coli DH10B (Invitrogen)served as the host for all in vitro recombinant DNA. Bacteria were grown intryptic soy broth (TSB) (Difco) and on tryptic soy agar (TSA) at 37°C. Bacterialcell cultures were incubated at 37°C with shaking at 200 rpm. The Sa371komutant was maintained on TSA plates with tetracycline (Tc) at a concentrationof 5 �g/ml.

Construction of saeS, SA1000, and efb gene replacement null mutants, saeS,SA1000, and efb gene complementary strains, and agrA promoter-gfp reporterfusion system. The S. aureus vector pSA7755 was used to generate gene replace-ment mutants as described previously (15). A cassette containing the tet gene(flanked by chromosomal fragments from upstream and downstream of the saeS,SA1000, or efb gene to be replaced) was constructed in pBluescript and insertedinto pSA7755. An S. aureus strain, RN4220, carrying this hybrid plasmid wasgrown at a restrictive temperature (40°C) for pSA7755 replication in the pres-ence of 5 �g/ml tetracycline. These conditions allow the growth of only cells withtet inserted into the chromosome by homologous recombination. To obtain gene

replacement, the mutated locus was transduced into the wild-type strainWCUH29 by �11 transduction. We determined that allelic replacement hadoccurred and resulted in the saeS null mutant strain Sa371ko, the SA1000 mutantstrain Sa1000ko, and the efb null mutant strain Efbko by selecting for tetracyclineresistance and screening for the loss of the erythromycin resistance markercarried by the vector. The mutation in saeS, SA1000, or efb was verified by PCRusing primers specific to the saeS, SA1000, efb, or tet gene. The results showedthat no PCR product from the saeS, SA1000, or efb gene was obtained from themutant strains, whereas there was a PCR product of the expected size obtainedfrom the wild-type strain using primers specific to saeS, SA1000, or efb (data notshown). In contrast, a PCR product was obtain from the mutant strains (but notthe wild-type strain) by using a primer specific to tet. To further confirm the saeS,SA1000, or efb mutation, we performed Southern blot analysis using digoxigenin(DIG)-labeled probes and found that no DNA hybridized with the saeS, SA1000,or efb probe in the mutant chromosomal DNA (Fig. 1).

In order to examine whether the expression of saeS, SA1000, or efb in trans cancomplement the effect of mutation of the corresponding endogenous gene, weconstructed the recombinant plasmids pYH4/saeS, pYH4/SA1000, and pYH4/efbby cloning the saeS, SA1000, and efb coding regions (obtained by PCR), respec-tively, into the AscI and PmeI sites of pYH4 (31) and electroporated them intothe Sa371ko, Sa1000ko, and Efbko strains, resulting in the Sa371com,Sa1000com, and Efbcom strains, respectively. The recombinant plasmid DNAsof pYH4/saeS, pYH4/SA1000, and pYH4/efb were isolated from the complemen-tary strains and confirmed by PCR and DNA sequencing (data not shown).

In order to confirm whether the mutation of saeS has an impact on agrAexpression, we utilized an agrA promoter-gfp reporter fusion system, which wasa kind gift from Philip Hill. Plasmid DNA was purified and electroporated intoS. aureus WCUH29 and Sa371ko, resulting in the WCUH29/pCY1006 andSa371ko/pCY1006 strains, respectively. gfp expression was determined by West-ern blotting as described previously (31), using a green fluorescent protein (GFP)antibody (Abcam Inc., MA).

RNA isolation and purification. Overnight cultures of S. aureus were inocu-lated into 5% TSB medium and grown to the mid-exponential (3 h) phase ofgrowth. Cells were harvested by centrifugation, and RNAs were isolated by useof an RNAPrep kit (Promega, MI). Contaminating DNA was removed with aDNA-free kit (Ambion), and the RNA yield was determined spectrophotometri-cally at 260 nm.

cDNA synthesis, cDNA fragmentation, and labeling. The integrity of the RNApreparations was analyzed by electrophoresis in 1.2% agarose-0.66 M formalde-hyde gels. The 23S and 16S rRNA bands were clear, without any obvious smear-ing patterns. Briefly, a total of 10 �g of RNA was reverse transcribed to generatecDNA, using Superscript II reverse transcriptase (RT) and random primers(Invitrogen). The RNA was then removed by treatment at 65°C for 30 min withNaOH. The cDNA was purified by using a QIAquick PCR purification kit(QIAGEN). The purified cDNA was digested with DNase I and labeled withbiotin-ddUTP (Roach).

Hybridization and scanning. After determining that the fragmented cDNAswere labeled with biotin, the fragmented biotinylated cDNAs were hybridized toS. aureus chips (Affymetrix) containing probe sets for S. aureus genomic openreading frames (ORFs), and hybridization intensities for each of the genes/transcripts were collected from the scanned images.

Microarray analysis. The S. aureus array (Affymetrix) contained probe sets forover 3,300 S. aureus open reading frames based on the updated S. aureusgenomic sequences of strains N315, Mu50, NCTC 8325, and COL. Additionally,the array also contained probes to study both the forward and reverse orienta-tions of over 4,800 intergenic regions throughout the S. aureus genome. Asubsequent analysis suggested that these probes represent approximately 2,738,2,668, 2,773, and 2,810 individual genes of the S. aureus COL, N315, NTC8325,and Mu50 genomes, respectively.

To identify genes with significantly altered expression, microarray analyses anda series of statistical analyses (filtering) were performed as described previously(37). We selected the genes with significant differential expression (P � 0.05).Those genes negatively regulated by the SaeRS system were identified as ORFswith transcript titers at least 1.8-fold higher in Sa371ko (saeS negative) than inWCUH29. Genes whose transcript levels were at least 1.8-fold higher inWCUH29 than in Sa371ko (saeS negative) were categorized as being positivelyregulated by SaeRS.

RT-PCR and quantitative real-time RT-PCR analysis. In order to examinewhether the complementary strains expressed the knockout genes in trans, weperformed RT-PCR, using the saeS-, SA1000-, or efb-specific primers listed inTable 1, as described previously (29). In order to confirm the results obtainedfrom the microarray analyses, we employed quantitative real-time RT-PCR tocompare the RNA levels that showed significant changes of expression in the

4656 LIANG ET AL. INFECT. IMMUN.

on October 22, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

microarray assay (37). The first-strand cDNA was synthesized using reversetranscriptase with a SuperScript III Platinum two-step qRT-PCR kit (Invitro-gen). For each RNA sample, duplicate reverse transcription reactions wereperformed, as well as a control without reverse transcriptase, in order to deter-mine the level of DNA contamination. PCRs were set up in triplicate by usingSYBR green PCR master mix (Bio-Rad). Real-time sequence-specific detection

and relative quantitation were performed with the Stratagene Mx3000P real-timePCR system. Gene-specific primers were designed to yield �100-bp specificproducts (Table 1). Relative quantification of the product was calculated usingthe comparative cycle threshold method as described for the StratageneMx3000P system. The housekeeping 16S rRNA gene was used as an endogenouscontrol (37). All samples were analyzed in triplicate and normalized against 16SrRNA gene expression. The experiments were repeated at least twice and ana-lyzed for correlation to the microarray results.

Cell culture and epithelial cell adhesion and invasion assay. A549 human lungepithelial cells (ATCC CCL 185) were cultured in RPMI 1640 medium supple-mented with 10% fetal bovine serum (FBS; Invitrogen). Cultures of A549 cellswere maintained in a medium containing penicillin (5 �g/ml) and streptomycin(100 �g/ml) (Sigma).

Assays of bacterial invasion and adherence were performed as previouslydescribed (1, 10, 57). Assays were performed in RPMI 1640 medium supple-mented with of 10% FBS (RPMI-FBS). We used 0.025% Triton X-100 for bettercell lysis. Briefly, 1 day prior to infection, approximately 2 � 105 cells wereseeded in each well of 24-well plates and incubated overnight at 37°C in a CO2

incubator. Monolayers of A549 cells (2 � 105 cells/well) were infected by adding0.5 ml RPMI containing approximately 5 � 105 CFU of bacteria, followed bycentrifugation at 100 � g for 5 min, and were incubated for 1 h at 37°C in 5%CO2. To measure bacterial adherence, the culture medium was removed frommonolayers 1 h after infection and discarded. The monolayer cells were thenwashed three times with phosphate-buffered saline (PBS; pH 7.4) to removenonadherent bacteria. Epithelial cells were dispersed by the addition of 150 �l of0.25% trypsin-1 mM EDTA (Invitrogen) and then lysed by the addition of 400 �lof 0.025% Triton X-100. The numbers of bacterial CFU released from the lysedepithelial cells were determined by plating of diluted lysates on TSA plates. Forinvasion assays, the culture medium was collected from wells used for total

FIG. 1. Schematic diagrams and Southern blot analysis of allelic gene replacement mutants. (A) Chromosomal DNAs isolated from WCUH29(lane 2) and Sa371ko (lane 3) were digested with XbaI (X) and probed with DIG-labeled saeS. As expected, a 2.8-kb DNA fragment fromWCUH29 was detected in the Southern blot. (B) Chromosomal DNAs isolated from WCUH29 (lane 2) and Sa1000ko (lane 3) were digested withHindIII (H) and probed with DIG-labeled SA1000. As expected, a 1.9-kb DNA fragment from WCUH29 was detected in the Southern blot.(C) Chromosomal DNAs isolated from WCUH29 (lane 2) and Efbko (lane 3) were digested with PstI (P) and probed with DIG-labeled efb. Asexpected, a 2.5-kb DNA fragment from WCUH29 was detected in the Southern blot. Lanes M, 1-kb DNA markers labeled with DIG.

TABLE 1. Primers used in real-time RT-PCR

Primer Sequence (5–3)

SA1000for ................GTATCAACGTTTGCCGGTGAATCTCSA1000rev ................CAGCTCTTTGTGCTTTACGGTGTGTTSA1003for (Efb)......GTACAATGATGGTACTTTTAAATATCAAT

CTAGACSA1003rev (Efb) .....GTTCTTTTTTAATAGTTGCATCAGTTTT

CGCTSA1004for ................AAGGGAATAAAGCAGATGCAAGTAGTCTSA1004rev ................GTGCCGCTTTAGCTCTATATTCATTCATSA1007for ................CAACTGATAAAAAAGTAGGCTGGAAAG

TGATSA1007rev ................CTGGTGAAAACCCTGAAGATAATAGAGSA1844agrAfor........GTGAAATTCGTAAGCATGACCCAGTTGSA1844agrArev .......TGTAAGCGTGTATGTGCAGTTTCTAAACSA2290 fnbBfor.......GCAGTGAGCGACCATACAACAGTTSA2290 fnbBrev ......CAATCACGCCATAATTACCGTGACCA16S rRNAfor ...........CTGTGCACATCTTGACGGTA16S rRNArev...........TCAGCGTCAGTTACAGACCA

VOL. 74, 2006 EFFECTS OF SaeRS INACTIVATION ON S. AUREUS 4657

on October 22, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

counts and discarded from wells used for invasion assays after 2 h of incubation.The monolayer cells were washed three times with PBS (pH 7.4), followed by theaddition of 1 ml of RPMI-10% FCS containing 100 �g/ml gentamicin and 5�g/ml lysostaphin to invasion wells and 1 ml of RPMI-10% FBS to total wells,and were then incubated. After 2 h, the supernatants in the wells were removedand discarded. All wells were washed three times with 1 ml PBS. A total of 1 mlRPMI-10% FBS containing100 �g/ml gentamicin and 5 �g/ml lysostaphin wasadded to the invasion wells to kill outside bacteria, and 1 ml of RPMI-10% FBSwas added to the wells used for total counts. The supernatant was removed fromeach well after 1 h of incubation. All wells were washed three times with 1 mlwarm PBS, and then 150 �l of 0.25% trypsin-EDTA was added. After 5 min, thecells in each well were carefully collected, and then 400 �l of 0.025% TritonX-100 was added to the tubes on ice. The numbers of bacterial CFU releasedfrom the lysed epithelial cells were determined by plating of diluted lysates onTSA plates.

The bacterial adhesion in each well was determined as the CFU that adheredto and invaded into the cells and is expressed as a percentage of the CFU in theinoculum (58). The controls were wells pretreated with medium alone (RPMI) orwere the wild-type control strains, considered to have 100% adhesion. Adhesionand invasion were then normalized against controls (1, 58). Each experiment wasrepeated three times, and all of the relative adhesion and invasion values werecalculated and statistically analyzed by Student’s t test, using Microsoft Excel2003 software. P values of �0.05 were considered significant.

DNA fragmentation assay. Infected and control cells were collected, lysed withlysis buffer, and treated with RNase (10 mg/ml) and proteinase K (20 mg/ml).DNAs were purified from the epithelial cells 24 h after infection with the saeSmutant or its parent strain by using phenol, precipitated with ethanol, andwashed with 75% ethanol. DNAs were dissolved in 100 �l Tris-EDTA buffer,followed by electrophoresis using 1.2% agarose gels (43).

Cytotoxicity assays. All cells were grown in 96-well plates to 70% confluence.Inhibitors were diluted serially in complete medium and applied to the cells. Thesupernatants from the overnight cultures of WCUH29 and Sa371ko were imme-diately applied to the cells, and the treated cells were incubated at 37°C with 5%CO2 overnight (16 h). At the end of the experiment, cell viability was determinedby using the CellTiter 96 aqueous nonradioactive cell proliferation assay (Pro-mega) according to the manufacturer’s instructions.

Hematogenous pyelonephritis infection. Overnight cultures of bacteria werestarted from single colonies in 10 ml of TSB and grown at 37°C with shaking.Cells were centrifuged and washed twice in PBS. The A600 was adjusted toapproximately 0.2. Female CD-1 mice (18 to 20 g) were inoculated with 0.2 mlof this suspension (containing approximately 7 log10 CFU bacteria) by tail veininjection (29, 30). Mice were monitored twice daily for signs of illness, and anywhich appeared moribund were euthanized prior to the end of the experiment.All remaining animals were euthanized by carbon dioxide overdose at 3 dayspostinoculation. Both kidneys were removed, using aseptic technique, and ho-mogenized in 1 ml PBS before enumeration of viable bacteria following platingon TSA plates.

RESULTS

Construction and characterization of saeS gene replacementmutant of a human clinical S. aureus isolate. In order toexamine the effect of the SaeRS system on bacterial pathogen-esis, we created a saeS gene replacement mutant of the humanclinical isolate WCUH29, designated Sa371ko, by the homol-ogous gene recombination strategy described in Materials andMethods. To confirm the allelic replacement, chromosomalDNAs were isolated from the mutant Sa371ko and wild-typeWCUH29 strains, and diagnostic PCR analysis (see Materialsand Methods) and Southern blot analysis were used to confirmthe mutation (Fig. 1A). The results showed that the saeS genewas knocked out by allelic gene replacement (Fig. 1A).

To determine whether the allelic mutation of saeS had anyimpact on bacterial growth and morphology, we measured thegrowth curve of the saeS mutant strain in nutrient-rich TSBmedium with Tc and the CFU on TSA-Tc. The results showedthat there was no difference in growth pattern and CFU be-tween the saeS knockout strain and its parent strain. We did

not observe obvious differences in the morphological pheno-type, such as colony size and shape, after the mutation of saeS.

The SaeRS system regulates the expression of genes in-volved in adhesion, invasion, and toxicity in vitro. Previousstudies have demonstrated that the SaeRS system is involved inregulating the expression of several virulence genes, includinghla, hlb, and coa (21, 22). In order to comprehensively under-stand the saeRS regulon, we performed a microarray analysisto examine the effect of the mutation of saeS on gene tran-scription, using Affymetrix S. aureus oligonucleotide chips asdescribed in Materials and Methods. Unlike other two-com-ponent signal transduction systems, such as agr (12) and arl(37), where the mutation of agr or arl has a global effect on theexpression of numerous genes, our microarray data showedthat the mutation of saeS significantly affected the expressionof fewer than 20 genes (Table 2). The microarray data indi-cated that the SaeRS system is involved in positive regulationof genes encoding fibronectin-binding proteins (FnBPs) andfibrinogen-binding proteins, such as fnbB, fnb, and coa (themRNA levels in the wild type were 10-, 3-, and 7-foldhigher, respectively, than those in the null mutant), as well asgenes encoding a putative hypothetical protein, the SA1000protein, and a bifunctional protein (Efb) (the mRNA levels inthe wild type were 50- and 35-fold higher, respectively,than those in the null mutant) (Table 2). To validate theseresults, a real-time RT-PCR was performed, and the resultsdemonstrated that the levels of fnbB, SA1000, efb, and SA1004mRNAs in the wild-type strain were significantly higher thanthose in the saeS mutant strain (Table 3). We also found thatSaeRS is involved in the regulation of genes encoding toxins,including hla, hlb, hlgC, and set-15 (the levels of mRNA in thewild type were 23-, 4-, 1.7-, and 75-fold higher, respec-tively, than those in the null mutant) (Table 2). The observa-tion that SaeRS positively affected hla and set-15 expressionwas confirmed by a real-time RT-PCR analysis in which RNAsamples from the saeS mutant and the wild-type strain weretaken at the mid-exponential growth phase. The results dem-onstrated that hla and set-15 mRNAs were significantly higherin the wild-type strain than in the saeS mutant strain (Table 3).In addition, we found that SaeRS positively affects severalgenes with unknown functions (Table 2). Importantly, our mi-croarray and RT-PCR results indicate that even the magni-tudes of the changes of the identified gene mRNA levels in thesaeS mutant strain were highly consistent between the microar-ray and real-time RT-PCR data.

Very surprisingly, the microarray results indicated that theSaeRS system negatively affects agrA (Table 2). In addition,the mutation of saeS increased agrB (1.5- fold) and agrC (1-fold) expression (data not shown). Real-time RT-PCR analysisconfirmed that the agrA mRNA level was significantly lower inthe wild-type strain than in the null mutant (Table 3). Tofurther confirm the effect on agr expression, we introduced anagrA promoter-gfp reporter fusion and determined the impactof the mutation in saeS on agr expression by monitoring thereporter gene. The results showed that the production of GFPwas significantly up-regulated in the saeS null mutant(Sa371ko/pCY1006) compared to that in the wild-type control(WCUH29/pCY1006) (Fig. 2). Taken together, these data in-dicate that SaeRS may repress the agrA regulatory system instrain WCUH29.

4658 LIANG ET AL. INFECT. IMMUN.

on October 22, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

Inactivating the SaeRS system eliminates staphylococcaladhesion and internalization of epithelial cells. It has beenreported that the SaeRS system plays an important role instaphylococcal adhesion and internalization into endothelialcells (55). Our genomic mRNA analysis also indicated that theSaeRS system regulates gene products that are associated withbacterial adhesion and/or invasion. In order to more specifi-cally characterize the role of the SaeRS system in the bacteri-um’s interaction with epithelial cells, we examined the impactof the saeS null mutation on adhesion to and/or internalizationof lung epithelial cells. We found that the adherence of thesaeS mutant strain was significantly less than that of the parentstrain (Fig. 3A). Moreover, compared to the parent cells, lessthan 10% of the saeS mutant cells were able to internalize into

epithelial cells after 2 h of infection (Fig. 3B). To furtherconfirm the role of SaeRS in bacterial adherence and invasion,we examined the effect of the saeRS null mutation on adhesionto and invasion of epithelial cells. The results were similar tothose for the saeS null mutant in that both adhesion andinternalization of the null mutant (strain 15981 �saeRS) weresignificantly less than those of the parent strain (15981; datanot shown).

To investigate whether the expression of saeS in trans cancomplement the effect of the mutation of endogenous saeS, weconstructed a recombinant plasmid, pYH4/saeS, containing thesaeS coding region and transformed it into the Sa371ko strain.First, we examined saeS expression by RT-PCR using saeS-specific primers. The results showed that a specific saeS PCRproduct was yielded with RT; in contrast, no PCR product wasdetected in the reaction without RT (Fig. 3C). This indicatesthat there is saeS expression in the pYH4/saeS complementaryconstruct. We then examined the influence of the plasmid-borne saeS gene on adherence and internalization. The results

FIG. 2. Western blot analysis of gfp expression in an agrA promoter-gfp fusion. S. aureus strains were incubated in TSB. The same amountsof bacterial cells were harvested from cultures at an optical density at600 nm of 0.5 (log phase) by centrifugation. Whole-cell lysates wereprepared, and the same amounts of protein were loaded into 12%sodium dodecyl sulfate-polyacrylamide gels and probed with rabbitanti-Gfp antiserum in a Western blot assay. Lane 1, Sa371ko/pCY1006; lane 2, WCUH29/pCY1006; lane 3, WCUH29.

TABLE 2. S. aureus genes regulated by SaeRS

N315 ORF Gene Description Change (fold)a agr, sar, rot, and/or sigB effectb

Cell-associated protein genesSA0222 coa Coagulase precursor �23.4 � (agr, sarA, rot, and sigB)SA2290 fnbB Fibronectin-binding protein B �10.3 � (sarA)SA2291 fnb Homologue of FnbA �3.2 � (sarA and sigB)SA1003 efb Extracellular fibrinogen-binding protein (36) �35.1

Exported protein genesSA0393 set-15 Exotoxin 15 �75SA0746 Staphylococcal nuclease �3SA1007 hla Alpha-hemolysin �23.4 � (agr, sarA, rot, and sigB)SA1752 hlb Beta-hemolysin �4.6SA2208 hlgC Gamma-hemolysin component C �1.7

Regulatory system protein genesSA0660 saeS Histidine kinase sensor �176SA1844 agrA Response regulator �2.2

Hypothetical protein genesSA0394 Hypothetical protein �27SA0743 Similar to staphylocoagulase precursor �4.5SA0841 Similar to cell surface protein MapW �2.4SA1000 Hypothetical fibrinogen-binding protein �58.7SA1004 Hypothetical protein �16.9SA1813 Similar to leukocidin LukM precursor �5.7SA1804 Hypothetical transcriptional regulator �34

a Normalized values for the wild-type strain over those for the saeS null mutant.b agr and sar effects were described by Dunman et al. (12), rot effects were described by Saıd-Salim et al. (53), and sigB effects were described by Bischoff et al. (4).

�, up-regulation; �, down-regulation.

TABLE 3. Real-time RT-PCR analysis of geneexpression regulated by SaeRS

N315 ORF Gene DescriptionChange (fold)a

RT-PCR Microarray

SA2290 fnbB Fibronectin-bindingprotein B

�5.2 �10.3

SA1000 Hypothetical fibrinogen-binding protein

�69.6 �58.7

SA1003 efb (36) Extracellular fibrinogen-binding protein

�33.6 �35.1

SA1004 Hypothetical protein �22.6 �16.9SA0390 set-15 Exotoxin �89.3 �75SA1007 hla Alpha-hemolysin �4.8 �23.4SA1844 agrA Two-component

response regulator�11.2 �2.2

a Normalized values for the wild-type strain over those for the saeS nullmutant. Positive numbers denote up-regulation in the wild-type strain.

VOL. 74, 2006 EFFECTS OF SaeRS INACTIVATION ON S. AUREUS 4659

on October 22, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

showed that the bacterium’s capacity to adhere to and invadeepithelial cells was restored fully (Fig. 3A and B).

A putative fibrinogen-binding protein and the exportedfibrinogen-binding protein are involved in adherence and in-ternalization. It is well known that FnBPs are required for theinternalization of S. aureus by endothelial cells (47). We se-lected two genes (SA1000 and efb), encoding a hypotheticalprotein and a bifunctional protein, Efb (36) (which binds ex-tracellular fibrinogen and complement factor C3), respectively,due to their apparent regulation by SaeRS from our microar-ray data and investigated whether they are involved in bacterialadhesion and invasion. We created SA1000 and efb allelic genereplacement null mutants as described in Materials and Meth-ods and confirmed the replacement mutations by PCR (seeMaterials and Methods for details) and Southern blot analysis(Fig. 1B and C). We then examined the adhesion to and inva-sion of epithelial cells by the mutant strains Sa1000ko andEfbko. The results showed that both adhesion and internaliza-tion of the null mutant strains (Sa1000ko and Efbko) weresignificantly less than those of the parent strain (Fig. 4A to D).These results indicate that the putative hypothetical proteinand Efb may be involved in adhesion to and invasion of epi-thelial cells.

To investigate whether the expression of SA1000 or efb intrans can complement the effect of the mutation of endogenousSA1000 or efb, we constructed the recombinant plasmidspYH4/SA1000 and pYH4/efb and transformed them into theSa1000ko and Efbko strains, respectively. The RT-PCR resultsshowed that specific SA1000 and efb PCR products were

yielded with RT; in contrast, no PCR products were detectedin the reaction mixtures without RT (data not shown). More-over, the adhesion and invasion results showed that the bacte-ria’s capacity to adhere to and invade epithelial cells was re-stored fully (Fig. 4A to D). These results indicate that theputative SA1000 protein and the extracellular fibrinogen-bind-ing protein Efb also affect pathogen-host cell interactions, al-beit with an effect that is not as pronounced as that of the saeSknockout.

Inactivation of the SaeRS system inhibits S. aureus-inducedapoptosis of human lung epithelial cells. It has been reportedthat alpha-toxin is involved in S. aureus-induced apoptosisand cell death through certain host cell signaling pathways(2, 14, 24, 33). Previous studies as well as our microarrayresults demonstrated that the SaeRS system regulates hlagene expression (21, 22). Thus, it is likely that the SaeRSregulatory system may be involved in the S. aureus-inducedapoptosis of host cells. To test this hypothesis, we examinedthe effect of the null mutation in saeS on S. aureus-inducedapoptosis, using a standard DNA fragmentation assay. Weobserved no ladder fragmentation pattern of chromosomalDNA for cells infected with the saeS null mutant strain (Fig. 5). Incontrast, distinctive ladder patterns of chromosomal DNAwere observed for cells infected with the parent strain (Fig. 5).We then quantitatively measured the cell death induced by S.aureus, using a standard cytotoxic assay. The results showedthat �55% of the epithelial cells infected by the wild-typestrain survived 26 h of infection, whereas 95% of the epithe-

FIG. 3. (A and B) Effects of SaeRS on adherence to and internalization of S. aureus by epithelial cells. The bacterial cells were collected fromovernight cultures (wild-type strain WCUH29, saeS null mutant strain Sa371ko, and the complementary strain Sa371com with plasmid-borne saeS),washed with PBS, and diluted and resuspended in RPMI 1640 medium with 10% FBS just prior to infection of monolayer cells. (A) Staphylococcaladherence to A549 cells. (B) Intracellular invasion. Relative adherence and relative invasion were calculated as described in Materials andMethods. Data are the means � standard errors of the means for nine infected monolayers of cells (three experiments, each performed intriplicate). Asterisks indicate significant differences between the wild type and the mutant (P � 0.01). (C) RT-PCR detection of saeS expressionin trans. S. aureus strains were incubated overnight in TSB with appropriate antibiotics, and total RNAs were purified from the above cultures andtreated with DNA-free kits. RT-PCR was performed using saeS- and 16S rRNA-specific primers. RT-PCR products from 16S rRNA were usedas positive controls. The negative controls were samples prepared without RT or template DNA. M, 100-bp DNA marker.

4660 LIANG ET AL. INFECT. IMMUN.

on October 22, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

lial cells infected by the saeS mutant strain survived the sameperiod of infection (Fig. 6A).

To determine whether the higher survival rate of the epi-thelial cells infected by the mutant may be due at least in partto the reduced levels of exported toxins, we examined the effectof SaeRS on cytotoxicity, using the supernatants of S. aureuscultures. In correlation with the whole-cell infection results,the supernatants of the saeS mutant strain showed no detect-

able toxicity to epithelial cells, whereas the supernatants of theparent strain caused 50% cell death 24 h after being exposedto the cells (Fig. 6B).

To investigate whether the expression of saeS in trans cancomplement the effect of the mutation of endogenous saeS,we examined the influence of the plasmid-borne saeS geneon S. aureus-induced cell death. The results showed that thebacterium’s capacity to cause the death of epithelial cellswas restored fully (Fig. 6A and B, Sa371com). Taken to-gether, the above results demonstrate that the SaeRS systemis important for S. aureus-induced death of epithelial cells(A549).

The SaeRS system significantly affects bacterial survivalduring infection. To investigate the role of SaeRS in patho-genesis, we chose a murine model of hematogenous pyelone-phritis. This model represents a localized kidney infection fromwhich bacteria can be recovered and quantitated and has beenused successfully to examine the essentiality of genes in vivo(29, 30). We infected mice via the tail vein, removed theirkidneys at 3 days postinfection, and examined the bacterialloads. We compared the null mutant strain with the wild typeas a control, using identical bacterial CFU. The results showedthat approximately 4 log10 CFU of the saeS null mutant wererecovered from infected kidneys, whereas, �6 log10 CFU ofthe wild-type strain were recovered from kidneys (Fig. 7).These results indicate that the mutation in saeS significantly

FIG. 4. Effects of the putative fibrinogen-binding SA1000 protein (A and B) and the extracellular fibrinogen-binding protein Efb (C and D)on adherence to and internalization of a human clinical S. aureus isolate by epithelial cells. The bacterial cells were collected from overnightcultures (wild-type strain WCUH29, the SA1000 null mutant strain Sa1000ko, the SA1000 complementary strain Sa1000com, the efb null mutantEfbko, and the efb complementary strain Efbcom), washed with PBS, and diluted and resuspended in RPMI 1640 medium with 10% FBS just priorto infection of monolayer cells. (A and C) Staphylococcal adherence to A549 cells. (B and D) Intracellular invasion. Relative adherence andrelative invasion were calculated as described in Materials and Methods. Data are the means � standard errors of the means for nine infectedmonolayers of cells (three experiments, each performed in triplicate). Asterisks indicate significant differences between the wild-type and mutantstrains (P � 0.05).

FIG. 5. Effect of SaeRS on DNA fragmentation characteristic ofapoptosis of epithelial cells (A549) induced by S. aureus. Lanes: M,1-kb DNA laddering marker; 1, control without treatment; 2, cellsincubated overnight with wild-type S. aureus WCUH29; 3, cells incu-bated overnight with saeS null mutant strain Sa371ko.

VOL. 74, 2006 EFFECTS OF SaeRS INACTIVATION ON S. AUREUS 4661

on October 22, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

attenuated the virulence of S. aureus in the murine model ofhematogenous pyelonephritis.

DISCUSSION

Microarray analysis has been widely used to identify genesregulated by different regulators in S. aureus, including AgrAand SarA (12), Rot (53), Sigma B (4), and ArlRS (37). In thisstudy, we demonstrated that a two-component signal transduc-tion system, SaeRS, is a predominant regulator of virulencefactors in S. aureus. Using an Affymetrix oligonucleotide array,we identified the genes that are directly and/or indirectly reg-

ulated by the SaeRS system. Our microarray analysis showedthat inactivating the SaeRS system dramatically down-regu-lates the expression of virulence genes that encode both cellwall-associated proteins (including fibronectin- and fibrinogen-binding proteins) and exported proteins, such as toxins. Thesedata are highly consistent with our data from the semiquanti-tative RT-PCR analysis presented in this study, in which wesaw a decrease in saeS expression and a decrease in mRNAexpression of coa, fnbA, fnbB, efb, hla, and other genes. Ourresults are also consistent with previous findings that SaeRScontrols the expression of coa, fnb, and hla (20, 21), since nofnbA, fnbB, or coa mRNA was detectable in the saeS mutantstrain (55). We also found that SaeRS may control the tran-scription of a gene encoding a hypothetical fibrinogen-bindingprotein (SA1000). Therefore, our results indicate that theSaeRS system is likely an important regulator of virulencefactors in S. aureus.

Surprisingly, our microarray and RT-PCR results indicatedthat the SaeRS system negatively affects the expression of agrA,which was confirmed by using an agrA promoter-gfp reporterfusion. To investigate whether this negative impact is a directeffect is beyond the scope of this study. Our result is inconsis-tent with a previous report that Agr acts upstream of sae (46),indicating that the two regulatory systems may interact witheach other. In addition, our studies did not reveal significantdifferences in expression of the extracellular adherence protein(Eap) and the extracellular matrix protein-binding protein(Emp) (26) after inactivation of the SaeRS system. This isinconsistent with some previous reports (21, 24), which may bedue either to the different sensitivities of different approachesor to the use of different S. aureus isolates (5). However, genesthat can bind to a dephosphorylated response regulator, SaeR,may have been missed, since we examined only the effects ofknocking out the sensor of histidine kinase on gene expression.

FIG. 6. Effects of SaeRS on S. aureus-induced death of epithelial cells (A549) by bacterial cells (A) and supernatants (5 �l/ml) of overnightcultures (B). Cell viability was measured after overnight treatment and is expressed as the average for at least three experiments � standarddeviation. Asterisks indicate significant differences between the wild-type and mutant strains (P � 0.05).

FIG. 7. S. aureus recovered from infected kidneys. Five mice pergroup were infected with about 107 CFU of bacteria via an intravenousinjection of 0.2 ml of bacterial suspension into the tail vein, using atuberculin syringe. The mice were sacrificed by carbon dioxide over-dose 3 days after infection. Kidneys were removed aseptically andhomogenized in 1 ml of PBS for enumeration of viable bacteria. Thisexperiment was repeated three times with similar results. **, P � 0.01.

4662 LIANG ET AL. INFECT. IMMUN.

on October 22, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

Also, microarray expression analyses are limited by the factthat short-lived and unstable transcripts are often not mea-sured, as microarrays are essentially a “snapshot” of transcrip-tional activity occurring at a fixed point in time. Therefore,some genes that are differentially regulated by the SaeRS reg-ulator during different phases of growth may go undetected.Because the microarray data represent the steady-state aver-age levels of mRNA, whether these increased levels resultfrom a direct or indirect effect of SaeRS cannot be discerned inthis study.

Our results indicate that the SaeRS system regulates theexpression of genes required for S. aureus to initiate infection,as we found that interrupting the signaling pathway of SaeRSsignificantly diminished bacterial adherence to and internaliza-tion into epithelial cells. These results are consistent with thesignificantly lower transcription levels of fnbA and coa in thesaeS null mutant in this study and with previous findings thatthe SaeRS system plays an important role in bacterial adhesionto and invasion of human endothelial cells (55). S. aureusexpresses a series of adhesins which can facilitate the organ-ism’s adherence to and/or invasion of nonphagocytic cells byinteracting with extracellular matrix components of the host,such as collagen, fibrinogen, and fibronectin (17). Althoughthere are some contradictory results from different models ofinfection (7, 8, 16, 35, 40, 44, 48), fibronectin-binding proteinsare the main surface-associated proteins that function asadhesins and invasins by assembling the extracellular matrixprotein Fn, which bridges to host cell receptors, such as 5�1-integrin (13, 19, 54). It has been demonstrated that fibronectin-binding proteins play a critical role in S. aureus infective en-docarditis (35, 50, 60) and osteomyelitis (32, 42). Theprevalence of FnBPs in clinical isolates also indicates the im-portance of FnBPs for S. aureus infection (47, 48, 52). In thisstudy, we not only demonstrated that the SaeRS system regu-lates the production of FnBPs but also found that it controls aputative fibrinogen-binding protein (the SA1000 protein) andEfb (36), which also contributes in part to the bacterial adhe-sion to and invasion of epithelial cells. It should be pointed outthat we named the ORF SA1000 based on the S. aureus N315annotated genome. However, we are not sure if the SA1000protein is able to bind fibrinogen. Further studies to investigatepotential mechanisms of the effects of the SA1000 protein andEfb on adhesion and invasion are in progress.

Our results also demonstrate that the SaeRS system modu-lates the expression of genes required for S. aureus to causesevere infections, since inactivation of the SaeRS system re-duced the staphylococcal capability of inducing apoptosis andcell death. These results are consistent with the eliminatedtranscription of hla expression in vitro in this study and previ-ous reports (11, 22). Our unpublished data and studies by otherinvestigators have demonstrated that alpha-toxin is necessaryfor S. aureus to cause different types of cell apoptosis and deathvia different signaling transduction pathways (2, 14, 25, 33).Furthermore, in this study we demonstrated that the SaeRSsystem plays an important role in S. aureus pathogenesis in amurine model of infection. The contribution of sae to virulencehas also been demonstrated using a mouse intraperitonealinfection model (3). These observations are consistent withother findings showing that alpha-toxin plays an important rolein pathogenesis in different models of infection (29, 34). Thus,

activation of the SaeRS system is crucial for S. aureus to inducethe death of infected cells, which in turn may promote thespread of infection.

Previous studies have demonstrated that surface-associatedproteins, including FnBPs, are produced in great numbers inthe early log phase of growth in vitro and during the early stageof infection; in contrast, the exported alpha-toxin is generateddramatically in the stationary phase of growth in vitro andduring later stages of infection (11, 22). The expression of coaand hla is regulated by a series of global regulators, includingAgr, SarA, Sigma B, and Rot, while the transcription of fnb isregulated by SarA and Sigma B (4, 12, 45). Moreover, theglobal regulators Agr and SarA have been demonstrated toplay a role in the induction of apoptosis in epithelial cells by S.aureus, as agr and sar mutants are internalized but do notinduce apoptosis (59). Therefore, the SaeRS system mightcoordinate with these and/or other global regulators to differ-entially control the expression of virulence genes both in vitroand in vivo during infection.

In conclusion, in this study we identified the genes that aredirectly and/or indirectly regulated by SaeRS by using microar-ray analysis. Our results demonstrated that inactivation of theSaeRS system dramatically eliminates the capability of S. au-reus to adhere to and/or invade epithelial cells and to triggerapoptosis and death of epithelial cells. Moreover, we demon-strated that a novel hypothetical fibrinogen-binding protein(the SA1000 protein) and a well-studied extracellular fibrino-gen-binding protein, Efb (which are regulated by SaeRS andrevealed in our microarray assay), are also involved in adhe-sion and invasion during pathogen-host interactions. Thesedata indicate that activation of the SaeRS system is requiredfor S. aureus to adhere to and invade epithelial cells.

ACKNOWLEDGMENTS

We thank Li Zheng for his technical assistance, Aaron Becker for hisassistance with microarray analysis, and M. Rosenberg, K. Matchett,and the anonymous reviewers for critical readings of the manuscript.

Y. Ji’s laboratory is supported by grant AI057451 from the NationalInstitute of Allergy and Infectious Disease and by AHC Faculty Re-search Development grant 03-02 at the University of Minnesota.

REFERENCES

1. Agerer, F., A. Michel, K. Ohlsen, and C. R. Hauck. 2003. Integrin-mediatedinvasion of Staphylococcus aureus into human cells requires Src family pro-tein-tyrosine kinases. J. Biol. Chem. 278:42524–42531.

2. Bantel, H., B. Sinha, W. Domschke, G. Peters, K. Schulze-Osthoff, and R. U.Janicke. 2001. -Toxin is a mediator of Staphylococcus aureus-induced celldeath and activates caspases via the intrinsic death pathway independently ofdeath receptor signaling. J. Cell Biol. 155:637–647.

3. Benton, B. M., J. Zhang, S. Bond, C. Pope, T. Christian, L. Lee, K. Winter-berg, M. B. Schmid, and J. M. Buysse. 2004. Large-scale identification ofgenes required for full virulence of Staphylococcus aureus. J. Bacteriol. 186:8478–8489.

4. Bischoff, M., P. Dunman, J. Kormanec, D. Macapagal, E. Murphy, W.Mounts, B. Berger-Bachi, and S. Projan. 2004. Microarray-based analysis ofthe Staphylococcus aureus �B regulon. J. Bacteriol. 186:4085–4099.

5. Blevins, J. S., K. E. Beenken, M. O. Elasri, B. K. Hurlburt, and M. S.Smeltzer. 2002. Strain-dependent differences in the regulatory roles of sarAand agr in Staphylococcus aureus. Infect. Immun. 70:470–480.

6. Bronner, S., H. Monteil, and G. Prevost. 2004. Regulation of virulencedeterminants in Staphylococcus aureus: complexity and applications. FEMSMicrobiol. Rev. 28:183–200.

7. Brouillette, E., G. Grondin, L. Shkreta, P. Lacasse, and B. G. Talbot. 2003.In vivo and in vitro demonstration that Staphylococcus aureus is an intracel-lular pathogen in the presence or absence of fibronectin-binding proteins.Microb. Pathog. 35:159–168.

8. Brouillette, E., B. G. Talbot, and F. Malouin. 2003. The fibronectin-binding

VOL. 74, 2006 EFFECTS OF SaeRS INACTIVATION ON S. AUREUS 4663

on October 22, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

proteins of Staphylococcus aureus may promote mammary gland colonizationin a lactating mouse model of mastitis. Infect. Immun. 71:2292–2295.

9. Cheung, A. L., J. M. Koomey, C. A. Butler, S. J. Projan, and V. A. Fischetti.1992. Regulation of exoprotein expression in Staphylococcus aureus by alocus (sar) distinct from agr. Proc. Natl. Acad. Sci. USA 89:6462–6466.

10. Cue, D., P. Dombek, H. Lam, and P. P. Cleary. 1998. Streptococcus pyogenesserotype M1 encodes multiple pathways for entry into human epithelial cells.Infect. Immun. 66:4593–4601.

11. Da Silva, M., J. Zahm, D. Gras, O. Bjolet, M. Abely, J. Hinnrasky, M.Milliot, M. de Assis, C. Hologne, N. Bonnet, M. Merten, M. Plotkowski, andE. Puchelle. 2004. Dynamic interaction between airway epithelial cells andStaphylococcus aureus. Am. J. Physiol. Lung Cell Mol. Physiol. 287:L453–L551.

12. Dunman, P. M., E. Murphy, S. Haney, D. Palacios, G. Tucker-Kellogg, S.Wu, E. L. Brown, R. J. Zagursky, D. Shlaes, and S. J. Projan. 2001. Tran-scription profiling-based identification of Staphylococcus aureus genes regu-lated by the agr and/or sarA loci. J. Bacteriol. 183:7341–7353.

13. Dziewanowska, K., J. M. Patti, C. F. Deobald, K. W. Bayles, W. R. Trumble,and G. A. Bohach. 1999. Fibronectin binding protein and host cell tyrosinekinase are required for internalization of Staphylococcus aureus by epithelialcells. Infect. Immun. 67:4673–4678.

14. Essmann, F., H. Bantel, G. Totzke, I. Engels, B. Sinha, K. Schulze-Osthoff,and R. U. Janicke. 2003. Staphylococcus aureus alpha-toxin-induced celldeath: predominant necrosis despite apoptotic caspase activation. CellDeath Differ. 10:1260–1272.

15. Fan, F., R. D. Lunsford, D. Sylvester, J. Fan, H. Celesnik, S. Iordanescu, M.Rosenberg, and D. McDevitt. 2001. Regulated ectopic expression and allelic-replacement mutagenesis as a method for gene essentiality testing in Staphy-lococcus aureus. Plasmid 46:71–75.

16. Flock, J. I., S. A. Hienz, A. Heimdahl, and T. Schennings. 1996. Reconsid-eration of the role of fibronectin binding in endocarditis caused by Staphy-lococcus aureus. Infect. Immun. 64:1876–1878.

17. Foster, T. J., and M. Hook. 1998. Surface protein adhesins of Staphylococcusaureus. Trends Microbiol. 6:484–488.

18. Fournier, B., A. Klier, and G. Rapoport. 2001. The two-component systemArlS-ArlR is a regulator of virulence gene expression in Staphylococcusaureus. Mol. Microbiol. 41:247–261.

19. Fowler, T., E. R. Wann, D. Joh, S. Johansson, T. J. Foster, and M. Hook.2000. Cellular invasion by Staphylococcus aureus involves a fibronectin bridgebetween the bacterial fibronectin-binding MSCRAMMs and host cell �1integrins. Eur. J. Cell Biol. 79:672–679.

20. Giraudo, A., A. Calzolari, A. Cataldi, C. Bogni, and R. Nagel. 1999. The saelocus of Staphylococcus aureus encodes a two-component regulatory system.FEMS Microbiol. Lett. 177:15–22.

21. Giraudo, A. T., A. L. Cheung, and R. Nagel. 1997. The sae locus of Staphy-lococcus aureus controls exoprotein synthesis at the transcriptional level.Arch. Microbiol. 168:53–58.

22. Goerke, C., U. Fluckiger, A. Steinhuber, W. Zimmerli, and C. Wolz. 2001.Impact of the regulatory loci agr, sarA and sae of Staphylococcus aureus onthe induction of -toxin during device-related infection resolved by directquantitative transcript analysis. Mol. Microbiol. 40:1439–1447.

23. Goerke, C., U. Fluckiger, A. Steinhuber, V. Bisanzio, M. Ulrich, M. Bischoff,J. M. Patti, and C. Wolz. 2005. The role of Staphylococcus aureus globalregulators sae and �B in virulence gene expression during device-relatedinfection. Infect. Immun. 73:3415–3421.

24. Harraghy, N., J. Kormanec, C. Wolz, D. Homerova, C. Goerke, K. Ohlsen, S.Qazi, P. Hill, and M. Herrmann. 2005. sae is essential for expression of thestaphylococcal adhesins Eap and Emp. Microbiology 151:1789–1800.

25. Haslinger, B., K. Strangfeld, G. Peters, K. Schulze-Osthoff, and B. Sinha.2003. Staphylococcus aureus -toxin induces apoptosis in peripheral bloodmononuclear cells: role of endogenous tumor necrosis factor- and themitochondrial death pathway. Cell Microbiol. 5:729–741.

26. Hussain, M., K. Becker, C. von Eiff, G. Peters, and M. Herrmann. 2001.Analogs of Eap protein are conserved and prevalent in clinical Staphylococ-cus aureus isolates. Clin. Diagn. Lab. Immunol. 8:1271–1276.

27. Iandolo, J. J. 1990. The genetics of staphylococcal toxins and virulencefactors, p. 399–426. In B. H. Iglewski and V. L. Clark (ed.), Molecular basisof bacterial pathogenesis. New York Academic Press, New York, N.Y.

28. Ji, G., R. Beavis, and R. Novick. 1995. Cell density control of staphylococcalvirulence mediated by an octapeptide pheromone. Proc. Natl. Acad. Sci.USA 92:12055–12059.

29. Ji, Y., A. Marra, M. Rosenberg, and G. Woodnutt. 1999. Regulated antisenseRNA eliminates alpha-toxin virulence in Staphylococcus aureus infection. J.Bacteriol. 181:6585–6590.

30. Ji, Y., B. Zhang, S. Van Horn, P. Warren, M. Burnham, G. Woodnutt, andM. Rosenberg. 2001. Identification of critical staphylococcal genes usingconditional growth phenotypes generated by antisense RNA. Science 293:2266–2269.

31. Ji, Y., D. Yin, B. Fox, D. Holmes, D. Payne, and M. Rosenberg. 2004.Validation of antibiotic mechanism of action by regulated antisense RNAexpression in Staphylococcus aureus. FEMS Microbiol. Lett. 231:177–184.

32. Johansson, A., J. I. Flock, and O. Svensson. 2001. Collagen and fibronectin

binding in experimental staphylococcal osteomyelitis. Clin. Orthop. 382:241–246.

33. Jonas, D., I. Walev, T. Berger, M. Liebetrau, M. Palmer, and S. Bhakdi.1994. Novel path to apoptosis: small transmembrane pores created by staphy-lococcal -toxin in T lymphocytes evoke internucleosomal DNA degrada-tion. Infect. Immun. 62:1304–1312.

34. Kernodle, D. S., R. Voladri, B. E. Menzies, C. C. Hager, and K. M. Edwards.1997. Expression of an antisense hla fragment in Staphylococcus aureusreduces alpha-toxin production in vitro and attenuates lethal activity in amurine model. Infect. Immun. 65:179–184.

35. Kuypers, J. M., and R. A. Proctor. 1989. Reduced adherence to traumatizedrat heart valves by a low-fibronectin-binding mutant of Staphylococcus au-reus. Infect. Immun. 57:2306–2312.

36. Lee, L., X. Liang, M. Hook, and E. Brown. 2004. Identification and charac-terization of the C3 binding domain of the Staphylococcus aureus extracel-lular fibrinogen-binding protein (Efb). J. Biol. Chem. 279:50710–50716.

37. Liang, X., L. Zheng, C. Landwehr, D. Lunsford, D. Holmes, and Y. Ji. 2005.Global regulation of gene expression by ArlSR, a two-component signaltransduction system of Staphylococcus aureus. J. Bacteriol. 187:5486–5492.

38. Lowy, F. D. 1998. Staphylococcus aureus infections. N. Engl. J. Med. 339:520–532.

39. Luong, T. T., S. Newell, and C. Y. Lee. 2003. Mgr, a novel global regulator inStaphylococcus aureus. J. Bacteriol. 185:3703–3710.

40. McElroy, M. C., D. J. Cain, C. Tyrrell, T. J. Foster, and C. Haslett. 2002.Increased virulence of a fibronectin-binding protein mutant of Staphylococ-cus aureus in a rat model of pneumonia. Infect. Immun. 70:3865–3873.

41. McNamara, P., K. Milligan-Monroe, S. Khalili, and R. A. Proctor. 2000.Identification, cloning, and initial characterization of rot, a locus encoding aregulator of virulence factor expression in Staphylococcus aureus. J. Bacte-riol. 182:3197–3203.

42. Menzies, B. E. 2003. The role of fibronectin binding proteins in the patho-genesis of Staphylococcus aureus infections. Curr. Opin. Infect. Dis. 16:225–229.

43. Menzies, B. E., and I. Kourteva. 2000. Staphylococcus aureus -toxin inducesapoptosis in endothelial cells. FEMS Immunol. Med. Microbiol. 29:39–45.

44. Mongodin, E., O. Bajolet, J. Cutrona, N. Bonnet, F. Dupuit, E. Puchelle, andS. de Bentzmann. 2002. Fibronectin-binding proteins of Staphylococcus au-reus are involved in adherence to human airway epithelium. Infect. Immun.70:620–630.

45. Novick, R. P. 2003. Autoinduction and signal transduction in the regulationof staphylococcal virulence. Mol. Microbiol. 48:1429–1449.

46. Novick, R. P., and D. Jiang. 2003. The staphylococcal saeRS system coordi-nates environmental signals with agr quorum sensing. Microbiology 149:2709–2717.

47. Peacock, S. J., N. P. Day, M. G. Thomas, A. R. Berendt, and T. J. Foster.2000. Clinical isolates of Staphylococcus aureus exhibit diversity in fnb genesand adhesion to human fibronectin. J. Infect. 41:23–31.

48. Peacock, S. J., C. E. Moore, A. Justice, M. Kantzanou, L. Story, K. Mackie,G. O’Neill, and N. P. Day. 2002. Virulent combinations of adhesin and toxingenes in natural populations of Staphylococcus aureus. Infect. Immun. 70:4987–4996.

49. Projan, S., and R. Novick. 1997. The molecular basis of pathogenicity, p.55–81. In G. Archer and K. Crossley (ed.), Staphylococci in human diseases.Churchill Livingstone, New York, N.Y.

50. Que, Y. A., P. Francois, J. A. Haefliger, J. M. Entenza, P. Vaudaux, and P.Moreillon. 2001. Reassessing the role of Staphylococcus aureus clumpingfactor and fibronectin-binding protein by expression in Lactococcus lactis.Infect. Immun. 69:6296–6302.

51. Rampone, H., G. L. Martinez, A. T. Giraudo, A. Calzolari, and R. Nagel.1996. In vivo expression of exoprotein synthesis with a sae mutant of Staph-ylococcus aureus. Can. J. Vet. Res. 60:237–240.

52. Rice, K., M. Huesca, D. Vaz, and M. J. McGavin. 2001. Variance in fibronec-tin binding and fnb locus polymorphisms in Staphylococcus aureus: identifi-cation of antigenic variation in a fibronectin binding protein adhesin of theepidemic CMRSA-1 strain of methicillin-resistant S. aureus. Infect. Immun.69:3791–3799.

53. Saıd-Salim, B., P. M. Dunman, F. M. McAleese, D. Macapagal, E. Murphy,P. J. McNamara, S. Arvidson, T. J. Foster, S. J. Projan, and B. N. Kreiswirth.2003. Global regulation of Staphylococcus aureus genes by rot. J. Bacteriol.185:610–619.

54. Sinha, B., P. P. Francois, O. Nusse, M. Foti, O. M. Hartford, P. Vaudaux,T. J. Foster, D. P. Lew, M. Herrmann, and K. H. Krause. 1999. Fibronectin-binding protein acts as Staphylococcus aureus invasin via fibronectin bridgingto integrin- 5�1. Cell. Microbiol. 1:101–117.

55. Steinhuber, A., C. Goerke, M. G. Bayer, G. Doring, and C. Wolz. 2003.Molecular architecture of the regulatory locus sae of Staphylococcus aureusand its impact on expression of virulence factors. J. Bacteriol. 185:6278–6286.

56. Toledo-Arana, A., N. Merino, M. Vergara-Irigaray, M. Debarbouille, J. R.Penades, and I. Lasa. 2005. Staphylococcus aureus develops an alternative,ica-independent biofilm in the absence of the arlRS two-component system.J. Bacteriol. 187:5318–5329.

4664 LIANG ET AL. INFECT. IMMUN.

on October 22, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

57. van Wamel, W., Y. Q. Xiong, A. S. Bayer, M. R. Yeaman, C. C. Nast, andA. L. Cheung. 2002. Regulation of Staphylococcus aureus type 5 capsularpolysaccharides by agr and sarA in vitro and in an experimental endocarditismodel. Microb. Pathog. 33:73–79.

58. Wang, B., R. Yurecko, S. Dedhar, and P. P. Cleary. 2006. Integrin-linkedkinase is an essential link between integrins and uptake of bacterial patho-gens by epithelial cells. Cell. Microbiol. 8:257–266.

59. Wesson, C. A., L. E. Liou, K. M. Todd, G. A. Bohach, W. R. Trumble, andK. W. Bayles. 1998. Staphylococcus aureus Agr and Sar global regulatorsinfluence internalization and induction of apoptosis. Infect. Immun. 66:5238–5243.

60. Xiong, Y. Q., A. S. Bayer, M. R. Yeaman, W. van Wamel, A. C. Manna, andA. L. Cheung. 2004. Impacts of sarA and agr in Staphylococcus aureus strainNewman on fibronectin-binding protein A gene expression and fibronectinadherence capacity in vitro and in experimental infective endocarditis. In-fect. Immun. 72:1832–1836.

61. Yarwood, J. M., J. K. McCormick, and P. M. Schlievert. 2001. Identificationof a novel two-component regulatory system that acts in global regulation ofvirulence factors of Staphylococcus aureus. J. Bacteriol. 183:1113–1123.

62. Zhang, L., L. Gray, R. P. Novick, and G. Ji. 2002. Transmembrane topologyof AgrB, the protein involved in the post-translational modification of AgrDin Staphylococcus aureus. J. Biol. Chem. 277:34736–34742.

Editor: V. J. DiRita

VOL. 74, 2006 EFFECTS OF SaeRS INACTIVATION ON S. AUREUS 4665

on October 22, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

INFECTION AND IMMUNITY, Oct. 2006, p. 6020–6021 Vol. 74, No. 100019-9567/06/$08.00�0 doi:10.1128/IAI.01189-06

ERRATUM

Inactivation of a Two-Component Signal Transduction System,SaeRS, Eliminates Adherence and Attenuates Virulence of

Staphylococcus aureusXudong Liang, Chaunxin Yu, Junsong Sun, Hong Liu, Christina Landwehr,

David Holmes, and Yinduo JiDepartment of Veterinary and Biomedical Sciences, College of Veterinary Medicine, University ofMinnesota, 1971 Commonwealth Ave., St. Paul, Minnesota 55108, and Anti-Infective Research,

GlaxoSmithKline Research and Development, 1250 S. Collegeville Rd.,Collegeville, Pennsylvania 19426

Volume 74, no 8, p. 4655–4665, 2006. Page 4657: Figure 1 and the legend to Fig. 1 should appear as shown on the following page.

6020

FIG. 1. Schematic diagrams and Southern blot analysis of allelic gene replacement mutants. (A) Chromosomal DNAs isolated from WCUH29(lane 1) and Sa371ko (lane 2) were digested with XbaI (X) and probed with DIG-labeled saeS. As expected, a 2.8-kb DNA fragment fromWCUH29 was detected in the Southern blot. (B) Chromosomal DNAs isolated from WCUH29 (lane 1) and Sa1000ko (lane 2) were digested withHindIII (H) and probed with DIG-labeled SA1000. As expected, a 1.9-kb DNA fragment from WCUH29 was detected in the Southern blot. (C)Chromosomal DNAs isolated from WCUH29 (lane 1) and Efbko (lane 2) were digested with PstI (P) and probed with DIG-labeled efb. Asexpected, a 2.5-kb DNA fragment from WCUH29 was detected in the Southern blot. Lanes M, 1-kb DNA markers labeled with DIG.

VOL. 74, 2006 ERRATUM 6021