SarZ Promotes the Expression of Virulence Factors and Represses Biofilm Formation by Modulating SarA and agr in Staphylococcus aureus

Abstract

Staphylococcus aureus is a remarkably adaptable organism capable of multiple modes of growth in the human host, as a part of the normal flora, as a pathogen, or as a biofilm. Many of the regulatory pathways governing these modes of growth are centered on the activities of two regulatory molecules, the DNA binding protein SarA and the regulatory RNAIII effector molecule of the agr system. Here, we describe the modulation of these regulators and their downstream target genes by SarZ, a member of the SarA/MarR family of transcriptional regulators. Transcriptional and phenotypic analyses of a sarZ mutant demonstrated that the decreased transcription of mgrA and the agr RNAIII molecule was accompanied by increased transcription of spa (protein A) and downregulation of hla (alpha-hemolysin) and sspA (V8 protease) transcripts when compared to its isogenic parent. The decrease in protease activity was also associated with an increase in SarA expression. Consistent with an increase in SarA levels, the sarZ mutant displayed an enhanced ability to form biofilms. Together, our results indicate that SarZ may be an important regulator governing the dissemination phase of S. aureus infections, as it promotes toxin expression while repressing factors required for biofilm formation.

Staphylococcus aureus is a gram-positive organism that colonizes the anterior nares of approximately 30% of the normal population. Infections can occur when there is a breach in the immune defenses of the host. Diseases caused by S. aureus vary, depending on the site of colonization, and can range from superficial skin lesions to invasive syndromes, such as pneumonia, endocarditis, osteomyelitis, and septicemia (3, 8). Adding to the seriousness of these infections is the propensity of S. aureus to form biofilms, which often leads to the establishment of chronic infections that are difficult to treat (12). The switching from the commensal, to the invasive, to the biofilm modes of growth is due in part to the complex repertoire of regulatory molecules that S. aureus uses to sense and respond to its environment.

The complex nature of the regulatory pathways governing virulence and biofilm formation in S. aureus arises from the fact that multiple regulators can activate or repress a single target gene. Detailed examination of the regulation of several virulence genes, such as spa (protein A), hla (alpha-hemolysin), and sspA (V8 protease), has revealed the following trends (3, 26-28, 36). First, gene expression occurs in a temporal fashion such that cell wall proteins and surface adhesins are expressed during the early, colonizing stages of infection, whereas toxins and secreted proteins are expressed later, during the tissue-damaging phase of disease. Second, the agr quorum-sensing system and the SarA family of DNA binding proteins form the cornerstone of virulence gene regulation in staphylococci. Together, these regulators control the expression of over a hundred genes that are involved in a myriad of cellular functions (11).

The agr system is comprised of two divergent transcripts (RNAII and RNAIII) that are activated in response to cell density. RNAII encodes a quorum-sensing two-component regulatory system that is activated by the autoinducing peptide. The RNAIII transcript, presumably activated by its cognate two-component regulatory system, is the agr effector molecule that activates the transcription of toxin genes such as hla and sspA and represses surface protein genes, such as spa, through its direct action or via intermediary regulators such as SarS, SarT, Rot, and SaeRS (14, 23, 25, 32).

The sarA gene encodes a DNA binding protein belonging to the SarA protein family of winged-helix transcriptional regulators (7). The SarA protein binds and activates the promoters of a number of genes, including agr, and represses other genes, such as sspA and spa (5, 9). Like agr, SarA also controls gene expression indirectly through its effect on other regulatory molecules. For example, the positive effect of SarA on hla is believed to be mediated in part by the repressor SarT (27).

Many members of the SarA family have been characterized and the majority of them have roles in controlling the expression of genes involved in virulence (7). One SarA homologue, SarZ, was previously reported to restore hemolysis in a mutant lacking that capability (18). We have now expanded that role to the general promotion of virulence through the activation of agr and repression of SarA. Additionally, we demonstrate a role for SarZ in the repression of biofilm formation, presumably through its effect on sarA. Together, these results suggest that SarZ plays an important role in the maintenance of active S. aureus infections.

MATERIALS AND METHODS

Bacterial strains, growth conditions, and reagents.

The bacterial strains used in this study are listed in Table 1. S. aureus strains were routinely cultured on tryptic soy agar or broth (TSB) or 03GL broth (24). Escherichia coli strains were grown on Luria-Bertani agar or broth. Antibiotics for plasmid selection and maintenance were used at the following concentrations: S. aureus, 2.5 μg/ml erythromycin and 10 μg/ml chloramphenicol; E. coli, 100 μg/ml ampicillin. All strains were grown in 18-mm borosilicate glass tubes at 37°C in an Excella E24 incubator (New Brunswick Scientific, Edison, NJ) with shaking at 250 rpm unless indicated otherwise. Bacterial growth was monitored by measuring the optical density at 650 nm (OD₆₅₀) on a Spectronic 20D+ spectrophotometer (Spectronic Analytical Instruments, Garforth, England). Early exponential, late exponential, and postexponential growth phases corresponded to OD readings of 0.7, 1.1, and 1.7, respectively, at 650 nm.

TABLE 1.

Bacterial strains and plasmids used in this study

Strain or plasmid	Description	Reference
S. aureus strains
RN4220	Heavily mutagenized strain that accepts foreign DNA, agr-deficient	24
RN6390	Laboratory strain related to strain 8325-4	29
SH1000	8325-4 with rsbU restored	15
ALC6374	RN6390 ΔsarZ	This work
ALC6366	SH1000 ΔsarZ	This work
ALC6367	RN6390 ΔsarZ/sarZ complement	This work
ALC6368	RN6390 ΔsarZ with pEPSA5	This work
ALC6369	RN6390 ΔsarZ with pEPSA5::sspA	This work
ALC6370	RN6390 ΔsarZ with pEPSA5::mgrA	This work
ALC6410	RN6390 ΔsarZ ΔsarA	This work
ALC6409	RN6390 ΔsarZ with pRN6735 (blaZ promoter driving RNAIII)	This work
E. coli strain TOP10	General cloning strain
Plasmids
pMAD	E. coli-S. aureus shuttle vector containing a thermosensitve origin of replication, bgaB, Emr Apr	1
pEPSA5	E. coli-S. aureus shuttle vector containing a xylose-inducible promoter, Apr Cmr	13
pALC1484	E. coli-S. aureus shuttle vector derived from pSK236 containing the gfpuvr gene, Apr Cmr	17
pALC1540	sarA triple promoter driving the expression of gfpuvr in pALC1484	21
pALC1435	sarA P1 promoter driving the expression of gfpuvr in pALC1484	21
pALC1743	agr P3 promoter driving the expression of gfpuvr in pALC1484	17
pALC1741	spa promoter driving the expression of gfpuvr in pALC1484	19
pALC1740	hla promoter driving the expression of gfpuvr in pALC1484	19
pALC2831	sspA promoter driving the expression of gfpuvr in pALC1484	This work
pALC2566	mgrA promoter driving the expression of gfpuvr in pALC1484	This work
pALC5448	pMAD containing a sarA deletion fragment generated by PCR	34
pRN6735	blaZ promoter driving the expression of promoterless RNAIII, Cmr	35

All chemicals and reagents were obtained from either Fisher or Sigma unless noted otherwise. Enzymes and reagents used for genetic manipulations were obtained from New England Biolabs.

Oligonucleotides and strain construction.

A list of the oligonucleotides used in this study is available from the authors upon request. DNA isolation, electrophoresis, PCRs, transformation of E. coli and S. aureus, and other genetic manipulations were carried out according to standard laboratory protocols. Plasmid DNA from E. coli was first introduced into the heavily mutagenized, DNA restriction system-deficient S. aureus strain RN4220 prior to transfer into other S. aureus strains.

An unmarked, in-frame deletion of sarZ was constructed by employing a PCR splicing by overlap extension approach. Briefly, 1-kb regions up and downstream of the sarZ coding region were amplified using primers with complementary 9-bp overhangs. The resulting amplicons were used as the template to create a composite DNA fragment missing sarZ. This fragment was cloned into pMAD, a shuttle vector with a temperature-sensitive origin of replication in S. aureus (1). Through a series of temperature shifts from 30°C to 44°C, the sarZ deletion construct was obtained via a double-crossover event in the chromosome, replacing the native sarZ chromosomal region with the one containing the in-frame markerless deletion. A similar approach, wherein the native sarZ gene was crossed back into the chromosome of the sarZ mutant, was used to create the sarZ-complemented strain. The sarA sarZ double mutant was constructed by introducing the pMAD derivative pALC5448 into the sarZ mutant, allowing for the replacement of the native sarA gene with the mutant allele. Correct clones were identified by PCR and confirmed by sequencing.

The SspA- and MgrA-overproducing strains were created by amplifying the respective open reading frames and cloning them downstream of the xylose-inducible promoter of the S. aureus expression vector pEPSA5 (13). The respective promoter sequences of these genes were also amplified and cloned upstream of the gfp_uvr gene in the E. coli-S. aureus shuttle vector pALC1484 to create transcriptional fusions (17). The RNAIII-overproducing strain was constructed by introducing pRN6735 into the sarZ mutant.

Northern hybridizations.

RNA was isolated from approximately 1.6 × 10¹¹ CFU from early exponential phase (OD₆₅₀, 0.7), late exponential phase (OD₆₅₀, 1.1), or postexponential phase (OD₆₅₀, 1.7) cells grown in TSB. Cells were lysed in Trizol according to the manufacturer's protocol (Invitrogen, Irving, CA) using 0.1-mm silica-zirconia beads and a reciprocating shaker (BIO 101, Vista, CA). Ten micrograms of RNA was separated on a 1.5% agarose-0.66 M formaldehyde gel in 20 mM morpholinepropanesulfonic acid, 10 mM sodium acetate, 2 mM EDTA (pH 7) at 80 V for 3 h, transferred to a nylon membrane (Amersham HyBond XL; GE Healthcare, Piscataway, NJ) in 20× SSC (3 M NaCl, 0.3 M sodium citrate, pH 7), and fixed by baking the membranes at 80°C for 2 h. Membranes were prehybridized in 5× SSC, 0.5% sodium dodecyl sulfate, and 5× Denhardt's solution (0.1% bovine serum albumin, 0.1% Ficoll 400, 0.1% polyvinylpyrrolidone) for 4 h at 65°C. Purified DNA fragments were labeled with the random primed DNA labeling kit (Hoffmann-La Roche Inc., Nutley, NJ) and allowed to hybridize to the membranes overnight at 65°C. Posthybridization, the membranes were washed extensively and the bands were visualized by autoradiography. The relative intensities of the resulting bands were determined by using ImageJ (30). Data shown are representative of at least three hybridizations.

Green fluorescent protein (GFP)-promoter fusion analysis.

Shuttle plasmids containing the promoter regions of various genes driving the expression of GFP_uvr (Table 1) were electroporated into the wild-type, sarZ mutant, and complemented strains by electroporation as previously described (31). Overnight cultures of the resulting strains were diluted 1:100 into fresh TSB and incubated at 37°C with shaking. Aliquots of the cultures were removed at specified time points, and their growth and fluorescence were monitored using an FL600 microplate fluorescence reader (BioTek Instruments, Winooski, VT). Experiments were repeated three times using four independently isolated clones read in triplicate. Error bars represent standard deviations of the means. Statistical significance was calculated using Student's t test.

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis and Western transfer.

Cell lysates were prepared by harvesting approximately 1.6 × 10¹¹ CFU from exponential-phase cells, resuspending the pellets in TEG buffer (25 mM Tris-Cl, pH 8, 25 mM EGTA), and lysing the cells using 0.1-mm silica-zirconia beads and a reciprocating shaker. The proteins were separated from cellular debris by centrifugation at 12,000 × g for 5 min at 4°C and quantitated using the bicinchoninic acid protein assay kit (Pierce, Rockford, IL). An equivalent amount of protein (75 μg) was separated by electrophoresis on 12.5% acrylamide gels, transferred onto polyvinylidene difluoride membranes (Millipore, Billerica, MA), and analyzed by Western blotting according to standard protocols. The primary antibody used was murine anti-SarA monoclonal antibody (1:1,000 dilution) (21), and the secondary antibody was donkey anti-mouse immunoglobulin G conjugated to horseradish peroxidase (1:10,000 dilution; Jackson ImmunoResearch Laboratories, West Grove, PA). Binding of the antibodies to the membrane was detected using the ECL Western blotting detection system according to the manufacturer's instructions (GE Healthcare). Densitometry was performed using ImageJ (30).

Protease detection assays.

Protease production was tested by spotting 2 μl of 10¹⁰ cells/ml onto agar plates containing 5% skimmed milk followed by incubation at 37°C for 24 h. Alternatively, proteins from 30 μl of culture supernatant from overnight bacterial cultures normalized to 10¹⁰ cells/ml were separated electrophoretically on 12.5% acrylamide gels containing 1 mg/ml gelatin. After electrophoresis, the gel was washed with gentle shaking in phosphate-buffered saline containing 2.5% Triton X-100 for 1 hour at room temperature and incubated overnight in protease detection buffer (50 mM Tris-HCl, pH 7.5, 200 mM NaCl, 5 mM CaCl₂, 1 mM cysteine, 0.2% Triton X-100) at 37°C with shaking. The gel was then stained in Coomassie brilliant blue R-250 and destained according to standard procedures to visualize protease activity as zones of clearing on a blue background.

Biofilm formation and detachment assays.

Biofilms were formed in 96-well polystyrene plates according to the method of Caiazza and O'Toole (4). Cells were inoculated at a concentration of 10⁷ cells/ml into 100 μl of 66% TSB containing 0.2% glucose or a mixture of 2% xylose and 2% glucose and incubated for 16 h at 37°C. The culture supernatant and nonadherent bacteria were removed by decanting and then washing the wells three times with water. Biofilms were stained with 0.1% crystal violet for 15 min and washed again three times with water. The crystal violet was released from the adherent cells with absolute ethanol and quantified by reading its absorbance at 562 nm.

For detachment studies, preformed biofilms were treated with 100 μl 20 mM Tris-Cl (pH 7.5), 100 μg/ml protease K, or 0.14 U/ml DNase I (Roche) at 37°C for 2 h. Washing and staining of the biofilms were carried out as described above. Data shown are representative of three independent experiments with at least six replicates each. Error bars represent the standard deviations of the means. Statistical significance was calculated using Student's t test.

RESULTS

Transcriptional profile of the sarZ gene.

SarZ was originally identified based on its homology with members of the SarA protein family of S. aureus. The sarZ locus, SA2174, is located near genes involved in the nitrate/nitrite utilization pathways in S. aureus. However, despite this proximity, sarZ does not appear to be in an operon with any of them. Directly downstream of sarZ is an uncharacterized gene that is predicted to encode a chaperone based on its sequence similarity to genes of the heat shock protein family bearing an alpha-crystallin domain (10).

We initially chose to study sarZ in the RN6390 background because this strain has been well-characterized and is accompanied by an assortment of regulatory mutants which have been evaluated previously (7). However, recognizing that RN6390 has a defective rsbU, which is required to activate the alternative sigma factor B, we conducted relevant phenotypic studies in strain SH1000, which has a restored rsbU (16) (see below).

The transcriptional profile of the sarZ gene in the wild-type strain RN6390 during various points of the S. aureus cell cycle was assessed by Northern blotting (Fig. 1B). The sarZ transcript appeared as two hybridizing bands, with a predominant species at 450 bp, the expected size of the monocistronic sarZ transcript, and a larger species at 1.5 kb, which likely corresponds to the cotranscription of sarZ with the putative chaperone SA2175 (Fig. 1A). As the growth cycle progressed, the level of sarZ transcript reached a maximal level during the late exponential phase (OD₆₅₀, 1.1). The level of the smaller sarZ transcript decreased during the postexponential phase (OD₆₅₀, 1.7), while the larger 1.5-kb transcript was not detectable. As expected, both transcripts were undetectable in the sarZ mutant, and the temporal pattern of sarZ transcription was restored upon complementation of the sarZ mutation.

FIG. 1.

Transcription of sarZ during the exponential and postexponential phases of growth. A. Genomic context of the sarZ gene. Numbers within the block arrows indicate the lengths of the genes in base pairs, and numbers above the arrows indicate the length of the intergenic regions, also in base pairs. B, left panel. Northern blot analysis of sarZ transcription in RN6390 during the early exponential (OD₆₅₀, 0.7), late exponential (OD₆₅₀, 1.1), and postexponential (OD₆₅₀, 1.7) phases of growth. Right panel. Northern blot analysis of sarZ transcription in the wild-type RN6390, sarZ mutant, and complemented sarZ mutant strains of S. aureus during late exponential and postexponential growth. Panels below the Northern blots show the 23S and 16S rRNAs, which served as the internal loading controls.

Effect of a sarZ mutation on transcription of regulatory genes.

Based on its similarity to SarA, we hypothesized that SarZ may be involved in regulation of the S. aureus virulence cascade. To test this hypothesis, the transcriptional profile of a sarZ deletion mutant was compared to that of the parental strain RN6390 and a complemented sarZ mutant strain. The levels of SarA in two independently isolated sarZ mutants were first assessed by immunoblotting. Compared to the parent strain, the two sarZ mutants had a two- to threefold increase in SarA protein level, suggesting that SarZ may have had a repressive effect on SarA expression (Fig. 2A). To confirm this, we analyzed sarA transcription in the sarZ mutant. The sarA locus is composed of three overlapping transcripts initiating from three distinct promoters, P1, P2, and P3 (21), that all encompass the sarA coding region. Northern blotting data revealed that the sarZ mutant demonstrated a moderate increase in the sarA P1 transcript during late exponential phase (OD₆₅₀, 1.1) compared with the parental and complemented strains (Fig. 2B, left panel). The level of the sarA P1 transcript in the mutant was similar to that of the parent during the postexponential phase. GFP-promoter fusion analysis, which reflects the cumulative promoter activity, indicated that the three strains had equivalent levels of P1 promoter activity during growth. When the collective strength of the native triple sarA promoter was investigated, its activity in the sarZ mutant was moderately, but significantly (P = 0.001), increased in comparison with the wild-type and complemented strains (Fig. 2B, right panel).

FIG. 2.

Effect of sarZ on the expression of sarA and agr. A. Western blot analysis of SarA protein expression in the wild-type RN6390 and two independently isolated sarZ mutant clones of S. aureus. B and C, left panels. Northern blot analysis of sarA (B) and agr RNA III (C) transcription in the wild-type RN6390, sarZ mutant, and complemented sarZ mutant strains of S. aureus. Right panels. GFP-fusion analysis of the sarA triple promoter (black), sarA P1 promoter (gray), and agr RNAIII promoter activity in RN6390, sarZ mutant, and complemented strains of S. aureus. Numbers underneath the panels indicate the relative intensities or sum of the intensities of the bands as determined with Image J. Panels below the Northern blots show the 23S and 16S rRNAs, which served as the internal loading controls. *, P < 0.01, determined with Student's t test.

As SarA is a key regulator of agr, the level of agr RNAIII transcription was assessed in the sarZ mutant (Fig. 2C). Both Northern blotting and GFP-promoter fusion analyses indicated that the RNAIII transcript was strongly downregulated in the sarZ mutant in comparison with the wild-type and complemented strains.

Given that the levels of SarA were increased in a sarZ-deficient background and that SarA is an activator of agr, we reasoned that SarZ may have exerted its effect on agr independently of SarA. To evaluate this, we investigated the possibility that SarZ may have acted on downstream genes via SarR, a repressor of SarA and activator of agr (22). However, the level of sarR transcription was unchanged in the sarZ mutant in comparison to the wild type (data not shown), implying that SarZ likely activated agr independently of SarR as well as SarA.

Effect of a sarZ mutation on transcription of virulence genes.

The repression of SarA and activation of agr by SarZ suggested that SarZ is an important player in the S. aureus virulence cascade. Therefore, the transcription of three virulence factor genes, controlled differentially by both SarA and agr, was examined in an attempt to delineate the regulatory network controlled by SarZ. The expression of the protein A gene, spa, is repressed by both SarA and agr. Northern blot analysis indicated that spa transcription was increased twofold in the sarZ mutant in comparison with the wild-type and complemented strains (Fig. 3A, left panel). This upregulation was confirmed by GFP-promoter fusion analysis. Given that SarA levels were increased in the sarZ mutant while RNAIII levels were decreased, these data indicated that SarZ probably represses spa expression by acting through a pathway mediated by agr. Indeed, provision of RNAIII in trans to the sarZ mutant halved the transcription of spa. As expected, removing sarA, a strong repressor of spa, from the sarZ mutant increased spa transcription by approximately 2.4-fold (Fig. 3A, center panel), indicating that these two regulators exert their effects upon spa independently of each other.

FIG. 3.

Effect of sarZ on transcription of spa, hla, and sspA. (Left and center panels) Northern blot analyses of spa (A), hla (B), and sspA (C) in the RN6390 wild-type, sarZ mutant, and complemented sarZ mutant strains (left panel) and sarZ mutant, RN6390, sarZ mutant expressing RNAIII in trans, and sarZ sarA double mutant (center panel). RNA was harvested from late exponential phase (spa) or postexponential phase (hla, sspA) S. aureus cells. The lightness of the spa band in the middle portion of panel A was due to underexposure of the film, which allowed us to visualize the intensity of the spa band in the sarZ sarA double mutant. Numbers underneath the panels indicate the relative intensities of the bands as determined with ImageJ. Right panels. GFP-fusion analyses of the spa (A), hla (B), and sspA (C) promoter activities in RN6390, sarZ mutant, and complemented strains of S. aureus. *, P < 0.01 determined by Student's t test. D. Protease activity of the wild-type RN6390, sarZ mutant, and complemented sarZ mutant strains of S. aureus. Top: proteolysis on skimmed milk agar. Bottom: proteolysis on a gelatin-containing polyacrylamide gel.

The transcription of hla, which encodes alpha-hemolysin, was moderately reduced in postexponentially growing sarZ mutant cells in comparison to the wild-type and complemented strains (0.7 versus 1) (Fig. 3B). The decrease in hla promoter activity was more evident when analyzing the GFP-promoter fusion activities during the growth cycle in the mutant versus the parent and complemented mutant strains. Deleting sarA from the sarZ mutant decreased hla transcription by fivefold, while providing RNAIII in trans increased hla transcription back to wild-type levels. Since sarZ activated agr while repressing sarA, and both of these activate hla, these data implied that the net effect of SarZ on hla transcription may be due to competing forces of agr and SarA.

The gene encoding the serine protease, sspA, is differentially regulated by agr and SarA. SarA represses the protease gene while agr activates it. As predicted, from the effect of sarZ upon SarA and agr, the transcription of sspA was markedly reduced (by approximately 10-fold) upon deletion of the sarZ gene (Fig. 3C, left panel). However, when sarA was deleted from the sarZ mutant or when RNAIII was provided in trans, the level of sspA transcription did not increase as expected (Fig. 3C, center panel). These data indicate that sarZ may activate sspA independently of both SarA and agr. Consistent with the transcriptional data, the sarZ mutant did not exhibit any detectable protease activity on either a skimmed milk agar plate or polyacrylamide gel containing gelatin (Fig. 3D).

Involvement of SarZ in biofilm formation.

During the course of this work, it was noted that liquid cultures of the sarZ mutant settled out of solution more readily than either the wild-type or complemented strains, a phenotype that is frequently seen in biofilm-producing strains. Accordingly, a biofilm formation assay in a 96-well polystyrene plate format was performed; the sarZ mutant was found to have an increased ability to form biofilms in comparison to the parent and complemented mutant (Fig. 4A).

FIG. 4.

A sarZ mutant hyperproduces biofilms rich in proteins. A. Quantitation of biofilm formation on 96-well polystyrene plates. *, P < 0.01 compared to the wild-type strain RN6390 determined by Student's t test. B. Detachment of biofilms formed by RN6390 and the sarZ mutant after treatment with proteinase K or DNase I for 2 h. *, P < 0.01 compared to the buffer control; **, P > 0.05 compared to the buffer control. Panels beneath the graphs show representative wells with the crystal violet staining phenotype of the biofilms.

In S. aureus, the major biofilm exopolysaccharide, poly-N-acetylglucosamine (PNAG), is encoded by the icaADBC operon (12). However, Northern blot and GFP-promoter fusion analyses failed to demonstrate an alteration in the transcription of the ica genes in a sarZ mutant (data not shown), suggesting that the major constituent of biofilms formed by the sarZ mutant was not PNAG. Treatment of preformed biofilms with proteinase K promoted detachment of the biofilms, whereas treatment with DNase I did not. This result suggested that the biofilm-forming capacity of the sarZ mutant was likely protein mediated (Fig. 4B).

We also reasoned that the ability of the sarZ mutant to produce robust biofilms may have been due to the marked reduction in endogenous protease production. To confirm this hypothesis, the sspA gene was provided in trans to the mutant. The resultant strain was able to produce protease (Fig. 5, middle panel) and more importantly, exhibited reduced biofilm formation, to the level of the parental strain (Fig. 5, top and bottom panels).

FIG. 5.

Protease-mediated detachment of biofilms produced by a sarZ mutant. Top, biofilm formation of RN6390, the sarZ mutant, and the sarZ mutants containing either pEPSA5 (VC) or pEPSA5::sspA (SspA). Cells were grown on 66% TSB, 2% glucose, and 2% xylose to induce both biofilm formation and SspA production. *, P < 0.01 compared to the parental strain RN6390 as determined with Student's t test; **, P > 0.05 compared to the RN6390 parental strain. Middle panels, protease activity of the indicated strains on a gelatin-containing polyacrylamide gel. Bottom panels, representative wells showing the crystal violet staining phenotype of the biofilms.

Effect of SarZ on MgrA expression.

Recently, unpublished data from our lab has demonstrated that another SarA homologue, MgrA, is involved in the repression of biofilm formation in an ica-independent manner, akin to that of SarZ. Hence, we sought to determine whether SarZ was involved in the regulation of MgrA. Both Northern and GFP-promoter fusion analyses demonstrated a noticeable decrease in mgrA transcriptional activity in the sarZ mutant compared with the parent, consistent with the notion that SarZ is an activator of mgrA (Fig. 6A). Provision of mgrA in trans under the control of an exogenous promoter was able to curtail biofilm formation of the sarZ mutant down to the levels of the wild-type strain (Fig. 6B). Although sarZ and mgrA are both positive regulators of agr and sspA, the effect of SarZ on biofilms likely occurred independently of agr and sspA, since provision of mgrA to the sarZ mutant in trans did not complement the transcription of these two genes, nor did it restore protease production in the sarZ mutant (Fig. 6C).

FIG. 6.

Effect of sarZ on mgrA. A. Left panel. Northern blot analysis of mgrA transcription in RN6390, sarZ mutant, and complemented sarZ mutant strains of S. aureus during late exponential phase. Numbers underneath the panels indicate the relative intensities of the bands as determined with ImageJ. Right panel. GFP-fusion analysis of the mgrA promoter activity in RN6390, sarZ mutant, and complemented strains of S. aureus. *, P < 0.01 as determined by Student's t test. B. Top, biofilm formation of RN6390, the sarZ mutant, and the sarZ mutants containing pEPSA5 (VC) or pEPSA5::mgrA (MgrA). Cells were grown on 66% TSB, 2% glucose, and 2% xylose to induce both biofilm formation and MgrA production. *, P < 0.01 compared to RN6390; **, P > 0.05 compared to RN6390. Bottom panels, representative wells showing the crystal violet staining phenotype of the biofilms. C. Top two panels, transcription of agr and sspA in RN6390, sarZ mutant, and complemented sarZ mutant strains of S. aureus as determined by Northern blotting. Numbers beside the panels indicate the relative intensities of the bands as determined with ImageJ. Bottom panels, protease activity of the indicated strains on a gelatin-containing polyacrylamide gel.

Role of SarZ in S. aureus strain SH1000.

Our current understanding of virulence gene regulation in S. aureus is primarily through the study of laboratory strains derived from or closely related to RN6390 (16, 22, 26, 36). RN6390, however, has a deletion in rsbU, which encodes a phosphatase required for the activation of the alternative sigma factor B. As SigB has been implicated in the control of several virulence genes and biofilm formation, we sought to determine whether the effects we observed in the sarZ mutant of RN6390 were applicable to strain SH1000, a strain similar to RN6390 but with a restored rsbU (15). As observed with strain RN6390, sarZ was transcribed maximally in SH1000 during the late exponential phase of growth (Fig. 7A). Analysis of the levels of SarA, agr, and mgrA in two independently isolated sarZ mutants in SH1000 also mirrored results obtained with strain RN6390. In particular, the SarA protein level appeared to be repressed by SarZ, while transcription of agr and mgrA was activated by sarZ (Fig. 7B). In association with these regulatory changes, the sarZ mutant of SH1000 also demonstrated enhanced biofilm production compared with the parent (Fig. 7C). These data suggest that the regulatory profile and the biofilm-positive phenotype of the sarZ mutants occurred independently of the alternative sigma factor B.

FIG. 7.

Effects of sarZ are conserved in SH1000. A. Transcription of sarZ during the early exponential (OD₆₅₀, 0.7), late exponential (OD₆₅₀, 1.1), and postexponential (OD₆₅₀, 1.7) phases of growth. B. Effects of sarZ upon SarA, agr, and mgrA expression during late exponential (SarA, mgrA) and postexponential (agr) growth. Top panel, protein levels of SarA in the SH1000 wild-type strain and two independently isolated sarZ mutant clones of S. aureus as determined by Western blot analysis. Middle and bottom panels, Northern blot analysis of the transcription of agr RNAIII (middle) and mgrA (bottom) in the wild-type SH1000 strain and two independently isolated sarZ mutant clones of S. aureus. Numbers beneath the panels indicate the relative intensities of the bands as determined with ImageJ. C. Biofilm formation of SH1000 and an isogenic sarZ mutant in a 96-well polystyrene plate. *, P < 0.05 as determined by Student's t test.

DISCUSSION

In this study, we have shown by transcriptional and phenotypic analyses that SarZ plays an important role in the S. aureus virulence cascade (Fig. 8). Inactivation of sarZ in two strains of S. aureus resulted in elevated production of SarA in comparison to the respective parental strains. Using strain RN6390 as a model and building upon the accumulated regulatory data on this strain, we found that a mutation in sarZ decreased agr transcription. Consistent with an agr-deficient phenotype, the transcription of spa was increased in the sarZ mutant, while the transcription of hla was decreased. Given that SarA upregulates hla and downmodulates spa, it is likely that the effects of SarZ on protein A and alpha-hemolysin gene expression are mediated primarily via agr rather than SarA (27). The drastic reduction of sspA transcription was first thought to be due to the augmented SarA expression and/or decreased agr RNAIII levels in the sarZ mutant. However, subsequent results with a sarA sarZ double mutant and a sarZ mutant overexpressing RNAIII demonstrated that the effect of sarZ upon sspA likely occurred independently of these two global regulators. Previous work by Kaito et al. (18) demonstrated the direct binding of SarZ to the hla and agr promoter, thus providing evidence that SarZ may directly regulate these genes. Whether this DNA binding activity also applies to the sarA, sspA, or spa promoters will require further study, especially in light of the finding that the binding of SarZ to DNA may be nonspecific (18).

FIG. 8.

Proposed mechanism of virulence gene regulation by SarZ. SarZ represses sarA and activates mgrA, leading to the downregulation of genes important to biofilm formation. Activation of agr by SarZ also leads to the upregulation of hla (alpha-hemolysin) and repression of spa (protein A). Additionally, SarZ activates the production of the SspA protease independently of both agr and sarA. These changes in gene expression result in a net decrease in surface adhesins and increase in toxins, which likely enable the bacteria to spread to new sites of infection.

Our observation that sarZ represses sarA (an activator of agr) and upregulates agr suggested that the activation of agr by SarZ occurred independently of SarA. We initially hypothesized that SarZ may act via SarR, a repressor of SarA and activator of agr (22). However, analysis of sarR transcription in a sarZ mutant did not reveal any reproducible changes with respect to the wild-type strain, suggesting that SarZ likely regulates agr and sarA independently of SarR. Based on the observed decrease in mgrA transcription in the sarZ mutant and its role in agr activation (16), MgrA was next examined as a potential intermediary between SarZ and agr. Provision of MgrA under the control of an exogenous promoter to the sarZ mutant repressed biofilm formation but was not sufficient in restoring the transcription of agr and its downstream target genes, sspA and hla (Fig. 6; hla data not shown), suggesting that SarZ or another factor controlled by SarZ was required for optimal agr activation. Detailed analysis of the binding activity of SarZ to the RNAIII promoter, alone or in conjunction with other regulators, such as SarA and MgrA, is required to fully appreciate the mechanistic details of agr regulation by SarZ.

The effect of sarZ on sarA expression is quite complex due to the intricacy of the triple promoters that drive the expression of sarA (5). Northern blot analysis revealed that sarZ may repress sarA due to its effect on the proximal P1 promoter during late exponential phase, which coincides with maximal sarZ transcription. GFP-promoter fusion analysis, however, suggested that the effect of SarZ upon sarA transcription may be attributable to the cumulative actions of SarZ on all three promoters throughout the growth cycle rather than on the sarA P1 promoter alone. Whether the discrepancy between sarA promoter activity and SarA protein expression levels in the sarZ mutant is attributable to an additional level of posttranscriptional control or occurs in conjunction with another regulator(s) remains to be defined. Indeed, both SarA and SarR can bind to and repress the sarA P1 promoter, thus indicating the complexity through which SarA protein expression may be controlled in S. aureus (6, 20).

As the sarZ mutant exhibited elevated SarA protein expression, we speculated that the sarZ mutant may have an enhanced biofilm phenotype. Previously, SarA has been shown to be essential to biofilm formation by activating the ica genes, which are required to synthesize PNAG, a major biofilm matrix component, and bap, encoding a protein adhesin found only in bovine isolates of S. aureus (33). However, the transcription of the ica genes was not altered in the sarZ mutant compared with the parental strain. Detachment studies with proteinase K and DNase I indicated that the major constituents of biofilms of sarZ mutants are likely proteins rather than a carbohydrate such as PNAG. Accordingly, there appear to be two plausible scenarios by which SarZ may repress biofilm formation in S. aureus: (i) by repressing surface adhesin expression via downregulation of SarA and (ii) through the activation of an inhibitory or detachment factor such as a protease. One likely candidate for a detachment factor was the V8 protease, SspA, since it was strongly repressed in the sarZ mutant. Provision of sspA in trans to the mutant restored the biofilm phenotype to the parental level. However, a recent publication as well as unpublished data from our lab have demonstrated that an sspA mutant alone does not result in an enhanced biofilm phenotype, contrary to what one would expect if SspA were a major detachment factor (2). Therefore, we conclude that while sspA is not essential for biofilm formation, high levels of the protease can contribute to biofilm detachment by acting as a general protease, much like proteinase K.

MgrA has also been described as a negative regulator of biofilm formation that acts in an ica-independent manner (34). Given the similarities between the biofilms of sarZ and mgrA mutants, we sought to determine whether SarZ controlled mgrA with respect to biofilm formation. Indeed, mgrA transcription was decreased in the sarZ mutant in comparison to the wild-type strain. Importantly, the enhanced biofilm phenotype of the sarZ mutant returned to near parental levels when mgrA was provided in trans via a plasmid with a xylose-inducible promoter. We also monitored agr and sspA transcription in this sarZ mutant derivative and determined that this MgrA-overproducing construct did not restore the transcription of these genes to the levels of the parental strain, RN6390. The lack of protease activity in this sarZ mutant derivative implied that increased expression of a surface adhesin(s), possibly via the downregulation of mgrA and/or agr, may account for the enhanced biofilm phenotype in the absence of SarZ. Whether the upregulation of SarA in the sarZ mutant contributes to further enhancement of biofilm formation via the expression of additional surface protein adhesin(s) remains to be determined.

Taken together, our results demonstrate that SarZ upregulates agr and represses SarA expression to promote the expression of virulence genes, such as hla and sspA. Additionally, the sarZ locus also controls mgrA and agr to repress biofilm formation. Repression of biofilm formation prevents the establishment of sessile bacterial communities and hence promotes active infections. In addition, concomitant expression of toxic proteins, such as alpha-hemolysin and V8 protease in a sarZ-positive strain, would lead to tissue damage and promote the spread of bacteria to new infection sites. Accordingly, we propose that SarZ is an important regulator required for the maintenance and spread of active S. aureus infections.