Global Regulation of Staphylococcus aureus Genes by Rot

Abstract

Staphylococcus aureus produces a wide array of cell surface and extracellular proteins involved in virulence. Expression of these virulence factors is tightly controlled by numerous regulatory loci, including agr, sar, sigB, sae, and arl, as well as by a number of proteins with homology to SarA. Rot (repressor of toxins), a SarA homologue, was previously identified in a library of transposon-induced mutants created in an agr-negative strain by screening for restored protease and alpha-toxin. To date, all of the SarA homologues have been shown to act as global regulators of virulence genes. Therefore, we investigated the extent of transcriptional regulation of staphylococcal genes by Rot. We compared the transcriptional profile of a rot agr double mutant to that of its agr parental strain by using custom-made Affymetrix GeneChips. Our findings indicate that Rot is not only a repressor but a global regulator with both positive and negative effects on the expression of S. aureus genes. Our data also indicate that Rot and agr have opposing effects on select target genes. These results provide further insight into the role of Rot in the regulatory cascade of S. aureus virulence gene expression.

Staphylococcus aureus is an important human pathogen that causes a wide spectrum of diseases ranging from benign skin infections to life-threatening endocarditis and toxic shock syndrome (20, 32). The large array of virulence determinants produced by S. aureus is central to the pathogenesis of this organism. The role of these factors in pathogenesis has been demonstrated mainly by comparing the virulence of a mutant defective in one gene with that of its wild-type parent strain in specific animal models of infection. Recently, the use of signature-tagged mutagenesis (STM) identified several factors that play a role in pathogenesis, as their disruption made the mutants unable to survive within the host (10, 25). These virulence factors can be broadly subdivided into cell surface and secreted proteins. Cell surface proteins include microbial surface components recognizing adhesive matrix molecules (MSCRAMMs), which enable the bacteria to colonize host tissues, and factors that facilitate the establishment of the infection by allowing the bacteria to evade the host immune system. Secreted proteins include various enzymes that degrade host components with antimicrobial activity, as well as different hemolysins that damage the host tissue (32).

Production of S. aureus virulence determinants is controlled by several global regulatory loci, which include agr (1), sar (5), sigB (3, 41), sae (14), arl (13), and six SarA homologues (2, 24). These regulators are parts of an important network modulating the expression of S. aureus virulence genes. One target virulence gene can be under the influence of several regulators that “cross talk” to ensure that the specific gene is expressed only when conditions are favorable. For instance, in vitro studies have demonstrated that agr negatively regulates the expression of spa, encoding protein A (32), whereas SarS binds to the spa promoter and activates its expression (38). Interestingly, agr downregulates sarS expression (6, 38). It has thus been proposed that agr downregulates spa expression by downregulating the expression of its activator, sarS (38). Therefore, virulence gene regulators could affect the expression of target genes directly, by binding to their promoters, or indirectly, via other regulators.

The specific role and/or importance of a virulence factor or a virulence gene regulator in S. aureus pathogenesis may vary from one infection type to another. For instance, fibronectin-binding proteins have been shown to mediate invasion of mammalian cells by S. aureus (36). In contrast, McElroy et al. have demonstrated that fibronectin-binding proteins decrease virulence in S. aureus-induced pneumonia (22). Similarly, the agr locus has been shown to be important in S. aureus pathogenesis in various animal infection models such as the mouse arthritis and abscess models (28, 29). However, a recent study by Yarwood et al. has shown that agr expression is not necessary for the development of toxic shock syndrome (40). Taken together, these observations illustrate the multifactorial aspect of S. aureus pathogenesis. In view of the fact that an increasing number of S. aureus strains are becoming more difficult to treat due to their resistance to multiple antibiotics (39), elucidation of S. aureus pathogenesis at the molecular level is crucial in the battle against this important human pathogen, as it may help in the design of new therapeutic strategies.

Among the regulators described above, SarA and agr are the best characterized. SarA is a transcriptional regulator that binds to a consensus motif in the promoter of its target genes (8). Recent studies have shown that SarA is a member of a large family of transcriptional regulators that share a common amino acid motif, KXRXXXDER (2). The agr system is self-activating in a cell density-dependent manner (29). Expression of the agr effector molecule, RNAIII, correlates with downregulation of select genes encoding cell surface proteins, such as protein A and fibronectin-binding proteins, and upregulation of genes encoding secreted proteins, such as alpha- and beta-toxins, lipases, and proteases (32). In vitro, RNAIII-regulated factors follow a temporal trend, with surface proteins being produced during the early-exponential phase of growth and secreted proteins synthesized predominantly during the post-exponential phase. The sequential expression of these determinants correlates with the production of RNAIII, whose levels peak during the post-exponential phase (20, 32). The mechanism by which RNAIII regulates the expression of these different genes is not completely understood. Several studies have suggested the existence of cofactors that interact directly or indirectly with RNAIII to control the expression of the different target genes (18, 24, 30). These factors include members of the Sar family of transcriptional regulators (2).

Recently, a novel SarA homologue, Rot, was identified as a repressor of toxins based on the following observation. An agr-negative strain produces low levels of alpha-toxin and protease activity. However, in a rot agr double mutant, expression of these proteins is increased due to the lack of repression by Rot. Understanding how Rot regulates these and other virulence factors is important in order to further discern S. aureus virulence gene regulation. Most S. aureus virulence gene modulators have both negative and positive effects on their target genes (8, 32, 38). For these reasons, we hypothesized that Rot may have both up- and downregulatory functions. To identify additional target genes regulated by Rot, we compared the transcriptional profile of a rot agr double mutant to that of its parental agr strain during the late-exponential phase of growth by use of the Affymetrix GeneChip. The GeneChip used in this study covers 86% of the S. aureus COL genome and contains approximately 4,600 DNA fragments, representing 2,339 individual S. aureus genes (12).

In this study, we demonstrate that Rot is a global regulator that negatively regulates the transcription of several known virulence genes. Some of these genes encode secreted proteins such as lipase, hemolysins, and proteases, all of which are postulated to play a role in tissue invasion. Rot also positively regulates the expression of a number of genes including those encoding cell surface adhesins. In addition, Rot modulates the expression of genes not previously implicated in pathogenesis. A relationship between Rot and agr in the regulation of S. aureus virulence genes is proposed.

MATERIALS AND METHODS

Bacterial strains, media, and growth conditions.

All S. aureus strains used in this study are listed in Table 1. Bacteria were grown in 0.3 GL agar and CY broth, supplemented with β-glycerophosphate (GP), at 37°C (27). Bacterial cell cultures were grown with rotary agitation at 200 rpm. All mutants were maintained on erythromycin or tetracycline at a concentration of 5 μg ml⁻¹.

TABLE 1.

Bacterial strains used in this study

Strain	Description	Reference or source
RN6390	agr+ laboratory strain	29
RN6911	agrΔ::tetM (Tcr)	29
KT201	sarH1::pKT200 (Emr)	38
PM614	agr-null chr::Tn917::rot (Tcr Emr)	24
BK9986	tetM::agr Tn917::rot (Tcr Emr)	This study
BK9987	tetM::agr sarH1::pKT200 (Tcr Emr)	This study

Construction of rot agr and sarH1 (sarS) agr mutants.

The rot agr double mutant (BK9986) was constructed by bacteriophage φ11-mediated transduction of the rot::Tn917 insertion from strain PM614 into the agr mutant strain RN6911. The sarH1 agr double mutant (BK9987) was made by transduction of the agr::tmn mutation from RN6911 into KT201 (sarH1 mutant). Transfer of the rot and agr mutations into strains BK9986 and B9987 was verified by Southern blot analysis (data not shown).

RNA isolation and Northern blot analysis.

Overnight cultures of S. aureus were diluted 1:50 in CY-GP medium and grown to the late-exponential (5 h) phase of growth. Since our aim was to determine genes whose expression is affected by Rot at any time point, we chose the late-exponential-growth phase because at this time point, we can still observe transcripts of genes expressed early, such as cell surface protein genes, as well as those expressed later, such as genes encoding secreted enzymes and toxins. Cells were harvested by centrifugation, and RNA was isolated by the FastPrep system (developed by Savant [Farmingdale, N.Y.] and Bio 101 [Vista, Calif.]) (4). RNA samples (10 μg) were electrophoresed through a 1.5% agarose-0.66 M formaldehyde gel in morpholinepropanesulfonic acid (MOPS) running buffer. RNA was blotted onto a Hybond N⁺ membrane (Amersham Pharmacia Biotech Inc., Piscataway, N.J.) by using the VacuGene XL blotting system (Amersham Pharmacia). The transfer was performed with 20× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate),, pH 7.0, for 2 h. After the RNA transfer, the RNA ladder (Invitrogen, Carlsbad, Calif.) was manually marked before the hybridization step. Membranes were hybridized overnight with a PCR-amplified probe derived from the target gene (Table 2). Specific transcripts were detected by using the enhanced chemiluminescence (ECL) detection kit (Amersham Pharmacia). In all the Northern blot experiments, the same RNA samples were hybridized with 16S rRNA as an internal control (38). New membranes were used for each Northern blot experiment.

TABLE 2.

Primers used in this study

Primer	Sequence (5′→3′)	GenBank accession no.	Nucleotide positions
hlbF	GGTGCACTTACTGAC	X13404	824-838
hlbR	CGCATATACATCCCATGGC		1678-1660
16S rRNAF	GCGTGGGGATCAAACAGG	X68417	777-794
16S rRNAR	CCCCAATCATTTGTCCCACC		1500-1481
splAF	CGAAGGAGGTAAAATATGAATAAAAATG	AF271715	987-1013
splAR	GTGGTGTGAAATAAACACCGAAATTC		1670-1645
sspAF	CAACGAATGGTCATTATGCACCC	AF309515	598-620
sspAR	GGTACACCGCCCCAATGAATTCC		1123-1101
gehF	GCACAAGCCTCGG	M12715	808-820
gehR	GACGGGGGTGTAG		1280-1268
ureCF	CTCGTGTAACACGTGATGACGTGAACG	AP003136	251461-251487
ureCR	GGCGCATGACCGCCACCAGCACCTTC		252139-252114
spaF	AGACGATCCTTCGGTGAGCAAAGA	U54636	829-852
spaR	GCTTTTGCAATGTCATTTACTGTAT		1203-1179
sarSF	GTCAAGCCTGAAGTCGATATG	AB035454	911-931
sarSR	GCATGGTCTTGCTGCGCGTC		1518-1499
rotF	GTTTTGGGATTGTTGGGATG	AF189239	478-497
rotR	GCATTGCTGTTGCTCTACTTGC		924-903

RNA labeling.

The integrity of RNA preparations was analyzed by electrophoresis in 1.2% agarose-0.66 M formaldehyde gels. Samples with defined 16S and 23S rRNA bands were subjected to a series of reactions that resulted in 5′ biotin-labeled fragmented RNA for GeneChip analysis, as previously described (12). Briefly, total bacterial RNA was first enriched for mRNA. This was accomplished by reverse transcription with rRNA-specific primers, followed by RNase H digestion to remove the rRNA of RNA-cDNA hybrid molecules and DNase I treatment. Enriched mRNA was then chemically fragmented and 5′-thiolated by T4 polynucleotide kinase treatment in the presence of γ-S-ATP. Samples were then biotinylated by addition of polyethylene oxide (PEO)-iodoacetyl-biotin.

GeneChip analysis.

The custom-made S. aureus GeneChips used in this study have been described previously (12). Each chip contains 4,528 open reading frames (ORFs), 12 tRNAs, and 3 rRNAs, with an average of 25 probe sets per ORF. Due to the preliminary state of the genome sequence at the time of design, many genes are represented on the chip as partial, duplicate, or overlapping fragments. Subsequent analyses based on the updated S. aureus genomic sequences suggest that these 4,528 tilings represent approximately 2,700 to 2,900 individual genes and that >86% of the S. aureus COL genome is represented on the chip. The missing genes (14%) include those encoding staphylococcal accessory regulator A (sarA), the sensor histidine kinase SaeS (saeS), d-alanine-d-alanyl carrier protein ligase (dltA), serine protease (sspA), leukotoxin D (lukD), sigma factor B regulators (rsbU, rsbV, and rsbW), and staphylococcal accessory regulator R (sarR).

GeneChip analysis was performed by the method described by Dunman and colleagues (12). Accordingly, to address the issue of reproducibility, two separate RNA samples from each strain were prepared from two different experiments, and for each sample, 1.5 μg of biotinylated RNA was hybridized to at least two separate GeneChips for 14 h at 45°C. GeneChips were then washed and subjected to a series of staining procedures that included the successive binding of streptavidin, biotinylated anti-streptavidin, and phycoerythrin-conjugated streptavidin. Each GeneChip was washed and scanned at a 570-nm wavelength and a 3-μm resolution in an Affymetrix GeneChip scanner. Affymetrix Microarray Suite 4.0 algorithms calculated signal intensities (average difference) and present or absent determinations for each ORF (19). GeneChips were then normalized, and background was defined by using GeneSpring 4.0 (Silicon Genetics) as previously described (12). GeneSpring software was used to further analyze the transcription patterns of genes. Genes negatively regulated by Rot were identified as ORFs with transcript titers at least twofold higher in BK9986 (rot negative) than in RN6911 (rot positive). Genes whose transcript levels were at least twofold higher in RN6911 (rot positive) than in BK9986 (rot negative) were categorized as being positively regulated by Rot. These lists were further filtered to include only genes with signal intensities above background level in the rot agr (rot-negative) strain for Rot-downregulated genes and in the agr (rot-positive) strain for Rot-upregulated genes. Genes represented more than once on the chip are represented only once on the list.

Urease activity.

Urease production was assayed on urea agar slants (Remel, Lenexa, Kans.) as described by the manufacturer. In this test, urease-positive strains show a color change from orange to pink while urease-negative strains have an intact orange color.

RESULTS

Rot is a transcriptional modulator.

GeneChip results indicate that Rot behaves as both a positive and a negative modulator of numerous virulence genes. We have identified 60 genes whose expression is negatively regulated by Rot (Table 3) and 86 genes whose expression is positively regulated by Rot (Table 4). From both categories of regulated genes, we chose seven genes with a well-known or postulated virulence function for further validation (Table 5).

TABLE 3.

Genes downregulated by Rot

Chip ORF no.a	N315 ORF no.b	N315 geneb	N315 descriptionb	Fold changec	agr or sar effectd	Role categoryd
4877	SA1811	hlb	Truncated beta-hemolysin	6.9
4218	SA0143	adhE	Alcohol-acetaldehyde dehydrogenase	3.8
1726	SA0471	cysK	Cysteine synthase (o-acetylserine sulfhydrylase) homologue	4.6
1664	SA2312	ddh	D-specific d-2-hydroxyacid dehydrogenase	11.5	agr, down	Carbohydrate metabolism
904	SA1267	ebhA	HP, similar to streptococcal adhesin Emb	567.4
1885	SA1268	ebhB	HP, similar to streptococcal adhesin Emb	Upregulated	agr, up	Virulence factors
1434	SA1091	fmhC (eprh)	FmhC protein	Upregulated
1527	SA0309	geh	Glycerol ester hydrolase	8.4	agr, up	Virulence factors
4046	SA2294	gntK	Gluconokinase	3.1
5281	SA2293	gntP	Gluconate permease	4.6	sar, down	Transport
1928	SA2209	hlgB	Gamma-hemolysin component B	31.9	agr, up	Virulence factors
1927	SA2208	hlgC	Gamma-hemolysin component C	19.6	agr, up	Virulence factors
1700	SA1881	kdpA	Probable potassium-transporting ATPase A chain	2.8
4715	SA1879	kdpC	Probable potassium-transporting ATPase C chain	3.7
4583	SA1090	lytN	LytN protein	Upregulated
5046	SA0549	mvaK2	Phosphomevalonate kinase	Upregulated
2268	SA2334	mvaS	3-Hydroxy-3-methylglutaryl CoA synthase	8.1
3524	SA2185	narG	Respiratory nitrate reductase alpha chain	6.2	agr, down	Electron transport
431	SA2435	pmi	Mannose-6-phosphate isomerase	3.0
4537	SA1659	prsA	Peptidyl-prolyl cis/trans isomerase homolog	6.5
3551	SA2326	ptsG	PTS system, glucose-specific IIABC component	2.7
1071	SA1589	ribD	Riboflavin-specific deaminase	Upregulated
5329	SA1631	splA	Serine protease SplA (pathogenicity island SaPIn3)	Upregulated	agr, up	Virulence factors
2929	SA1630	splB	Serine protease SplB (pathogenicity island SaPIn3)	Upregulated	agr, up	Virulence factors
3742	SA1629	splC	Serine protease SplC (pathogenicity island SaPIn3)	62.5
324	SA1628	splD	Serine protease SplD (pathogenicity island SaPIn3)	Upregulated	agr, up	Virulence factors
325		splE	Serine protease SplE	133.4
2377	SA1627	splF	Serine protease SplF (pathogenicity island SaPIn3)	Upregulated	agr, up	Virulence factors
2174	SA0900	sspB	Cysteine protease precursor	10.4	agr, up	Virulence factors
2175	SA0899	sspC	Cysteine protease	5.2	agr, up	Virulence factors
4823	SA2082	ureA	Urease gamma subunit	Upregulated
4822	SA2083	ureB	Urease beta subunit	Upregulated
1900	SA2084	ureC	Urease alpha subunit	Upregulated
4817	SA2088	ureD	Urease accessory protein UreD	2.9
1899	SA2085	ureE	Urease accessory protein UreE	6.7
1898	SA2086	ureF	Urease accessory protein UreF	6.0
1897	SA2087	ureG	Urease accessory protein UreG	23.7
401	SA0124		HP, similar to glycosyltransferase TuaA	5.6
657	SA2455		Capsular polysaccharide biosynthesis, CapC	11.1
849	SA2414		HP, similar to glutathione peroxidase	10.8
1386	SA0368		HP, similar to proton/sodium-glutamate symport protein	6.4	agr, up	Transport
1574	SA2062		HP	3.5
1815	SA2200		HP, similar to ABC transporter, ATP binding subunit	3.0
1816	SA2201		HP, similar to ABC transporter, permease protein	3.2
1994	SA1812		HP, similar to synergohymenotropic toxin precursor of Staphylococcus intermedius	3.1
2064	SA2218		Reverse complement of HP	4.4
2067	SA1437		Conserved HP	3.6
2238	SA2434		Fructose phosphotransferase system enzyme FruA homolog	5.4
2543	SA1725		Staphopain, cysteine proteinase	11.9
2612	SA1093		DNA topoisomerase I TopA homolog	Upregulated
2860	SA0989		Conserved HP	7.2
2926	SA0663		HP	3.0
3017	SA0127		HP, similar to capsular polysaccharide synthesis protein 14L	50.7
3078	SA0165		HP, similar to alpha-helical coiled-coil protein SrpF	2.7
3373	SA0904		HP, probable autolysin transcription regulator	Upregulated
3768	SA0850		HP, similar to oligopeptide ABC transporter oligopeptide- binding protein	Upregulated
4061	SA1007		Alpha-hemolysin precursor	24.5	agr, up	Virulence factors
4069	SA0123		HP, similar to UDP-glucose 4-epimerase (Gale-1)	Upregulated
4530	SA0710		Conserved HP	Upregulated
5131	SA2297		HP, similar to GTP-pyrophosphokinase	2.8

^aS. aureus GeneChip ORF number.

^bBased on the published sequence of strain N315 (accession no. NC_002745). For genes not present in N315, the gene name and description are from the COL genome, available from The Institute for Genomic Research Comprehensive Microbial Resource website (http://www.tigr.org). HP, hypothetical protein; CoA, coenzyme A; PTS, phosphotransferase system.

^cNormalized values in the agr rot mutant over values in the agr strain. “Upregulated” denotes genes highly downregulated in the agr strain, such that the transcripts were below detectable levels and the fold change could not be accurately calculated.

^dAs described by Dunman et al. (12). up, upregulated; down, downregulated.

TABLE 4.

Genes upregulated by Rot

Chip ORF no.a	N315 ORF no.b	N315 geneb	N315 descriptionb	Fold changec	agr or sar effectd	Role categoryd
3994	SA1736	aldH	Aldehyde dehydrogenase	2.1
5372		clfB	Clumping factor B	2.9
2462	SA2336	clpL	ATP-dependent Clp proteinase chain ClpL	2.7	sar, down	Adaptation
3833		coa	Coagulase	4.3
2163	SA1253	ctpA	Probable carboxy-terminal processing proteinase CtpA	2.9
5110	SA1164	dhoM	Homoserine dehydrogenase	2.3
4873	SA0796	dltD	Poly(glycerophosphate chain) d-alanine transfer protein	2.0	agr, down	Transport
3383		epiA	Lantibiotic gallidermin precursor EpiA	2.6
84	SA0602	fhuA	Ferrichrome transport ATP-binding protein	2.1
4551	SA2172	gltT	Proton/sodium-glutamate symport protein	2.2
4042	SA2288	gtaB	UTP-glucose-1-phosphate uridyltransferase	2.4
3984	SA0189	hsdR	Probable type I restriction enzyme restriction chain	2.4
2883	SA1505	lysP	Lysine-specific permease	2.4	agr, up	Transport
4486	SA0250	lytS	Two-component sensor histidine kinase	2.3
1047	SA1194	msrA	Peptide methionine sulfoxide reductase homolog	2.4
854	SA2410	nrdD	Anaerobic ribonucleoside-triphosphate reductase	2.1
4205	SA0685	nrdI	NrdI protein involved in ribonucleotide reductase function	2.9
4390	SA0374	pbuX	Xanthine permease	2.3
4905	SA0923	purM	Phosphoribosylformylglycinamidine cyclo-ligase PurM	2.2	sar, down	Nucleotide and nucleic acid metabolism
2800	SA0676	recQ	Probable DNA helicase	2.8	agr, up	DNA replication
1079	SA1583	rot	Repressor of toxins Rot	12.0
2343		sdrC	SdrC protein	2.7	sar, up	Virulence factors
2119	SA0107	spa	Immunoglobulin G binding protein A precursor	15.6	agr, down	Virulence factors
5112	SA1166	thrB	Homoserine kinase homolog	3.9
3239	SA1165	thrC	Threonine synthase	3.8
5075	SA0373	xprT	Xanthine phosphoribosyltransferase	3.0
17			HP	2.0
18	SA0220		HP, similar to glycerophosphodiester phosphodiesterase	2.9
237	SAS013		Reverse complement of HP (pathogenicity island SaPIn2)	2.3	agr, up	Unknown
392	SA0739		Conserved HP	2.4
442	SA0651		HP	2.2
457	SA1613		Conserved HP	2.5
477	SA0428		Conserved HP	2.5
493	SA2378		Conserved HP	3.2	agr, down	Unknown
534	SA1717		Glutamyl-tRNAGln amidotransferase subunit C	2.6
600	SA0523		HP, similar to poly (GP) alpha-glucosyltransferase (TA biosynthesis)	2.2	sar, up	Cell wall
616	SA2133		Conserved HP	2.8
678	SA0682		HP, similar to di-tripeptide ABC transporter	2.1
748	SA2261		HP, similar to efflux pump	3.0
873	SAS088		HP	2.2
1027	SA2007		HP, similar to alpha-acetolactate decarboxylase	2.1
1082	SA0620		Secretory antigen SsaA homologue	2.0
1175	SA2284		HP, similar to accumulation-associated protein	2.5	agr, up	Virulence factors
1353			HP	2.7
1356	SA2170		HP, similar to general stress protein 26	2.1
1357	SA2171		HP	2.0
1398	SA0914		Reverse complement of HP, similar to chitinase B	3.0	agr, up	Miscellaneous
1563	SA0407		Conserved HP	2.0
1564	SA0408		HP	2.3
1592			HP	3.0
1597			HP	2.0
1666	SA2303		HP, similar to membrane-spanning protein	3.2
1765			Epidermin immunity protein F	4.7	agr, up	Antibiotic production
1991	SA0439		HP, similar to lysine decarboxylase	2.5
2032	SA1059		Methionyl-tRNA formyltransferase	2.4	agr, up	Aminoacyl tRNA synthetases
2076				2.4
2077	SA0291		HP	3.1
2078	SA0292		HP	2.5
2103	SA0271		Conserved HP	6.3	agr, up	Unknown
2132	SA2131		Conserved HP	4.9
2133	SA2132		HP, similar to ABC transporter (ATP-binding protein)	2.4
2232	SA2440		HP	2.0
2255			HP	2.1
2337	SA2001		HP, similar to oxidoreductase, aldo/ketoreductase family	2.0
2454	SA1320		HP	4.4
2507	SA0380		Conserved HP (pathogenicity island SaPIn2)	3.5
2697	SA2442		Preprotein translocase SecA homolog	2.2
3259	SA0675		HP, similar to ABC transporter ATP-binding protein	3.0
3313	SA2436		HP, similar to phage infection protein	2.1	sar, up	Bacteriophage related
3391			Conserved HP	2.0
3426	SA1924		HP, similar to aldehyde dehydrogenase	2.5
3599			ABC transporter, ATP-binding protein	2.6
3698	SA0173		HP, similar to surfactin synthetase	2.9	agr, up	Antibiotic production
3888	SA2339		HP, similar to antibiotic transport-associated protein	2.1
4153	SA0310		HP	2.1
4253	SA0003		Conserved HP	2.2
4282	SA1942		Conserved HP	2.5
4525	SA1016		Reverse complement of conserved HP	2.4
4550			HP	2.3
4667	SA0973		Phosphopantetheine adenyltransferase homolog	3.2
4752	SA1056		HP	3.5
5065	SA2256		Conserved HP	15.8
5245	SA0827		HP, similar to ATP-dependent nuclease subunit B	2.3
5437	SA0246		Hypotheticl protein, similar to d-xylulose reductase	2.1
5471	SA1131		HP, similar to 2-oxoacid ferredoxin oxidoreductase, alpha subunit	2.3
5554	SA1679		HP, similar to d-3-phosphoglycerate dehydrogenase	2.0

^aS. aureus GeneChip ORF number.

^bBased on the published sequence of strain N315 (accession no. NC_002745). For genes not present in N315, the gene name and description are from the COL genome, available from The Institute for Genomic Research Comprehensive Microbial Resource website (NC_002745). For genes not present in N315, the gene name and description are from the COL genome, available from The Institute for Genomic Research Comprehensive Microbial Resource website (http://www.tigr.org). HP, hypothetical protein; TA, teichoic acid.

^cNormalized values in the agr strain over values in the agr rot double mutant.

^dAs described by Dunman et al. (12). up, upregulated; down, downregulated.

TABLE 5.

Classification of Rot-regulated genes

Gene/gene producta	Function	Effectb of:
		Rot	agr
Secreted enzymes
geh/lipasec	Degrades lipids	−	+
splA through splF/serine protease operonc	Degrades proteins	−	+
sspB, sspC/within serine protease operonc	Degrades proteins	−	+
Secreted toxins
hla/alpha-toxin	Disrupts red blood cells	−	+
hlb/β-hemolysinc	Sphingomyelinase activity	−	+
Cell surface proteins
clfB/clumping factor	Binds fibrinogen	+	NA
sdrC	Putative cell surface adhesin	+	ND
spa/protein A	Binds Fc portion of immuno- globulin	+	−
Potential virulence factors
thrB	Threonine synthase	+	ND
ureA through ureGc,d	Urease complex	−	+
Regulator, sarS	Virulence gene regulator	+	−

^aGenes are classified based on their known or postulated virulence function.

^b+, activation; −, repression; NA, not affected; ND, not determined.

^cValidated by Northern blot analysis.

^dValidated by protein activity.

Virulence genes negatively regulated by Rot.

Initial characterization of Rot has indicated that it is a repressor of hla, encoding alpha-toxin (24). Our GeneChip data showed that Rot negatively regulates another toxin gene, hlb, which encodes beta-toxin (Tables 3 and 5). The GeneChip result for hlb was confirmed by Northern blot analysis using the rot agr double mutant (rot negative) and its parental agr strain (rot positive). As shown in Fig. 1A, the hlb transcript is present at higher levels in the rot agr strain than in the agr mutant. The detection of two bands that hybridize with hlb is consistent with previous findings (7, 41). The 16S rRNA transcript (internal control) is present at similar levels in the two strains (Fig. 1B).

FIG. 1.

Northern blot analysis of hlb, spl, and ssp. RNA was isolated from BK9986 (rot agr) and RN6911 (agr) at the late-exponential-growth phase and hybridized with hlb (A), splA (C), or sspA (D). Transcription of 16S rRNA was used as an internal control (B).

Rot was previously shown to have a phenotypic effect on protease production. (24). Unlike alpha-toxin and beta-toxin, which are solely encoded by hla and hlb, respectively, staphylococcal protease activity is associated with the gene products of at least two operons, splABCDEF (33) and sspABC (34). Our GeneChip results demonstrated that Rot negatively regulates both operons (Tables 3 and 5). The sspA gene, encoding V8 protease (34), is not listed in Table 3 because it is not represented on the GeneChip used in this study.

The GeneChip data for spl were confirmed by Northern blot analysis, using splA as a probe. The transcript of the spl operon (5.5 kb) is detectable only in the rot agr strain, BK9986, not in the agr strain, RN6911 (Fig. 1C). We tested ssp operon expression in the two strains by Northern blot analysis, using sspA as a probe, and found that the transcript levels of the ssp operon (2.8 kb) were higher in the rot agr double mutant than in the agr strain (Fig. 1D).

Rot also negatively regulates the expression of geh, encoding lipase (Tables 3 and 5). The negative regulation of geh expression by Rot was confirmed by Northern blot analysis in which RNA samples from the rot agr double mutant and the agr strain taken at late-exponential growth were hybridized with a geh-specific probe. The results demonstrated that geh mRNA is present at higher levels in the rot agr strain than in the agr strain (Fig. 2A).

FIG. 2.

Expression analysis of geh (A) and ureC (B and C). (A) RNA samples from BK9986 (rot agr) and RN6911 (agr) were taken during the late-exponential phase of growth and hybridized with a geh-specific probe. (B) ureC transcripts in BK9986 (rot agr) and its parental agr strain (RN6911) were analyzed. (C) Urease activity in both strains was tested on urea agar slants.

As shown in Table 3, GeneChip data indicated that Rot negatively regulates various urease genes such as ureA, ureB, and ureC, encoding different subunits of the urease enzyme, as well as accessory genes such as ureD and ureE, required for the enzyme activity (17) (GenBank accession number AP003136). We performed Northern blot analysis using the ureC gene as a probe and found that the rot agr strain had higher levels of ureC transcripts than the agr strain (Fig. 2B). One of the urease genes downregulated by Rot, ureD, was previously identified as a potential virulence gene, because a ureD transposon mutant was strongly attenuated in an animal model (10). We also tested for urease activity on urea agar slants for the two strains. As shown in Fig. 2C, the rot agr strain was urease positive but the agr strain was urease negative.

Virulence genes upregulated by Rot.

Although Rot was originally identified as being a repressor of certain virulence factors, our results indicate that Rot also acts as a positive modulator of staphylococcal virulence genes. Among the genes upregulated in the presence of Rot are those encoding cell surface proteins. One of these genes is the well-studied spa gene, which encodes protein A (Tables 3 and 5). To validate this result, Northern blot analysis was performed on rot agr and agr strains, and the results demonstrated that spa mRNA is detected only in the agr (rot-positive) strain, not in the rot agr double mutant (rot negative) (Fig. 3A).

FIG. 3.

Northern blot analysis of spa (A), rot (B), and sarS (C). (A) Analysis for the spa transcript in BK9986 (rot agr) and RN6911 (agr) at the late-exponential phase of growth. (B and C) RNA samples from BK9986 (rot agr), BK9987 (sarS agr), RN6911 (agr), and RN6390 (wild-type [wt] strain) at the late-exponential phase of growth were hybridized with rot- and sarS-specific probes, respectively.

Previous studies have demonstrated that SarH1, also called SarS, positively regulates the expression of the protein A gene, spa (6, 38). Since our results indicate that spa is also positively regulated by Rot, we sought to determine if Rot positively regulates spa expression by positively regulating the expression of the spa activator SarS. Using Genespring software analysis, we found that the sarS transcript was considered present in the rot-positive strain (RN6911) but absent in the rot mutant (BK9986), suggesting that Rot may positively regulate sarS expression.

To validate the sarS result, and to analyze the effects of Rot and SarS on each other, we created a sarH1 agr double mutant (BK9987) and compared the transcription of both rot and sarS in the two strains, i.e., BK9986 (rot agr) and BK9987 (sarS agr), as well as in RN6911 (agr) and RN6390 (wild-type strain), in a Northern blot assay. We found that rot is expressed in all the strains tested except the rot mutant, as expected (Fig. 3B). On the other hand, sarS is expressed only in the presence of Rot and in the absence of RNAIII (Fig. 3C). The fact that the sarS transcript is detectable in RN6911 (agr) but not in BK9986 (rot agr) indicates that Rot is indeed required for sarS expression. The higher levels of sarS transcript in RN6911 compared to RN6390 are in good agreement with previous data that showed that agr represses sarS expression (6, 38).

Rot and agr have opposing effects on the expression of target genes.

Our results indicate that Rot-regulated factors follow a general expression pattern opposite that of agr, that is, secreted proteins generally upregulated by agr are negatively regulated by Rot (Table 5). For example, Rot negatively regulates hla and hlb while agr positively regulates them (32). Since Rot negatively regulates the expression and activity of urease, we tested to see if agr will upregulate them. For this, we tested the urease activity of an agr mutant (RN6911) and its parental wild-type strain (RN6390) on urea agar slants and found that indeed the agr-positive strain was urease positive whereas the agr-negative strain was urease negative (data not shown). Conversely, cell surface proteins generally downregulated by agr appear to be positively modulated by Rot (Table 5). For instance, Rot positively regulates spa, encoding protein A, whereas agr negatively regulates it. In addition to structural genes, Rot and agr also have opposing effects on the expression of the virulence gene regulator sarS. Our results indicate that Rot positively regulated the expression of sarS, which has been shown to be repressed by agr (6, 38).

Additional Rot-regulated genes.

Table 3 lists additional virulence genes negatively regulated by Rot as determined by GeneChip analysis, and Table 4 lists all genes shown to be positively modulated by Rot. These tables include genes which encode factors previously shown to promote the growth and survival of S. aureus in an infection model. For instance, Rot downregulates hlgB and hlgC, which encode gamma-hemolysins B and C, respectively. Gamma-hemolysin is an S. aureus virulence factor that has been shown to play a role in S. aureus endophthalmitis and corneal pathogenesis (11, 37). Rot positively regulates thrB, a gene involved in threonine biosynthesis. Disruption of this gene reduces virulence in animal infection models (25). From both lists, it is clear that in addition to virulence genes, Rot also regulates the expression of genes that play different biological roles, including amino acid and carbohydrate metabolism, cell wall biosynthesis, and transport.

DISCUSSION

S. aureus is a versatile human pathogen capable of causing diverse infections in different sites of the body. This versatility can be attributed to its impressive arsenal of virulence factors (20). It is now apparent that S. aureus has numerous regulators that, alone and in combination, control the expression of these factors. Although the list of these virulence factors is quite substantial, it is believed to be incomplete and limited to determinants that are easily detectable by available assays. The recent advent of genomewide expression profiling techniques such as DNA arrays and proteomics provides an unprecedented ability to identify genes regulated in a coordinated manner, thus increasing our understanding of biological processes. In a recent study, Dunman and colleagues used DNA GeneChip technology to demonstrate that the two well-studied virulence gene regulators agr and SarA regulate the expression of known virulence genes as well as novel genes not previously known to play a role in S. aureus pathogenesis (12). Likewise, using a proteomic approach, Ziebandt et al. identified new members of the SarA regulon (41).

Virulence factors whose expression is controlled by the different regulators generally fall into two categories: cell surface proteins and secreted proteins. The agr system generally downregulates cell surface proteins such as protein A and upregulates secreted proteins such as alpha-toxin. By use of DNA array technology, we have clearly demonstrated in this report that Rot does not act exclusively as a repressor. The work presented here indicates that Rot also positively regulates the expression of genes, many of which have previously been determined to be virulence factors. We also demonstrated that Rot and agr have opposing effects on the expression of certain virulence determinants. Collectively, these findings suggest that Rot is a more global virulence regulator than previously recognized. Whether the effects of Rot on its target genes are direct or indirect is currently unknown.

Rot positively regulates the expression of sarS, a transcriptional modulator of virulence gene expression (6, 38). Arvidson and colleagues have shown that SarS binds to the promoter of spa (38), a gene activated by both Rot and SarS. Thus, Rot may affect spa transcription via SarS. Other SarA homologues have also been shown to regulate the expression of virulence gene regulators. For instance, SarR represses sarA expression (21) and SarA represses sarT expression (35). We could postulate that these modulators affect certain target genes by influencing the expression of regulators of these genes.

Besides genes with well-characterized or postulated virulence functions, Rot also regulates genes not previously linked to virulence. These genes could either be potential virulence genes or additional accessory genes with other biological functions. In the latter case, as Dunman and colleagues point out, the regulators involved in the agr-SarA regulatory pathway may be more global than previously recognized and may affect processes that have yet to be established in S. aureus pathogenesis (12). Given the ability of S. aureus to cause a broad range of clinical manifestations, it is likely that many of these gene products are involved in various, as yet unrecognized stages of infection.

Although this study identified numerous members of the Rot regulon, it is very likely that the list is not complete due to the fact that 14% of the staphylococcal genes are missing from the chip. For example, the sspA gene from the serine protease operon is missing from the GeneChip and hence from Table 3. However, Northern blot analysis showed that the whole operon is downregulated by Rot. With the complete genome available, new GeneChips are being designed to cover 100% of the genome, and these will be used in future studies.

Based on the virulence function of Rot-regulated genes, we hypothesize that Rot plays an important role in the early steps of infection. Factors positively modulated by Rot include cell surface adhesins such as ClfB (clumping factor B), which binds fibrinogen (26), and SdrC, a cell surface adhesin (15). Rot also positively regulates the expression of a gene represented by ORF 1175 on the chip, which corresponds to N315 ORF 2284 (Table 4). This ORF contains a cell wall sorting signal motif, LPXTG, which is characteristic of most adhesins (17). Rot positively regulates the production of protein A, a factor involved in evasion of the host immune system (32). Another gene upregulated by Rot is dltD (Table 4). This gene is a member of the dlt operon, which consists of dltA, dltB, dltC, and dltD (31). This operon encodes proteins involved in d-alanine incorporation into teichoic acid in the S. aureus cell wall. Interestingly, dltA is not present on the list because it is among the genes missing from the GeneChip used in this study (see Materials and Methods). In addition, dltB and dltC are not listed because their fold difference (1.3 for dltB and 1.4 for dltC) did not make the cutoff. The dlt operon has been shown to confer resistance to human antimicrobial agents such as defensins (31). A recent study by Collins et al. has demonstrated that dlt mutants are more efficiently killed by human neutrophils than wild-type strains (9). The same study has also shown that mortality and arthritis frequencies were significantly lower in mice infected with dlt mutants than in mice infected with the wild type strains, indicating that the strains lacking the dlt operon were less virulent than dlt-positive strains. All these factors positively modulated by Rot are hypothesized to be required at an early stage of infection. Some factors, such as the adhesins, allow the bacteria to attach to host cells, while others, such as protein A and DltA, -B, -C, and -D facilitate establishment of the infection through inhibition of the host immune response.

Simultaneously, Rot negatively regulates factors that may interfere with the establishment of the infection. For instance, Rot negatively regulates the production of proteases. One role for proteases is to degrade surface proteins such as protein A, fibronectin-binding proteins, and clumping factor B, which facilitate adherence of the bacteria to host tissues. Degradation of these factors will allow the bacteria to detach from the initial site of infection and start the infection at another location (16, 23). Rot also represses the synthesis of hla and hlb, which encode alpha- and beta-hemolysin, respectively. These hemolysins are known to disrupt a variety of mammalian cells, and because of their function, they are classified as virulence factors involved in invasion and tissue penetration (32). Thus, Rot promotes early stages of infection by stimulating factors needed at this stage and by inhibiting those that interfere with the early processes of infection or that are needed for later processes.

During late stages of infection, which correspond to the post-exponential growth phase, there is a peak of the levels of RNAIII, the agr effector molecule (32). Since our results indicate that Rot and agr have opposing effects on the expression of S. aureus virulence genes, it is likely that RNAIII either directly or indirectly exerts an inhibitory effect on Rot. Since rot is transcribed throughout the growth curve (data not shown) and the rot transcript is present in both RNAIII-positive and RNAIII-negative strains (Fig. 3B), Rot is likely to be inhibited by RNAIII posttranscriptionally. The inhibitory effects of RNAIII on Rot result in the repression of the cell surface proteins and the activation of the secreted proteins that promote the spread of the bacteria and allow them to initiate infection at other sites. Given that Rot was identified in an agr-negative strain, we proceeded to determine the extent of S. aureus gene regulation by Rot in the same background. However, preliminary results indicate that Rot effects on select target genes such as ureC, geh, spl, and ssp operons could be validated in an agr-positive background as well.

In conclusion, the present body of work has employed transcription profiling technology to investigate the spectrum of genes that are regulated by the S. aureus repressor of toxin, Rot. The results demonstrate that Rot is a global transcriptional modulator of virulence genes and that its influence on the expression of certain target genes is opposite that of agr. Genes whose expression is influenced by Rot can be grouped in three major categories: (i) genes that have been extensively studied and have well-defined virulence functions, (ii) genes postulated to play a role in virulence, as their disruption results in attenuation of the mutant in a specific infection model, and (iii) genes not previously known to play a role in S. aureus pathogenesis. By unveiling additional regulatory levels, this study gives us better insight into the agr-SarA regulatory network of S. aureus virulence genes. Additional studies are needed to investigate the molecular interaction of Rot with its target genes and with other regulators, as well as the role of the novel genes regulated by Rot in S. aureus virulence.