Article available online at http://www.idealibrary.com on Microbial Pathogenesis 2001; 31: 81–89

PATHOGENESISMICROBIAL

doi:10.1006/mpat.2001.0450

Identification of a major, CsrRS-regulatedsecreted protein of Group A streptococcusAndrew Heatha, Alita Millerb,c, Victor J. DiRitab,c &N. Cary Engleberga,b∗

Departments of aInternal Medicine, bMicrobiology & Immunology and cLaboratory Animal Medicine,University of Michigan School of Medicine, Ann Arbor, MI 48109, U.S.A.

(Received November 23, 2000; accepted in revised form April 23, 2001)

CsrR/CsrS (CovR/CovS) is a two-component regulator of extracellular virulence factors in GroupA streptococcus, but the full range of regulated exoproteins is unknown. Since CsrR repressesexpression of regulated factors, culture supernates of wild-type and CsrR− mutant strains werecompared by two-dimensional gel electrophoresis (2DGE) to identify regulated exoproteins.Supernates of �csrRS− mutant, but not wild-type, bacteria contained an abundant 23 kDa protein.The N-terminal sequence of this spot corresponded to a putative open reading frame (ORF) in thestreptococcal genome. In a mobility shift assay, phosphorylated CsrR bound to a PCR ampliconthat included sequences upstream of this ORF. By primer extension analysis, the ORF (designatedmspA, for mucoidy-associated secreted protein) was expressed in mid- and late-exponential phasein a �csrRS− mutant. The presence of an in-frame deletion in mspA did not affect colonyappearance, mucoidy or in vitro growth, and there was no difference between �mspA and wild-type strains in a mouse model of skin infection. MspA is co-regulated with other factors requiredfor dermonecrosis (e.g. capsule, streptolysin S and purogenic exotoxin B); however, deletion ofthis gene does not affect expression of hyaluronic acid capsule or severity of skin infection inmice. 2001 Academic Press

Key words: Group A streptococcus, Streptococcus pyogenes, exoprotein, two-component regulators, skininfection, animal models, exotoxins, mspA, site-directed mutagenesis.

regulators have also been termed CovR andIntroductionCovS [3]. The genetic locus encoding CsrR andCsrS was discovered when mutagenesis with

CsrR and CsrS are elements of a two-component Tn916 yielded colonies with a mucoid pheno-regulatory system in Group A streptococcus that type, and subsequent analysis revealed alteredrepresses expression of hyaluronic acid cap- regulation of the capsular synthetic genes, has-sule [1] and several exoproteins, including strep- ABC. Regulation of the exoproteins listed abovetolysin S [2, 3], pyrogenic exotoxin B [2], strep- was demonstrated by performing transcriptionaltokinase and other factors [3]. These two analysis of genes known to encode potential

virulence factors. In its phosphorylated state,∗Author for correspondence. E-mail: cengleb@umich.edu

0882–4010/01/080081+09 $35.00/0 2001 Academic Press

82 A. Heath et al.

the response regulator CsrR binds upstream ofthe capsular synthetic genes and represses theirexpression [4, 5]. Other known regulated geneproducts are also controlled by transcriptionalrepression [5].

We and others have shown that mutations inthe csrRS locus that abrogate expression of CsrRresult in streptococci with mucoid colony mor-phology and the capacity to produce rapidlyexpanding dermonecrotic lesions and lethalityin a mouse model of skin and soft tissue infection[1, 2, 6]. Enhanced expression of capsule isclearly important in this model, since a strainlacking both CsrRS and the capsular syntheticgenes exhibits substantially reduced levels ofskin lesions and necrosis. Preliminary evidencesuggests that the other regulated exoproteinsalso contribute to enhanced virulence [7, 8]. Therepression of a set of secreted virulence factorsby CsrR led us to ask whether there are other,currently unrecognized, exoproteins in this re-gulon that may also play a role in disease.

Results

Identification of CsrR-regulatedexoproteins

To detect streptococcal exoproteins that are reg-ulated by CsrR/CsrS, we examined broth culture

MGAS166 (wt)

AA2392 (∆csrRS–)

supernates of the wild-type strain, MGAS166, Figure 1. Coomassie blue-stained two-dimensionaland a csrR mutant, AA2392, by two-dimensional SDS-PAGE of bacterial culture supernates harvestedgel electrophoresis (2DGE). Among the protein at mid exponential phase. (a) 80-fold concentratedspots resolved by this technique, at least two supernate of wild-type strain MGAS166, (b) 80-foldwere present and abundant in supernates from concentrated supernate of �csrRS− mutant. The blackthe csrR mutant but undetectable in the wild- arrow indicates the position of the intense, CsrR-

repressed protein, MspA. The white arrow indicatestype strain. The densest spot at 23 kDa mobilitya less intense spot that is also not apparent in the(Fig. 1, black arrow) was isolated from the gelsupernate of MGAS166.and submitted to our Biomedical Research Core

Facility for Edman degradation analysis. The N-terminal sequence was reported as: H D N I DE K G K V. A search of the streptococcal genomic neutral protein. The estimated pI of 7.1 is con-

sistent with the migration of the spot on 2DGE.database [9] with a degenerate oligonucleotideyielded a match within a theoretical open read- Searches of protein databases failed to reveal

any significant homologies. The gene encodinging frame at coordinate 1725799 (as of May23, 2000). The N-terminal sequence matched this putative protein was designated as mspA, for

mucoidy-associated secreted protein. The sameperfectly with a string of amino acids beginningat residue 28 of a putative >24 kDa protein. protein was recently identified by Lei et al., in a

study of secreted streptococcal proteins resolvedAnalysis of the first 27 residues of the poly-peptide showed a hydrophobic start of seven by 2DGE [10]. These authors designated this

spot as ‘‘protein 27’’.residues followed by a strongly hydrophobicspan of 15 residues, consistent with a signal A second spot that was present only on gels

of the csrRS mutant (Fig. 1, white arrow) wassequence. The remainder of the sequence (res-idue 28–222) suggests a largely hydrophilic and also isolated and found to have the N-terminal

CsrRS-regulated exoprotein of S. pyogenes 83

a b c d e f a b c d e

Figure 2. Mobility shift analysis of CsrR binding tosequences from the 5′ end of mspA (left). 32P-endlabelled PCR amplicons comprising 5′ sequences ofmspA were mixed with no added CsrR (lane a), two-fold increasing concentrations (from 300 ng to 2.5 �g)of phosphorylated CsrR (lanes b–e), and 2.4 �g un-phosphorylated CsrR (lane f). Note the concentration-dependent shift in the mobility of band when phos-phorylated CsrR is added. As a control (right), 32P-endlabelled PCR amplicons including the rofA promoterwere mixed with no added CsrR (lane a), two-foldincreasing concentrations (from 600 ng to 2.4 �g)of phosphorylated CsrR (lanes b–d), and un-

–35

T G C A a b c

–10

*Start oftranscription

phosphorylated CsrR (lane e). Note that phos-phorylated CsrR did not bind non-specifically to this Figure 3. DNA footprint of the promoter region ofstreptococcal promoter. MspA. Sequencing lanes (TCGA) are shown on the

left. The transcriptional start site and −10 and −35promoter sequences are indicated. Lanes a, b and care DNAse digestions of this fragment; a, no proteinsequence: I F/P G/S A N G E I G/R. Due to added; b, phosphorylated CsrR; c, unphosphorylated

the apparent ambiguity, we were unable to find a CsrR. Note that the protected area in lanes b and cmatch for this sequence in the genomic database. include the −10 and −35 sequences and extendHowever, Lei et al. described a protein (protein upstream.22) that matches the mobility and pI of this spotand has the N-terminal sequence: D S F S A NQ E I R [10]. This protein is a homolog of human of mspA fragments were detected using phos-

phorylated CsrR. However, only minimal shiftCD11b.was observed when non-phosphorylated CsrRwas used, a similar pattern to that observed withother CsrR-regulated promoters [5]. Addition ofRegulation of mspAphosphorylated CsrR to the streptococcal rofApromoter (not regulated by CsrR) did not induceCharacterization of CsrR/CsrS regulation of

other genes has shown that phosphorylated a mobility shift, confirming that CsrR does notbind non-specifically to DNA.CsrR binds at sites overlapping the −10 and

−35 elements of promoters [5]. To confirm that To confirm that CsrR binds to the mspA pro-moter, we performed footprint analysis usingmspA is regulated by CsrR/CsrS, we performed a

mobility shift assay to detect binding of purified both the DNAse I and hydroxyl radical cleavageprotection methods. The results of the DNAseCsrR to sequences upstream of the gene (Fig. 2).

Concentration-dependent shifts in the migration protection assays (Fig. 3) show a footprint of

84 A. Heath et al.

a b c d

MGAS166(wt)

a b c d

AA2392(∆csrRS–)

Figure 4. Transcriptional analysis of mspA in wild-type (MGAS166) and a �csrR mutant strain using primerextension analysis. RNA was extracted from the indicated bacteria during early exponential phase (lane a),mid-exponential phase (lane b), late exponential (lane c) and early stationary phase (lane d). Agarose gelelectrophoresis of the RNA samples used for primer extension is shown directly below the autoradiograms.DNA mobility markers (leftmost lane) were a 1 kb ladder-plus purchased from Gibco BRL (Rockville, MD,U.S.A.).

both the phosophorylated and unphos- findings confirm that expression of mspA is re-pressed by CsrR, and like some other CsrR-reg-phorylated CsrR directly over the −10 and

−35 sequences of the mspA promoter, again in ulated genes, mspA is transcribed mostabundantly from mid to late log phase.keeping with what has been observed with other

CsrR-regulated genes. The peak of mspA transcription correlatedwell with expression of a 24 kDa protein inExpression of mspA at various points in the

broth culture curve were assessed by primer ex- broth culture supernates. SDS-PAGE analysisof supernates from the csrR mutant AA2392tension analysis (Fig. 4). From cultures of the csrR

mutant AA2392, mspA mRNA was detected in all revealed a band with the expected mobilitybeginning in log phase and disappearing insamples but reached peak concentrations in mid

to late log phase. By early stationary phase, the early stationary phase [Fig. 5(a)]. We concludedthat the band migrating at 24 kDa (black ar-amount of specific message was markedly dimin-

ished.mRNAfromthewild-typestraincontained rows) in one-dimensional SDS-PAGE wasMspA based upon the absence of other visibly-little or no mspA RNA; a faint band was barely

discernible in the late log phase sample. Analysis stained supernate proteins with similar mo-bility by 2DGE and in one-dimensional SDS-of rofA transcripts from similar times points in

both the wild-type and CsrRS mutant strain PAGE of �csrRS−, �mspA supernates [Fig. 5(c)].The difference in intensity between SBmuc5showed no comparable variability in the amount

of message (data not shown), suggesting that the and the two other CsrRS mutants may reflectthe slightly earlier entry of this strain intoabsence of CsrR is not associated with a non-spe-

cific increase in late log phase transcription. These stationary phase [see growth curve, Fig. 5(b)].

CsrRS-regulated exoprotein of S. pyogenes 85

a

0.0124

10

Hour of cultivation

AB

S60

0 n

m (

nea

t)

0

0.1

1

12

(b)

LE

ESS

6 18

(a) MGAS166(b) SBmuc5(c) SBmuc7

(d) AA2392 (∆csrRS–)

(a)

b c d e a b c d e a b c d e

LE ES S

a

(c) ∆csrRS

b c d a

∆csrRS, ∆mspA

b c d

Figure 5. SDS-PAGE of culture supernates from wild-type and csrRS− mutants. (a) Wild-type and mutantsupernates were sampled at late exponential phase (LE), early stationary phase (ES) and stationary phase(S). The times at which individual cultures were sampled are shown in the growth curve (b). Samplesinclude supernates of MGAS166 (lane a), two csrR::Tn916 mutants (lanes b and c), AA2392 (�csrRS− ) (laned), and molecular weight markers (lane e). The closed arrow and open arrows indicate the positions of theMspA protein and SpeB. (c) Coomassie blue stain of supernates from cultures of MGAS166 and AA2392sampled at four intervals between late exponential and late stationary phase: lane a, 5 h of cultivation(OD600=0.5–0.6); lane b, 6.75 h (OD600=1.0); lane c, 7.75 h (OD600=1.0); lane d, 8.75 h (OD600=1.0). The blackarrow on the left indicates the mobility of MspA; the open arrow indicates a band previously identified asSpeB by immunoblotting (data not shown).

86 A. Heath et al.

(Table 1). All of the mice inoculated with thewild-type or �mspA derivative gained weighton each of the first two days of the experiment.One of the nine mice that received AA2392(�csrR, csrS− ) gained weight in spite of havinga necrotizing skin lesion. Two others recoveredsome weight by 48 h. These animals contributeto the large variance for this group at 48 h. Allmice in the group that received AA3033 (�csrR,csrR−; �mspA ) lost weight at 24 h; one regainedsome weight at 48 h. The presence of a csrRSmutation in either an MspA+ or MspA− strainyielded dramatic weight loss at both 24 and 48 h(P<0.001 for each isogenic pair at both timepoints; Students t-test).

Discussion

The CsrR/CsrR system regulates several genes0

800

mm

2

100

200

300

400

500

600

700

Wt ∆mspA ∆csrRS– ∆csrRS–,∆mspA

required for virulence in streptococcal skin andsoft tissue infection. We observed that elim-Figure 6. Size and appearance of skin lesions 24 h

after inoculation of >5×106 cfu of MGAS166, csrRS ination of CsrR also results in enhanced ex-and mspA mutants in a mouse model of skin and soft pression of an abundant exoprotein of unknowntissue infection (see Table 1). (Β) Indicates abscess, function, designated MspA. The density of theerythema or no visible lesion; (Χ) indicate lesions spot produced by this protein in a two-di-with necrotic centres. mensional gel electrophoresis analysis suggests

that it is among the most highly expressed pro-teins detected in log phase culture supernatesof the mutant strain. By sequencing of the N-terminus of the protein, we were able to identifyAfter a few hours in stationary phase, none ofthe relevant open reading frame in the strep-the supernatants had detectable MspA.tococcal genomic database. The position of theN-terminal sequence within the translated ORFand the hydrophobicity profile of the leadersequence suggest that this protein is secretedand processed. Two additional features of theVirulence of �mspA mutantsDNA sequence confirm that the ORF we iden-tified encodes MspA. First, the predicted mo-Identical large in-frame deletions were in-

troduced into the mspA genes of wild-type strain lecular mass and charge are both consistent withthe migration of MspA by two-dimensional gelMGAS166 and csrR mutant strain AA2392 to

generate strains AA3111 (�mspA) and AA3033 electrophoresis. Second, purified and phos-phorylated CsrR binds quantitatively to se-(�csrR−, csrS; �mspA), respectively.

In animal inoculation studies, each mspA mut- quences upstream of mspA, suggesting that thisgene is repressed by the CsrR/CsrS system. Theant was compared with its immediate parent

strain injected at the same inoculum. MGAS166 sequential shifting of the mspA probe in themobility shift assays (Fig. 2) is consistent withand its �mspA derivative were given at 106 cfu/

mouse; AA2392 and its �mspA derivative were our recent observation that CsrR binds and oli-gomerizes at target promoters when it is phos-given at >5×106 cfu/mouse. As shown in Fig.

6, the occurrence of lesions and their mean size phorylated [5]. The concentration-dependentshifting of the probe represents increasing num-were comparable in each of the paired groups.

Similarly, there was no significant difference bers of copies of phosphorylated CsrR bound toeach oligonucleotide. Unphosphorylated CsrRin the daily weight change between the mspA

mutants and their immediate parent strains also binds to this probe, but this binding results

CsrRS-regulated exoprotein of S. pyogenes 87

Table 1. Weight change associated with a>5×106 cfu inoculumof MGAS166, csrRS and mspA mutants in a mouse model of skinand soft tissue infection (Fig. 5)

Weight gain at

24 h 48 hStrain n (%±SD) (%±SD)

MGAS166 (wt) 8 15.4±2.0 22.5±3.8AA3111 (�mspA ) 8 11.6±6.4 18.1±7.3AA2392 (�csrRS− ) 9 −7.0±5.4 −9.0±10.2AA3033 (�csrRS−, �mspA ) 8 −8.2±1.1 −10.8±5.1

in a smaller shift at the same concentration of In spite of our lack of knowledge about thefunction of MspA, we sought to determineprotein in the mobility shift assay, suggesting

that fewer copies of the protein are bound to whether this protein contributes to streptococcalpathogenesis, since its expression is tightly con-the promoter in this state. The DNA footprint

experiments (Fig. 3) confirm that both forms of trolled by the CsrR/CsrS regulon. In-frame de-letion of mspA in either the wild-type or CsrRS−CsrR bind to and protect the mspA promoter

from DNAse I digestion, a finding that would background failed to alter the course of skininfection in the hairless mouse model. We con-not have been predicted by the mobility shift

assays. It is possible that both unphosphorylated clude that MspA is not a critical factor in theproduction of murine skin lesions, and thereforeand phosphorylated CsrR bind in the same man-

ner and protect the same region in the protection may have no role in streptococcal virulence. Itis more likely, however, that MspA has a func-assay, but the attraction between the un-

phosphorylated proteins may be insufficient to tion that is not apparent in the animal model thatwe used and might be clarified by evaluating themaintain oligomerization during gel electro-

phoresis [5]. mutant in a different model of streptococcalinfection.The complete peptide sequence of MspA in-

ferred from the ORF did not help us to assigna function to this protein. Scrutinizing the listof partial homologies for those that might have Materials and Methodspotential significance, we noted that MspA doeshave some similarity to the functional domains Bacterial strains and plasmidsof two Bacillus subtilis metalloproteases. We havenot purified MspA to homogeneity and are The streptococcal strains used for this study

were derived from wild-type strain MGAS166.therefore not able to test whether it has pro-teolytic activity. Moreover, comparison of pro- A �csrR, csrS− strain, AA2392, has a 441 bp in-

frame deletion in the csrR reading frame andteolytic activity in the crude supernates is notlikely to yield a dramatic difference since the single base substitution of the start codon of

csrS. Construction of this strain is describedother highly abundant protein seen on our gelsis pyrogenic exotoxin B (SpeB), which is itself a elsewhere [2]. Streptococci were grown in Todd–

Hewitt broth (Difco, Detroit, MI, U.S.A.) sup-cysteine proteinase. Since MspA rapidly dis-appears from culture supernates at the same time plemented with 0.2 yeast extract (THYB) or on

Todd–Hewitt agar plates. When antibiotic se-that SpeB appears, we considered the possibilitythat MspA is a substrate for SpeB. However, lection of streptococci was required, ery-

thromycin 0.02 �g/ml, streptomycin 100 �g/mlrecent analysis of culture supernates of a�CsrRS−, speB mutant generated in our laborat- or spectinomycin 100 �g/ml (Sigma Laborat-

ories, St. Louis, MO, U.S.A.) were added to theory suggests that MspA disappears from themedia in stationary phase even in the absence appropriate media. Escherichia coli strain DH5�

was used for cloning. For selective growth,of SpeB (unpublished data). We conclude thateither MspA has a very short half-life, or there Luria–Bertani media was supplemented with

ampicillin 50 �g/ml or erythromycin 300 �g/ml,is at least one other mechanism of regulationother than CsrRS. as required.

88 A. Heath et al.

In-frame, site-directed deletions of mspA in Transcriptional analysesMGAS166 and AA2392 were generated by anallelic exchange method described by Perez– mspA transcripts were visualized using primer

extension analysis of lysates taken at early-logCasal and co-workers [11]. With this method, in-frame deletions of target genes were constructed (OD600=0.3), mid-log (OD600=0.6), late log

(OD600=0.9), and early stationary phase (OD600=and inserted into pJRS233. This temperature-sensitive plasmid confers erythromycin re- 1.2). RNA was isolated using the method of

Cheung and co-workers [14] and the FastPrepsistance. Insertions of the plasmid onto the strep-tococcal chromosome were selected by growth system (Bio101, Vista, CA, U.S.A.). Bacteria were

shaken with beads for 45 s at 6000 rpm to lyseat the non-permissive temperature on ery-thromycin. Co-integrates were then passed on cells in CRSR 455 (chaotropic RNA stabilizing

reagent 455). Comparability of RNA yield wasmedium without erythromycin and screened forplasmid deletion by PCR. The presence of the confirmed by comparison of samples after

electrophoresis in agarose gels and ethidiumdeletions on the mutant chromosomes was alsoconfirmed by PCR analysis. bromide staining. The DNA primer used for

extension was 5′-AGCATCGAAGCTGCTAG-3′,Deletions in mspA were generated by ligatingamplicons from the upstream PCR primer pair, and the reverse transcription was catalysed by

Superscript⊂ (GIBCO/BRL).5′-GCCGGCGGATCCGATAATCCCGCCCGT-CTTT-3′ and 5′-GCGGCCGCTGGGCCCTCG-AAGCTGCTGCTAGTAGTGAAGTGA-3′ withamplicons from the downstream primer pair, 5′- DNA binding studiesCCCGCGCTGCAGCTGGAGTCATGGGGAAA-GATTA-3′ and 5′-GGGCCCAGCGGCCGCA- DNA binding studies were performed using a

200 bp fragment of mspA generated by per-CAACCCATTAGCAGCATCATTCC-3′ in a PCRreaction with splicing by overlapping extension forming PCR on wild-type genomic DNA. To

amplify the promoter region of mspA, 5′-(‘‘SOEing’’) [12]. The resulting fragment wascloned between the BamHI and PstI sites of TTGCTTCCATAACCC-3′ was the forward

primer, and the reverse primer was the samepNEB193 to generate pAA2696, and then sub-cloned between the BamHI and PstI sites of one used in the primer extension analysis. PCR

amplicons were end-labelled �32P-ATP (Amer-pJRS233 to generate pAA2952. Ligation of thesetwo fragments generated a copy of mspA with sham) and 10 u T4 kinase (Gibco-BRL) in 70 mM

Tris pH 6.4, 10 mM MgCl2 and 5 mM DTT fora 564 bp inframe deletion which was confirmedby DNA sequencing across the deletion site in 30 min at 37°C. Free nucleotide was removed by

passage of the labelled probe through a G-25the plasmid pAA2696. In the translated protein,188 residues are deleted from the predicted poly- spin column (Roche). As a control for non-spe-

cific DNA binding, a similar labelled probe ofpeptide of 221 residues (deletion of codons 17–204). the rofA promoter was prepared using the

primers 5′-CCTTTAAGCTATTTC-3′ (forward)and 5′-GTGCTACCTCAGTTATTGGC-3′ (re-verse).

To generate purified CsrR for these assays,Gel electrophoresisthe csrR gene was amplified by PCR and clonedinto pQE30 (Qiagen) to create a translationalSupernates from broth cultures were con-

centrated 80-fold by a series of ultrafiltrations fusion between a His6 N-terminal tag and wild-type CsrR protein. The recombinant plasmidusing Centriprep⊂ and Microcon⊂ devices with

a 10 kDa cutoff (Amicon Inc., Beverly, MA, was then overexpressed in E. coli M15/pREP4,and His6-CsrR was purified from the bacterialU.S.A.). Sodium dodecyl sulfate polyacrylamide

gel electrophoresis (SDS-PAGE) was performed lysate on a Ni++-agarose column following themanufacturer’s protocol. Purified His6-CsrR wasusing the discontinuous system of Laemmli [13].

For two-dimensional gel electrophoresis, sep- phosphorylated as described by Quon et al. [15].For mobility shift assays, 10 000 cpm of thearation and detection of proteins was achieved

by performing equilibrium isoelectric focusing labelled probe was incubated for 20 min at roomtemperature with various concentrations of His6-followed by separation in the second dimension

on 15% SDS-PAGE. The gels were stained with CsrR in DNA binding buffer (25 mM Tris pH 8.0,50 mM KCl, 0.5 mM EDTA, 0.5 mM DTT, 5 mMCoomassie blue to visualize protein spots.

CsrRS-regulated exoprotein of S. pyogenes 89

synthesis in group A streptococcus. Mol Microbiol 1998;MgCl2, 3 mM CaCl2, 4% glycerol, 1 mg/ml BSA,30: 209–19.10 �g/ml salmon sperm DNA). Samples were

2 Heath A, DiRita VJ, Barg NL, Engleberg NC. A two-analysed by native gel electrophoresis at 4°C in component regulatory system, CsrR/CsrS, repressesa 6% agarose 0.5× TBE gel. expression of three Streptococcus pyogenes virulence fac-

For DNAse I footprinting, 30 000 cpm of the tors, hyaluronic acid capsule, streptolysin S, and pyro-genic exotoxin B. Infect Immun 1999; 67: 5298–305.mspA probe was incubated for 20 min at room

3 Federle M, McIver K, Scott J. A response regulator thattemperature with 9.6 �g of His6-CsrR in DNArepresses transcription of several virulence operons inbinding buffer (see above) at a total volume ofthe group A streptococcus. J Bacteriol 1999; 181: 3649–57.30 �l. Forty mU of DNAse I (Roche) was added 4 Bernish B, Van de Rijn I. Characterization of a two-

in a 1 �l volume, and after 2 min the reaction component system in Streptococcus pyogenes which iswas stopped by addition of 200 �l of 200 mM involved in regulation of hyaluronic acid production.

J Biol Chem 1999; 274: 4786–93.NaCl, 2 mM EDTA, 1% SDS. The reaction prod-5 Miller AA, Engleberg NC, DiRita VJ. Repression ofucts were extracted with phenol:chloroform, eth-

virulence genes by phosphorylation-dependent oli-anol precipitated and resuspended in TE.gomerization of CsrR at target promoters in S. pyogenes.Samples were analysed in a 6% sequencing gel Mol Microbiol 2001; 40: 976–90.

by autoradiography. 6 Bunce C, Wheeler L, Reed G, Musser J, Barg N. Amurine model of cutaneous infection with gram positivecocci. Infect Immun 1992; 60: 2636–40.Assessment of virulence in an animal

7 Betschel SD, Borgia SM, Barg NL, Low DE, De AzavedomodelJC. Reduced virulence of group A streptococcal Tn916mutants that do not produce streptolysin S. Infect ImmunVirulence of the �mspA strains was assessed 1998; 66: 1671–9.

using mouse model of infection and der- 8 Kulomeski S, Montgomery CA, Rurangirwa J et al.monecrosis, as previously described [6]. Briefly, Extracellular cysteine protease produced by Strepto-

coccus pyogenes participates in the pathogenesis of in-streptococci were harvested at mid-log phasevasive skin infection and dissemination in mice. Infectand diluted to produce a suspension ofImmun 1999; 67: 1779–88.5×105–5×106 cfu/ml in different experiments.

9 Roe BA, Linn SP, Song L et al. Streptococcal genomeA 200 �l volume of this suspension was injected sequencing project. Vol. 2000: University of Oklahomasubcutaneously into the right flank of hairless, (USPHS/NIH grant #AI38406), 2000.4 week old male crl:SKH1 (hrhr) Br mice (Charles 10 Lei B, Mackie S, Lukomski S, Musser JM. IdentificationRiver, Wilmington, MA, U.S.A.). Mice were and immunogenicity of Group A streptococcus culture

supernatant proteins. Infect Immun 2000; 68: 6807–18.weighed prior to inoculation and every 24 h.11 Perez-Casal J, Price JA, Maguin E, Scott JR. An MNecrotic wounds were measured daily using

protein with a single C repeat prevents phagocytosisthe equation: Area=�×L×W/4, where L is the of Streptococcus pyogenes: use of a temperature-sensitivelong axis and W is the short axis of the lesion. shuttle vector to deliver homologous sequences to the

chromosome of S. pyogenes. Mol Microbiol 1993; 8: 809–19.Acknowledgements

12 Higuchi R. Recombinant PCR. In: Innis MA, GelfandDH, Sninksy JJ, White TJ, Eds. PCR Protocols: a guide tomethods and applications. San Diego, CA: Academic Press,N.C.E. is supported by Public Health Service GrantInc., 1990: 177–183.RO1-141682. A.M. is supported by Public Health

13 Laemmli UK. Cleavage of structural proteins duringService Grant F32-GM20420. Technical assistance wasthe assembly of the head of bacteriophage T4. Natureprovided by GCRC grant no. MO1-RR00042. 1970; 117: 680–5.

14 Cheung AL, Eberhardt KJ, Fischetti VA. A method toisolate RNA from Gram-positive bacteria and my-Referencescobacteria. Anal Biochem 1994; 222: 511–4.

15 Quon KC, Marczynski GT, Shapiro L. Cell cycle controlby an essential bacterial two-component signal trans-1 Levin JC, Wessels MR. Identification of csrR/csrS, a

genetic locus that regulates hyaluronic acid capsule duction protein. Cell 1996; 84: 83–93.