Present addresses: Eijkman Winkler Institute for Medical Microbiology, Infectious Diseases and Inflammation, University Hospital Utrecht, Utrecht, the Netherlands.
Synthesis of group A streptococcal virulence factors is controlled by a regulatory RNA molecule
Article first published online: 16 JUL 2004
Volume 53, Issue 5, pages 1515–1527, September 2004
How to Cite
Mangold, M., Siller, M., Roppenser, B., Vlaminckx, B. J. M., Penfound, T. A., Klein, R., Novak, R., Novick, R. P. and Charpentier, E. (2004), Synthesis of group A streptococcal virulence factors is controlled by a regulatory RNA molecule. Molecular Microbiology, 53: 1515–1527. doi: 10.1111/j.1365-2958.2004.04222.x
- Issue published online: 3 AUG 2004
- Article first published online: 16 JUL 2004
- Accepted 11 May, 2004.
The capacity of pathogens to cause disease depends strictly on the regulated expression of their virulence factors. In this study, we demonstrate that the untranslated mRNA of the recently described streptococcal pleiotropic effect locus (pel), which incidentally contains sagA, the structural gene for streptolysin S, is an effector of virulence factor expression in group A beta-haemolytic streptococci (GAS). Our data suggest that the regulation by pel RNA occurs at both transcriptional (e.g. emm, sic, nga) and post-transcriptional (e.g. SpeB) levels. We could exclude the possibility that the pel phenotype was linked to a polar effect on downstream genes (sagB-I). Remarkably, the RNA effector is regulated in a growth phase-dependent fashion and we provide evidence that pel RNA expression is induced by conditioned media.
Streptococcus pyogenes or group A beta-haemolytic streptococcus (GAS) is an important human pathogen that causes a wide variety of diseases, ranging from throat and skin infections, such as pharyngitis and erysipelas, to severe invasive diseases, such as necrotizing fasciitis and streptococcal toxic shock syndrome (Carapetis et al., 1995; Efstratiou, 2000; Stevens, 2002). This is reflected by a large number of secreted proteins, e.g. streptococcal pyrogenic exotoxins (Spe’s), the haemolysins streptolysin S (SLS) and streptolysin O (SLO), and an array of matrix-binding proteins that all contribute to virulence by mediating adhesion to host tissues, evasion of host defence mechanisms, invasion, survival and tissue damage (Darmstadt et al., 2000; Talay et al., 2000; Madden et al., 2001).
The availability of four different annotated genome sequences of an M1, M3, M5 and M18 strain has revealed, in addition to more than 40 putative virulence-associated genes, 13 potential two-component signal transduction systems (Ferretti et al., 2001; Beres et al., 2002; Smoot et al., 2002). These findings point out the existence of highly sophisticated regulatory networks, in addition to the already described regulatory components (Wessels, 1999; Kreikemeyer et al., 2003).
The best studied control element in GAS is the multiple gene activator Mga, which represents a typical response regulator component of classical signalling systems, activating the expression of a number of genes encoding surface-associated and secreted proteins such as emm (M protein), mrp, arp and enn (M-related proteins) sclA (the collagen-like proteins SclA), scpA (C5a peptidase), sof (serum opacity factor), sic (streptococcal inhibitor of complement) and speB (cysteine protease) (Caparon and Scott, 1987; Podbielski et al., 1995; McIver and Scott, 1997; Kreikemeyer et al., 1999; 2003; Rasmussen et al., 2000; Frick et al., 2003). However, the putative sensor protein that links environmental signals and activation of Mga remains unknown.
Among the complete two-component systems of GAS, the best described is CsrRS/CovRS, which negatively controls expression of several virulence factors, including the hyaluronic acid capsule, SpeB, streptokinase, SLS and streptodornase (Levin and Wessels, 1998; Federle et al., 1999; Heath et al., 1999). More recently, a study using global expression profiling showed that CsrR/CovR influences on a grand scale expression of genes encoding exoproteins, genes involved in adaptive responses and transcriptional regulators including other two-component systems (Graham et al., 2002). A second two-component regulatory system, FasBCA, is involved in the regulation of expression of fbp54 and mrp (adhesins), as well as expression of sagA (SLS) and ska (streptokinase) (Kreikemeyer et al., 2001). Interestingly, the authors suggest that fasX, a putative non-translated RNA, could be an effector of the fas regulon.
Until recently, control of transcription by repressor or activator proteins was considered to be the main regulatory mechanism for gene expression. In bacterial pathogens, only in the last few years, evidence has been obtained that RNA-regulated gene expression might be more common than previously anticipated (Anderson and Schneewind, 1997; Altuvia and Wagner, 2000; Gottesman et al., 2001; Ma et al., 2001; Shimizu et al., 2002; Johansson and Cossart, 2003). In addition to the well-known agr-system in Staphylococcus aureus with RNAIII being its effector molecule (Novick et al., 1993; Morfeldt et al., 1995), the very recent work by Johansson and coworkers has shown that the control of virulence gene expression in Listeria monocytogenes is dependent on an RNA thermosensor, the untranslated 5′-end of the prfA transcript (Johansson et al., 2002).
Recently, the analysis of a Tn917 transposon library of an M49 GAS strain revealed a gene locus, pleiotropic effect locus (pel), with a pleiotropic effect on the expression of a number of streptococcal exoproteins (Li et al., 1999). The pel locus comprises sagA, which is the structural gene for SLS (Nizet et al., 2000). The genes located downstream of the sagA gene, sagB to sagI, have been shown to be required for the proper secretion of SLS, because they are involved in processing and export of SLS. As described, the pel locus exerts its effects on virulence factors, including M-protein, SpeB and Ska (Li et al., 1999; Biswas et al., 2001). Another report showed that in vivo passage of a pel-deficient mutant selected for mutants with restored production of SLS and M-protein but not SpeB or Ska (Eberhard et al., 2001). All these isolates had retained the original Tn917 insertion in sagA, suggesting the involvement of another regulatory system or the formation of by-pass mutations. In contrast to these reports, an earlier study of a Tn916 sagA mutation failed to show any effects, in serotype M1 and M18 GAS strains, on the production of virulence factors other than SLS (Betschel et al., 1998). In a sagA-deficient M6-type mutant, emm transcription was not affected; however, the M protein was truncated, lacking its C-terminal domain required for anchoring it to the cell surface (Biswas et al., 2001). These discrepancies, which could be strain dependent, require further investigation.
In the present study, we have established the role of the pel locus as a positive regulator of important streptococcal virulence factors, like M-protein, Sic and SpeB in an M1 GAS background. Furthermore, we could demonstrate that the pel-encoded regulator is the untranslated pel RNA, and that the synthesis of pel RNA is induced by conditioned media. To avoid any confusion, we will refer in this manuscript to sagA as the structural gene coding for SLS and to the pel locus as the entire 459 bp large transcript, which comprises the sagA gene from position 147–308.
Transcriptional and post-transcriptional regulation of virulence factors by the streptococcal pel locus
The pel locus, comprising the structural gene sagA coding for the streptococcal haemolysin SLS, has been described to be crucial for the regulation of several virulence factors. To investigate the mechanism behind this phenomenon, we mutated the pel locus in the sequenced M1 GAS strain, RDN29, by replacing the sagA gene with the kanamycin resistance cassette gene, aphIII. The orientation of the aphIII gene, which contains its own promoter and transcriptional terminator, was inverted towards the original direction of transcription of the sagA gene. SDS–PAGE analysis of exoprotein patterns of RDN29 and its isogenic haemolysis-negative, sagA-deficient mutant RDN17 demonstrated striking differences (Fig. 1). The expression of at least nine proteins was clearly reduced in the sagA-deficient mutant. After tryptic digest of the respective gel bands, the protein fragments were subjected to mass-spectroscopy and sequencing. Six proteins could be assigned unambiguously to genes on the M1 genome (Table 1). These data need to be confirmed by performing immunoassays and activity assays.
|1||Not identified||≈ 140|
|2||Not identified||≈ 80|
|3||NAD-glycohydrolase (nga)||≈ 55|
|4||Streptokinase A (ska)||≈ 52|
|5||Not identified||≈ 40|
|6||Streptococcal inhibitor of complement (sic)||≈ 38|
|7||Putative deoxyribonuclease (mf3)||≈ 34|
|8||Putative DNase (mf2)||≈ 28|
|9||Streptococcal pyrogenic exotoxin C (speC)||≈ 27|
To examine whether the reduced level of virulence protein expression in the sagA-deficient mutant RDN17 was the result of transcriptional alterations, Northern blot analysis was used. The genes coding for Sic, M-protein and NAD-glycohydrolase showed reduced transcriptional expression levels in RDN17 (Fig. 2A), whereas transcription of speB and speC was not affected (data not shown). To further determine the level of post-transcriptional regulation of SpeB expression by pel, secreted SpeB was examined for its ability to proteolyse casein (Tsai et al., 1998). In the wild-type strain, specific cysteine protease activity appeared abruptly during late logarithmic phase, peaked during the transition into stationary phase and then started to decline after several hours in stationary phase. In the sagA-deficient mutant RDN17, the onset of cysteine protease activity was clearly delayed compared to that in the wild-type strain (Fig. 2B). SpeB activity in RDN17 appeared at a time point (late stationary phase) when the activity started to decay in the wild-type strain. Western blot analysis using specific antibodies against the zymogen and the mature form of SpeB (gift of Lars Björck) (Lyon and Caparon, 2003) demonstrated that the sagA-deficient mutant RDN17 secreted efficiently the zymogen of SpeB but its processing to the active mature form appeared to be delayed. These results suggest a possible effect of pel on processing and maturation of SpeB. Taken together, our data indicate that pel influences virulence factor production at transcriptional and post-transcriptional level.
The sagA-downstream genes, sagB-I, are not responsible for the pel phenotype
The sagA gene replacement in strain RDN17 could exhibit a polar effect on the downstream sagB-I operon. To rule out the possibility that sagB-I genes would be responsible for the observed phenotypes, we created the sagBC-, sagDEF- and sagGHI-deficient strains, EC540, EC560 and EC566 respectively. Northern blot analysis of the mutants revealed a transcriptional expression pattern of emm and sic identical to the one in the parent strain RDN29 (Fig. 3). This result clearly indicated that the sagB-I genes are not responsible for the effects seen in RDN17. To further provide evidence for the role of the pel locus in expression of virulence factors, we investigated whether in the sagA-deficient strain RDN17, transcription of the downstream sagB-I operon was affected. For this purpose, reverse transcriptase polymerase chain reaction (RT-PCR) experiments of RDN17 were performed and compared to RDN29, using for the reverse transcription reactions primer 177, which anneals downstream of the sagI gene. The analysis showed that the genes sagE and sagH were transcribed in both strains, RDN17 and RDN29 (data not shown). To unambiguously link the pel locus to the described phenotype, RDN17 was complemented with the wild-type allele of sagA under the control of its own promoter. A DNA fragment containing the transcriptional start site of pel, determined by primer extension (data not shown), plus the −35 and −10 regions as well as the putative transcriptional terminator, was cloned into the multiple cloning site (MCS) of pMSP3535 (gift of Gary Dunny) to create pRDN50. RDN17 was transformed with pRDN50 to obtain strain RDN435. Northern blot analysis demonstrated a restoration of emm and sic transcription to wild-type levels in strain RDN435 (data not shown). In addition, restoration of β-haemolysis was observed. Taken together, the data indicate that the observed effects on virulence factor expression are linked to the pel locus itself and do not result from a polar effect on downstream-located genes.
Untranslated pel RNA is the effector for changes in virulence factor expression
To investigate whether a translational product of pel RNA or the untranslated pel RNA regulates virulence factor expression, translation of sagA was prevented by mutating its RBS and start codon directly on the chromosome. For this purpose, a fragment of pel, containing a mutated RBS (GGAGG to GTAAA) and start codon (ATG to CTG), was cloned into the MCS of a temperature-sensitive shuttle vector pRDN18, creating pRDN20. Plasmid pRDN20 was then used to mutate the sagA RBS and start codon on the chromosome of wild-type strain RDN29, creating RDN165. Because of the lack of sagA mRNA translation, the β-haemolysis phenotype was lost in this strain. To exclude the possibility that the mutation would affect expression of the sagB-I downstream genes, we complemented strain RDN165 with pRDN50, which expresses the wild-type allele of pel. The complemented strain was β-haemolytic, thus indicating that in strain RDN165, the sagB-I genes were expressed. RT-PCR analysis, performed as described above, confirmed transcription of sagB-I genes in this strain (data not shown).
To exclude the possibility of accelerated RNA degradation because of the missing protection of ribosome binding in the SLS-deficient mutant RDN165, RNA stability assays were performed. For this purpose, wild-type strain RDN29 and its isogenic mutant RDN165 were grown to late mid-logarithmic phase and exposed to the inhibitor of transcription, rifampicin (250 µg ml−1). After exposure to rifampicin, six RNA samples of both strains were obtained from time point zero up to 45 min. Northern blot analysis demonstrated no differences in transcript stability between strains RDN29 and RDN165 (Fig. 4). Interestingly, pel mRNA was very stable. The half-life was about 30 min, considerably longer than that of the mRNA of speC, 10 min.
RDN165 was then analysed for the pel phenotype. The azocasein proteolytic assay demonstrated that the kinetics of SpeB-associated cysteine protease activity of RDN165 supernatants were identical to the one of wild-type RDN29 supernatants (Fig. 5). Similarly, Northern blot analysis showed no differences in emm and sic transcription, comparing strains RDN29 and RDN165 (Fig. 6). Furthermore, complementation of the sagA-deficient mutant RDN17 with pRDN51, which expresses the pelRBSI*ATG* mutant form of the pel allele, demonstrated a restoration of emm and sic transcription to wild-type levels (data not shown). These findings suggest that the untranslated pel RNA and not a translational product of pel RNA functions as a regulator for GAS virulence factor expression.
Analysis of a possible alternative translational start site in the pel sequence
Inspection of the pel locus sequence revealed in another reading frame, within the coding sequence of sagA, a potential second coding sequence with a plausible Shine-Dalgarno site, GGAGG, designed RBSII, followed by a theoretical start codon TTG, 7 bp downstream. A stop codon, TAA, is located 45 bp further downstream. Translation of this sequence would correspond to a 15-amino-acid peptide, which might have regulatory activity. Coincidentally, the respective potential Shine-Dalgarno sequence encodes in the sagA coding sequence two glycine amino acids (GGA GGC), which have been suggested to function as the leader peptide cleavage site of the SLS protein (Nizet et al., 2000). Taking advantage of this set-up, the putative RBSII was mutated using the same approach used for the sagA RBS, in a way that the coding region encodes valine and serine (GTA AGC), abolishing the initiation of translation at the respective second RBS as well as cleavage of the hypothesized SLS leader peptide. A PCR-generated pel mutant allele containing these mutations was cloned into the MCS of the temperature-sensitive plasmid pRDN18, creating pRDN32. Plasmid pRDN32 was used to mutate the potential second RBS on the chromosome of wild-type strain RDN29, creating strain RDN423. RDN423 showed a β-haemolysis-deficient phenotype presumably because of its incapacity to cleave the SLS leader sequence. Northern blot analysis of emm and sic transcription showed the same phenotype for RDN29 and RDN423 (Fig. 6). Furthermore, the SpeB activity assay displayed identical kinetics of cysteine protease activity of supernatants of RDN423 and wild-type parent RDN29 (Fig. 5). This suggests that a possible translational product originating from the tentative second RBS does not act as the regulator of virulence factor expression.
Temporal regulation of pel transcription and induction by conditioned media
To determine whether pel transcription is regulated in a temporal fashion, RNA extracts were prepared during growth of strain RDN29 from lag to stationary phase. Northern blot analysis revealed that the pel transcript was upregulated during mid-logarithmic phase followed by an exponential increase of the signal reaching a maximum at early stationary phase (Fig. 7A). The temporal expression of the pel transcript suggested a possible regulation of gene expression in response to changes in cell density. To test for a possible effect of conditioned media on the expression of pel RNA, culture supernatants of strains RDN29 and RDN17 were collected at different time points during growth. The conditioned media were added in a ratio of 1:1 to lag phase cultures of RDN29, a time where no pel transcript could be detected. After 60 min of incubation, the cells were harvested for Northern blot analysis, which demonstrated a clear activation of the pel transcript. Conditioned media derived from strain RDN17 also induced transcription of pel indicating that the pel locus itself does not encode the activator substance (Fig. 7B).
A recent study showed that a Tn917 insertion (designated pel-1) in the sagA promoter of the serotype M49 GAS strain CS101 led to reduced transcription of several virulence factors (Li et al., 1999). Similar effects on virulence factor expression were observed in three different M1 serotype isolates after having been transduced with the Tn917-associated mutation in the sagA gene (Li et al., 1999). As outlined in Introduction, other phenotypes of pel mutants, some of them contradictory, have been described (Betschel et al., 1998; Biswas et al., 2001; Fontaine et al., 2003). However, the possible mechanism involved has not been addressed.
Our study showed a clear effect of pel on virulence factor expression in an M1 GAS strain. Three possible mechanisms are: (i) SLS or a processed form of the secreted peptide could be the regulatory molecule, (ii) the untranslated pel RNA itself could act as a regulator, similar to RNAIII in S. aureus and (iii) other downstream-located sag genes could be responsible for the observed effects. To exclude the possibility that the observed phenotype in the sagA-deficient mutant RDN17 could be the result of a polar effect on the downstream-located sag-operon, we created isogenic sagBC-, sagDEF- and sagGHI-deficient strains. Transcription of the emm and sic genes to wild-type levels in these strains clearly linked the observed phenotype to the pel locus. This was further confirmed by complementation of RDN17 with the wild-type allele of pel and by RT-PCR. Notably, a very recent paper by Fontaine and colleagues seems to support our findings, because a non-polar deletion mutation of the sagB gene did not lead to pleiotropic effect on the expression of virulence factors (Fontaine et al., 2003). However, the authors have proposed, on this basis, that the pel-associated phenotypes described earlier are not necessarily linked to the sagA gene. We understand the findings of Fontaine and colleagues differently. As to be expected, their sagB mutant had no effect on transcription of sagA and lacked SLS activity. This is in accordance with the data of Nizet and colleagues that sagB is an accessory gene required for the proper processing of SLS (Nizet et al., 2000).
As outlined in our study, we used a point mutation approach of the respective RBS and start codon of sagA, which allowed us to assign the observed effects directly to the mutated gene, avoiding the potential problems of polar effects. Furthermore, in contrast to vector complementation studies, the observed effects are independent of plasmid stability, copy numbers and the presence of antibiotics. This seems to be in particular important because it has been described that in the presence of antibiotics such as erythromycin effects of, e.g. RNAIII target gene transcription, are mimicked (Novick et al., 1993). By preventing sagA from being translated, it could be established that pel RNA rather than a sagA gene product is an effector of virulence factor expression in GAS. This finding suggested that the pel RNA shares at least at first sight some characteristics with the staphylococcal RNAIII molecule. First and foremost both RNA molecules are effectors of virulence factor expression, encoding at the same time a haemolysin, the staphylococcal δ-haemolysin and the streptococcal β-haemolysin SLS (Peng et al., 1988; Morfeldt et al., 1995). Similar to RNAIII, the pel RNA seems to mediate its effect on both the transcriptional and post-transcriptional levels. Furthermore, RNAIII and pel RNA are of similar size, 512 bp and 459 bp respectively. Thermodynamic calculations of the RNA structure using the mfold RNA algorithm of Zuker (http://www.bioinfo.rpi.edu/applications/mfold/old/rna) predicted a considerable RNA stability of pel (−139 kcal per mole) very similar to that of RNAIII (−129 kcal per mole).
However, at the same time some characteristics of RNAIII are obviously not shared by the pel locus. RNAIII is the most downstream element of the staphylococcal global regulator agr, characterized by the classical set-up of Gram-positive signalling cascades (Ji et al., 1995). Beside a number of other factors, activation of RNAIII transcription is triggered by the release of the autoinducing thiolactone peptide, AgrD, which is sensed by the dedicated two-component regulatory system (Ji et al., 1997; Mayville et al., 1999). Classical structural genetic elements similar to those of the staphylococcal agr-system, e.g. two-component systems or genes encoding signalling molecules, have not been identified in the vicinity of the pel locus.
In addition to the staphylococcal RNAIII, recent studies in bacterial pathogens show a growing number of examples, which demonstrate that RNA can control virulence gene expression (Anderson and Schneewind, 1997; Altier et al., 2000; Ma et al., 2001; Johansson et al., 2002; Shimizu et al., 2002). However, in many cases the mode of action by which RNA mediates its role on gene expression is not well understood. This also applies to recent work by Kreikemeyer and colleagues, which has suggested the existence of a regulatory RNA molecule, fasX, in GAS (Kreikemeyer et al., 2001). The Fas operon shares some features with the agr-system. Despite the presence of a two-component system, the Fas system does not contain an autoinducing peptide. In addition, although it is regulated in a growth phase-dependent manner, its regulation is not driven by a quorum-sensing mechanism, a key feature of, for example, the staphylococcal agr-system. Although, as mentioned above, we were unable to localize any accessory components, like two-component systems or transporters, in the vicinity of the pel RNA, our data still raise the question whether the identified pel locus is part of a more globally acting streptococcal regulatory system. One clue for the existence of such a system is the finding that transcription of pel is growth phase-dependent. Addition of conditioned media to early logarithmic cells triggered efficiently transcription of pel, a phenomenon characteristic for the existence of soluble activator substances. It is important to point out that this observation is not shared in a study performed by Kreikemeyer and colleagues. They investigated the effects of conditioned media on luciferase reporter activities using a pel-luciferase gene fusion and observed no relevant effect (Kreikemeyer et al., 2001). The differences are possibly explained by a different experimental set-up or the use of different strains. In case conditioned growth media is added at a time where transcription of pel has been already endogenously initiated, we were also not able to induce transcription, possibly because of a refractory period similar to the one described for RNAIII transcription. The answer to this question can be only addressed conclusively after the purification and characterization of the activator substance(s).
The results presented in this study establish the pel locus as a positive regulator of virulence factor expression in GAS. In particular the pronounced effects on emm and sic transcription underscore the importance of this locus. Although many questions remain to be answered, in particular how pel RNA mediates its effect, we clearly demonstrated that the untranslated pel RNA is an effector of regulation of virulence factor expression in GAS. Growth phase-dependent regulation of the pel transcript and induction of the transcript by conditioned media suggests the presence of other components that influence the regulation of this RNA molecule. Further insights in the obviously complex function of the pel/sagA locus should not only aid to understand the pathogenesis of GAS, but might also contribute to smooth the way for a different perception of heterogeneous bacterial species.
Bacterial strains and growth conditions
Bacterial strains used in this study are described in Table 2. Escherichia coli DH5α and TOP10 were used as hosts for plasmid construction. The strains were grown in Luria-Bertani medium either in liquid with shaking or on agar plates at 37°C. S. pyogenes was cultured in Todd-Hewitt broth supplemented with 0.2% yeast extract without agitation or on tryptic soy agar supplemented with 3% sheep blood; minimal medium was also used for liquid cultures (Van de Rijn and Kessler, 1980). Cultures were grown at 37°C in a 5% CO2 to 20% O2 atmosphere. When required, antibiotics were added to the medium to the following final concentrations: erythromycin 300 µg ml−1 for E. coli and 3 µg ml−1 for S. pyogenes; kanamycin 25 µg ml−1 for E. coli and 300 µg ml−1 for S. pyogenes; spectinomycin 100 µg ml−1 for both E. coli and S. pyogenes. Growth of S. pyogenes cells was turbidimetrically monitored at 620 nm with a microplate reader (SLT Spectra Reader).
|Strain or plasmid||Relevant characteristics||Source|
|RDN29||M1 serotype||ATCC 700294|
|RDN17||Isogenic sagA-deficient mutant of RDN29; sagA::aphIII||This study|
|EC540||Isogenic sagBC-deficient mutant of RDN29; sagBC::aphIII||This study|
|EC560||Isogenic sagDEF-deficient mutant of RDN29; sagDEF::aphIII||This study|
|EC566||Isogenic sagGHI-deficient mutant of RDN29; sagGHI::aphIII||This study|
|RDN163||RDN29 (pRDN20)||This study|
|RDN164||RDN29 (pRDN20, integrated)||This study|
|RDN165||Isogenic pel/sagA mutant of RDN 29; RBSI* (GGAGGGTAAA) and start codon* (ATGCTG)||This study|
|RDN289||RDN29 (pRDN32)||This study|
|RDN420||RDN29 (pRDN32, integrated)||This study|
|RDN423||Isogenic pel/sagA mutant of RDN 29; RBSII* (GGAGGGTAAG)||This study|
|RDN253||RDN17 (pMSP3535)||This study|
|RDN435||RDN17 (pRDN50)||This study|
|RDN436||RDN17 (pRDN51)||This study|
|EC576||RDN165 (pMSP3535)||This study|
|EC575||RDN165 (pRDN50)||This study|
|DH5α||Host for cloning||Laboratory strain collection|
|TOP10||Host for cloning||Invitrogen Life Technologies|
|pUC19||ColE1ori, AmpR., lacZ||New England Biolabs|
|pBCSK+||ColE1ori, CmR., f1(+) ori, lacZ||Stratagene|
|pRDN18||repAts-pWV01, aad9, ColE1ori||This study|
|pMSP3535||ermB, repD-E-G-pAMβ1, nisRK, ColE1ori, PnisA||Gary Dunny|
DNA restriction endonuclease digestions, ligations, agarose gel electrophoresis and DNA amplifications by PCR were performed according to standard techniques (Sambrook et al., 1990). DNA purification and plasmid preparations were performed using kits from Qiagen (Qiagen GmbH) according to the manufacturer's instructions. Preparation of chromosomal DNA from S. pyogenes was carried out according to standard protocols (Caparon and Scott, 1991). PCR reactions were performed using Pwo polymerase (Roche Diagnostics GmbH). The primers used in this study are listed in Table 3. Sequencing reactions were performed by VBC-Genomics Bioscience Research GmbH.
|150b||F||CTAAAGATACTGATCAAG||Upstream region of pel/sagA|
|138||F||ATTAGATAAGGAGGTAAACC||Internal fragment pel/sagA|
|165||F||CAAACCACTTGTCCTTATC||Internal fragment sagB|
|148||F||ATTATCGGATCCCTGCAGTAGTTTCACAGTTGGTTAAGGGAG||Internal fragment sagC|
|163||F||ATCAATCTCAAGTTCCAG||Internal fragment sagD|
|188||F||GCAACTCTGATTTATGAC||Internal fragment sagE|
|190||F||CTTATGACTCGATAGCAG||Internal fragment sagH|
|133||F||TATTCGCTTAGAAAATGAA||Internal fragment emm|
|248||F||AGGAATAAATTGGTCCTCTT||Internal fragment nga|
|246||F||ACTAGGAGCTACACAACCAG||Internal fragment sic|
|45||F||GATAACCATACGATTCAGCT||Internal fragment speB|
|57||F||TATGCATACACTATAACTCC||Internal fragment speC|
|242||F||CGGTAACTAACCAGAAAGGG||Internal fragment rrs|
|269||F||TCTATTATAGAATTCAGTATATTAG||Used for sagA-deficient mutant|
|145||F||AATGAAGGTACCACTCTACAACCTTGATTGATTCGC||Used for sagBC-deficient mutant|
|168||F||ATCAAAGGTACCGTCTAATTCACAGACACAAGGAGC||Used for sagDEF-deficient mutant|
|172||F||AATGAAGGTACCGCTTATATTATTAGGAGAACAGTC||Used for sagGHI-deficient mutant|
|273||F||CGTATGTAAGGCCTTCAGGGGG||Amplification of aphIII cassette for sagA-deficient mutant|
|57||F||GTTTAAGGATCCGGTTTCAAAATCGGCTCCGTC||Amplification of aphIII cassette for sagBC-, sagDEF- and sagGHI-deficient mutants|
|265||R||AATAATCTGCAGCTATTATAAAATTCAGTATAT||Used to construct the complementation plasmids|
|150a||F||CACCAGTGTGGATCCTCCTTGGTTTTAAGG||Used for chromosomal mutation of pel/sagA|
|177||R||ACAAGACAGCTTTATAGC||Used for RT-PCR|
|149||R||TCCAGGAGCAACTTGAGTTGTTTCAGCTACACTAGTAGC||Used for primer extension|
Transformation of bacteria
Chromosomal deletion of the sagA, sagBC, sagDEF and sagGHI genes
Replacement of the sagA gene with a kanamycin resistance cassette, aphIII (Trieu-Cuot and Courvalin, 1983) with transcriptional terminator, was performed in the M1 strain ATCC 700294, RDN29 (Suvorov and Ferretti, 1996; Ferretti et al., 2001), selecting for a double cross-over event. For this purpose, amplification of a 257 bp fragment upstream of sagA was performed using primer 269, containing an EcoRI site, and primer 270, containing a BamHI site. Amplification of a 315 bp fragment downstream of sagA was performed using primer 271, containing a StuI site, and primer 272, containing a XbaI site. The aphIII cassette was amplified using primer 273, containing a StuI site, and primer 274, containing a BamHI site. After digestion with the appropriate enzymes, the three fragments (sagAupstream-aphIII-sagAdownstream DNA) were ligated and cloned into pUC19 (Table 2). The resulting plasmid was purified, linearized with the restriction enzyme ScaI (which cuts within the ampicillin resistance cassette) and used to transform S. pyogenes as linear DNA. Kanamycin-resistant clones were then selected and further analysed. The same strategy was used to construct the sagBC-, sagDEF- and sagGHI-deficient strains, EC540, EC560 and EC566 respectively. For these mutants, the primers were designed to contain the restriction sites KpnI and BamHI (amplification of upstream fragments) and the restriction sites PstI and SphI (amplification of downstream fragments). The upstream fragments were amplified using RDN29 as template and 145/146 (sagBC-deficient strain), 168/169 (sagDEF-deficient strain) or 172/173 (sagGHI-deficient strain) as primer pairs. The downstream fragments were amplified using RDN29 as template and 148/149 (sagBC-deficient strain), 170/171 (sagDEF-deficient strain) or 174/175 (sagGHI-deficient strain) as primer pairs. In contrast to RDN17, the aphIII cassette was inserted into the chromosome in a way that the aphIII gene was transcribed in the same direction as the sagB-I operon. The aphIII cassette was amplified using primer 57, containing a BamHI site, and primer 58, containing a PstI site. In the sagBC-deficient strain, 656 bp of sagB (3′) and 143 bp of sagC (5′) were deleted. In the sagDEF-deficient strain, 466 bp of sagD (3′), the sagE coding sequence and 278 bp of sagF (5′) were deleted. In the sagGHI-deficient strain, 231 bp of sagG (3′), the coding sequence of sagH and 295 bp of sagI (5′) were deleted. To check for the correct deletion and insertion of the aphIII gene in the sagA-, sagBC-, sagDEF- and sagGHI-deficient strains, PCR analysis was performed. The replacement events were further verified by Southern blot analysis.
Mutation of sagA RBSI-start codon and putative RBSII
For the introduction of point mutations in streptococcal genomes, we constructed a new temperature-sensitive shuttle vector, pRDN18, based on PCR-generated interchangeable cassettes. Plasmid pRDN18 contains repAts (a streptococcal pWVO1-based thermosensitive origin of replication), aad9 (a spectinomycin resistance gene with its own promoter and transcriptional terminator), ColE1ori (a pUC19-based ColE1 origin of replication for E. coli) and an expanded MCS (Oka et al., 1979; Leenhouts et al., 1991; Maguin et al., 1992). A 635 bp fragment starting within the sagA/pel upstream region and ending 9 bp downstream of the sagA start codon was amplified using the primer pair 150a/219. Primer 150a introduces a BamHI restriction site while primer 219 introduces point mutations in the ribosome-binding site (RBSI) (GGAGGGTAAA) and in the start codon (ATGCTG). A 501 bp fragment starting with the RBS of the sagA gene and ending downstream of the putative transcriptional terminator was amplified using primer pair 218/220. 218 is complementary to 219 and introduces the same mutations in the RBS and the start codon; 220 introduces the EcoRI restriction site. Both PCR fragments were used in equimolar concentrations as template for a PCR with the primer pair 150a/220. After digestion with BamHI and EcoRI, the resulting fragment of 1085 bp was cloned into pRDN18 to create pRDN20. The construct was confirmed by sequencing. S. pyogenes RDN29 was transformed with pRDN20 by electroporation at permissive temperature of 28°C to create RDN163. Growing RDN163 at the restrictive temperature of 40°C in the presence of spectinomycin promoted integration of pRDN20 into the chromosome via a single cross-over event thus obtaining the SpecR and β-haemolysis-positive strain RDN164. Correct insertion of pRDN20 was checked by PCR analysis. Growing RDN164 at permissive temperature (28°C) in the absence of spectinomycin stimulated a second recombination event either upstream or downstream of the introduced mutations (Noirot et al., 1987; Petit et al., 1992). Scoring for β-haemolysis-deficient and SpecS isolates led to the population of mutants, which had lost the entire vector but carried the RBSI and start codon mutations on the chromosome. In those isolated clones, which lost SpecS but remained β-haemolytic, homologous recombination had occurred such that the wild-type genotype was regained. Using the same approach the putative second ribosome-binding site (RBSII) in the sagA sequence was mutated. Using the primer pairs 150a/261 and 260/220 two sagA/pel fragments were amplified and used as templates for a PCR with the primer pair 150a/220. Primers 260 and 261 are complementary and introduce two point mutations in the RBSII (GGAGGGTAAG). After digestion with BamHI and EcoRI, the resulting fragment of 1085 bp was cloned into pRDN18 to create pRDN32. RDN29 was transformed at permissive temperature with pRDN32 to obtain RDN289. The temperature shift to the restrictive temperature of 40°C led to integration of the plasmid into the chromosome leading to strain RDN420. Shifting the temperature back to 28°C led to excision of pRDN32 and mutation of RBSII in the chromosome of RDN29. For further evaluation, SpecS and, probably because of the loss of the SLS glycine–glycine cleavage site, β-haemolysis negative isolates were selected. Loss of the vector and chromosomal mutation of the sagA RBSI and start codon in strain RDN165 and mutation of the RBSII mutation in strain RDN423 was confirmed by PCR analysis and by sequencing of the pel locus. Furthermore, Southern blot analysis confirmed that in RDN165 and RDN423 recombination events did not affect the upstream region of pel and the downstream-located sag locus and that the vector was lost.
Construction of plasmids for complementation studies
A 751 bp large DNA fragment containing the entire pel sequence including the promoter region and the putative transcriptional terminator was amplified from genomic DNA of RDN29 with primers 265 and 266. Blunt ends were generated using the T4 DNA polymerase and the pel fragment was cloned into PstI-digested and -blunted pMSP3535 to create pRDN50 (Bryan et al., 2000). To create pRDN51 that contains pelRBSI*ATG*, the same strategy was used, using DNA of RDN165 as a template for the PCR reaction instead of DNA of the wild-type strain.
Southern blot analysis
Genomic DNA was prepared according to a previously described method (Caparon and Scott, 1991). For Southern blot analysis, 5 µg of genomic DNA was digested with HindIII and/or SpeI, separated by electrophoresis and further processed for blotting, prehybridization and hybridization as described (Charpentier et al., 2000). α-32P-dATP-labelled PCR-generated internal fragments of sagA, sagB, sagC, sagD, sagE, sagH, upstream region of pel/sagA, aphIII, repAts, aad9 and ColE1ori were used as specific DNA probes (Table 3). Labelling of the probes was performed using the random primed DNA labelling kit (Boehringer Mannheim).
Northern blot analysis
Total RNA was prepared according to a previously described protocol (Li et al., 1999). In brief, total RNA was prepared from samples equalized to the same number of cells. The amount of RNA in each sample was further normalized to that of the rRNA 16S, which was quantified by scanning an ethidium bromide-stained agarose gel with a video imager and by spectrophotometry measurements. Northern blot analysis was carried out as described elsewhere (Herbert et al., 2001). PCR-derived DNA-probes were amplified from strain RDN29 using primers described in Table 3 and labelled as described above. α-32P-dATP-labelled PCR product specific for 16S rRNA served as a loading control. In experiments of inhibition of mRNA synthesis with rifampicin, mRNA half-lives were determined by automated pixel counting.
Primer extension analysis and nucleotide sequence determination
Total RNA was isolated as described above and primer extension was performed according to standard protocols. Primer 149 was end-labelled with γ-32P-dATP according to standard methods (Sambrook et al., 1990) and used as primer for reverse transcription. A 630 bp large DNA fragment corresponding to the 5′ region of pel of the RDN29 genome was amplified with primers 149 and 155 (Table 3) and cloned into the SmaI and EcoRI sites pBCSK+ (Table 2) thus creating pRDN9. Sequencing was carried out with plasmid pRDN9 and primer 149 using the T7 DNA polymerase sequencing kit (USB). The pel reverse transcribed product and the sequencing reactions were analysed on a 6% PAA/8 M urea gel.
Reverse-transcriptase-PCR (RT-PCR) experiments
Total RNA prepared from mid-logarithmic phase culture of S. pyogenes strains was treated with RNase-free DNase I to remove contaminating genomic DNA. The DNA-free RNA was converted to cDNA using reverse transcriptase and a reverse primer specific to the sagC, sagE or sagI region (Sambrook et al., 1990). cDNA was then used as a template for PCR reactions. As positive control, a chromosomal DNA template was used. As negative control, the DNase I-treated RNA template was used.
Streptococcus pyogenes culture supernatants were collected, filter sterilized (0.45 µm filter), incubated on ice for 30 min with 1/10 volume 4% (w/v) DOC in TCA and centrifuged for 10 min at 14000 g at 4°C. The pellet was resuspended in 1 ml of ice-cold acetone and incubated on ice for 15 min. The acetone-extraction step was repeated two more times and finally the pellet was resuspended in 100 µl of 1 M Tris/HCl, pH 8.0.
In gel-tryptic digestion of proteins and MALDI-TOF mass spectrophotometry
In gel-tryptic digestion of proteins and MALDI-TOF mass spectrophotometry were performed as previously described (Hoffmann et al., 2001) at the Biologisch-Medizinisches Forschungszentrum, Heinrich-Heine-Universität, Düsseldorf, Germany.
Western blot analysis
Proteins were separated by sodium dodecyl sulphate 10% polyacrylamide gel electrophoresis (SDS–PAGE) and electroblotted onto a nitrocellulose membrane. This was followed by blocking the membrane with 2% bovine serum albumin in 0.1% Tween in TBS (TBST) for 1 h. The membrane was incubated overnight at 4°C with a SpeB-specific antiserum (against the zymogen and the mature form) diluted 1 : 5000 in TBST (Berge and Björck, 1995). A goat anti-rabbit horseradish peroxidase-conjugate was used as secondary antibody and the blot was developed using the ECL Chemiluminescence Kit, Amersham, Buckinghamshire, England.
Streptococcal cysteine protease assay
Cysteine protease activity associated with SpeB was quantified in culture supernatants based on an azocasein proteolysis assay developed earlier (Tsai et al., 1998). Briefly, bacterial supernatants were collected at different time points during growth and filtered (0.45 µm) to remove any cells. The reaction was initiated by addition of 200 µl of supernatant to 400 µl of prewarmed reaction mixture containing 2.7 mg of azocasein (Sigma) per ml in 50 mM Tris-HCl (pH 8.0). After incubation at 37°C for 20 min, the reaction was stopped by addition of 100 µl of 15% ice-cold trichloroacetic acid. The reaction mixture was kept on ice for 15 min, centrifuged and an equal volume of 0.5 M NaOH was added to the supernatant. The absorbance at 450 nm of the sample was measured with a microplate reader. The activity of medium was used to determine background values, which were insignificant under the conditions of the assay. To verify that the observed proteolytic activity was specifically the result of SpeB, samples were also incubated with 20 µM of the cysteine-protease inhibitor E-64 (Sigma). E-64 abolished the protease activity of the filtered supernatants by ≈ 90%.
Preparation of conditioned medium
Conditioned medium was prepared by removing cells by centrifugation for 30 min at 4°C at 10.000 r.p.m. Supernatants were used directly or stored at −20°C.
We thank Lars Björck, Lund University, Lund, Sweden, for his generous gift of streptococcal pyrogenic exotoxin B antiserum. We are grateful to Gary Dunny, University of Minnesota Medical School, Minneapolis, USA, for providing us the plasmid pMSP3535 plus sequence. This work was supported by a grant from the Austrian Science Fund Project P15041 (to R. Novak) and a grant from the Austrian National Bank No. 9601 (to R. Novak).
- 2000) Characterization of two novel regulatory genes affecting Salmonella invasion gene expression. Mol Microbiol 35: 635–646. , , , , and (
- 2000) Switching on and off with RNA. Proc Natl Acad Sci USA 97: 9824–9826. , and (
- 1997) A mRNA signal for the type III secretion of Yop proteins by Yersinia enterocolitica. Science 278: 1140–1143. , and (
- 2002) Genome sequence of a serotype M3 strain of group A Streptococcus: phage-encoded toxins, the high-virulence phenotype, and clone emergence. Proc Natl Acad Sci USA 99: 10078–10083. , , , , , , et al. (
- 1995) Streptococcal cysteine proteinase releases biologically active fragments of streptococcal surface proteins. J Biol Chem 270: 9862–9867. , and (
- 1998) Reduced virulence of group A streptococcal Tn916 mutants that do not produce streptolysin S. Infect Immun 66: 1671–1679. , , , , and (
- 2001) Generation and surface localization of intact M protein in Streptococcus pyogenes are dependent on sagA. Infect Immun 69: 7029–7038. , , , and (
- 2000) Improved vectors for nisin-controlled expression in gram-positive bacteria. Plasmid 44: 183–190. , , , and (
- 1987) Identification of a gene that regulates expression of M protein, the major virulence determinant of group A streptococci. Proc Natl Acad Sci USA 84: 8677–8681. , and (
- 1991) Genetic manipulation of pathogenic streptococci. Methods Enzymol 204: 556–586. , and (
- 1995) Increasing severity of invasive group A streptococcal disease in Australia: clinical and molecular epidemiological features and identification of a new virulent M-nontypeable clone. Clin Infect Dis 21: 1220–1227. , , , , and (
- 2000) Plakoglobin suppresses epithelial proliferation and hair growth in vivo. J Cell Biol 149: 503–520. , , , and (
- 2000) Role of group A streptococcal virulence factors in adherence to keratinocytes. Infect Immun 68: 1215–1221. , , , and (
- 2001) Mouse skin passage of a Streptococcus pyogenes Tn917 mutant of sagA/pel restores virulence, beta-hemolysis and sagA/pel expression without altering the position or sequence of the transposon. BMC Microbiol 1: 33. , , and (
- 2000) Group A streptococci in the 1990s. J Antimicrob Chemother 45(Suppl.): 3–12. (
- 1999) A response regulator that represses transcription of several virulence operons in the group A streptococcus. J Bacteriol 181: 3649–3657. , , and (
- 2001) Complete genome sequence of an M1 strain of Streptococcus pyogenes. Proc Natl Acad Sci USA 98: 4658–4663. , , , , , , et al. (
- 2003) Combined contributions of streptolysin O and streptolysin S to virulence of serotype M5 Streptococcus pyogenes strain Manfredo. Infect Immun 71: 3857–3865. , , and (
- 2003) SIC, a secreted protein of Streptococcus pyogenes that inactivates antibacterial peptides. J Biol Chem 278: 16561–16566. , , , , and (
- 2001) Small RNA regulators of translation: mechanisms of action and approaches for identifying new small RNAs. Cold Spring Harb Symp Quant Biol 66: 353–362. , , , , , and (
- 2002) Virulence control in group A Streptococcus by a two-component gene regulatory system: global expression profiling and in vivo infection modeling. Proc Natl Acad Sci USA 99: 13855–13860. , , , , , , et al. (
- 1999) A two-component regulatory system, CsrR-CsrS, represses expression of three Streptococcus pyogenes virulence factors, hyaluronic acid capsule, streptolysin S, and pyrogenic exotoxin B. Infect Immun 67: 5298–5305. , , , and (
- 2001) Subinhibitory clindamycin differentially inhibits transcription of exoprotein genes in Staphylococcus aureus. Infect Immun 69: 2996–3003. , , and (
- 2001) Neutral loss analysis in MALDI-MS using an ion-gate scanning mode. Anal Chem 73: 4845–4851. , , and (
- 1995) Cell density control of staphylococcal virulence mediated by an octapeptide pheromone. Proc Natl Acad Sci USA 92: 12055–12059. , , and (
- 1997) Bacterial interference caused by autoinducing peptide variants. Science 276: 2027–2030. , , and (
- 2003) RNA-mediated control of virulence gene expression in bacterial pathogens. Trends Microbiol 11: 280–285. , and (
- 2002) An RNA thermosensor controls expression of virulence genes in Listeria monocytogenes. Cell 110: 551–561. , , , , , and (
- 1999) SfbII protein, a fibronectin binding surface protein of group A streptococci, is a serum opacity factor with high serotype-specific apolipoproteinase activity. FEMS Microbiol Lett 178: 305–311. , , and (
- 2001) Group A streptococcal growth phase-associated virulence factor regulation by a novel operon (Fas) with homologies to two-component-type regulators requires a small RNA molecule. Mol Microbiol 39: 392–406. , , , , and (
- 2003) Virulence factor regulation and regulatory networks in Streptococcus pyogenes and their impact on pathogen–host interactions. Trends Microbiol 11: 224–232. , , and (
- 1991) Nucleotide sequence and characterization of the broad-host range lactococcal plasmid pWV01. Plasmid 26: 55–56. , , , , , and (
- 1998) Identification of csrR/csrS, a genetic locus that regulates hyaluronic acid capsule synthesis in group A Streptococcus. Mol Microbiol 30: 209–219. , and (
- 1999) Identification of pel, a Streptococcus pyogenes locus that affects both surface and secreted proteins. J Bacteriol 181: 6019–6027. , , , , and (
- 2003) Trigger factor-mediated prolyl isomerization influences maturation of the Streptococcus pyogenes cysteine protease. J Bacteriol 185: 3661–3667. , and (
- 2001) Molecular characterization of global regulatory RNA species that control pathogenicity factors in Erwinia amylovora and Erwinia herbicola pv. gypsophilae. J Bacteriol 183: 1870–1880. , , , , , and (
- 1997) Role of mga in growth phase regulation of virulence genes of the group A streptococcus. J Bacteriol 179: 5178–5187. , and (
- 2001) Cytolysin-mediated translocation (CMT): a functional equivalent of type III secretion in gram-positive bacteria. Cell 104: 143–152. , , and (
- 1992) New thermosensitive plasmid for gram-positive bacteria. J Bacteriol 174: 5633–5638. , , , , and (
- 1999) Structure-activity analysis of synthetic autoinducing thiolactone peptides from Staphylococcus aureus responsible for virulence. Proc Natl Acad Sci USA 96: 1218–1223. , , , , , , and (
- 1995) Activation of alpha-toxin translation in Staphylococcus aureus by the trans-encoded antisense RNA, RNAIII. EMBO J 14: 4569–4577. , , , and (
- 2000) Genetic locus for streptolysin S production by group A streptococcus. Infect Immun 68: 4245–4254. , , , , , , and (
- 1987) Plasmid replication stimulates DNA recombination in Bacillus subtilis. J Mol Biol 196: 39–48. , , and (
- 1993) Synthesis of staphylococcal virulence factors is controlled by a regulatory RNA molecule. EMBO J 12: 3967–3975. , , , , , and (
- 1979) Nucleotide sequence of small ColE1 derivatives structure of the regions essential for autonomous replication and colicin E1 immunity. Mol Gen Genet 172: 151–159. , , , , , and (
- 1988) Cloning, characterization, and sequencing of an accessory gene regulator (agr) in Staphylococcus aureus. J Bacteriol 170: 4365–4372. , , , , and (
- 1992) Induction of DNA amplification in the Bacillus subtilis chromosome. EMBO J 11: 1317–1326. , , , , and (
- 1995) The group A streptococcal virR49 gene controls expression of four structural vir regulon genes. Infect Immun 63: 9–20. , , and (
- 2000) SclA, a novel collagen-like surface protein of Streptococcus pyogenes. Infect Immun 68: 6370–6377. , , and (
- 1990) Molecular Cloning. A Laboratory Manual, 2nd edition. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory. , , and (
- 2002) Clostridial VirR/VirS regulon involves a regulatory RNA molecule for expression of toxins. Mol Microbiol 43: 257–265. , , , , and (
- 2002) Genome sequence and comparative microarray analysis of serotype M18 group A Streptococcus strains associated with acute rheumatic fever outbreaks. Proc Natl Acad Sci USA 99: 4668–4673. , , , , , , et al. (
- 2002) Streptococcal toxic shock syndrome. Clin Microbiol Infect 8: 133–136. (
- 1996) Physical and genetic chromosomal map of an M type 1 strain of Streptococcus pyogenes. J Bacteriol 178: 5546–5549. , and (
- 2000) Co-operative binding of human fibronectin to Sfbl protein triggers streptococcal invasion into respiratory epithelial cells. Cell Microbiol 2: 521–535. , , , , , , et al. (
- 1983) Nucleotide sequence of the Streptococcus faecalis plasmid gene encoding the 3′5′′-aminoglycoside phosphotransferase type III. Gene 23: 331–341. , and (
- 1998) Effect of group A streptococcal cysteine protease on invasion of epithelial cells. Infect Immun 66: 1460–1466. , , , , , , et al. (
- 1980) Growth characteristics of group A streptococci in a new chemically defined medium. Infect Immun 27: 444–448. , and (
- 1999) Regulation of virulence factor expression in group A streptococcus. Trends Microbiol 7: 428–430. (