|Home | About | Journals | Submit | Contact Us | Français|
Streptococcus pyogenes (group A streptococcus [GAS]) is a versatile human pathogen, and emm1/sequence type 28 (ST28) is the most frequently isolated type from GAS infections. The emm1/ST28 strain is associated with necrotizing fasciitis and streptococcal toxic shock syndrome. Growth-phase regulation is one of the important regulatory mechanisms in GAS, which controls gene expression at restricted phases of growth. CovRS, a two-component regulatory system, is considered the regulator of streptococcal pyrogenic exotoxin B (SpeB) and is thought to be activated in the exponential phase of growth. In the present study, Northern hybridization analysis showed that 52% of the analyzed GAS strains expressed covR at the exponential phase, but 48% of the strains expressed covR at the early stationary phase of growth. Strains transcribing covR at the early stationary phase showed better growth and earlier SpeB expression than the other group of strains. Multilocus sequence typing and pulsed-field gel electrophoresis analysis showed only emm1/ST28 strains (which comprise a clonal cluster) were expressing covR at the early stationary phase of growth, indicating that emm1/ST28 strains have special characteristics which may be related to their worldwide distribution.
Streptococcus pyogenes (group A streptococcus [GAS]) is an important human pathogen which causes a wide spectrum of diseases such as pharyngitis, cellulitis, necrotizing fasciitis, and streptococcal toxic shock syndrome (STSS). Although distinct serotypes of GAS have been isolated from invasive infections, M1T1 is the most persistent and frequently isolated serotype from invasive GAS infections worldwide (4, 16).
Currently, sequence-based emm typing is used to predict the Lancefield M serotypes and is one of most used typing methods with GAS. More than 177 emm types and 750 emm subtypes have been identified so far (36). The other two major typing methods for GAS are multilocus sequence typing (MLST) and pulsed-field gel electrophoresis (PFGE) typing. MLST is a method based on the nucleotide sequence of housekeeping genes and is primarily used for the identification of clusters among isolates (14). MLST sequence type 28 (ST28) is the most dominant type in the MLST GAS database (http://www.mlst.net/). Since horizontal gene transfer has frequently occurred within the GAS population, PFGE typing, which can probe the genomic DNA organization, is also an important tool for GAS typing (18).
Growth-phase regulation is one of the most important regulatory mechanisms in GAS (9, 44). Under the control of growth-phase regulation, any specific gene in GAS is transcribed during a restricted phase of growth. Since GAS has no functional alternative sigma factors, it is believed that the control of gene expression is dependent on the complex interactions among transcriptional regulators and two-component regulatory systems (9, 34). Transcriptional regulators Mga, Rgg, the RofA/Nra regulator family, and the two-component regulatory system CovRS are associated with growth-phase-dependent virulence gene regulation in GAS (26). CovRS-regulated genes are expressed in both the exponential phase and stationary phase (17). In addition, the CovRS system has important roles for bacterial growth during general stress and amino acid starvation conditions (13, 37). Graham et al. further suggested that CovRS has important roles in linking key biosynthesis, catabolic, and virulence functions during transcriptome restructuring (21).
Streptococcal pyrogenic exotoxin B (SpeB) is a cysteine protease of GAS which degrades human immunoglobulins, complement components, and extracellular matrix proteins (15, 23, 27, 42). In addition, SpeB also degrades several bacterial proteins, including surface M proteins, protein F1 and C5a peptidase, and secreted proteins such as DNase and EndoS (2, 6, 33, 45). It is thought that SpeB enhances host tissue destruction and bacterial dissemination during GAS infection (6, 7, 23).
In the present study, we found that different GAS strains have different covR expression patterns during bacterial growth. The difference in the covR expression pattern among GAS strains is associated with different bacterial growth and SpeB expression patterns. In addition, covR expression at the early stationary phase is highly associated with emm1/ST28 strains. Our results show that epidemiologically important emm1/ST28 strains present unique covR and speB expression patterns and exhibit faster growth.
The GAS clinical isolates analyzed in this study were randomly selected from the collection consecutively archived between 1994 and 2003 at National Cheng Kung University Hospital, Tainan, Taiwan (Table (Table1).1). GAS strain NZ131 (M49) was described previously (43). SF370 (M1T1) is strain ATCC 700294. All GAS strains were grown in tryptic soy broth (Becton, Dickinson and Company, Sparks, MD) supplemented with 0.5% yeast extract (TSBY).
RNA extraction from GAS was performed by the method described previously (12). Briefly, the bacterial pellet was washed with ice-cold 0.2 M sodium acetate twice and resuspended in buffer (100 mM of Tris-HCl [pH 7.0], 1 mM EDTA, 25% [wt/vol] glucose) containing 40 μl of lysozyme (20 mg/ml) and 10 μl of mutanolysin (1 U/μl; Sigma Chemicals, St. Louis, MO) and then incubated at 37°C for 1 h. After incubation, bacterial pellets were collected and resuspended in 500 μl of acetic buffer (20 mM sodium acetate [pH 5.5], 1 mM EDTA, 0.5% [wt/vol] sodium dodecyl sulfate) with 500 μl of hot phenol. After vortexing, the mixtures were centrifuged at 14,000 × g for 20 min. The upper aqueous layer was collected, and RNAs were precipitated with isopropanol. RNAs were dissolved in diethyl pyrocarbonate-H2O and treated with DNase (Promega, Madison, WI) at 37°C for 2 h. After DNase treatment, RNAs were obtained by using phenol-chloroform extraction and isopropanol precipitation. Northern hybridization analysis was performed as described previously (12). The probes used for detecting covR, speB, sagA, emm, and scpA expression are described in Table Table2.2. After hybridization, the membranes were washed and the hybridized RNA transcripts were detected by a LAS3000 Intelligent Dark Box (Fujifilm, Japan). All Northern hybridization analyses were performed at least twice independently.
Three hundred microliters of bacterial culture supernatant was mixed with 1.2 ml of acetone and incubated at −20°C for 1 h to precipitate the total protein. The precipitated protein was collected by centrifugation at 14,000 × g for 20 min at 4°C. The protein pellet was dissolved in 50 mM NaOH, and the protein concentration was determined by the Bio-Rad protein assay (Bio-Rad Laboratories, Hercules, CA). Two micrograms of total protein was subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis followed by transfer to polyvinylidene difluoride membranes. The membrane was blocked with phosphate-buffered saline (PBS) containing 5% (wt/vol) skim milk at 37°C for 2 h and probed with anti-C192S mouse serum prepared as previously described (28). After three washes with PBS containing 0.05% Tween 20 (PBST), the membrane was incubated with horseradish peroxidase-conjugated anti-mouse immunoglobulin antibody (Chemicon International, Inc., Temecula, CA). After incubation, the membrane was washed with PBST solution three times and then was treated with chemiluminescent substrate (Pierce ECL Western blotting substrate; Thermo Fisher Scientific, Inc., Rockford, IL) to visualize the signal. The signal was detected with the LAS 3000 Intelligent Box (Fujifilm, Japan). All Western hybridization analyses were performed at least twice independently.
To obtain bacterial growth curves, GAS strains were cultured in TSBY broth at 37°C for 16 h. One milliliter of overnight culture was transferred to 100 ml of fresh TSBY and incubated at 37°C without agitation. The optical density of the bacterial culture at 600 nm (OD600) was measured by UV/visible spectrophotometer (GE Healthcare, United Kingdom) after 3 to 17 h of incubation.
The bacterial growth curve was fitted by using a modified logistic model to estimate the growth parameters asymptotes (A), maximum specific growth rate (μm), and lag time (λ), as described by Zwietering et al. (47). A is defined as the maximal value of bacterial growth, μm is defined as the tangent in the inflection point of the bacterial growth curve, and λ is defined as the x-axis intercept of this tangent. Experiments in duplicate were performed, and the averages were used for estimating growth parameters.
emm typing and PFGE typing were performed as previous described (46). The PFGE fragment patterns were interpreted in accordance with the criteria of Tenover et al. (41). PFGE fragment patterns were compared with the use of GelCompar II software (Unimed Healthcare, Inc., Houston, TX), by using the unweighted pair group method using average linkages (UPGMA) based on the Dice coefficient with a position tolerance of 1.5%. PFGE-based clusters were defined as isolates with a genetic relatedness of >80% on a dendrogram.
GAS genomic DNA was prepared as previously described (12). Seven housekeeping genes of GAS (gki, gtr, murI, mutS, recP, xpt, and yiqL) were amplified and sequenced using the primers and conditions described in the MLST database (www.mlst.net). New housekeeping allele sequences and STs generated from this study were submitted to the MLST database. To generate dendrograms, sequences of the seven housekeeping genes used in the MLST scheme were joined into a 3,134-bp concatenated sequence and a dendrogram was constructed from concatenated sequences by the UPGMA method and the Kimura two-parameter nucleotide substitution model with 1,000 bootstrap trials (25). A clonal cluster was defined by bootstrap value of >95 and further verified by the eBURST program (eburst.mlst.net).
Statistical analysis was performed by using Student's t test. A P value of <0.05 was taken as significant.
To investigate the covR expression pattern in SF370 (emm1) and NZ131 (emm49) during bacterial growth, the RNAs were extracted from SF370 and NZ131 at different phases of growth and covR expression was analyzed by Northern hybridization with an alkaline phosphatase-labeled covR probe. covR and covS consist of an operon, the covR- and covS-cotranscribed RNA (covRS), and covR monocistronic transcripts (covR) were detected by a covR probe (Fig. (Fig.1A)1A) (12). The results showed that SF370 had maximal covR RNA expression at the early stationary phase (after 5 h of incubation), whereas NZ131 had its maximal covR expression in the exponential phase of growth (Fig. 1A and B). No difference was found in covRS-cotranscribed RNA between SF370 and NZ131 (Fig. (Fig.1A1A).
Nineteen clinical strains collected from 1994 to 2003 were analyzed for their covR expression pattern by Northern hybridization with an alkaline phosphatase-labeled covR probe. Bacteria were cultured in TSBY broth, and covR expression was analyzed after 3, 5, and 7 h of incubation. We found nine strains expressed covR RNA maximally at the early stationary phase (covR-ST), and 10 strains expressed covR RNA at the exponential phase (covR-E). Among the covR-E strains, six different emm types were found (two of emm4, four of emm12, one of emm22, one of emm73, one of emm89, and one of emm95). All of the emm1 strains were the covR-ST phenotype. In addition, one emm75 strain and one emm89 strain were the covR-ST type (Table (Table11).
SF370 and NZ131 expressed covR RNA at different phases of growth (Fig. (Fig.1A).1A). In addition, the growth curves of SF370 and NZ131 had different characteristics (Fig. (Fig.1B).1B). Apparently, SF370 had the better growth activity and spent less time reaching the stationary phase than did NZ131 (Fig. (Fig.1B).1B). To further demonstrate that the covR expression pattern was associated with bacterial growth activity, we analyzed the growth curves of 21 strains (SF370, NZ131, and 19 clinical isolates) (Table (Table1).1). Bacteria were cultured in TSBY broth at 37°C, and the bacterial density was determined by spectrophotometer at 600 nm after 3, 5, 7, and 9 h of incubation. The average of the bacterial density of covR-E or covR-ST strains at each time point was calculated and plotted on a chart to represent the growth curve of covR-E and covR-ST strains (Fig. (Fig.2A).2A). This chart showed that the covR-ST strains had faster growth and reached the stationary phase faster than did the covR-E strains (Fig. (Fig.2A2A).
To demonstrate the growth curve characteristics mathematically, we used the mathematic model to analyze the bacterial growth kinetics of our clinical isolates. The results showed the lag time (λ of 4.096 for covR-E strains and 4.234 for covR-ST strains) and the maximal value of bacterial growth (A of 1.977 for covR-E strains and 1.947 for covR-ST strains) were no different between covR-E and covR-ST strains (P > 0.5). However, the maximum specific growth rate (μm) of the covR-ST strains was significantly higher than that of the covR-E strains (Fig. (Fig.2B)2B) (P < 0.05).
Gene expression in GAS is controlled by growth-phase regulation (9, 43). The transcription of speB and sagA is restricted to the stationary phase, and expression of emm and scpA is restricted to the exponential phase. We hypothesized that the covR-ST strains, which had better growth activity, would express speB and sagA genes earlier than did the covR-E strains. To verify this hypothesis, SF370 and A-20 (covR-ST strains) and NZ131 and GAS516 (covR-E strains) were cultured for 6 to 10 h in TSBY broth and speB and sagA transcripts were analyzed by Northern hybridization with the alkaline phosphatase-labeled speB and sagA probes, respectively. The results showed that speB expression in SF370 was detected after 6 h of incubation, whereas speB expression in NZ131 was only detected after 10 h of incubation (Fig. (Fig.3A).3A). Similarly, A-20 speB expression was detected after 6 h of incubation, and GAS516 speB expression was detected after 8 h of incubation (Fig. (Fig.3B).3B). However, the sagA expression pattern in these strains was similar. Expression was detected after 6 h of incubation and increased after 8 and 10 h of incubation in SF370, NZ131, and GAS516 (Fig. (Fig.3C3C).
To further clarify the speB expression pattern in SF370 and NZ131, SpeB protein secretion was detected by Western hybridization. SF370 and NZ131 were cultured in TSBY broth for 3 to 20 h, and SpeB protein in culture supernatant was analyzed by Western hybridization with anti-C192S mouse serum (C192S is a protease-negative recombinant SpeB protein). The results showed that the zymogen form of SpeB protein (42 kDa) was detected in SF370 culture supernatant after 8 h of incubation, whereas it was detected after 12 h of incubation in NZ131 (Fig. (Fig.3D).3D). In addition, after 12 h of incubation, SF370 accumulated more SpeB protein in culture supernatant than did NZ131 (Fig. (Fig.3D3D).
emm and scpA expression was restricted to the exponential phase of growth. We also hypothesized that covR-ST strains expressed these genes in a more restricted pattern during bacterial growth than did covR-E strains. The SF370 and A-20 strains (covR-ST) and NZ131 (covR-E) were cultured for 3 to 9 h, and emm and scpA transcripts were analyzed by Northern hybridization with alkaline phosphatase-labeled emm1 and scpA probes, respectively. The results showed that SF370, A-20, and NZ131 maximally expressed emm and scpA genes at 5 h of incubation and decreased expression after 7 h of incubation (data not shown).
We found that all emm1 strains expressed covR at the early stationary phase of growth. Therefore, we hypothesized that the strains expressing covR at the early stationary phase were clonally related. To verify this hypothesis, MLST analysis was used to analyze all 21 strains. The allelic profile, emm type, and epidemiologic information of strains in this study were posted on the MLST database (www.mlst.net). Four new MLST STs (465, 466, 467, and 468) were found in the present study. Thirteen STs were found in 21 GAS strains. Among them, all emm1 strains were ST28 (Table (Table1).1). Strains 663 and 276, which expressed covR at the early stationary phase, were ST468 and ST384, respectively.
The MLST-based dendrogram analysis showed four major clonal clusters (Fig. (Fig.4).4). Eight emm1 strains with identical STs (ST28) were defined as cluster 1 (Fig. (Fig.4).4). Cluster 2 and cluster 3 were composed of four emm12 strains with three different STs (36, 465, and 467) and two emm4 strains with two different STs (391 and 39), respectively (Fig. (Fig.4).4). L257 (emm95/ST400) was cluster 4. Furthermore, strains that shared more than five identical housekeeping loci were classified into one clonal cluster by eBURST analysis. eBURST analysis showed similar classification results except for cluster 4 (L257; data not shown). Six strains (276, 625, 516, 663, L259, and NZ131) were singletons (Fig. (Fig.4).4). Other than strains 276 and 663, clusters 2, 3, and 4 and other singleton strains all expressed covR at the exponential phase of growth (Fig. (Fig.44).
A PFGE-based dendrogram showed four major clusters (Fig. (Fig.5).5). The largest cluster (cluster P2) comprised 6 emm1/ST28 strains. Another emm1/ST28 strain, L395, has >75% genetic relatedness to cluster P2 strains (Fig. (Fig.5).5). Strains SF370 (emm1/ST28), 276 (emm89/ST384), and 663 (emm75/ST468), expressed covR at the early stationary phase and were not related to any specific clusters. Clusters P1 and P4 comprised two emm4 and three emm12 strains, respectively (Fig. (Fig.5).5). Cluster P3 comprised emm73 and emm95 strains with different STs. Except for cluster P2 strains, which expressed covR at the early stationary phase, cluster P1, P3, and P4 strains all expressed covR at the exponential phase of growth (Table (Table11).
Although emm1/ST28 is the predominant type isolated from GAS infections in many countries and associated with severe diseases such as necrotizing fasciitis and STSS, the underlying mechanism for these important epidemiological observations is still not clear. In the present study, we show two different covR expression patterns among GAS strains during their growth. Strains with covR expression at the early stationary phase are highly correlated with the emm1/ST28 type and associated with faster growth than the strains with covR expression at the exponential phase. In addition, we found that the different growth activity affects the speB expression pattern, but not sagA, emm, and scpA expression. To our knowledge, this is the first demonstration that covR is expressed at different growth phases among the GAS population and that the different expression is significantly associated with different growth rates and speB expression patterns.
CovRS is a global regulatory system and regulates about 15% of gene expression in GAS. Previously, we found that the covRS operon transcribes three different RNA transcripts, which are covRS cotranscribed RNA and two different sizes of covR monocistronic transcripts (12). CovRS was thought to be expressed at the exponential phase (20). In this study, we found two covR expression patterns among GAS and showed that the different covR expression patterns are significantly associated with bacterial growth rates (Fig. (Fig.2).2). The different growth kinetics and covR expression patterns between covR-E and covR-ST strains suggest that the virulence gene expression patterns may be different between these two groups of strains. However, the results showed that emm, scpA, and sagA expression patterns in SF370, A-20, NZ131, and GAS516 were similar, whereas the speB expression pattern was different between covR-ST (SF370 and A-20) and covR-E (NZ131 and GAS516) strains (Fig. (Fig.3).3). One possibility to explain the different SpeB expression patterns in covR-ST and covR-E strains is that the bacterial-density-related factors are required for triggering SpeB expression. In culture broth, bacteria consume the carbohydrates and release metabolic products, causing a decrease of the broth pH. The acidic pH of the broth and carbohydrate starvation are signals for triggering SpeB expression (10, 29). Therefore, these positive signals might induce SpeB expression at an early culture stage in the covR-ST strains. Although the interaction between CovR and SpeB is controversial, CovR is considered the regulator to control SpeB expression (17, 21, 38). We have constructed the covR isogenic mutation in SF370, A-20, NZ131, and GAS516 strains and analyzed speB and sagA expression pattern. In SF370, A-20, and GAS516 covR mutants, speB and sagA expression were increased (data not shown). However, we found in NZ131 (covR-E strain) covR isogenic mutant, speB expression was similar when compared with that of the wild-type strain, but sagA expression was increased in the mutant strain (data not shown). These results suggest that the different speB expression patterns in these strains are not directly responsible for different covR expression patterns.
SpeB is a cysteine protease which degrades several host and bacterial proteins (35). Recently, SpeB was shown to degrade GAS-secreted DNase to decrease the neutrophil resistance ability of GAS (45). These reports indicate that SpeB has broad biological effects during GAS infection. Therefore, the timing of SpeB expression during infection would be critical for GAS to establish a successful infection. In this study, we showed that the SpeB expression pattern is different in covR-ST and covR-E strains. SF370 expresses SpeB earlier than NZ131 and also accumulates more SpeB protein in culture supernatant during the same incubation period (Fig. 3A and D). Different SpeB expression patterns could cause different levels of bacterial cell surface protein integrity and might also contribute to different degrees of tissue damage.
emm1 GAS is the predominant clone isolated from GAS infections in many countries and is associated with severe infections such as necrotizing fasciitis and STSS (4, 16). In general, emm1 strains have several superantigen genes, such as speA, speF, speG, and smeZ (8, 24). Furthermore, the presence of the prophage-associated speA gene is statistically associated with strains recovered from patients with severe or recurrent GAS infections (5, 32, 40). In addition, the chromosomal sic gene (coding for streptococcal inhibitor of complement) is also considered to be present in all emm1 but not other emm type strains (1, 22). In the present study, we found covR expression at the early stationary phase of growth is highly associated with emm1/ST28 strains (Table (Table1).1). These results suggest that, along with the presence of the speA and sic genes, covR expression at the early stationary phase of growth is one of the specific characteristics of emm1/ST28 strains.
MLST analysis showed all emm1 strains comprised a clonal cluster with identical STs (Fig. (Fig.4).4). PFGE analysis further showed that, except for SF370, all emm1 strains have similar PFGE band patterns (Fig. (Fig.5).5). SF370 has no speA gene, which is carried on a prophage (3). The lack of the prophage integration in the SF370 genome could be the reason for the completely different PFGE band pattern compared with other emm1/ST28 strains. emm1 isolates were randomly chosen from our collections from 1994 to 2002, suggesting that these isolates were not collected from a single clonal outbreak (Table (Table1).1). These results indicate that the emm1/ST28 strains should be highly homogenous, which supports the observations of previous studies (8, 31, 39). Importantly, emm1/ST28/PFGE type A is also the dominant type found in Canada, Portugal, and northern Taiwan, reinforcing its epidemiologic significance on a global scale (8, 11, 19). However, due to the limited geographical origin of these strains, the phenomenon of our finding requires further study on strains from different parts of the world.
In summary, we found that two different covR expression patterns were distributed among GAS strains. Strains with covR expression at the early stationary phase are highly correlated with the emm1/ST28 type and exhibited faster bacterial growth and earlier SpeB expression patterns than the other group of strains. Since emm1/ST28 is a predominant type of GAS isolates, we suggest that the specific covR expression pattern, faster growth, and earlier SpeB expression in emm1/ST28 strains could be a potential key phenotype found in global distribution of emm1 S. pyogenes.
We are very grateful to Robert M. Jonas for helpful comments on the manuscript.
This work was supported in part by grant NSC96-2320-B-006-008 from the National Science Council and by grant NHRI-EX96-9429SP from the National Health Research Institutes, Taiwan. We acknowledge the use of the Streptococcus pyogenes MLST database which is located at the Imperial College London and is funded by the Wellcome Trust.
Published ahead of print on 26 August 2009.