• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of iaiPermissionsJournals.ASM.orgJournalIAI ArticleJournal InfoAuthorsReviewers
Infect Immun. Jan 2003; 71(1): 446–455.
PMCID: PMC143243

Cytotoxic Effects of Streptolysin O and Streptolysin S Enhance the Virulence of Poorly Encapsulated Group A Streptococci

Abstract

Although the toxicity of streptolysin O (SLO) and streptolysin S (SLS) in purified group A streptococci (GAS) has been established, the effect of these molecules in natural infection is not well understood. To identify whether biologically relevant concentrations of SLO and SLS were cytotoxic to epithelial and phagocytic cells that the bacteria would typically encounter during human infection and to characterize the influence of cell injury on bacterial pathogenesis, we derived GAS strains deficient in SLO or SLS in the background of an invasive GAS M3 isolate and determined their virulence in in vitro and in vivo models of human disease. Whereas bacterial production of SLO resulted in lysis of both human keratinocytes and polymorphonuclear leukocytes, GAS expression of SLS was associated only with keratinocyte injury. Expression of SLO but not SLS impaired polymorphonuclear leukocyte killing of GAS in vitro, but this effect could only be demonstrated in the background of acapsular organisms. In mouse invasive soft-tissue infection, neither SLO or SLS expression significantly influenced mouse survival. By contrast, in a mouse model of bacterial sepsis after intraperitoneal inoculation of GAS, SLO expression enhanced the virulence of both encapsulated and acapsular GAS, whereas SLS expression increased the virulence only of acapsular GAS. We conclude that the cytotoxic effects of SLO protect GAS from phagocytic killing and enhance bacterial virulence, particularly of strains that may be relatively deficient in hyaluronic acid capsule. Compared to SLO, SLS in this strain background has a more modest influence on GAS pathogenicity and the effect does not appear to involve bacterial resistance to phagocytosis.

Group A streptococci (GAS) are important human pathogens. Millions of children each year develop GAS pharyngitis. Although throat infection is typically benign, it can be complicated by the development of acute and chronic rheumatic heart disease. Invasive GAS infection occurs less frequently than pharyngitis but causes significantly more morbidity. The incidence of severe invasive GAS disease associated with shock and organ failure appears to be increasing (14, 38, 40).

GAS produce two hemolysins that may contribute to pathogenesis. Streptolysin O (SLO) is a well-characterized oxygen-labile prototype of a cholesterol-binding bacterial exotoxin. When cultured in broth, GAS express SLO during exponential-phase and early stationary-phase growth (1). Streptolysin S (SLS) is an oxygen-stable oligopeptide primarily responsible for the characteristic zone of beta-hemolysis surrounding GAS colonies grown on blood-agar medium (17). SLS production occurs when cells are in stationary-phase growth conditions in broth culture (9).

Early studies of purified SLO and SLS demonstrated that these hemolysins were toxic to a variety of human cells in vitro and in vivo (3, 10, 16, 19, 20). In rabbits, injection of SLO caused blood vessel contraction, increased capillary permeability, massive intravascular thrombosis, dermal necrosis, cardiotoxicity, and death (1). Intravenous injection of SLS in rabbits resulted in massive intravascular hemolysis and death (15). The severity of injury in these studies suggests that the injected concentrations of SLO and SLS may have exceeded those achieved during natural infection.

More recently, investigators have attempted to assess the effects of biologically relevant concentrations of SLO and SLS in GAS pathogenesis by determining the virulence of hemolysin-deficient mutants in animal models of human infection. Limbago and colleagues reported that SLO-deficient GAS injected subcutaneously in mice were attenuated (24). Similarly, Betschel et al. noted that SLS-deficient GAS subcutaneously injected in mice were less virulent than the wild-type parent strain (11).

To characterize further the pathogenic effects associated with bacterial expression of SLO and SLS, we derived GAS deficient in the production of SLO or SLS in the background of an invasive GAS M3 isolate and used the strains in models of human infection. Our results demonstrate that GAS expression of SLO and, to a lesser extent, of SLS is associated with human keratinocyte and polymorphonuclear leukocyte (PMN) injury in vitro and, particularly in the background of acapsular GAS strains, enhances bacterial virulence in murine models of invasive infection.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

GAS strain 950771 is a moderately encapsulated M-type 3 invasive isolate from a child with necrotizing fasciitis and sepsis. Strain 950771Sm is a spontaneous streptomycin-resistant variant of 950771. GAS strain 950771hasA is an acapsular derivative of 950771 (6). GAS strain 950771SLO is an SLO-deficient derivative of 950771Sm (13). GAS strains were grown in Todd-Hewitt broth (Difco Laboratories, Detroit, Mich.) supplemented with yeast extract (Difco) to a final concentration of 2.5% (wt/vol) (THY broth) or, for SLS-hemolysis assays, in brain-heart infusion (BHI) broth (Difco) supplemented with 2% (wt/vol) sodium bicarbonate and 1% (wt/vol) maltose. GAS were grown on either Trypticase-soy blood agar medium (Becton Dickinson Microbiological Systems, Franklin Lakes, N.J.) (TSA-blood) or THY agar supplemented with 5% (vol/vol) defibrinated sheep blood (PML Microbiologicals, Inc., Wilsonville, Oreg.) (THY-blood). When required, THY-blood medium was supplemented with 1.0 μg of erythromycin per ml. GAS were cultured at 37°C in the presence of 5% CO2.

DNA manipulation.

Restriction endonuclease digests, DNA ligations, transformation of Escherichia coli, agarose gel electrophoresis, and Southern hybridization analysis were performed according to standard techniques (8). Restriction enzymes and DNA ligase were purchased commercially (Invitrogen, Carlsbad, Calif.). Plasmid DNA was recovered from E. coli strain DH5α (New England Biolabs, Beverly, Mass.) and purified with Qiagen mini- or maxipreps according to the manufacturer's protocol (Qiagen, Inc., Valencia, Calif.). GAS chromosomal DNA was purified by the method of O'Connor and Cleary (31). DNA probes for Southern hybridization analysis were conjugated to horseradish peroxidase, and hybridization was detected with a chemiluminescent substrate according to the manufacturer's instructions (ECL kit; Amersham Biosciences, Inc., Piscataway, N.J.).

Derivation of SLS- and SLO-deficient GAS. (i) 950771SLO, 950771hasASLO, and 950771hasASLORV.

The SLO-deficient GAS strain 950771SLO and the acapsular GAS strain 950771hasA were derived previously (6, 13). Plasmid pJSLOΔ, containing an internally deleted slo allele (6), was used to derive the acapsular SLO-deficient mutant 950771hasASLO by allelic exchange as described previously (6). In addition, strain 950771SLORV, in which the plasmid initially integrated into slo and was then excised from the bacterial chromosome to generate the wild-type slo genotype, was recovered as a control for passage effects on bacterial virulence. Southern hybridization analysis of restriction endonuclease-digested chromosomal DNA confirmed that there was a 1-kb deletion in the mutant slo allele and that the revertant slo gene was similar in size to the wild-type allele.

(ii) 950771SLS, 950771hasASLS, and 950771hasASLSRV.

The GAS sagA gene encodes SLS (30). The oligonucleotide primer pair 5′-CGCGGATCCCACATAGTTATTGATAGAAT (forward) and 5′-TCCAGGAGCAACTTGAGTTG (reverse) and the primer pair 5′-CAACTCAAGTTGCTCCTGGACAAGGTGGTAGCGGAAGTTA (forward) and 5′-GCGAAGCTTGTAATCCGATAAGGACAAGT (reverse) were used to PCR amplify a 359-bp segment and a 392-bp segment from the 5′ and 3′ termini of the GAS sagA sequence, respectively. The reverse primer for the first segment and the forward primer for the second segment had complementary sequences to generate DNA fragments with overlapping ends. These fragments were used as the DNA template in a subsequent overlap PCR in which the first and fourth primers listed above were used to amplify a final product composed of the sagA sequence with an internal 60-bp deletion. Unique BamHI and HindIII restriction endonuclease sites included in the PCR primers (underlined) were used to clone the amplification product into pJRS233 (33) to form plasmid pJsagAΔ. Determination of insert sequence in plasmid pJsagAΔ confirmed the absence of 60 bp.

Plasmid pJsagAΔ was purified from E. coli strain DH5α and used to derive the SLS-deficient mutant strains 950771SLS and 950771hasASLS by allelic exchange mutagenesis as described previously (6). In addition, strain 950771hasASLSRV, in which the plasmid initially integrated into sagA and was then excised from the 950771hasA chromosome to generate the wild-type sagA genotype, was recovered as a control for passage effects on bacterial virulence. Southern hybridization analysis of restriction endonuclease-digested chromosomal DNA confirmed a small deletion in the mutant SLS allele and that the revertant sagA gene was similar in size to the wild-type allele.

Hemolysis assays. (i) SLO-hemolysis assay.

Serial dilutions of GAS culture supernatants recovered during the transition from exponential-phase to stationary-phase growth were analyzed for the ability to lyse a standardized preparation of sheep erythrocytes as previously described (13). To confirm that hemolysis was due to SLO, control reactions included culture supernatants to which water-soluble cholesterol (cholesterol/methyl-β-cyclodextrin; Sigma-Aldrich, St. Louis, Mo.), a specific SLO inhibitor, had been added to yield a final concentration of 250 μg/ml (estimated cholesterol, 17 μg/ml).

(ii) SLS-hemolysis assay.

SLS-hemolysis assays were performed essentially as described by Loridan and Alouf (25). GAS were grown in BHI broth supplemented with 2% (wt/vol) Na2CO3 and 1% (wt/vol) maltose for 16 h. Cells were washed in 100 mM K2HPO4, pH 7.0 (potassium phosphate buffer). SLS expression into the supernatant was induced in 100 mM potassium phosphate buffer supplemented with 2 mM MgSO4, 30 mM maltose, and RNA core (Sigma-Aldrich) at 0.5 mg/ml. To stabilize SLS, ammonium acetate (100 mM) was immediately added to the sterilized supernatants. A standardized suspension of sheep erythrocytes was prepared by diluting 6 ml of fresh defibrinated sheep blood (PML Microbiologicals, Inc.) in 100 ml of phosphate-buffered saline (PBS), pH 6.5, supplemented with 0.1% (wt/vol) bovine serum albumin (Sigma). Additional PBS-bovine serum albumin buffer was added as required so that the lysis of a 500-μl aliquot of the erythrocyte suspension by its dilution in 14.5 ml of 0.1% (wt/vol) Na2CO3 resulted in an A541 of 0.2.

A 1.0-ml aliquot of serially diluted culture supernatant in THY broth was incubated with 500 μl of the standardized erythrocyte suspension at 37°C for 45 min. Intact erythrocytes were removed by centrifugation, and the A541 of the supernatant was determined. Absorbance values were compared with a standard curve of erythrocytes lysed in water. To confirm that hemolysis was not due to SLO, control reactions included culture supernatants to which water-soluble cholesterol (cholesterol/methyl-β-cyclodextrin; Sigma-Aldrich) had been added to a final concentration of 250 μg/ml (estimated cholesterol, 17 μg/ml).

Isolation of human PMNs.

Human PMNs were isolated from freshly drawn heparinized venous blood by gradient density centrifugation with Histopaque-1077 and -1119 according to the manufacturer's instructions (Sigma-Aldrich). Erythrocytes contaminating the neutrophil preparation were lysed by adding hypotonic saline. The final preparation contained 95% polymorphonuclear leukocytes, based on analysis of morphology after Wright staining. After staining with trypan blue, 95% of the cells were viable, as determined by dye exclusion. In all assays including leukocytes, the final concentration was 106 cells per ml.

Keratinocyte cytotoxicity assays.

Human keratinocytes, derived from a well-differentiated squamous cell carcinoma (SCC-13 cells) (35), were seeded into 24-well culture plates (Corning Incorporated Life Sciences, Acton, Mass.) at ≈104 cells per well in serum-free, antibiotic-free keratinocyte medium (K-SFM; Invitrogen) supplemented with bovine pituitary extract (50 μg/ml), epidermal growth factor (0.1 ng/ml), and 0.3 mM CaCl2 and grown to confluence (≈6 × 104 cells per well) at 37°C in the presence of 5% CO2 and 100% relative humidity. The monolayers were washed with warmed K-SFM to remove unattached cells, then inoculated with 3 × 106 CFU of GAS at a multiplicity of infection of 50:1, and suspended in 0.1 ml of K-SFM without supplements. Additional K-SFM was added to a final volume of 0.5 ml. The cells were incubated at 37°C in the presence of 5% CO2. At 3 h, the cell supernatants were removed, and fresh medium was applied to each well. After a 24-h incubation, the cells were examined microscopically before and after staining with 0.1% (wt/vol) trypan blue (Sigma). The number of keratinocytes per field before trypan blue staining and the number of cells incorporating dye after staining were determined microscopically and compared to these numbers of cells in uninfected wells.

Human PMN cytotoxicity assays.

To determine whether either SLO or SLS is cytotoxic to human neutrophils, we incubated the wild-type, SLS-deficient, and SLO-deficient GAS strains at a multiplicity of infection of 5:1 with freshly isolated human neutrophils in the presence of 10% serum as a complement source. After a 1-h end-over-end rotation at 37°C, the samples were fixed with 2% (vol/vol) paraformaldehyde in PBS overnight at 4°C. Flow cytometry was performed on a MoFlo flow cytometer (Cytomation, Fort Collins, Colo.), analyzing 100,000 events for each sample. The proportion of events detected in the intact cell and debris window were determined with Summit software version 7.19.

To demonstrate that the reduction in numbers of intact PMNs in infected samples is associated with a shift of PMNs into cell debris due to cell lysis, we labeled the samples with a fluorescent monoclonal antibody to CD66b, a molecule originally described as a granulocyte-specific activation antigen. Uninfected and infected PMNs were incubated and fixed as described previously above. PMNs were incubated for 20 min at 4°C with a fluorescein isothiocyanate-conjugated monoclonal mouse anti-human CD66b antibody (Immunotech). Cells were washed three times with PBS, resuspended in 0.3 ml of PBS, and applied to the flow cytometer, analyzing 10,000 events for each sample. The proportion of events detected in the intact cell and debris windows was determined with Summit software version 7.19

Phagocytic assays.

The direct bactericidal test of Lancefield (21) was used to determine the ability of GAS strains to resist phagocytic killing in whole blood. In addition, opsonophagocytic assays were performed to test the ability of GAS to resist killing in the presence of neutrophils and 10% human serum as a complement source as described previously (42). Results from the phagocytic assays were reported as the logarithm of the change in CFU, defined as the total CFU after incubation divided by the total starting CFU.

Murine invasive infection.

Murine invasive soft-tissue infection and intraperitoneal infection were performed as previously described (7). In brief, female CD1 mice between 6 and 8 weeks old (Charles River Laboratories, Inc., Wilmington, Mass.) were inoculated with exponential-phase GAS either in the subcutaneous tissue or in the peritoneum. The challenge dose for each experiment is noted in the figure legends. Animals were observed daily. Moribund animals and animals surviving to day 10 (soft-tissue infection) or days 3 to 5 (intraperitoneal infection) were euthanized, and their spleens were removed and cultured to enable enumeration of GAS. PCR amplification of DNA purified from spleen isolates recovered from animals challenged with SLO-deficient GAS was performed with primers specific to slo, as noted above.

Statistical analysis.

Statistical calculations were performed with GraphPad Prism version 2.0 (GraphPad, San Diego, Calif.). Differences between strains in cytotoxicity were compared for significance with repeated-measures analysis of variance with Dunnett's posttest analysis (2). Differences between strains in resistance to phagocytic killing were compared for significance with a paired t test (2). Analysis for significant differences in mouse survival following GAS challenge employed the log-rank test (23).

RESULTS

Derivation and characterization of SLO and SLS deletion mutants in GAS strain 950771.

We derived SLO and SLS deletion mutants in the background of the M3 invasive GAS isolate 950771 or an isogenic streptomycin-resistant derivative, 950771Sm. Internally deleted slo and sagA sequences were amplified by PCR and cloned in the temperature-sensitive gram-positive shuttle vector pJRS233 (33). The cloned constructs were used to introduce the mutation in the GAS chromosome by allelic exchange mutagenesis (6). The genotypes of the parent and derived strains are shown in Table Table1.1. All genotypes were confirmed by Southern hybridization analysis of restriction endonuclease-digested chromosomal DNA probed with DNA from the target gene (data not shown).

TABLE 1.
GAS strains used in this study

When cultured on blood-agar plates and incubated aerobically, wild-type and SLO-deficient GAS produced large zones of beta-hemolysis surrounding the colonies. By contrast, SLS-deficient GAS produced markedly reduced zones of beta-hemolysis that were only visible at the edge of the bacterial colony. To demonstrate that the phenotypes observed in the blood-plate assays reflected activity specific to SLO and SLS, either late-exponential-phase culture supernatants (SLO) or late-stationary-phase cell extracts (SLS) were tested for the ability to lyse sheep erythrocytes in the presence and absence of cholesterol, a specific inhibitor of SLO (Table (Table2).2). As expected, wild-type GAS demonstrated both SLO and SLS hemolytic activity; SLO-deficient GAS lacked SLO hemolytic activity but had SLS hemolytic activity; and SLS-deficient GAS lacked SLS activity but retained SLO hemolytic activity. These results confirmed the targeted mutations and the association of the mutant genotypes with the anticipated hemolytic phenotypes.

TABLE 2.
Hemolytic activity of encapsulated GAS strains used in this study

SLO and SLS are cytotoxic to human keratinocytes in vitro.

The acute inflammation characteristic of both GAS pharyngitis and GAS skin infection is associated with epithelial cell injury. To identify the cytotoxic effects of SLO and SLS on keratinocytes, we incubated either wild-type or mutant GAS with squamous carcinoma cells and then observed the cells over time for changes in cell morphology and the ability to exclude the vital dye trypan blue as a marker of membrane integrity. After 24 h of exposure, wild-type and mutant GAS expressing SLO but not SLS caused extensive keratinocyte destruction compared with mutant GAS expressing SLS but not SLO, which produced modest injury of cells (Fig. (Fig.1).1). Uninfected keratinocytes remained intact (Fig. (Fig.1).1). Trypan blue dye uptake by these cells demonstrated a similar correlation between GAS hemolysin expression and cell injury: uptake was observed in 100% of cells exposed to wild-type GAS, in 63.2% ± 18.6% of cells exposed to SLS-deficient GAS. and in 39.3% ± 6.7% of cells exposed to SLO-deficient GAS (Fig. (Fig.1).1). These results indicated that GAS expression of SLO or SLS was independently associated with significant keratinocyte injury.

FIG. 1.
Keratinocyte morphology and uptake of trypan blue dye after exposure to GAS. SCC13 monolayers were incubated with GAS for 24 h. The bacteria were removed by washing. Microscopic photographs of the same fields were taken before (top) and after (bottom) ...

Influence of SLO and SLS on human PMN viability and phagocytic killing in vitro.

The ability of GAS to resist PMN killing is an important component of their pathogenicity. The observation that GAS expression of SLO and SLS contributed to keratinocyte injury led us to examine whether these molecules similarly impair PMN viability and their ability to kill GAS in vitro. To determine whether GAS expression of SLO and/or SLS effects PMN lysis, we used flow cytometry to detect significant decreases in PMN size and granularity after incubation with GAS and 10% normal human serum in conditions similar to a standard opsonophagocytic assay (42). When the values were normalized to cells that were not exposed to bacteria, incubation with wild-type or SLS-deficient GAS resulted in an equivalent, marked reduction in the percentage of intact PMNs detected compared to the percentage of intact PMNs detected after incubation with SLO-deficient GAS: cells incubated without GAS, 100% of PMNs intact; cells incubated with wild-type GAS, 47.1% ± 7.9% of PMNs intact; cells incubated with SLS-deficient GAS, 45.9% ± 10.3% of PMNs intact; and cells infected with SLO-deficient GAS, 77.4% ± 6.1% of PMNs intact (data are from three experiments, and the results of one representative experiment are depicted in Fig. Fig.22).

FIG. 2.
PMN lysis after exposure to GAS. Human PMNs were isolated from fresh whole blood and incubated with GAS in the presence of 10% human serum. After a 1-h incubation, the PMN population was analyzed by flow cytometry for changes in cell size and granularity ...

The number of intact PMNs after incubation with wild-type or SLS-deficient GAS was significantly less than the number of intact PMNs after incubation with SLO-deficient GAS (P = 0.042). Analysis of cellular debris for the presence of CD66b, a white blood cell-specific marker protein (41), confirmed that the decrease in PMN size and granularity detected in the assay was due to cell fragmentation (data not shown). These results indicated that GAS expression of SLO but not of SLS was associated with PMN lysis in vitro.

To determine whether GAS expression of SLO or SLS inhibits the ability of PMNs to kill bacteria, we quantified GAS survival after incubation with human whole blood or human PMNs and 10% human serum as a source of complement. In whole blood, GAS resistant to PMN killing typically increase exponentially in number over time. In the presence of PMNs and 10% serum, GAS that are resistant to PMN killing typically neither increase nor decrease in number. In human whole blood, the net changes in numbers of wild-type and SLO-deficient and wild-type and SLS-deficient GAS were equivalent (Fig. (Fig.3A).3A). Similarly, after incubation in the presence of PMNs and human serum, the change in the number of wild-type and SLO-deficient and wild-type and SLS-deficient GAS was not significantly different (Fig. (Fig.3B).3B). These results indicated that expression of neither SLO nor SLS significantly influenced the susceptibility of wild-type GAS to phagocytic killing in vitro.

FIG. 3.
Influence of SLO and SLS on resistance of encapsulated GAS to phagocytic killing. (A) Net changes in bacterial numbers after incubation in whole blood. Values represent the change in CFU after a 3-h incubation (mean ± standard error of the mean). ...

The ability of GAS deficient in SLO to resist PMN killing despite the lytic effects of SLO on these cells in vitro suggested a dominant protective effect of antiphagocytic molecules present on the bacterial surface. To determine whether hemolysin expression enhanced bacterial resistance to PMN killing in GAS deficient in one critical antiphagocytic factor, the hyaluronic acid capsule, we tested acapsular GAS deficient in SLO or SLS in the phagocytic assays. SLO- or SLS-deficient GAS were derived by allellic exchange mutagenesis in the background of an acapsular GAS mutant previously derived from GAS strain 950771 (Table (Table1)1) (6). The anticipated genotypes of the mutants were confirmed by Southern hybridization (data not shown), and the phenotypes of the mutants were confirmed by analysis of the pattern of hemolysis on blood agar.

Similar to the effect seen with encapsulated GAS, significant PMN lysis occurred in vitro after incubation with acapsular bacteria expressing SLO but not SLS (data not shown). Compared to the net growth of acapsular GAS in whole blood, the net growth of acapsular SLO-deficient GAS was less, although the magnitude of the decrease was not statistically significant (Fig. (Fig.4A).4A). Similarly, there was no significant difference between the net bacterial growth in whole blood of acapsular GAS and the net bacterial growth in whole blood of acapsular SLS-deficient GAS (Fig. (Fig.4A).4A). After incubation of the bacteria with purified human PMNs and serum, significantly fewer acapsular SLO-deficient GAS were recovered compared to the acapsular GAS parent strain (Fig. (Fig.4B).4B). There was no significant difference in the number of recovered acapsular SLS-deficient GAS compared to the number of recovered acapsular GAS in the assay (Fig. (Fig.4B).4B). These results indicated that expression of SLO but not SLS enhanced the resistance of acapsular bacteria to phagocytic killing.

FIG. 4.
Influence of SLO and SLS on resistance of acapsular GAS to phagocytic killing. (A) Net changes in bacterial numbers after incubation in whole blood. Values represent the change in CFU after a 3-h incubation (mean ± standard error of the mean). ...

Effect of SLO and SLS in murine invasive infection.

To characterize the influence of SLO and SLS on invasive infection, we assessed mouse survival after subcutaneous or intraperitoneal inoculation with wild-type and mutant GAS. The survival rates of mice inoculated subcutaneously with wild-type, SLO-, or SLS-deficient GAS were similar (Fig. (Fig.5A).5A). In these and all other mouse experiments, spleen cultures from animals that died were positive for GAS. Analysis of DNA PCR amplicons and hemolysis assays of culture supernatants confirmed that the isolates recovered from animals challenged with SLO-deficient GAS lacked SLO activity. Similarly, the absence of beta-hemolysis after culture on blood-agar medium confirmed that the splenic isolates recovered from animals challenged with SLS-deficient GAS lacked SLS activity. These results indicated that neither SLO nor SLS was required for full bacterial virulence in this mouse model of GAS invasive soft-tissue infection.

FIG. 5.
Mouse survival after subcutaneous or intraperitoneal challenge with encapsulated GAS. (A) The curves represent Kaplan-Meier estimates of mouse survival after subcutaneous inoculation with 3 × 107 CFU of either 950771 (60 animals), 950771SmSLO ...

Because the anatomical site of inoculation may influence the impact of bacterial factors on the course of infection, we performed experiments similar to those described above, except that mice were inoculated intraperitoneally. In contrast to the soft-tissue infection model, the survival of animals inoculated intraperitoneally with SLO-deficient GAS was significantly longer than that of animals challenged with either wild-type or SLS-deficient GAS (Fig. (Fig.5B).5B). Although these results demonstrated a contribution of SLO to mouse mortality after intraperitoneal inoculation with GAS, the magnitude of the effect was small and of uncertain biological importance. Similar to the data from mice after subcutaneous inoculation, there was no significant contribution of SLS to mouse mortality after intraperitoneal inoculation with GAS.

Since the results of the in vitro opsonophagocytic assay indicated that SLO expression increased the survival of acapsular GAS, we sought to determine whether acapsular GAS expressing SLO or SLS were more virulent than acapsular SLO- or SLS-deficient GAS in murine invasive infection. Because even maximal doses of acapsular GAS are avirulent in the skin model (6, 37), only intraperitoneal challenge studies were performed. The median survival time of animals challenged intraperitoneally with acapsular GAS was 36 h. By contrast, all animals challenged with the acapsular SLO-deficient strain survived (Fig. (Fig.6A).6A). Animals also survived significantly longer after challenge with the acapsular SLS-deficient strain than after challenge with the isogenic acapsular parent strain, although the increase in survival time was less than that observed after challenge with SLO-deficient GAS (Fig. (Fig.6B).6B). In each experiment, animals challenged with revertant strains (to control for possible nonspecific attenuation of bacterial virulence related to mutagenesis) had survival rates similar to those observed for animals challenged with the parent strains (Fig. 6A and 6B). These results indicated that GAS production of SLO and of SLS enhanced the virulence of acapsular GAS in mice after intraperitoneal challenge.

FIG. 6.
Mouse survival after intraperitoneal challenge with acapsular GAS. The curves represent Kaplan-Meier estimates of mouse survival after intraperitoneal inoculation. Each curve represents combined data from independent experiments with 10 mice per strain. ...

DISCUSSION

We derived isogenic GAS mutants deficient in SLO or SLS and characterized their phenotypes in assays of cytotoxicity and mouse virulence. Bacterial expression of both SLO and SLS was associated with significant keratinocyte injury, but the effect of SLO predominated. GAS production of SLO but not of SLS was associated with PMN lysis in vitro. Expression of neither SLO nor SLS inhibited PMN killing of wild-type GAS or was required for full virulence of wild-type GAS in mouse invasive soft-tissue infection. By contrast, SLO expression by encapsulated GAS enhanced bacterial virulence in mice after intraperitoneal challenge. In the background of acapsular GAS, SLO expression reduced PMN killing of bacteria in vitro and increased the virulence of GAS in mice after intraperitoneal challenge. Although SLS expression did not increase the survival of acapsular GAS in the phagocytic assays, the expression of SLS did enhance the virulence of the acapsular strain after intraperitoneal inoculation in mice.

Early work characterizing streptococcal virulence factors indicated that GAS were toxic to a variety of cultured cell lines, including oral epithelial cells, cardiac myocytes, cervical carcinoma cells, hepatocytes, renal cells, and synovial cells (26, 28, 34). In general, the results of these studies demonstrated that the cytotoxic effects observed required direct bacterial contact with the target cell. In some instances, the molecule effecting cell lysis was concluded to be SLS, since the cell injury could be inhibited by SLS antagonists and because GAS naturally deficient in SLS production produced little or no cell damage (28). The use of purified protein preparations also supported SLS cytotoxicity but in addition indicated that SLO mediated a similar effect (16, 28, 39). The conclusions from these early studies and their relevance to human infection were limited either by the inability to control for the expression of confounding cytotoxins by GAS or by the use of potentially supraphysiological concentrations of purified protein.

To directly address a cytotoxic effect of SLO on a clinically relevant cell line in conditions similar to natural infection, Ruiz et al. assessed the ability of isogenic SLO-deficient GAS to damage keratinocytes in vitro (36). Their experiments clearly demonstrated that GAS expression of SLO was associated with progressive keratinocyte damage and that adherence of the bacteria to the target cell was necessary for cell injury. The results of the current study confirm that GAS expression of SLO is associated with keratinocyte injury. In addition, our data demonstrate a cytotoxic effect that is directly attributable to SLS. Earlier reports demonstrating more complete cell lysis after incubation with purified SLS presumably reflect peptide concentrations significantly higher than those produced by GAS in vitro (28, 39).

Our results do not address the molecular events leading to keratinocyte injury. SLO- and SLS-mediated pore formation in the eukaryotic membrane could account for the observed uptake of trypan blue dye in injured cells and undoubtedly contributes to keratinocyte damage. In addition to direct membrane damage, two recent reports indicate that the cytotoxicity attributable to SLO also may reflect its participation in the intracellular translocation of another toxin, the GAS NAD+-glycohydrolase (13, 27).

Epithelial cell injury by GAS elicits an inflammatory reaction characterized in part by PMN recruitment to the site of infection. Over 30 years ago, Hirsch and colleagues analyzed the cytopathic effects of purified SLO and SLS preparations on rabbit PMNs by motion picture analysis (18). Exposure to SLO induced a series of changes in PMN morphology characterized by initial disruption of cytoplasmic granules, followed by the development of filamentous processes on the cell surface and eventual cytoplasmic and nuclear liquefaction. PMN degranulation in response to SLS exposure was similar to that seen after incubation with SLO, but the effect was delayed. Changes in the cell surface and nuclear fusion in response to SLS were less pronounced than the changes in response to SLO (18). Subsequent investigations demonstrated that preparations of SLO and SLS also decreased murine macrophage viability and ability to ingest GAS. In addition, sublethal concentrations of purified SLO impaired human PMN chemotaxis (4, 32).

The results of the present study extend these observations by investigating the toxic effects of SLO and SLS on PMNs under conditions that more accurately reflect natural infection. Our results indicate that the interaction between GAS and PMNs is associated with PMN destruction and that the majority of this effect is attributable to the bacterial expression of SLO. The absence of a lytic effect of bacterial SLS expression on PMNs contrasts with the findings of prior studies using purified peptide and is likely reflective of lower SLS concentrations in the bacterial culture supernatants compared to those achieved by addition of purified SLS to the assay medium.

Because SLO- or SLS-associated inhibition of PMN function without actual cell lysis may be the predominant pathogenic effect of these molecules on leukocytes, we also assessed the ability of GAS to survive in the presence of PMNs in vitro. In fresh human blood and in 10% human serum, wild-type GAS and GAS deficient in SLO or SLS were fully resistant to killing by PMNs. By contrast, acapsular GAS deficient in SLO but not SLS were more susceptible to killing by PMNs in both whole blood and human serum. These findings suggest that although bacterial production of SLO can destroy a significant fraction of the PMN population, this effect does not contribute to net bacterial survival because GAS produce a capsule that is very effective in inhibiting bacterial phagocytosis. In the absence of the antiphagocytic hyaluronic acid capsule, SLO production provides a second line of defense against phagocytosis through its cytotoxic effect on PMNs. We anticipate that SLO expression by GAS deficient in M protein, another critical antiphagocytic surface molecule, would also enhance bacterial resistance to phagocytic killing.

In contrast to our results, Biswas et al. noted that GAS deficient in SLS were more susceptible to phagocytic killing than the wild-type parent strain (12). Although the exact mechanism was unclear, these authors concluded that sagA inactivation in their mutant resulted in aberrant proteolytic processing of M protein; the protein was not attached to the cell wall but instead was secreted into the extracellular environment (12). The SLS-deficient mutants used in the current study grew as well as the parent strains in human whole blood, a test that is quite sensitive to the loss of M protein (22, 29). Therefore, it is unlikely that attachment of M protein to the cell wall was reduced by inactivation of sagA in our mutants.

It is possible that there are strain-specific differences in the influence of SLS on the linkage of M protein to the cell wall. Alternatively, the loss of surface-linked M protein in association with sagA inactivation noted by Biswas et al. may have been unrelated to the sagA mutation or, as the authors suggested, may have been due to polar effects of the mutation on a gene downstream of sagA. In any event, our findings indicate that loss of SLS is not necessarily associated with impaired surface localization of M protein and increased susceptibility of GAS to PMN killing.

The results of the in vitro phagocytic assays predicted the virulence of mutant GAS strains in the murine models. In mouse invasive skin infection, encapsulated GAS deficient in SLO or SLS were as virulent as wild-type GAS. SLO expression did enhance the virulence of encapsulated GAS after intraperitoneal challenge, but the magnitude of the effect was small and of uncertain biological significance. By contrast, in the background of acapsular GAS, SLO and, to a lesser extent, SLS expression significantly enhanced bacterial virulence, although the strains remained attenuated compared with the fully encapsulated wild-type strain. These results are consistent with a model in which the cytotoxic effect of SLO enhances the survival of bacteria that would otherwise be more susceptible to phagocytic clearance. Because the amount of both surface hyaluronic acid and M protein varies between GAS strains (22, 37), the contribution of SLO to the pathogenicity of wild-type GAS may be more critical in strains that are poorly encapsulated or express reduced levels of M protein. The mechanism by which SLS expression contributes to GAS pathogenicity does not appear to involve inhibition of bacterial killing by PMNs. It is possible that the cytotoxic effects of SLS on keratinocytes that we observed in vitro are reflective of epithelial cell injury that promotes bacterial invasion after intraperitoneal challenge in mice.

In the experiments described, we evaluated the role of SLS in GAS pathogenesis with bacteria recovered initially from exponential-phase growth in broth culture. The influence of bacterial growth phase on the experimental outcome was not addressed. It is possible that the use of GAS recovered from a late-stationary-phase broth culture under conditions of maximal SLS production in vitro may have demonstrated more pronounced SLS effects. The clinical practice of using protein synthesis inhibitors in the treatment of toxic shock syndrome caused by GAS is in part based on the assumption that this targets slowly dividing organisms expressing toxins that are relatively resistant to antibiotics that inhibit cell wall synthesis. It is important to note, however, that at the initiation of infection, relatively few organisms are present and these undergo rapid expansion to generate the large bacterial populations achieved during fulminant disease. We elected not to use stationary-phase GAS in these experiments because we believe that the clinically explosive nature of invasive GAS infections is predominantly initiated by a rapidly dividing bacterial population.

Although the cytotoxicity of purified SLO and SLS has been established, the influence of these molecules on GAS infection is still being defined. The current study demonstrates that bacterial expression of each hemolysin in vitro results in keratinocyte injury and that SLO expression mediates PMN lysis. PMN injury induced by SLO contributes to bacterial resistance to phagocytosis and is associated with enhanced GAS virulence in murine models of invasive GAS disease. SLS expression also enhances the virulence of GAS, but the mechanism of the effect remains undefined. The impact of SLO-mediated leukotoxicity on the course of bacterial infection is likely to be strain dependent and is predicted to be greatest in GAS that are relatively deficient in other mechanisms of resistance to phagocytic killing.

Acknowledgments

We thank Christopher Schiro and Jennifer Christianson for expert technical assistance.

This work was supported by Public Health Service grants AI29952 (M.R.W.) and R21 AI48743 (C.D.A.) and contract AI75326 from the National Institute of Allergy and Infectious Diseases.

Notes

Editor: V. J. DiRita

REFERENCES

1. Alouf, J. E. 1980. Streptococcal toxins (streptolysin O, streptolysin S, erythrogenic toxin). Pharmacol. Ther. 11:661-717. [PubMed]
2. Altman, D. G. 1991. Practical statistics for medical research, 1st ed. Chapman and Hall, New York, N.Y.
3. Andersen, B. R., and J. L. Duncan. 1980. Activation of human neutrophil metabolism by streptolysin O. J. Infect. Dis. 141:680-685. [PubMed]
4. Anderson, B. R., and D. F. VanEpps. 1980. Suppression of chemotactic activity of human neutrophils by streptolysin O. J. Infect. Dis. 125:353-359. [PubMed]
5. Ashbaugh, C. D., T. J. Moser, M. H. Shearer, G. L. White, R. C. Kennedy, and M. R. Wessels. 2000. Bacterial determinants of persistent throat colonization and the associated immune response in a primate model of human group A streptococcal infection. Cell. Microbiol. 2:283-292. [PubMed]
6. Ashbaugh, C. D., H. B. Warren, V. J. Carey, and M. R. Wessels. 1998. Molecular analysis of the role of the group A streptococcal cysteine protease, hyaluronic acid capsule, and M protein in a murine model of human invasive soft-tissue infection. J. Clin. Investig. 102:550-560. [PMC free article] [PubMed]
7. Ashbaugh, C. D., and M. R. Wessels. 2001. Absence of a cysteine protease effect on bacterial virulence in two murine models of human invasive group a streptococcal infection. Infect. Immun. 69:6683-6688. [PMC free article] [PubMed]
8. Ausubel, F. M., R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, J. A. Smith, and K. Struhl (ed.). 1987. Current protocols in molecular biology. John Wiley & Sons, New York, N.Y.
9. Bernheimer, A. W. 1949. Formation of a bacterial toxin (streptolysin S) by resting cells. J. Exp. Med. 90:373-392. [PMC free article] [PubMed]
10. Bernheimer, A. W., and L. L. Schwartz. 1964. Lysosomal disruption by bacterial toxins. J. Bacteriol. 87:1100-1104. [PMC free article] [PubMed]
11. Betschel, S. D., S. M. Borgia, N. L. Barg, D. E. Low, and J. C. S. De Azavedo. 1998. Reduced virulence of group A streptococcal Tn916 mutants that do not produce streptolysin S. Infect. Immun. 66:1671-1679. [PMC free article] [PubMed]
12. Biswas, I., P. Germon, K. McDade, and J. R. Scott. 2001. Generation and surface localization of intact M protein in Streptococcus pyogenes are dependent on sagA. Infect. Immun. 69:7029-7038. [PMC free article] [PubMed]
13. Bricker, A., C. Cywes, C. D. Ashbaugh, and M. R. Wessels. 2002. NAD+-glycohydrolase acts as an intracellular toxin to enhance the extracellular survival of group A streptococci. Mol. Microbiol. 44:257-269. [PubMed]
14. Davies, H. D., A. McGeer, B. Schwartz, K. Green, D. Cann, A. E. Simor, D. E. Low, and the Ontario Group A Streptococcal Study Group. 1996. Invasive group A streptococcal infections in Ontario, Canada. N. Engl. J. Med. 335:547-554. [PubMed]
15. Ginsburg, I. 1972. Microbial toxins, p. 99-171. Academic Press, New York, N.Y.
16. Ginsburg, I. 1972. Mechanism of cell and tissue injury by group A streptococci: relation to poststreptococcal sequelae. J. Infect. Dis. 126:294-456. [PubMed]
17. Herbert, D., and E. W. Todd. 1944. The oxygen-stable haemolysin of group A haemolytic streptococci (streptolysin S). Br. J. Exp. Pathol. 25:242-254.
18. Hirsch, J. G., A. W. Bernheimer, and G. Weissman. 1963. Motion picture study of the toxic action of streptolysins on leukocytes. J. Exp. Med. 118:223-227. [PMC free article] [PubMed]
19. Hryniewicz, W., and J. Pryjma. 1977. Effect of streptolysin S on human and mouse T and B lymphocytes. Infect. Immun. 16:730-733. [PMC free article] [PubMed]
20. Keiser, H., G. Weissman, and A. W. Bernheimer. 1964. Studies on lysosomes. IV. Solubilization of enzymes during mitochondrial swelling and disruption of lysosomes by streptolysin S and other hemolytic agents. J. Cell Biol. 22:101-113. [PMC free article] [PubMed]
21. Lancefield, R. 1957. Differentiation of group A streptococci with a common R antigen into three serological types with special reference to the bactericidal test. J. Exp. Med. 107:525-544. [PMC free article] [PubMed]
22. Lancefield, R. C. 1962. Current knowledge of the type specific M antigens of group A streptococci. J. Immunol. 89:307-313. [PubMed]
23. Lawless, J. F. 1982. Statistical models and methods for lifetime data. J. Wiley, New York, N.Y.
24. Limbago, B., V. Penumalli, B. Weinrick, and J. R. Scott. 2000. Role of streptolysin O in a mouse model of invasive group A streptococcal disease. Infect. Immun. 68:6384-6390. [PMC free article] [PubMed]
25. Loridan, C., and J. E. Alouf. 1986. Purification of RNA-core induced streptolysin S, and isolation and haemolytic characteristics of the carrier-free toxin. J. Gen. Microbiol. 132:307-315. [PubMed]
26. Lowry, P. N., and R. W. Quinn. 1964. Effect of hemolytic streptococci on tissue cells. Proc. Soc. Exp. Biol. Med. 116:46-51. [PubMed]
27. Madden, J. C., N. Ruiz, and M. Caparon. 2001. Cytolysin-mediated translocation (CMT): a functional equivalent of type III secretion in gram-positive bacteria. Cell 104:143-152. [PubMed]
28. Marcus, Z., A. M. Davies, and I. Ginsburg. 1965. Oxygen-stable hemolysins of group A streptococci. V. Effect on rat heart and kidney cells grown in tissue culture. Exp. Mol. Pathol. 4:93-107. [PubMed]
29. Moses, A. E., M. R. Wessels, K. Zalcman, S. A. Alberti, S. Natanson-Yaron, T. Menes, and E. Hanski. 1997. Relative contributions of hyaluronic acid capsule and M protein to virulence in a mucoid strain of group A streptococci. Infect. Immun. 65:64-71. [PMC free article] [PubMed]
30. Nizet, V., B. Beall, D. J. Bast, V. Datta, L. Kilburn, D. E. Low, and J. C. S. De Azavedo. 2000. Genetic locus for streptolysin S production by group A streptococci. Infect. Immun. 68:4245-4254. [PMC free article] [PubMed]
31. O'Connor, S. P., and P. P. Cleary. 1987. In vivo Streptococcus pyogenes C5a peptidase activity: analysis with transposon- and nitrosoguanidine-induced mutants. J. Infect. Dis. 156:495-504. [PubMed]
32. Ofek, I., S. Bergner-Rabinowitz, and I. Ginsberg. 1972. Oxygen-stable hemolysins of group A streptococci. 8. Leukotoxic and antiphagocytic effects of streptolysins S and O. Infect. Immun. 6:459-464. [PMC free article] [PubMed]
33. Perez-Casal, J., J. A. Price, E. Maguin, and J. R. Scott. 1993. An M protein with a single C repeat prevents phagocytosis of Streptococcus pyogenes: use of a temperature-sensitive shuttle vector to deliver homologous sequences to the chromosome of S. pyogenes. Mol. Microbiol. 8:809-819. [PubMed]
34. Quinn, R. W., and P. N. Lowry. 1967. Effect of Streptococcus pyogenes on tissue cells. J. Bacteriol. 93:1825-1831. [PMC free article] [PubMed]
35. Rheinwald, J. G., and M. A. Beckett. 1981. Tumorigenic keratinocyte lines requiring anchorage and fibroblast support cultures from human squamous cell carcinomas. Cancer Res. 41:1657-1663. [PubMed]
36. Ruiz, N., B. Wang, A. Pentland, and M. Caparon. 1998. Streptolysin O and adherence synergistically modulate proinflammatory responses of keratinocytes to group A streptococci. Mol. Microbiol. 27:337-346. [PubMed]
37. Schrager, H., and M. R. Wessels. 1996. Hyaluronic acid capsule and the role of streptococcal entry into keratinocytes in invasive skin infection. J. Clin. Investig. 98:1954-1958. [PMC free article] [PubMed]
38. Sharkawy, A., D. E. Low, R. Saginur, D. Gregson, B. Schwartz, P. Jessamine, K. Green, and A. McGeer. 2002. Severe group a streptococcal soft-tissue infections in Ontario: 1992-1996. Clin. Infect. Dis. 34:454-460. [PubMed]
39. Snyder, I. S., and T. R. Hamilton. 1963. Effect of streptolysin S on mammalian cells. J. Pathol. Bacteriol. 86:242-247. [PubMed]
40. Stevens, D. L. 1996. Invasive group A streptococcal disease. Infect. Agents Dis. 5:157-166. [PubMed]
41. Stocks, S. C., M. H. Ruchaud-Sparagano, M. A. Kerr, F. Grunert, C. Haslett, and I. Dransfield. 1996. CD66: role in the regulation of neutrophil effector function. Eur. J. Immunol. 26:2924-2932. [PubMed]
42. Wessels, M. R., A. E. Moses, J. B. Goldberg, and T. J. DiCesare. 1991. Hyaluronic acid capsule is a virulence factor for mucoid group A streptococci. Proc. Natl. Acad. Sci. USA 88:8317-8321. [PMC free article] [PubMed]

Articles from Infection and Immunity are provided here courtesy of American Society for Microbiology (ASM)
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...