H2O2-nonproducing Streptococcus pyogenes strains: survival in stationary phase and virulence in chronic granulomatous disease

Mitsumasa Saito1,2, Shouichi Ohga2, Miyoko Endoh3, Hiroaki Nakayamaa,4, Yoshimitsu Mizunoe1, Toshiro Hara2 and Shin-ichi Yoshida1

Departments of Bacteriology1 and Pediatrics2, Faculty of Medical Sciences, Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka 812-8582, Japan
Department of Microbiology, Tokyo Metropolitan Research and Laboratory of Public Health, Tokyo 169-0073, Japan3
Department of Microbiology, Faculty of Dentistry, Kyushu University, Higashi-ku, Fukuoka 812-8582, Japan4

Author for correspondence: Mitsumasa Saito. Tel: +81 92 642 6130. Fax: +81 92 642 6133. e-mail: msaito{at}bact.med.kyushu-u.ac.jp


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The production of hydrogen peroxide (H2O2) and related phenotypes were studied with Streptococcus pyogenes strains isolated from cases of pharyngitis or severe group A streptococcal infections. Of the 46 strains examined (34 from severe infections and 12 from pharyngitis cases), 25 strains accumulated H2O2 in the culture medium when grown under glucose-limited, aerobic conditions, whereas the rest of the strains did not. There was no correlation between these traits and the type of disease from which each strain had been isolated. The H2O2-nonproducing strains tested in this study belonged to T type 3 or T type 12. The accumulation of H2O2 started when the culture reached the late exponential phase. A rapid loss of cell viability accompanied H2O2 accumulation but was completely prevented by the addition of a catalase, indicating that the lethality was actually caused by H2O2. Cells of H2O2-nonproducing strains were resistant to killing by phagocytes from patients with chronic granulomatous disease (CGD), whereas those of H2O2-producing strains were subject to killing. Subcutaneous inoculation of 105 c.f.u. H2O2-nonproducing S. pyogenes strains into the hind footpads of CGD mice provoked more prominent swelling of the footpad than did H2O2-producing strains. The mortality rate in the CGD mice infected with the H2O2-nonproducing strains was higher than that produced by the H2O2-producing strains. It is suggested that H2O2-nonproducing S. pyogenes strains are prevalent in humans and that they may be a potential threat to the health of CGD patients.

Keywords: group A streptococci, hydrogen peroxide, CGD, gp91-phox knockout mice

Abbreviations: BHI, brain-heart infusion; CGD, chronic granulomatous disease

a Emeritus. Present address: 4395-2 Fukuma, Fukuoka-ken 811-3213, Japan.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Streptococcus pyogenes is the causative agent of a variety of human diseases, including pharyngitis, impetigo, necrotizing fasciitis, toxic-shock-like syndrome, and post-streptococcal diseases (acute glomerulonephritis and rheumatic fever) (Holm et al., 1992 ; Shanley et al., 1996 ; Talkington et al., 1993 ). Of these, necrotizing fasciitis and toxic-shock-like syndrome, often referred to as severe group A streptococcal infections, are particularly notable in that they are extremely life-threatening. Although a number of factors have been implicated in the virulence of this micro-organism, relatively little is known about what distinguishes this severe form from other common streptococcal diseases.

During the course of our study on pathogenicity factors of S. pyogenes strains in relation to the severe form of infection, we found that some strains were killed in the early stationary phase of growth under aerobic conditions, while others survived for more than 24 h after entering the stationary phase. Subsequent studies revealed that this loss of viability could be accounted for by a build-up of hydrogen peroxide (H2O2) in the culture medium, and established the presence of a dichotomy between positive and negative H2O2 accumulation in S. pyogenes. Streptococci possess superoxide dismutase, and many strains are known to produce H2O2. There have been many studies on H2O2 production by oral streptococci (Garcia-Mendoza et al., 1993 ; Holmberg & Hallander, 1973 ; LeBien & Bromel, 1975 ; Liebana et al., 1993 ; Willcox & Drucker, 1988 ), and this characteristic has often been important for the classification of these species (McLeod & Gordon, 1922 ; Whiley et al., 1990 ). However, there are few reports on H2O2 production by S. pyogenes (Gibson et al., 2000 ).

It has been known that H2O2-nonproducing organisms are common causes of infections in patients with chronic granulomatous disease (CGD) (Segal et al., 2000 ). CGD is an inherited disease involving phagocyte dysfunction, and is characterized by recurrent pyogenic infections that usually begin early in life (Gallin et al., 1983 ; Roos, 1994 ). The NADPH oxidase complex in CGD granulocytes produces deficient levels of H2O2. This results in impaired intracellular killing (Roos, 1994 ), which is compensated for partially by the H2O2 production of some infecting micro-organisms. Here we report the impaired bactericidal activity of CGD granulocytes against H2O2-nonproducing S. pyogenes strains.

Recently, gene targeting has led to the development of mouse models for both X-linked (gp91-phox-) and autosomal recessive (p47-phox-) forms of CGD (Jackson et al., 1995 ; Pollock et al., 1995 ). To compare the virulence of H2O2-producing and H2O2-nonproducing strains for CGD in vivo, we infected X-linked CGD mice with S. pyogenes strains (via the footpads of the animals). H2O2-nonproducing S. pyogenes strains provoked prominent footpad swelling and were more lethal to CGD mice.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains, culture and characterization.
The S. pyogenes strains used in this study comprised 34 isolates from severe group A streptococcal infections and 12 from pharyngitis cases. All of the strains were epidemiologically unrelated to each other. Strains were cultured in brain-heart infusion (BHI) broth (pH 7·4, 0·2% glucose concentration; Eiken Chemical) at 37 °C with shaking under aerobic conditions. C medium [0·5% Protease Peptone no. 3 (Difco), 1·5% yeast extract (Difco), 10 mM K2HPO4, 0·4 mM MgSO4 and 17 mM NaCl], was adjusted to pH 7·5 (Lyon et al., 1998 ) and used as a low-glucose basal medium to study the effects of glucose on H2O2 accumulation (see below). To monitor growth at intervals, the OD660 of bacterial cultures was measured. To count c.f.u., 10% sheep-blood agar plates were used. To test for catalase activity, cells were scraped from a culture plate and were then suspended in a drop of 3% H2O2 on a slide. Strains were judged catalase-positive when bubbles were observed in the suspension. Thioredoxin peroxidase activity in a crude cell extract was measured by the method described in a previous report (Jeong et al., 1999 ). T typing was performed by the method described in a manual published by the World Health Organization (Rotta & Facklam, 1980 ).

H2O2 measurement.
Samples (0·1 ml) were taken from test cultures, and the cells were removed by centrifugation. H2O2 was measured essentially by the method reported by Guilbault et al. (1966) . A solution containing homovanillic acid at 18 µg ml-1 and horseradish peroxidase at 2 µg ml-1 was prepared in 0·1 M sodium phosphate buffer (pH 7·0), and a 1 ml portion of it was mixed with 10 µl supernatant of each sample. The fluorescence intensity of the mixture was recorded at room temperature in a spectrophotofluorometer (model RF-500; Shimadzu) at a {lambda}ex of 315 nm and a {lambda}em of 425 nm (Guilbault et al., 1966 ). When a plateau was reached, a known amount of authentic H2O2 was added as internal reference, and the resulting increment in the fluorescence intensity was recorded; this was then used to convert the final fluorescence intensities for the samples into H2O2 concentrations.

Effects of glucose on H2O2 accumulation.
To test the effect of glucose on H2O2 accumulation, strains to be tested were cultured in C medium and glucose-supplemented C medium (the final glucose concentration was 1%, w/v) at 37 °C with shaking under aerobic conditions. At intervals, samples were taken to measure c.f.u. and the concentration of H2O2. The samples were serially diluted 10-fold with PBS then plated onto 10% sheep blood agar. The plates were incubated at 37 °C and the c.f.u. were counted after 18 h. The concentrations of H2O2 in the supernatants of the samples were measured by using the above method.

H2O2 sensitivity of organisms.
The H2O2 sensitivity of organisms was measured using the method described by Cleary & Larkin (1979) . Strains to be tested were cultured in BHI broth at 37 °C. At early stationary phase, the bacteria in 5 ml culture broth were washed with sterile PBS then resuspended in 5 ml PBS. H2O2 was added to each suspension to a final concentration of 5 µmol ml-1. The suspension was incubated at 37 °C under anaerobic conditions; at 0, 2, 4, 6 and 8 h after the addition of H2O2, samples were taken, and catalase (final concentration, 200 U ml-1) was added to each sample to eliminate the H2O2. After serial 10-fold dilution of the suspension with PBS, 0·1 ml was plated onto blood agar. The plates were incubated at 37 °C, and c.f.u. were counted after 18 h.

Lactate-sensitivity assays at different pH values.
To study the lactate sensitivity of the S. pyogenes strain, cells were grown in BHI broth (pH 7·4) at 37 °C. The cells (at early stationary phase) were washed twice and resuspended in PBS with or without sodium lactate, which was adjusted to the desired pH with HCl. The suspensions were incubated at 37 °C with gentle shaking. At intervals, 200 µl aliquots of the samples were transferred to test tubes, and, after serial 10-fold dilution with PBS, 0·1 ml was plated onto blood agar. The plates were incubated at 37 °C, and c.f.u. were counted after 18 h.

Intracellular bacterial killing by granulocytes of CGD patients.
Human peripheral blood was obtained from three patients with CGD (patients 1, 2 and 3) and from three healthy adults (controls). The granulocytes were separated aseptically by the dextran method (Kaplan et al., 1968 ). Patients 1 and 2 were diagnosed as having X-CGD (gp91-phox-). Patient 3 was diagnosed with autosomal recessive-type CGD (p47-phox-). Early stationary-phase S. pyogenes cells were used in the assay. Cells in broth cultures were collected by centrifugation then washed twice with sterile PBS. Intracellular killing by granulocytes was studied by using the method described by Hiemstra et al. (1992) . In vitro phagocytosis of serum-opsonized S. pyogenes was carried out for 5 min at 37 °C at a bacterium (c.f.u)/granulocyte ratio of 1:1 in Hanks’ balanced salt solution containing 0·1% (w/v) gelatin (gelatin-HBSS). Non-phagocytized bacteria were removed, and phagocytosis was stopped by washing twice with cold gelatin-HBSS. Approximately 5x106 granulocytes containing phagocytosed bacteria were suspended in 1 ml gelatin-HBSS containing 10% (v/v) fresh human serum from healthy adults of the AB blood group, and 0·25 ml portions were transferred to four sterile 1·5 ml microtubes (12310MTB1.5; Iwaki Glass). The granulocytes in the microtubes were cultured at 37 °C under slow rotation (10 r.p.m). At 0, 30, 60 and 120 min, one sample tube each was taken and its inner wall scraped with a sterile scraper (Techno Plastic Products) to remove adhering granulocytes. The cells were well suspended and then disrupted by dilution with distilled water containing 0·01% (w/v) BSA followed by vigorous mixing. The lysate was spread on blood-agar plates to determine the numbers of viable bacteria.

Animals.
Male and female X-linked CGD mice (Pollock et al., 1995 ), which were produced by knockout of gp91-phox of C57BL/6 mice, were donated by Dr Dinauer (Riley Children’s Hospital, Indianapolis, IN, USA) and Dr Kume (Division of Genetic Therapeutics, Center for Molecular Medicine, Jichi Medical School, Tochigi, Japan), and were reproduced in a specific pathogen-free animal room in our laboratory. Male, 6-week-old, X-linked CGD mice were used in the experiment. C57BL/6 mice of the same age and sex were also used as controls. They were purchased from Japan SLC.

Experimental infection, footpad swelling and mortality.
One group consisted of three mice infected with one S. pyogenes strain; they were kept in one cage. Subcutaneous infection was performed by inoculating the suspension of S. pyogenes strains (105 c.f.u. in a volume of 0·05 ml PBS) into the left hind footpads of mice, using a 27-gauge needle and a 1 ml syringe. As a control, 0·05 ml PBS was injected into the right hind footpad. The footpad thickness was measured by using a dial thickness gauge caliper (model G; Ozaki) at 24 and 48 h after inoculation; the difference between the thicknesses of left and right footpads was taken as the footpad swelling. The mortality of the CGD and control mice was observed until 10 d after infection. The mortality rates were determined by counting the number of dead mice in each group (consisting of 12 mice infected with four H2O2-producing and four H2O2-nonproducing strains, respectively).

Statistics.
The correlation between H2O2 production and the type of disease was analysed using Fisher’s exact test. Statistical analysis of the results of intracellular killing assay and animal experiments was performed with analysis of variance followed by a Student’s t-test. Data monitored over time were compared by using the area under the curve.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial growth curves and H2O2 production
Bacterial growth in BHI broth and the concentration of H2O2 accumulated in four representative strains are shown in Fig. 1. H2O2 accumulated in the culture media of strains SP2 and MK5 in the late exponential and early stationary phases, whereas it was undetectable in those of strains SP1 and ME198 throughout the whole culture period. The turbidity of the cultures of the four strains was maintained or found to decrease only slowly during the stationary phase. The numbers of viable cells of strains SP2 and MK5, but not those of strains SP1 and ME198, decreased rapidly in the stationary phase. When catalase (final concentration, 200 U ml-1) was added to the culture, the viability loss in strains SP2 and MK5 was prevented (data not shown). These results indicated that S. pyogenes strains were heterogeneous with respect to H2O2 production, which was responsible for the rapid loss of viability observed in the stationary phase of the culture. We therefore examined all of the 46 test strains for H2O2 production; the results obtained are shown, with reference to T types, in Table 1. Of the 34 strains that were isolated from patients with severe streptococcal disease, 15 strains (44·1%) did not produce H2O2. Six strains out of 12 isolates (50%) from patients with pharyngitis did not produce H2O2. There was no correlation between H2O2 production and the type of disease from which each strain had been isolated (P=0·75). All of the H2O2-nonproducing strains, which were epidemiologically unrelated to each other, belong to T types 3 or 12. Neither catalase activity nor thioredoxin peroxidase activity was detected in any strain (data not shown).



View larger version (30K):
[in this window]
[in a new window]
 
Fig. 1. Bacterial growth in BHI broth and H2O2 production of S. pyogenes strains SP1 (a), SP2 (b), ME198 (c) and MK5 (d). SP1 and SP2 are strains isolated from patients with severe group A streptococcal infection; ME198 and MK5 are strains isolated from patients with pharyngitis. {bullet}, Concentration of H2O2 in the culture medium; {circ}, OD660; {blacksquare}, numbers of viable bacteria.

 

View this table:
[in this window]
[in a new window]
 
Table 1. H2O2 production and T type of 46 S. pyogenes strains examined in this study

 
H2O2 sensitivity
The sensitivity to H2O2 of strains SP2 and MK5 (H2O2 producers) and strains SP1 and ME198 (H2O2 nonproducers) was compared. All strains were killed at equal rates in the presence of H2O2 at a concentration of 5 µmol ml-1, and a complete loss of viability was observed by 8 h (data not shown). It was also found that when H2O2 was added directly to stationary-phase cultures of strains SP1 and ME198 (H2O2 nonproducers) to a final concentration of 5 µmol ml-1, a complete loss of viability was observed by 8 h, and the H2O2 concentration of the cultures had been maintained for this period (data not shown).

Effects of glucose on H2O2 production and bacterial survival
It is known that H2O2 production by S. pyogenes is depressed when the micro-organism is cultured in a glucose-rich medium (Gibson et al., 2000 ). To test the effects of glucose in the culture medium on H2O2 production, two strains (SP1 and SP2) were cultured in C medium and glucose-supplemented C medium. When C medium was used, the results were the same as those with BHI broth (Fig. 1). Thus, H2O2 accumulated in the C medium of strain SP2, but not at all in that of strain SP1 (Fig. 2a, c). The numbers of viable cells of strain SP2, but not of strain SP1, decreased rapidly in the stationary phase because of H2O2 accumulation, as they did not decrease in the presence of catalase (data not shown). After cultivation in C medium for 36 h, the cultures showed final pH values in the range 6·6–6·7.



View larger version (28K):
[in this window]
[in a new window]
 
Fig. 2. Effects of glucose on H2O2 production. Strains SP1 (a, b) and strain SP2 (c, d) were cultured in C medium (a, c) and in glucose-supplemented C medium (b, d). {blacksquare}, Numbers of viable bacteria; {bullet}, concentration of H2O2 in the culture medium.

 
When glucose was added to C medium at a final concentration of 1% (w/v), the cultures of strain SP2, like those of strain SP1, did not accumulate H2O2 (Fig. 2b, d). However, despite the fact that two strains produced no detectable H2O2, they were killed very rapidly in the stationary phase (Fig. 2b, d). This apparent deviation from H2O2-dependent cell killing may be largely explained by the effect of the lactic acid derived from the glucose supplementation. Thus, the pH of the culture in glucose-enriched C medium was found to drop as low as 4·8 in the stationary phase. It was also found that 0·75% sodium lactate at pH 5·0 (a condition thought to be somewhat less severe than that actually present in the stationary phase culture) was capable of lowering the viability of SP1 cells by two orders of magnitude within a 3 h period (data not shown).

Intracellular killing by granulocytes of CGD patients
Intracellular killing of S. pyogenes strains was assayed by using human granulocytes from three CGD patients and three healthy control adults. S. pyogenes strains used in this assay were two H2O2-producers (strains SP2 and MK5) and two H2O2 nonproducers (strains SP1 and ME198). Strains SP1 and SP2 were isolated from the patients with severe group A infection. Strains MK5 and ME198 were isolated from the patients with pharyngitis. Fig. 3 shows the mean percentage of viable intracellular bacteria in normal and CGD granulocytes over 120 min after in vitro phagocytosis. In granulocytes from three CGD patients, strains SP2 and MK5 (H2O2-producers) showed similar change over 120 min, and the killing rates did not differ significantly (all P>0·05) at all three time points. Therefore, the data for these two strains phagocytosed by the granulocytes from three CGD patients were pooled, and their mean data (‘H2O2-producing strains phagocytosed by CGD granulocytes’) are shown in Fig. 3. Likewise, as strains SP1 and ME198 (H2O2 nonproducers) showed similar changes (all P>0·05), their data were pooled as ‘H2O2-nonproducing strains phagocytosed by CGD granulocytes’ (Fig. 3). In granulocytes from three normal control granulocytes, the difference between the two strains and the individual difference were also not significant (all P>0·05). Therefore, the results fell into four groups, i.e. H2O2-nonproducing and H2O2-producing strains phagocytosed by normal and CGD granulocytes.



View larger version (16K):
[in this window]
[in a new window]
 
Fig. 3. Intracellular killing of S. pyogenes strains by human granulocytes from three CGD patients (dashed lines) and three normal control adults (solid lines). S. pyogenes strains used in this assay were two H2O2-producers (strains SP2 and MK5) and two H2O2 nonproducers (strains SP1 and ME198). {circ}, two H2O2-nonproducing strains phagocytosed by CGD granulocytes; {square}, two H2O2-producing strains phagocytosed by CGD granulocytes; {bullet}, two H2O2-nonproducing strains phagocytosed by normal granulocytes; {blacksquare}, two H2O2-producing strains phagocytosed by normal granulocytes. Data are expressed as the mean (±SD) of percentages of viable intracellular bacteria obtained from six experiments, that is, three individuals and two bacterial strains.

 
In CGD granulocytes, the survival levels of intracellular H2O2-nonproducing organisms were almost 100% after a 120 min period, whereas those of H2O2-producing organisms decreased by 50% at 120 min (Fig. 3). There was a significant difference between the two groups in terms of the killing by CGD granulocytes (P=0·0001). These results demonstrated that the ability of S. pyogenes cells to produce H2O2 rendered them susceptible to intracellular killing by CGD granulocytes. On the other hand, normal granulocytes killed both H2O2-producers and H2O2 nonproducers at the same rate (P=0·90) (Fig. 3). The percentages of viable H2O2-nonproducing and H2O2-producing organisms in normal granulocytes were significantly lower than those in CGD granulocytes over 120 min, respectively (both P=0·0001). Thus, cells of H2O2-nonproducing strains were totally resistant to killing by phagocytes from patients with CGD, whereas those of H2O2-producing strains were subject to moderate killing.

Mouse footpad infection
To clarify the virulence of H2O2-nonproducing strains for CGD in vivo, we used X-linked CGD mice and performed subcutaneous infection by inoculating S. pyogenes strains into the footpads of the mice. Footpad swelling has been used for the assessment of inflammation and may reflect the defect of initial killing of bacteria by neutrophils. C57BL/6 mice were used as the control. S. pyogenes strains used in this experimental infection were four H2O2-producers (strains SP2, MK5, ME1410 and ME1123) and four H2O2 nonproducers (strains SP1, ME198, ME206 and ME1125). Strains SP1, SP2, ME206 and ME1410 were isolated from patients with severe group A streptococcal infection; the others were isolated from patients with pharyngitis. Fig. 4 shows the footpad swelling of CGD and control mice at 48 h after infection with eight strains. The mean (±SD) footpad swelling of CGD mice inoculated with H2O2-nonproducing and H2O2-producing strains was 2·42 (±0·49) mm and 0·34 (±0·46) mm at 24 h, and 3·16 (±0·43) mm and 0·58 (±0·65) mm at 48 h after inoculation, respectively. In CGD mice, the swelling of the footpads infected with H2O2-nonproducing strains was more prominent than that of footpads infected with H2O2-producing strains at both time points (both P=0·001). On the other hand, in control C57BL/6 mice, the mean (±SD) swelling of footpads infected with H2O2 nonproducers and H2O2-producers was, respectively, 0·20 (±0·20) mm and 0·17 (±0·18) mm at 24 h, and 0·33 (±0·36) mm and 0·15 (±0·15) mm at 48 h after infection. There were no differences between the two groups at both time points (P=0·82 and 0·31, respectively). The footpad swelling of CGD mice, in comparison to control mice, was significantly greater for challenges with both H2O2-nonproducing (P=0·0001) and H2O2-producing strains (P=0·0174) at 48 h after infection. Within 10 d after inoculation, some CGD mice were dead because of systemic infection. Fig. 5 shows the survival curves of the CGD and control mice after the infection of footpads. The mortality rate at day 10 in CGD mice infected with H2O2-nonproducing strains was higher than that for mice infected with H2O2-producing strains (92% versus 42%, respectively). No control mice died within the same period.



View larger version (33K):
[in this window]
[in a new window]
 
Fig. 4. Footpad swelling of CGD mice (hatched columns) and control C57BL/6 mice (solid black columns) at 48 h after infection with S. pyogenes strains. The S. pyogenes strains used in this experimental infection comprised four H2O2-producers (strains SP2, MK5, ME1410 and ME1123) and four H2O2 nonproducers (strains SP1, ME198, ME206 and ME1125). Strains SP1, SP2, ME206 and ME1410 were isolated from patients with severe group A infection; the others were isolated from patients with pharyngitis. Data are expressed as the means of results from three mice (±SD).

 


View larger version (14K):
[in this window]
[in a new window]
 
Fig. 5. Survival curves of CGD (dotted lines) and control (solid lines) mice after inoculation of their footpads with H2O2-producing and H2O2-nonproducing S. pyogenes strains. The mice and the S. pyogenes strains used in this experiment were the same as those used in the experiments presented in Fig. 4. {circ}, Four H2O2-nonproducing strains inoculated into the footpads of 12 CGD mice; {square}, four H2O2-producing strains inoculated into the footpads of 12 CGD mice; {bullet}, four H2O2-nonproducing strains inoculated into the footpads of 12 control C57BL/6 mice; {blacksquare}, four H2O2-producing strains inoculated into the footpads of 12 control C57BL/6 mice.

 

   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The ability to produce and excrete H2O2, manifested at low glucose concentrations, has been documented recently for a strain of S. pyogenes (Gibson et al., 2000 ). In the present study, we found that clinical isolates of this bacterium could be divided into two classes characterized by the presence or absence of H2O2 accumulation in the culture medium. Of the 46 strains examined, 21 were H2O2 nonproducers. The H2O2 accumulation was growth-phase dependent; it started when the culture was in the late exponential phase, the maximum level being attained in the early stationary phase. The rapid viability loss in H2O2-producing strains during the stationary phase was due to the toxicity of endogenously produced H2O2 (since it was abolished by the addition of catalase to the medium). Furthermore, there was no difference between H2O2-producing and H2O2-nonproducing strains in terms of sensitivity to exogenous H2O2.

The rapid cell death in the glucose-rich medium during the stationary phase was apparently independent of H2O2 production. This cell death was probably due to undissociated lactic acid (thought to be capable of crossing the plasma membrane) that was increased in the bacterial cells under weakly acid conditions (pH <5) and caused cell damage. The toxicity of organic acids produced by fermentation is generally explained by the transmembrane flux of undissociated acids under the acidic conditions (Russell & Diez-Gonzalez, 1998 ; Shimokawa & Nakayama, 1999 ).

One potentially significant finding was that all the H2O2-nonproducing strains, which were epidemiologically unrelated to each other, belonged to either T type 3 or T type 12. This apparent correlation must be studied with an expanded array of clinical isolates before its meaning is considered any further; however, it must be pointed out that both T types have H2O2-producing members as well. There was no correlation between the ability of a strain to accumulate H2O2 and the severity of the infection from which it had been isolated (Table 1).

The H2O2-producing phenotype was shown to affect the fate of bacterial cells within phagocytes. Cells of H2O2-producers were moderately susceptible to intracellular killing by CGD granulocytes deficient in H2O2 generation by the NADPH oxidase system, whereas those of H2O2 nonproducers were totally resistant to it. This finding is consistent with the well-known fact that bacteria that do not produce H2O2 constitute the main cause of pyogenic infections in CGD patients (Segal et al., 2000 ). Furthermore, we noted that H2O2-nonproducing S. pyogenes strains caused prominent inflammation (Fig. 4) and higher mortality (Fig. 5) in CGD mice. This suggests the possibility that the H2O2-nonproducing S. pyogenes strains would be important causative agents of severe infections in the patients with CGD. However, this relationship is not readily recognizable in clinical cases of streptococcal infections. Thus, Gallin et al. (1983) reported that out of 119 major febrile episodes in 14 CGD patients, Streptococcus was isolated only rarely (three Streptococcus pneumoniae strains). Regelmann et al. (1983) found only five Streptococcus strains (three S. pneumoniae and two Streptococcus intermedius) among 93 micro-organisms isolated from 15 CGD patients. It is known that S. pneumoniae produces H2O2 (Spellerberg et al., 1996 ; Ukada et al., 1959 ). However, the significance of its isolation from CGD patients is unclear, because it is the most common cause of pneumonia, even in healthy individuals. On the other hand, our previous study demonstrated that S. intermedius, an H2O2 nonproducer (Whiley et al., 1990 ), was the causative agent of severe infection (brain abscess) in a CGD patient, and its pathogenicity was attributed to the lack of H2O2 production (Nagatomo et al., 1999 ).

The rarity of S. pyogenes infection in CGD patients may be due to an early start of antimicrobial therapy and the virtual absence of drug-resistant strains. Although S. pyogenes infections are presently not common in CGD patients, it would be of interest to examine the H2O2-production phenotypes of isolates from such rare cases. In contrast to CGD granulocytes, normal granulocytes killed H2O2-producers and H2O2 nonproducers with practically equal efficiency. This may be taken to suggest that the H2O2 phenotype of S. pyogenes should be of little significance as a virulence factor for infections in healthy individuals.

Also interesting is the biochemistry of H2O2 production and related matters in this organism. Accumulation of H2O2 in the culture medium in large amounts is a unique feature common among lactic acid bacteria (streptococci, lactococci, enterococci and lactobacilli). At variance with most other organisms (in which H2O2 is derived largely from the superoxide anion), lactic acid bacteria are generally believed to produce H2O2 mainly through oxidation of substrates such as NADH by H2O2-producing oxidases (Thomas & Pera, 1983 ; Murphy & Condon, 1984 ; Marty-Teysset et al., 2000 ; Gibson et al., 2000 ). In fact, a homology search of the S. pyogenes genome database has revealed the genes of putative H2O2-producing oxidases for NADH, {alpha}-glycerophosphate and lactate, and their involvement in H2O2 production has been suggested (Gibson et al., 2000 ). In accordance with the widely held view, we are also of the opinion that at least a major part of the H2O2 accumulated in S. pyogenes cultures is produced by one or more of the above-mentioned oxidases.

The accumulation of H2O2 was repressed by the presence of high concentrations of glucose, and was manifested only after cell growth had slowed down in the late exponential phase. This apparent ‘glucose effect’ has also been well documented for lactic acid bacteria. Although precise mechanisms for these phenomena still remain unclear, their biological significance might be evident if one takes account of the fact that the above-mentioned oxidase reactions are functionally linked to the ‘thioclastic’ pathway of pyruvate metabolism. This ATP-coupled pathway is known to operate under glucose-limited conditions in a variety of lactic acid bacteria, and is energetically advantageous (for a review, see Stouthamer, 1978 ). In fact, the S. pyogenes genome database shows the presence of the complete set of genes for the thioclastic pathway. Finally, the dichotomy between the presence and absence of H2O2 accumulation, a situation also widely seen among lactic acid bacteria, awaits explanation. Although not discussed here, a role for the H2O2-degrading system must be considered when such a dichotomy is addressed. Attempts to answer these unsolved questions are under way in the authors’ laboratory.

In conclusion, we found that H2O2-nonproducing S. pyogenes strains resisted intracellular killing by CGD phagocytes, and we demonstrated their virulence to CGD mice in vivo. Our findings raise the possibility that H2O2-nonproducing S. pyogenes strains could be a potential threat to CGD patients, though such cases have not yet been documented.


   ACKNOWLEDGEMENTS
 
We thank M. Kurokawa for providing the clinical S. pyogenes strains, K. Ohga, H. Kajiwara and T. Ishikawa for technical assistance, W. Kong for analysis of thioredoxin peroxidase activity, S. Kono for statistical analysis, and K. Nakayama for helpful discussion and advice. We also thank Linda Saza for her valuable editorial advice on the manuscript.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Cleary, P. P. & Larkin, A. (1979). Hyaluronic acid capsule: strategy for oxygen resistance in group A streptococci. J Bacteriol 140, 1090-1097.[Medline]

Gallin, J. I., Buescher, E. S., Seilingmann, B. E., Nath, J., Gaither, T. E. & Katz, P. (1983). Recent advances in chronic granulomatous disease. Ann Intern Med 99, 657-674.[Medline]

Garcia-Mendoza, A., Liebana, J., Castillo, A. M., de la Higuera, A. & Piedrola, G. (1993). Evaluation of the capacity of oral streptococci to produce hydrogen peroxide. J Med Microbiol 39, 434-439.[Abstract]

Gibson, C. M., Mallett, T. C., Claiborne, A. & Caparon, M. G. (2000). Contribution of NADH oxidase to aerobic metabolism of Streptococcus pyogenes. J Bacteriol 182, 448-455.[Abstract/Free Full Text]

Guilbault, G. G., Brignac, P. J. & Zimmer, M. (1966). Homovanillic acid as a fluorometric substrate for oxidative enzyme. Anal Chem 38, 190-196.

Hiemstra, P. S., Annema, A., Schippers, E. F. & van Furth, R. (1992). Pertussis toxin partially inhibits phagocytosis of immunoglobulin G-opsonized Staphylococcus aureus by human granulocytes but does not affect intracellular killing. Infect Immun 60, 202-205.[Abstract]

Holm, S. E., Norrby, A., Bergholm, A. M. & Norgren, M. (1992). Aspects of pathogenesis of serious group A streptococcal infections in Sweden, 1988–1989. J Infect Dis 166, 31-37.[Medline]

Holmberg, K. & Hallander, H. O. (1973). Production of bactericidal concentrations of hydrogen peroxide by Streptococcus sanguis. Arch Oral Biol 18, 423-434.[Medline]

Jackson, S. H., Gallin, J. I. & Holland, S. M. (1995). The p47phox mouse knock-out model of chronic granulomatous disease. J Exp Med 182, 751-758.[Abstract]

Jeong, J. S., Kwon, S. J., Kang, S. W., Rhee, S. G. & Kim, K. (1999). Purification and characterization of a second type thioredoxin peroxidase (type II TPx) from Saccharomyces cerevisiae. Biochemistry 38, 776-783.[Medline]

Kaplan, E. L., Laxdal, T. & Quie, P. G. (1968). Studies of polymorphonuclear leukocytes from patients with chronic granulomatous disease of childhood: bactericidal capacity for streptococci. Pediatrics 41, 591-599.[Abstract]

LeBien, T. W. & Bromel, M. C. (1975). Antibacterial properties of a peroxidogenic strain of Streptococcus mitior (mitis). Can J Microbiol 21, 101-103.[Medline]

Liebana, J., Parejo, E., Castillo, A. & Gutierrez, J. (1993). Phenotypic characterization of oral streptococci by classical methods. Microbios 76, 7-18.[Medline]

Lyon, W. R., Gibson, C. M. & Caparon, M. G. (1998). A role for trigger factor and an rgg-like regulator in the transcription, secretion and processing of the cysteine proteinase of Streptococcus pyogenes. EMBO J 17, 6263-6275.[Abstract/Free Full Text]

McLeod, J. W. & Gordon, J. (1922). Production of hydrogen peroxide by bacteria. Biochem J 16, 499-504.

Marty-Teysset, C., de la Torre, F. & Garel, J.-R. (2000). Increased production of hydrogen peroxide by Lactobacillus delbrueckii subsp. bulgaricus upon aeration: involvement of an NADH oxidase in oxidative stress. Appl Environ Microbiol 66, 262-267.[Abstract/Free Full Text]

Murphy, M. G. & Condon, S. (1984). Correlation of oxygen utilization and hydrogen peroxide accumulation with oxygen induced enzymes in Lactobacillus plantarum cultures. Arch Microbiol 138, 44-48.[Medline]

Nagatomo, T., Ohga, S., Saito, M., Takada, H., Sasaki, Y., Okada, K., Inamura, T. & Hara, T. (1999). Streptococcus intermedius-brain abscess in chronic granulomatous disease. Eur J Pediatr 158, 872-873.[Medline]

Pollock, J. D., Williams, D. A., Gifford, M. A., Li, L. L., Du, X., Fisherman, J., Orkin, S. H., Doerschuk, C. M. & Dinauer, M. C. (1995). Mouse model of X-linked chronic granulomatous disease, an inherited defect in phagocyte superoxide production. Nat Genet 9, 202-209.[Medline]

Regelmann, W., Hays, N. & Quie, P. G. (1983). Chronic granulomatous disease: historical perspective and clinical experience at the University of Minnesota Hospital. Adv Host Def Mech 3, 3-23.

Roos, D. (1994). The genetic basis of chronic granulomatous disease. Immunol Rev 138, 121-157.[Medline]

Rotta, J. & Facklam, R. R. (1980). Type identification of group A streptococci. Manual of Microbiological Diagnostic Methods for Streptococcal Infections and their Sequelae , 25-35. Geneva:World Health Organization.

Russell, J. B. & Diez-Gonzalez, F. (1998). The effects of fermentation acids on bacterial growth. Adv Microb Physiol 39, 205-234.[Medline]

Segal, B. H., Leto, T. L., Gallin, J. I., Malech, H. L. & Holland, S. M. (2000). Genetic, biochemical, and clinical features of chronic granulomatous disease. Medicine 79, 170-200.[Medline]

Shanley, T. P., Schrier, D., Kapur, V., Kehoe, M., Musser, J. M. & Ward, P. A. (1996). Streptococcal cysteine protease augments lung injury induced by products of group A streptococci. Infect Immun 64, 870-877.[Abstract]

Shimokawa, O. & Nakayama, H. (1999). Acetate-mediated growth inhibition in sterol 14 alpha-demethylation-deficient cells of Candida albicans. Antimicrob Agents Chemother 43, 100-105.[Abstract/Free Full Text]

Spellerberg, B., Cundell, D. R., Sandros, J., Pearce, B. J., Idanpaan-Heikkila, I., Rosenow, C. & Masure, R. (1996). Pyruvate oxidase, as a determinant of virulence in Streptococcus pneumoniae. Mol Microbiol 19, 803-813.[Medline]

Stouthamer, A. H. (1978). Energy-yielding pathways. In The Bacteria , pp. 389-462. Edited by L. N. Ornston & J. R. Sokatch. New York:Academic Press.

Talkington, D. F., Schwartz, B., Black, C. M., Todd, J. K., Elliott, J., Breiman, R. F. & Facklam, R. R. (1993). Association of phenotypic and genotypic characteristics of invasive Streptococcus pyogenes isolates with clinical components of streptococcal toxic shock syndrome. Infect Immun 61, 3369-3374.[Abstract]

Thomas, E. L. & Pera, K. A. (1983). Oxygen metabolism of Streptococcus mutans: uptake of oxygen and release of superoxide and hydrogen peroxide. J Bacteriol 154, 1236-1244.[Medline]

Ukada, S., Koukol, J. & Vennesland, B. (1959). Lactic oxidase of the pneumococcus. J Bacteriol 78, 714-725.[Medline]

Whiley, R. A., Fraser, H., Hardie, J. M. & Beighton, D. (1990). Phenotypic differentiation of Streptococcus intermedius, Streptococcus constellatus, and Streptococcus anginosus strain within the ‘‘Streptococcus milleri Group’’. J Clin Microbiol 28, 1497-1501.[Medline]

Willcox, M. D. P. & Drucker, D. B. (1988). Partial characterisation of the inhibitory substances produced by Streptococcus oralis and related spieces. Microbios 55, 135-145.[Medline]

Received 2 March 2001; revised 10 May 2001; accepted 18 May 2001.