- Email: [email protected]
Streptococcal Pyrogenic Exotoxin A, Streptolysin O, Exoenzymes, Serotype and Biotype Profiles of Streptococcus pyogenes Isolates from Patients with Toxic Shock Syndrome and other Severe Infections
Zbl. Bakt. 286, 421-433 (1997) © Gustav Fischer Verlag
Zentralblatt fur
BakterioiOSie
Streptococcal Pyrogenic Exotoxin A, Streptolysin 0, Exoenzymes, Serotype and Biotype Profiles of Streptococcus pyogenes Isolates from Patients with Toxic Shock Syndrome and other Severe Infections Heide Miiller-Alouf 1, 2, Christiane Geoffroy 1, P. Geslin 3, Anne Bouvet\ Annie Felten s, Elisabeth Giinther6 , J.-H. Ozegowski 6, and J. E. AlouP,2 1, 2
3 4
5 6
Unite des Toxines Microbiennes (URA 1858, Centre National de la Recherche Scientifique), Institut Pasteur, F-75724 Paris Cedex 15 and Institut Pasteur de Lille, F-59019 Lille Cedex, France Centre Hospitalier Intercommunal, F-94010 Crete ii, France Universite Paris VI, Hopital Hotel-Dieu, F-75181 Paris 04, France Hopital Saint Louis, F-75010 Paris, France Institut fiir experimentelle Mikrobiologie, Friedrich Schiller Universitat, Jena, Germany
Received February 5, 1997 . Accepted February 13, 1997
Summary The determination of protein M and T serotypes, biotypes and pyrogenic (erythrogen ic) exotoxin A (SPE A), streptolysin 0 (SLO), streptokinase (SK), hyaluronidase (HA) and cysteine proteinase release by 212 S. pyogenes isolates from patients with severe invasive group A streptococcal (GAS) infections, among them 74 cases of streptococ cal toxic shock syndrome (STSS) has been investigated. Ml or M3 serotypes were ex pressed by 25% of the isolates (53/212), whereas 59% (125/212) belonged to 15 oth er different serotypes and 16% (34/212) were untypeable. Of the 74 isolates from STSS patients, 42% (31/74) expressed Ml and to a lesser extent M3 serotypes versus 19% of the non STSS isolates (26/138) . Among the ten different biotypes known, biotypes 1 and 3 were prevalent, particularly the former in the case of STSS isolates. SPE A was detectably produced by about 25% (54/212) of the strains. However, as high as 40.5% of the STSS isolates (30/74) versus 17.4% of non STSS isolates (24/138) released SPE A. Moreover, 67% of the SPE A producing strains were of serotype Ml or M3. SK and HA were released by 71 % and 10% of the isolates respectively. All
422
H. Miiller-Alouf et al.
strains released SLO (4 to 256 HU/ml) and 85% cysteine proteinase. No relationship between toxin or enzyme titer and the type of disease or clinical origin of the strains was found. Culture supernatants of all isolates showed moderate to high lymphocyte trans forming activity with index values ranging from 14.5 to 50.3 including those strains which did not release detectable amounts of SPE A suggesting that SPE C and other mitogenic factor(s) are released by the isolates investigated.
Introduction After a four-decade decline of the prevalence of severe invasive group A strep tococcal (GAS) diseases in developed countries, a worldwide increase since the mid 1980s of the incidence of these infections (bacteremia, myositis, cellulitis, fasciitis, ... ) with mortality rates ranging from 20 to 60% has been reported, particularly in North America and Europe (4, 19,29). These severe GAS dis eases also include the clinical entity called streptococcal toxic shock syndrome (STSS) suspected in 1983 (42) and widely described and documented since 1987 (19, 23, 36, 37, 38, 43). This syndrome characterized by hypotension, shock, fever, multiorgan failure, skin desquamation and other biological per turbations (4, 43) corresponds very likely to the cases of scarlatina maligna described in older literature (20, 23) during outbreaks of scarlet fever. The bacterial effectors which are believed to be causally involved in the se verity of GAS infections are certain serotypes of M protein, a major cell sur face virulence factor (9), streptococcal pyrogenic (erythrogenic) exotoxins A and probably C (SPE A, SPE C) which belong to the family of the bacterial superantigens (1). In this respect, many recent reports concerning the S.pyo genes strains currently isolated from patients with severe invasive GAS infec tions and particularly STSS pointed to the predominance of strains expressing Ml and M3 serotypes (as well as M12 and M18 in certain reports), concom itantly with the production of SPE A (4, 6, 7, 15,20, 24, 29, 34, 36, 37, 38, 40). Other extracellular toxins such as streptolysin 0 (SLO), a membrane damaging sulfhydryl-activated toxin (2), and cysteine proteinase (interleukin IP convertase) which has been shown on immunochemical (10) and genetic grounds (5, 15) to be identical to or a variant of pyrogenic (erythrogenic) exo toxin B (5, 10,35) are also thought to contribute to the pathogenesis of GAS diseases (16,20). These considerations and recent work in our laboratory (27, 28) and by others (14) on cytokine induction by streptococcal toxins in relation to STSS and other diseases, led us to analyze 212 S. pyogenes strains isolated in a num ber of hospitals from various areas of France between 1990 and 1994 from patients with STSS and/or other invasive GAS infections. The strains provid ed to the laboratory of one of us (J. E. A.) were tested with the cooperation of the other authors for M and/or T serotypes, the recently defined streptococ cal biotypes (3), and for the release in culture medium of SPE A, SLO, strep-
Invasive S. pyogenes strains from severe infections
423
tokinase, hyaluronidase and cysteine proteinase. The transforming activity of culture supernatants (mitogenic effect) on human lymphocytes has been also undertaken.
Materials and Methods S. pyogenes clinical isolates and patients. The strains investigated were isolated from 212 patients (108 males and 114 females, mean age 47 years, range 3 months to 94 years) who all met the criteria for invasive GAS infections established by the Working Group on Severe Streptococcal Infections (43): (i) STSS (ii), other infections defined by isolation of S. pyogenes from normal sterile sites not meeting criteria for STSS (bacte remia with no identified focus, focal infections with or without bacteremia particular ly skin and deep-soft tissue infection), (iii) scarlet fever. On the basis of this classification, there were among the 212 patients, 74 confirmed cases (35%) of STSS and 138 cases (65%) with other severe (invasive) GAS infections. The patients with STSS (28 males and 46 females, mean age 36 years, range 5 months to 94 years) presented the following associated clinical features: bacteremia without identified focus (49 patients), bacteremia associated to cellulitis or necrotizing fasciitis (15 patients), necrotizing fasciitis (4 patients), postpartum infection (6 patients). The other severe GAS infections without associated STSS (64 males and 74 females, mean age 56 years, range 5 months to 94 years) included severe scarlet fever (9 cases), bac teremia without identified focus (40 cases), bacteremia with associated pneumonia (14 cases), deep soft-tissue infections (29 cases), or erysipelas (21 cases), infected burn wounds (14 cases), varicella gangrenosa (11 cases). The majority of the strains isolat ed from patients with bacteremia were provided by P. Geslin (coordinator of the mul ticenter study on bacteremia). An overall fatality rate of 13.7% was recorded among the 212 patients (29/212). Seventeen deaths occurred among the 74 patients with STSS (23% fatality rate). Bacterial culture. The S. pyogenes isolates were purified by streaking on 5% sheep blood Columbia agar plates and then cultured on Todd-Hewitt broth (Difco Labora tories) at 37 °C to the stationary phase of growth. The cultures were centrifuged and bacterial pellets separated for typing purposes. The supernatant fluids were sterilized by filtration through 22-[,Im-pore size Millipore filters (Millex-GV, Millipore). The fil trates were stored at -20 °C for exotoxin, exoenzyme, and mitogenic activity assay or detection. Serotyping of M and T antigens. M typing was determined by double immunodif fusion on agar gel using Lancefield extracts of streptococcal cells and specific antisera prepared in the laboratory of one us (E. G. National Reference Center for Streptococ ci, Jena). T-typing was done by agglutination of trypsinized bacterial suspensions. Biotyping. This new typing method of S. pyogenes strains of epidemiological inter est based on enzymatic and carbohydrate fermentation reactions was determinated by one of us (A. B.) as recently described (3). The testing for the ten biotypes character ized to date was carried out with the rapid ID 32 STREP identification system (Bio Merieux, La-Balme-Ies-Grottes, France). Immunoassay of streptococcal pyrogenic exotoxin A (SPE A). The supernatant fluids of the strains cultured on Todd-Hewitt medium were 100-fold concentrated by ethanol precipitation at -20°C as described (22). The precipitate was dissolved in
424
H. Miiller-Alouf et al.
acetate-buffered saline, pH 4.0. SPE A was determined quantitatively in the concen trated material by a sensitive and specific enzyme linked immunosorbent assay (ELISA) using two mono specific polyclonal anti SPE A (goat and rabbit) hyperimmune serum as described (22). The detection limit for SPE A is ca.10 pg/ml. Streptolysin 0 (SLO) assay. This membrane damaging toxin was assayed by the de termination of its hemolytic activity on sheep red blood cells expressed as hemolytic units (HU) per ml of culture supernatants after activation with 20 mM cysteine as de scribed (2). Streptokinase (SK) assay. The quantitation of this enzyme in culture supernatants was determined by the skim milk/plasminogen agarose plate technique (25). The area of casein lysis into the agarose medium around the wells filled with test samples and a standard purified SK solution can be correlated to the amount of enzyme released in bacterial cultures. Hyaluronidase (HA) assay. Enzyme determination was undertaken by the turbidi metric method previously described (32). Serial dilutions of culture supernatants (500 !tl) in 500 !tl of 100 mM acetate buffer pH 6.0 were incubated at 37°C for 30 min with 500 !tl of 0.2 mg/ml hyaluronic acid in 10 mM acetate buffer purified from Strep tococcus zooepidemicus. The reaction was stopped by adding 3 ml of 2.5% (w/v) ce tyltrimethylammonium bromide in 0.5 N sodium hydroxide. After standing 20 min at room temperature, the turbidity was photometrically measured at 600 nm. One HA unit is that amount which splits 50 ng of hyaluronic acid within 30 min. Proteinase evaluation. The plate assay method of Haynes and Tagg (17) based upon the digestion of skim milk (1.5% w/v) contained within Columbia agar base medium was used for detection of proteinase production by individual surface grown colonies of the strains. The diameter of the proteolysis zones was measured after 24 h of aero bic incubation of the plates at 3rc. Protease activity was calculated as the ratio of the diameter of the proteolysis zone to the diameter of the corresponding colony. Mitogenic activity of culture supernatants. The lymphocyte blast transforming ac tivity of the supernatants were determined on human peripheral blood mononuclear cells isolated and cultivated on RPMI 1640 medium (Gibco) as described (27). The evaluation of lymphocyte transformation after cell pulse with tritiated thymidine (Amersham) was undertaken as described (22). The results were expressed as stimula tion index based on the ratio of cpm in lymphocyte culture with bacterial supernatant! cpm in lymphocyte culture with sterile Todd-Hewitt broth. Statistical analysis. The Student's t test was used to determine the significance of dif ferences between STSS associated isolates and the other isolates from GAS infections. A P value of 0.05 or less was considered statistically significant.
Results M and T serotypes of the clinical isolates. Of the 212 isolates investigated, 34 (16 %) were untypeable, whereas the other 178 isolates (84 %) belonged to 17 different M and (or) T serotypes (Table 1). Despite that all strains were iso lated from severe invasive GAS infections, only 53 isolates (25%) expressed either serotype M1 (40 isolates) or M3 (13 isolates). Interestingly, 42% of the STSS strains (31/74) belonged to the serotypes M1 or M3 whereas only 19% of the strains (26/138) isolated from those patients without associated STSS
Invasive S. pyogenes strains from severe infections
425
Table 1. Mff serotype distribution among S. pyogenes isolates from patients with invasive GAS infections with or without associated STSS Serotype
All isolates (%) n =212
STSS isolates (%) n = 74
Other isolates (%) n = 138
M1 M2 M3rr3 M4rr4 T5 M9 Mll M12rr12 M13rr13 T25 R28rr28 B3264 T33 M41 M48 T56 T78 Untypeable
40 3 13 8 5 1 16 22 12 6 3 22 4 5 1 15 2 34
20 (0.35) 0 (0.0) 5 (6.7) 4 (5.4) 5 (6 .7) 0 (0 .0) 4 (5.4) 4 (5.4) 3 (4.0) 2 (2.7) 1 (1.3) 3 (4.0) 1 (1.3) 1 (1.3) 0 (0.0) 5 (6.7) 1 (1.3) 9 (12.2)
14 3 8 4 0 1 12 18 9 4 2 19 3 4 1 10 1 25
(18.9) 0.4) (6 .1) (3.7) (2.3) (0.5) (7.5) (10.4) (5.6) (2.8) (1.4) (10.8) (1.9) (2.4) (0.5) (7.1) (0.9) (16.0)
(10.4) (2.2) (5.8) (2.9) (0.0) (0.7) (8.7) (13.0) (6.5) (2.9) (1.2) (13.8) (2.2) (2.9) (0.7) (7.3) (0.7) (18.0)
expressed these two markers. As shown in Fig. 1, the percentage of strains ex pressing the serotype Ml was significantly much higher (p < 0.01) for STSS strains whereas the percentage of strains expressing the serotype M3 was practically the same for both groups of isolates. Biotypes of the clinical isolates. Biotypes 1 and 3 and to a lesser extent bio type 5 were the most preponderant among the strains tested (Fig. 1). Biotype 1 expression was significantly higher (p < 0.01) in the isolates from STSS pa tients than in the other isolates whereas biotype 3 appeared to be more fre quently expressed in the non STSS associated isolates. Release of pyrogenic (erythrogenic) exotoxin A (SPE A). This superantigen ic toxin was detectably released in culture supernatants (ELISA test: 10 pg/ml detection threshold) by 25.5% of the strains (54/212). Among the 74 isolates from STSS patients, the percentage of SPE A producing strains was 40.5% (30174) as compared to the 17.4 % of SPE A producing strains from patients without associated STSS (241138) (p < 0.01). At the quantitative level, the amounts of toxin released varied over a wide concentration range (1 to 17000 ng/ml). Thirteen strains (6 from patients with STSS) produced 5000 to 17000ng of SPE Nml, 14 strains (9 from patients with STSS) produced 500 to 5000 nglml and 19 strains (17 from patients with STSS) produced 100 to 500 ng/ml respectively.
426
H. Miiller-Alouf et al. % 40
SEROTYPE 30
20
10
% 60
BIOTYPE
50
40
30
20
10
o Fig.1. Frequencies percentage of S. pyogenes serotypes M and T (upper panel) and biotypes (lower panel) among 212 clinical isolates from patients with severe invasive group A streptococcal infections including 74 cases of associated streptococcal toxic shock syndrome (STSS).
Invasive S. pyogenes strains from severe infections
427
Relation between SPE A production and M serotypes of the clinical isolates. Of the 54 isolates which released detectable amounts of SPE A, 36 strains (67%) belonged to serotypes M1 (28 strains) or M3 (8 strains). Twenty two were from patients with STSS. Among the other SPE A producing isolates (18/ 54), 10 belonged to other serotypes and 8 were untypeable. On the basis of these results, 68% of the isolates of serotypes M1 or M3 produced SPE A (36/ 53) whereas only 11.3% of the isolates from other serotypes or untypeable (18/159) released this toxin. In relation to strain biotypes, most SPE A expressing isolates were found associated to biotypes 1 and 5 (data not shown). Release of streptolysin 0 (SLO), streptokinase (SK) and hyaluronidase (HA). All isolates released SLO in amounts ranging from 4 to 256 hemolytic units (HU)/ml. SLO titers varied from one strain to another regardless the or igin of the isolates, the clinical type of disease and its outcome. Streptokinase release was detectable for 71 % of the isolates. The amounts of SK relased widely varied (3 to 1000 units/ml) from strain to strain. As for SLO release, no clear relationship was found between SK production and titer and the type of disease or the clinical origin of the strains. Hyaluronidase was detectable in a low percentage (ca. 10%) of all isolates tested regardless of strain origin and type of disease. Release of cysteine proteinase. Most isolates released this enzyme. The per centage of proteinase positive strains evaluated by the formation of a proteo lytic zone on skim milk agar plates ranged from 84 to 86% regardless of strain serotypes and clinical origin except for most isolates from patients with cellu litislfasciitis for which apparently a higher percentage (93%) of proteinase positive strains was observed (statistically not significant). The diameters of the lytic zones ranged from 10.2 ± 5 mm to 12 ± 3.5 mm. Mitogenic activity. All culture supernatants of the 212 isolates investigated showed moderate to high lymphocyte transforming activity. Index values var ied from strain to strain within a range of 14.5 ± 4.6 to 50.3 ± 5.8 regardless of the origin of the isolates, the type of disease and its outcome (survival ver sus death). Interestingly, the supernatants of many strains which did not re lease detectable amounts of SPE A or SPE C (immunological testing for SPE C undertaken only on a limited number of the isolates, data not shown) showed SI values higher than those found for the strains producing these toxins sug gesting that other mitogenic factor (s) are produced (see Discussion).
Discussion The history of severe invasive group A streptococcal (GAS) infections recent ly reviewed by Stevens (37) and Bronze and Dale (4) has shown changing pat terns of these infections in modern times. Since the mid 1980s, a worldwide increase of severe GAS skin/deep tissue infections, bacteremia and streptococ cal toxic shock syndrome (STSS) with high mortality rates has been reported
428
H. Muller-Alouf et al.
(4, 13, 24, 37). The severity of these invasive infections has been attributed besides host related factors (particularly the immune status) to an increase in the pathogenicity and virulence of the invading S. pyogenes strains (see Intro duction). In the present work, we investigated the phenotypic expression of some cell bound and extracellular pathogenicity and virulence factors of 212 clinical strains isolated in France (1990-1994) from patients with severe GAS infec tions among them 74 cases (35%) of STSS. Significant differences with respect to age were observed between those pa tients with STSS and the others. The mean age of the former was 36 years, whereas that of the latter was clearly higher (56 years) indicating that among other factors, patient age is an important parameter in STSS versus other in vasive GAS infections such as skin/soft tissue infections or bacteremia as also shown by other authors (4, 37). The analysis of the isolates investigated in this work involved M protein serotyping (in parallel with T protein serotyping and serum opacity factor determination in certain cases as complementary ap proaches for M serotyping), biotype determination, the assay in culture super natants of the pyrogenic (erythrogenic) exotoxin A (SPE A), streptolysin 0, streptokinase, hyaluronidase, transforming activity on human blood lympho cytes and the evaluation of the release of cysteine proteinase by GAS colonies on agar plates. The data were particularly examined with respect to STSS and non STSS isolates to find out whether significant differences in these virulence or pathogenicity markers could be identified. M protein serotyping has been widely used for more than 50 years as a method for studying the epidemiology of S. pyogenes, particularly in relation with infection. On this basis, a significant predominance of Ml and M3 sero types and to a lesser extent M12 and MIS has been observed in the past ten years among the strains isolated from patients with STSS and other severe GAS infections in North America, Europe and elsewhere (see Introduction). However, other M serotypes and non M serotypeable strains have been iso lated in such cases (29,40). In the present work, 25% of the isolates (53/212) expressed serotypes Ml and to a lesser extent M3, whereas 16% were untypeable (34/212) and 59% (125/212) belonged to 15 other serotypes without obvious prevalence of any of them reflecting probably the non epidemic background of the isolates. Howev er, among the 74 isolates from patients with confirmed STSS, 42% (31/74) be longed to serotype Ml and some of them to M3. These data, other reports (29, 40) as well as the isolation of many strains expressing Ml or M3 serotypes from asymptomatic carriers or during uncomplicated GAS infections (6, 16) suggest that the emergence of severe GAS infections is the result of strain-specific viru lence factors rather than virulence broadly defined by an individual serotype (4). Furthermore, the immune status of the host is certainly critical with regard to infection by strains expressing M 1, M3 or other serotypes. As shown by Holm et al. (20) the antibody titers against Ml antigen were found lower in fa tal cases than in patients with bacteremia or in uncomplicated infections.
Invasive S. pyogenes strains from severe infections
429
SPE A is historically known as primarily responsible for the skin rash in scarlet fever (1, 23, 24). The initial observations of Cone et al. (8) and Ste vens et al. (38) at the end of the 1980s which showed a close association between SPE A expression by the strains isolated from patients with severe GAS infections including STSS led to the contention, that this toxin plays a major role in the pathogenesis of these diseases and may be linked to their apparent resurgence (16,23,29,31,38,40). An accepted concept points to the direct implication of SPE A and probably SPE C (4) and other newly dis covered superantigens (11, 12,21,31,33) in severe GAS infections through the massive release of cytokines and other inflammatory factors by the immune system cells of the host (18). These effectors which are thought to mediate shock, tissue damage and fever are released as a consequence of the super antigen induced MHC class II dependant polyclonal proliferation of those T lymphocyte clones bearing appropriate motifs in the variable region of the ~ chain of T cell receptor (1, 26, 31). The concept is supported by the (i) in vitro release of various cytokines (interleukins-1, 2, 6, 8, 10, tumor ne crosis factors a, and ~, interferon y) by superantigen stimulated human pe ripheral blood mononuclear cells (14, 20, 27, 28, 31, 39) (ii) increased levels of cytokines in patients sera during STSS and other streptococcal diseases (14, 30), (iii) marked in vivo reduction of cognate V~-bearing T lymphocytes (26, 41), (iv) development of various pathophysiological disturbances, cytokine induction, organ impairment and death in experimental animal models (mouse, rabbit, baboon), challenged with SPE A or SPE C expressing strains (4,35,39). In the present work, SPE A was detectably produced by about 26% (541 212) of the isolates. However, SPE A release (40.5%) was higher for the iso lates from STSS patients. Moreover, this toxin was predominantly released by 67% of the strains of serotypes M1 or M3. Similar studies in the past few years on clinical isolates from North America (29) and Europe (7, 20, 34) showed great variations (30 to 80%) in the frequency of spe A gene and (or) of its phenotypic expression by the strains. This frequency was clearly higher among the isolates from patients with scarlet fever and other severe infections particularly STSS than in isolates from patients with non invasive infections or from healthy carriers (6, 29). On the other hand, as reported here, in a previous work (34) and by others (6,29,40), many isolates from severely diseased patients lacked spe A gene or did not produce the toxin suggesting that other bacterial products may be critically involved in patho genesis (6, 16). This is supported by the discovery of a number of novel GAS superantigens as mentioned above. The finding in the present work of po tent mitogenic activity in the supernantants of those strains with did not de tectably produce SPE A or SPE C (data not shown for the latter) supports this view. Streptococcal pyrogenic exotoxin B (SPE B) which is identical to or a vari ant of streptococcal cysteine proteinase SCP (5, 10, 15) has been shown to be released alone or in association with SPE A by clinical isolates during severe
430
H. Miiller-Alouf et al.
GAS infections (20, 24, 31). This chromosomally encoded enzymatic toxin appears to be associated with both shock and organ disorders in STSS patients (6, 40) as also supported by experiments in animal models (35). Moreover, SCP cleaves human fibronectin, vitronectin and interleukin (IL-) IB precursor with the generation of the pro inflammatory bioactive IL-IB (6). In the present work, ca. 85% of the strains released SCP regardless of strain serotypes and clinical orgin. Talkington et al. (40), Musser et al. (29) and Hauser et al. (16) similarly showed that SCP was produced by 63%, 51.8% and 59% of the strains respectively. Hyaluronidase (HA), streptolysin 0 (SLO) and streptokinase (SK) released in culture medium were also evaluated. The former was released by only 10% of the strains. Its possible involvement in pathogenicity has not been so far es tablished to our knowledge. SK was released in varying amounts (3 to 1000 U/ml) and no relationship between enzyme titer and the type of disease or clin ical origin of the strains was found. These results are in agreement with those reported by Chaussee et al. (6) who did not observe association between these two exoproteins and disease. All strains released SLO in amounts ranging from 4 to 256 HU/ml. This cy tolysin is expressed by virtually all isolates so far tested (2), in accordance with the permanent presence of slo gene in S. pyogenes strains (2). No apparent rel evance was found between the amounts of toxin released and patients diseas es as also found by Holm et al. (20) for swedish bacteremia strains. Whether this toxin plays a role in severe GAS infections remains to be established. However, several lines of evidence suggest its involvement in the pathogenesis of these diseases (14,40) and on account of the remarkable in vitro membrane damaging and disrupting effects on eukaryotic cells, particularly leucocytes (2). On the other hand, SLO has been shown by Hackett and Stevens (14) to behave as a potent in vitro inducer of the proinflammatory cytokines TNF a and IL-IB, and to increase synergistically in association with ETA the release of the latter cytokine by monocytes. SLO also acts jointly with SCP and SPE A to enhance tissue injury (35) suggesting that these proteins may function as important virulence and pathogenicity factors, particularly as shock media tors in severe GAS infections. Acknowledgement. We wish to thank P. Desreumaux for helpful discussion and assi stance with the statistical methods.
References 1. Alouf,]. E., H. Knoll, and W. Kohler: The family of mitogenic, shock-inducing and superantigenic toxins from staphylococci and streptococci. In: Sourcebook of Bac terial Protein Toxins. (J. E. Alouf and ]. H. Freer, eds.), pp.367-414. Academic Press, Ltd., London (1982) 2. Alouf, J. E. and C. Geoffroy: Production, purification and assay of streptolysin O. Meth. Enzymol. 165 (1988) 52-59
Invasive S. pyogenes strains from severe infections
431
3. Bouvet, A., P. Geslin, P. Kriz-Kuzemenska, V. Blanc, C. Devine, and F. Grimont: Restricted association between biotypes and serotypes within group A streptococ ci. J. Clin. Microbiol. 32 (1994) 312-317 4. Bronze, M. S. and]. D. Dale: The reemergence of serious group A streptococcal in fections and acute rheumatic fever. Am. J. Med. Sci. 311 (1996) 41-54 5. Chaussee, M. S., D. Gerlach, C. -E. Yu, and ].]. Ferretti: Inactivation of the strep tococcal erythrogenic toxin B gene (spe B) in Streptococcus pyogenes. Infect. Im mun.61 (1993) 3719-3723 6. Chaussee, M. S., J. Liu, D. L. Stevens, and].]. Ferretti: Genetic and phenotypic di versity among isolates of Streptococcus pyogenes from invasive infections. J. In fect. Dis. 173 (1996) 901-908 7. Colman, G., A. Tanna, A. Efstratiou, and E. T. Gawozerwska: The serotypes of Streptococcus pyogenes present in Britain during 1980-1990 and their association with disease. J. Med. Microbiol. 39 (1993) 165-178 8. Cone, L. A., D. R. Voodard, P. M. Schlievert, and G. S. Tomory: Clinical and bac teriologic observation of a toxic shock-like syndrome due to Streptococcus pyo genes. N. Engl. J. Med. 317 (1987) 146-149 9. Fischetti, V. A.: Streptococcal M protein: molecular design and biological behavior. C1in. Microbiol. Rev. 2 (1989) 285-314 10. Gerlach, D., H. Knoll, W Kohler,]. H. Ozegowski, and V. Hiibalova: Isolation and characterization of erythrogenic toxins. V. Identity of erythrogenic toxin type B and streptococcal proteinase precursor. Zbl. Bakt. 255 (1983) 221-233 11. Gerlach, D., H. Alouf, L. Moravek, M. Pavlick, and W Kohler: The characteriza tion of two new low molecular weight proteins (LMPs) from Streptococcus pyo genes. Zbl. Bakt. 277 (1992) 1-9 12. Geoffroy, c., H. Muller-Alouf, E. Champagne, ]. M. Cavaillon, and ]. E. Alouf: Identification of a new extracellular mitogen from group A streptococci. Zbl. Bakt. Suppl. 24 (1994) 90-91 13. Geslin, P., P. Kritz-Kuzmenska, A. Fremaux, H. Hav[ickova, A. Dublanchet, and A. Bouvet: Septicaemia caused by phenotypic variants of group A streptococci. Res. Microbiol. 147 (1996) 273-277 14. Hackett, S. P. and D. L. Stevens: Streptococcal toxic shock syndrome synthesis of tumor necrosis factor and IL-1 by monocytes stimulated with pyrogenic exotoxin A and streptolysin O. J. Infect. Dis. 165 (1992) 879-885 15. Hauser, A. R. and P. M. Schlievert, P. M.: Nucleotide sequence of the streptococcal pyrogenic exotoxin type B gene and relationship between the toxin and the strep tococcal proteinase precursor. J. Bacteriol. 172 (1990) 4536-4542 16. Hauser, A. R., D. L. Stevens, E. L. Kaplan, and P. M. Schlievert: Molecular analy sis of pyrogenic exotoxins from Streptococcus pyogenes isolates associated with toxic shock-like syndrome. J. C1in. Microbiol. 29 (1991) 1562-1567 17. Haynes, W L. and]. R. Tagg: A simple plate assay for detection of group A strep tococcus proteinase. J. Microbiol. Meth. 4 (1985) 25-31 18. Hensler, T., M. Koller, C. Geoffroy,]. E. Alouf, and W Konig: Staphylococcus au reus toxic shock syndrome toxin 1 and Streptococcus pyogenes erythrogenic tox in A modulate inflammatory mediator release from human neutrophils. Infect. Im mun. 61 (1993) 1055-1061 19. Hoge, C. W, B. Schwarz, D. F. Talkington, R. F. Breiman, E. M. MacNeill, and].]. Englender: The changing epidemiology of invasive group A streptococcal infec tions and the emergence of streptococcal toxic chock-like syndrome. A retrospec tive population-based study. J. Amer. Med. Ass. 269 (1993) 384-389
432
H. Miiller-Alouf et al.
20. Holm, S. E., A. Norrby, A.-M. Bergholm, and M. Norgren: Aspects of pathogene sis in serious group A streptococcal infections in Sweden 1988-1989. J. Infect. Dis.
166 (1992) 31-37
21. Iwasaki, M., H. Igarashi, Y. Hinuma, and T. Yutsudo: Cloning, characterization and overexpression of a Streptococcus pyogenes gene encoding a new type of mito genic factor. FEBS Lett. 331 (1993) 187-192 22. Knoll, H., j. Sramek, K. Vrbova, D. Gerlach, W. Reichardt, and W. Kohler: Scarlet fever and types of erythrogenic toxins produced by the infecting streptococcal strains. Zbl. Bakt. 276 (1991) 94-106 23. Kohler, w., D. Gerlach, and H. Knoll: Streptococcal outbreaks and erythrogenic toxin type A. Zbl. Bakt. 266 (1987) 104-115 24. Kohler, w.: Streptococcal toxic shock syndrome. Zbl. Bakt. 272 (1990) 527-564 25. Malke, H. and ].j. Ferretti: Streptokinase: cloning, expression and excretion by Escherichia coli. Proc. Natl. Acad. Sci. USA 81 (1984) 3557-3561 26. Michie, c., A. Scott, J. Cheesborough, P. Berverley, and G. Pasvol: Streptococcal toxic shock-like syndrome: evidence of superantigen activity and its effects on T lymphocyte subsets in vivo. Clin. Exp. Immunol. 98 (1994) 140-144 27. Muller-Alouf, H., ]. E. Alouf, D. Gerlach, J.-H. Ozegowski, C. Fitting, and J.-M. Cavaillon: Comparative study of cytokine release by human peripheral blood mononuclear cells stimulated with Streptococcus pyogenes superantigenic erythro genic toxins, heat-killed streptococci, and lipopolysaccharide. Infect. Immun. 62
(1994) 4915-4921
28. Muller-Alouf, H., ]. E. Alouf, D. Gerlach, ].-H. Ozegowski, C. Fitting, and ].-M. Cavaillon: Human pro- and anti-inflammatory cytokine pattern induced by Strep tococcus pyogenes erythrogenic (pyrogenic) exotoxins A and C superantigens. In fect. Immun. 64 (1996) 1450-1453 29. Musser, ]. M., A. R. Hauser, M. H. Kim, P. M. Schlievert, K. Nelson, and R. K. Se lander: Streptococcus pyogenes causing toxic shock-like syndrome and other inva sive disease: clonal diversity and pyrogenic exotoxin expression. Proc. Natl. Acad. Sci. USA 88 (1991) 2668-2672 30. Nadal, D., R. P. Lauener, C. P. Braegger, A. Kaufhold, B. Simma, R. Lutticken, and A. Seger: T cell activation and cytokine release in streptococcal toxic shock-like syndrome. J. Pediatr. 122 (1993) 727-729 31. Norrby- Teglund, A., M. Norgren, S. E. Holm, U. Andersson, and ]. Andersson: Similar cytokine induction profiles of a novel streptococcal exotoxin, MF, and py rogenic exotoxins A and B. Infect. Immun. 62 (1994) 3731-3738 32. Ozegowski, ].-H., E. Gunther, and W. Reichardt: Purification and characteriza tion of hyaluronidase from Streptococcus agalactiae. Zbl. Bakt. 280 (1994) 497-
506
33. Reda, K. B., V. Kapur,]. A. Mollick, j. G. Lamphear, J. M. Musser, and R. R. Rich: Molecular characterization and phylogenetic distribution of the streptococcal superantigen gene (ssa) from Streptococcus pyogenes. Infect. Immun. 62 (1994)
1867-1874
34. Reichardt, w., H. Muller-Alouf,]. E. Alouf, and W. Kohler: Erythrogenic toxins A, Band C: occurrence of the genes and exotoxin formation from clinical Streptococ cus pyogenes strains associated with streptococcal toxic shock like syndrome. FEMS Microbiol. Lett. 100 (1992) 313-322 35. Schanley, T. P., D. Schrier, V. Kapur, M. Kehoe,]. M. Musser, and P. E. Ward: Strep tococcal cysteine protease augments lung injury induced by products of group A streptococci. Infect. Immun. 64 (1996) 870-874
Invasive S. pyogenes strains from severe infections
433
36. Schwartz, B., R. Facklam, and R. F. Breiman: Changing epidemiology of group A streptococcal infection in the USA. The Lancet 336 (1990) 1167-1171 37. Stevens, D. L.: Invasive group A streptococcus infections. Clin. Infect. Dis. 14 (1992) 2-13 38. Stevens, D. L., M. H. Tanner, ]. Winship, R. Swarts, K. M. Ries, P. M. Schlievert, and E. Kaplan: Severe group A streptococcal infections associated with toxic shock-like syndrome and scarlet fever toxin A. N. Engl. J. Med. 321 (1989) 1-7 39. Stevens, D. L., A. E. Bryant, S. P. Hackett, A. Chang, G. Peer, S. Kosanke, T. Emer son, and L. Hinshaw: Group A streptococcal bacteremia: the role of tumor necro sis factor in shock and organ failure. J. Infect. Dis. 173 (1996) 619-626 40. Talkington, D. F., B. Schwartz, C. M. Black, ]. K. Todd, ]. Elliott, R. F. Breiman, and R. R. Facklam: Association of phenotypic and genotypic characteristics of in vasive Streptococcus pyogenes isolated with clinical components of streptococcal toxic shock syndrome. Infect. Immun. 61 (1993) 3369-3374 41. Watanabe-Ohnishi, R., D. E. Low, A. McGeer, D. L. Stevens, P. M. Schlievert, D. Newton, B. Schwartz, B. Kreiswirth, Ontario Streptococcal study project, and M. Kotb: Selective depletion of V~-bearing T cells in patients with severe invasive group A streptococcal infections and streptococcal toxic shock syndrome. J. Infect. Dis. 171 (1995) 74-84 42. Willoughby, R. and R. N. Greenberg: Toxic shock syndrome and streptococcal py rogenic exotoxins. Ann. Intern. Med. 98 (1983) 559 43. Working Group on Severe Streptococcal Infections: Defining the group A strepto coccal toxic shock syndrome. J. Amer. Med. Ass. 269 (1993) 390-391 Corresponding author: Dr. Heide Muller-Alouf, Centre d'Immunologie et de Biologie Parasitaires, Institut Pasteur, 1, Rue du Professeur A. Calmette, F-59019 Lille, France, Tel.: 33320877965, Fax: 33320877888
29
Zbl. Bakt. 28613
Zentralblatt fur
BakterioiOSie
Streptococcal Pyrogenic Exotoxin A, Streptolysin 0, Exoenzymes, Serotype and Biotype Profiles of Streptococcus pyogenes Isolates from Patients with Toxic Shock Syndrome and other Severe Infections Heide Miiller-Alouf 1, 2, Christiane Geoffroy 1, P. Geslin 3, Anne Bouvet\ Annie Felten s, Elisabeth Giinther6 , J.-H. Ozegowski 6, and J. E. AlouP,2 1, 2
3 4
5 6
Unite des Toxines Microbiennes (URA 1858, Centre National de la Recherche Scientifique), Institut Pasteur, F-75724 Paris Cedex 15 and Institut Pasteur de Lille, F-59019 Lille Cedex, France Centre Hospitalier Intercommunal, F-94010 Crete ii, France Universite Paris VI, Hopital Hotel-Dieu, F-75181 Paris 04, France Hopital Saint Louis, F-75010 Paris, France Institut fiir experimentelle Mikrobiologie, Friedrich Schiller Universitat, Jena, Germany
Received February 5, 1997 . Accepted February 13, 1997
Summary The determination of protein M and T serotypes, biotypes and pyrogenic (erythrogen ic) exotoxin A (SPE A), streptolysin 0 (SLO), streptokinase (SK), hyaluronidase (HA) and cysteine proteinase release by 212 S. pyogenes isolates from patients with severe invasive group A streptococcal (GAS) infections, among them 74 cases of streptococ cal toxic shock syndrome (STSS) has been investigated. Ml or M3 serotypes were ex pressed by 25% of the isolates (53/212), whereas 59% (125/212) belonged to 15 oth er different serotypes and 16% (34/212) were untypeable. Of the 74 isolates from STSS patients, 42% (31/74) expressed Ml and to a lesser extent M3 serotypes versus 19% of the non STSS isolates (26/138) . Among the ten different biotypes known, biotypes 1 and 3 were prevalent, particularly the former in the case of STSS isolates. SPE A was detectably produced by about 25% (54/212) of the strains. However, as high as 40.5% of the STSS isolates (30/74) versus 17.4% of non STSS isolates (24/138) released SPE A. Moreover, 67% of the SPE A producing strains were of serotype Ml or M3. SK and HA were released by 71 % and 10% of the isolates respectively. All
422
H. Miiller-Alouf et al.
strains released SLO (4 to 256 HU/ml) and 85% cysteine proteinase. No relationship between toxin or enzyme titer and the type of disease or clinical origin of the strains was found. Culture supernatants of all isolates showed moderate to high lymphocyte trans forming activity with index values ranging from 14.5 to 50.3 including those strains which did not release detectable amounts of SPE A suggesting that SPE C and other mitogenic factor(s) are released by the isolates investigated.
Introduction After a four-decade decline of the prevalence of severe invasive group A strep tococcal (GAS) diseases in developed countries, a worldwide increase since the mid 1980s of the incidence of these infections (bacteremia, myositis, cellulitis, fasciitis, ... ) with mortality rates ranging from 20 to 60% has been reported, particularly in North America and Europe (4, 19,29). These severe GAS dis eases also include the clinical entity called streptococcal toxic shock syndrome (STSS) suspected in 1983 (42) and widely described and documented since 1987 (19, 23, 36, 37, 38, 43). This syndrome characterized by hypotension, shock, fever, multiorgan failure, skin desquamation and other biological per turbations (4, 43) corresponds very likely to the cases of scarlatina maligna described in older literature (20, 23) during outbreaks of scarlet fever. The bacterial effectors which are believed to be causally involved in the se verity of GAS infections are certain serotypes of M protein, a major cell sur face virulence factor (9), streptococcal pyrogenic (erythrogenic) exotoxins A and probably C (SPE A, SPE C) which belong to the family of the bacterial superantigens (1). In this respect, many recent reports concerning the S.pyo genes strains currently isolated from patients with severe invasive GAS infec tions and particularly STSS pointed to the predominance of strains expressing Ml and M3 serotypes (as well as M12 and M18 in certain reports), concom itantly with the production of SPE A (4, 6, 7, 15,20, 24, 29, 34, 36, 37, 38, 40). Other extracellular toxins such as streptolysin 0 (SLO), a membrane damaging sulfhydryl-activated toxin (2), and cysteine proteinase (interleukin IP convertase) which has been shown on immunochemical (10) and genetic grounds (5, 15) to be identical to or a variant of pyrogenic (erythrogenic) exo toxin B (5, 10,35) are also thought to contribute to the pathogenesis of GAS diseases (16,20). These considerations and recent work in our laboratory (27, 28) and by others (14) on cytokine induction by streptococcal toxins in relation to STSS and other diseases, led us to analyze 212 S. pyogenes strains isolated in a num ber of hospitals from various areas of France between 1990 and 1994 from patients with STSS and/or other invasive GAS infections. The strains provid ed to the laboratory of one of us (J. E. A.) were tested with the cooperation of the other authors for M and/or T serotypes, the recently defined streptococ cal biotypes (3), and for the release in culture medium of SPE A, SLO, strep-
Invasive S. pyogenes strains from severe infections
423
tokinase, hyaluronidase and cysteine proteinase. The transforming activity of culture supernatants (mitogenic effect) on human lymphocytes has been also undertaken.
Materials and Methods S. pyogenes clinical isolates and patients. The strains investigated were isolated from 212 patients (108 males and 114 females, mean age 47 years, range 3 months to 94 years) who all met the criteria for invasive GAS infections established by the Working Group on Severe Streptococcal Infections (43): (i) STSS (ii), other infections defined by isolation of S. pyogenes from normal sterile sites not meeting criteria for STSS (bacte remia with no identified focus, focal infections with or without bacteremia particular ly skin and deep-soft tissue infection), (iii) scarlet fever. On the basis of this classification, there were among the 212 patients, 74 confirmed cases (35%) of STSS and 138 cases (65%) with other severe (invasive) GAS infections. The patients with STSS (28 males and 46 females, mean age 36 years, range 5 months to 94 years) presented the following associated clinical features: bacteremia without identified focus (49 patients), bacteremia associated to cellulitis or necrotizing fasciitis (15 patients), necrotizing fasciitis (4 patients), postpartum infection (6 patients). The other severe GAS infections without associated STSS (64 males and 74 females, mean age 56 years, range 5 months to 94 years) included severe scarlet fever (9 cases), bac teremia without identified focus (40 cases), bacteremia with associated pneumonia (14 cases), deep soft-tissue infections (29 cases), or erysipelas (21 cases), infected burn wounds (14 cases), varicella gangrenosa (11 cases). The majority of the strains isolat ed from patients with bacteremia were provided by P. Geslin (coordinator of the mul ticenter study on bacteremia). An overall fatality rate of 13.7% was recorded among the 212 patients (29/212). Seventeen deaths occurred among the 74 patients with STSS (23% fatality rate). Bacterial culture. The S. pyogenes isolates were purified by streaking on 5% sheep blood Columbia agar plates and then cultured on Todd-Hewitt broth (Difco Labora tories) at 37 °C to the stationary phase of growth. The cultures were centrifuged and bacterial pellets separated for typing purposes. The supernatant fluids were sterilized by filtration through 22-[,Im-pore size Millipore filters (Millex-GV, Millipore). The fil trates were stored at -20 °C for exotoxin, exoenzyme, and mitogenic activity assay or detection. Serotyping of M and T antigens. M typing was determined by double immunodif fusion on agar gel using Lancefield extracts of streptococcal cells and specific antisera prepared in the laboratory of one us (E. G. National Reference Center for Streptococ ci, Jena). T-typing was done by agglutination of trypsinized bacterial suspensions. Biotyping. This new typing method of S. pyogenes strains of epidemiological inter est based on enzymatic and carbohydrate fermentation reactions was determinated by one of us (A. B.) as recently described (3). The testing for the ten biotypes character ized to date was carried out with the rapid ID 32 STREP identification system (Bio Merieux, La-Balme-Ies-Grottes, France). Immunoassay of streptococcal pyrogenic exotoxin A (SPE A). The supernatant fluids of the strains cultured on Todd-Hewitt medium were 100-fold concentrated by ethanol precipitation at -20°C as described (22). The precipitate was dissolved in
424
H. Miiller-Alouf et al.
acetate-buffered saline, pH 4.0. SPE A was determined quantitatively in the concen trated material by a sensitive and specific enzyme linked immunosorbent assay (ELISA) using two mono specific polyclonal anti SPE A (goat and rabbit) hyperimmune serum as described (22). The detection limit for SPE A is ca.10 pg/ml. Streptolysin 0 (SLO) assay. This membrane damaging toxin was assayed by the de termination of its hemolytic activity on sheep red blood cells expressed as hemolytic units (HU) per ml of culture supernatants after activation with 20 mM cysteine as de scribed (2). Streptokinase (SK) assay. The quantitation of this enzyme in culture supernatants was determined by the skim milk/plasminogen agarose plate technique (25). The area of casein lysis into the agarose medium around the wells filled with test samples and a standard purified SK solution can be correlated to the amount of enzyme released in bacterial cultures. Hyaluronidase (HA) assay. Enzyme determination was undertaken by the turbidi metric method previously described (32). Serial dilutions of culture supernatants (500 !tl) in 500 !tl of 100 mM acetate buffer pH 6.0 were incubated at 37°C for 30 min with 500 !tl of 0.2 mg/ml hyaluronic acid in 10 mM acetate buffer purified from Strep tococcus zooepidemicus. The reaction was stopped by adding 3 ml of 2.5% (w/v) ce tyltrimethylammonium bromide in 0.5 N sodium hydroxide. After standing 20 min at room temperature, the turbidity was photometrically measured at 600 nm. One HA unit is that amount which splits 50 ng of hyaluronic acid within 30 min. Proteinase evaluation. The plate assay method of Haynes and Tagg (17) based upon the digestion of skim milk (1.5% w/v) contained within Columbia agar base medium was used for detection of proteinase production by individual surface grown colonies of the strains. The diameter of the proteolysis zones was measured after 24 h of aero bic incubation of the plates at 3rc. Protease activity was calculated as the ratio of the diameter of the proteolysis zone to the diameter of the corresponding colony. Mitogenic activity of culture supernatants. The lymphocyte blast transforming ac tivity of the supernatants were determined on human peripheral blood mononuclear cells isolated and cultivated on RPMI 1640 medium (Gibco) as described (27). The evaluation of lymphocyte transformation after cell pulse with tritiated thymidine (Amersham) was undertaken as described (22). The results were expressed as stimula tion index based on the ratio of cpm in lymphocyte culture with bacterial supernatant! cpm in lymphocyte culture with sterile Todd-Hewitt broth. Statistical analysis. The Student's t test was used to determine the significance of dif ferences between STSS associated isolates and the other isolates from GAS infections. A P value of 0.05 or less was considered statistically significant.
Results M and T serotypes of the clinical isolates. Of the 212 isolates investigated, 34 (16 %) were untypeable, whereas the other 178 isolates (84 %) belonged to 17 different M and (or) T serotypes (Table 1). Despite that all strains were iso lated from severe invasive GAS infections, only 53 isolates (25%) expressed either serotype M1 (40 isolates) or M3 (13 isolates). Interestingly, 42% of the STSS strains (31/74) belonged to the serotypes M1 or M3 whereas only 19% of the strains (26/138) isolated from those patients without associated STSS
Invasive S. pyogenes strains from severe infections
425
Table 1. Mff serotype distribution among S. pyogenes isolates from patients with invasive GAS infections with or without associated STSS Serotype
All isolates (%) n =212
STSS isolates (%) n = 74
Other isolates (%) n = 138
M1 M2 M3rr3 M4rr4 T5 M9 Mll M12rr12 M13rr13 T25 R28rr28 B3264 T33 M41 M48 T56 T78 Untypeable
40 3 13 8 5 1 16 22 12 6 3 22 4 5 1 15 2 34
20 (0.35) 0 (0.0) 5 (6.7) 4 (5.4) 5 (6 .7) 0 (0 .0) 4 (5.4) 4 (5.4) 3 (4.0) 2 (2.7) 1 (1.3) 3 (4.0) 1 (1.3) 1 (1.3) 0 (0.0) 5 (6.7) 1 (1.3) 9 (12.2)
14 3 8 4 0 1 12 18 9 4 2 19 3 4 1 10 1 25
(18.9) 0.4) (6 .1) (3.7) (2.3) (0.5) (7.5) (10.4) (5.6) (2.8) (1.4) (10.8) (1.9) (2.4) (0.5) (7.1) (0.9) (16.0)
(10.4) (2.2) (5.8) (2.9) (0.0) (0.7) (8.7) (13.0) (6.5) (2.9) (1.2) (13.8) (2.2) (2.9) (0.7) (7.3) (0.7) (18.0)
expressed these two markers. As shown in Fig. 1, the percentage of strains ex pressing the serotype Ml was significantly much higher (p < 0.01) for STSS strains whereas the percentage of strains expressing the serotype M3 was practically the same for both groups of isolates. Biotypes of the clinical isolates. Biotypes 1 and 3 and to a lesser extent bio type 5 were the most preponderant among the strains tested (Fig. 1). Biotype 1 expression was significantly higher (p < 0.01) in the isolates from STSS pa tients than in the other isolates whereas biotype 3 appeared to be more fre quently expressed in the non STSS associated isolates. Release of pyrogenic (erythrogenic) exotoxin A (SPE A). This superantigen ic toxin was detectably released in culture supernatants (ELISA test: 10 pg/ml detection threshold) by 25.5% of the strains (54/212). Among the 74 isolates from STSS patients, the percentage of SPE A producing strains was 40.5% (30174) as compared to the 17.4 % of SPE A producing strains from patients without associated STSS (241138) (p < 0.01). At the quantitative level, the amounts of toxin released varied over a wide concentration range (1 to 17000 ng/ml). Thirteen strains (6 from patients with STSS) produced 5000 to 17000ng of SPE Nml, 14 strains (9 from patients with STSS) produced 500 to 5000 nglml and 19 strains (17 from patients with STSS) produced 100 to 500 ng/ml respectively.
426
H. Miiller-Alouf et al. % 40
SEROTYPE 30
20
10
% 60
BIOTYPE
50
40
30
20
10
o Fig.1. Frequencies percentage of S. pyogenes serotypes M and T (upper panel) and biotypes (lower panel) among 212 clinical isolates from patients with severe invasive group A streptococcal infections including 74 cases of associated streptococcal toxic shock syndrome (STSS).
Invasive S. pyogenes strains from severe infections
427
Relation between SPE A production and M serotypes of the clinical isolates. Of the 54 isolates which released detectable amounts of SPE A, 36 strains (67%) belonged to serotypes M1 (28 strains) or M3 (8 strains). Twenty two were from patients with STSS. Among the other SPE A producing isolates (18/ 54), 10 belonged to other serotypes and 8 were untypeable. On the basis of these results, 68% of the isolates of serotypes M1 or M3 produced SPE A (36/ 53) whereas only 11.3% of the isolates from other serotypes or untypeable (18/159) released this toxin. In relation to strain biotypes, most SPE A expressing isolates were found associated to biotypes 1 and 5 (data not shown). Release of streptolysin 0 (SLO), streptokinase (SK) and hyaluronidase (HA). All isolates released SLO in amounts ranging from 4 to 256 hemolytic units (HU)/ml. SLO titers varied from one strain to another regardless the or igin of the isolates, the clinical type of disease and its outcome. Streptokinase release was detectable for 71 % of the isolates. The amounts of SK relased widely varied (3 to 1000 units/ml) from strain to strain. As for SLO release, no clear relationship was found between SK production and titer and the type of disease or the clinical origin of the strains. Hyaluronidase was detectable in a low percentage (ca. 10%) of all isolates tested regardless of strain origin and type of disease. Release of cysteine proteinase. Most isolates released this enzyme. The per centage of proteinase positive strains evaluated by the formation of a proteo lytic zone on skim milk agar plates ranged from 84 to 86% regardless of strain serotypes and clinical origin except for most isolates from patients with cellu litislfasciitis for which apparently a higher percentage (93%) of proteinase positive strains was observed (statistically not significant). The diameters of the lytic zones ranged from 10.2 ± 5 mm to 12 ± 3.5 mm. Mitogenic activity. All culture supernatants of the 212 isolates investigated showed moderate to high lymphocyte transforming activity. Index values var ied from strain to strain within a range of 14.5 ± 4.6 to 50.3 ± 5.8 regardless of the origin of the isolates, the type of disease and its outcome (survival ver sus death). Interestingly, the supernatants of many strains which did not re lease detectable amounts of SPE A or SPE C (immunological testing for SPE C undertaken only on a limited number of the isolates, data not shown) showed SI values higher than those found for the strains producing these toxins sug gesting that other mitogenic factor (s) are produced (see Discussion).
Discussion The history of severe invasive group A streptococcal (GAS) infections recent ly reviewed by Stevens (37) and Bronze and Dale (4) has shown changing pat terns of these infections in modern times. Since the mid 1980s, a worldwide increase of severe GAS skin/deep tissue infections, bacteremia and streptococ cal toxic shock syndrome (STSS) with high mortality rates has been reported
428
H. Muller-Alouf et al.
(4, 13, 24, 37). The severity of these invasive infections has been attributed besides host related factors (particularly the immune status) to an increase in the pathogenicity and virulence of the invading S. pyogenes strains (see Intro duction). In the present work, we investigated the phenotypic expression of some cell bound and extracellular pathogenicity and virulence factors of 212 clinical strains isolated in France (1990-1994) from patients with severe GAS infec tions among them 74 cases (35%) of STSS. Significant differences with respect to age were observed between those pa tients with STSS and the others. The mean age of the former was 36 years, whereas that of the latter was clearly higher (56 years) indicating that among other factors, patient age is an important parameter in STSS versus other in vasive GAS infections such as skin/soft tissue infections or bacteremia as also shown by other authors (4, 37). The analysis of the isolates investigated in this work involved M protein serotyping (in parallel with T protein serotyping and serum opacity factor determination in certain cases as complementary ap proaches for M serotyping), biotype determination, the assay in culture super natants of the pyrogenic (erythrogenic) exotoxin A (SPE A), streptolysin 0, streptokinase, hyaluronidase, transforming activity on human blood lympho cytes and the evaluation of the release of cysteine proteinase by GAS colonies on agar plates. The data were particularly examined with respect to STSS and non STSS isolates to find out whether significant differences in these virulence or pathogenicity markers could be identified. M protein serotyping has been widely used for more than 50 years as a method for studying the epidemiology of S. pyogenes, particularly in relation with infection. On this basis, a significant predominance of Ml and M3 sero types and to a lesser extent M12 and MIS has been observed in the past ten years among the strains isolated from patients with STSS and other severe GAS infections in North America, Europe and elsewhere (see Introduction). However, other M serotypes and non M serotypeable strains have been iso lated in such cases (29,40). In the present work, 25% of the isolates (53/212) expressed serotypes Ml and to a lesser extent M3, whereas 16% were untypeable (34/212) and 59% (125/212) belonged to 15 other serotypes without obvious prevalence of any of them reflecting probably the non epidemic background of the isolates. Howev er, among the 74 isolates from patients with confirmed STSS, 42% (31/74) be longed to serotype Ml and some of them to M3. These data, other reports (29, 40) as well as the isolation of many strains expressing Ml or M3 serotypes from asymptomatic carriers or during uncomplicated GAS infections (6, 16) suggest that the emergence of severe GAS infections is the result of strain-specific viru lence factors rather than virulence broadly defined by an individual serotype (4). Furthermore, the immune status of the host is certainly critical with regard to infection by strains expressing M 1, M3 or other serotypes. As shown by Holm et al. (20) the antibody titers against Ml antigen were found lower in fa tal cases than in patients with bacteremia or in uncomplicated infections.
Invasive S. pyogenes strains from severe infections
429
SPE A is historically known as primarily responsible for the skin rash in scarlet fever (1, 23, 24). The initial observations of Cone et al. (8) and Ste vens et al. (38) at the end of the 1980s which showed a close association between SPE A expression by the strains isolated from patients with severe GAS infections including STSS led to the contention, that this toxin plays a major role in the pathogenesis of these diseases and may be linked to their apparent resurgence (16,23,29,31,38,40). An accepted concept points to the direct implication of SPE A and probably SPE C (4) and other newly dis covered superantigens (11, 12,21,31,33) in severe GAS infections through the massive release of cytokines and other inflammatory factors by the immune system cells of the host (18). These effectors which are thought to mediate shock, tissue damage and fever are released as a consequence of the super antigen induced MHC class II dependant polyclonal proliferation of those T lymphocyte clones bearing appropriate motifs in the variable region of the ~ chain of T cell receptor (1, 26, 31). The concept is supported by the (i) in vitro release of various cytokines (interleukins-1, 2, 6, 8, 10, tumor ne crosis factors a, and ~, interferon y) by superantigen stimulated human pe ripheral blood mononuclear cells (14, 20, 27, 28, 31, 39) (ii) increased levels of cytokines in patients sera during STSS and other streptococcal diseases (14, 30), (iii) marked in vivo reduction of cognate V~-bearing T lymphocytes (26, 41), (iv) development of various pathophysiological disturbances, cytokine induction, organ impairment and death in experimental animal models (mouse, rabbit, baboon), challenged with SPE A or SPE C expressing strains (4,35,39). In the present work, SPE A was detectably produced by about 26% (541 212) of the isolates. However, SPE A release (40.5%) was higher for the iso lates from STSS patients. Moreover, this toxin was predominantly released by 67% of the strains of serotypes M1 or M3. Similar studies in the past few years on clinical isolates from North America (29) and Europe (7, 20, 34) showed great variations (30 to 80%) in the frequency of spe A gene and (or) of its phenotypic expression by the strains. This frequency was clearly higher among the isolates from patients with scarlet fever and other severe infections particularly STSS than in isolates from patients with non invasive infections or from healthy carriers (6, 29). On the other hand, as reported here, in a previous work (34) and by others (6,29,40), many isolates from severely diseased patients lacked spe A gene or did not produce the toxin suggesting that other bacterial products may be critically involved in patho genesis (6, 16). This is supported by the discovery of a number of novel GAS superantigens as mentioned above. The finding in the present work of po tent mitogenic activity in the supernantants of those strains with did not de tectably produce SPE A or SPE C (data not shown for the latter) supports this view. Streptococcal pyrogenic exotoxin B (SPE B) which is identical to or a vari ant of streptococcal cysteine proteinase SCP (5, 10, 15) has been shown to be released alone or in association with SPE A by clinical isolates during severe
430
H. Miiller-Alouf et al.
GAS infections (20, 24, 31). This chromosomally encoded enzymatic toxin appears to be associated with both shock and organ disorders in STSS patients (6, 40) as also supported by experiments in animal models (35). Moreover, SCP cleaves human fibronectin, vitronectin and interleukin (IL-) IB precursor with the generation of the pro inflammatory bioactive IL-IB (6). In the present work, ca. 85% of the strains released SCP regardless of strain serotypes and clinical orgin. Talkington et al. (40), Musser et al. (29) and Hauser et al. (16) similarly showed that SCP was produced by 63%, 51.8% and 59% of the strains respectively. Hyaluronidase (HA), streptolysin 0 (SLO) and streptokinase (SK) released in culture medium were also evaluated. The former was released by only 10% of the strains. Its possible involvement in pathogenicity has not been so far es tablished to our knowledge. SK was released in varying amounts (3 to 1000 U/ml) and no relationship between enzyme titer and the type of disease or clin ical origin of the strains was found. These results are in agreement with those reported by Chaussee et al. (6) who did not observe association between these two exoproteins and disease. All strains released SLO in amounts ranging from 4 to 256 HU/ml. This cy tolysin is expressed by virtually all isolates so far tested (2), in accordance with the permanent presence of slo gene in S. pyogenes strains (2). No apparent rel evance was found between the amounts of toxin released and patients diseas es as also found by Holm et al. (20) for swedish bacteremia strains. Whether this toxin plays a role in severe GAS infections remains to be established. However, several lines of evidence suggest its involvement in the pathogenesis of these diseases (14,40) and on account of the remarkable in vitro membrane damaging and disrupting effects on eukaryotic cells, particularly leucocytes (2). On the other hand, SLO has been shown by Hackett and Stevens (14) to behave as a potent in vitro inducer of the proinflammatory cytokines TNF a and IL-IB, and to increase synergistically in association with ETA the release of the latter cytokine by monocytes. SLO also acts jointly with SCP and SPE A to enhance tissue injury (35) suggesting that these proteins may function as important virulence and pathogenicity factors, particularly as shock media tors in severe GAS infections. Acknowledgement. We wish to thank P. Desreumaux for helpful discussion and assi stance with the statistical methods.
References 1. Alouf,]. E., H. Knoll, and W. Kohler: The family of mitogenic, shock-inducing and superantigenic toxins from staphylococci and streptococci. In: Sourcebook of Bac terial Protein Toxins. (J. E. Alouf and ]. H. Freer, eds.), pp.367-414. Academic Press, Ltd., London (1982) 2. Alouf, J. E. and C. Geoffroy: Production, purification and assay of streptolysin O. Meth. Enzymol. 165 (1988) 52-59
Invasive S. pyogenes strains from severe infections
431
3. Bouvet, A., P. Geslin, P. Kriz-Kuzemenska, V. Blanc, C. Devine, and F. Grimont: Restricted association between biotypes and serotypes within group A streptococ ci. J. Clin. Microbiol. 32 (1994) 312-317 4. Bronze, M. S. and]. D. Dale: The reemergence of serious group A streptococcal in fections and acute rheumatic fever. Am. J. Med. Sci. 311 (1996) 41-54 5. Chaussee, M. S., D. Gerlach, C. -E. Yu, and ].]. Ferretti: Inactivation of the strep tococcal erythrogenic toxin B gene (spe B) in Streptococcus pyogenes. Infect. Im mun.61 (1993) 3719-3723 6. Chaussee, M. S., J. Liu, D. L. Stevens, and].]. Ferretti: Genetic and phenotypic di versity among isolates of Streptococcus pyogenes from invasive infections. J. In fect. Dis. 173 (1996) 901-908 7. Colman, G., A. Tanna, A. Efstratiou, and E. T. Gawozerwska: The serotypes of Streptococcus pyogenes present in Britain during 1980-1990 and their association with disease. J. Med. Microbiol. 39 (1993) 165-178 8. Cone, L. A., D. R. Voodard, P. M. Schlievert, and G. S. Tomory: Clinical and bac teriologic observation of a toxic shock-like syndrome due to Streptococcus pyo genes. N. Engl. J. Med. 317 (1987) 146-149 9. Fischetti, V. A.: Streptococcal M protein: molecular design and biological behavior. C1in. Microbiol. Rev. 2 (1989) 285-314 10. Gerlach, D., H. Knoll, W Kohler,]. H. Ozegowski, and V. Hiibalova: Isolation and characterization of erythrogenic toxins. V. Identity of erythrogenic toxin type B and streptococcal proteinase precursor. Zbl. Bakt. 255 (1983) 221-233 11. Gerlach, D., H. Alouf, L. Moravek, M. Pavlick, and W Kohler: The characteriza tion of two new low molecular weight proteins (LMPs) from Streptococcus pyo genes. Zbl. Bakt. 277 (1992) 1-9 12. Geoffroy, c., H. Muller-Alouf, E. Champagne, ]. M. Cavaillon, and ]. E. Alouf: Identification of a new extracellular mitogen from group A streptococci. Zbl. Bakt. Suppl. 24 (1994) 90-91 13. Geslin, P., P. Kritz-Kuzmenska, A. Fremaux, H. Hav[ickova, A. Dublanchet, and A. Bouvet: Septicaemia caused by phenotypic variants of group A streptococci. Res. Microbiol. 147 (1996) 273-277 14. Hackett, S. P. and D. L. Stevens: Streptococcal toxic shock syndrome synthesis of tumor necrosis factor and IL-1 by monocytes stimulated with pyrogenic exotoxin A and streptolysin O. J. Infect. Dis. 165 (1992) 879-885 15. Hauser, A. R. and P. M. Schlievert, P. M.: Nucleotide sequence of the streptococcal pyrogenic exotoxin type B gene and relationship between the toxin and the strep tococcal proteinase precursor. J. Bacteriol. 172 (1990) 4536-4542 16. Hauser, A. R., D. L. Stevens, E. L. Kaplan, and P. M. Schlievert: Molecular analy sis of pyrogenic exotoxins from Streptococcus pyogenes isolates associated with toxic shock-like syndrome. J. C1in. Microbiol. 29 (1991) 1562-1567 17. Haynes, W L. and]. R. Tagg: A simple plate assay for detection of group A strep tococcus proteinase. J. Microbiol. Meth. 4 (1985) 25-31 18. Hensler, T., M. Koller, C. Geoffroy,]. E. Alouf, and W Konig: Staphylococcus au reus toxic shock syndrome toxin 1 and Streptococcus pyogenes erythrogenic tox in A modulate inflammatory mediator release from human neutrophils. Infect. Im mun. 61 (1993) 1055-1061 19. Hoge, C. W, B. Schwarz, D. F. Talkington, R. F. Breiman, E. M. MacNeill, and].]. Englender: The changing epidemiology of invasive group A streptococcal infec tions and the emergence of streptococcal toxic chock-like syndrome. A retrospec tive population-based study. J. Amer. Med. Ass. 269 (1993) 384-389
432
H. Miiller-Alouf et al.
20. Holm, S. E., A. Norrby, A.-M. Bergholm, and M. Norgren: Aspects of pathogene sis in serious group A streptococcal infections in Sweden 1988-1989. J. Infect. Dis.
166 (1992) 31-37
21. Iwasaki, M., H. Igarashi, Y. Hinuma, and T. Yutsudo: Cloning, characterization and overexpression of a Streptococcus pyogenes gene encoding a new type of mito genic factor. FEBS Lett. 331 (1993) 187-192 22. Knoll, H., j. Sramek, K. Vrbova, D. Gerlach, W. Reichardt, and W. Kohler: Scarlet fever and types of erythrogenic toxins produced by the infecting streptococcal strains. Zbl. Bakt. 276 (1991) 94-106 23. Kohler, w., D. Gerlach, and H. Knoll: Streptococcal outbreaks and erythrogenic toxin type A. Zbl. Bakt. 266 (1987) 104-115 24. Kohler, w.: Streptococcal toxic shock syndrome. Zbl. Bakt. 272 (1990) 527-564 25. Malke, H. and ].j. Ferretti: Streptokinase: cloning, expression and excretion by Escherichia coli. Proc. Natl. Acad. Sci. USA 81 (1984) 3557-3561 26. Michie, c., A. Scott, J. Cheesborough, P. Berverley, and G. Pasvol: Streptococcal toxic shock-like syndrome: evidence of superantigen activity and its effects on T lymphocyte subsets in vivo. Clin. Exp. Immunol. 98 (1994) 140-144 27. Muller-Alouf, H., ]. E. Alouf, D. Gerlach, J.-H. Ozegowski, C. Fitting, and J.-M. Cavaillon: Comparative study of cytokine release by human peripheral blood mononuclear cells stimulated with Streptococcus pyogenes superantigenic erythro genic toxins, heat-killed streptococci, and lipopolysaccharide. Infect. Immun. 62
(1994) 4915-4921
28. Muller-Alouf, H., ]. E. Alouf, D. Gerlach, ].-H. Ozegowski, C. Fitting, and ].-M. Cavaillon: Human pro- and anti-inflammatory cytokine pattern induced by Strep tococcus pyogenes erythrogenic (pyrogenic) exotoxins A and C superantigens. In fect. Immun. 64 (1996) 1450-1453 29. Musser, ]. M., A. R. Hauser, M. H. Kim, P. M. Schlievert, K. Nelson, and R. K. Se lander: Streptococcus pyogenes causing toxic shock-like syndrome and other inva sive disease: clonal diversity and pyrogenic exotoxin expression. Proc. Natl. Acad. Sci. USA 88 (1991) 2668-2672 30. Nadal, D., R. P. Lauener, C. P. Braegger, A. Kaufhold, B. Simma, R. Lutticken, and A. Seger: T cell activation and cytokine release in streptococcal toxic shock-like syndrome. J. Pediatr. 122 (1993) 727-729 31. Norrby- Teglund, A., M. Norgren, S. E. Holm, U. Andersson, and ]. Andersson: Similar cytokine induction profiles of a novel streptococcal exotoxin, MF, and py rogenic exotoxins A and B. Infect. Immun. 62 (1994) 3731-3738 32. Ozegowski, ].-H., E. Gunther, and W. Reichardt: Purification and characteriza tion of hyaluronidase from Streptococcus agalactiae. Zbl. Bakt. 280 (1994) 497-
506
33. Reda, K. B., V. Kapur,]. A. Mollick, j. G. Lamphear, J. M. Musser, and R. R. Rich: Molecular characterization and phylogenetic distribution of the streptococcal superantigen gene (ssa) from Streptococcus pyogenes. Infect. Immun. 62 (1994)
1867-1874
34. Reichardt, w., H. Muller-Alouf,]. E. Alouf, and W. Kohler: Erythrogenic toxins A, Band C: occurrence of the genes and exotoxin formation from clinical Streptococ cus pyogenes strains associated with streptococcal toxic shock like syndrome. FEMS Microbiol. Lett. 100 (1992) 313-322 35. Schanley, T. P., D. Schrier, V. Kapur, M. Kehoe,]. M. Musser, and P. E. Ward: Strep tococcal cysteine protease augments lung injury induced by products of group A streptococci. Infect. Immun. 64 (1996) 870-874
Invasive S. pyogenes strains from severe infections
433
36. Schwartz, B., R. Facklam, and R. F. Breiman: Changing epidemiology of group A streptococcal infection in the USA. The Lancet 336 (1990) 1167-1171 37. Stevens, D. L.: Invasive group A streptococcus infections. Clin. Infect. Dis. 14 (1992) 2-13 38. Stevens, D. L., M. H. Tanner, ]. Winship, R. Swarts, K. M. Ries, P. M. Schlievert, and E. Kaplan: Severe group A streptococcal infections associated with toxic shock-like syndrome and scarlet fever toxin A. N. Engl. J. Med. 321 (1989) 1-7 39. Stevens, D. L., A. E. Bryant, S. P. Hackett, A. Chang, G. Peer, S. Kosanke, T. Emer son, and L. Hinshaw: Group A streptococcal bacteremia: the role of tumor necro sis factor in shock and organ failure. J. Infect. Dis. 173 (1996) 619-626 40. Talkington, D. F., B. Schwartz, C. M. Black, ]. K. Todd, ]. Elliott, R. F. Breiman, and R. R. Facklam: Association of phenotypic and genotypic characteristics of in vasive Streptococcus pyogenes isolated with clinical components of streptococcal toxic shock syndrome. Infect. Immun. 61 (1993) 3369-3374 41. Watanabe-Ohnishi, R., D. E. Low, A. McGeer, D. L. Stevens, P. M. Schlievert, D. Newton, B. Schwartz, B. Kreiswirth, Ontario Streptococcal study project, and M. Kotb: Selective depletion of V~-bearing T cells in patients with severe invasive group A streptococcal infections and streptococcal toxic shock syndrome. J. Infect. Dis. 171 (1995) 74-84 42. Willoughby, R. and R. N. Greenberg: Toxic shock syndrome and streptococcal py rogenic exotoxins. Ann. Intern. Med. 98 (1983) 559 43. Working Group on Severe Streptococcal Infections: Defining the group A strepto coccal toxic shock syndrome. J. Amer. Med. Ass. 269 (1993) 390-391 Corresponding author: Dr. Heide Muller-Alouf, Centre d'Immunologie et de Biologie Parasitaires, Institut Pasteur, 1, Rue du Professeur A. Calmette, F-59019 Lille, France, Tel.: 33320877965, Fax: 33320877888
29
Zbl. Bakt. 28613