Zbl. Bakt. Hyg. A 266, 104-11 5 (1987 )

Streptococcal Outbreaks and Erythrogenic Toxin Type A *

WERNE R KO HLE R, DIETE R GERLAC H, an d HEIDE KNO LL

Akademie der Wissenscha ften der DDR, Zentralinstitut fur Mikrob iologie und experimen­telle Therap ie, l ena

Wit h 3 Figures· Received March 25 , 1987

Abstract

Reference strains of Streptococcus pyagenes and strai ns from recent epidemics andsporadic cases of scarlet fever were examined for their ability to produce erythrogeni c toxintype A (£T A) by ELISA and doubl e immunodiffusion (a uchter/any) using an anti-ET Aantibody purified by affinity chromatography. Of the reference strains (most of them iso­lated before 1945) 16/5 1 pro duced more or less ET A (Table 1). ET A synthesis is strain­specific, but no t type-specific. Well-known toxin produ cers like the strains NY-5; 594 or"Smith" produ ce up to 16.000 ilg/l under optima l culture conditions. Type 3 strains isolatedfrom scarlet fever patient s duri ng the outbreak 1972/73 seem to belong to one clone asevidenced by the uniform 50S-PAGE patt ern: They were found to produce 5-200 ilgll(mean 68 ilg/l) £T A only. Type 3 strains from sporadic cases, isolated 10 years lat er,produ ced 0-138 ilg/l (mean 30 ug/l), Strains of the type 1 clone, causing the epidemic in1982/83 produced only 0.75- 10 ilg/l (mean 8 ilg/I) £T A (Table 3). On ly a few strains of S.pyogenes isolated 1984 or later synthesized ET A but they were found more often toprodu ce £T B (proteinase precursor ) in batch cultures. S. pyagenes strains seem to have losttheir abili ty to produce large amounts of ET A dur ing the last decades. Because th is tox inmust be considered as a path ogenicity factor the decrease in toxin production may be onereaso n for the present mild form of scarlet fever.

Zusammenfassung

Referenzst iimme des Streptococcus pyogenes sowie Srarnrne von Scharl achep idemien undsporadischen Fallen jungeren Datums wurden auf ihr Verrnogen gepr ufr, erythrogenes To­xin Typ A (£T A) zu bilden. Die qualitati ven und quantirariven Untersuchungen wurden miteinem ELISA und der Agargelpriizipitation (a uchter/any) unter Verwendung von af­finitiitschro matographisch-gereinigtem ET A-Antikorper durchgefiihrt . Von den meist vor1945 isolierten Referenzstiimmen bildeten 16151 ET A, (Tab. 1). Die ET A-Synrhese iststarnm-, aber nicht typenabhangig. Internation al bekannte Toxinbildner, wie die StamrneNY-5, 594 und Smith bildeten unter optimalen Kulturbedingungen his zu 16000 ilg/1ET A.

Article invited on the occasion of the 100th anniversar y of the "Zentralblart" .

Streptococcal Outbreaks and Erythrogenic Toxin Type A 105

Die wahrend der Scharlachepidemie 1972/73 isolierten Typ 3-Stamme gehoren, wie dieSDS-PAGEbeweist, einem Cion an. Siebildeten nur zwischen 5-200 ilg/I (Mittelwert 68 ilg/I), Typ 3-Stamme, die 10 Jahre sparer von sporadischen Fallen isoliert wurden, 0-138 ilg/I(Mittelwert 30 ilg/l). S. pyogenes-Stamme des Typ 1, der die Epidemie 1982/83 verursachte,erwiesen sich als noch schwachere Toxinproduzenten (0,75-10 (Mittelwert 8) ilg/I) (Tab.3). Nur wenige der nach 1984 isolierten S. pyogenes-Stamme bildeten ET A, dafur hatten siedie Eigenschaft gewonnen, in Standkulturen ohne pH-Regulation ET B (Proteinase Precur­sor) zu produzieren. Dies lalSt sich bei vielen Referenzstarnmen oft nur durch Ziichtung beipH Werten unter 6,5 erreichen. - S. pyogenes hat offenbar in den letzten Dekaden dieFahigkeit zur Bildung grolSerer Mengen ET A verloren. Da ET A nach unseren Unter­suchungen mit dem Gewebekafig-Infektionsmodell (14) als Pathogenitatsfaktor anzusehenist, kann die verminderte ET A-Bildungeiner der Griinde fur den derzeit leichten Verlauf desScharlachs sein.

Introduction

The nature of the scarlatiniform rash is still obscure. Even in most of the moderntextbooks, this erythema is defined as a result of the activity of erythrogenic toxin(s)produced by certain strains of Streptococcus pyogenes. Originally, Dick and Dick (5)proposed a toxic activity of erythrogenic toxin as the cause of scarlet fever erythema.They probably worked with a toxin which is now known as erythrogenic toxin A (syn.streptococcal [pyrogenic] exotoxin type A; SPE A). Some discrepancies in toxin-antito­xin flocculation reactions and skin neutralizing activity were explained by Hooker andFollensby (15) as being due to the existence of more than one type of toxin. Theyprobably detected erythrogenic toxin type B. Fifty years later this was found to beidentical with streptococcal proteinase precursor by Gerlach et al. (10).

Shortly after the Dick reaction had been published and explained as a reaction ofprimary toxicity, Dochez and Stevens (6) also reported on substances present in fil­trates of "scarlet fever streptococci" but they held the skin reaction to be a hypersen­sitivity reaction.

Kim and Watson (17) revived this hypothesis and tried to combine both primary andsecondary toxicity. According to Kim and Watson (17) primary toxicity (pyrogenicity,lethality, immunosuppression, enhancement of susceptibility to endotoxin shock) mayplaya part in the pathogenesis of streptococcal infections and the secondary toxicity is"an allergic reaction to a heat-stable portion which is antigenically common to apyrogenic exotoxin" i.e. secondary toxicity is responsible for the rash in scarlet feverby inducing a skin hypersensitivity. This concept as well as biochemical studies of thetoxins have been reviewed among others by Watson and Kim (30), Parker (27) andAlouf (1).

A second problem in scarlet fever is allied to the epidemiology of the disease. Thehigh mortality of foregoing centuries (3.8%0 in Germany in 1845) decreased to nearlyzero in the middle of this century but the morbidity was similar during this centuryprovided that a correct reporting system existed. Because of the lack of reliable dataexplaining this change in the clinical picture of scarlet fever we started a study on theproduction of erythrogenic toxins by S. pyogenes strains isolated from patients duringlarge outbreaks of scarlet fever. Hallas (13) found the production of erythrogenic toxintype A (ET A) only in strains isolated before 1940. Our results are somewhat different.

106 W. Kohler, D. Gerlach, and H. Knoll

Material and Methods

Bacteria

S. pyogenes reference strains were available at the National Streptococcus ReferenceLaboratory of the GDR, at the authors institute. Most of the strains were obtained from theStreptococcus Department, State Serum Institute Copenhagen in 1954, a few others fromDr. j. Rotta, Prague.

The patient strains were isolated from 1960 to 1986 from cases of scarlet fever byregional laboratories, sent to the Reference Laboratory for typing (21,22) and kept here in alyophilized state. In 1961 and 1962, type 49 and type 22 strains, respectively, were isolatedin local outbreaks. The epidemics in 1972/73 and 1982/83 were caused by types 3 or 1,respectively.

Determination of erythrogenic toxin type A

Cultivation and concentration of supernatants: The streptococci were grown as batchcultures, without regulation of pH for 20 h at 3rC in 100 ml Todd-Hewitt broth (Difeo)and killed by addition of 0.5% H20 2 (final concentration). After centrifugation the super­natants containing erythrogenic toxins and other streptococcal exoproducts were concen­trated by differential precipitation with -20 °C ethanol and resolubilization in acetate­buffered saline according to Kim and Watson (17). Insoluble material was discarded, thesupernatant dialyzed against distilled water and lyophilized. For determination of ery­throgenic toxins this material (from 100 ml original culture supernatant) was dissolved in0.5 to 0.8 ml distilled water.

Preparation and purification ofantibodies to ET A: Antibodies to ET A were prepared ina goat as described by Louvard et al. (24). Briefly, purified ET A from strain NY-5 was usedfor immunization. With the total amount of 11 mg ET A, a hyperimmune antiserum wasachieved after 13 injections, starting with an immunizing dose of 400 ug ET A and increasedat 14-28 day intervals. After six months the titer of the antiserum was found to be 0.55 mg!ml defined as ET A binding capacity in the Mancini test (Becker, 2).

Booster injection were given later consisting of 3 mg ET A with complete Freund adjuv­ant, by the subcutaneous route.

The antibodies were purified by affinity chromatography using 7 ml AH-Sepharose cou­pled with 30 mg cloned ET A (11). The toxin was coupled by glutaraldehyde (Cambiaso etaI., 4), followed by reduction with sodium borohydnte, A total of 157 mg pure antibodywas obtained from 162 rnl antiserum with an antibody titer of 0.27 mg/ml as determined bythe Mancini technique (Becker, 2). The antibodies were stored in 50% glycerol (v/v) at-20°e.

Determination of ET A by Ouchterlony test: The double immunodiffusion test wasperformed in a 1.5')10 agar gel prepared with 0.05 M Tris (hydroxymethyl) aminomethane(TRIS)-HCI buffer, pH 8.0. The wells had a diameter of 4 rnrn, the distance was 5 rnrn, andwere filled with 20 fAI antibody or sample solution. The concentration of the purified anti­ET A-antibody was 1 mg/rnl, the samples were lOO-200 fold concentrated culture supernat­ants of the strains to be examined for the production of ET A (see above). As a controlhighly purified ET A was used in a concentration of 0.1 mg/ml.

Enzyme-linked immunosorbent assay (ELISA): This method (Engvall and Perlmann, 7)was used for determination of the amount of ET A in culture filtrates. For coating ofmicrotitration plates purified IgG of goat antibody against ET A was used in a concentrationof 15 fAg/m!. After 3 h incubation at 37°(; the plates were washed and stored at 4°C untiluse (up to two weeks). Culture filtrate concentrates, diluted 1:5 to 1:625 were added andincubated for 1 h at room temperature. After washing a rabbit antiserum to ET A (diluted1:100) was added and incubated for 1 h at room temperature. After washings an alkalinephosphatase labelled anti-rahbit-IgG conjugate prepared in swine, diluted l:300, was ad­ded. The plates were allowed to incubate overnight at 4°e. After washings the substrate p-

Streptococcal Outbreaks and Erythrogenic Toxin Type A 107

nitrophenyl phosphate was added and after 30 min, the reaction was stopped by 3 MNaOH. The absorbance was measured at 405 nm in a Specol 11 (VEB Carl Zeiss lena)equipped with a scanner for microtiter plates. The system was calibrated with a dilutionseries of cloned ET A between 100 ltg and I ltg. The culture supernatant concentrates wereconsidered to be positive when the absorbance was equal to or higher than the valueobtained with 10 ng/ml purified ET A.

Sodium dodecylsulfate (SDS) polyacrylamide gel electrophoresis (SDS-PAGE)

The method was performed as described by Laemmli (23) using a 15% polyacrylamidegel concentration. The samples for SDS-PAGE were prepared from 0.25 to 0.4 ml of theculture supernatant concentrates. After addition of 4.75 to 4.6 ml 0.05 M TRIS buffer (pH

rr >

0 ...- C 'J

1 2 3 4 5 Cr.,---

(~

D " :1::J --- ---- ---."

c= ~

"---

7 8 9 10 11 12 13, ( :J ~)--- ---

Fig. 1. Immunodiffusion (Ouchterlony) of culture concentrates. Each slide (1-13) carriestwo rosettes. The central wells were filled with anti- ET A antibody (1 mg/rnl) and the wellson the top and the bottom with purified ET A (0.1 mg/ml). The left and right wells werefilled with culture concentrates.The culture concentrates are denoted by the [MET Collection number (Table 1) or other­wise, where not identical with strains listed in Table 1.Slide 1: ET A ET A

3000 Antibody 3000 3005 Antibody 3000ETA ETA

upper rosette lower rosetteSlide 1: 3000-3000/3005-3000; Slide 2: 3045-AM4/3059-3058;Slide 3: T19(Watson)-305913067-AKopI9;Slide 4: 3084-AT27/31 06-3005;Slide 5: Group C-3059/B220 (Elliot)-594C;Slide 6: 3122-SmithlType 6-Tvpe 6 (routine strains);'Slide 7: C203U-S. sanguis with cloned ET A13009-Type 11;Slide 8: 3027-3022/3043-3031; Slide '1: 3044-Type 1113045-AM12;Slide 10: AT14-3045/3057-ET B;Slide 11: 3073-3071;Slide 12: AT25-AT26/A Kop 38-A Kop 30;Slide 13: S. sanguis-So sanguls/ET A-Toxoid-empty.

108 W. Kohler, D. Gerlach, and H. Knoll

7.5) to a final volume of 5.0 ml the samples were precipitated with 1 ml of 30% (w/v)trichloracetic acid (TCA). The centrifugated precipitates were dissolved in sample bufferaccording to Laemm/i (23), and boiled for 5 min. Twenty microliters of these samples wereapplied to 80 x 80 x 0.8 mm slab gels. Staining with Coomassie Blue R 250 and destainingas described by Weber and Osborn (31).

Results

Sensitivity and specificity af Ouchterlany and ELISA tests.

Experiments with dilutions of purified ET A showed that a positive precipiationreaction occurred with the purified antibody when the concentration of toxin was 20ug/ml or more. Regarding the concentration of supernatants, strains producing 0.2 mg/I or more could be identified by the Ouchterlany test. The ELISA was about 1000 timesmore sensitive than the double immunodiffusion. A positive reaction was obtainedwith 10 ng ET A/ml. Strains producing 0.1 Ilg/l reacted positively in the ELISA there­fore, some strains were positive in the ELISA only.

Unspecific reactions were excluded by the use of an antibody purified by affinitychromatography with immobilized cloned ET A, produced by a heterologous host (S.sanguis, 32). The concentrates of all strains were examined in the Duchterlony testwith reference toxin type A. No unspecific reactions were observed (Fig. 1).

Table 1. Production of ET A by reference strains

Reaction

Ouchter/any positiveELISA positive

Ouchter/ony negativeELISA positive

Ouchter/any negativeELISA negative

No IMET'f No (Internat. Designation; type)

11 3000 (SF 130; 1); 3005 (TZ/44/Rb4/91; 2); 3027 (T8/52/3; 8); 3033 (U/447; 11); 3036 (T12/46/1; 12); 3043(NY-5; 12); ATZ7 (46/1; 27); 3084 (D23; 29); AKap41(-; 41); AKop47 (-; 47); 3106 (DSC 300; 49)

5 3057 (T88A/0/5; 15); ATZ6 (SF13; 26); AKop38 (-;38); 3097 (C95/128/5; 39); AKop46 (-; 46)

35 3009 (D58-Richards; 3); 3017 (T4/95/Rb5; 4); 3021(T5B; 5); 3022 (Lambert; 6); 3031 (T9/101/4; 9); 3044(K56; 12); 3045 (T13/51; 13); AT14 (-; 14); 3051 (S23/101lW5; 14); 3058 (J17E; 17); 3059 (R147; 18); 3065(J17D; 19); AKop19 (5310; 19); 3067 (63T; 22); 3071(Barts; 23); 3073 (115; 24); ATZ5 (TZ5/34/3; 25); 3081(Hv glossy; 28); AKop30 (-; 30); 3090 (C107/24/6; 33);3101 (C126/59/2; 43); AKop44 (-; 44); 3105 (B403/48/1; 48); 3108 (B514/33/3; 50); 3111 (A309/31/5; 51);3115 (Schonborn; 52); 3117 (Hanson; 53); 3119 (King­bird; 54); 3120 (-; 55); 3122 (Baker; 56); 3123 (SS790;57); 3127 (0335; 60); 3129 (0336/56/1; 61); 3113 (Im­petigo 19); 3114 (B3264).

Culture Collection of the Zentralinstitut fur Mikrobiologie und experimentelle TherapieJena.

Streptococcal Outbreaks and Erythrogenic Toxin Type A 109

Table 2. Qualitative ET A production

Number of strainsType Year Epidemic or E':' P':"f E P E P E P

sporadic + + + (+) +

49 1961 local outbreak 222 1962 local outbreak 7

3 1972/3 epidemic 20 41 1982/3 epidemic 123 1982 sporadic 11

'f E = ELISA*'f P = Precipitation (Ouchterlony)+ = positive, (+) = weakly positive, - = negative

Production of ET A by reference strains

Representatives of S. pyogenes types 1-61 were examined for their ability to pro­duce ET A in the Ouchter/any test and ELISA (Table 1). Eleven out of 51 strainsexamined were positive in both reactions, 5 strains were positive in ELISA only, (i.e.weak ET A producers), and 35 did not produce detectable amounts of ET A (i.e. lessthan 0.1 ug/l). Most of the strains producing ET A were isolated between 1910 and1944 (types 1-48) (20). In Table 1 the strain numbers are given in detail since produc­tion of ET A is not type- but strain-specific. A striking example is type 12: strain NY-5was one of the most potent producers of ET A and strain K56 (Kjems, 19) served as anegative control.

Production of ET A by epidemic strains and strains of sporadic cases

S. pyogenes strains of two local outbreaks and two epidemics, representing fourdifferent types, as well as sporadic strains were examined qualitatively for the produc­tion of £T A by precipitation and EUSA (Table 2). Two strains of type 49 from a localoutbreak were nonproducers of ET A. One year later another small outbreak of scarletfever occurred in another district and all type 22 strains examined were found to bestrong producers. The type 3 epidemic started in 1968 and reached its peak with anincidence of 47.3% 00 in I972/Tl (29). All strains produced relatively large amounts ofET A (Table 3). The type 1 epidemic which occurred 10 years later (1982/83) was

Table 3. Quantitative ET A production (Measured by ELISA)

ET A f-lg/lYear Type n x SO Range

Epidemic 1972/73 J 23 67.9 63.75 5-200Epidemic 1982/83 I 8 7.9 7.75 0.75-10Sporadic 1982 J 12 29.5 51.0 0-138

110 W. Kohler, D. Gerlach, and H. Knoll

u, u,3en 3 3 33 3 3 3 0 3 3 3 3 3 3

<,NU.'" u.1ll<D U. t::u..~ u. O\ LL CO 1.L,u.~ !(U.~LL~U. ~u.'" U.

<D ...... N~<, <,

CI)~Q\~........ Ill <D <D ..........0 0 __

N N M M...,..., U') III <D~ r:::r-..", .......... ..... e-, ............... r--.""'""" r--. Cl)COQ)CO ee '" '" CDCOClOc;D '" '"

Fig. 2. Percentage of S. pyogenes type in scarlet fever in the GDR 1968- 1986.S type 4/24 . . . I!IlI type 12 • type 3 fIJ1 type 1 0 agglutination pattern 12/B3264, SO R +F = spring; W = autumn-winter

characterized by stra ins synthesizing low amo unts of ET A, being detectable onl y byELISA but not by the O uchter /ony test. Stra ins of type 3 disappeared after the epidemic1972/73 until 1980 (a few strai ns were isolated in 1975/76; Fig. 2).

Ty pe .1 st ra ins, isolated in 1982 from sporadic cases of scarlet fever, pr od uced ET Aonly in low amo unts or no t at al l. Part of these strains was exa mined for quanti tativeET A production in the ELISA. The va lues must be carefully interpreted since we haveused the same culture conditions but did no t determine th e number of microorgan isms.It can be clearly seen th at there have been differences between the two epidemics andbetween type 3 strains isolat ed fro m epid emic and sporadic cases of scarlet fever (Table3).

5DS-PAGE or culture con centrat es

f or better demonstr a tion so me results of SDS-PAGE of culture concentra tes areshown as drawin gs instead of photograph s in Fig. 3. Prefixes T or M mean refere ncestra ins of so me types. Types 49, 22, .1 and 1 refer to epidemic stra ins and K56 is anontoxin producer (type (2).

All epide mic stra ins (Table 2) were examined by SDS-PAGE and showed absolutelyidentica l pat tern s. It co uld be con cluded tha t the epidemic strai ns of type 1 belonged to

the same clon e. The same was true for types 3 and 49 ; for type 22 two examined strain swere no t sufficient to suppo rt such an assumption.

Strept ococcal Outbreaks and Erythro genic Toxin Type AlII

kD

50 •

1.0

30

26 •

21

11.12

6

- . -

-. -Typ Typ Typ Typ Typ

T1 M4 T8 TlO T12 TI3 MIl. Tl7 T18 56 1.9 22 3 I K56

Fig. 3. SOS-PAGE of culture supernata nt concentrates.T l - strain SF130 used in the tissue cage infection experiments;T 12 - stra in NY-5; for others, see text.

Otherwise, SDS-PAGE pattern s were different. Some strains showed onl y a few,others up to 8 protein bands. Stre ptococcal proteinase proenzyme appeared at 40 kD,protein ase at 30 kD , both erythrogenic to xin s A and C at 26 kD and the low-molecularweight mitogen LMP- I 0 k (12) at 10 kD. The band patterns and th eir int ensity werespecific markers for each strain.

Discussion

It has been the aim o f thi s study to exa mine a ma jor number of stra ins isolated fromsca rlet fever outbreaks for their ability to produce ET A and to compare the resultswith those obtained from reference strai ns and from single cases of sca rlet fever. Only afew papers dealt with simil ar results but they were restri cted to qu alit at ive aspects. Toour knowledge this is the first pap er considering also qu antitat ive dat a of ET A synth­esis of more than a few str ain s.

The examination of more than 100 strains by SDS-PAGE, ELISA and agar gelprecipitation required the preparation of culture supernatant concentrates by a simple,rep roducible method. Th eref ore the cultivation conditions could be not optimal. It isknown th at a maximal yield can be obta ined only with pH-regulat ed cultures; batchcultures are less effective. Furthermore, the optimum for product ion of erythrogenictox in type B (st reptococcal proteinase proenzyme) is abo ut pH 6.0 . The optimal cul­ture conditio ns for the synt hesis of erythrogenic to xin C are unknown with regard toth e medium. Therefore, the report is limited to ET A although the stra ins were alsoexamined for th eir ability to prod uce erythrogenic toxins B and C. It was for the samereason - the great number of strains - that only a simple meth od of concent rationcould be used, i.e. precipit at ion by ethanol. In thi s wa y the concentrat ion was limitedto 100-200 fold, otherwi se the concent rate became to o viscous du e to th e accompany­ing concentration of hyalur on ic acid .

112 W. Kohler, D. Gerlach, and H. Knoll

The quality of antibody is of great importance for such studies. In order to obtain ahigh specificity of the agar gel precipitation reaction and ELISA we used an antibodypreparation purified by affinity chromatography on a Sepharose-cloned ET A-column.Therefore precipitating cross-reactions were not likely to occur. The specificity of theELISA test was confirmed by the examination of 20 strains (not included in this study),isolated in 1984. These strains, isolated from scarlet fever patients, were examined byFerretti et al. (8) using a specific probe consisting of 606 bases within the speA gene. ByDNA-DNA hybridization, two strains were found to contain the speA gene. Whenexamining the same strains in a blind trial by ELISA only these two were found toproduce ET A. Bloomster and Watson (3) examined 14 recently isolated S. pyogenesstrains by an Elek-like method. A macro-colony was grown on Todd-Hewitt agarplates and after incubation wells were made around the colony and filled with anti­bodies against ET A, ET Band ET C. They found 12/14 strains to be ET B producers,three produced also ET C and two produced no ET. Strains synthesizing ET A were notseen. Hallas (13) proved 40 recently isolated strains from scarlet fever and found 26 toproduce ET B, 12 additionally also ET C, 7 were ET C producers and 2 were negative.On the other hand, 4/10 strains isolated before 1940 produced ET A (partly in combi­nation with ET B and/or C). Hallas concluded "that there appears to be a difference inthe type of SPE [streptococcal pyrogenic exotoxin] produced by recent isolates fromscarlet fever cases and some of the strains isolated before 1940". The same can beconcluded from the studies of Bloomster and Watson: they thought it to be anevolutionary process. We also found among recently isolated strains more B toxinproducers in batch-cultures. This seems to be the main point. More S. pyogenes strainsisolated in the eighties seem to be able to produce ET B - or correctly: streptococcalproteinase precursor - under the conditions of a batch culture. According to ourexperience almost all strains of S. pyogenes are able to produce proteinase (ET B) inpH-regulated cultures at pH 6.0.

The results of this study point out that recently isolated strains have been found toproduce less ET A than some of the reference strains isolated more than 50 years ago.In 1961-198343/45 strains isolated from scarlet fever outbreaks in the GDR producedET A and 11/12 strains from sporadic cases did so. In 1984, only 2/20 S. pyogenesstrains from sporadic scarlet fever cases produced this toxin. A difference was foundalso in the amount of ET A produced by strains isolated during different epidemics andfrom sporadic cases. The quantitative determination was done by ELISA. The amountof ET A determined by ELISA was 4 to 20 times lower than the results obtained by theMancini technique. The higher the amount of ET A produced by the culture was, thesmaller the difference between both methods became. - Type 3 strains of the epidemic1972/73 produced about 68~tg/1 ET A (mean ELISA value), strains of sporadic cases ofthe same type isolated in 1982: 30 ~g/I; type 1 strains from the epidemic 1982/83 8 ~g/l

only (and therefore gave negative reactions in the precipitation reaction) (Table 3). Forthe period before 1983 we found only a very few strains not producing ET A. In bothepidemics the clinical picture did not differ, regardless of the amount of ET A producedby the infecting strain. - Reference strains, i.e. strains from severe scarlet fever cases atthe beginning of this century produced much more ET A. Cultivated under optimalconditions with pH-regulation and determined by the Mancini technique, strain NY-5(isolated 1924?) produced 3-8 mg/l; strain 594 (isolated 1910?) 5 mg/l, and strain"Smith" (isolated before 1935) up to 16 mg/l (9). This is in agreement with the SDS­PAGE pattern. In Fig. 3, the strain designated as no is strain NY-5 showing a broadband at 26 kD while ET A appears at 26 kD. On the other hand the 26 kD bands of the

Streptococcal Outbreaks and Erythrogenic Toxin Type A 113

epidemic strains (Type 1 and Type 3) were very small. - The severity of scarlet fevermay be influenced by the amount of ET A produced by the infecting strains. ET A mustbe considered as a pathogenicity factor. Using the tissue cage model, 4/10 rabbits diedafter infection with strain SF 130 (up to 9 mg/l ET A; 9) but no deaths occurred afterimmunization with ET A toxoid prior to infection (14).

When analyzing the results of this study one has to consider, however, that thestrains of the epidemics 1972/73 (type 3) and 1982/83 (type 1) seem to belong to twospecific clones. Only the range of ET A production varied within the strains of anepidemic but the SDS-PAGE results were absolutely identical. This fits to the cloneconception described by 0rskov et al. (26) for enterotoxigenic E. coli. The structuralgene for ET A (speA) is located in the bacteriophage genome of phage T12 (andpossibly other phages, too; 16, 32). Lysogenization by appropriate phages resulted inthe conversion of non-toxigenic strains to ET A producers, similar to conversion of C.diphtheriae to toxin synthesis hv the bacteriophage ~. Nida and Ferretti (25) indicatedthat transfer of the speA gene by bacteriophage was a general phenomenon. It isunknown whether or not the first strain of the type 3 epidemic was toxigenic. Asalready mentioned, this outbreak started in 1968/69. At that time, 6.5% of S. pyogenesstrains isolated from scarlet fever patients belonged to type 3 (Fig. 2). At the peak of theoutbreak, in the winter 1972/73, nearly 60'/"0 of the strains were identified as type 3. Atthe same time about 60% more cases of scarlet fever than in endemic years appeared,and the so-called "endemic types" 4/24... and 12 disappeared, which normally rep­resented 50-70% of the group A streptococci isolated from scarlet fever patients. Thesituation was similar in the type I epidemic in 1982/83. About 46% of group Astreptococci were type 1 strains and this is also the percentage of additional casescompared with epidemic years (The incidence in 1972/73 was 47.3%00, in 1982/8333.9%00 (29)). The epidemic strains were grafted on the normal endemic process, andone strain (clone) was responsible for the outbreak, i.e. type 3 in 1972/73 and type 1 in1982/83. Interestingly, in the endemic years 1976 to 1978 a "prewave" of type 1occurred but did not increase the cases of scarlet fever to epidemic dimensions (Fig. 2).The SDS-PAGE patterns of type 3 strains (the same applies for type 1) gives evidence ofthe uniformity of these strains. As far as we have seen each strain shows an individualSDS-PAGE pattern, but strains of the same type isolated from the same event (hospitalinfections) were identical (data not shown).

The results of this study do not answer the question why a rash appears in scarlatinaor even whether the SO-GlUed ervthrogenic toxins, A, Band C are the only onesresponsible for the rash. Up to 1983 a few strains only were found to be non-producersof ET A (within the limitations of the methods). Other streptococcal T cell mitogens(like LMP-lOk, 12) may he able to cause a rash, too. In lymphocyte transformationtests all supernatants tested contained mitogenic substances and the relative stimula­tion (expressing the stimulation hy the culture filtrate examind in relation to thestimulation achieved by I [tg purified ET A) was not statistically different in scarlatinalstrains with high, low or no FT A production (data not shwon).

The real difference between recent epidemic strains and scarlet fever strains fromperiods of high mortality (heglflning of this century) are quantitative. Quinn (28)speculated whv the severity ot streptococcal infections has declined but not the preva­lence. He did not come to am tinal conclusion hut believed that "there are indicationsthat repeated streptococcal infections due to similar M types (worldwide) occurring inyoung children over many decades (generations) result in immunity and possibly theevolution of less virulent but nor less infectious strains of group A streptococci". This is

8 Zbl. Bakt. Hyg. A 266/1-2

114 W. Kohler, D. Gerlach, and H. Knoll

not in agreement with the isolation of "new" types from scarlet fever. We isolatedtypes 76 and provo type 180, and we got strains from Dr. Sramek, Prague, of the types73, 75 and 77. The clinical picture of the disease was the same mild form, but thesetypes have been (until now) rare and it is unlikely that an immunity exists in thepopulation.

We present another hypothesis. The decline in mortality is caused by the evolution ofstrains producing less toxin. It is most unlikely that this limitation of toxin synthesis isthe single reason but it may be an important factor. Present knowledge does not sufficeto explain the existence of strong and weak toxin producers. It may be that thenumbers of copies of the toxine gene are different changes in the mechanism of excre­tion occurrence etc. The pathogenicity of group A streptococci shows cyclic trends andif the quantity of toxin production influences the severity of the disease then somedaystrains with a high Et A production may reappear. At present we are facing the samesituation as Thomas Sydenham, when he described "febris scarlatinae" as a verybenign disease (18). It may change to a severe disease again as he experienced it in 1689when scarlet fever changed to malignancy.

References

1. Alouf, ]. E.: Streptococcal toxins (streptolysin 0, streptolysin S, erythrogenic toxin).Pharm. Ther. 11 (1980) 661-717

2. Becker, W.: Single radial immunodiffusion. A method for determination of antiserumtiters. Clin. Chim. Acta 23 (1969) 509-510

3. Bloomster, T. G. and D. W. Watson: Recent trends in streptococcal pyrogenic exotoxinproduction by group A streptococci. In: S. E. Holm and P. Christensen (Eds.), BasicConcepts of Streptococci and Streptococcal Diseases, pp. 139-141. Reedbooks, Chert­sey (1982)

4. Cambiaso, C. L., A. Goffinet, ]. P. Vaerman, and ]. H. Heremans: Glutaraldehyde­activated aminohexyl-derivate of Sephadex 4B as a new versabile immunosorbent. Im­munochemistry 12 (1975) 273-275

5. Dick, G. F. and G. H. Dick: Etiology of scarlet fever. ]. Amer. Med. Ass. 82 (1924)301-302

6. Dochez, A. R. and F. A. Stevens: Studies on the biology of the Streptococcus. VII.Allergic reactions with strains from erysipelas. ]. expo Med. 46 (1927) 487-495

7. Engvall, E. and P. Perlmann: Enzyme-linked immunosorbent assay, ELISA. III. Quanti­tation of specific antibodies by enzyme-labeled anti-immunoglobulin in antigen-coatedtubes. ]. Immunol. 109 (1972) 129-139

8. Ferretti,]. t.. C. E. Yu, W. L. Hynes, and C. R. Weeks: Molecular characterization ofthe group A streptococcal exotoxin type A (erythrogenic toxin) gene and product.Infect. Immun. (in press)

9. Gerlach, D., H. Knoll, W. Kohler und ]. H. Ozegowski: Isolierung und Charak­terisierung erythrogener Toxine von Streptococcus pyogenes. 3. Mitt.: VergleichendeUntersuchung erythrogener Toxine vom Typ A. Zbl. Bakt. Hyg., I. Abt. Orig. A 250(1981) 277-286

10. Gerlach, D., H. Knoll, W. Kohler, j.-H. Ozegowski, and V. Hribalova: Isolation andcharacterization of erythrogenic toxins. V. Identity of erythrogenic toxin type Bandstreptococcal proteins precursor. Zbl. Bakt. Hyg., I. Abt, Orig. A 255 (1983) 221-233

11. Gerlach, D., W. Kohler, H. Knoll, L. Moravek, C. R. Weeks, and].]. Ferretti: Purifica­tion and characterization of Streptococcus pyogens erythrogenic toxin type A producedby a cloned gene in Streptococcus sanguis. Zbl. Bakt. Hyg, A 266 (1987) in press

Streptococcal Outbreaks and Erythrogenic Toxin Type A 115

12. Gerlach, D., ].-H. Ozegowski, H. Knoll und W. Kohler: Isolierung und Charak­terisierung von erythrogenen Toxinen. VIII. Die Reinigung eines biologisch aktivenProteins vom Molekulargewicht 10000 (LMP-I0k) aus Kulturfiltraten des Streptococ­cus pyogenes, Stamm NY-5. Beziehung zum erythrogenen Toxin Typ A. Zb!. Bakt. Hyg.A 261 (1986) 75-84

13. Hallas, G.: The production of pyrogenic exotoxins by group A streptococci. ]. Hyg.(Camb.) 95 (1985) 47-57

14. Holm, S. E., H. Knoll, D. Gerlach, ].-H. Ozegowski, and W. Kohler: The influence ofimmunization with ET type A on the infection of rabbits with viable streptococci. In:Kimura, Y., S. Kotami, and Y. Shiokawa (Eds.), Recent Advances in Streptococci andStreptococcal Diseases, pp. 141-143. Reedbooks, Chertsey (1985)

15. Hooker, S. B. and E. M. Follensby: Studies of scarlet fever. II. Different toxins producedby hemolytic streptococci of scarlatinal origin.]. Immuno!. 27 (1934) 177-193

16. Johnson, 1. P. and P. M. Schlievert: Phage T12 carries the structural gene for strep­tococcal pyrogenic exotoxin type A. Mol. Gen. Genet. 194 (1984) 52-56

17. Kim, Y. B. nd D. W. Watson: A purified group A streptococcal pyrogenic exotoxin.]'Exp. Med. 131 (1970) 611-628

18. Kirchner,]. 1. C. F. G.: Scarlatinae. Litteratura et historia. Diss., Halle (1844)19. Kjems, E.: Studies on streptococcal bacteriophages. 2. Adsorption, lysogenization, and

one-step growth experiments. Acta path. microbiol. scand. 42 (1958) 56-6620. Kohler, W.: Typisierung harnolysierender Steptokokken der Gruppe A. I. Mitt.: Metho­

den der Typisierung und Herstellung agglutinierender Sereno Z. Immun.-Forsch. 120(1960) 253-278

21. Kohler, W.: Typenverbreitung des Streptococcus pyogenes in der DDR. 1960 bis 1973.Dt. Gesundh.-Wes. 29 (1974) 966-976

22. Kohler, W.: Results of the second International Streptococcus pyogens type distributionsurvey. In: M. ]. Haverkorn (Ed.), Streptococcal Disease and the Community, pp.10-13. Elsevier, Amsterdam (1974)

23. Laemmli, U. K.: Cleavage of structural proteins during assembly of the head of bac­teriophage T4. Nature 227 (1970) 680-685

24. Louvard, D., C. Vannier, S. Marroux, ].-M. Pages, and C. Lazdunski: A quantitativeimmunochemical technique for evaluation of the extent of integration of membraneproteins and of protein conformational changes and homologies. Analyt. Biochem. 76(1976) 83-94

25. Nida, S. K. and]. ]. Ferretti: Phage influence on the synthesis of extracellular toxins ingroup A streptococci. Infect. Immun. 36 (1982) 745-750

26. 0rskov, F., I. 0rskov, D.]. Evans ir., R. B. Sack, D. A. Sack, and T. Wadstr6m: SpecialEscherichia coli serotypes among enterotoxigenic strains from diarrhoea in adults andchildren. Med. Microbiol. Immuno!. 162 (1976) 73-80

27. Parker, M. T.: Streptococcal disease.ln: Topley and Wilson's Principles of Bacteriology,Virology and Immunity, Vol. III, pp. 225-253 (p. 230). 7th ed. Arnold, London (1984)

28. Quinn, R. W.: Epidemiology of group A streptococcal infections - Their changingfrequency and severity. Yale J. BioI. Med. 55 (1982) 265-270

29. Statistisches ]ahrbuch der Deurschen Demokratischen Republik. Berlin (1986)30. Watson, D. W. and Y. B. Kim: Erythrogenic toxin. In: Montie, T. C, S. Kadis, and S.].

Ajl (Eds.), Microbial Toxins, '101. III, pp. 173-187. Academic Press New York-London(1970)

31. Weber, K. and M. Osborn: The reliability of molecular weight determination by dodecylsulfate-polyacrylamide gel electrophoresis.]. BioI. Chern. 244 (1969) 4406--4412

32. Weeks, C. R. and].]. Ferretti: The gene of type A streptococcal exotoxin (erythrogenictoxin) is located in bacteriophage T 12. Infect. Immun. 46 (1984) 531-536

Prof. Dr. Dr. Dr. h. C. Werner Kohler, Zentralinstitut fur Mikrobiologie und experimen­telle Therapie, Beutenbergstr. 11, DDR-6900 lena