Differentiation and Distribution of Three Types of Exfoliative Toxin Produced

FEMS Immunology and Medical Microbiology 20 (1998) 301^310

Di¡erentiation and distribution of three types of exfoliative toxin produced by Staphylococcus hyicus from pigs with exudative epidermitis

Lars Ole Andresen *

Department of Microbiology, Danish Veterinary Laboratory, Buëlowsvej 27, DK-1790 Copenhagen V, Denmark Received 20 December 1997; revised 19 February 1998; accepted 20 February 1998

Abstract

Exfoliative toxins of approximately 30 kDa produced by Staphylococcus hyicus strains NCTC 10350, 1289D-88 and 842A-88 were purified and specific polyclonal antisera were raised against each of the toxins. It was shown by immunoblot analysis and ELISA that three exfoliative toxins from S. hyicus were antigenically distinct. The three toxins were designated ExhA, ExhB and ExhC. From 60 diseased pigs, each representing an outbreak of exudative epidermitis, a total of 584 isolates of S. hyicus were phage typed and tested for production of exfoliative toxin. ExhA-, ExhB- and ExhC-producing S. hyicus isolates were found in 12 (20%), 20 (33%) and 11 (18%), respectively, of the 60 pig herds investigated. Production of the different types of exfoliative toxin was predominantly associated with certain phage groups. However, toxin production was found in all of the six phage groups defined by the phage typing system. Some changes in the distribution of isolates between phage groups were observed when the results of this study were compared to previous investigations. In this study two new antigenically distinct exfoliative toxins were isolated and tools for in vitro detection of toxin producing S. hyicus isolates and for further studies on the exfoliative toxins from S. hyicus have been provided. z 1998 Federation of European Microbiological Societies. Pub- lished by Elsevier Science B.V.

Keywords: Staphylococcus hyicus; Exfoliative toxin; Exudative epidermitis; Antigenic diversity; Autogenous vaccine

1. Introduction foliative toxin has been shown to be the factor re- sponsible for the alterations of the skin in EE in pigs The disease exudative epidermitis (EE) in pigs is [4] and the production of exfoliative toxin correlates caused by infection with Staphylococcus hyicus. Stud- with the ability of the strains to induce EE in pigs ies have shown that both virulent and avirulent S. [5]. hyicus can be isolated from diseased pigs [1]. The Staphylococcal scalded skin syndrome (SSSS) in virulent strains produce an approximately 27-kDa humans is a skin disease with aspects in common [2,3] or 30-kDa [1,4] exfoliative toxin. S. hyicus ex- with EE in pigs. SSSS and EE are both generalised skin infections that a¡ect young individuals and pig- * Tel.: +45 35 30 02 82; Fax: +45 35 30 01 20; lets, respectively. Patients with SSSS are usually E-mail: [email protected] under the age of 5 years [6] and EE a¡ects piglets

0928-8244 / 98 / $19.00 ß 1998 Federation of European Microbiological Societies. Published by Elsevier Science B.V. PII S0928-8244(98)00026-1

FEMSIM 864 27-4-98 302 L.O. Andresen / FEMS Immunology and Medical Microbiology 20 (1998) 301^310 in the age of 1^5 weeks, but mild forms of the dis- genically distinct exfoliative toxins were isolated and ease can be observed among older pigs [7]. The his- designated ExhA, ExhB and ExhC, respectively. Fur- topathological alterations of the skin in the two dis- thermore, the prevalence and distribution of the dif- eases are similar [6,8] and are in both cases caused by ferent toxins among S. hyicus isolates from pigs with toxins [3,4,9]. SSSS in humans is caused by infection EE were investigated. with Staphylococcus aureus producing antigenically distinct exfoliative toxins ETA and ETB, which are the factors that cause exfoliation of the skin [9]. 2. Materials and methods Studies on S. hyicus have shown that several phage types could be isolated simultaneously from the skin 2.1. Bacterial strains, growth media and culturing of diseased pig [1,10] and it was shown that only one conditions of the clones from diseased pigs colonised by multiple clones of S. hyicus could induce EE [1]. During Eighteen strains of S. hyicus previously reported recent years phage typing have been used for selec- [1] to be either virulent or avirulent were used in this tion of strains for production of autogenous vaccine study (see Table 1). Furthermore, a total of 584 S. in our laboratory. Eight to 10 isolates from submit- hyicus isolates were used for investigating the preva- ted material have been phage typed and one isolate lence and distribution of the three types of exfolia- of each phage type was selected for preparation of a tive toxin among di¡erent phage groups of S. hyicus. mixed autogenous vaccine. This procedure was chos- These 584 isolates were from 60 specimens from dis- en because there was no suitable laboratory method eased pigs each representing an outbreak of EE. Iso- available for selection of disease-causing clones, e.g. lates were randomly picked from primary plates in by testing for production of exfoliative toxin, and the sets of 8^10 isolates from each piglet with EE or skin use of multiple phage types from the same specimen samples from piglets with EE submitted to the Dan- should increase the probability of including the dis- ish Veterinary Laboratory during the period Decem- ease-causing clone in the vaccine. ber 1996 to July 1997. Recently, antigenic diversity among exfoliative After isolation and identi¢cation, S. hyicus was toxins produced by S. hyicus has been reported. Ta- grown on Columbia agar base (Oxoid, Unipath nabe et al. [5] have distinguished between two types Ltd, Basingstoke, UK) plates containing 5% bovine of exfoliative toxin by immunodi¡usion using anti- blood (C-blood agar). Liquid growth medium con- sera against exfoliative toxin produced by the S. hy- sisted of 30 g l31 Trypticase soy broth (BBL 11768, icus strains P-1 and P-23. In a recent study [4] it was Becton Dickinson and Co., Cockeysville, MD, USA) shown that polyclonal and monoclonal antibodies supplemented with 10 g l31 yeast extract (Oxoid against the exfoliative toxin produced by S. hyicus L21), pH was 7.2. Liquid cultures were grown in strain 1289D-88 only reacted with two of nine well tightly closed 10-ml tubes containing 5 ml of Trypti- characterised virulent S. hyicus strains in both immu- case-yeast extract medium at 37³C with shaking at noblot analysis and indirect ELISA. These observa- 130 rpm. Strains were stored at 380³C as overnight tions are important since the need for better diagnos- culture scraped from C-blood agar plates and sus- tic tools for selection of isolates for production of pended in liquid growth medium supplemented autogenous vaccines has become more pressing as with 10% glycerol. Strains from which exfoliative the number of outbreaks of EE in Danish pig herds toxin was puri¢ed were grown under small-scale fer- has increased during the past 10 years. mentation conditions as described previously [4]. The aim of the present study was to purify the putative exfoliative toxin produced by selected viru- 2.2. Isolation and identi¢cation lent strains of S. hyicus and to raise polyclonal and monoclonal antibodies against the toxins in order to Non-pigmented colonies isolated on selective/indi- identify the exfoliative toxin from these strains and cative medium [11] as described by Wegener [12] to investigate the antigenic diversity of the exfoliative were selected for further identi¢cation. In addition toxins from di¡erent strains of S. hyicus. Three anti- to being positive for lipase, and non-haemolytic on

FEMSIM 864 27-4-98 L.O. Andresen / FEMS Immunology and Medical Microbiology 20 (1998) 301^310 303

C-blood agar, isolates which were positive for hya- ammonium sulfate at 75% saturation, and the toxin luronidase [13], heat-stable nuclease [14] and catalase was puri¢ed by hydrophobic interaction chromatog- and gave a negative reaction for oxidase [15] were raphy on a phenyl-Sepharose CL-4B (Pharmacia, identi¢ed as S. hyicus. These criteria were su¤cient Uppsala, Sweden) column eluded by a gradient to distinguish S. hyicus from other relevant Sta- from 0.8 to 0.3 M (NH4)2SO4 in 25 mM KH2PO4, phylococci [16]. In particular S. hyicus was distin- pH 7.0. The exfoliative toxin was identi¢ed as a dis- guished from S. chromogenes by S. hyicus being tinct red-brown coloured band of approximately 30 able to hydrolyse Tween 80 (lipase activity) and hav- kDa in silver staining of SDS-PAGE gels [4]. Frac- ing a positive reaction for heat-stable nuclease and tions containing exfoliative toxin were pooled, dia- hyaluronidase [17]. lysed and further puri¢ed by anion exchange chromatography on a diethylaminoethyl (DEAE)- 2.3. Phage typing Sepharose CL-6B (Pharmacia) matrix using a linear elution gradient from 0.0 to 0.2 M NaCl in 10 mM

Phage typing and interpretation of phage typing NaH2PO4, pH 6.0. Fractions were analysed by SDS- results was performed according to [12] using a PAGE and proteins were visualised by silver staining phage typing system derived from a phage typing as described below. system described by Wegener [12]. Changes were that the phage typing system used in the present 2.5. Sodium dodecyl sulfate-polyacrylamide gel study comprised only 10 di¡erent lytic phages and electrophoresis (SDS-PAGE) and silver staining that the criteria for di¡erentiation between phage of proteins types were one `strong reaction di¡erence' (a `strong reaction di¡erence' being when a phage produced Analysis of the puri¢cation procedure was per- more than 50 plaques in one strain and no plaques formed by SDS-PAGE by the method of Laemmli in another strain). This phage typing system was [18]. SDS-PAGE gels were prepared from a 40% used for routine phage typing of the 584 S. hyicus acrylamide:N,NP-methylene-bis-acrylamide solution isolates included in this study. Each phage was de- (29:1, 3.3% C) (Bio-Rad, Hercules, CA, USA). Sam- nominated by a letter, and the letters of phages pro- ples were stacked in a 4% polyacrylamide (50 V, ducing more than 50 plaques with a given isolate constant voltage) gel and separated in a gel contain- designated the phage type of the isolate. The 10 ing 16.6% polyacrylamide (100 V, constant voltage). phages (B, C, D, H, F, G, P, U, X, and Q) ensured Protein bands were visualised by staining with silver the identi¢cation of phage types belonging to the based on the method of Morrissey [19] with some major phage groups de¢ned in the original 23-phage modi¢cations: the gel was ¢xed in 30% ethanol, typing system characterised by the following typical 12% acetic acid, 0.14% formaldehyde for a minimum phage reactions: I, A/B/C/W; II, D/H/I/R/S; III, E/ of 60 min, omitting the step of ¢xing the gel in glu- F/G/L/O/V; IV, long phage types. In the 10-phage taraldehyde. The gel was washed three times in 30% typing system the typical phage reactions of these ethanol for 10 min and soaked in 0.8 mM Na2S2O3 phage groups were: I, B/C; II, D; III, F/G; IV, D/ for 1 min. The gel was then washed three times for F/G/P/U. In both phage typing systems the phage 20 s in water and soaked in 0.2% silver nitrate, 0.20% groups Miscellaneous types and not typable (NT) formaldehyde for 20 min in the dark. The gel was were included. washed three times for 20 s in water and developed

with 566 mM Na2CO3,16WMNa2S2O3, 0.14% 2.4. Puri¢cation of exfoliative toxin formaldehyde. Staining was stopped by pouring o¡ the solution and soaking the gel in 50% ethanol, 12% Exfoliative toxin was puri¢ed from strains NCTC acetic acid for 10 min and then washing the gels in 10350 (ExhA), 1298D-88 (ExhB) and 842A-88 30% ethanol for 20 min. The protein size was deter- (ExhC), respectively, by a method previously de- mined by semi-logarithmic plots against standard scribed [4]. Brie£y, proteins from the culture super- molecular mass marker proteins (Bio-Rad Laborato- natant of the S. hyicus strain were precipitated by ries, Richmond, CA, USA).

FEMSIM 864 27-4-98 304 L.O. Andresen / FEMS Immunology and Medical Microbiology 20 (1998) 301^310

2.6. Immunoblotting mM Na2CO3, 140 mM NaHCO3, pH 9.6) was used for dilution of antigen preparations when coating Electrophoretic transfer of proteins from SDS- ELISA plates with heat-treated culture supernatant

PAGE gels to nitrocellulose membranes for immu- or (NH4)2SO4-precipitated culture supernatant. noblot analysis was performed by the semi-dry blot- ting method as described by Kyhse-Andersen [20]. 2.9. Screening for production of exfoliative toxin by Staining of immunoblots was performed using the ELISA and immunoblotting dioctyl sodium sulfosuccinate and 3,5,3P,5P-tetrame- thylbenzidine staining method described by Koch et A total of 584 phage typed S. hyicus isolates were al. [21] when horseradish peroxidase-conjugated sec- included in the investigation of the distribution of ondary antibodies were used and using 5-bromo-4- the three types of exfoliative toxin. All the isolates chloro-3-indolyl phosphate and nitroblue tetrazolium were tested for production of ExhA and ExhB in staining as described in [22] when alkaline phospha- indirect ELISA and for production of ExhA, ExhB tase-conjugated secondary antibodies were used. In and ExhC in immunoblotting. immunoblotting prestained SDS-PAGE standard Screening for production of exfoliative toxin by molecular mass marker proteins (Bio-Rad Laborato- ELISA was performed by the following procedure: ries) were used. SDS-PAGE and immunoblot analy- S. hyicus isolates were inoculated in 5 ml of liquid sis of the production of the three di¡erent exfoliative growth medium using a single colony from an over- toxins by the nine virulent and nine avirulent strains night culture on C-blood agar. After overnight incu- of S. hyicus was performed with samples of the bation the culture supernatant was isolated by cen- supernatant from overnight broth cultures and trifugation. Samples of supernatant were heated for from culture supernatant precipitated with 10 min in a boiling water bath and used for coating

(NH4)2SO4 as previously described [4] using ab- of ELISA plates. The heat-treated samples were ap- sorbed polyclonal antisera and monoclonal antibod- plied to the ELISA plates in a 1:2 dilution in 200 ies to the respective exfoliative toxins as primary mM carbonate bu¡er, pH 9.6 in duplicate (100 Wl/ antibodies. well) on each of two ELISA plates. On each ELISA plate toxin-positive control samples from S. hyicus 2.7. Preparation of polyclonal rabbit antisera and strains NCTC 10350 and 1289D-88 were included. mouse monoclonal antibodies Coated ELISA plates were left at 4³C overnight. Monoclonal antibodies MabEXH7.7 and Mab- Polyclonal rabbit antisera and mouse monoclonal EXH5.1 were used separately as primary antibodies antibodies to the puri¢ed exfoliative toxins from for detecting ExhA and ExhB, respectively. Bound strains NCTC 10350 (ExhA), 1289D-88 (ExhB), primary antibodies were detected using a horseradish and polyclonal rabbit antiserum to the exfoliative peroxidase-conjugated F(abP)2 fragment of polyclo- toxin from strain 842A-88 (ExhC) were prepared nal sheep antibodies to mouse Ig (NA9310, Amer- by methods previously described [4]. Monoclonal sham, UK) and o-phenylenediamine and H2O2 as antibodies to ExhA and ExhB used in this study substrates. were designated MabEXH7.7 and MabEXH5.1, re- Samples of culture supernatant from all of the 584 spectively. Polyclonal rabbit antisera were absorbed S. hyicus isolates were also tested in immunoblotting with (NH4)2SO4-precipitated culture supernatant using MabEXH7.7, MabEXH5.1 and polyclonal from the avirulent S. hyicus strain 1403B-88 before rabbit antiserum raised against puri¢ed ExhC as pri- they were used in immunoblotting. mary antibodies for detection of ExhA, ExhB and ExhC, respectively. Screening for production of ex- 2.8. Indirect enzyme-linked immunosorbent assay foliative toxin by immunoblotting was performed as (ELISA) described above using precast 10% NuPAGE gels (NOVEX, San Diego, CA, USA) run in MOPS run- Indirect ELISA was performed as previously de- ning bu¡er according to the manufacturer's instruc- scribed [4], except that 200 mM carbonate bu¡er (60 tions. Secondary antibodies to either mouse or rabbit

FEMSIM 864 27-4-98 L.O. Andresen / FEMS Immunology and Medical Microbiology 20 (1998) 301^310 305 immunoglobulins were alkaline phosphatase conju- exfoliative toxin from strain 1289D-88 and polyclo- gated (D0486 or D0487, Dako, Glostrup, Denmark). nal rabbit antibodies were raised against this protein. Immunoblot analysis showed that the antibodies raised against the 30-kDa protein from strain 3. Results NCTC10350 reacted with a 30-kDa protein band produced by ¢ve out of the nine virulent strains, as 3.1. Di¡erentiation of exfoliative toxins shown in Fig. 1A. The polyclonal and monoclonal antibodies raised against this protein did not react It has previously been shown [4] that both poly- with any 30-kDa proteins produced by any of the clonal and monoclonal antibodies raised against the nine avirulent strains (Table 1). The puri¢cation pro- exfoliative toxin from S. hyicus strain 1289D-88 only cedure and the results of the immunoblot analysis reacted with strains 1403E-88 and 1289D-88 of the strongly indicated that this 30-kDa protein was an nine virulent strains examined, as illustrated in Fig. exfoliative toxin, similar to but antigenically distinct 1B. Therefore, an approximately 30-kDa protein from the exfoliative toxin from strain 1289D-88. from S. hyicus strain NCTC 10350 was puri¢ed by Strains 842A-88 and A2869C produced toxin that the same procedure as used for puri¢cation of the did not react with the antibodies raised against the

Table 1 Origin of strains and type of exfoliative toxin produced by known virulent and avirulent strains of S. hyicus Straina Origin Reference Type of toxin producedd Countryb Sourcec Virulent strains NCTC 10350 DK 1 [28,29] ExhA 9390-88 DK 2 [30] ExhA P411 UK 3 [31] ExhA 1403E-88 DK 2 [1] ExhB 1289D-88 DK 2 [1] ExhB 842A-88 DK 2 [1] ExhC P119 UK 3 [32] ExhA S3588 G 4 [33] ExhA A2869C G 4 [33] ExhDe Avirulent strains A3793/76 G 4 [33] ntdf A4596/76 G 4 [33] ntd A72/75 G 4 [33] ntd 842G-88 DK 2 [1] ntd 842J-88 DK 2 [1] ntd 1403B-88 DK 2 [1] ntd 1403H-88 DK 2 [1] ntd 1289E-88 DK 2 [1] ntd SK170 UK 3 [31] ntd aAll strains were isolated from pigs with exudative epidermitis, except strains A3793 (arthritis), A4569/76 (necrosis) and A72/75 (mastitis) [1]. bDK = Denmark, UK = United Kingdom, G = Germany. c1: D. Sombolinsky and 2: H.C. Wegener, Danish Veterinay Laboratory, Copenhagen, Denmark; 3: R. Allaker, Royal Veterinary College, London, UK; 4: G. Amtsberg, Tieraërtzliche Hochschule Hannover, Germany. dAntigenically distinct exfoliative toxins puri¢ed from strains NCTC 10350, 1289D-88 and 842A-88 were designated ExhA, ExhB and ExhC, respectively. Strains producing exfoliative toxin reactive with speci¢c antibodies to ExhA, ExhB or ExhC were assigned the respective toxin type. eStrain A2869C produced a toxin antigenically distinct from ExhA, ExhB and ExhC which provisionally was designated ExhD. This toxin was not puri¢ed or charaterised in this study. f ntd = no toxin detected.

FEMSIM 864 27-4-98 306 L.O. Andresen / FEMS Immunology and Medical Microbiology 20 (1998) 301^310

Fig. 1. Immunoblot analysis showing the production of the exfoliative toxins ExhA, ExhB and ExhC among nine virulent S. hyicus strains. Overnight culture supernatants from nine well de¢ned virulent S. hyicus strains were subjected to SDS-PAGE and immunoblotting using polyclonal rabbit antiserum raised against the exfoliative toxin puri¢ed from S. hyicus strain NCTC 10350 (ExhA) in A, strain 1289D-88 (ExhB) in B, strain 842A-88 (ExhC) in C and a monoclonal antibody MabEXH7.7 raised against exfoliative toxin from strain NCTC 10350 (ExhA) in D. Lane M: molecular mass marker; lane 1: strain NCTC 10350; lane 2: strain 9390-88; lane 3: strain P411; lane 4: strain 1403E-88; lane 5: strain 1289D-88; lane 6: strain 842A-88; lane 7: strain P119; lane 8: strain S3588; lane 9: strain A2896C. exfoliative toxin from strain NCTC 10350 or from ExhC, respectively. The molecular mass of the puri- strain 1289D-88. The exfoliative toxin from strain ¢ed exfoliative toxins ExhA, ExhB and ExhC was 842A-88 was then puri¢ed by the same method as estimated to be approximately 30 kDa by SDS- the other two exfoliative toxins and the polyclonal PAGE analysis. In Table 1 the distribution of the rabbit antibodies raised against this toxin only re- three antigenically distinct types of exfoliative toxin acted with the toxin from strain 842A-88 (Fig. 1C). among the nine virulent strains is shown. None of Attempts to purify the exfoliative toxin from strain the polyclonal or monoclonal antibodies that were A2869C by the same method as used for the other raised against the puri¢ed exfoliative toxins for toxins resulted in isolating the toxin in fractions con- strains NCTC 10350, 1289D-88 or 842A-88 reacted taining other proteins that did not bind to the with any 30-kDa proteins produced by the avirulent DEAE-Sepharose matrix. Isolation of the toxin in strains included in this study (Table 1, immunoblots pure form was not achieved by substituting the not shown). The results of the immunoblot analyses DEAE-Sepharose anion exchange chromatography were con¢rmed by indirect ELISA using the mono- puri¢cation step with cation exchange chromatogra- clonal antibodies against ExhA and ExhB. phy using CM-Sepharose. In this study a monoclonal antibody, designated The antigenically distinct exfoliative toxins pro- MabEXH7.7, was produced against the puri¢ed ex- duced by S. hyicus strains NCTC 10350, 1298D-88 foliative toxin from strain NCTC 10350. In ELISA, and 842A-88 were designated ExhA, ExhB and MabEXH7.7 reacted strongly with supernatant from

FEMSIM 864 27-4-98 L.O. Andresen / FEMS Immunology and Medical Microbiology 20 (1998) 301^310 307

Table 2 Distribution of the three di¡erent types of exfoliative toxin produced by di¡erent S. hyicus phage groups Phage groupa No. (%) of isolates producingb No. (%) of isolates not producing toxin Total no. (%) of isolates ExhA ExhB ExhC I 19 (3.3) 95 (16.3) 0 (0.0) 8 (1.4) 122 (20.9) II 0 (0.0) 0 (0.0) 3 (0.5) 54 (9.2) 57 (9.8) III 2 (0.3) 0 (0.0) 1 (0.2) 19 (3.3) 22 (3.8) IV 6 (1.0) 0 (0.0) 48 (8.2) 36 (6.2) 90 (15.4) Miscellaneous 0 (0.0) 2 (0.3) 3 (0.5) 70 (12.0) 75 (12.8) NTc 39 (6.7) 36 (6.2) 14 (2.4) 129 (22.1) 218 (37.3) Total 66 (11.3) 133 (22.8) 69 (11.8) 316 (54.1) 584 (100) From 60 diseased pigs 8^10 colonies of S. hyicus were randomly isolated, phage typed, and tested for production of exfoliative toxin. aPhage groups were de¢ned acording to Wegener [12]. bExhA and ExhB were detected in both ELISA and immunoblotting. ExhC was detected in immunoblotting. cNT = not typable.

ExhA producing isolates (OD values s 1.0) and had were both toxin-producing isolates and isolates not weak reactions (OD values typically in the range of producing exfoliative toxin. In four cases, NT iso- 0.1^0.5) with supernatant from isolates producing lates from the same specimen were either producing ExhB or ExhC. Therefore, isolates were considered ExhA or not producing ExhA, and in one case this positive for ExhA if OD was more than 1.0. Immu- was observed for NT isolates either producing or not noblot analysis (Fig. 1D) of culture supernatant producing ExhB. The results of the phage typing, from the nine virulent strains using MabEXH7.7 as indirect ELISA and immunoblotting analysis are primary antibody showed that this monoclonal anti- summarised in Table 2. Altogether, 248 (45.9%) of body reacted with ExhA, ExhB and ExhC. The re- the 584 S. hyicus isolates were able to produce one of action with ExhA was more intense compared to the the three exfoliative toxins tested for in this study. reaction with ExhB and ExhC similar to what was The majority of the toxin producing isolates be- observed in the indirect ELISA. Furthermore, immu- longed to phage groups I (19.6%), IV (9.2%) and noblot analysis showed that ExhC was slightly small- to the NT group (15.3%). er in molecular mass compared to ExhA and ExhB, and that the toxin produced by strain A2869C was not detected by MabEXH7.7. 4. Discussion

3.2. Distribution of exfoliative toxins In this study the production of exfoliative toxin from 18 strains of S. hyicus was studied. Nine of The 584 isolates of S. hyicus representing 60 pig these 18 strains were previously by experimental in- herds with EE were screened for the production of fection shown to be virulent and nine strains were exfoliative toxins ExhA, ExhB and ExhC. The results avirulent [1]. We have recently shown that the exfo- of the screening showed that ExhA, ExhB and ExhC liative toxin from S. hyicus strain 1289D-88 is a pro- were produced in 12 (20%), 20 (33%) and 11 (18%), tein of approximately 30 kDa [4]. This toxin could in respectively, of the 60 pig herds investigated. In 17 pure form induce the characteristic skin alterations (28%) of the herds no toxin-producing isolates were of EE in pigs. In the present study a 30-kDa protein found among the 8^10 S. hyicus isolated from the from each of the well described virulent S. hyicus submitted material. Production of more than one strains NCTC 10350 and 842A-88 was puri¢ed by type of exfoliative toxin among the 8^10 isolates the same method as used for puri¢cation of the 30- from any of the specimens was not observed. In kDa exfoliative toxin from S. hyicus strain 1289D- three cases, one for each type of toxin, isolates 88. It was also shown by immunoblotting that the with the same phage type from one diseased pig 30-kDa protein from strain NCTC 10350 was com-

FEMSIM 864 27-4-98 308 L.O. Andresen / FEMS Immunology and Medical Microbiology 20 (1998) 301^310 mon to ¢ve of the nine known virulent strains (Table types of S. aureus have been studied [26,27] and it 1 and Fig. 1A) and the monoclonal antibody Mab- was shown that the ability to produce exfoliative EXH7.7 raised against this 30-kDa protein reacted toxin was not restricted to particular phage groups with proteins of approximately 30 kDa produced by of S. aureus. This is consistent with the ¢ndings for eight of the nine virulent strains (Fig. 1D). The ex- S. hyicus in this study. Furthermore, it has been foliative toxin from S. hyicus strain NCTC 10350 shown that isolates of S. aureus can produce both was previously partially puri¢ed and its biological ETA and ETB concurrently [26,27]. Simultaneous activity was demonstrated in a fraction containing expression of two types of exfoliative toxin from S. eight proteins including two proteins of 27 and 30 hyicus has been reported by Tanabe et al. [5]. How- kDa which were unique to strain NCTC 10350 com- ever, this phenomenon was not observed in the pared to an avirulent strain [23]. It was concluded present study. that either of these proteins could be the exfoliative In Fig. 1D, MabEXH7.7 reacts with what appears toxin of strain NCTC 10350. In the present study, to be a triple band in the lanes with supernatant none of the antibodies raised against the 30-kDa from ExhA-producing strains which could be due proteins from strains NCTC 10350, 1289D-88 or either to some minor heterogeneity in the toxin or 842A-88 reacted with any 30-kDa proteins produced to degradation of the toxin. This also demonstrates by the nine avirulent strains. Altogether, these results that the toxins have at least one epitope in common strongly indicate that the 30-kDa proteins from which elicited an antibody response in mouse but strains NCTC 10350 and 842A-88 isolated and anti- apparently not in rabbit. genically characterised in this study are two new In 72% of the 60 pig herds investigated toxin-pro- antigenic variants of the exfoliative toxin from S. ducing (toxigenic) S. hyicus isolates were found hyicus. However, it remains to be shown that the among the 8^10 isolates tested. In the remaining puri¢ed 30-kDa proteins from strains NCTC 10350 28% of the herds the reason for not detecting toxi- and 842A-88 can also induce the characteristic skin genic S. hyicus isolates could either be that the num- alterations of EE in pigs. ber of isolates tested was too small to ensure detec- The antigenically distinct exfoliative toxins from tion of the toxigenic isolates or that the toxigenic strains NCTC 10350, 1289D-88 and 842A-88 were isolates in these herds were producing a toxin or designated ExhA, ExhB and ExhC, respectively, in toxins that were not detected by the antibodies order to clearly distinguish between the S. hyicus used in this study. exfoliative toxins and the exfoliative toxins of S. au- The investigation of the prevalence and distribu- reus. The exfoliative toxin from S. hyicus strain tion of the three toxins among S. hyicus isolates from A2869C did not react with any of the polyclonal pigs with EE showed that exfoliative toxin was pro- or monoclonal antibodies raised against ExhA, duced by several phage types and that at least one of ExhB or ExhC. Thus, at least one more antigenically the types of exfoliative toxin was produced in each of distinct exfoliative toxin, provisionally designated the six phage groups de¢ned in the phage typing ExhD, may be produced by virulent strains of S. system. It was also shown that occasionally isolates hyicus. with the same phage type isolated from the same In the case of S. aureus, strains can be divided into diseased pig could be either toxigenic or not produc- producers and non-producers with respect to exfolia- ing toxin (non-toxigenic). The distribution of toxi- tive toxin. The existence of di¡erent types of exfolia- genic isolates and of the di¡erent types of exfoliative tive toxin from S. aureus has been recognised and so toxin among the di¡erent phage types and phage far, three types of exfoliative toxin have been re- groups may indicate that the ability to produce toxin ported. The exfoliative toxins of S. aureus, ETA is associated with a mobile genetic element, e.g. and ETB, and their signi¢cance in SSSS in humans phage, plasmid or transposon, but this remains to have been thoroughly reviewed [6,9,24]. A third type be elucidated. However, the production of exfoliative of exfoliative toxin from a S. aureus strain of equine toxin was predominantly associated with a few fre- origin has been identi¢ed and designated sETC [25]. quently occurring phage types. Production of ExhA The distribution of ETA and ETB among phage and ExhB was predominantly associated with phage

FEMSIM 864 27-4-98 L.O. Andresen / FEMS Immunology and Medical Microbiology 20 (1998) 301^310 309 group I and the NT isolates. Production of ExhC Acknowledgments was predominantly associated with phage group IV (long types). This observation may re£ect that cer- The excellent technical assistance of Heidi Pia tain clones belonging to these phage groups may be Andersen throughout this study is gratefully appre- more widespread or epidemic among the pig herds ciated. I am grateful to Regina Lund for preparation investigated. In previous studies [10,12] isolates be- and culturing of hybridoma cells, to Ole SÖrensen longing to phage group I were also found more fre- and Susanne M. Ranebo for identi¢cation and quently on diseased pigs. In these investigations the phage typing S. hyicus isolates, to Jane N. Larsen long phage types in phage group IV were not as for assisting in screening for toxin production frequently isolated from pigs with EE (7% and and to Henrik C. Wegener Ph.D. for valuable 3.4%, respectively) as in the present study (15.4%). discussions and critical reading of the manuscript. Of the phage group IV isolates found in this study 53.3% were producing ExhC. These results may indicate that the ExhC-producing clones belonging to References phage group IV may be clones that have emerged since the other investigations which were conducted [1] Wegener, H.C., Andresen, L.O. and Bille-Hansen, V. (1993) on isolates collected during 1985^1988 [12] (Wege- Staphylococcus hyicus virulence in relation to exudative epi- ner, personal communication) and 1989^1991 [10], dermitis in pigs. Can. J. Vet. Res. 57, 119^125. respectively. [2] Sato, H., Tanabe, T., Kuramoto, M., Tanaka, K., Hashimoto, The antibodies and screening techniques developed T. and Saito, H. (1991) Isolation of exfoliative toxin from Staphylococcus hyicus subsp. hyicus and its exfoliative activity during this study may be used for routine selection in the piglet. Vet. Microbiol. 27, 263^275. of toxigenic S. hyicus isolates for preparation of au- [3] Tanabe, T., Sato, H., Kuramoto, M. and Saito, H. (1993) togenous vaccine. Puri¢cation of exfoliative toxin produced by Staphylococcus In conclusion, this study presents the isolation of hyicus and its antigenicity. Infect. Immun. 61, 2973^2977. what seem to be two new types of exfoliative toxin of [4] Andresen, L.O., Bille-Hansen, V. and Wegener, H.C. (1997) Staphylococcus hyicus exfoliative toxin: puri¢cation and dem- approximately 30 kDa from S. hyicus. Although the onstration of antigenic diversity among toxins from virulent biological activity of the exoproteins isolated in this strains. Microb. Pathogen. 22, 113^122. study was not demonstrated the indirect evidence for [5] Tanabe, T., Sato, H., Sato, H., Watanabe, K., Hirano, M., them being exfoliative toxins, i.e. the antigenic cross- Hirose, K., Kurokawa, S., Nakano, K., Saito, H. and Mae- reactivity of MabEXH7.7, the common puri¢cation hara, N. (1996) Correlation between occurrence of exudative epidermitis and exfoliative toxin-producing ability of Sta- procedure and the combined results of [4], [23] and phylococcus hyicus. Vet. Microbiol. 48, 9^17. the present study, was regarded as su¤cient for des- [6] Elias, P.M., Fritsch, P. and Epstein, E.H. (1977) Staphylococ- ignating the three proteins as exfoliative toxins cal scalded skin syndrome. Arch. Dermatol. 113, 207^219. ExhA, ExhB and ExhC, respectively. It was shown [7] Taylor, D.J. (1995) Pig Diseases, 6th edn. D.J. Taylor, Glas- that virulent S. hyicus produces one of at least three gow. [8] Obel, A.-L. (1968) Epithelial changes in porcine exudative antigenically distinct exfoliative toxins. The produc- epidermitis. The light-microscopical picture. Pathol. Vet. 5, tion of the three types of exfoliative toxin were pre- 253^269. dominantly associated with certain S. hyicus phage [9] Arbuthnott, J.P. (1983) Epidermolytic toxins. In: Sta- groups, however toxigenic isolates could be assigned phylococci and Staphylococcal Infections 2 (Easmon, C.S.F to each of the six phage groups de¢ned by the phage and Adlam, C., Eds.), pp. 599^617. Academic Press, Lon- don. typing system. Some changes in the distribution of [10] Wegener, H.C. (1993) Diagnostic value of phage typing, anti- isolates between the phage groups were observed and biogram typing, and plasmid pro¢ling of Staphylococcus hyi- may be due to the emergence of speci¢c phage types. cus from piglets with exudative epidermitis. J. Vet. Med. B 40, Finally, this study has provided reagents and meth- 13^20. ods for selecting toxigenic S. hyicus isolates for the [11] Devriese, L.A. (1977) Isolation and identi¢cation of Staphy- lococcus hyicus. Am. J. Vet. Res. 38, 787^792. production of autogenous vaccines and for further [12] Wegener, H.C. (1993) Development of a phage typing system studies on the prevalence, diversity and nature of for Staphylococcus hyicus. Res. Microbiol. 144, 237^244. the exfoliative toxins of S. hyicus. [13] Devriese, L.A., Schleifer, K.H. and Adegoke, G.O. (1985)

FEMSIM 864 27-4-98 310 L.O. Andresen / FEMS Immunology and Medical Microbiology 20 (1998) 301^310

Identi¢cation of coagulase-negative staphylococci from farm [24] Iandolo, J.J. (1989) Genetic analysis of extracellular toxins animals. J. Appl. Bacteriol. 58, 45^55. of Staphylococcus aureus. Annu. Rev. Microbiol. 43, 375^ [14] Lachica, R.V.F., Genigeorgis, C. and Hoeprich, P.D. (1971) 402. Metachromatic agar-di¡usion methods for detecting staphylo- [25] Sato, H., Matsumori, Y., Tanabe, T., Saito, H., Shimizu, A. coccal nuclease activity. Appl. Microbiol. 21, 585^587. and Kawano, J. (1994) A new type of staphylococcal exfolia- [15] Faller, A. and Schleifer, K.-H. (1981) Modi¢ed oxidase and tive toxin from a Staphylococcus aureus strain isolated from a benzidine tests for separation of staphylococci from micrococ- horse with phlegmon. Infect. Immun. 62, 3780^3785. ci. J. Clin. Microbiol. 13, 1031^1035. [26] Kondo, I., Sakurai, S., Sarai, Y. and Futaki, S. (1975) Two [16] Holt, J.G., Kreig, N.R., Sneath, P.H.A., Staley, J.T. and Wil- serotypes of exfoliatin and their distribution in staphylococcal liams, S.T. (1994) Bergey's Manual of Determinative Bacteri- strains isolated from patients with scalded skin syndrome. ology, 9th edn. Williams and Wilkins, Baltimore, MD. J. Clin. Microbiol. 1, 397^400. [17] Haèjek, V., Devriese, L.A., Mordarski, M., Goodfellow, M., [27] Azavedo, J.d. and Arbuthnott, J.P. (1981) Prevalence of epi- Pulverer, G. and Varaldo, P.E. (1986) Elevation of Staphylo- dermolytic toxin in clinical isolates of Staphylococcus aureus. coccus hyicus subsp. chromogenes (Devriese et al., 1978) to J. Med. Microbiol. 14, 341^344. species status: Staphylococcus chromogenes (Devriese et al., [28] Sompolinsky, D. (1950) Impetigo contagiosa suis. Maa- 1978) comb. nov. Syst. Appl. Microbiol. 8, 169^173. nedsskr. Dyrlaeger 61, 401^453. [18] Laemmli, U.K. (1970) Cleavage of structural proteins during [29] Devriese, L.A., Haèjek, V., Oeding, P., Meyer, S.A. and Schlei- the assembly of the head of bacteriophage T4. Nature 227, fer, K.H. (1978) Staphylococcus hyicus (Sompolinsky 1953) 680^685. comb. nov. and Staphylococcus hyicus subsp. chromogenes [19] Morrissey, J.H. (1981) Silver stain for proteins in polyacryl- subsp. nov. Int. J. Syst. Bacteriol. 28, 482^490. amide gels: a modi¢ed procedure with enhanced uniform sen- [30] Wegener, H.C. (1990) Studies on Staphylococcus hyicus viru- sitivity. Anal. Biochem. 117, 307^310. lence in relation to exudative epidermitis in piglets. In: [20] Kyhse-Andersen, J. (1984) Electroblotting of multiple gels: a Tagung der Fachgruppe Bakteriologie und Bakterielle Krank- simple apparatus without bu¡er tank for rapid transfer of heiten (Blobel, H., Ed.), pp. 118^131. Deutsche Vete- proteins from polyacrylamide to nitrocellulose. J. Biochem. rinaërmedizinische Gesellschaft e.V., Giessen. Biophys. Methods 10, 203^209. [31] Allaker, R.P., Lloyd, D.H. and Smith, I.M. (1988) Prevention [21] Koch, C., SkjÖdt, K. and Laursen, I. (1985) A simple immu- of exudative epidermitis in gnotobiotic piglets by bacterial noblotting method after separation of proteins in agarose gel. interference. Vet. Rec. 123, 597^598. J. Immunol. Methods 84, 271^278. [32] Allaker, R.P., Whitlock, M. and Lloyd, D.H. (1991) Cytotox- [22] Harlow, E. and Lane, D. (1988) Antibodies, A Laboratory icity activity of Staphylococcus hyicus. Vet. Microbiol. 26, Manual. Cold Spring Harbor Laboratory, Cold Spring Har- 161^166. bor, NY. [33] Amtsberg, G. (1979) Nachweis von Exfoliation ausloësenden [23] Andresen, L.O., Wegener, H.C. and Bille-Hansen, V. (1993) Substanzen in Kulturen von Staphylococcus hyicus des Staphylococcus hyicus skin reactions in piglets caused by crude Schweines und Staphylococcus epidermidis Biotyp 2 des extracellular products and by partially puri¢ed exfoliative tox- Rindes. Zbl. Vet. Med. B. 26, 257^272. in. Microb. Pathogen. 15, 217^225.

FEMSIM 864 27-4-98