Comparative antigenic analysis of extracellular proteins of Bacteroides nodosus isolated from virulent and benign ovine footrot

Veterinary Microbiology,12 (1986) 135--145 Elsevier Science Publishers B.V., Amsterdam -- Printed in The Netherlands

135

COMPARATIVE ANTIGENIC ANALYSIS OF EXTRACELLULAR PROTEINS OF BACTEROIDES NODOSUS ISOLATED FROM VIRULENT AND BENIGN OVINE FOOTROT

W.K. YONG and L.M. GORDON Department of Agriculture and Rural Affairs, Regional Veterinary Laboratory, P.O. Box 406, Hamilton, Victoria, 3300 (Australia) (Accepted for publication 15 October 1985)

ABSTRACT Yong, W.K. and Gordon, L.M., 1986. Comparative antigenic analysis of extracellular proteins of Bacteroides nodosus isolated from virulent and benign ovine footrot. Vet. Microbiol., 12: 135--145. Antigens in the extracellular protein (ECP) complexes of Bacteroides nodosus, isolated from sheep with either benign or virulent footrot, were studied by immunoelectrophoresis (IEP). Rabbit antisera against ECP from virulent and benign strains, were used in homologous and heterologous crossed IEP. Four precipitin peaks unique to the virulent strain, and five peaks unique to the benign strain were identified. In an attempt to characterize the different antigens in ECP, rabbit antisera were raised against an outer membrane protein (OMP, tool. wt. 35 000 daltons), pill and various proteases of virulent and benign strains of B. nodosus. No precipitin band was observed when ECP from both B. nodosus strains were reacted against anti-OMP and anti-pilus antisera. However, single precipitin bands unique to one protease from the benign strain and one protease from the virulent strain were identified. The results suggest that specific antigens other than proteases or pill are important in determining whether a B. nodosus isolate is virulent or benign.

INTRODUCTION B a c t e r o i d e s n o d o s u s has been c h a r a c t e r i z e d f r o m cases o f virulent f o o t r o t in sheep (Beveridge, 1 9 4 1 ) , benign f o o t r o t in sheep ( T h o m a s , 1 9 6 2 ; E g e r t o n and P a r s o n s o n , 1 9 6 9 ) and interdigital dermatitis in cattle (Laing and Egert o n , 1978). C o m p a r a t i v e studies have been m a d e on B. n o d o s u s isolates to d e t e r m i n e w h e t h e r the virulent ovine strains have a n y distinguishing u l t r a s t r u c t u r a l or b i o c h e m i c a l characteristics t h a t m a y be associated with their p a t h o g e n i c i t y in sheep ( E g e r t o n and Parsonson, 1 9 6 9 ; Walker et al., 1 9 7 3 ; S h o r t et al., 1976). E l e c t r o n m i c r o s c o p y has revealed t h a t organisms w h i c h caused virulent f o o t r o t have a b u n d a n t pili or fimbrial structures, diffuse p o l a r material and an a d d i t i o n a l layer o n the cell surface (Walker et al., 1 9 7 3 ; S h o r t et al., 1 9 7 6 ; Stewart, 1 9 7 8 ; E v e r y and S k e r m a n , 1980,

0378-1135/86/$03.50

© 1986 Elsevier Science Publishers B.V.

136 1983). Virulence is also associated with the stability and range of extracellular proteases produced (Egerton and Parsonson, 1969; Depiazzi and Richards, 1979; Stewart, 1979; Skerman et al., 1981; Every 1982; Kortt et al., 1983). This paper describes a different approach in determining whether there are other associated virulence factors which correlate with the pathogenicity of the B. nodosus isolate in question. A comparative immunoelectrophoretic study of the extracellular proteins (ECP) excreted/secreted by B. nodosus was made using antisera from rabbits immunized with these proteins. Partial success in identification of ECP antigens unique to either a virulent or a benign strain is reported here. MATERIALS AND METHODS Bacterial strains and culture methods B. nodosus strains 198 (virulent) and 305 (benign) of ovine origin were kindly supplied by D.J. Stewart, CSIRO Division of Animal Health, Parkville, Victoria, Australia. These were grown anaerobically in an atmosphere of 90% H2 and 10% CO2 at 37°C in modified trypticase-arginine-serine broth (Gordon et al., 1985). Prepara tion o f ex tracellular pro teins (ECP)

Four litres of culture supernatant were concentrated to 100 ml at 4°C in a DC2 hollow fibre diafiltration/concentration system fitted with a HIP5-20 cartridge (Amicon Corporation, USA). The concentrate was clarified by centrifugation at 10,000 × g for 20 min and the proteins extracted by precipitation with 60% saturated ammonium sulphate (pH 7.2). The precipitate was dissolved in 20 ml 20 mM Tris-HC1/5 mM CaC12 (pH 8) and exhaustively dialyzed with the same buffer. The protein concentration was determined by the method of Lowry et al. (1951) using bovine serum albumin as the standard. Enzyme activity was measured by radial diffusion into 1-mm thick agatose (1.5% w/v) containing gelatin (0.5%, w/v) and using Clostridium histolyticum collagenase Type III (Sigma Chemical Co., USA) as the standard (Santarius and Ryan, 1977). The ECP was stored in 1-ml aliquots at -20°C. Isolation o f proteases

Separation of individual proteases was performed using an electrophoreticzymogram method (Kortt et al., 1983) modified for large scale isolation of each enzyme. A central sample well (90 mm wide) and two outside wells (10 mm wide) were cut in the PAGE slab gel (180 × 160 × 3 mm; Protean Cell, Bio-Rad Laboratories, USA). ECP preparation was mixed

137 1 : 10 with glycerol and 50/~1 of this mixture (2 units of protease activity) was applied to each of the outside wells, and 500 pl (20 units of protease activity) to the central well. Electrophoresis was carried out at 25 mA constant current per gel at 12°C until the tracking dye (bromophenol blue) was 1 cm from the bottom of the gel. After electrophoresis, proteolytic activity was detected by a modified zymogram technique (Foissy, 1974; Kortt et al., 1983). The gel slab was overlaid with two narrow gelatin-agarose (0.5%--1.5% w/v) strips 2 mm thick. These were placed on the slab gel so as to coincide with the two lateral 10-mm well lanes. Enzyme bands appeared on the gelatin-agarose overlay as clear zones against a milky-white background. Individual bands were identified in the parallel strips by their relative mobility (Rf)values calculated using the tracking dye front as the reference point and the top of the resolving gel as the origin. The same enzymes were located and excised from the 90 mm central lane by extrapolation of Rf values from the two laterally placed developed gelatin-agarose strips. The purity of each isolated enzyme was checked by the presence of a single band of enzyme activity and its Rf value when a sample of the excised gel was loaded onto another slab gel and re-electrophoresed. The homogeneity of each excised enzyme band was further demonstrated by staining the PAGE slab gel with 2% Commassie Blue R250 (Bio-Rad Laboratories, USA). The protein content of individual enzymes from a single gel was estimated (Lowry et al., 1951) after recovery by an electroelution technique (Walker et al., 1982), and was shown to range from 10 to 50 pg. The protein concentration of enzymes subsequently isolated under similar conditions and from the same batch of ECP was assumed to be identical. The excised gel slices were stored at -20°C until used for immunization. A n tisera

ECP from virulent B. n o d o s u s were emulsified with an equal volume of complete Freund's adjuvant (Difco Laboratories, USA). Each of two 6-month-old rabbits was immunized with 0.5 ml (0.1 mg protein) of this mixture by intramuscular injection. The rabbits were given booster injections containing Freund's incomplete adjuvant weekly for 6 weeks, starting 3 weeks after the initial injection. Each of two rabbits was similarly immunized with ECP from benign B. nodosus. Five ml of blood were collected from the ear vein of each rabbit before and 18 days after the first injection, and then 4 days after each booster injection. Each enzyme-containing gel slice was homogenized in a glass tissue grinder with 0.5 ml 20 mM Tris-HC1/5 mM CaC12 and 0.5 ml incomplete Freund's adjuvant. Rabbits were immunized intramuscularly with 0.5 ml of this mixture, using the same protocol as for ECP preparations. Rabbit antisera specific against pilus and an outer membrane protein

138 (OMP, mol. wt. 35 000 daltons) of virulent strain 198, prepared as described by Mattick et al. (1984) and Emery et al. (1984), respectively, were kindly supplied by E.K. Moses, Regional Veterinary Laboratory, Hamilton. All serum samples were stored at -20°C after removal from blood clot.

Immunoelectrophoretic analysis Specific antibodies in rabbit antisera were analyzed by agarose immunoelectrophoresis (IEP) (Grabar and Williams, 1953), crossed IEP (Clarke and Freeman, 1968), and tandem crossed IEP (Kroll, 1969). For the IEP test, 5 pl of antigen (1.5 mg protein m1-1) were put in the well and antiserum in the trough was used undiluted. The antigen-antibody reaction was allowed to proceed at room temperature in a humid chamber for 48 h. The slide was washed in saline, rinsed with distilled water, dried at 37°C, stained with 0.01% (w/v) amido black and then photographed. For the crossed IEP and tandem crossed IEP tests the antiserum was used at a final concentration of 10% (v/v) in the second dimension gel, and the slides were washed and processed as described before. RESULTS

Comparative antigenic characterization of ECP Antigens in ECP preparations from a benign and a virulent strain of B. nodosus were analyzed and compared after one- and two-dimensional agarose gel electrophoresis and reaction with either the homologous or heterologous rabbit antiserum. Using ECP from the virulent strain and homologous antiserum, a complex series of at least nine immunoprecipitin bands was seen. These immunoprecipitates were easily identifiable in the crossed IEP test (Fig. la). C o m m o n identity of five of these immunoprecipitates was revealed when the same antiserum was reacted with benign ECP (Fig. lb) and confirmed by the fusion of immunoprecipitin peaks in a tandem crossed IEP test (Fig. lc). Four immunoprecipitates (Peaks 1, 2, 5 and 9) were unique to virulent B. nodosus antigens. Similarly, using the benign ECP and homologous antiserum, nine immunoprecipitates were usually detected (Fig. 2a) and c o m m o n identity of four of these immunoprecipitates was revealed in a heterologous immunoelectrophoretic reaction (Fig. 2b). A broad precipitate (Peak 10) was often seen in the heterologous reaction (Fig. 2b), but could be detected in the homologous reaction only when the antigen-antibody ratio was changed by using 20--25% antiserum in the second dimension gel. Five immunoprecipitates (Peaks 1, 2, 7, 8 and 9) were unique to benign B. nodosus antigens. No precipitin reaction was observed with either antiserum when each was tested against sterile uninoculated culture medium as the antigen.

139

i

i:C i

Fig. 1. Antigenic profile of extracellular proteins (ECP) prepared from culture supernarant of Bacteroides nodosus virulent strain 198. Demonstration of common antigens in ECP of strains 198 and 305 which are precipitated by antiserum to the former. (a) Crossed immunoelectrophoretic patterns and schematic representation of virulent (V) ECP antigens in well against antiserum from rabbit immunized with homologous ECP in the second dimension gel. (b) Same as in (a), but the well in the first dimension gel was loaded with ECP from benign (B) strain 305. (c) Tandem crossed immunoelectrophoretic patterns and schematic representation of V and B antigens in wells against antiserum from rabbit immunized with ECP of virulent strain 198 in the second dimension gel. All of the second dimension gels contained 10% antiserum. Immunoprecipitates of common identity are represented by the same number. Precipitates with two peaks are indicated by arrows.

Nature o f individual antigenic ECP molecules Figure 3 shows the protease banding patterns after electrophoresis of ECP from virulent and benign strains and the definition of individual enz y m e s b y t h e i r R f v a l u e s as u s e d in t h e p r e s e n t s t u d y . T h r e e e n z y m e s ( R f 4 8 , 5 3 a n d 6 3 ) s p e c i f i c t o t h e v i r u l e n t B. nodosus s t r a i n a n d f o u r e n z y m e s ( R f 3 5 , 4 4 , 5 0 a n d 5 5 ) s p e c i f i c t o t h e b e n i g n B. nodosus s t r a i n w e r e i s o l a t e d by preparative electrophoresis zymogram and used to immunize rabbits.

140

a

I

I

Fig. 2. Antigenic profile of extracellular proteins (ECP) prepared from culture supernatant of Bacteroides nodosus benign strain 305. Demonstration of common antigens in ECP of strains 305 and 198 which are precipitated by antiserum to the former. (a) Crossed immunoelectrophoretic patterns and schematic representation of benign (B) ECP antigens in well against antiserum from rabbit immunized with homologous ECP in the second dimension gel. (b) Same as in (a), but the well in the first dimension gel was loaded with ECP from virulent (V) strain 198. All of the second dimension gels contained 10% antiserum. Immunoprecipitates of common identity are represented by the same number. Precipitates with two peaks are indicated by arrows.

A t t e m p t s were m a d e to define the n a t u r e o f the d i f f e r e n t antigens in ECP o f virulent and benign strains o f B. n o d o s u s using antisera raised specifically against OMP, pili and various extracellular proteases. In t h e IEP test antisera against the benign strain e n z y m e R f 5 5 gave o n e precipitin b a n d w h e n reacted with ECP f r o m the h o m o l o g o u s strain b u t no precipitin r e a c t i o n was observed with ECP f r o m the virulent strain (Fig. 4a). F u r t h e r analysis using crossed IEP p r o c e d u r e s d e m o n s t r a t e d t h a t this precipitin band c o r r e s p o n d e d to i m m u n o p r e c i p i t a t e Peak 8 s h o w n in Fig. 2a. Similarly, antisera against virulent strain e n z y m e R f 6 3 gave o n e precipitin b a n d in the h o m o l o g o u s r e a c t i o n and n o b a n d w h e n reacted against the h e t e r o l o g o u s antigen, and this b a n d c o r r e s p o n d e d to i m m u n o p r e c i p i t a t e Peak 5 s h o w n in Fig. l a . All o f the o t h e r a n t i - e n z y m e antisera gave crossed r e a c t i o n s with h e t e r o logous antigens, revealing the same single precipitin b a n d originally d e t e c t e d

141

Fig. 3. Protease patterns after polyacrylamide gel electrophoresis of extracellular proteins excreted/secreted into the culture supernatant by virulent Bacteroides nodosus strain 198 and benign B. nodosus strain 305. Each enzyme was identified by its electrophoretic relative mobility (Rf) value. Hatched rectangles represented the enzymes which were excised from preparative gel and used for immunization of rabbits in this study (see Materials and Methods).

Fig. 4. Immunoelectrophoresis to test for specificity of antisera (AS) from rabbits immunized with protease band Rf55 (a), and protease band Rf35 (b) of benign Bncteroides nodosus strain 305. Wells were filled with extracellular proteins of either homologous strain 305 (B) or virulent strain 198 (V).

in homologous strain enzyme precipitin band reacted against

reactions (Fig. 4b), except for antiserum against virulent Rf53 which gave two cross-reacting precipitin bands. No was observed when ECP from either B. nodosus strain was anti-OMP and anti-pilus antisera.

DISCUSSION

Virulent and benign strains of B. nodosus can be differentiated by their colony morphology as well as by the biological properties of the ECP ex-

!42 creted/secreted by the different isolates (Thomas, 1962; Egerton and Parsonson, 1969; Thorley, 1976; Depiazzi and Richards, 1979; Stewart, 1979; Skerman et al., 1981). These biological differences suggest that ECP from virulent and benign strains of B. nodosus may differ at the molecular level. Recently, Every (1982) and Kortt et al. (1983) demonstrated differences in the electrophoretic pattern of ECP between virulent and benign isolates and suggested that these patterns could be used in their differentiation. Our data extend these findings and demonstrate the presence of both immunologically c o m m o n and unique ECP antigens between virulent and benign isolates from ovine footrot. A total of 9--10 immunoprecipitin peaks was observed in antisera when rabbits were immunized against ECP from either virulent or benign strains, and at least four immunoprecipitin peaks specific to each of the strains were identified. These antigenic profiles were obtained from study of only two strains: there will almost certainly be a wide range of antigenic profiles when more strains are studied. However, it is interesting that despite analysis of more than 30 different B. nodosus strains for extracellular protease band patterns there have been to date only three basic patterns recognized, one benign and two virulent (Every, 1982; Kortt et al., 1983; Gordon et al., 1985). Doubling of precipitin peaks in the second dimension gel is thought to be a technical artefact (Clarke and Freeman, 1968). This effect was noted in the present study (Figs. 1 and 2) and unsuccessful attempts made to eliminate these double peaks included allowing the first dimension gel to mature overnight in a humidified box before using it and reducing the voltage on the first dimension run in order to ensure no water appeared on the surface of the gel. The double peaks are probably due to causes other than those already known (Clarke and Freeman, 1968). The bacterial origin of all immunoprecipitates detected was confirmed as no precipitin reaction was observed when the uninoculated medium was run against each antiserum. One of the best physicochemically and immunologically characterized B. nodosus antigens is the pilus (Walker et al., 1973; Egerton, 1973; Every, 1979). Pili are often present in cell-free culture supernatants of B. nodosus as they are easily detached from the bacterial cell by physical stress such as that produced by centrifugation (Mattick et al., 1984). The isolation and purification of pili from cell-free culture supernatant is normally performed by isoelectric and magnesium chloride precipitation procedures (Brinton, 1965; Every, 1979). The present results suggest that the ammonium sulphate precipitation employed in the preparation of ECP was able to remove pili present in culture supernatants of B. nodosus, as attempts to demonstrate the presence of pili in the ECP antigens were unsuccessful. Similarly, proteins of the outer membrane complex, particularly OMP (mol. wt. 35 000 daltons) were not detected in the ECP preparation. Both anti-pilus and anti-OMP antisera had precipitating antibodies against the homologous immunogen in agar gel diffusion plates (E.K. Moses, personal communication, 1985).

143 In contrast, using the same approach, partial success in characterizing the antigens in ECP was achieved with antisera from rabbits immunized with discrete proteolytic enzyme bands electrophoretically separated from ECP complex of virulent and benign isolates of B. nodosus. Five protease antisera identified c o m m o n antigens shared between the two strains which suggested similar epitopes among the idividual enzymes. However, the remaining protease antisera (i.e. antisera against bands Rf55 and Rf63) reacted only with antigens specific to the homologous strain. These results indicate that specific antisera can be raised against certain individual proteins, such as protease Rf55 from benign strain and protease Rf63 from virulent strain, using polyacrylamide gel slices containing these enzymes after electrophoretic separation. Comparisons of the numbers of specific immunoprecipitates identified by antisera against crude ECP and antisera against individual proteases indicate that there are other non-protease molecules excreted/secreted by B. nodosus which are immunogenic when injected into rabbits. There are four immunoprecipitates specific to ECP from the virulent strain and five immunoprecipitates specific to the benign strain we examined. However, only one from each was identified with anti-protease antisera. Various workers have investigated other ultrastructural features of B. nodosus in attempts to define virulence-associated antigens of different isolates (Short et al., 1976; Stewart, 1978; Every and Skerman, 1980). Diffuse polar material and an additional layer on the cell surface were suggested to be associated with virulence, but conclusive evidence was lacking (Every and Skerman, 1983). Our data also suggest the involvem e n t of factors other than pili and proteases in determining whether a B. nodosus isolate is virulent or benign. These factors are excretory/secretory products of B. nodosus but the identity of these molecules is not yet defined. In addition, more work is needed to relate these differences to virulence. Their molecular structure and mode of action if any, in infection, especially in relation to differentiation of virulent and benign footrot, are of great current interest and warrant further investigation and characterization. ACKNOWLEDGEMENTS This work was supported by a grant from the Australian Wool Corporation. We are grateful to J. Bain for technical assistance, K. Aldridge for photographic work and W. Franklin for typing the manuscript.

REFERENCES Beveridge, W.I.B., 1941. Footrot in sheep: a transmissible disease due to infection w i t h Fusiforrnis nodosus (n.sp.). Studies on its cause, epidemiology and control. Bull. Counc. Sci. Ind. Res. (Aust.), 140: 1--56.

144 Brinton, C.C., 1965. The structure, function, synthesis and genetic control of bacterial pill and a molecular model for DNA and RNA transport in gram negative bacteria. Trans. N.Y. Acad. Sci., 27: 1003--1054. Clarke, M.H.G. and Freeman, T., 1968. Quantitative immunoelectrophoresis of human serum proteins. Clin. Sci., 35: 403--413. Depiazzi, L.J. and Richards, A.B., 1979. A degrading proteinase test to distinguish benign and virulent ovine isolates of B a c t e r o i d e s nodosus. Aust. Vet. J.~ 55: 25--28. Egerton, J.R., 1973. Surface and somatic antigens of F u s i f o r m i s nodosus. J. Comp. Pathol., 83: 151--159. Egerton, J.R. and Parsonson, I.M., 1969. Benign footrot. A specific interdigital dermatitis of sheep associated with infection by less proteolytic strains of F u s i f o r m i s n o d o sus. Aust. Vet. J., 45: 345--349. Emery, D.L., Clark, B.L., Stewart, D.J., O'Donnelt, I.J. and Hewish, D.R., 1984. Analysis of the outer membrane proteins of B a c t e r o i d e s nodosus, the causal organism of ovine footrot. Vet. Microbiol., 9: 155--168. Every, D., 1979. Purification of pili from B a c t e r o i d e s n o d o s u s and an examination of their chemical, physical and serological properties. J. Gen. Microbiol., 115: 309-316. Every, D., 1982. Proteinase isoenzyme patterns of B a c t e r o i d e s n o d o s u s : distinction between ovine virulent isolates, ovine benign isolates and bovine isolates. J. Gen. Microbiol., 128: 809--812. Every, D. and Skerman, T.M., 1980. Ultrastructure of the B a c t e r o i d e s n o d o s u s cell envelope layers and surface. J. Bacteriol., 141: 845--857. Every, D. and Skerman, T.M., 1983. Surface structure o f B a c t e r o i d e s n o d o s u s in relation to virulence and immunoprotection in sheep. J. Gen. Microbiol., 129: 225--234. Foissy, H., 1974. A method for demonstrating bacterial proteolytic isoactivities after electrophoresis in acrylamide gels. J. Appl. Bacteriol., 37 : 133--135. Gordon, L.M., Yong, W.K. and Woodward, C.A.M., 1985. The temporal relationships and characterization of extracellular proteases from benign and virulent strains of B a c t e r o i d e s n o d o s u s as detected in zymogram gels. Res. Vet. Sci., 3 9 : 1 6 5 - - 1 7 2 . Grabar, P. and Williams, C.A., Jr., 1953. Methode permittant l'etude conjugee des proprietes electrophoretiques et immunochemique d'um melange de proteines; application au serum sanguin. Biochim. Biophys. Acta, 10: 193--194. Kortt, A.A., Burns, J.E. and Stewart, D.J., 1983. Detection of the extracellular proteases of B a c t e r o i d e s n o d o s u s in polyacrylamide gels: a rapid method of distinguishing virulent and benign ovine isolates. Res. Vet. Sci., 35 : 171--174. Kroll, J., 1969. Immunochemical identification of specific precipitin lines in quantitative immunoelectrophoresis patterns. Scand. J. Clin. Lab. Invest., 24: 55--60. Laing, E.A. and Egerton, J.R., 1978. The occurence, prevalence and transmission of B a c t e r o i d e s n o d o s u s infection in cattle. Res. Vet. Sci., 24: 300--304. Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J., 1951. Protein measurements with the Folin phenol reagent. J. Biol. Chem., 193: 265--275. Mattick, J.S., Anderson, B.J., Mott, M.R. and Egerton, J.R., 1984. Isolation and characterization of B a c t e r o i d e s n o d o s u s fimbriae: structural subunit and basal protein antigens. J. Bacteriol., 160: 740--747. Santarius, K. and Ryan, C., 1977. Radial diffusion as a sensitive method for screening endopeptidase activity in plant extract. Anal. Biochem., 77 : 1--9. Short, J.A., Thorley, C.M. and Walker, P.D., 1976. An electron microscope study of B a c t e r o i d e s n o d o s u s : ultrastructure of organisms from primary isolates and different colony types. J. Appl. Bacteriol., 40: 311--315. Skerman, T.M., Erasmuson, S.K. and Every, D., 1981. Differentiation of B a c t e r o i d e s n o d o s u s biotypes and colony variants in relation to their virulence and immunoprotective properties in sheep. Infect. Immun., 32: 788--795.

145 Stewart, D.J., 1978. The role of various antigenic fractions of Bacteroides nodosus infections in eliciting protection against footrot in vaccinated sheep. Res. Vet. Sci., 24: 14--19. Stewart, D.J., 1979. The role of elastase in the differentiation of Bacteroides nodosus infections in sheep and cattle. Res. Vet. Sci., 27 : 99--105. Thomas, J.H., 1962. The differential diagnosis of footrot in sheep. Aust. Vet. J., 38: 159--163. Thorley, C.M., 1976. A simplified method for the isolation of Bacteroides nodosus from ovine footrot and studies on its colony morphology and serology. J. Appl. Bacteriol., 40: 301--309. Walker, P.D., Short, J., Thompson, R.O. and Roberts, D.S., 1973. The fine structure of Fusiformis nodosus with special reference to the antigens associated with immunogenicity. J. Gen. Microbiol., 77 : 351--361. Walker, J.E., Auffret, A.D., Carne, A., Gurnett, A., Hanisch, P., Hill, D. and Saraste, M., 1982. Solid-phase sequence analysis of polypeptides etuted from polyacrylamide gels. Eur. J. Biochem., 123: 253--260.