Veterinary Microbiology, 31 ( 1992 ) 351-362 Elsevier Science Publishers B.V., Amsterdam

351

Serotype specificity of immunological assays for the capsular polymer of Actinobacillus pleuropneumoniae serotypes 1 and 9 Thomas J. Inzana a, Jean Todd a, Carol Koch a and Jacques Nicolet b a

Veterinary Microbiology Research Laboratories, Department of Pathobiology, Virginia Polytechnic Institute and State University, Blacksburg, Virginia, USA bInstitute for Veterinary Bacteriology, University of Berne, Berne, Switzerland (Accepted 23 October 1991 )

ABSTRACT Inzana, T.J., Todd, J., Koch, C. and Nicolet, J., 1992. Serotype specificity of immunological assays for the capsular polymer ofActinobacillus pleuropneumoniae serotypes 1 and 9. Vet. Microbiol. 31: 351-362. The cross-reactivity of the purified polysaccharides of Actinobacillus pleuropneumoniae serotypes 1 and 9 were examined using a variety of highly sensitive assays, such as radioimmunoassay, latex agglutination, enzyme-linked immunosorbent assay (ELISA), and immunoblotting. In addition, conventional immunodiffusion was included for comparison. Latex agglutination, utilizing affinity-purified lgG to capsule, was also used to serotype whole cells. Agglutination or precipitation tests (radioimmunoassay, latex agglutination, and immunodiffusion) indicated no cross-reactivity between the capsules of serotypes 1 and 9, and no cross-reactivity between whole cells by latex agglutination. Assays that required binding of the capsule to a solid support (ELISA and immunoblotting) did demonstrate cross-reactions between serotypes 1 and 9 capsules, although reactions with the heterologous serotype were weaker than with the homologous serotype. The cross-reactivity could not be attributed solely to nonspecific factors because similar cross-reactivity did not occur with serotype 5 or 7 capsules by any assay. Reactivity of antisera with homologous or heterologous capsule was reduced, but not completely eliminated, by adsorption with washed, live bacteria of the heterologous serotype. Thus, the assay, as well as the antigen or specificity of the antibody reagent used, may influence the results ofA. pleuropneumoniae serotyping or serological tests.

INTRODUCTION

Actinobacillus pleuropneumoniae has been serologically grouped into 12 serotypes based on indirect hemagglutination and gel diffusion (Nielsen, 1986). The serotype-specific antigen ofA. pleuropneumoniae is the capsular Correspondence to: T.J. Inzana, Veterinary Research Laboratories, Department of Pathobiology, Virginia-Maryland Regional College of Veterinary Medicine, Virginia Polytechnic Institute and State University, Blacksburg, VA 24061, USA.

0378-1135/92/$05.00

© 1992 Elsevier Science Publishers B.V. All rights reserved.

352

T.J. INZANA ETAL.

polysaccharide (Inzana and Mathison, 1987), although other antigens have been shown to contain serotype-specific epitopes in the limited number of isolates examined (Fenwick and Osburn, 1986). Many somatic antigens (lipopolysaccharide (LPS) and membrane proteins) are known to cross-react among different serotypes of A. pleuropneumoniae (Fenwick and Osburn, 1986; Inzana and Mathison, 1987; Perry et al., 1990; Rapp and Ross, 1986; Rosendal and Mittal, 1985). Cross-reactivity has been particularly prominent among isolates of serotypes 3, 6, and 8 (Nielsen and O'Connor, 1984; Mittal et al., 1987); 4 and 7 (Mittal et al., 1987); and 1, 9, and 11 (Kamp et al., 1987; Mittal, 1990; Nielsen, 1985; Nicolet, 1988). This cross-reactivity could be explained by the common LPS structures shared within the serotypes (Perry et al., 1990). Serotype 1 is recognized as one of the most common serotypes isolated in many countries (Mittal et al., 1983; Rapp et al., 1985; Rosendal and Boyd, 1982; Schultz et al., 1983). However, serotype 9 has been reported to be the most common serotype isolated in the Netherlands (Kamp et al., 1987). Because current killed vaccines are serotype-specific, it is essential that isolates are precisely typed. As serological tests become more sensitive, cross-reactions are more likely to occur, necessitating the use of purified antigens, or monoclonal or affinity-purified polyclonal antibodies in assays to obtain adequate specificity. Recently, the use of formalinized whole cells, heated whole cells, and whole cell- and heated whole cell-saline extracts as antigens have been evaluated in several serotyping tests to identify serotype-specific and cross-reacting antigens of serotypes 1 and 9 (Mittal, 1990). The tests included tube agglutination, 2-mercaptoethanol tube agglutination, coagglutination (CoA), slide agglutination, indirect hemagglutination, co-indirect hemagglutination, immunodiffusion, and ring precipitation tests with antisera to whole cells of serotype 1 and 9 adsorbed or not adsorbed with the homologous and heterologous serotypes. The author concluded that the cross-reactive antigens were probably of cell wall origin. The structure of the serotype 1 and 9 capsules has been determined, and although distinct, both capsules share O~-D-galactose and phosphate residues (Perry et al., 1990). Ideally, purified antigen or monospecific antibodies should be used in diagnostic tests, particularly in the more sensitive assays. In this study we were interested in comparing how the use of purified serotype 1 and 9 capsular polysaccharide (CP), and/or affinity-purified IgG to each capsule, affected the specificity of a variety of highly sensitive assays. While assays that required precipitation or agglutination of antigen or antibody (eg. radioimmunoassay (RIA), latex agglutination (LA), and conventional gel immunodiffusion) were absolutely serotype specific, assays that required binding of capsular antigen to a solid support (eg. enzyme-linked immunosorbent assay (ELISA) and immunoblotting) demonstrated some cross-reactivity.

CROSS-REACTIVITY OF A. PLEUROPNEUMONL~IE CAPSULES

353

MATERIALS AND METHODS

Bacterial strains and growth conditions The A. pleuropneumoniae strains used in this study included: serotype 1, strains 4074, 4045 (Inzana, 1987 ), and eight additional field isolates ( 1 from Czechoslovakia, 4 from Switzerland, 1 from Canada, 1 from the United States, and 1 from Argentina), serotype 5, strains KI 7, J45, 178 (Inzana, 1987 ), and two additional field isolates from the United States, serotype 7, strains 29628 and five additional field isolates (obtained from L. Hoffman, Iowa State University, Ames, IA), serotype 9, strain 13261 and 6 additional field isolates (2 from Switzerland, 3 from Czechoslovakia, and 1 from Holland), and serotype 11, strain 56153 and 2 additional field isolates from Switzerland. Strains were grown in casamino acid-yeast extract broth supplemented with 5 #g/ml of nicotinamide adenine dinucleotide to late stationary phase for purification of capsule or immunization (Inzana, 1987).

Purification of CP The CP of each strain was purified from clarified culture supernatant by precipitation with hexadecyltrimethylammonium bromide, NaC1 extraction, phenol extraction, and ultracentrifugation. If the CP contained greater than 1% contamination with protein, nucleic acid, or LPS, it was further purified by Sepharose-2B column chromatography (Inzana, 1987 ).

Preparation of antiserum and purification of lgG Rabbit antiserum to whole cells was prepared as previously described (Inzana, 1987 ). Rabbits were immunized until their ELISA titer to purified, homologous CP was greater than 1 : 10 000. Antiserum to serotype 1 was thoroughly adsorbed with an isogenic, noncapsulated mutant until there was no reactivity to the non-capsulated cells by ELISA (Inzana and Mathison, 1987 ). The IgG fraction of the adsorbed serum was then obtained by Protein-A/G column chromatography. The serum was diluted with an equal volume of 10 mM phosphate buffer, pH 7.0 containing 150 mM NaC1 and 0.005% NaN3 (PBS), and passed slowly ( 5 m l / h / c m 2 ) through a hybrid column containing Protein A-Sepharose and Protein G-Sepharose (Pharmacia, Piscataway, N J). The column was washed with PBS until the OD280 returned to baseline, the flow direction reversed, and the IgG fraction eluted with 100 mM glycine HCI, pH 2.8, into tubes containing an equal volume of PBS. The IgG fractions were pooled, dialyzed against PBS, and concentrated by ultrafiltration to about 10 mg/ml protein. The IgG component of antiserum to serotype 9 whole cells was also isolated by Protein A/G-affinity chromatography. The IgG fraction specific for the serotype 9 capsule was then purified from an affinity column of amino-Sepharose 2B that was coupled with purified serotype 9 capsule activated with CNBr (Lucas, 1988). Removal of reactivity of homologous or

354

T.J. INZANA ET AL.

heterologous rabbit sera by absorption with homologous or heterologous purified capsule was unsuccessful. Therefore, antibodies to capsule were removed by adsorption with live, whole cells of the homologous and heterologous serotype. One liter of serotype 1 and serotype 9 bacteria in late log phase were washed twice with excess PBS, the cells equally divided into four 15-ml tubes, and each bacterial pellet was resuspended with 0.5 ml of the homologous or heterologous serum and incubated overnight at 4 ° C. The sera were then centrifuged at 8000 g to remove the bacteria, and the second set of bacterial pellets were resuspended with the adsorbed sera. After 4 h incubation at 4 °C with occasional mixing, the bacteria were removed by centrifugation and the sera-filter sterilized. Hyperimmune swine antiserum to A. pleuropneumoniae was made as described (Inzana and Mathison, 1987). Convalescent swine antisera to serotype 1 was obtained from Drs. Richard Ross (College of Veterinary Medicine, Iowa State University, Ames, IA) and Martha Mulks (College of Veterinary Medicine, Michigan State University, East Lansing, MI).

Serological assays The RIA (Inzana et al., 1990) and ELISA (Inzana and Mathison, 1987) have been previously described in detail. For the RIA, bacteria were intrinsically radiolabelled with 3H-glucose, the 3H-capsule was purified, and used in a working solution containing 2 nCi of 3H-capsule and 3 nCi of 36C1/25 ill. Equal aliquots of working solution and diluted serum were mixed, precipitated with saturated a m m o n i u m sulfate, the precipitate solubilized, and the radioactivity determined in a liquid scintillation counter. Percent binding of antibody to 3H-capsule was determined by a formula used with Lotus 123 computer software (Inzana et al., 1990). For ELISAs, purified capsule ( 1 ¢tg) was bound to polystyrene microtiter plates (Costar, Cambridge, MA), nonspecific binding was blocked with 3% nonfat dry milk, and dilutions of antiserum in PBS containing 0.05% Tween 20 (PBS-Tween) added. Specific binding of antibody to capsule was determined with a 1 : 1000 dilution of peroxidase-conjugated antibody to the host species Ig (H & L chain) diluted in PBS-Tween, followed by substrate (5-aminosalicylic acid in 0.005% H202). For detection of swine antibodies to capsule, biotinylated goat antibody to swine Ig was used ( 1 : 1000 ) in combination with strepavidin-conjugated horseradish peroxidase (1:2000) (Jackson Immunoresearch Labs, Inc., West Grove, PA). The plates were washed at least three times between addition of each reagent with PBS-Tween. Each reagent was incubated for 1 h at 37 ° C, except for capsule, which was incubated at 4 °C overnight. The concentration of reagents used were determined by checkerboard titration to maximize sensitivity and specificity. The ELISA titer was considered the dilution of antiserum that resulted in an absorbance >i 0.100. This value was based on studies in which the titer to capsule was considered the highest dilution with an

CROSS-REACTIVITY OF A. PLEUROPNEUMONIAE CAPSULES

3 55

absorbance greater than 3 times the standard deviation of the mean of negative control sera at the same dilution. Immunoblotting was done as described (Inzana and Mathison, 1987; Ma and Inzana, 1990) with the following changes: 8 #g of purified capsule was electrophoresed in a 10% separating gel, and 5% nonfat dry milk in PBS was used to block nonspecific binding and to dilute antisera. For the latex agglutination test, affinity purified antibody ( 1600 Ftg/ml ) to capsule was conjugated to 2.5% carboxylated latex beads (0.8 # m diameter) with water soluble carbodiimide (Bio-Rad, Richmond, CA) by modification of the procedure recommended by the manufacturer (Inzana, unpublished data; Polysciences, Corp. Warrington, PA). Overnight plate cultures of bacteria were diluted to a McFarland standard of 0.5 in PBS. Bacteria or purified capsule in 25 #1 of PBS were mixed with 10 #1 of latex reagent on a black slide (Polysciences Corp. ), and the slide rotated for 1 min. Agglutination was scored as 0, 1 + , 2 + , 3 + , or 4 + based on 0, 25, 50, 75, or 100% agglutination. A score of 2 + , 3 + , or 4 + was considered a positive test. Gel immunodiffusion was performed by modification of described procedures (Gunnarsson et al., 1978). Briefly, wells in 1% Seakem LE agarose in 0.8% saline were filled with 10 Ftl of undiluted serum (center well) or a 1 m g / ml solution of purified capsule in distilled water (peripheral wells). The gels were incubated overnight at room temperature in a sealed, moist chamber prior to examination for precipitation bands. RESULTS

Binding of radiolabelled serotype 1 capsule to homologous antiserum was inhibited only slightly by a 1 m g / m l solution of purified serotype 9 capsule by competitive RIA, but not at all by a 500 # g / m l solution. In contrast, a 100 n g / m l solution of serotype 1 capsule inhibited binding of radiolabelled serotype 1 capsule by about 50%. Hyperimmune rabbit serum to serotype 9 capsule did not react with radiolabelled serotype I capsule (titer < 1 : 10 ), whereas almost 98% binding of labelled antigen occurred at a 1:20 000 dilution of antiserum to serotype 1 strain 4045 (Table 1 ). Furthermore, antisera to strains 4074 and 4045 capsules could both detect as little as 1 n g / m l of serotype 1 capsule (not shown). When whole cells were used for inhibition, at least 2 . 8 X l 0 9 C F U / m l of serotype 9 cells did not inhibit binding of serotype 1 antiserum. In contrast, almost 4 logs fewer cells of serotype 1 (4.7 X 105 C F U / ml) inhibited binding of homologous antiserum by about 47%. Therefore, even though the RIA is highly sensitive and nonspecific cross-reactions are possible when concentrated reagents are used (Inzana et al., 1990), serotype 1 and 9 capsules were not cross-reactive by RIA. Latex beads coated with affinity purified antibodies to serotype 1 capsule were agglutinated (2 + ) in the presence of at least 10 n g / m l of purified sero-

356

T.J. 1NZANA ET AL.

TABLE1 Cross-reactivity o f purified capsules or whole cells o f serotypes 1 and 9 by radioimmunoassay 1 Antiserum 2

Inhibiting Antigen 3

Binding 4 (%)

Serotype 1 (1:20,000)

Serotype Serotype Serotype Serotype Serotype Serotype Serotype None None None

79.7 97.0 2.2 47.6 9.5 54.4 97.7 97.8 2.5 0.3

Serotype 9 ( 1 : 10) Serotype 9 ( 1 : 100)

9 CP ( 1 m g / m l ) 9 CP (500~tg/ml) 1 CP (100/zg/ml) 1 CP (100 n g / m l ) 1 cells (4.73< 106/ml) 1 cells (4.7X 105/ml) 9 cells (2.8X 109/ml)

~Purified, radiolabelled serotype 1 capsule from strain 4045 was used for all assays unless otherwise stated. Similar results were obtained with radiolabelled CP from serotype 1 strain 4074. 2Antiserum was to serotype 1 strain 4045, and serotype 9 strain 13261. 3Unlabelled CP or cells o f s e r o t y p e 1 strain 4045 or serotype 9 strain 13261 were used for inhibition. 4Less than 10% binding is a negative reaction without inhibitor, whereas greater than 90% binding is a negative reaction with inhibitor (Inzana et al., 1990). TABLE 2 Latex agglutination test ofA. pleuropneumoniae whole cells or purified capsule with latex beads conjugated to IgG specific for serotypes 1 or 9 capsule I Antigen

Serotype 1 Latex reagent

Serotype 9 Latex reagent

Serotype 1 cells ( 10 strains) Serotype 5 cells (5 strains) Serotype 7 cells (6 strains) Serotype 9 cells (7 strains) Serotype 11 cells (3 strains) Serotype 1 capsule (1 m g / m l ) ( 100 n g / m l ) ( 10 n g / m l ) Serotype 9 capsule ( 1 m g / m l ) ( 100 n g / m l ) (10 n g / m l )

4+ 2 4+ 4+ 2+ 4 -

_ 3 4+ 4+ 4+ 2+

ITwenty-five/zl o f a 0.5 McFarland suspension o f bacterial cells or diluted, purified capsule in phosphate buffered saline were mixed with 10/d o f latex beads conjugated to affinity purified lgG to serotypes 1 or 9 capsule on a slide. The slide was rotated for 1 min and read for agglutination. 24 + = 100% agglutination, large particles. 3_ = 0% agglutination, fine particles only. 42 + = 50% agglutination, fine and m e d i u m size particles.

CROSS-REACTIVITYOF A. PLEUROPNEUMONIAE CAPSULES

357

TABLE 3 Cross-reactivity of rabbit or swine sera to A. pleuropneumoniae serotypes with purified capsules by

ELISA Serum~

Antigen

Titer

Rabbit serum to serotype 1

Serotype 1 capsule Serotype 9 capsule Serotype 5 or 7 capsule Serotype l capsule Serotype 9 capsule Serotype l capsule Serotype 9 capsule Serotype 1 capsule Serotype 9 capsule Serotype 5 capsule Serotype 7 capsule Serotype 1 capsule Serotype 9 capsule Serotype l capsule Serotype 9 capsule Serotype 1 capsule Serotype 9 capsule

25,600 3,200 < 100 < 100 < 100 3,200 400 3,200 12,800 400 < 100 100 400 < 100 < 100 3,200-6,400 3,200

Adsorbed with serotype 12 Adsorbed with serotype 1 Adsorbed with serotype 92

Adsorbed with serotype 9 Rabbit serum to serotype 9

Adsorbed with serotype 12 Adsorbed with serotype 1 Adsorbed with serotype 92 Adsorbed with serotype 9 Swine serum to serotype 1~

t Rabbit sera to serotypes 5 and 7 reacted with only the homologous capsule. 2Antiserum was adsorbed with a pellet from 250 ml of bacterial culture of the homologous or heterologous serotype twice at 4 °C prior to assay. 3The swine sera used included hyperimmune (titer 6400) and convalescent (titer 3200). Swine serum to serotype 9 was not available.

type 1 capsule, whereas a 1 m g / m l solution of serotype 9 capsule did not agglutinate the serotype 1-specific beads. Identical, but reciprocal, results were obtained with latex beads coated with IgG to serotype 9 capsule in the presence of serotype 9 and 1 capsules. Latex beads to serotype 1 were strongly agglutinated (4 + ) in the presence of all l0 serotype 1 strains tested, but were not agglutinated ( - ) by any of the 7 serotype 9 strains, 6 serotype 7 strains, 5 serotype 5 strains, or 3 serotype 11 strains. Likewise, agglutination occurred with all 7 serotype 9 strains by latex reagent specific for serotype 9 capsule, but no agglutination occurred with any of the strains of serotypes l, 5, 7, or I 1 (Table 2 ). A strong band of precipitation occurred with antiserum to serotype 1 capsule and homologous capsular antigen by immunodiffusion; there was no reaction with serotype 9 capsule. Identical, but reciprocal, results were obtained with antiserum to serotype 9 capsule and serotypes 9 and 1 capsular polysaccharides (data not shown). ELISA results of cross reactivity between serotypes 1 and 9 capsules, however, were very different from those obtained with agglutination or precipitation assays. Rabbit antisera to two strains of serotype l had endpoint titers

358

T.J. INZANA ET AL.

B

A

1

2

3

4

1

2

3

4

Fig. 1. Immunoblot of purified A. pleuropneurnoniae capsules with homologous and heterologous rabbit antiserum. Capsules were electrophoresed in a 10% separating gel, electroblotted to nitrocellulose, and the paper strips reacted with a 1:200 dilution of hyperimmune rabbit serum to serotype 1 (A) or serotype 9 (B). Bound antibodies were detected with goat anti-rabbit IgG coupled to horseradish peroxidase followed by 4-chloro-l-naphthol substrate. (A) lanes and antigens: 1, serotype 1 capsule; 2, serotype 5 capsule; 3, serotype 7 capsule; 4, serotype 9 capsule. (B) lanes and antigens: 1, serotype 1 capsule; 2, serotype 5 capsule; 3, serotype 7 capsule; 4, serotype 9 capsule. o f 25 600 to s e r o t y p e 1 capsule a n d 1 : 3 2 0 0 to s e r o t y p e 9 capsule ( T a b l e 3). In c o n t r a s t , a n t i s e r u m to s e r o t y p e 1 d i d n o t react with s e r o t y p e s 5 or 7 capsules at 1 : 100 dilution. T h e titer o f r a b b i t a n t i s e r u m to s e r o t y p e 9 with serotype 9 capsule was 1 : 12 800, b u t was also 1 : 3200 with s e r o t y p e 1 capsule. T h e r e was also n o r e a c t i v i t y o f s e r o t y p e 9 a n t i s e r u m with s e r o t y p e 7 capsule, a n d v e r y w e a k r e a c t i v i t y with s e r o t y p e 5 capsule ( t i t e r 1 : 4 0 0 ) . R a b b i t antisera to serotypes 5 a n d 7 also failed to react with s e r o t y p e 9 capsule ( d a t a n o t s h o w n ) . Swine a n t i s e r a to s e r o t y p e 1 r e a c t e d a l m o s t equally as well to serotype 9 as to s e r o t y p e I capsule ( T a b l e 3).

CROSS-REACTIVITY OF A. PLEUROPNEUMONIAE CAPSULES

3 59

Adsorption of antiserum with washed, live bacteria of the homologous serotype removed all reactivity to the homologous and heterologous capsule of each serotype. Adsorption of serotype 1 antiserum with serotype 9 cells reduced, but did not completely remove, reactivity with serotype 1 or 9 capsule. The same results occurred when antiserum to serotype 9 was adsorbed with serotype 1 cells, except the adsorption of serotype 9 antiserum with serotype 1 cells removed more cross-reactive antibodies than adsorption of serotype 1 antiserum with serotype 9 cells (Table 3 ). Immunoblotting of purified serotypes 1 and 9 capsule with antisera to the homologous and heterologous serotype confirmed that the cross-reactivity was specific for the capsule and was not due to contaminating somatic components. Serotypes 5 and 7 capsules did not react with antiserum to serotype 1 or serotype 9 (Fig. 1 ). DISCUSSION

Most standard assays that have been used for serotyping ofA. pleuropneumoniae have utilized agglutination or precipitation of bacteria in solution (Gunnarsson et al., 1978; Kamp et al., 1987; Mittal, 1990; Nielsen, 1986; Rapp et al., 1985; Rosendal, Mittal, 1985; Schultz, Ross, et al., 1983). Using such assays, however, investigators (Nielsen, 1985; Kamp et al., 1987) have noted cross-reactivity between serotypes 1 and 9, as well as the infrequently isolated serotype 11 (Nicolet, 1988). Mittal (1990) used a variety of similar assays with whole cells or crude extracts to determine that the cross-reactive antigens were most likely present in the cell wall. The capsule of A. pleuropneumoniae, however, is the serotype-specific antigen (Inzana and Mathison, 1987; Perry et al., 1990). Therefore, we were interested in determining if cross-reactivity would occur between serotypes 1 and 9 when purified capsules or affinity purified antibodies were used in a new generation of more sensitive assays, as well as by traditional immunodiffusion. No significant cross-reactivity occurred between the capsule or specific antibodies in any of the assays that involved agglutination or precipitation of antigen-antibody complexes (RIA, latex agglutination, or immunodiffusion), provided reagents were used at normal working concentrations. Unexpectedly, however, cross-reactions did occur between antigens and antisera of serotypes 1 and 9 by ELISA and immunoblotting; purified capsule of serotype 11 was not available for testing in these assays. Capsules and antisera to serotypes 5 and 7 did not cross-react with antisera or capsules to serotypes 1 or 9 (a very weak cross-reactivity did occur between serotypes 9 and 5 by ELISA). The cross-reactivity cannot be explained by greater sensitivity of ELISA-type assays because RIA has been shown to be more sensitive than ELISA for A. pleuropneumoniae capsular serology (Inzana et al., 1990 ). The cross-reactivity was not due to contaminating somatic antigens because pur-

360

T.J. INZANA ET AL.

ified capsules immunoblotted with antiserum to whole cells reacted only with material having an electrophoretic profile identical to capsule, and not to any proteins or LPS (Inzana et al., 1990). To confirm that the cross-reactivity of the antiserum with the heterologous antigen was specific to capsular epitopes, each serum was incubated with an excess of the homologous or heterologous capsule prior to testing by ELISA. However, this procedure did not remove the cross-reactivity for the heterologous antigen, nor did it reduce activity for the homologous antigen. Although a precipitate did form, it is possible that the solubilized capsule was a poor antigen for adequately removing specific antibodies for an ELISA. Adsorption of the antisera with washed, live whole cells was far more efficient for removing antibodies. Since the antigen used in the ELISA was purified capsule, removal of cross-reactivity due to adsorption of cross-reacting somatic antibodies was not a factor. The thoroughness of the adsorption was confirmed by the removal of all reactivity with homologous or heterologous capsule following adsorption with homologous bacteria. Of interest was that adsorption with heterologous bacteria reduced, but did not eliminate, reactivity with the homologous or heterologous capsule. Furthermore, adsorption of serotype 9 antiserum with serotype 1 cells was somewhat more efficient at removing specific and cross-reactive activity than adsorption of serotype 1 antiserum with serotype 9 cells. This may indicate in the ELISA, more shared epitopes are accessible in the serotype 9 capsule than in the serotype 1 capsule. The basis for the discrepancy between the results of agglutination/precipitation and ELISA/immunoblotting is not clear, but may be related to the physical state and epitope accessibility of the antigens in the assays. In precipitation and agglutination assays the antigen is in suspension, while in ELISA and immunoblotting the antigen is bound to a solid phase (polystyrene well or nitrocellulose paper). The binding of antigen to a solid phase may allow c o m m o n cross-reactive epitopes to be exposed for reactivity with antibody that are not available when the antigen is in suspension. We did not examine if cross-reactions would occur by indirect fluorescence immunoassay, in which whole cells are dried onto a glass slide. Perry and co-workers (Altman et al., 1986; Perry et al., 1990) have shown that the capsule of serotype 1 is an Nacetylglucosamine-galactose-phosphate polymer, and the serotype 9 capsule is a galactose-glycerol-phosphate polymer; both capsules are partly O-acetylated. Thus, c o m m o n components, and the possibility of cross-reactive epitopes, exist between the two capsules. Our ELISA and immunoblotting results do not explain the cross-reactions observed by others (Nielsen, 1985; K a m p et al., 1987) because both groups used agglutination/precipitation-type assays. Rather, the use of multi-component reagents and the detection of cross-reactive cell wall antigens, as suggested by Mittal (Mittal, 1990), are more likely reasons for the initial reports of cross-reactivity.

CROSS-REACTIVITY OF A. PLEUROPNEUMONIAE CAPSULES

361

Although ELISA and immunoblotting have not generally been used for serotyping A. pleuropneumoniae, they have been used for measuring the serotype-specific antibody response of pigs suspected of pleuropneumonia (Boss6 et al., 1990; Inzana and Mathison, 1987; Nicolet, Paroz, et al., 1981 ). Therefore, ELISA and immunoblotting assays used to measure the antibody response of swine to A. pleuropneumoniae serotypes 1 or 9 may be prone to cross-reactions. Our results indicate, however, that cross-reactions with the heterologous antigen are weaker than with the homologous antigen. Thus, comparative use of both serotypes 1 and 9, and/or the use of precipitation or agglutination assays that utilize purified capsule (eg. RIA or latex agglutination), will prevent misinterpretation of ELISA or immunoblotting results. ACKNOWLEDGEMENTS

This study was supported, in part, by grants 88-34116-3641 from the U.S. Department of Agriculture, and BIO-88-006 from the Virginia Center for Innovative Technology.

REFERENCES Altman, E., Brisson, J.-R. and Perry, M,B., 1986. Structural studies of the capsular polysaccharide from Haernophiluspleuropneumoniae serotype 1. Biochem. Cell Biol., 64:707-716. Boss6, J.T., Johnson, R.P. and Rosendal, S., 1990. Capsular polysaccharide antigens for detection of serotype-specific antibodies to Actinobacillus pleuropneumoniae. Can. J. Vet. Res., 54: 320-325. Fenwick, B.W. and Osburn, B.I., 1986. Immune response to the lipopolysaccharides and capsular polysaccharides of Haernophilus pleuropneumoniae in convalescent and immunized pigs. Infect. Immun., 54: 575-582. Gunnarsson, A., Hurvell, B. and Biberstein, E.L., 1978. Serologic studies of porcine strains of Haemophilus parahaemolyticus (pleuropneumoniae): antigen specificity and relationship between serotypes. Am. J. Vet. Res., 39: 1286-1292. Inzana, T,J., 1987. Purification and partial characterization of the capsular polymer of Haemophiluspleuropneumoniae serotype 5. Infect. Immun., 55:1573-1579. Inzana, T.J. and Mathison, B., 1987. Serotype specificity and immunogenicity of the capsular polymer ofHaemophiluspleuropneumoniae serotype 5. Infect. Immun., 55:1580-1587. Inzana, T.J., Clark, G.F. and Todd, J., 1990. Detection of serotype-specific antibodies or capsular antigen of Actinobacillus pleuropneumoniae by a double-label radioimmunoassay. J. Clin. Microbiol., 28: 312-318. Kamp, E.M., Popma, J.K. and Van Leengoed, L.A.M.G., 1987. Serotyping of Haemophilus pleuropneumoniae in the Netherlands: with emphasis on heterogeneity within serotype 1 and (proposed) serotype 9. Vet. Microbiol., 13: 249-257. Lucas, A.H., 1988. Expression of crossreactive idiotypes by human antibodies specific for the capsular polysaccharide of Haemophilus influenzae type b. J. Clin. Invest., 81: 480-486. Ma, J. and Inzana, T.J., 1990. Indirect enzyme-linked immunosorbent assay for detection of antibody to a 110,000-molecular-weight hemolysin of Actinobacillus pleuropneumoniae. J. Clin. Microbiol., 28: 1356-1361.

362

Y.J. INZANA ET AL.

Mittal, K.R., 1990. Cross-reactions between Actinobacillus (Haemophilus) pleuropneumoniae strains of serotypes 1 and 9. J. Clin. Microbiol., 28: 535-539. Mittal, K.R., Higgins, R., Larivibre, S., 1983. Identification and serotyping of Haemophilus pleuropneumoniae by coagglutination test. J. Clin. Microbiol., 18:1351-1354. Mittal, K.R., Higgins, R., Larivibre, S., 1987. An evaluation of agglutination and coagglutination techniques for serotyping of Haernophilus pleuropneumoniae isolates. Am. J. Vet. Res., 48: 219-226. Nicolet, J., 1988. Taxonomy and serological identification ofActinobacillus pleuropneumoniae. Can. Vet. J., 29: 578-580. Nicolet, J., Paroz, P., Krawinkler, M. and Baumgartner, A., 1981. An enzyme-linked immunosorbent assay, using an EDTA-extracted antigen for the serology ofHaemophilus pleuropneumoniae. Can. Vet. J., 29: 578-580. Nielsen, R., 1985. Serological characterization of Haemophilus pleuropneumoniae (Actinobacillus pleuropneumoniae) strains and proposal of a new serotype: serotype 9. Acta. Vet. Scand., 26: 581-585. Nielsen, R., 1986. Serological characterization of Actinobacilluspleuropneumoniae strains and proposal of a new serotype: serotype 12. Acta. Vet. Scand., 27: 452-455. Nielsen, R. and O'Connor, P.J., 1984. Serological characterization of 8 Haemophilus pleuropneumoniae strains and proposal of a new serotype: serotype 8. Acta. Vet. Scand., 25: 96106. Perry, M.B., Altman, E., Brisson, J.-R., Beynon, L.M. and Richards, J.C., 1990. Structural characteristics of the antigenic capsular polysaccharides and lipopolysaccharides involved in the serological classification of Actinobacillus (Haemophilus) pleuropneumoniae strains. Serodiag. lmmunother. Infect. Dis., 4: 299-308. Rapp, V.J. and Ross, R.F., 1986. Antibody response of swine to outer membrane components of Haemophilus pleuropneumoniae during infection. Infect. lmmun. 54:751-760. Rapp, V.J., Ross, R.F., and Erickson, B.Z., 1985. Serotyping ofHaemophilus pleuropneumoniae by rapid slide agglutination and indirect fluorescent antibody tests in swine. Am. J. Vet. Res., 46: 185-192. Rosendal, S. and Boyd, D.A., 1982. Haemophilus pleuropneumoniae serotyping. J. Clin. Microbiol., 16: 840-843. Rosendal, S. and Mittal, K.R., 1985. Serological cross-reactivity between a porcine Actinobacil/us strain and Actinobacillus pleuropneumoniae. Can. J. Comp. Med., 49:164-170. Schultz, R.A., Ross, R.F., Gunnarsson, A., and Nielsen, R., 1983. Serotyping of 50 different isolates of Hemophilus pleuropneurnoniae from swine pneumonia in Iowa and surrounding states. Vet. Med. Small Anim. Clin. 78:1451-1453.

Serotype specificity of immunological assays for the capsular polymer of Actinobacillus pleuropneumoniae serotypes 1 and 9.

The cross-reactivity of the purified polysaccharides of Actinobacillus pleuropneumoniae serotypes 1 and 9 were examined using a variety of highly sens...
979KB Sizes 0 Downloads 0 Views