- Email: [email protected]
Binding of Chicken, Bovine, and Rabbit Immunoglobulins by Avian, Bovine, and Human Strains of Staphylococcus aureus1
Binding of Chicken, Bovine, and Rabbit Immunoglobulins by Avian, Bovine, and Human Strains of Staphylococcus aureus1 WILLIAM R. SCHWAN and PAUL A. HARTMAN2 Department of Microbiology, 205 Science I, Iowa State University, Ames, Iowa 50011 (Received for publication April 5, 1985)
1986 Poultry Science 65:696-703
INTRODUCTION
Staphylococcus aureus, a common inhabitant of poultry and most mammals, is responsible for a number of pathological conditions (Cole, 1984). Although 5. aureus is a single species, biotypes isolated from different animals differ in the cellular products they produce. Animal and human strains differ, for example, in the type of hemolysin produced, pigmentation, coagulation of plasma, staphylokinase production, growth on crystal violet (CV) agar, deoxyribonuclease (DNase) production, and bacteriophage type (Devriese, 1984; Devriese and Oeding, 1976). Many human strains of S. aureus produce the cellular product protein A, which possesses the capability of binding the Fc region of immunoglobulins, especially immunoglobulin G (IgG) (Sjoquist and Stalenheim, 1969). Protein A production is a variable property of poultry, and other origins of S. aureus (Hajek and Marsalek, 1976) and differences in IgG binding exist for human strains of S. aureus (Richman
'Journal Paper Number J-11750 of the Iowa Agriculture and Home Economics Experiment Station, Ames, IA. Projects 2377 and 2678. 1 To whom correspondence should be addressed.
696
et al., 1982). These observations as well as the differences in other cellular products led us to believe that preparations of protein A that possess different immundglobulin-binding properties might exist. The purpose of the present investigation, therefore, was to study the binding of various IgG molecules to strains of S. aureus obtained from different origins, including poultry. An enzyme immunoassay (EIA) was developed to assess qualitative differences in IgG binding and to correlate binding to the origin of the S, aureus.
MATERIALS AND METHODS
Bacterial Cultures. Chicken S. aureus and coagulase-negative staphylococci were obtained from the exterior of live White Leghorn hens and processed chickens. The S. aureus cultures were confirmed by biochemical and morphological tests. Bovine isolates of S. aureus were obtained from R. A. Packer (Iowa State University, Ames, IA). Strains of human S. aureus, including Cowan I, which produces high amounts of protein A, and Wood 46, which produces low amounts of protein A, were supplied by P. A. Pattee (Iowa State University, Ames, IA). Two strains of Staphylococcus epidermidis (protein A-negative) and six nonstaphylococcal species (two Micrococcus sp., Bacillus subtilis,
Downloaded from http://ps.oxfordjournals.org/ at New York University on April 24, 2015
ABSTRACT Sixty-six strains of Staphylococcus aureus of avian, bovine, and human origin were tested for their ability to bind chicken, bovine, and rabbit immunoglobulin G (IgG). A microtitration plate hemagglutination assay and a direct-tube enzyme immunoassay were used to determine qualitative differences. Twice as many chicken and bovine S. aureus isolates than human strains reacted positively to chicken IgG. Mean binding values of chicken IgG were also twice as high for chicken and bovine S. aureus isolates when compared with human-derived strains. Isolates of bovine and human origin displayed a high affinity for bovine IgG and rabbit IgG, whereas few isolates from chickens bound substantial quantities of the nonavian IgG species. These results demonstrate that preparations of staphylococcal protein A with affinities for immunoglobulins from poultry and other animals can be obtained by screening large numbers of isolates, especially those obtained from an animal species, such as chickens, that is the same as the source of the immunoglobulin one wishes to bind. (Key words: staphylococcal protein A, enzyme immunoassay, immunoglobulin G binding)
697
IMMUNOGLOBULIN BINDING
Preparation of Enzyme-Conjugated Antibody. A one step glutaraldehyde method of Voller et al. (1976) was used to conjugate alkaline phosphatase (Sigma Type VII-T) to antibody. Antibodies used for conjugation were chromatographically purified chicken IgG,
bovine IgG, and rabbit IgG (Sigma Chemical Company, St. Louis, MO). After conjugation and dialysis, the 1-ml volume of IgG-alkaline phosphatase conjugate was diluted to 3 ml with Tris-HCl buffer (.05 M, pH 8.0) containing 1.0% bovine serum albumen and .02% NaN 3 . The conjugate was stored in the dark at 4 C. Tube Enzyme Immunoassay. Staphylococcal cultures grown in 8 ml of BHI broth at 3-7 C for 24 hr were pelleted by centrifugation at 1000 X g for 10 min. Supernatant fluids were decanted, and each pellet was resuspended in .5 ml of PBS (.01 M, pH 7.4). A .5-ml volume of enzyme-conjugated IgG diluted in PBS was added to each tube. After incubation for 1 hr at 35 C, the tubes were rinsed three times with .5-ml volumes of PBS by using centrifugation at 1000 X g for 10 min after each wash. Then .5 ml of substrate (p-nitrophenyl phosphate, Sigma #104-105) was added to each tube. The substrate was previously diluted to 1 mg/ml in substrate buffer (10% diethanolamine, pH 9.8) before use in the assay (Voller et al., 1976). Hydrolysis was stopped after 30 min at room temperature by adding .3 ml of 3 N NaOH. Cells were pelleted at 1600 X g, and 150 (i\ of supernatant was transferred from each tube to a well of a microtitration plate. Each culture was analyzed spectrophotometrically at 405 nm with a Micro-ELISA Plate Reader (Dynatech Laboratories, Alexandria, VA). Controls were run with each analysis. Optimal dilutions of enzyme-conjugated antibody for use in the tube EIA were determined for each of the three conjugates (chicken IgG, bovine IgG, and rabbit IgG). To determine these optimal concentrations, dilutions of enzyme-labeled antibody from 1:100 to 1:1000 were analyzed by the tube EIA method described previously. The ratio of the positive control (Cowan I) to the negative control (S. epidermidis) was used to determine the optimal dilution of each conjugate. RESULTS
Analysis of Growth on Crystal Violet Agar and Deoxyribonuclease Production. Percentages of S. aureus of bovine, chicken, and human origin that were DNase-positive and percentages that were CV Type A or C are shown in Table 1. A high percentage of bovine isolates and human strains produced heat-stable DNase (93 and 100%, respectively) as compared with chicken isolates (74%). A majority of the
Downloaded from http://ps.oxfordjournals.org/ at New York University on April 24, 2015
Escherichia coli, Streptococcus zooepidemicus, and a Salmonella sp.) were obtained from the Iowa State University Microbiology Department Culture Collection (Ames, IA). A strain of turkey and a strain of pheasant S. aureus were supplied by L. J. Hoffman (Veterinary Diagnostic Laboratory, Ames, IA). Comparison of Deoxyribonuclease Production and Growth on Crystal Violet Agar. All S. aureus cultures were grown on CV agar (Difco tryptose agar containing 8 jug/ml CV). The color of growth after 24 and 48 hr was recorded according to Devriese (1984) to distinguish Type A from Type C cultures. Every S. aureus culture was also grown on Difco DNase agar containing .03 g/liter of toluidine blue, which was used as an indicator of deoxyribonucleic acid hydrolysis (Devriese and Oeding, 1976). The cultures were incubated for 24 hr, flooded with 1 N HC1, and observed for DNase activity. Micro titration Hemagglutination Assay. A microtitration hemagglutination procedure was used to test the chicken S. aureus and coagulase-negative staphylococci. Rabbit anti-sheep red blood cell sera prepared according to Winblad and Ericson (1973) and chicken antisheep red blood cell sera prepared according to Williams and Chase (1971) were used to sensitize sheep red blood cells as reported by Anderson et al. (1970). These sensitized sheep red blood cells were used to detect antibody binding to staphylococcal cells in a microtitration hemagglutination assay according to Sjoquist and Stalenheim (1969) as modified by Rollo (E. E. Rollo, 1982, M.S. thesis, Iowa State University, Ames, IA). The technique was as follows: 50 /ill of 1% sensitized sheep red blood cells was added to each well of a U-bottomed microtitration plate (Cooke Engineering Co., San Mateo, CA) that contained a loopful of culture from a brain-heart infusion (BHI) agar plate (Difco) suspended in 50 jul of phosphatebuffered saline (PBS) (.01 Af, pH 7.4). Microtitration plates were incubated for 2 hr at room temperature, examined, incubated overnight at 4 C, and examined again. Cowan I served as a positive control, and S. epidermidis served as a negative control.
698
SCHWAN AND HARTMAN TABLE 1. Origin of Staphylococcus aureus strains and percentages that produced heat-stable deoxyribonuclease (DNase) and crystal violet (CV) reactions
Origin of S. aureus
Number of strains
Bovine Chicken Human
14 100 23
CV Type
(% positive)
(%A:%C) 43:57 59:41 13:87
93 74
100
origin bound less rabbit IgG (A405 = .26). For chicken IgG, chicken and bovine S. aureus displayed comparable A40s of .12 and.14, respectively, whereas human S. aureus bound chicken IgG less well (A 40 s = .07). Reactions with bovine IgG were greatest for human strains (A405 = .79), followed by bovine (A40S = .51), and then by chicken (A405 = .38) isolates. Coagulase-negative staphylococci, turkey, and pheasant strains of S. aureus, and nonstaphylococcal species displayed low absorbance values (less than A405 = .10) for all three types of IgG. The two coagulase-negative isolates and the one S. aureus isolate that showed positive responses and autoagglutination in the microtitration hemagglutination assay also exhibited low absorbance values (less than A405 = .10) in the direct tube EIA. To categorize further the isolates within each group as producing strong or weak reactions, an absorbance value equal to or greater than .30 was considered a strong positive response, and an absorbance value of .10 to .29 was considered a weak positive response (Table 4). Human and bovine strains of S. aureus again showed similarities when reacted with rabbit IgG (total percentages of positive results were 96 and 92%, respectively) and bovine IgG (91 and 100%, respectively). Only 74% of chicken S. aureus isolates bound rabbit or bovine IgG. Only 26% of human S. aureus bound chicken IgG, whereas 57% of bovine and 55% of chicken isolates bound chicken IgG. Binding Patterns of Selected Staphylococcus aureus Cultures. Additional differences in binding patterns are shown in Table 5. Some cultures bound only rabbit IgG, others bound only chicken IgG, and others bound IgG preparations from two or three species. DISCUSSION Among strains of S. aureus obtained from
Downloaded from http://ps.oxfordjournals.org/ at New York University on April 24, 2015
bovine isolates and human strains were CV Type C, whereas a majority of the chicken isolates were CV Type A. Preliminary Identification of Antibody Binding. A qualitative microtitration hemagglutination assay was used to determine how many coagulase-negative and coagulase-positive chicken staphylococci would bind rabbit and chicken antibodies (Table 2). Of 175 coagulasenegative chicken staphylococci tested, only 1% yielded a positive test. Of 100 chicken S. aureus isolates, 27% showed reactivity to one or both antisera. Most S. aureus isolates that displayed a positive response did so only with rabbit antiserum (18%). However, 9% of the S. aureus isolates bound chicken antiserum alone (4%) or in combination with rabbit antiserum (5%). Cowan I reacted only with rabbit antiserum, and the 5. epidermidis strains showed no response to either antisera. The two coagulasenegative chicken isolates and one of the S. aureus isolates that exhibited positive responses also displayed sporadic autogglutination. Analysis with the Tube EIA. A sensitive tube EIA was used to determine which specific class of antibody, especially chicken IgG, was bound to the strains of S. aureus. Optimal dilutions of the three conjugates (rabbit, chicken, and bovine) based on the ratio of binding to Cowan IIS. epidermidis were 1/500 for rabbit IgG (P/N = .68/.01), 1/200 for chicken IgG (P/N = .12/.01), and 1/600 for bovine IgG (P/N = .75/.01). When the optimum dilutions of each species of IgG were used, a wide range of absorbance values was detected (Table 3). Comparisons could be made of mean binding differences between origin of the staphylococci and the type of IgG that was bound. Similar mean values of absorbance can be seen for human (A405 = .47) and bovine (A405 = 49) S. aureus when reacted with rabbit IgG; isolates of S. aureus of chicken
DNase
IMMUNOGLOBULIN BINDING
699
Downloaded from http://ps.oxfordjournals.org/ at New York University on April 24, 2015
different ecological niches, there are differences in the amounts and types of cellular products produced. Fewer strains isolated from poultry produce heat-stable DNase as detected with the conventional plating procedure (Devriese and Van de Kerckhove, 1979), an observation confirmed in the present study. Devriese (1984) biotyped S. aureus, based in part on CV reactions. Poultry and bovine strains of S. aureus were classified primarily as CV Type A, whereas human biotypes were grouped primarily into CV Type C. The results presented in this study correlate with Devriese's typing schemes for human and chicken S. aureus. Although the CV reactions of the bovine S. aureus did not correlate as well as in Devriese's study, a substantial percentage of the isolates (43%) still fell into the CV Type A group. Because different strains of S. aureus vary in production of some cellular products, it was thought that strains of S. aureus obtained from poultry and other animal species might also produce protein A with different binding affinities to IgG. These differences could result from adaptation of the S. aureus to specific ecological niches (animal species). This adaptation could include modification of the protein A produced so that the protein A would bind immunoglobulin(s) of the host species. Therefore, we attempted to determine if there was an association between species of origin and immunoglobulin-binding properties of different strains of S. aureus. Immunoglobulin-binding patterns of human strains of protein A-bearing S. aureus have been well characterized (Goudswaard et al., 1978; Kronvall et al., 1974), but little has been done to characterize antibody binding by poultry and other animal strains of S. aureus. Differences exist, however, in immunoglobulin binding by several groups of streptococci (Christensen and Oxelius, 1974; Myhre et al, 1979). The micro titration hemagglutination assay used as an initial screen for detection of antibody binding to the chicken staphylococci provided sensitivity but also had limitations because of autoagglutination. Autoagglutination has been seen by other investigators working with coagulase-negative staphylococci (Hovelius and Mardh, 1979) and S. aureus (Carret et al., 1981). Because of the autoagglutination problem, a direct tube EIA was used to detect antibody binding differences. Most human strains of S. aureus bound rabbit IgG within the range noted by others (Forsgren,
SCHWAN AND HARTMAN
700
TABLE 3. Numbers of staphylococci tested by using a tube enzyme immunoassay and mean absorbance values obtained with immunoglobulin G (IgG) preparations from three different animals T y p e of isolate or strain
Total
Rabbit IgG
Chicken IgG
Bovine IgG
.26 .47 .49 .04 .02
.12 .07 .14 .05 .00
.38 .79 .51 .04 .04
(n) Staphylococcus Chicken Human Bovine
aureus
Pheasant Chicken coag — s t a p h . 2 N o n s t a p h . species 3
4 6
± ± ± ± ±
.16' .24 .29 .01 .00
.01 ± .00 .00 + .00
± ± ± ± ±
.08 .05 .10 .01 .00
. 0 1 ± .02 .01 ± .01
+ ± ± + +
.25 .24 .29 .01 .01
.02 ± .03 .02 + .01
Downloaded from http://ps.oxfordjournals.org/ at New York University on April 24, 2015
Turkey
27 23 14 1 1
'Mean absorbance ± standard deviation measured at 405 nm. 2 Coagulase-negative staphylococci. 3 Species other than Staphylococcus (two Micrococcus sp., Escherichia coli, Bacillus subtilis, one Salmonella, Streptococcus zooepidemicus).
1970; Lachica et al, 1979), but few bound chicken IgG. This limited reactivity to chicken IgG, and chicken immune complexes has been noted before (Barkas and Watson, 1979; Richmond et al., 1982). Many human strains also bound bovine IgG, although the overall percentage of positive responses was slightly decreased as compared with rabbit IgG. Fewer chicken isolates bound bovine IgG and rabbit IgG when compared with human strains. A substantial proportion of the chicken and bovine isolates, however, reacted with chicken IgG (Table 4) compared with human strains of S. aureus. These qualitative differences in antibody binding patterns suggest that the type of protein A produced by S. aureus of different origins may differ slightly. Miiller et al. (1983) also observed differences in protein A activity; they tested porcine IgG with S. aureus isolated from different animals. Species-to-species differences in immunoglobulin structure have evolved through the years, resulting from changes in amino acid composi-
tion. This has led to differences in the Fc regions of immunoglobulins of different species and immunoglobulin classes and subclasses. Strains of 5, aureus may also have adapted to various hosts by producing types of protein A with subtle differences in Fc receptor domains. This may offer a competitive advantage to the S. aureus associated with chickens and other animals. The ability to bind a chicken immunoglobulin may enhance the survival of S. aureus. Protein A has been suggested to be linked with pathogenesis by inhibiting or reducing opsonization because of its Fc-binding ability (Forsgren and Quie, 1974; Peterson et al., 1977). Such a mechanism as binding the chicken's immunoglobulins may shield the staphylococcal cells from the chicken's immune system. In addition to qualitative differences in binding of poultry IgG by S. aureus of different origins, quantitative differences were observed and will be reported elsewhere (Schwan and Hartman, unpublished).
aureus
1
4
3
2
6 6 0 0
48 74 71 0 0
(%)
0 0
7 5 3 0 0
(n)
0 0
26 22 21 0 0
(%)
Rabb it IgG (±) 2
0 0
3 0 2 0 0
(n)
(+)
0 0
11 0 14 0 0
(%)
Chicken IgG
0 0
12 6 6 0 0
(n)
(
Chi I
Species other than Staphylococcus (twoMicrococcus sp., Escherichia coli, Bacillus subtilis, one Salmonella, Strepto
Coagulase-negative staphylococci.
± Values of .10 to .29 at 405 nm.
+ Values of >.30 at 405 nm.
13 17 10 0 0
27 23 14 1
0 0
(n)
(n)
Total
T-i
Coag—staph.3 Nonstaph. species4
Staphylococcus Chicken Human Bovine Turkey Pheasant
Origin of cultures
Rabbit IgG (+) 1
TABLE 4. The numbers and percentages of positive reactions obtained when cultures were reacted with rabbit, by using a tube enzyme immunoassay
ed from http://ps.oxfordjournals.org/ at New York University on April 24, 2015
SCHWAN AND HARTMAN
702
TABLE 5. Specific IgG binding patterns of selected Staphylococcus aureus cultures as determined by using a tube enzyme immunoassay Origin and isolate or strain
Rabbit IgG
Chicken IgG
Bovine IgG
Human Cowan I Wood 46 ISP 474 ISP 893 ISP 1066
Bovine Drivold 6 P-48 Davis 6 Denison 8 Cockrum 1
Determined from absorbance values at 405 nm; + = >.30, ± = .10 to .29, — = <.09.
REFERENCES Anderson, S. G., M. W. Bentzon, V. Houba, and P. Krag, 1970. International reference preparation of rheumatoid arthritis serum. Bull. World Health Org. 42:311-318. Barkas, T., and C.M.J. Watson, 1979. Induction of an Fc conformational change by binding of antigen: the generation of protein A-reactive sites in chicken immunoglobulin. Immunology 36:557— 561. Carret, G., J. P. Flandrois, and J. Fougerat, 1981. Detection of Staphylococcus aureus protein A by passive hemagglutination test using microtiter plates. Zentralbl. Bakteriol. Hyg., Abt. I. Orig. A 249:32-38. Christensen, P., and V.-A. Oxelius, 1974. Quantitation of the uptake of human IgG by some streptococci groups A, B, C, and G. Acta Pathol. Microbiol. Scand., Sect. B 82:475-483. Cole, J. R., Jr., 1984. Micrococcus and Staphylococcus, Pages 161—166 in Diagnostic Procedures in Veterinary Bacteriology and Mycology. 4th ed. G. R. Carter, ed. Charles C. Thomas, Springfield, IL. Devriese, L. A., 1984. A simplified system for biotyping Staphylococcus aureus strains isolated from different animal species. J. Appl. Bacteriol. 49:1-11. Devriese, L. A., and P. Oeding, 1976. Characteristics of Staphylococcus aureus strains isolated from different animal species. Res. Vet. Sci. 21:284— 291. Devriese, L. A., and A. Van de Kerckhove, 1979.
A comparison of methods and the validity of deoxyribonuclease tests for the characterization of staphylococci isolated from animals. J. Appl. Bacteriol. 4 6 : 3 8 5 - 3 9 3 . Forsgren, A., 1970. Significance of protein A production by staphylococci. Infect. Immun. 2:672— 673. Forsgren, A., and P. G. Quie, 1974. Effects of staphylococcal protein A on heat labile opsonins. J. Immunol. 112:1177-1180. Goudswaard, J., J. A. Van der Donk, A. Noordzij, R. H. Van Dam, and J.-P. Vaerman, 1978. Protein A reactivity of various mammalian immunoglobulins. Scand. J. Immunol. 8:21—28. Hajek, V., and E. Marsalek, 1976. Evaluation of classificatory criteria for staphylococci. Pages 11—21 in Staphylococci and Staphylococcal Diseases. J. Jeljaszewicz, ed. Gustav Fischer Verlag, New York, NY. Hovelius, B., and P.-A. Mardh, 1979. Haemagglutination by Staphylococcus saprophyticus and other staphylococcal species. Acta Pathol. Microbiol. Scand., Sect. B 87:45-50. Kronvall, G., U. S. Seal, S. Svensson, and R. C. Williams, Jr., 1974. Phylogenetic aspects of staphylococcal protein A-reactive serum globulins in birds and mammals. Acta Pathol. Microbiol. Scand., Sect. B 82:12-18. Lachica, R.V.F., C. A. Genigeorgis, and P. D. Hoeprich, 1979. Occurrence of protein A in Staphylococcus aureus and closely related Staphylococcus species. J. Clin. Microbiol. 10:752-753. Miiller, H. P., O. M. Litke, and H. Blobel, 1983.
Downloaded from http://ps.oxfordjournals.org/ at New York University on April 24, 2015
Chicken BS2 BS5 BS13 BS14 BS16
IMMUNOGLOBULIN BINDING
Protein A activity of staphylococci of different animal origins with immunoglobulin G from domestic and experimental animals and from man. Zentralbl. Veterinaermed. Reihe B 30:305 — 312. Myhre, E. B., O. Holmberg, and G. Kronvall, 1979. Immunoglobulin-binding structure on bovine group G streptococci different from type III Fc receptor on human group G streptococci. Infect. Immun. 2 3 : 1 - 7 . Peterson, P. K., J. Verhoff, L. D. Sabath, and P. G. Quie, 1977. Effect of protein A on staphylococcal opsonization. Infect. Immun. 15:760— 764. Richman, D. D., P. H. Cleveland, M. N. Oxman, and K. M. Johnson, 1982. The binding of staphylo-
703
coccal protein A by the sera of different animal species. J. Immunol. 128:2300-2305. Sjoquist, J., and G. Staleriheim, 1969. Protein A from Staphylococcus aureus. IX. Complement-fixing activity of protein A-IgG complexes. J. Immunol. 103:467-473. Voller, A., D. E. Bidwell, and A. Bartlett, 1976. Enzyme immunoassays in diagnostic medicine. Bull World Health Org. 53:55-65. Williams, C. A., and M. W. Chase, 1971. Methods in Immunology and Immunochemistry. Vol. 1. Academic Press, Inc., New York, NY. Winblad, S., and C. Ericson, 1973. Sensitized sheep red blood cells as a reactant for Staphylococcus aureus protein A. Acta Pathol. Microbiol. Scand., Sect. B 81:150-156. Downloaded from http://ps.oxfordjournals.org/ at New York University on April 24, 2015
1986 Poultry Science 65:696-703
INTRODUCTION
Staphylococcus aureus, a common inhabitant of poultry and most mammals, is responsible for a number of pathological conditions (Cole, 1984). Although 5. aureus is a single species, biotypes isolated from different animals differ in the cellular products they produce. Animal and human strains differ, for example, in the type of hemolysin produced, pigmentation, coagulation of plasma, staphylokinase production, growth on crystal violet (CV) agar, deoxyribonuclease (DNase) production, and bacteriophage type (Devriese, 1984; Devriese and Oeding, 1976). Many human strains of S. aureus produce the cellular product protein A, which possesses the capability of binding the Fc region of immunoglobulins, especially immunoglobulin G (IgG) (Sjoquist and Stalenheim, 1969). Protein A production is a variable property of poultry, and other origins of S. aureus (Hajek and Marsalek, 1976) and differences in IgG binding exist for human strains of S. aureus (Richman
'Journal Paper Number J-11750 of the Iowa Agriculture and Home Economics Experiment Station, Ames, IA. Projects 2377 and 2678. 1 To whom correspondence should be addressed.
696
et al., 1982). These observations as well as the differences in other cellular products led us to believe that preparations of protein A that possess different immundglobulin-binding properties might exist. The purpose of the present investigation, therefore, was to study the binding of various IgG molecules to strains of S. aureus obtained from different origins, including poultry. An enzyme immunoassay (EIA) was developed to assess qualitative differences in IgG binding and to correlate binding to the origin of the S, aureus.
MATERIALS AND METHODS
Bacterial Cultures. Chicken S. aureus and coagulase-negative staphylococci were obtained from the exterior of live White Leghorn hens and processed chickens. The S. aureus cultures were confirmed by biochemical and morphological tests. Bovine isolates of S. aureus were obtained from R. A. Packer (Iowa State University, Ames, IA). Strains of human S. aureus, including Cowan I, which produces high amounts of protein A, and Wood 46, which produces low amounts of protein A, were supplied by P. A. Pattee (Iowa State University, Ames, IA). Two strains of Staphylococcus epidermidis (protein A-negative) and six nonstaphylococcal species (two Micrococcus sp., Bacillus subtilis,
Downloaded from http://ps.oxfordjournals.org/ at New York University on April 24, 2015
ABSTRACT Sixty-six strains of Staphylococcus aureus of avian, bovine, and human origin were tested for their ability to bind chicken, bovine, and rabbit immunoglobulin G (IgG). A microtitration plate hemagglutination assay and a direct-tube enzyme immunoassay were used to determine qualitative differences. Twice as many chicken and bovine S. aureus isolates than human strains reacted positively to chicken IgG. Mean binding values of chicken IgG were also twice as high for chicken and bovine S. aureus isolates when compared with human-derived strains. Isolates of bovine and human origin displayed a high affinity for bovine IgG and rabbit IgG, whereas few isolates from chickens bound substantial quantities of the nonavian IgG species. These results demonstrate that preparations of staphylococcal protein A with affinities for immunoglobulins from poultry and other animals can be obtained by screening large numbers of isolates, especially those obtained from an animal species, such as chickens, that is the same as the source of the immunoglobulin one wishes to bind. (Key words: staphylococcal protein A, enzyme immunoassay, immunoglobulin G binding)
697
IMMUNOGLOBULIN BINDING
Preparation of Enzyme-Conjugated Antibody. A one step glutaraldehyde method of Voller et al. (1976) was used to conjugate alkaline phosphatase (Sigma Type VII-T) to antibody. Antibodies used for conjugation were chromatographically purified chicken IgG,
bovine IgG, and rabbit IgG (Sigma Chemical Company, St. Louis, MO). After conjugation and dialysis, the 1-ml volume of IgG-alkaline phosphatase conjugate was diluted to 3 ml with Tris-HCl buffer (.05 M, pH 8.0) containing 1.0% bovine serum albumen and .02% NaN 3 . The conjugate was stored in the dark at 4 C. Tube Enzyme Immunoassay. Staphylococcal cultures grown in 8 ml of BHI broth at 3-7 C for 24 hr were pelleted by centrifugation at 1000 X g for 10 min. Supernatant fluids were decanted, and each pellet was resuspended in .5 ml of PBS (.01 M, pH 7.4). A .5-ml volume of enzyme-conjugated IgG diluted in PBS was added to each tube. After incubation for 1 hr at 35 C, the tubes were rinsed three times with .5-ml volumes of PBS by using centrifugation at 1000 X g for 10 min after each wash. Then .5 ml of substrate (p-nitrophenyl phosphate, Sigma #104-105) was added to each tube. The substrate was previously diluted to 1 mg/ml in substrate buffer (10% diethanolamine, pH 9.8) before use in the assay (Voller et al., 1976). Hydrolysis was stopped after 30 min at room temperature by adding .3 ml of 3 N NaOH. Cells were pelleted at 1600 X g, and 150 (i\ of supernatant was transferred from each tube to a well of a microtitration plate. Each culture was analyzed spectrophotometrically at 405 nm with a Micro-ELISA Plate Reader (Dynatech Laboratories, Alexandria, VA). Controls were run with each analysis. Optimal dilutions of enzyme-conjugated antibody for use in the tube EIA were determined for each of the three conjugates (chicken IgG, bovine IgG, and rabbit IgG). To determine these optimal concentrations, dilutions of enzyme-labeled antibody from 1:100 to 1:1000 were analyzed by the tube EIA method described previously. The ratio of the positive control (Cowan I) to the negative control (S. epidermidis) was used to determine the optimal dilution of each conjugate. RESULTS
Analysis of Growth on Crystal Violet Agar and Deoxyribonuclease Production. Percentages of S. aureus of bovine, chicken, and human origin that were DNase-positive and percentages that were CV Type A or C are shown in Table 1. A high percentage of bovine isolates and human strains produced heat-stable DNase (93 and 100%, respectively) as compared with chicken isolates (74%). A majority of the
Downloaded from http://ps.oxfordjournals.org/ at New York University on April 24, 2015
Escherichia coli, Streptococcus zooepidemicus, and a Salmonella sp.) were obtained from the Iowa State University Microbiology Department Culture Collection (Ames, IA). A strain of turkey and a strain of pheasant S. aureus were supplied by L. J. Hoffman (Veterinary Diagnostic Laboratory, Ames, IA). Comparison of Deoxyribonuclease Production and Growth on Crystal Violet Agar. All S. aureus cultures were grown on CV agar (Difco tryptose agar containing 8 jug/ml CV). The color of growth after 24 and 48 hr was recorded according to Devriese (1984) to distinguish Type A from Type C cultures. Every S. aureus culture was also grown on Difco DNase agar containing .03 g/liter of toluidine blue, which was used as an indicator of deoxyribonucleic acid hydrolysis (Devriese and Oeding, 1976). The cultures were incubated for 24 hr, flooded with 1 N HC1, and observed for DNase activity. Micro titration Hemagglutination Assay. A microtitration hemagglutination procedure was used to test the chicken S. aureus and coagulase-negative staphylococci. Rabbit anti-sheep red blood cell sera prepared according to Winblad and Ericson (1973) and chicken antisheep red blood cell sera prepared according to Williams and Chase (1971) were used to sensitize sheep red blood cells as reported by Anderson et al. (1970). These sensitized sheep red blood cells were used to detect antibody binding to staphylococcal cells in a microtitration hemagglutination assay according to Sjoquist and Stalenheim (1969) as modified by Rollo (E. E. Rollo, 1982, M.S. thesis, Iowa State University, Ames, IA). The technique was as follows: 50 /ill of 1% sensitized sheep red blood cells was added to each well of a U-bottomed microtitration plate (Cooke Engineering Co., San Mateo, CA) that contained a loopful of culture from a brain-heart infusion (BHI) agar plate (Difco) suspended in 50 jul of phosphatebuffered saline (PBS) (.01 Af, pH 7.4). Microtitration plates were incubated for 2 hr at room temperature, examined, incubated overnight at 4 C, and examined again. Cowan I served as a positive control, and S. epidermidis served as a negative control.
698
SCHWAN AND HARTMAN TABLE 1. Origin of Staphylococcus aureus strains and percentages that produced heat-stable deoxyribonuclease (DNase) and crystal violet (CV) reactions
Origin of S. aureus
Number of strains
Bovine Chicken Human
14 100 23
CV Type
(% positive)
(%A:%C) 43:57 59:41 13:87
93 74
100
origin bound less rabbit IgG (A405 = .26). For chicken IgG, chicken and bovine S. aureus displayed comparable A40s of .12 and.14, respectively, whereas human S. aureus bound chicken IgG less well (A 40 s = .07). Reactions with bovine IgG were greatest for human strains (A405 = .79), followed by bovine (A40S = .51), and then by chicken (A405 = .38) isolates. Coagulase-negative staphylococci, turkey, and pheasant strains of S. aureus, and nonstaphylococcal species displayed low absorbance values (less than A405 = .10) for all three types of IgG. The two coagulase-negative isolates and the one S. aureus isolate that showed positive responses and autoagglutination in the microtitration hemagglutination assay also exhibited low absorbance values (less than A405 = .10) in the direct tube EIA. To categorize further the isolates within each group as producing strong or weak reactions, an absorbance value equal to or greater than .30 was considered a strong positive response, and an absorbance value of .10 to .29 was considered a weak positive response (Table 4). Human and bovine strains of S. aureus again showed similarities when reacted with rabbit IgG (total percentages of positive results were 96 and 92%, respectively) and bovine IgG (91 and 100%, respectively). Only 74% of chicken S. aureus isolates bound rabbit or bovine IgG. Only 26% of human S. aureus bound chicken IgG, whereas 57% of bovine and 55% of chicken isolates bound chicken IgG. Binding Patterns of Selected Staphylococcus aureus Cultures. Additional differences in binding patterns are shown in Table 5. Some cultures bound only rabbit IgG, others bound only chicken IgG, and others bound IgG preparations from two or three species. DISCUSSION Among strains of S. aureus obtained from
Downloaded from http://ps.oxfordjournals.org/ at New York University on April 24, 2015
bovine isolates and human strains were CV Type C, whereas a majority of the chicken isolates were CV Type A. Preliminary Identification of Antibody Binding. A qualitative microtitration hemagglutination assay was used to determine how many coagulase-negative and coagulase-positive chicken staphylococci would bind rabbit and chicken antibodies (Table 2). Of 175 coagulasenegative chicken staphylococci tested, only 1% yielded a positive test. Of 100 chicken S. aureus isolates, 27% showed reactivity to one or both antisera. Most S. aureus isolates that displayed a positive response did so only with rabbit antiserum (18%). However, 9% of the S. aureus isolates bound chicken antiserum alone (4%) or in combination with rabbit antiserum (5%). Cowan I reacted only with rabbit antiserum, and the 5. epidermidis strains showed no response to either antisera. The two coagulasenegative chicken isolates and one of the S. aureus isolates that exhibited positive responses also displayed sporadic autogglutination. Analysis with the Tube EIA. A sensitive tube EIA was used to determine which specific class of antibody, especially chicken IgG, was bound to the strains of S. aureus. Optimal dilutions of the three conjugates (rabbit, chicken, and bovine) based on the ratio of binding to Cowan IIS. epidermidis were 1/500 for rabbit IgG (P/N = .68/.01), 1/200 for chicken IgG (P/N = .12/.01), and 1/600 for bovine IgG (P/N = .75/.01). When the optimum dilutions of each species of IgG were used, a wide range of absorbance values was detected (Table 3). Comparisons could be made of mean binding differences between origin of the staphylococci and the type of IgG that was bound. Similar mean values of absorbance can be seen for human (A405 = .47) and bovine (A405 = 49) S. aureus when reacted with rabbit IgG; isolates of S. aureus of chicken
DNase
IMMUNOGLOBULIN BINDING
699
Downloaded from http://ps.oxfordjournals.org/ at New York University on April 24, 2015
different ecological niches, there are differences in the amounts and types of cellular products produced. Fewer strains isolated from poultry produce heat-stable DNase as detected with the conventional plating procedure (Devriese and Van de Kerckhove, 1979), an observation confirmed in the present study. Devriese (1984) biotyped S. aureus, based in part on CV reactions. Poultry and bovine strains of S. aureus were classified primarily as CV Type A, whereas human biotypes were grouped primarily into CV Type C. The results presented in this study correlate with Devriese's typing schemes for human and chicken S. aureus. Although the CV reactions of the bovine S. aureus did not correlate as well as in Devriese's study, a substantial percentage of the isolates (43%) still fell into the CV Type A group. Because different strains of S. aureus vary in production of some cellular products, it was thought that strains of S. aureus obtained from poultry and other animal species might also produce protein A with different binding affinities to IgG. These differences could result from adaptation of the S. aureus to specific ecological niches (animal species). This adaptation could include modification of the protein A produced so that the protein A would bind immunoglobulin(s) of the host species. Therefore, we attempted to determine if there was an association between species of origin and immunoglobulin-binding properties of different strains of S. aureus. Immunoglobulin-binding patterns of human strains of protein A-bearing S. aureus have been well characterized (Goudswaard et al., 1978; Kronvall et al., 1974), but little has been done to characterize antibody binding by poultry and other animal strains of S. aureus. Differences exist, however, in immunoglobulin binding by several groups of streptococci (Christensen and Oxelius, 1974; Myhre et al, 1979). The micro titration hemagglutination assay used as an initial screen for detection of antibody binding to the chicken staphylococci provided sensitivity but also had limitations because of autoagglutination. Autoagglutination has been seen by other investigators working with coagulase-negative staphylococci (Hovelius and Mardh, 1979) and S. aureus (Carret et al., 1981). Because of the autoagglutination problem, a direct tube EIA was used to detect antibody binding differences. Most human strains of S. aureus bound rabbit IgG within the range noted by others (Forsgren,
SCHWAN AND HARTMAN
700
TABLE 3. Numbers of staphylococci tested by using a tube enzyme immunoassay and mean absorbance values obtained with immunoglobulin G (IgG) preparations from three different animals T y p e of isolate or strain
Total
Rabbit IgG
Chicken IgG
Bovine IgG
.26 .47 .49 .04 .02
.12 .07 .14 .05 .00
.38 .79 .51 .04 .04
(n) Staphylococcus Chicken Human Bovine
aureus
Pheasant Chicken coag — s t a p h . 2 N o n s t a p h . species 3
4 6
± ± ± ± ±
.16' .24 .29 .01 .00
.01 ± .00 .00 + .00
± ± ± ± ±
.08 .05 .10 .01 .00
. 0 1 ± .02 .01 ± .01
+ ± ± + +
.25 .24 .29 .01 .01
.02 ± .03 .02 + .01
Downloaded from http://ps.oxfordjournals.org/ at New York University on April 24, 2015
Turkey
27 23 14 1 1
'Mean absorbance ± standard deviation measured at 405 nm. 2 Coagulase-negative staphylococci. 3 Species other than Staphylococcus (two Micrococcus sp., Escherichia coli, Bacillus subtilis, one Salmonella, Streptococcus zooepidemicus).
1970; Lachica et al, 1979), but few bound chicken IgG. This limited reactivity to chicken IgG, and chicken immune complexes has been noted before (Barkas and Watson, 1979; Richmond et al., 1982). Many human strains also bound bovine IgG, although the overall percentage of positive responses was slightly decreased as compared with rabbit IgG. Fewer chicken isolates bound bovine IgG and rabbit IgG when compared with human strains. A substantial proportion of the chicken and bovine isolates, however, reacted with chicken IgG (Table 4) compared with human strains of S. aureus. These qualitative differences in antibody binding patterns suggest that the type of protein A produced by S. aureus of different origins may differ slightly. Miiller et al. (1983) also observed differences in protein A activity; they tested porcine IgG with S. aureus isolated from different animals. Species-to-species differences in immunoglobulin structure have evolved through the years, resulting from changes in amino acid composi-
tion. This has led to differences in the Fc regions of immunoglobulins of different species and immunoglobulin classes and subclasses. Strains of 5, aureus may also have adapted to various hosts by producing types of protein A with subtle differences in Fc receptor domains. This may offer a competitive advantage to the S. aureus associated with chickens and other animals. The ability to bind a chicken immunoglobulin may enhance the survival of S. aureus. Protein A has been suggested to be linked with pathogenesis by inhibiting or reducing opsonization because of its Fc-binding ability (Forsgren and Quie, 1974; Peterson et al., 1977). Such a mechanism as binding the chicken's immunoglobulins may shield the staphylococcal cells from the chicken's immune system. In addition to qualitative differences in binding of poultry IgG by S. aureus of different origins, quantitative differences were observed and will be reported elsewhere (Schwan and Hartman, unpublished).
aureus
1
4
3
2
6 6 0 0
48 74 71 0 0
(%)
0 0
7 5 3 0 0
(n)
0 0
26 22 21 0 0
(%)
Rabb it IgG (±) 2
0 0
3 0 2 0 0
(n)
(+)
0 0
11 0 14 0 0
(%)
Chicken IgG
0 0
12 6 6 0 0
(n)
(
Chi I
Species other than Staphylococcus (twoMicrococcus sp., Escherichia coli, Bacillus subtilis, one Salmonella, Strepto
Coagulase-negative staphylococci.
± Values of .10 to .29 at 405 nm.
+ Values of >.30 at 405 nm.
13 17 10 0 0
27 23 14 1
0 0
(n)
(n)
Total
T-i
Coag—staph.3 Nonstaph. species4
Staphylococcus Chicken Human Bovine Turkey Pheasant
Origin of cultures
Rabbit IgG (+) 1
TABLE 4. The numbers and percentages of positive reactions obtained when cultures were reacted with rabbit, by using a tube enzyme immunoassay
ed from http://ps.oxfordjournals.org/ at New York University on April 24, 2015
SCHWAN AND HARTMAN
702
TABLE 5. Specific IgG binding patterns of selected Staphylococcus aureus cultures as determined by using a tube enzyme immunoassay Origin and isolate or strain
Rabbit IgG
Chicken IgG
Bovine IgG
Human Cowan I Wood 46 ISP 474 ISP 893 ISP 1066
Bovine Drivold 6 P-48 Davis 6 Denison 8 Cockrum 1
Determined from absorbance values at 405 nm; + = >.30, ± = .10 to .29, — = <.09.
REFERENCES Anderson, S. G., M. W. Bentzon, V. Houba, and P. Krag, 1970. International reference preparation of rheumatoid arthritis serum. Bull. World Health Org. 42:311-318. Barkas, T., and C.M.J. Watson, 1979. Induction of an Fc conformational change by binding of antigen: the generation of protein A-reactive sites in chicken immunoglobulin. Immunology 36:557— 561. Carret, G., J. P. Flandrois, and J. Fougerat, 1981. Detection of Staphylococcus aureus protein A by passive hemagglutination test using microtiter plates. Zentralbl. Bakteriol. Hyg., Abt. I. Orig. A 249:32-38. Christensen, P., and V.-A. Oxelius, 1974. Quantitation of the uptake of human IgG by some streptococci groups A, B, C, and G. Acta Pathol. Microbiol. Scand., Sect. B 82:475-483. Cole, J. R., Jr., 1984. Micrococcus and Staphylococcus, Pages 161—166 in Diagnostic Procedures in Veterinary Bacteriology and Mycology. 4th ed. G. R. Carter, ed. Charles C. Thomas, Springfield, IL. Devriese, L. A., 1984. A simplified system for biotyping Staphylococcus aureus strains isolated from different animal species. J. Appl. Bacteriol. 49:1-11. Devriese, L. A., and P. Oeding, 1976. Characteristics of Staphylococcus aureus strains isolated from different animal species. Res. Vet. Sci. 21:284— 291. Devriese, L. A., and A. Van de Kerckhove, 1979.
A comparison of methods and the validity of deoxyribonuclease tests for the characterization of staphylococci isolated from animals. J. Appl. Bacteriol. 4 6 : 3 8 5 - 3 9 3 . Forsgren, A., 1970. Significance of protein A production by staphylococci. Infect. Immun. 2:672— 673. Forsgren, A., and P. G. Quie, 1974. Effects of staphylococcal protein A on heat labile opsonins. J. Immunol. 112:1177-1180. Goudswaard, J., J. A. Van der Donk, A. Noordzij, R. H. Van Dam, and J.-P. Vaerman, 1978. Protein A reactivity of various mammalian immunoglobulins. Scand. J. Immunol. 8:21—28. Hajek, V., and E. Marsalek, 1976. Evaluation of classificatory criteria for staphylococci. Pages 11—21 in Staphylococci and Staphylococcal Diseases. J. Jeljaszewicz, ed. Gustav Fischer Verlag, New York, NY. Hovelius, B., and P.-A. Mardh, 1979. Haemagglutination by Staphylococcus saprophyticus and other staphylococcal species. Acta Pathol. Microbiol. Scand., Sect. B 87:45-50. Kronvall, G., U. S. Seal, S. Svensson, and R. C. Williams, Jr., 1974. Phylogenetic aspects of staphylococcal protein A-reactive serum globulins in birds and mammals. Acta Pathol. Microbiol. Scand., Sect. B 82:12-18. Lachica, R.V.F., C. A. Genigeorgis, and P. D. Hoeprich, 1979. Occurrence of protein A in Staphylococcus aureus and closely related Staphylococcus species. J. Clin. Microbiol. 10:752-753. Miiller, H. P., O. M. Litke, and H. Blobel, 1983.
Downloaded from http://ps.oxfordjournals.org/ at New York University on April 24, 2015
Chicken BS2 BS5 BS13 BS14 BS16
IMMUNOGLOBULIN BINDING
Protein A activity of staphylococci of different animal origins with immunoglobulin G from domestic and experimental animals and from man. Zentralbl. Veterinaermed. Reihe B 30:305 — 312. Myhre, E. B., O. Holmberg, and G. Kronvall, 1979. Immunoglobulin-binding structure on bovine group G streptococci different from type III Fc receptor on human group G streptococci. Infect. Immun. 2 3 : 1 - 7 . Peterson, P. K., J. Verhoff, L. D. Sabath, and P. G. Quie, 1977. Effect of protein A on staphylococcal opsonization. Infect. Immun. 15:760— 764. Richman, D. D., P. H. Cleveland, M. N. Oxman, and K. M. Johnson, 1982. The binding of staphylo-
703
coccal protein A by the sera of different animal species. J. Immunol. 128:2300-2305. Sjoquist, J., and G. Staleriheim, 1969. Protein A from Staphylococcus aureus. IX. Complement-fixing activity of protein A-IgG complexes. J. Immunol. 103:467-473. Voller, A., D. E. Bidwell, and A. Bartlett, 1976. Enzyme immunoassays in diagnostic medicine. Bull World Health Org. 53:55-65. Williams, C. A., and M. W. Chase, 1971. Methods in Immunology and Immunochemistry. Vol. 1. Academic Press, Inc., New York, NY. Winblad, S., and C. Ericson, 1973. Sensitized sheep red blood cells as a reactant for Staphylococcus aureus protein A. Acta Pathol. Microbiol. Scand., Sect. B 81:150-156. Downloaded from http://ps.oxfordjournals.org/ at New York University on April 24, 2015