J Vet Diagn Invest 4:419-422 (1992)
Use of ELISA to detect toxigenic Pasteurella multocida in atrophic rhinitis in swine Terry L. Bowersock, Tom Hooper, Ronald Pottenger Abstract. The use of an enzyme-linked immunosorbent assay (ELISA) as a means of detecting dermonecrotoxin-producing strains of Pasteurella multocida was investigated. The assay was evaluated as a means to identify toxigenic P. multocida isolates recovered from nasal secretions of swine with atrophic rhinitis. The sensitivity and specificity of the ELISA for detecting dermonecrotoxin-producing P. multocida strains were compared to those of mouse-inoculation and cytotoxicity assays. The ELISA was highly sensitive and more specific than animal inoculation or tissue culture assay and is thus a more effective method for screening swine herds for the presence of toxigenic strains of P. multocida. The ELISA is a rapid, effective, economical way to identify toxigenic P. multocida isolates.
Pasteurella multocida is an important pathogen of the respiratory tract in many species and the primary pathogen associated with atrophic rhinitis (AR) in swine. Pasteurella multocida can be isolated from swine with AR and from unaffected animals. Attempts to experimentally produce AR often fail, indicating that a difference in virulence among P. multocida isolates may exist. The ability of some P. multocida strains to cause AR is associated with the elaboration of an exotoxin. This exotoxin, called dermonecrotoxin, has been purified, and the purified dermonecrotoxin induces atrophy of turbinates when administered directly onto the nasal mucosa or when injected parenterally into pigs. 7,11,16 Differentiation of toxigenic and nontoxigenic strains of P. multocida is essential for the accurate diagnosis, treatment, and prevention of AR in swine. Several tests are used to detect toxigenic strains, including mouse lethality, dermonecrotic effect on guinea pigs, and in 6,8,17,18,20 A differential test that vitro cytotoxicity assays. does not require live animals or the time and labor involved with cell culture is needed. The development of monoclonal antibodies to the dermonecrotoxin of P. multocida has resulted in an enzyme-linked immunosorbent assay (ELISA) that is highly specific and sensitive in detecting this toxin. Other advantages of the ELISA over the other procedures are that it allows testing of a large number of isolates quickly, efficiently, and inexpensively. The purpose of this experiment was to evaluate the usefulness of the ELISA described to
identify toxin-producing isolates obtained from nasal passages of pigs suspected of harboring P. multocida and to compare the sensitivity and specificity of the ELISA to the mouse-inoculation and cytotoxicity methods. Materials and methods Specimens. Nasal secretions were obtained from 322 pigs raised on 11 different commercial farms. During the farm visits, all pigs were observed for any evidence of AR, including deformed snouts, bloody or mucus-laden noses, sneezing, and watery eyes. All farms were farrow-to-finish operations except one unaffected farm that sold feeder pigs. All farms except the feeder pig operation used a commercial P. multocida toxoid vaccine. Pigs from 4 farms showed typical clinical signs of moderate to severe AR. Pigs from the other 7 farms showed no signs of AR at the time of sampling and were considered unaffected. The pigs varied in age from suckling ( 8 wk old). Rayon-tippeda or calcium alginateb swabs were used to collect the nasal secretions. The use of the 2 types of swabs was matched as closely as possible relative to the size of each pig. For the < 4-wk-old pigs, 33 affected and 36 unaffected pigs were swabbed; for the 4-8-wk-old pigs, 70 affected and 65 unaffected pigs were swabbed; and for the > 8-wk-old pigs, 59 affected and 59 unaffected pigs were swabbed. The swabs were transported to the laboratory on ice and streaked on blood agar (trypticase agar basec plus 5% defibrinated sheep blood) supplemented with 3.5 mg/liter bacitracind and neomycindd,2 and on MacConkey agarc within 4 hr of collection. Identification and typing ofbacteria. Each inoculated plate was incubated overnight at 37 C. Colonies exhibiting morphology typical of P. multocida were picked and identified by standard methods3 and assayed for dermonecrotoxin production. Colonies exhibiting morphology typical of Bordetella bronchiseptica were also identified using standard methods3 and recorded. All plates that did not demonstrate typical colonies of P. multocida or B. bronchiseptica were
From The Department of Veterinary Pathobiology, School of Veterinary Medicine, Purdue University, W. Lafayette, IN 47907-1243 (Bowersock), and The Animal Disease Diagnostic Laboratory, Southern Indiana Purdue Agricultural Center, Dubois, IN 475279666 (Hooper, Pottenger). Received for publication November 13, 1991. 419
Downloaded from vdi.sagepub.com by guest on March 29, 2015
420
Bowersock, Hooper, Pottenger Table 1. Culture results of atrophic rhinitis-affected swine herds and -unaffected herds.
incubated for an additional 24 hr and examined for further growth of these two bacteria. Pasteurella multocida isolates were typed as A or D by subculture onto dextrose starch agar,c which was crossstreaked with a S. aureus known to produce hylauronidase. Isolates that were sensitive to hyaluronidase were classified as type A, whereas those that were resistant to hyaluronidase were classified as type D.4 Toxin extraction. The technique to identify the dermonecrotoxin was performed as previously described. 10 Each isolate of P. multocida was transferred to a blood agar plate and streaked for heavy growth. After incubation for 24 hr at 37 C, the cultures were checked for purity, and all growth was scraped from the plate and suspended in 2 ml of sterile distilled water. Approximately 300-400 µ1 of this suspension was placed in a microtiter plate and held 24 hr at 37 C for use in the ELISA. In some experiments, an aliquot of the extract was frozen at -70 C so samples could be batched for future assays. Frozen samples were thawed, placed in a microtiter plate, and held 24 hr at 37 C for easy transfer to the ELISA test plates. In some cases, Escherichia coli was intentionally added to cultures of a known toxigenic and a nontoxigenic isolate of P. multocida prior to making the extract to test the specificity and sensitivity of the test. The extracted toxin was transferred to the ELISA test plate as indicated below. ELISA test. The ELISA method used was a simple sandwich format with avidin-biotin amplification. Microtiter platese were coated with 50 µ1/well of monoclonal antibody #M815f to the dermonecrotoxin of P. multocida and incubated overnight at 4 C. Plates were then incubated for 30 min at 37 C with 50 µl/well of bovine serum albumin c in phosphate-buffered saline (PBS), pH 7.2. The plates were washed twice with PBS containing Tween-20d and incubated at 22 C for 1 hr with 50 µl of the bacterial (toxin) extract to
be tested. After 2 washings, the monoclonal antibody #M816f to the dermonecrotoxin of P. multocida conjugated to biotin was added, and the plate was incubated for 1 hr at 22 C. After washing the plate twice with PBS-Tween, the plates were incubated at 22 C for 45 min with extravidin-peroxidase,d and color was developed with the substrate O-phenylenediamine.d The color reaction was stopped after 5 min with 1 M sulphuric acid.d The plates were read by visual observation after the color change was stopped; no detectable background color reaction occurred. Any orange color in a well was reported as positive for the presence of exotoxin. Comparison of ELISA to other toxicity assays. A total of 79 isolates that were screened for dermonecrotoxin production by other laboratories were tested using the ELISA protocol described above. Thirteen of these isolates had been screened by a mouse-inoculation test.g The in vitro cytotoxicity assay8 using feline fetal lung cells in an agar gel overlayh was used to screen the other 66 isolates. Statistical tests. Prevalence of toxigenic isolates was evaluated by determining the confidence interval for the proportion (n = 162) statistical test. 21 The total number of isolates from all pigs from affected herds (n = 162) was compared to the number of isolates from all pigs from unaffected herds (n = 160). Parameters in which the confidence interval for each group did not overlap were determined to be significantly different. The sensitivity, specificity, and positive predictive value of the ELISA results were determined by comparing results of isolates detected as positive or negative by either mouseinoculation or in vitro cytotoxicity assays to results obtained by ELISA using the statistical method previously described. 13 Isolates tested by other assays were twice as likely to be reported positive incorrectly than negative incorrectly compared with the ELISA test. The sensitivity of the ELISA test in this report is the probability of obtaining a positive result
Downloaded from vdi.sagepub.com by guest on March 29, 2015
ELISA to detect toxigenic P. multocida
in positive individuals. It was calculated by the ratio of the true-positive results (TP; those positive by ELISA and by another assay) to the sum of the TP and the false-negative results (FN; those negative by ELISA and positive by another assay). The specificity of the ELISA test is the probability of obtaining a negative result in individuals who are negative. It was calculated by the ratio of the true-negative results (TN; those negative by ELISA and another assay) to the sum of the TN plus false-positive results (FP; those positive by ELISA and negative by another assay). The positive predictive value (the probability that a test result is correct for a population) was also calculated by the ratio of TP to the sum of TP and FP.
421
Table 2. Specificity and sensitivity of ELISA assay compared with those of other biological assays used to detect the exotoxin of Pasteurella multocida.
Results The prevalence of toxigenic isolates, distribution of types A and D, and total number of P. multocida isolates as well as the prevalence of B. bronchiseptica recovered from AR-affected and -unaffected herds are shown in Table 1. The sensitivity and specificity of the ELISA assay were 80% and 85%, respectively, and the positive predictive value was 88% (Table 2). Discussion Pasteurella multocida was isolated with equal frequency from pigs that showed signs of AR and from pigs without apparent signs of AR. However, the herds with clinical signs of AR had a higher prevalence of group D toxigenic P. multocida and a lower prevalence of group A nontoxigenic P. multocida. The overall difference in the prevalence of P. multocida strains found in this study is consistent with those of prior reports 12,15,19 The prevalence of toxigenic P. multocida in AR-affected herds has been reported to vary from 0 to 60%, depending on the size of the herd. 12,15 Prevalence has been reported as high as 50-60% only in pigs with snout deformities in Europe. 19 The evaluation of this assay for sensitivity and specificity compared with other tests is somewhat deceptive because the ELISA may be more accurate. Animal tests and especially in vitro cytotoxicity tests are subject to both false-positive and false-negative interpretation. All the results that disagreed with the ELISA were found in the cytotoxicity assay. Animal-inoculation tests agreed with ELISA, although mouse-inoculation results were in some instances more difficult to interpret. The disagreement in interpretation occurs because of the inherent subjective evaluation of cytotoxicity assays, differences in the way many laboratories interpret their findings, and the presence of other substances in the supernatant fluids. With the monoclonal antibody test, the presence of the toxin is detected by a highly specific reaction to the toxin without any subjective interpretation of the assay. The disagreement between ELISA and other toxicity tests in 14 of 79 tests may be attributed to the subjective interpretation
and decreased sensitivity of other tests compared with ELISA. The sensitivity and specificity of ELISA may actually be much better than indicated in Table 2. Use of a DNA probe test on isolates of toxigenic and nontoxigenic P. multocida indicated 100% agreement with ELISA in the detection of toxigenic isolates (S. Singh, personal communication). Bordetella bronchiseptica was equally prevalent in affected and unaffected pigs and has been implicated in some studies as a cause of reversible turbinate atrophy and as a cause of AR.5 These results suggest that B. bronchiseptica is not likely to cause AR alone but may contribute to the severity of AR by enchancing the growth of P. multocida (Bemis DA: 1989, Abstr Int Conf HAP Organisms, Guelph, Ontario, Canada #27; Vert HP, et al.: 1989, Abstr Int Conf HAP Organisms, Guelph, Ontario, Canada #28). Vaccination with a Pasteurella multocida type D toxoid apparently did not protect pigs against type D P. multocida infection. All 4 affected herds reported using toxoid vaccines, and 6 of 7 of the unaffected herds reported using toxoid vaccines. The reason for the apparent failure of vaccination on the affected farms is not clear. Vaccination of gilts and sows against P. multocida and B. bronchiseptica has greatly reduced the incidence and severity of AR, especially in severely affected herds. 1,9,14 However, the recovery of toxigenic P. multocida and B. bronchiseptica from pigs of vaccinated dams may not be significantly different than the recovery from nonvaccinated dams.14 The history of AR in the herds used in this study is not known, and the severity of AR and prevalence of toxigenic P. multocida may have been greater prior to the use of the vaccine. Production records were not available; therefore, the effect of the vaccines could not be evaluated. Selection of affected herds was made on the basis of obvious clinical signs of AR; therefore vaccination probably was not highly effective in reducing the incidence of AR in these herds. These results suggest that
Downloaded from vdi.sagepub.com by guest on March 29, 2015
Bowersock, Hooper, Pottenger
422
management methods, genetics, nutrition, and environmental factors are as important as vaccination in the control of AR. There are several advantages to using the ELISA. The ELISA is easy to perform, can be completed within a workday, can be used to screen many samples at a time, is easy to interpret, produces objective results, and is economical. The ELISA was useful for screening herds for the prevalence of toxigenic P. multocida and for evaluating isolates for dermonecrotoxin production. Acknowledgements We thank Mary Jo Fisher for technical assistance and Ms. Denise Riley for manuscript preparation.
Sources and manufacturers a. Culturette, Baxter Healthcare, McGaw Park, IL. b. Transette 1, Spectrum Laboratories, Houston, TX. c. Difco Laboratories, Detroit, MI. d. Sigma Chemical Company, St. Louis, MO. e. Cooke Laboratory Products, Alexandria, VA. f. Dakopatts a/s, Clostrup, Denmark. g. Pasteurella multocida types A & D, National Animal Disease Laboratory, Ames, IA. h. Pasteurella multocida types A & D, Oxford Laboratories, Worthington, MN.
References 1. Barfod K, Pedersen KB: 1982, Synergism between Bordetella bronchiseptica and a toxin-producing strain of Pasteurella multocida in the causation of atrophic rhinitis in SPF pigs. Proc Int Pig Vet Soc Conf, Mexico City 7:112. 2. Barfod K, Pedersen KB: 1984, Influence of vaccination of sows with Bordetella-Pasteurella vaccines on the occurrence of atrophic rhinitis among their offspring after experimental infection with Bordetella bronchiseptica and toxigenic Pasteurella multocida. Nord Veterinaermed 36:337-345. 3. Carter GR, Cole JR Jr, ed.: 1990, Diagnostic procedure in veterinary bacteriology and mycology, pp. 87-93, 129-142. Academic Press, San Diego, CA. 4. Carter GR, Rundell SW: 1975, Identification of type A strains of Pasteurella multocida using staphylococcal hyaluronidase. Vet Rec 96:343. 5. Chanter N, Meygar T, Rutter JM: 1989, Interaction between B. bronchiseptica and toxigenic P. multocida in atrophic rhinitis in pigs. Res Vec Sci 47:48-53.
6. de Jong MF: 1983, In: Atrophic rhinitis in pigs, ed. Pedersen KB, Nielsen NC, pp. 136-146. CEC Agriculture EUR 8643 EN, EEC, Luxembourg. 7. Dominick MA, Remler RB: 1988, Turbinate osteoporosis in pigs following intranasal inoculation of purified Pasteurella toxin: histomorphometric and ultrastructural studies. Vet Pathol 25: 17-27. 8. Eamens GJ, Kirkland PD, Turner MJ, Gardner IA, et al.: 1988, Identification of toxigenic P. multocida in atrophic rhinitis of pigs by in vitro characterization. Aust Vet J 65:120-123. 9. Foged NT, Nielsen JP, Jorsal SE: 1989, Protection against progressive atrophic rhinitis by vaccination with Pasteurella multocida toxin purified by monoclonal antibodies. Vet Rec 125: 7-11. 10. Foged NT, Nielsen JP, Pedersen KB: 1988, Differentiation of toxigenic from nontoxigenic isolates of Pasteurella multocida by enzyme-linked immunosorbent assay. J Clin Microbiol 26: 1419-1420. 11. Foged NT, Pedersen KB, Elling F: 1987, Characterization and biological effects of Pasteurella multocida toxin. FEMS Microbiol Lett 43:45-51. 12. Gardner IA, Eamens GJ, Turner MJ, Homitzky CL: 1989, Toxigenic type D Pasteurella multocida in New South Wales pig herds-prevalence and factors associated with infection. Aust Vet J 66:318-321. 13. Gerstman BB, Cappucci OT: 1986, Evaluating the reliability of diagnostic tests. J Am Vet Med Assoc 188:248-251. 14. Giles CJ, Smith IM: 1983, Vaccination of pigs with Bordetella bronchiseptica. Vet Bull 53:327-338. 15. Goodwin RFW, Chanter N, Rutter JM: 1990, Detection and distribution of toxigenic Pasteurella multocida in pig herds with different degrees of atrophic rhinitis. Vet Rec 126:452-456. 16. Kamp EM, Kimman TG: 1988, Induction of nasal turbinate atrophy in germ-free pigs, using Pasteurella multocida as well as bacterium free crude and purified dermonecrotic toxin of P. multocida. Am J Vet Res 49: 1844-1849. 17. Pedersen KB: 1983, Cultural and serological diagnosis of atrophic rhinitis in pigs. In: Atrophic rhinitis in pigs, ed. Pedersen KB, Nielsen NC, pp. 22-31. CEC Agriculture EUR 8643 EN, EEC, Luxembourg. 18. Pedersen KB, Barfod K: 1981, The aetiological significance of Bordetella bronchiseptica and Pasteurella multocidia in atrophic rhinitis of swine. Nord Veterinaermed 33:513-522. 19. Rutter JM: 1987, Atrophic rhinitis in pigs. Pig News & Info 8: 385-387. 20. Rutter JM, Luther PD: 1984, Cell culture assay for toxigenic Pasteurella multocida from atrophic rhinitis in pigs. Vet Rec 114:393-396. 21. Samuels ML: 1989, Statistics for the life sciences, pp. 329-331. Dellen, San Francisco, CA.
Downloaded from vdi.sagepub.com by guest on March 29, 2015
Use of ELISA to detect toxigenic Pasteurella multocida in atrophic rhinitis in swine Terry L. Bowersock, Tom Hooper, Ronald Pottenger Abstract. The use of an enzyme-linked immunosorbent assay (ELISA) as a means of detecting dermonecrotoxin-producing strains of Pasteurella multocida was investigated. The assay was evaluated as a means to identify toxigenic P. multocida isolates recovered from nasal secretions of swine with atrophic rhinitis. The sensitivity and specificity of the ELISA for detecting dermonecrotoxin-producing P. multocida strains were compared to those of mouse-inoculation and cytotoxicity assays. The ELISA was highly sensitive and more specific than animal inoculation or tissue culture assay and is thus a more effective method for screening swine herds for the presence of toxigenic strains of P. multocida. The ELISA is a rapid, effective, economical way to identify toxigenic P. multocida isolates.
Pasteurella multocida is an important pathogen of the respiratory tract in many species and the primary pathogen associated with atrophic rhinitis (AR) in swine. Pasteurella multocida can be isolated from swine with AR and from unaffected animals. Attempts to experimentally produce AR often fail, indicating that a difference in virulence among P. multocida isolates may exist. The ability of some P. multocida strains to cause AR is associated with the elaboration of an exotoxin. This exotoxin, called dermonecrotoxin, has been purified, and the purified dermonecrotoxin induces atrophy of turbinates when administered directly onto the nasal mucosa or when injected parenterally into pigs. 7,11,16 Differentiation of toxigenic and nontoxigenic strains of P. multocida is essential for the accurate diagnosis, treatment, and prevention of AR in swine. Several tests are used to detect toxigenic strains, including mouse lethality, dermonecrotic effect on guinea pigs, and in 6,8,17,18,20 A differential test that vitro cytotoxicity assays. does not require live animals or the time and labor involved with cell culture is needed. The development of monoclonal antibodies to the dermonecrotoxin of P. multocida has resulted in an enzyme-linked immunosorbent assay (ELISA) that is highly specific and sensitive in detecting this toxin. Other advantages of the ELISA over the other procedures are that it allows testing of a large number of isolates quickly, efficiently, and inexpensively. The purpose of this experiment was to evaluate the usefulness of the ELISA described to
identify toxin-producing isolates obtained from nasal passages of pigs suspected of harboring P. multocida and to compare the sensitivity and specificity of the ELISA to the mouse-inoculation and cytotoxicity methods. Materials and methods Specimens. Nasal secretions were obtained from 322 pigs raised on 11 different commercial farms. During the farm visits, all pigs were observed for any evidence of AR, including deformed snouts, bloody or mucus-laden noses, sneezing, and watery eyes. All farms were farrow-to-finish operations except one unaffected farm that sold feeder pigs. All farms except the feeder pig operation used a commercial P. multocida toxoid vaccine. Pigs from 4 farms showed typical clinical signs of moderate to severe AR. Pigs from the other 7 farms showed no signs of AR at the time of sampling and were considered unaffected. The pigs varied in age from suckling ( 8 wk old). Rayon-tippeda or calcium alginateb swabs were used to collect the nasal secretions. The use of the 2 types of swabs was matched as closely as possible relative to the size of each pig. For the < 4-wk-old pigs, 33 affected and 36 unaffected pigs were swabbed; for the 4-8-wk-old pigs, 70 affected and 65 unaffected pigs were swabbed; and for the > 8-wk-old pigs, 59 affected and 59 unaffected pigs were swabbed. The swabs were transported to the laboratory on ice and streaked on blood agar (trypticase agar basec plus 5% defibrinated sheep blood) supplemented with 3.5 mg/liter bacitracind and neomycindd,2 and on MacConkey agarc within 4 hr of collection. Identification and typing ofbacteria. Each inoculated plate was incubated overnight at 37 C. Colonies exhibiting morphology typical of P. multocida were picked and identified by standard methods3 and assayed for dermonecrotoxin production. Colonies exhibiting morphology typical of Bordetella bronchiseptica were also identified using standard methods3 and recorded. All plates that did not demonstrate typical colonies of P. multocida or B. bronchiseptica were
From The Department of Veterinary Pathobiology, School of Veterinary Medicine, Purdue University, W. Lafayette, IN 47907-1243 (Bowersock), and The Animal Disease Diagnostic Laboratory, Southern Indiana Purdue Agricultural Center, Dubois, IN 475279666 (Hooper, Pottenger). Received for publication November 13, 1991. 419
Downloaded from vdi.sagepub.com by guest on March 29, 2015
420
Bowersock, Hooper, Pottenger Table 1. Culture results of atrophic rhinitis-affected swine herds and -unaffected herds.
incubated for an additional 24 hr and examined for further growth of these two bacteria. Pasteurella multocida isolates were typed as A or D by subculture onto dextrose starch agar,c which was crossstreaked with a S. aureus known to produce hylauronidase. Isolates that were sensitive to hyaluronidase were classified as type A, whereas those that were resistant to hyaluronidase were classified as type D.4 Toxin extraction. The technique to identify the dermonecrotoxin was performed as previously described. 10 Each isolate of P. multocida was transferred to a blood agar plate and streaked for heavy growth. After incubation for 24 hr at 37 C, the cultures were checked for purity, and all growth was scraped from the plate and suspended in 2 ml of sterile distilled water. Approximately 300-400 µ1 of this suspension was placed in a microtiter plate and held 24 hr at 37 C for use in the ELISA. In some experiments, an aliquot of the extract was frozen at -70 C so samples could be batched for future assays. Frozen samples were thawed, placed in a microtiter plate, and held 24 hr at 37 C for easy transfer to the ELISA test plates. In some cases, Escherichia coli was intentionally added to cultures of a known toxigenic and a nontoxigenic isolate of P. multocida prior to making the extract to test the specificity and sensitivity of the test. The extracted toxin was transferred to the ELISA test plate as indicated below. ELISA test. The ELISA method used was a simple sandwich format with avidin-biotin amplification. Microtiter platese were coated with 50 µ1/well of monoclonal antibody #M815f to the dermonecrotoxin of P. multocida and incubated overnight at 4 C. Plates were then incubated for 30 min at 37 C with 50 µl/well of bovine serum albumin c in phosphate-buffered saline (PBS), pH 7.2. The plates were washed twice with PBS containing Tween-20d and incubated at 22 C for 1 hr with 50 µl of the bacterial (toxin) extract to
be tested. After 2 washings, the monoclonal antibody #M816f to the dermonecrotoxin of P. multocida conjugated to biotin was added, and the plate was incubated for 1 hr at 22 C. After washing the plate twice with PBS-Tween, the plates were incubated at 22 C for 45 min with extravidin-peroxidase,d and color was developed with the substrate O-phenylenediamine.d The color reaction was stopped after 5 min with 1 M sulphuric acid.d The plates were read by visual observation after the color change was stopped; no detectable background color reaction occurred. Any orange color in a well was reported as positive for the presence of exotoxin. Comparison of ELISA to other toxicity assays. A total of 79 isolates that were screened for dermonecrotoxin production by other laboratories were tested using the ELISA protocol described above. Thirteen of these isolates had been screened by a mouse-inoculation test.g The in vitro cytotoxicity assay8 using feline fetal lung cells in an agar gel overlayh was used to screen the other 66 isolates. Statistical tests. Prevalence of toxigenic isolates was evaluated by determining the confidence interval for the proportion (n = 162) statistical test. 21 The total number of isolates from all pigs from affected herds (n = 162) was compared to the number of isolates from all pigs from unaffected herds (n = 160). Parameters in which the confidence interval for each group did not overlap were determined to be significantly different. The sensitivity, specificity, and positive predictive value of the ELISA results were determined by comparing results of isolates detected as positive or negative by either mouseinoculation or in vitro cytotoxicity assays to results obtained by ELISA using the statistical method previously described. 13 Isolates tested by other assays were twice as likely to be reported positive incorrectly than negative incorrectly compared with the ELISA test. The sensitivity of the ELISA test in this report is the probability of obtaining a positive result
Downloaded from vdi.sagepub.com by guest on March 29, 2015
ELISA to detect toxigenic P. multocida
in positive individuals. It was calculated by the ratio of the true-positive results (TP; those positive by ELISA and by another assay) to the sum of the TP and the false-negative results (FN; those negative by ELISA and positive by another assay). The specificity of the ELISA test is the probability of obtaining a negative result in individuals who are negative. It was calculated by the ratio of the true-negative results (TN; those negative by ELISA and another assay) to the sum of the TN plus false-positive results (FP; those positive by ELISA and negative by another assay). The positive predictive value (the probability that a test result is correct for a population) was also calculated by the ratio of TP to the sum of TP and FP.
421
Table 2. Specificity and sensitivity of ELISA assay compared with those of other biological assays used to detect the exotoxin of Pasteurella multocida.
Results The prevalence of toxigenic isolates, distribution of types A and D, and total number of P. multocida isolates as well as the prevalence of B. bronchiseptica recovered from AR-affected and -unaffected herds are shown in Table 1. The sensitivity and specificity of the ELISA assay were 80% and 85%, respectively, and the positive predictive value was 88% (Table 2). Discussion Pasteurella multocida was isolated with equal frequency from pigs that showed signs of AR and from pigs without apparent signs of AR. However, the herds with clinical signs of AR had a higher prevalence of group D toxigenic P. multocida and a lower prevalence of group A nontoxigenic P. multocida. The overall difference in the prevalence of P. multocida strains found in this study is consistent with those of prior reports 12,15,19 The prevalence of toxigenic P. multocida in AR-affected herds has been reported to vary from 0 to 60%, depending on the size of the herd. 12,15 Prevalence has been reported as high as 50-60% only in pigs with snout deformities in Europe. 19 The evaluation of this assay for sensitivity and specificity compared with other tests is somewhat deceptive because the ELISA may be more accurate. Animal tests and especially in vitro cytotoxicity tests are subject to both false-positive and false-negative interpretation. All the results that disagreed with the ELISA were found in the cytotoxicity assay. Animal-inoculation tests agreed with ELISA, although mouse-inoculation results were in some instances more difficult to interpret. The disagreement in interpretation occurs because of the inherent subjective evaluation of cytotoxicity assays, differences in the way many laboratories interpret their findings, and the presence of other substances in the supernatant fluids. With the monoclonal antibody test, the presence of the toxin is detected by a highly specific reaction to the toxin without any subjective interpretation of the assay. The disagreement between ELISA and other toxicity tests in 14 of 79 tests may be attributed to the subjective interpretation
and decreased sensitivity of other tests compared with ELISA. The sensitivity and specificity of ELISA may actually be much better than indicated in Table 2. Use of a DNA probe test on isolates of toxigenic and nontoxigenic P. multocida indicated 100% agreement with ELISA in the detection of toxigenic isolates (S. Singh, personal communication). Bordetella bronchiseptica was equally prevalent in affected and unaffected pigs and has been implicated in some studies as a cause of reversible turbinate atrophy and as a cause of AR.5 These results suggest that B. bronchiseptica is not likely to cause AR alone but may contribute to the severity of AR by enchancing the growth of P. multocida (Bemis DA: 1989, Abstr Int Conf HAP Organisms, Guelph, Ontario, Canada #27; Vert HP, et al.: 1989, Abstr Int Conf HAP Organisms, Guelph, Ontario, Canada #28). Vaccination with a Pasteurella multocida type D toxoid apparently did not protect pigs against type D P. multocida infection. All 4 affected herds reported using toxoid vaccines, and 6 of 7 of the unaffected herds reported using toxoid vaccines. The reason for the apparent failure of vaccination on the affected farms is not clear. Vaccination of gilts and sows against P. multocida and B. bronchiseptica has greatly reduced the incidence and severity of AR, especially in severely affected herds. 1,9,14 However, the recovery of toxigenic P. multocida and B. bronchiseptica from pigs of vaccinated dams may not be significantly different than the recovery from nonvaccinated dams.14 The history of AR in the herds used in this study is not known, and the severity of AR and prevalence of toxigenic P. multocida may have been greater prior to the use of the vaccine. Production records were not available; therefore, the effect of the vaccines could not be evaluated. Selection of affected herds was made on the basis of obvious clinical signs of AR; therefore vaccination probably was not highly effective in reducing the incidence of AR in these herds. These results suggest that
Downloaded from vdi.sagepub.com by guest on March 29, 2015
Bowersock, Hooper, Pottenger
422
management methods, genetics, nutrition, and environmental factors are as important as vaccination in the control of AR. There are several advantages to using the ELISA. The ELISA is easy to perform, can be completed within a workday, can be used to screen many samples at a time, is easy to interpret, produces objective results, and is economical. The ELISA was useful for screening herds for the prevalence of toxigenic P. multocida and for evaluating isolates for dermonecrotoxin production. Acknowledgements We thank Mary Jo Fisher for technical assistance and Ms. Denise Riley for manuscript preparation.
Sources and manufacturers a. Culturette, Baxter Healthcare, McGaw Park, IL. b. Transette 1, Spectrum Laboratories, Houston, TX. c. Difco Laboratories, Detroit, MI. d. Sigma Chemical Company, St. Louis, MO. e. Cooke Laboratory Products, Alexandria, VA. f. Dakopatts a/s, Clostrup, Denmark. g. Pasteurella multocida types A & D, National Animal Disease Laboratory, Ames, IA. h. Pasteurella multocida types A & D, Oxford Laboratories, Worthington, MN.
References 1. Barfod K, Pedersen KB: 1982, Synergism between Bordetella bronchiseptica and a toxin-producing strain of Pasteurella multocida in the causation of atrophic rhinitis in SPF pigs. Proc Int Pig Vet Soc Conf, Mexico City 7:112. 2. Barfod K, Pedersen KB: 1984, Influence of vaccination of sows with Bordetella-Pasteurella vaccines on the occurrence of atrophic rhinitis among their offspring after experimental infection with Bordetella bronchiseptica and toxigenic Pasteurella multocida. Nord Veterinaermed 36:337-345. 3. Carter GR, Cole JR Jr, ed.: 1990, Diagnostic procedure in veterinary bacteriology and mycology, pp. 87-93, 129-142. Academic Press, San Diego, CA. 4. Carter GR, Rundell SW: 1975, Identification of type A strains of Pasteurella multocida using staphylococcal hyaluronidase. Vet Rec 96:343. 5. Chanter N, Meygar T, Rutter JM: 1989, Interaction between B. bronchiseptica and toxigenic P. multocida in atrophic rhinitis in pigs. Res Vec Sci 47:48-53.
6. de Jong MF: 1983, In: Atrophic rhinitis in pigs, ed. Pedersen KB, Nielsen NC, pp. 136-146. CEC Agriculture EUR 8643 EN, EEC, Luxembourg. 7. Dominick MA, Remler RB: 1988, Turbinate osteoporosis in pigs following intranasal inoculation of purified Pasteurella toxin: histomorphometric and ultrastructural studies. Vet Pathol 25: 17-27. 8. Eamens GJ, Kirkland PD, Turner MJ, Gardner IA, et al.: 1988, Identification of toxigenic P. multocida in atrophic rhinitis of pigs by in vitro characterization. Aust Vet J 65:120-123. 9. Foged NT, Nielsen JP, Jorsal SE: 1989, Protection against progressive atrophic rhinitis by vaccination with Pasteurella multocida toxin purified by monoclonal antibodies. Vet Rec 125: 7-11. 10. Foged NT, Nielsen JP, Pedersen KB: 1988, Differentiation of toxigenic from nontoxigenic isolates of Pasteurella multocida by enzyme-linked immunosorbent assay. J Clin Microbiol 26: 1419-1420. 11. Foged NT, Pedersen KB, Elling F: 1987, Characterization and biological effects of Pasteurella multocida toxin. FEMS Microbiol Lett 43:45-51. 12. Gardner IA, Eamens GJ, Turner MJ, Homitzky CL: 1989, Toxigenic type D Pasteurella multocida in New South Wales pig herds-prevalence and factors associated with infection. Aust Vet J 66:318-321. 13. Gerstman BB, Cappucci OT: 1986, Evaluating the reliability of diagnostic tests. J Am Vet Med Assoc 188:248-251. 14. Giles CJ, Smith IM: 1983, Vaccination of pigs with Bordetella bronchiseptica. Vet Bull 53:327-338. 15. Goodwin RFW, Chanter N, Rutter JM: 1990, Detection and distribution of toxigenic Pasteurella multocida in pig herds with different degrees of atrophic rhinitis. Vet Rec 126:452-456. 16. Kamp EM, Kimman TG: 1988, Induction of nasal turbinate atrophy in germ-free pigs, using Pasteurella multocida as well as bacterium free crude and purified dermonecrotic toxin of P. multocida. Am J Vet Res 49: 1844-1849. 17. Pedersen KB: 1983, Cultural and serological diagnosis of atrophic rhinitis in pigs. In: Atrophic rhinitis in pigs, ed. Pedersen KB, Nielsen NC, pp. 22-31. CEC Agriculture EUR 8643 EN, EEC, Luxembourg. 18. Pedersen KB, Barfod K: 1981, The aetiological significance of Bordetella bronchiseptica and Pasteurella multocidia in atrophic rhinitis of swine. Nord Veterinaermed 33:513-522. 19. Rutter JM: 1987, Atrophic rhinitis in pigs. Pig News & Info 8: 385-387. 20. Rutter JM, Luther PD: 1984, Cell culture assay for toxigenic Pasteurella multocida from atrophic rhinitis in pigs. Vet Rec 114:393-396. 21. Samuels ML: 1989, Statistics for the life sciences, pp. 329-331. Dellen, San Francisco, CA.
Downloaded from vdi.sagepub.com by guest on March 29, 2015
