- Email: [email protected]
An interferon-gamma assay for diagnosis of Corynebacterium pseudotuberculosis infection in adult sheep from a research flock
Veterinary Microbiology 88 (2002) 287–297
An interferon-gamma assay for diagnosis of Corynebacterium pseudotuberculosis infection in adult sheep from a research flock J.F. Prescotta,*, P.I. Menziesb, Y.-T. Hwangb a
b
Department of Pathobiology, University of Guelph, Guelph, Ont., Canada N1G 2W1 Department of Population Medicine, University of Guelph, Guelph, Ont., Canada N1G 2W1 Received 18 December 2001; received in revised form 3 June 2002; accepted 3 June 2002
Abstract The optimal method of control of caseous lymphadenitis of sheep caused by Corynebacterium pseudotuberculosis is eradication of infection by identification and removal of infected carrier animals. Current serological approaches to identification of infected sheep are generally hampered by low sensitivity and specificity of available tests. The objective of this study was to develop a whole blood assay for detection of C. pseudotuberculosis-infected sheep, based on detection of IFN-g response to whole cell C. pseudotuberculosis antigens, and to determine the reliability of the assay. A commercially available bovine interferon-g assay enzyme-linked immunosorbent assay was used and the test optimised using experimentally infected sheep. The assay was also tested on known CLAnegative sheep. Setting a IFN-g optical density cut-off at 0.100 as positive under the conditions used, the test detected C. pseudotuberculosis experimentally infected sheep over a 450-day period with a reliability of 95.7%. It identified known non-infected sheep with a reliability of 95.5%. Repeated vaccination of three uninfected sheep with a commercially available bacterin-toxoid vaccine did not interfere with the assay. The IFN-g response of sheep whole blood to C. pseudotuberculosis antigens offers promise for use in a test-and-removal approach to eradication of CLA in sheep. # 2002 Elsevier Science B.V. All rights reserved. Keywords: Corynebacterium pseudotuberculosis; Sheep-bacteria; Diagnosis; Interferon-gamma
1. Introduction Corynebacterium pseudotuberculosis is a widely distributed Gram-positive bacterium that causes caseous lymphadenitis (CLA) in sheep and goats, as well as occasional * Corresponding author. Tel.: þ1-519-823-4716; fax: þ1-519-767-0809. E-mail address: [email protected] (J.F. Prescott).
0378-1135/02/$ – see front matter # 2002 Elsevier Science B.V. All rights reserved. PII: S 0 3 7 8 - 1 1 3 5 ( 0 2 ) 0 0 1 2 1 - 9
288
J.F. Prescott et al. / Veterinary Microbiology 88 (2002) 287–297
suppurative infection in other species (Pe´ pin et al., 1999; Williamson, 2001). Current methods of diagnosis and control of CLA are not fully satisfactory, although vaccination may significantly reduce the spread of infection (Paton et al., 1995). The optimal method of control would be eradication of infection by identification and removal of infected carrier animals. A large number of serodiagnostic tests have been described (Brown et al., 1987; Menzies and Muckle, 1989; Sutherland et al., 1987; ter Laak et al., 1992; Menzies et al., 1994; Sting et al., 1998), but their sensitivity and specificity are inadequate to provide a reliable control approach based on identification and removal of infected individuals. There may be exceptions to this statement (Schreuder et al., 1994; Dercksen et al., 2001). Because C. pseudotuberculosis is a facultative intracellular pathogen, cell-mediated immunity is an important component of the protective immune response (Lan et al., 1998). An alternative approach to immunodiagnosis of CLA is therefore to assess the cellmediated rather than humoral immune response to infection. If such a test was available, it might be possible to use it on its own, or together with humoral response, in an eradication program. A monoclonal antibody assay to bovine interferon-gamma (IFN-g), using whole blood from test animals incubated with Mycobacterium bovis purified-protein derivative (PPD), has been used extensively and successfully in the eradication of bovine tuberculosis using a test-and-slaughter approach (Rothel et al., 1990; Wood et al., 1991, 1992; Whipple et al., 1995). The assay is more sensitive than the intradermal tuberculin test for the diagnosis of bovine tuberculosis (Wood et al., 1991). The cross-reactivity of the monoclonal antibody to bovine IFN-g with sheep and goat IFN-g has been demonstrated previously (Rothel et al., 1990) and also used previously to detect an IFN-g response of infected sheep to C. pseudotuberculosis antigen in purified peripheral blood lymphocytes (Pe´ pin et al., 1997). The assay has been applied to the detection of bovine tuberculosis in goats but the study was too limited to determine the sensitivity and specificity of the assay (Cousins et al., 1993). The purpose of the work described here was to develop a whole blood assay for detection of C. pseudotuberculosis-infected sheep, based on detection of IFN-g response to whole cell antigens, and to determine the values of the assay in comparison with the infection status of the sheep.
2. Materials and methods 2.1. Source of sheep blood 2.1.1. Infected sheep Three young adult sheep (numbered 2, 16, 20) from a research flock (Ponsonby Sheep Research Station) free of C. pseudotuberculosis, were used. Two millilitres of 2 107 /ml of C. pseudotuberculosis (strain 662/82) was injected just above the coronet of the right hind limb (Pe´ pin et al., 1997), and sheep were maintained in an isolation facility. Another three uninfected control sheep were injected in a similar site with 2 ml phosphate buffer saline, pH 7.2 (PBS) and maintained separately from the first group in an isolation facility.
J.F. Prescott et al. / Veterinary Microbiology 88 (2002) 287–297
289
2.1.2. Vaccinated, uninfected sheep Three additional uninfected control sheep maintained at a research station were vaccinated with 1 ml of a commercially available C. pseudotuberculosis bacterin-toxoid (Glanvac 6: Vetrepharm Canada, London, Ontario), and were re-vaccinated after 31, 161 and 331 days, and heparinised blood obtained for periodic testing in the IFN-g assay. 2.1.3. Uninfected sheep Heparinised blood was obtained from 73 sheep from a flock of carefully monitored research sheep known to be free of infection with C. pseudotuberculosis. Fourteen of these sheep were sampled twice, four were sampled three times, and one was sampled four times, for a total of 99 samples. The flock had been maintained in an isolated state at a research station for 13 years. In addition, three uninfected non-vaccinated sheep from the same flock were maintained in the isolation facility over a period of 338 days, and were bled periodically for the IFN-g assay. 2.2. IFN-g enzyme-linked immunosorbent assay (ELISA) and standardisation 2.2.1. Production of antigen C. pseudotuberculosis strain 662/82 was grown in 500 ml trypticase soy broth (Difco, Becton Dickinson, Sparts, MD) for 72 h at 37 8C and shaking at 50 rpm. The culture was harvested by centrifugation at 10 000 g for 30 min, and the pellet resuspended in 30 ml of PBS, and lysed by a French Press. The supernatant was lyophilised and redissolved to give a protein concentration of 500 mg/ml as assessed by the Lowry method. The antigen was frozen in small aliquots at 70 8C until used. 2.2.2. Whole blood culture: standardisation of the IFN-g assay One millilitre sample of heparinised blood was placed within 2 h of collection in each of 24 wells of tissue culture plates, and 50 mg antigen added to each of three wells containing blood from one animal and incubated at 37 8C for 24 h in a 5% CO2 in air, humidified, atmosphere. Culture plates were centrifuged for 10 min at 1400 rpm and the separated plasma stored at 20 8C for quantitation of IFN-g by ELISA (Bovigam, bovine gamma interferon test EIA, CSL Veterinary Ltd., Parkville, Victoria, Australia). This was used according to the manufacturer’s recommendations. The optical density of plasma was read at 450 nm and results expressed as an optical density reading. A positive IFN-g control optical density was run in triplicate on each plate. In testing individual sheep’s blood, each blood was run in triplicate in the presence of C. pseudotuberculosis antigen as well as in triplicate in the absence of antigen (negative control). Normalisation of the IFN-g optical density results for individual sheep to the positive IFN-g control and adjustment for the negative control optical density is described in the results. 2.2.3. Statistical methods The data were analysed using SAS version 6.12 software. Data were transformed to log values and confidence intervals calculated from the log mean and standard deviation and then transformed to the antilog. Zero values were excluded. Reliability statistics for individual sheep values were calculated using proc varcomp.
290
J.F. Prescott et al. / Veterinary Microbiology 88 (2002) 287–297
3. Results 3.1. Experimental infection After injection of C. pseudotuberculosis, the three sheep each developed a painful swelling above the coronary band at the site of infection, and showed associated lameness. The inflammation lasted 10 days. In two sheep a local subcutaneous abscess developed which ruptured to release thickly purulent material. A small hard nodule persisted at the sites of infection for 2–3 months. On post-mortem examination 454 days after infection, sheep 2 had six C. pseudotuberculosis-infected pea-sized abscesses in the lungs. In addition, the right leg had two small abscesses as well as caseous pus at the injection site, from which C. pseudotuberculosis was isolated in large numbers. Sheep 20 had a small abscess in the right inguinal lymph node from which C. pseudotuberculosis was also isolated in large numbers. By contrast, sheep 16 had no evidence of infection at the site of injection or elsewhere in the body, and no C. pseudotuberculosis were recovered from the injection site. 3.2. Optimisation and standardisation of the IFN-g assay The optical density results of the whole blood IFN-g to a range of antigen concentrations 3 weeks after infection are shown in Table 1. A concentration of 50 mg/ml of antigen was chosen as the quantity to be used in all subsequent whole blood IFN-g assays. Because of variation between the positive IFN-g control values tested on different days, we made the following correction to individual results for all sheep. Results were only accepted for a plate if the positive IFN-g control optical density was between 1.0 and 1.3; results for individual sheep blood were then normalised by dividing the mean optical density of the triplicate wells with antigen for that sheep by the positive control optical density value. For example, if the actual optical density value for the blood of an individual sheep with 50 mg/ml was 0.5 and the positive IFN-g control value was 1.2, the normalised result for that sheep was 0.5 divided by 1:2 ¼ 0:42. A further correction was to subtract the normalised mean optical density of the three negative control wells for that sheep’s blood. This was then the final ‘‘normalised, adjusted’’ optical density value recorded for that sheep. Optical densities 0.000 were recorded as 0. The results of day-to-day variation in IFN-g responses for individual infected sheep, 72 days after infection, are shown in Table 2.
Table 1 Bovine IFN-g ELISA optical density response of sheep whole blood cultures to different concentrations of C. pseudotuberculosis antigens Sheep
0 mg/ml
10 mg/ml
20 mg/ml
50 mg/ml
100 mg/ml
2 16 20
0 0 0
0.34 0.164 0.176
0.363 0.217 0.119
0.406 0.207 0.126
0.324 0.168 0.126
J.F. Prescott et al. / Veterinary Microbiology 88 (2002) 287–297
291
Table 2 Day-to-day variation in IFN-g ELISA optical density response to 50 mg/ml of C. pseudotuberculosis antigen in infected sheep whole blood cultures, normalised adjusted values Sheep
Day 1
Day 2
Day 3
Day 4
Day 5
2 16 20
0.202 0.244 0.407
0.291 0.299 0.436
0.303 0.232 0.541
0.102 0.124 0.319
0.186 0.171 0.416
3.3. IFN-g response in experimentally infected sheep The three experimentally infected sheep were followed for 454 days following infection. Results of bovine IFN-g optical densities are given in Fig. 1. 3.4. Effect of vaccination of sheep on IFN-g response There was no effect of vaccination of three sheep with a commercially available C. pseudotuberculosis bacterin-toxoid on the IFN-g response of whole blood cultures (Fig. 2). 3.5. Assessment of reliability of IFN-g assay Three uninfected, non-vaccinated control sheep were tested over time (Fig. 3). In addition, uninfected sheep maintained at a research station were bled on a single occasion for testing (Fig. 4).
Fig. 1. Bovine IFN-g ELISA optical density response to 50 mg of C. pseudotuberculosis antigen in whole blood cultures of experimentally infected sheep over time.
292
J.F. Prescott et al. / Veterinary Microbiology 88 (2002) 287–297
Fig. 2. Bovine IFN-g ELISA optical density response to 50 mg of C. pseudotuberculosis antigen in whole blood cultures from vaccinated sheep over time.
Fig. 3. Bovine IFN-g ELISA optical density response to 50 mg of C. pseudotuberculosis antigen in whole blood cultures from uninfected sheep over time.
J.F. Prescott et al. / Veterinary Microbiology 88 (2002) 287–297
293
Fig. 4. IFN-g ELISA optical density response to 50 mg of C. pseudotuberculosis antigen in whole blood cultures from individual uninfected sheep at a research farm, first sampling (n ¼ 73).
The mean value for the uninfected sheep (vaccinated and control) was 0.0216 (CI ¼ 0:0168; 0.0278) and the mean value for the infected sheep was 0.2343 (CI ¼ 0:1902; 0.3039). Based on these results, an optical density cut-off of 0.100 was selected as indicating a positive test. Three of 70 samples obtained from the experimentally infected sheep were classified as negative (Fig. 1), giving a reliability of the assay in detecting infected animals of 95.5%. Of 56 results from experimentally vaccinated uninfected sheep, all but one sample were classified as negative for a reliability of 98.2% (Fig. 2). Of the three non-vaccinated, non-infected sheep bled repeatedly over time, one of 46 samples was positive, for a reliability of 97.8% (Fig. 3). Of 73 additional uninfected sheep that were tested, seven samples were positive (of 99 samples) for a reliability of 92.9%. If only the first test result per animal is used, then the test reliability for non-exposed animals would be five positive tests from the 73 animals tested, or a test reliability of 93.2% (Fig. 4). Overall, nine of 201 tests gave falsepositive results, giving an estimated reliability of 95.5%. When the cut-off point is changed, the proportion of sheep being classified correctly varies (Table 4). The cut-off point of 0.1 appears to optimise the proportion of sheep correctly classified in both categories. 3.6. Repeatability of test results from individual animals from a known negative flock Nineteen of the sheep from a known negative flock were sampled more than once (Tables 3 and 4). Sixteen sheep remained negative, one positive sheep remained positive, one negative sheep became positive, and one weakly positive sheep became negative. The sheep J 119, which gave two strongly positive reactions, was euthanized; a detailed post-mortem examination failed to reveal any evidence of CLA.
294
J.F. Prescott et al. / Veterinary Microbiology 88 (2002) 287–297
Table 3 Comparison of results for bovine IFN-g ELISA optical density response to 50 mg/ml of C. pseudotuberculosis antigen in whole blood cultures from individual animals that were sampled at repeated intervals from an uninfected research flock ID
First sample
Second sample
Third sample
C 32 D 142 E 193 E 256 E3 H 131 H 134 H 88 J 114 J 119 J 138 J 142 K 128 K 151 K 42 K 47 K 54 K 64 K 70
0 0.043 0 0 0 0.004 0.027 0 0 0.256 0 0 0.006 0.044 0 0.019 0.109 0 0.083
0 0.034 0 0.041 0 0.001 0.051 0.174 0 0.322 0 0.005 0 0.003 0 0.022 0 0 0.025
0.009 0.007 0
0.08 0
3.7. Variation in the test results within individual sheep The Rho value for the three experimentally infected sheep is 0.0846 and for the six noninfected sheep (vaccinated and control) is 0.0579. This indicates that most of the variation in titre is within the individual sheep rather than between. Table 4 Reliability of test classification at selected IFN-g ELISA optical density cut-off points IFN-g optical density cut point
Proportion of negative sheep classified as negative
Proportion of infected sheep classified as positive
0.25 0.2 0.15 0.14 0.13 0.12 0.11 0.1a 0.09 0.08 0.07 0.06
0.985 0.975 0.97 0.965 0.965 0.965 0.96 0.955 0.945 0.94 0.925 0.915
0.471 0.6 0.729 0.743 0.8 0.9 0.943 0.957 0.971 0.986 1.0 1.0
a
0.1, selected cut-off point.
J.F. Prescott et al. / Veterinary Microbiology 88 (2002) 287–297
295
4. Discussion This is the first study to use a commercially available monoclonal antibody to bovine IFN-g in a whole blood assay for detection of cell-mediated immunity in C. pseudotuberculosis-infected sheep. The study suggests that the assay, based on detection of IFN-g response to whole cell antigens, has value in detection of infection in sheep. The study showed that immunisation with a vaccine commonly used in the control of CLA in sheep did not interfere with the assay. Using a positive ‘‘cut-off’’ optical density value of 0.1, the whole blood assay correctly classified infected animals as positive and non-exposed sheep as negative with an acceptable reliability for use as a diagnostic test in a test-and-control program. These figures are the equivalent of 95.7% sensitivity and 95.5% specificity, but because the samples were taken from repeatedly tested individuals (apart from one large group of uninfected sheep) we assessed reliability rather than sensitivity and specificity. Further work is required to define the sensitivity of the assay using blood from sheep flocks known to be naturally infected with C. pseudotuberculosis, and to improve definition of the specificity of the assay using blood from sheep from flocks known to be free of infection. The end point of the assay was normalised to the positive control, to reduce some of the variation between readings in different tests. In earlier studies with the assay for the detection of bovine tuberculosis, the optical density was not normalised but rather the test was ‘‘accepted’’ if the positive optical density control was above 0.7 (Wood et al., 1991). Wood et al. (1991) regarded an animal as positive if the optical density value in the antigen containing wells was 0.05 above the nil antigen negative control. Whether the assay is less sensitive in the detection of sheep than in cattle was not investigated, but the optical density values recorded for sheep never reached some of those recorded in blood from some cattle with bovine tuberculosis (Rothel et al., 1990; Doherty et al., 1995). This difference may relate to the greater severity of infection in some cattle naturally infected with M. bovis. The most heavily infected sheep, sheep 2, gave the highest IFN-g response, and the sheep (16) shown not to be infected at time of euthanasia, gave the lowest response (Fig. 1). These results suggest that there is a relation between severity of infection of sheep and IFN-g response, but must be confirmed with a larger number of animals; however, severely infected sheep could become anergic. Factors known to affect the activity of the test include the time within which the blood sample is placed in culture, the quantity of the antigen used, and the use of heparinised blood (Rothel et al., 1992). Apart from assessing the effect of quantity of C. pseudotuberculosis antigen used, we did not investigate other factors affecting the variability of the assay results. Samples were incubated within 2 h of collection, well within the time known to be suitable for optimal performance of the assay (Rothel et al., 1992). The studies with experimentally infected sheep (Fig. 1) and confirmed by the low Rho values, showed that there is fluctuation in individual animal responses over both the shortand the long-term. All sheep varied in the same manner, suggesting that other factors may influence day-to-day results, such a stage of disease, possible bacterial shedding and health of the animal. However, change in disease classification from positive to negative was not a common event. It may be that when applying this test in a test-and-remove eradication program within a flock, negative sheep should be tested at repeated intervals, perhaps every 2 months to enhance correct classification. In the eradication of bovine tuberculosis, the
296
J.F. Prescott et al. / Veterinary Microbiology 88 (2002) 287–297
advantage of the whole blood IFN-g test over the intradermal tuberculin test is that it is not followed by the ‘‘anergy’’ observed with skin testing (Wood et al., 1992; Doherty et al., 1995), which means that a skin test cannot be repeated within 6 weeks. The fluctuation in negative animals also suggests that there may be other factors that cause a significant rise in response in some sheep (Table 3). Again Rho values were low indicating that individual sheep results varied but all negative sheep varied the same amount. However, not all sheep varied in the same way at the same sampling time. For example, between the first and second sampling, six values decreased, seven remained unchanged, and six values increased. It is possible that infection of sheep with mycolic acid-containing bacteria related to C. pseudotuberculosis, such as environmental Mycobacterium or Nocardia species, might give rise to cross-reacting antibodies. The sheep flock is free of M. avium subspecies paratuberculosis infection. The test might be improved by the use of defined antigens rather than supernatant obtained after disruption of whole cells. A marked rise in optical density values occurred in some infected (Fig. 1) and an uninfected sheep (Fig. 2); we were unable to determine the reason for this. Could this test be useful in the eradication of CLA in sheep? The test will have to be assessed in the field before an answer can be given. Although humoral tests have apparently been successful in some instances in eradication of infection (Schreuder et al., 1994; Dercksen et al., 2001), others have found that their sensitivity and specificity have been relatively low so that they have been difficult to apply to detection of individual infected animals (Pe´ pin et al., 1999). One potential advantage of the IFN-g assay over humoral assays is that it could be used in vaccinated animals, which had developed titres to phospholipase D or other antigens in response to immunisation (Eggleton et al., 1991). Since vaccination is a useful approach to control CLA (Eggleton et al., 1991), it could be combined with an eradication procedure based on the IFN-g assay which could in the long-term be the method of choice for control of the infection.
Acknowledgements We thank the Ontario Ministry of Agriculture and Food (OMAF) for supporting this work. We also thank Pam Jordan from the Arkell Research Station, and David Bridle and Tony Cengija from the OMAF Isolation Unit for their help with maintaining and bleeding sheep.
References Brown, C.C., Olander, H.J., Alves, S.F., 1987. Synergistic hemolysis-inhibition titers associated with caseous lymphadenitis in a slaughterhouse survey of goats and sheep in northeastern Brazil. Can. J. Vet. Res. 51, 46–49. Cousins, D.V., Francis, B.R., Casey, R., Mayberry, C., 1993. Mycobacterium bovis infection in a goat. Aust. Vet. J. 70, 262–263. Dercksen, D.P., Brinkhof, J.M., Dekker-Nooren, T., Maanen, K., Bode, C.F., Baird, G., Kamp, E.M., 2001. A comparison of four serological tests for the diagnosis of caseous lymphadenitis in sheep and goats. Vet. Microbiol. 75, 167–175.
J.F. Prescott et al. / Veterinary Microbiology 88 (2002) 287–297
297
Doherty, M.L., Monaghan, M.L., Bassett, H.F., Quinn, P.J., 1995. Effect of a recent injection of purified protein derivative on diagnostic tests for tuberculosis in cattle infected with Mycobacterium bovis. Res. Vet. Sci. 58, 217–221. Eggleton, D.G., Haynes, J.A., Middleton, H.D., Cox, J.C., 1991. Immunization against ovine caseous lymphadenitis: correlation between Corynebacterium pseudotuberculosis toxoid content and protective efficacy in combined clostridial–corynebacterial vaccines. Aust. Vet. J. 68, 322–325. Lan, D.T.B., Taniguchi, S., Makino, S., Shirahata, T., Nakane, A., 1998. Role of endogenous tumour necrosis factor alpha: an gamma interferon in resistance to Corynebacterium pseudotuberculosis in mice. Microbiol. Immunol. 42, 863–870. Menzies, P.I., Muckle, C.A., 1989. The use of a microagglutination assay for the detection of antibodies to Corynebacterium pseudotuberculosis in naturally infected sheep and goat flocks. Can. J. Vet. Res. 53, 313–318. Menzies, P.I., Muckle, C.A., Hwang, Y.T., Songer, J.G., 1994. Evaluation of an enzyme-linked immunosorbent assay using an Escherichia coli recombinant phospholipase D antigen for the diagnosis of Corynebacterium pseudotuberculosis infection. Small Rum. Res. 13, 193–198. Paton, M.W., Sutherland, S.S., Rose, I.R., Hart, R.A., Mercy, A.R., Ellis, T.M., 1995. The spread of Corynebacterium pseudotuberculosis infection to unvaccinated and vaccinated sheep. Aust. Vet. J. 72, 266–269. Pe´ pin, M., Seow, H.F., Corner, L., Rothel, J.S., Hodgson, A.L.M., Wood, P.R., 1997. Cytokine expression in sheep following experimental infection with various strains of Corynebacterium pseudotuberculosis differing in virulence. Vet. Res. 28, 149–163. Pe´ pin, M., Sanchis, R., Paton, M., 1999. La lymphade´ nite case´ euse des ovins et des caprins. Point Vet. 30, 33–40. Rothel, J.S., Jones, S.L., Corner, L.A., Cox, J.C., Wood, P.R., 1990. A sandwich enzyme immunoassay for bovine interferon-g and its use for the detection of tuberculosis in cattle. Aust. Vet. J. 67, 134–137. Rothel, J.S., Jones, S.L., Corner, L.A., Cox, J.C., Wood, P.R., 1992. The gamma-interferon assay for diagnosis of bovine tuberculosis in cattle: conditions affecting the production of gamma-interferon in whole blood culture. Aust. Vet. J. 69, 1–4. Schreuder, B.E.C., ter Laak, E.A., Dercksen, D.P., 1994. Eradication of caseous lymphadenitis in sheep with the help of a newly developed ELISA technique. Vet. Rec. 135, 174–176. Sting, R., Steng, G., Spengler, D., 1998. Serological studies on Corynebacterium pseudotuberculosis infections in goats using enzyme-linked immunosorbent assay. Zentralbl. Veterinarmed. B 45, 209–216. Sutherland, S.S., Ellis, T.M., Mercy, A.R., Paton, M., Middleton, H., 1987. Evaluation of an enzyme-linked immunosorbent assay for the detection of Corynebacterium pseudotuberculosis infection in sheep. Aust. Vet. J. 64, 263–266. ter Laak, E.A., Bosch, J., Bijl, G.C., Schreuder, B.E.C., 1992. Double-antibody sandwich enzyme-linked immunosorbent assay and immunoblot analysis used for control of caseous lymphadenitis in goats and sheep. Am. J. Vet. Res. 53, 1125–1132. Whipple, D.L., Bolin, C.A., Davis, A.J., Jarnigan, J.L., Johnson, D.C., Nabors, R.S., Payeur, J.B., Saari, D.A., Wilson, A.J., Wolf, M.M., 1995. Comparison of the sensitivity of the caudal fold skin test and a commercial g-interferon assay for diagnosis of bovine tuberculosis. Am. J. Vet. Res. 56, 415–419. Williamson, L.H., 2001. Caseous lymphadenitis in small ruminants. Vet. Clin. North Am. Food Anim. Pract. 17, 359–371. Wood, P.R., Corner, L.A., Rothel, J.S., Baldock, C., Jones, S.L., Cousins, D.R., McCormick, B.S., Francis, B.R., Creeper, J., Tweddle, N.E., 1991. Field comparison of the interferon-gamma assay and the intradermal tuberculin test for the diagnosis of bovine tuberculosis. Aust. Vet. J. 68, 286–290. Wood, P.R., Corner, L.A., Rothel, J.S., Ripper, J.L., Fifis, T., McCormick, B.S., Francis, B., Melville, L., Small, K., De Witte, K., Tolson, J., Ryan, T.J., de Lisle, G.W., Cox, J.C., Jones, S.L., 1992. A field evaluation of serological and cellular diagnostic tests for bovine tuberculosis. Vet. Microbiol. 31, 71–79.
An interferon-gamma assay for diagnosis of Corynebacterium pseudotuberculosis infection in adult sheep from a research flock J.F. Prescotta,*, P.I. Menziesb, Y.-T. Hwangb a
b
Department of Pathobiology, University of Guelph, Guelph, Ont., Canada N1G 2W1 Department of Population Medicine, University of Guelph, Guelph, Ont., Canada N1G 2W1 Received 18 December 2001; received in revised form 3 June 2002; accepted 3 June 2002
Abstract The optimal method of control of caseous lymphadenitis of sheep caused by Corynebacterium pseudotuberculosis is eradication of infection by identification and removal of infected carrier animals. Current serological approaches to identification of infected sheep are generally hampered by low sensitivity and specificity of available tests. The objective of this study was to develop a whole blood assay for detection of C. pseudotuberculosis-infected sheep, based on detection of IFN-g response to whole cell C. pseudotuberculosis antigens, and to determine the reliability of the assay. A commercially available bovine interferon-g assay enzyme-linked immunosorbent assay was used and the test optimised using experimentally infected sheep. The assay was also tested on known CLAnegative sheep. Setting a IFN-g optical density cut-off at 0.100 as positive under the conditions used, the test detected C. pseudotuberculosis experimentally infected sheep over a 450-day period with a reliability of 95.7%. It identified known non-infected sheep with a reliability of 95.5%. Repeated vaccination of three uninfected sheep with a commercially available bacterin-toxoid vaccine did not interfere with the assay. The IFN-g response of sheep whole blood to C. pseudotuberculosis antigens offers promise for use in a test-and-removal approach to eradication of CLA in sheep. # 2002 Elsevier Science B.V. All rights reserved. Keywords: Corynebacterium pseudotuberculosis; Sheep-bacteria; Diagnosis; Interferon-gamma
1. Introduction Corynebacterium pseudotuberculosis is a widely distributed Gram-positive bacterium that causes caseous lymphadenitis (CLA) in sheep and goats, as well as occasional * Corresponding author. Tel.: þ1-519-823-4716; fax: þ1-519-767-0809. E-mail address: [email protected] (J.F. Prescott).
0378-1135/02/$ – see front matter # 2002 Elsevier Science B.V. All rights reserved. PII: S 0 3 7 8 - 1 1 3 5 ( 0 2 ) 0 0 1 2 1 - 9
288
J.F. Prescott et al. / Veterinary Microbiology 88 (2002) 287–297
suppurative infection in other species (Pe´ pin et al., 1999; Williamson, 2001). Current methods of diagnosis and control of CLA are not fully satisfactory, although vaccination may significantly reduce the spread of infection (Paton et al., 1995). The optimal method of control would be eradication of infection by identification and removal of infected carrier animals. A large number of serodiagnostic tests have been described (Brown et al., 1987; Menzies and Muckle, 1989; Sutherland et al., 1987; ter Laak et al., 1992; Menzies et al., 1994; Sting et al., 1998), but their sensitivity and specificity are inadequate to provide a reliable control approach based on identification and removal of infected individuals. There may be exceptions to this statement (Schreuder et al., 1994; Dercksen et al., 2001). Because C. pseudotuberculosis is a facultative intracellular pathogen, cell-mediated immunity is an important component of the protective immune response (Lan et al., 1998). An alternative approach to immunodiagnosis of CLA is therefore to assess the cellmediated rather than humoral immune response to infection. If such a test was available, it might be possible to use it on its own, or together with humoral response, in an eradication program. A monoclonal antibody assay to bovine interferon-gamma (IFN-g), using whole blood from test animals incubated with Mycobacterium bovis purified-protein derivative (PPD), has been used extensively and successfully in the eradication of bovine tuberculosis using a test-and-slaughter approach (Rothel et al., 1990; Wood et al., 1991, 1992; Whipple et al., 1995). The assay is more sensitive than the intradermal tuberculin test for the diagnosis of bovine tuberculosis (Wood et al., 1991). The cross-reactivity of the monoclonal antibody to bovine IFN-g with sheep and goat IFN-g has been demonstrated previously (Rothel et al., 1990) and also used previously to detect an IFN-g response of infected sheep to C. pseudotuberculosis antigen in purified peripheral blood lymphocytes (Pe´ pin et al., 1997). The assay has been applied to the detection of bovine tuberculosis in goats but the study was too limited to determine the sensitivity and specificity of the assay (Cousins et al., 1993). The purpose of the work described here was to develop a whole blood assay for detection of C. pseudotuberculosis-infected sheep, based on detection of IFN-g response to whole cell antigens, and to determine the values of the assay in comparison with the infection status of the sheep.
2. Materials and methods 2.1. Source of sheep blood 2.1.1. Infected sheep Three young adult sheep (numbered 2, 16, 20) from a research flock (Ponsonby Sheep Research Station) free of C. pseudotuberculosis, were used. Two millilitres of 2 107 /ml of C. pseudotuberculosis (strain 662/82) was injected just above the coronet of the right hind limb (Pe´ pin et al., 1997), and sheep were maintained in an isolation facility. Another three uninfected control sheep were injected in a similar site with 2 ml phosphate buffer saline, pH 7.2 (PBS) and maintained separately from the first group in an isolation facility.
J.F. Prescott et al. / Veterinary Microbiology 88 (2002) 287–297
289
2.1.2. Vaccinated, uninfected sheep Three additional uninfected control sheep maintained at a research station were vaccinated with 1 ml of a commercially available C. pseudotuberculosis bacterin-toxoid (Glanvac 6: Vetrepharm Canada, London, Ontario), and were re-vaccinated after 31, 161 and 331 days, and heparinised blood obtained for periodic testing in the IFN-g assay. 2.1.3. Uninfected sheep Heparinised blood was obtained from 73 sheep from a flock of carefully monitored research sheep known to be free of infection with C. pseudotuberculosis. Fourteen of these sheep were sampled twice, four were sampled three times, and one was sampled four times, for a total of 99 samples. The flock had been maintained in an isolated state at a research station for 13 years. In addition, three uninfected non-vaccinated sheep from the same flock were maintained in the isolation facility over a period of 338 days, and were bled periodically for the IFN-g assay. 2.2. IFN-g enzyme-linked immunosorbent assay (ELISA) and standardisation 2.2.1. Production of antigen C. pseudotuberculosis strain 662/82 was grown in 500 ml trypticase soy broth (Difco, Becton Dickinson, Sparts, MD) for 72 h at 37 8C and shaking at 50 rpm. The culture was harvested by centrifugation at 10 000 g for 30 min, and the pellet resuspended in 30 ml of PBS, and lysed by a French Press. The supernatant was lyophilised and redissolved to give a protein concentration of 500 mg/ml as assessed by the Lowry method. The antigen was frozen in small aliquots at 70 8C until used. 2.2.2. Whole blood culture: standardisation of the IFN-g assay One millilitre sample of heparinised blood was placed within 2 h of collection in each of 24 wells of tissue culture plates, and 50 mg antigen added to each of three wells containing blood from one animal and incubated at 37 8C for 24 h in a 5% CO2 in air, humidified, atmosphere. Culture plates were centrifuged for 10 min at 1400 rpm and the separated plasma stored at 20 8C for quantitation of IFN-g by ELISA (Bovigam, bovine gamma interferon test EIA, CSL Veterinary Ltd., Parkville, Victoria, Australia). This was used according to the manufacturer’s recommendations. The optical density of plasma was read at 450 nm and results expressed as an optical density reading. A positive IFN-g control optical density was run in triplicate on each plate. In testing individual sheep’s blood, each blood was run in triplicate in the presence of C. pseudotuberculosis antigen as well as in triplicate in the absence of antigen (negative control). Normalisation of the IFN-g optical density results for individual sheep to the positive IFN-g control and adjustment for the negative control optical density is described in the results. 2.2.3. Statistical methods The data were analysed using SAS version 6.12 software. Data were transformed to log values and confidence intervals calculated from the log mean and standard deviation and then transformed to the antilog. Zero values were excluded. Reliability statistics for individual sheep values were calculated using proc varcomp.
290
J.F. Prescott et al. / Veterinary Microbiology 88 (2002) 287–297
3. Results 3.1. Experimental infection After injection of C. pseudotuberculosis, the three sheep each developed a painful swelling above the coronary band at the site of infection, and showed associated lameness. The inflammation lasted 10 days. In two sheep a local subcutaneous abscess developed which ruptured to release thickly purulent material. A small hard nodule persisted at the sites of infection for 2–3 months. On post-mortem examination 454 days after infection, sheep 2 had six C. pseudotuberculosis-infected pea-sized abscesses in the lungs. In addition, the right leg had two small abscesses as well as caseous pus at the injection site, from which C. pseudotuberculosis was isolated in large numbers. Sheep 20 had a small abscess in the right inguinal lymph node from which C. pseudotuberculosis was also isolated in large numbers. By contrast, sheep 16 had no evidence of infection at the site of injection or elsewhere in the body, and no C. pseudotuberculosis were recovered from the injection site. 3.2. Optimisation and standardisation of the IFN-g assay The optical density results of the whole blood IFN-g to a range of antigen concentrations 3 weeks after infection are shown in Table 1. A concentration of 50 mg/ml of antigen was chosen as the quantity to be used in all subsequent whole blood IFN-g assays. Because of variation between the positive IFN-g control values tested on different days, we made the following correction to individual results for all sheep. Results were only accepted for a plate if the positive IFN-g control optical density was between 1.0 and 1.3; results for individual sheep blood were then normalised by dividing the mean optical density of the triplicate wells with antigen for that sheep by the positive control optical density value. For example, if the actual optical density value for the blood of an individual sheep with 50 mg/ml was 0.5 and the positive IFN-g control value was 1.2, the normalised result for that sheep was 0.5 divided by 1:2 ¼ 0:42. A further correction was to subtract the normalised mean optical density of the three negative control wells for that sheep’s blood. This was then the final ‘‘normalised, adjusted’’ optical density value recorded for that sheep. Optical densities 0.000 were recorded as 0. The results of day-to-day variation in IFN-g responses for individual infected sheep, 72 days after infection, are shown in Table 2.
Table 1 Bovine IFN-g ELISA optical density response of sheep whole blood cultures to different concentrations of C. pseudotuberculosis antigens Sheep
0 mg/ml
10 mg/ml
20 mg/ml
50 mg/ml
100 mg/ml
2 16 20
0 0 0
0.34 0.164 0.176
0.363 0.217 0.119
0.406 0.207 0.126
0.324 0.168 0.126
J.F. Prescott et al. / Veterinary Microbiology 88 (2002) 287–297
291
Table 2 Day-to-day variation in IFN-g ELISA optical density response to 50 mg/ml of C. pseudotuberculosis antigen in infected sheep whole blood cultures, normalised adjusted values Sheep
Day 1
Day 2
Day 3
Day 4
Day 5
2 16 20
0.202 0.244 0.407
0.291 0.299 0.436
0.303 0.232 0.541
0.102 0.124 0.319
0.186 0.171 0.416
3.3. IFN-g response in experimentally infected sheep The three experimentally infected sheep were followed for 454 days following infection. Results of bovine IFN-g optical densities are given in Fig. 1. 3.4. Effect of vaccination of sheep on IFN-g response There was no effect of vaccination of three sheep with a commercially available C. pseudotuberculosis bacterin-toxoid on the IFN-g response of whole blood cultures (Fig. 2). 3.5. Assessment of reliability of IFN-g assay Three uninfected, non-vaccinated control sheep were tested over time (Fig. 3). In addition, uninfected sheep maintained at a research station were bled on a single occasion for testing (Fig. 4).
Fig. 1. Bovine IFN-g ELISA optical density response to 50 mg of C. pseudotuberculosis antigen in whole blood cultures of experimentally infected sheep over time.
292
J.F. Prescott et al. / Veterinary Microbiology 88 (2002) 287–297
Fig. 2. Bovine IFN-g ELISA optical density response to 50 mg of C. pseudotuberculosis antigen in whole blood cultures from vaccinated sheep over time.
Fig. 3. Bovine IFN-g ELISA optical density response to 50 mg of C. pseudotuberculosis antigen in whole blood cultures from uninfected sheep over time.
J.F. Prescott et al. / Veterinary Microbiology 88 (2002) 287–297
293
Fig. 4. IFN-g ELISA optical density response to 50 mg of C. pseudotuberculosis antigen in whole blood cultures from individual uninfected sheep at a research farm, first sampling (n ¼ 73).
The mean value for the uninfected sheep (vaccinated and control) was 0.0216 (CI ¼ 0:0168; 0.0278) and the mean value for the infected sheep was 0.2343 (CI ¼ 0:1902; 0.3039). Based on these results, an optical density cut-off of 0.100 was selected as indicating a positive test. Three of 70 samples obtained from the experimentally infected sheep were classified as negative (Fig. 1), giving a reliability of the assay in detecting infected animals of 95.5%. Of 56 results from experimentally vaccinated uninfected sheep, all but one sample were classified as negative for a reliability of 98.2% (Fig. 2). Of the three non-vaccinated, non-infected sheep bled repeatedly over time, one of 46 samples was positive, for a reliability of 97.8% (Fig. 3). Of 73 additional uninfected sheep that were tested, seven samples were positive (of 99 samples) for a reliability of 92.9%. If only the first test result per animal is used, then the test reliability for non-exposed animals would be five positive tests from the 73 animals tested, or a test reliability of 93.2% (Fig. 4). Overall, nine of 201 tests gave falsepositive results, giving an estimated reliability of 95.5%. When the cut-off point is changed, the proportion of sheep being classified correctly varies (Table 4). The cut-off point of 0.1 appears to optimise the proportion of sheep correctly classified in both categories. 3.6. Repeatability of test results from individual animals from a known negative flock Nineteen of the sheep from a known negative flock were sampled more than once (Tables 3 and 4). Sixteen sheep remained negative, one positive sheep remained positive, one negative sheep became positive, and one weakly positive sheep became negative. The sheep J 119, which gave two strongly positive reactions, was euthanized; a detailed post-mortem examination failed to reveal any evidence of CLA.
294
J.F. Prescott et al. / Veterinary Microbiology 88 (2002) 287–297
Table 3 Comparison of results for bovine IFN-g ELISA optical density response to 50 mg/ml of C. pseudotuberculosis antigen in whole blood cultures from individual animals that were sampled at repeated intervals from an uninfected research flock ID
First sample
Second sample
Third sample
C 32 D 142 E 193 E 256 E3 H 131 H 134 H 88 J 114 J 119 J 138 J 142 K 128 K 151 K 42 K 47 K 54 K 64 K 70
0 0.043 0 0 0 0.004 0.027 0 0 0.256 0 0 0.006 0.044 0 0.019 0.109 0 0.083
0 0.034 0 0.041 0 0.001 0.051 0.174 0 0.322 0 0.005 0 0.003 0 0.022 0 0 0.025
0.009 0.007 0
0.08 0
3.7. Variation in the test results within individual sheep The Rho value for the three experimentally infected sheep is 0.0846 and for the six noninfected sheep (vaccinated and control) is 0.0579. This indicates that most of the variation in titre is within the individual sheep rather than between. Table 4 Reliability of test classification at selected IFN-g ELISA optical density cut-off points IFN-g optical density cut point
Proportion of negative sheep classified as negative
Proportion of infected sheep classified as positive
0.25 0.2 0.15 0.14 0.13 0.12 0.11 0.1a 0.09 0.08 0.07 0.06
0.985 0.975 0.97 0.965 0.965 0.965 0.96 0.955 0.945 0.94 0.925 0.915
0.471 0.6 0.729 0.743 0.8 0.9 0.943 0.957 0.971 0.986 1.0 1.0
a
0.1, selected cut-off point.
J.F. Prescott et al. / Veterinary Microbiology 88 (2002) 287–297
295
4. Discussion This is the first study to use a commercially available monoclonal antibody to bovine IFN-g in a whole blood assay for detection of cell-mediated immunity in C. pseudotuberculosis-infected sheep. The study suggests that the assay, based on detection of IFN-g response to whole cell antigens, has value in detection of infection in sheep. The study showed that immunisation with a vaccine commonly used in the control of CLA in sheep did not interfere with the assay. Using a positive ‘‘cut-off’’ optical density value of 0.1, the whole blood assay correctly classified infected animals as positive and non-exposed sheep as negative with an acceptable reliability for use as a diagnostic test in a test-and-control program. These figures are the equivalent of 95.7% sensitivity and 95.5% specificity, but because the samples were taken from repeatedly tested individuals (apart from one large group of uninfected sheep) we assessed reliability rather than sensitivity and specificity. Further work is required to define the sensitivity of the assay using blood from sheep flocks known to be naturally infected with C. pseudotuberculosis, and to improve definition of the specificity of the assay using blood from sheep from flocks known to be free of infection. The end point of the assay was normalised to the positive control, to reduce some of the variation between readings in different tests. In earlier studies with the assay for the detection of bovine tuberculosis, the optical density was not normalised but rather the test was ‘‘accepted’’ if the positive optical density control was above 0.7 (Wood et al., 1991). Wood et al. (1991) regarded an animal as positive if the optical density value in the antigen containing wells was 0.05 above the nil antigen negative control. Whether the assay is less sensitive in the detection of sheep than in cattle was not investigated, but the optical density values recorded for sheep never reached some of those recorded in blood from some cattle with bovine tuberculosis (Rothel et al., 1990; Doherty et al., 1995). This difference may relate to the greater severity of infection in some cattle naturally infected with M. bovis. The most heavily infected sheep, sheep 2, gave the highest IFN-g response, and the sheep (16) shown not to be infected at time of euthanasia, gave the lowest response (Fig. 1). These results suggest that there is a relation between severity of infection of sheep and IFN-g response, but must be confirmed with a larger number of animals; however, severely infected sheep could become anergic. Factors known to affect the activity of the test include the time within which the blood sample is placed in culture, the quantity of the antigen used, and the use of heparinised blood (Rothel et al., 1992). Apart from assessing the effect of quantity of C. pseudotuberculosis antigen used, we did not investigate other factors affecting the variability of the assay results. Samples were incubated within 2 h of collection, well within the time known to be suitable for optimal performance of the assay (Rothel et al., 1992). The studies with experimentally infected sheep (Fig. 1) and confirmed by the low Rho values, showed that there is fluctuation in individual animal responses over both the shortand the long-term. All sheep varied in the same manner, suggesting that other factors may influence day-to-day results, such a stage of disease, possible bacterial shedding and health of the animal. However, change in disease classification from positive to negative was not a common event. It may be that when applying this test in a test-and-remove eradication program within a flock, negative sheep should be tested at repeated intervals, perhaps every 2 months to enhance correct classification. In the eradication of bovine tuberculosis, the
296
J.F. Prescott et al. / Veterinary Microbiology 88 (2002) 287–297
advantage of the whole blood IFN-g test over the intradermal tuberculin test is that it is not followed by the ‘‘anergy’’ observed with skin testing (Wood et al., 1992; Doherty et al., 1995), which means that a skin test cannot be repeated within 6 weeks. The fluctuation in negative animals also suggests that there may be other factors that cause a significant rise in response in some sheep (Table 3). Again Rho values were low indicating that individual sheep results varied but all negative sheep varied the same amount. However, not all sheep varied in the same way at the same sampling time. For example, between the first and second sampling, six values decreased, seven remained unchanged, and six values increased. It is possible that infection of sheep with mycolic acid-containing bacteria related to C. pseudotuberculosis, such as environmental Mycobacterium or Nocardia species, might give rise to cross-reacting antibodies. The sheep flock is free of M. avium subspecies paratuberculosis infection. The test might be improved by the use of defined antigens rather than supernatant obtained after disruption of whole cells. A marked rise in optical density values occurred in some infected (Fig. 1) and an uninfected sheep (Fig. 2); we were unable to determine the reason for this. Could this test be useful in the eradication of CLA in sheep? The test will have to be assessed in the field before an answer can be given. Although humoral tests have apparently been successful in some instances in eradication of infection (Schreuder et al., 1994; Dercksen et al., 2001), others have found that their sensitivity and specificity have been relatively low so that they have been difficult to apply to detection of individual infected animals (Pe´ pin et al., 1999). One potential advantage of the IFN-g assay over humoral assays is that it could be used in vaccinated animals, which had developed titres to phospholipase D or other antigens in response to immunisation (Eggleton et al., 1991). Since vaccination is a useful approach to control CLA (Eggleton et al., 1991), it could be combined with an eradication procedure based on the IFN-g assay which could in the long-term be the method of choice for control of the infection.
Acknowledgements We thank the Ontario Ministry of Agriculture and Food (OMAF) for supporting this work. We also thank Pam Jordan from the Arkell Research Station, and David Bridle and Tony Cengija from the OMAF Isolation Unit for their help with maintaining and bleeding sheep.
References Brown, C.C., Olander, H.J., Alves, S.F., 1987. Synergistic hemolysis-inhibition titers associated with caseous lymphadenitis in a slaughterhouse survey of goats and sheep in northeastern Brazil. Can. J. Vet. Res. 51, 46–49. Cousins, D.V., Francis, B.R., Casey, R., Mayberry, C., 1993. Mycobacterium bovis infection in a goat. Aust. Vet. J. 70, 262–263. Dercksen, D.P., Brinkhof, J.M., Dekker-Nooren, T., Maanen, K., Bode, C.F., Baird, G., Kamp, E.M., 2001. A comparison of four serological tests for the diagnosis of caseous lymphadenitis in sheep and goats. Vet. Microbiol. 75, 167–175.
J.F. Prescott et al. / Veterinary Microbiology 88 (2002) 287–297
297
Doherty, M.L., Monaghan, M.L., Bassett, H.F., Quinn, P.J., 1995. Effect of a recent injection of purified protein derivative on diagnostic tests for tuberculosis in cattle infected with Mycobacterium bovis. Res. Vet. Sci. 58, 217–221. Eggleton, D.G., Haynes, J.A., Middleton, H.D., Cox, J.C., 1991. Immunization against ovine caseous lymphadenitis: correlation between Corynebacterium pseudotuberculosis toxoid content and protective efficacy in combined clostridial–corynebacterial vaccines. Aust. Vet. J. 68, 322–325. Lan, D.T.B., Taniguchi, S., Makino, S., Shirahata, T., Nakane, A., 1998. Role of endogenous tumour necrosis factor alpha: an gamma interferon in resistance to Corynebacterium pseudotuberculosis in mice. Microbiol. Immunol. 42, 863–870. Menzies, P.I., Muckle, C.A., 1989. The use of a microagglutination assay for the detection of antibodies to Corynebacterium pseudotuberculosis in naturally infected sheep and goat flocks. Can. J. Vet. Res. 53, 313–318. Menzies, P.I., Muckle, C.A., Hwang, Y.T., Songer, J.G., 1994. Evaluation of an enzyme-linked immunosorbent assay using an Escherichia coli recombinant phospholipase D antigen for the diagnosis of Corynebacterium pseudotuberculosis infection. Small Rum. Res. 13, 193–198. Paton, M.W., Sutherland, S.S., Rose, I.R., Hart, R.A., Mercy, A.R., Ellis, T.M., 1995. The spread of Corynebacterium pseudotuberculosis infection to unvaccinated and vaccinated sheep. Aust. Vet. J. 72, 266–269. Pe´ pin, M., Seow, H.F., Corner, L., Rothel, J.S., Hodgson, A.L.M., Wood, P.R., 1997. Cytokine expression in sheep following experimental infection with various strains of Corynebacterium pseudotuberculosis differing in virulence. Vet. Res. 28, 149–163. Pe´ pin, M., Sanchis, R., Paton, M., 1999. La lymphade´ nite case´ euse des ovins et des caprins. Point Vet. 30, 33–40. Rothel, J.S., Jones, S.L., Corner, L.A., Cox, J.C., Wood, P.R., 1990. A sandwich enzyme immunoassay for bovine interferon-g and its use for the detection of tuberculosis in cattle. Aust. Vet. J. 67, 134–137. Rothel, J.S., Jones, S.L., Corner, L.A., Cox, J.C., Wood, P.R., 1992. The gamma-interferon assay for diagnosis of bovine tuberculosis in cattle: conditions affecting the production of gamma-interferon in whole blood culture. Aust. Vet. J. 69, 1–4. Schreuder, B.E.C., ter Laak, E.A., Dercksen, D.P., 1994. Eradication of caseous lymphadenitis in sheep with the help of a newly developed ELISA technique. Vet. Rec. 135, 174–176. Sting, R., Steng, G., Spengler, D., 1998. Serological studies on Corynebacterium pseudotuberculosis infections in goats using enzyme-linked immunosorbent assay. Zentralbl. Veterinarmed. B 45, 209–216. Sutherland, S.S., Ellis, T.M., Mercy, A.R., Paton, M., Middleton, H., 1987. Evaluation of an enzyme-linked immunosorbent assay for the detection of Corynebacterium pseudotuberculosis infection in sheep. Aust. Vet. J. 64, 263–266. ter Laak, E.A., Bosch, J., Bijl, G.C., Schreuder, B.E.C., 1992. Double-antibody sandwich enzyme-linked immunosorbent assay and immunoblot analysis used for control of caseous lymphadenitis in goats and sheep. Am. J. Vet. Res. 53, 1125–1132. Whipple, D.L., Bolin, C.A., Davis, A.J., Jarnigan, J.L., Johnson, D.C., Nabors, R.S., Payeur, J.B., Saari, D.A., Wilson, A.J., Wolf, M.M., 1995. Comparison of the sensitivity of the caudal fold skin test and a commercial g-interferon assay for diagnosis of bovine tuberculosis. Am. J. Vet. Res. 56, 415–419. Williamson, L.H., 2001. Caseous lymphadenitis in small ruminants. Vet. Clin. North Am. Food Anim. Pract. 17, 359–371. Wood, P.R., Corner, L.A., Rothel, J.S., Baldock, C., Jones, S.L., Cousins, D.R., McCormick, B.S., Francis, B.R., Creeper, J., Tweddle, N.E., 1991. Field comparison of the interferon-gamma assay and the intradermal tuberculin test for the diagnosis of bovine tuberculosis. Aust. Vet. J. 68, 286–290. Wood, P.R., Corner, L.A., Rothel, J.S., Ripper, J.L., Fifis, T., McCormick, B.S., Francis, B., Melville, L., Small, K., De Witte, K., Tolson, J., Ryan, T.J., de Lisle, G.W., Cox, J.C., Jones, S.L., 1992. A field evaluation of serological and cellular diagnostic tests for bovine tuberculosis. Vet. Microbiol. 31, 71–79.