- Email: [email protected]
Antibodies to mycobacteria in cattle not infected with mycobacterium bovis
Veterinary Microbiology, 18 (1988) 51-61 Elsevier Science Publishers B.V., Amsterdam - - Printed in The Netherlands
51
A n t i b o d i e s to M y c o b a c t e r i a in Cattle not I n f e c t e d w i t h Mycobacterium bovis L.A. AUER and S.M. SCHLEEHAUF
Queensland Department of Primary Industries, Animal Research Institute, YeerongpiUy, Qld. (Australia) (Accepted for publication 1 March 1988)
ABSTRACT Auer, L.A. and Schleehauf, S.M., 1988. Antibodies to mycobacteria in cattle not infected with Mycobacterium boris. Vet. Microbiol., 18: 51-61. An indirect anti-IgG enzyme-linked immunosorbent assay (ELISA) using a whole cell sonicate of Mycobacterium boris as the coating antigen, was used to detect anti-mycobacterial antibodies in cattle not infected with M. boris. False positive M. boris ELISA scores were produced in 6 cattle experimentally inoculated with Mycobacterium avium-intraceUulare-scrofulaceum (MAIS) serovars 2, 8, 9, 14 and 18 and Mycobacterium flavescens, respectively. False positive ELISA results were also found in 39.5% of cattle from which other mycobacteria were cultured and in 56.4% of necropsied cattle with other pathological conditions. No M. boris was cultured from these animals. Other groups of animals, with no pathological conditions, which had been tuberculintested negative, tuberculin-tested positive and never tuberculin tested showed positive ELISA results in 15.4%, 73.6% and 42.4% of the respective groups. The variation of these non-specific responses in uninfected cattle highlights the need for careful selection of negative controls in evaluating ELISAs for the diagnosis of bovine tuberculosis.
INTRODUCTION S h a r e d epitopes b e t w e e n the m y c o b a c t e r i a l species have been c o n s i d e r e d responsible for specificity p r o b l e m s in b o t h c e l l - m e d i a t e d a n d serological tests for b o v i n e t u b e r c u l o s i s ( C o r n e r a n d P e a r s o n , 1978; K e t t e r e r et al., 1981; Lepper a n d C o r n e r , 1983; Auer, 1987). I n f o r m a t i o n , however, is s c a n t y on the e x t e n t of this problem. I n e v a l u a t i n g d i a g n o s t i c tests, i n a p p r o p r i a t e s a m p l i n g of negative c o n t r o l g r o u p s m a y lead to a b e r r a t i o n s in test i n t e r p r e t a t i o n s . B a c k g r o u n d k n o w l e d g e of t h e p r e v a l e n c e a n d degree of i m m u n e r e s p o n s e s to m y c o b a c t e r i a o t h e r t h a n M y c o b a c t e r i u m boris ( h e r e a f t e r referred to as o t h e r m y c o b a c t e r i a ) is t h e r e f o r e of value in the a s s e s s m e n t a n d a p p l i c a t i o n of dia g n o s t i c tests. In h u m a n t u b e r c u l o s i s a n d leprosy, a s p e c t r u m of i m m u n e r e s p o n s e s has
0378-1135/88/$03.50
© 1988 Elsevier Science Publishers B.V.
52
demonstrated that low cell-mediated responses are generally associated with high serum antibody levels (Lenzini et al., 1977). The enzyme-linked immunosorbent assay (ELISA) offers the potential of a sensitive and convenient method to detect those infected cattle with negative cell-mediated immune responses (intradermal tuberculin test). An indirect E L I S A using a whole cell sonicate of M. bovis as the antigen was recently evaluated as a serodiagnostic aid for bovine tuberculosis (Auer, 1987). Sensitivity was 88.7% and 63% for infected cattle which gave positive intradermal tuberculin tests and tbr cattle which had never been tuberculin tested, respectively. Specificity, however, was low at 52.6%. The present study was undertaken to determine the prevalence of mycobacterial antibodies in uninfected cattle and to investigate and identify factors which may be implicated in false positive reactions. The role of other mycobacteria, the presence of pathological conditions and the influence of tuberculin testing were examined. MATERIALS AND METHODS
Antigen All cultures used in this study were obtained from lesions found in cattle at slaughter. Organisms were classified by cultural and drug sensitivity characteristics. M. avium complex strains were serotyped by the method of Reznikov and Leggo ( 1972 ). Strains other than M. bovis were grown in Dubos broth and plated onto Middlebrook 7 H l l agar. M. bovis was grown on a modified Middlebrook 7 H l l broth and plated onto a modified Middlebrook 7 H l l agar, according to the method of Gallagher and Horwill (1977). Growth was scraped o f f w i t h a pipette either at 3-4 weeks (M. bovis) or 5-10 days (other mycobacteria ), washed and suspended in saline at 2 mg m l - 1 (antigens for inoculation ) and 10 mg m1-1 (ELISA antigen) (wet weight). Mycobacteria used for the inoculations were killed by ultraviolet irradiation (M. bovis, 60 min; other mycobacteria, 30 min). For the ELISA, the mycobacteria were heat killed (120 rain, 60 ° C, 10 mg m l - 1). The antigen used in the ELISA was sonicated (2 min, continuous mode, repeated 10-12 times in a Sonifier Cell Disruptor, Branson Sonic Power Company ) and dispensed into i ml aliquots for storage at - 20 ° C.
E L I S A procedure The procedure used was an indirect anti-bovine IgG ELISA for M. bovis (Auer, 1987 ). Briefly, M. bovis sonicated whole cells ( 10 mg m l - 1) were diluted 1 in 50 and 50 pl transferred to each of the 96 wells of a flat-bottomed microtitre plate ( A / S Nunc, Kamstrup, DK-4000 Roskilde, Denmark). After incubation at 37°C (1 h) and washing in phosphate-buffered saline, 0.05% Tween 20 ( P B S / T 2 0 ) , test serum (50 ~1, 1/100 dilution in P B S / T 2 0 / E D T A ) was
53 added to each well. Affinity-purified goat anti-bovine IgG (H and L chains) conjugated to alkaline phosphatase (Kirkegaard and Perry Laboratories Inc., Gaithersburg, MD) was added (50 zl at 1 zg m1-1) after further incubation and washing steps. Finally, the enzyme was detected by the substrate P-nitrophenylphosphate ( P N P P , 100 zl, 1 mg m1-1) after 30 min incubation. The absorbance values were read at 405 nm and 630 nm on a dual wavelength spectrophotometric microplate reader (Dynatech MR600, Dynatech Laboratories Inc., Alexandria, VA).
Mathematical analyses The ELISA absorbance data were converted to a percentage of t h a t of a positive serum (from an infected animal) replicated 8 times across and down each plate. This score was more reproducible t h a n a percentage of a negative serum also replicated 8 times, because the coefficient of variation of the posirive control (0.095) was less t h a n t h a t of the negative control (0.165) measured over 50 plates. The negative serum came from an animal in a confirmed free herd which had never been tuberculin tested. Group data was analysed by one way analysis of variance to determine significant differences between the mean values of the different groups.
Experimental procedure The sera used were from 6 groups of cattle, designated 1, 2, 3, 4, 5 and 6. These animals were negative for M. bovis either by negative bacteriological culture or confirmed free herd status. The serum samples were tested with the ELISA using the M. bovis antigen. From previous work (Auer, 1987 ) scores of greater t h a n 35% of a known positive M. bovis serum (No. 1495) were considered positive and those below 35%, negative. Group I consisted of 8 animals, 7 of which were immunized 2 or 3 times with different mycobacteria at approximately 50-day intervals. These cattle were from cont~irmed free herds which had never been tuberculin tested. The cattle were inoculated subcutaneously with 5 ml of a 2 mg m l - 1 suspension of ultraviolet killed whole mycobacterial cells. Blood samples were taken at approximately weekly intervals over a 120-150-day period. The mycobacteria inoculated were M. bovis, Mycobacterium avium-intraceUulare-scrofulaceum (MAIS) serovars 2, 8, 9, 14, 18 and M. flavescens. One animal (control) was not inoculated. All sera were collected and stored at - 7 0 ° C and were tested together after the experimental period was over. Group 2 consisted of 38 animals which had been tuberculin tested positive and which had other mycobacteria, but no M. bovis isolated from tissues submitted for bacteriological culture. Only 9 cattle showed lesions at slaughter. The lymph nodes were collected as
54
cleanly as possible from field necropsies (32) and abattoirs (6). The abattoir lymph nodes were not incised. Group 3 comprised 78 cattle which had never been tuberculin tested and which exhibited a variety of other pathological changes (not M. boris-related) at necropsy, ranging from Nocardia granulomas to purulent lymphadenitis (see Table 3). No M. bovis and no other mycobacteria were isolated from tissues submitted from Group 3. Groups 4, 5 and 6 were groups of cattle with different tuberculin testing histories, but showing no pathological changes at slaughter (time of blood collection). Group 4 animals ( 78 cattle) were tuberculin-tested negative 56 days prior to blood collection. Group 5 comprised 87 animals tuberculin-tested positive with blood samples taken 3-7 days later. Tissue samples submitted from Group 5 cattle were negative on histological examination and
A.
1 O0
M.
boris
,$
Control
i-
--
....
m
80
60 0_ ~e 40
20
I
I
10
20
I
I
I
I
30
40
50
60
I
I
I
I
t
70
I
80
90
100
1 10
120
I 70
I 80
I 90
I 100
I 110
I !20
Days
B. I O0
MAIS MAIS
9 14
m- . . . .
,~ -n
--
80
--
60
--
40
--
20
--
,t',, o_ ~e
I 10
I 20
I 30
I 40
I 50
I 60 Days
Fig. 1. Graphs of M. boris ELISA scores against time of sera from: (A) one animal inoculated with killed M. bovis whole cells (5 ml, 2 mg ml i s.c.) and one control animal which was not inoculated; (B) two animals inoculated with killed whole cells of MAIS serovar 9 and M A I S serovar 14, respectively (5 ml, 2 mg ml 1 s.c.). Arrows indicate times of inoculation.
55
mycobacteriological culture. Group 6 was a control group of 66 cattle with no pathological conditions and never tuberculin tested. RESULTS
Cattle injected with other mycobacteria (Group 1) The ELISA scores of sera taken from cattle injected with other mycobacteria (Group 1) are graphed in Figs. 1 and 2. Following inoculation, increases in antibody titre to M. bovis antigen occurred in all animals. The mean antibody score of the 6 animals injected with other mycobacteria rose from 34.7 to 44.1% at 14 days post-injection. In comparison, the M. bovis-injected animal rose from 30.9 to 59.4% and the control animal 25.9 to 31.0%. Sera from the unA. 100
MAIS MAIS
2 8
¢ ~ ....
-
80 60
/
.,--_...
":
7---....,_ .... -.
40
20
I I0
I 20
L 30
L 40
I 50
I 60
I 70
L 80
I 90
I 100
I II0
I I 120 130
Days B.
100]
MAIS 18 M. f l a v e s c e n s
~ I- ....
41
80 r~
i r ~
/"
60 I
"~I
2O
l I0
I 20
I 30
I 40
I 50
I 60
I 70
I 80
] 90
I 100
I II0
I 120
I 130
Days
Fig. 2. Graphs of M. bovis E L I S A scores against time of sera from: (A) two animals inoculated with killed whole cells of M A I S serovar 2 a n d M A I S serovar 8, respectively (5 ml, 2 m g m l - 1 s.c. ); ( B ) two animals inoculated with killed whole cells of M A I S serovar 18 a n d M. flavescens, respectively (5 ml, 2 m g m l - ] s.c.). Arrows indicate times of inoculation.
56
inoculated control animal showed only minor fluctuations in titre over the period of the experiment. E L I S A scores in this animal ranged from 25.9 to 32.4%. The percentage increase in E L I S A scores rose upon subsequent injections in the inoculated cattle, except for the 2 cattle injected with MAIS serovars 9 and 14. The time taken for peak antibody responses to be reached varied from 3 to 45 days. There was no indication that antibody scores from second or third injections peaked earlier than those from initial injections. Some fluctuations occurred unrelated to the inoculations but these were generally smaller in magnitude than the other responses. Because of the small numbers in this group, statistical analysis was not appropriate for within-group differences. For between-group means (Groups 1-6) one way analysis of variance showed significant differences (F ratio 14.138, P < 0.001 ). However, there was no significant difference in the 14-day Group 1 mean (M. bovis and control animal excluded) and Group 6 animals never tuberculin tested with no pathological conditions (Table 1).
TABLE 1
M. bovis E L I S A a n t i b o d y scores of groups of cattle n o t i n f e c t e d w i t h M. boris Mean ELISA score (% P 0 S ) 1+ S E
% Cattle w i t h s c o r e s > 35% (positives)
6
44.1 _+5.2
83.3
Cattle w i t h o t h e r m y c o b a c t e r i a isolated T T P O S :~
38
36.6 + 2.5
39.5
Cattle w i t h o t h e r pathological c o n d i t i o n s N T T :~
78
44.04 _+2.5
56.4
Cattle w i t h no pathological c o n d i t i o n s T T N E G :~
78
27.44 + 2.1
15.4
Cattle w i t h no pathological c o n d i t i o n s ~ T T P O S :~
87
54.14 + 3.2
73.6
Cattle w i t h no pathological c o n d i t i o n s N T T :~
66
35.2 _+ 1.8
42.4
Group
1. 2. 3. 4. 5. 6.
No.
Cattle injected w i t h o t h e r mycobacteria'-' N T T :~
~%POS = % of positive control s e r u m No. 1495. SSamples t a k e n 14 days after first inoculation, M. bovis a n d control a n i m a l s excluded. :~NTT = never tuberculin tested; T T N E G = tuberculin-tested negative; T T P O S = tuberculin-tested positive. ~Mean scores significantly different ( P < 0.01 ) to G r o u p 6 m e a n . '-'Histological e x a m i n a t i o n a n d mycobacteriological culture negative.
57 TABLE2 Distribution of M. bovis E L I S A positive scores found in cattle from which other mycobacteria were isolated Mycobacteria isolated
Number of times Organisms isolated
M. m a r i n u m M. phlei M. flavescens MAIS unknown serovar M. terrae, MAIS unknown serovar M. terrae MAIS serovar 3 MAIS serovar 8 M. terrae, Gp. IV M. terrae, M. flavescens M. flavescens, Gp. IV M. flavescens, MAIS serovar 15 M. simmiae, MAIS serovar 10
1 1 6 8 2 12 1 1 1 2 1 1 1
Total numbers
38
ELISA positive1
1 1 4 4 1 4 0 0 0 0 0 0 0 15 (39.5)
Walues greater t h a n 35% of a known positive serum (Auer, 1987 ).
Cattle with other mycobacteria isolated (Group 2) Fifteen of 38 animals from which other mycobacteria were isolated (39.5 % ), including 4 of 9 animals with suspect M. boris lesions at slaughter (later cultured negative for M. boris), were positive in the M. boris E L I S A (Table 2). The mean antibody titre was 36.6% (Table 1). There was no significant difference between the mean of Group 2 cattle and the mean of Group 6 cattle (no pathological conditions, never tuberculin tested).
Cattle with other pathological conditions (Group 3) The mean E L I S A score of Group 3 was 44.0% with 56.4% of scores being positive (Table 1 ). The mean of this group was significantly higher ( P < 0.01 ) than the mean of Group 6 (no pathological conditions, never tuberculin tested). The distribution of negative and false positive results in animals with lesions such as club forming foreign body granulomas was similar to that found in animals from which Rhodococcus was isolated (Table 3). The 2 animals from which Streptomyces/Nocardia were diagnosed were negative while sera from animals with granulomatous lesions of undetermined cause gave 9 out of 12 positive E L I S A results.
58 TABLE 3 Positive M. boris E L I S A scores found in cattle with other pathological conditions, culture negative for mycobacteria and never t u b e r c u l i n - tested Pathology ~
N u m b e r of animals Total No.
Parasitic granulomas Granulomatous lesion - u n d e t e r m i n e d cause Squamous cell carcinoma Rhodococcus equi granulomas Club-forming granulomas (actinobacilloses/ actinomycoses ) Foreign body granuloma Fibrosis of lymph node Hydatid cyst Other neoplasms Streptomyces/Nocardia granuloma Purulent lymphadenitis Abscess Total numbers
E L I S A positive ~
2 12 4 8
2 9 3 5
28 9 2 5 4 2 1
16 5 1 2 1 0 0
1
0
78
44 (56.4%)
~Histological diagnosis of tissues with gross lesions s u b m i t t e d for examination for tuberculosis. ~Values greater t h a n 35% of a known positive serum (Auer, 1987).
Cattle with no lesions and different tuberculin testing histories (Groups 4, 5 and 6) The mean antibody E L I S A scores (Table 1 ) were 27.4, 54.1 and 35.2%, respectively, for cattle tuberculin-tested negative (Group 4), tuberculin-tested positive (Group 5 ) and never tuberculin tested (Group 6). The corresponding percentages of positive scores in each group were 15.4, 73.6 and 42.4%. There were significant differences between each of Groups 4, 5 and 6 ( P < 0.01 ). DISCUSSION
In this investigation, nonspecific antibodies to M. bovis were frequently found in cattle uninfected with this bacteria. The results indicated that exposure to other mycobacteria and the presence of other pathological conditions would increase this nonspecific response. In contrast, the prevalence of false positive results was significantly lower in cattle which had been tuberculin tested several times in eradication surveys. The variation seen in this study indicates the importance of using large numbers of randomly-selected negative controls in the evaluation of serological tests for bovine tuberculosis. The development of an accurate serological test for tuberculosis has been
59 pursued for many decades. With the advent of highly-sensitive and practical technology such as the ELISA, there has been renewed interest in this area. ELISAs to detect M. bovis antibodies in pigs and cattle (Thoen et al., 1981; Auer, 1987) have been developed but lack specificity even when a purified protein derivative ( P P D ) is used as the antigen (Thoen et al., 1981 ). A whole cell sonicate was used as the antigen in the E L I S A in the present study in order to detect the full extent of antibody responses to shared epitopes. If cattle infected with M. boris react predominantly to shared epitopes as has been reported to occur in human tuberculosis and leprosy (Stanford, 1983 ), the search for a unique antigen of M. boris to improve specificity in serological tests may become less relevant. It would then become more important to target and identify those factors which are most implicated in false positive reactions. Harboe and Nagai (1984) have reported an antigen M P B 7 0 in high concentrations in M. bovis BCG and in lesser concentrations in M. tuberculosis H 3 7 R U and in Nocardia asteroides. More specific antigens such as these could be used to determine their potential in diagnostic tests. Exposure to other mycobacteria was found in this study to result in markedly increased M. boris E L I S A scores but at different times post-inoculation in individual animals. These responses may not reflect the natural situation due to the artificial nature of the exposure. Nevertheless, the antigens of the other mycobacteria evoked M. bovis antibody responses as measured by the ELISA. However, in animals from which other mycobacteria were isolated, only 39.5% had false positive M. bovis E L I S A scores compared to 73.6% of tuberculin test positive, culture negative cattle with no pathological conditions. The lack of response by 60.5% of Group 2 could have been due to the cattle actually being unexposed to other mycobacteria, the environmental contamination during collection of the lymph node causing the positive culture results. Against this, all cattle in the group were false positive reactors to the tuberculin test. A possible explanation is that the antigen in the tuberculin injection mopped up circulating antibody and decreased titres. The high percentage of false positive results (56.4%) in cattle with a variety of pathological conditions (Table 3) indicates the possible existence of other nonspecific stimuli. The significantly higher mean E L I S A score of this group compared to the group with no pathological conditions and never tuberculin tested (Group 6) supports this idea. Minden et al. (1972) showed that sera from rabbits immunized with M. boris strain BCG bound not only radiolabelled homologous test antigens b u t also radiolabelled antigens from many other unrelated bacteria. Alternatively, high levels of anti-mycobacterial antibodies occurring in other infectious diseases may be due to a nonspecific elevation of antibody synthesis as occurs in some chronic diseases such as malaria (Grange, 1984 ). Significant differences were found between the groups of cattle with no pathological conditions and different tuberculin testing histories. The group
60 which had been tuberculin-tested negative 56 days prior to slaughter showed the lowest mean antibody levels while those cattle tuberculin-tested positive 3-7 days prior to blood sampling showed the highest scores (P<0.01). One explanation for this could be the culling of false positive tuberculin reactor cattle in repeated herd tests. The tuberculin test negative group had a long history of tuberculin testing and this may have decreased the percentage of cattle exposed to agents mimicking M. bovis immune responses. A beneficial consequence, if ELISA was applied for diagnosis, might be greatly improved specificity in extensively-tested herds. Another factor boosting the mean score in tuberculin-tested positive animals with no pathological conditions (Group 5) could be the immunogenic effect of the tuberculin. Previous work with infected animals tuberculin-tested positive 3-7 days earlier showed higher mean ELISA scores than infected animals never tuberculin tested (Auer, 1987). However, these results are at variance with the tuberculin-positive cattle with other mycobacteria isolated (Group 2), which had unexpectedly low antibody scores. A fall in antibody titre occurred in active human tuberculosis patients 2-4 days after a tuberculin test. Such a drop did not occur in patients with inactive tuberculosis or with other diseases (Algeorge and Stoian, 1973). In cattle, it has been shown that tuberculin does not increase antibody titres from 34 days to 1 year post-inoculation (Auer, 1987). If, however, there is a shortterm fluctuation in titres, the application of ELISAs to recently tuberculintested cattle should be closely examined. In the development ofa serodiagnostic test for M. bovis, accuracy will depend upon the prevalences in sera of specific and nonspecific antibodies to M. bovis. This paper reports on various factors influencing these prevalences. By studying the factors affecting nonspecific responses, strategies for developing and applying accurate tests based on this knowledge will be of benefit in eradicating bovine tuberculosis. ACKNOWLEDGEMENTS We thank T. Veerman, G. Smith, J. Pike and K. Mason of the Animal Research Institute (ARI), Yeerongpilly, for the preparation of M. bovis antigen and bacteriological diagnosis of lesions, G. Storie and R. Pierce (ARI), for histological diagnosis of lesions and G. Tudor (ARI) and staff for care of the experimental cattle. We also thank Dr. G. Simmons, J. Walthall and staff of Veterinary Services Branch, Queensland Department of Primary Industries ( QDPI ) and J. Anderson and staff, Veterinary Public Health Branch (QDPI), for their support in the establishment of a serum bank of samples from infected and non-infected animals. Funds for this work were provided by the Australian National Brucellosis and Tuberculosis Eradication Scheme.
61 REFERENCES Auer, L.A., 1987. Assessment of an enzyme linked immunosorbent assay for the detection of cattle infected with Mycobacterium bovis. Aust. Vet. J., 64:172 176. Algeorge, G. and Stoian, M., 1973. Serological diagnosis of tuberculosis by fluorescent antibody test and kaolin agglutination test. Probl. Tuberk., 12:99 113. Corner, L.A. and Pearson, C.W., 1978. Pathogenicity for cattle of atypical mycobacteria isolated from feral pigs and cattle and the correlation of lesions with tuberculin sensitivity. Aust. Vet. J., 54: 280-286. Gallagher, J. and Horwill, D.M., 1977. A selective oleic acid albumin agar medium for the cultivation of Mycobacterium bovis. J. Hyg., Camb., 79:155 160. Grange, J.M., 1984. The humoral immune response in tuberculosis: its nature, biological role and diagnostic usefulness. Adv. Tuberc. Res., 21: 1-78. Harboe, M. and Nagai, S., 1984. MPB70, a unique antigen of Mycobacterium boris BCG. Am. Rev. Resp. Dis., 129: 444-452. Ketterer, P.J., Rogers, R.J. and Donald, B., 1981. Pathology and tuberculin sensitivity in cattle inoculated with Myobacterium avium complex serotypes 6, 14 and 18. Aust. Vet. J., 57: 61-65. Lenzini, L., Rottoli, P. and Rottoli, L., 1977. The spectrum of human tuberculosis. Clin. Exp. hnmunol., 27: 230-237. Lepper, A.W.D. and Corner, L.A., 1983. Serological tests. In: C. Ratledge and J. Stanford (Editors), The Biology of the Mycobacteria, Vol. 2. Academic Press, Sydney, pp. 490-495. Minden, P., McClatchy, J.K. and Farr, R.S., 1972. Shared antigens between heterologous bacterial species. Infect. Immun., 6:574 582. Reznikov, M. and Leggo, J.H., 1972. Modification of Schaefer's procedure for serotyping of organisms of the Mycobacterium avium-M, intracellulare and M. scrofulaceum complex. Appl. Microbiol., 23: 819-823. Stanfl)rd, J.L., 1983. Immunologically important constituents of mycobacteria: antigens. In: C. Ratledge and J. Stanf'ord (Editors), The Biology of the Mycobacteria, Vol. 2. Academic Press, Sydney, pp. 85-127. Thoen, C.O., Ivey, C.R., Hines, E.M. and Harrington, Jr. R., 1981. Application of an enzyme linked immunosorbent assay fl)r detecting antibodies in sera of feral swine naturally exposed to Mycobacteriura bovis. Proceedings of the 85th Annual Meeting of the U.S. Animal Health Association, pp. 500 504.
51
A n t i b o d i e s to M y c o b a c t e r i a in Cattle not I n f e c t e d w i t h Mycobacterium bovis L.A. AUER and S.M. SCHLEEHAUF
Queensland Department of Primary Industries, Animal Research Institute, YeerongpiUy, Qld. (Australia) (Accepted for publication 1 March 1988)
ABSTRACT Auer, L.A. and Schleehauf, S.M., 1988. Antibodies to mycobacteria in cattle not infected with Mycobacterium boris. Vet. Microbiol., 18: 51-61. An indirect anti-IgG enzyme-linked immunosorbent assay (ELISA) using a whole cell sonicate of Mycobacterium boris as the coating antigen, was used to detect anti-mycobacterial antibodies in cattle not infected with M. boris. False positive M. boris ELISA scores were produced in 6 cattle experimentally inoculated with Mycobacterium avium-intraceUulare-scrofulaceum (MAIS) serovars 2, 8, 9, 14 and 18 and Mycobacterium flavescens, respectively. False positive ELISA results were also found in 39.5% of cattle from which other mycobacteria were cultured and in 56.4% of necropsied cattle with other pathological conditions. No M. boris was cultured from these animals. Other groups of animals, with no pathological conditions, which had been tuberculintested negative, tuberculin-tested positive and never tuberculin tested showed positive ELISA results in 15.4%, 73.6% and 42.4% of the respective groups. The variation of these non-specific responses in uninfected cattle highlights the need for careful selection of negative controls in evaluating ELISAs for the diagnosis of bovine tuberculosis.
INTRODUCTION S h a r e d epitopes b e t w e e n the m y c o b a c t e r i a l species have been c o n s i d e r e d responsible for specificity p r o b l e m s in b o t h c e l l - m e d i a t e d a n d serological tests for b o v i n e t u b e r c u l o s i s ( C o r n e r a n d P e a r s o n , 1978; K e t t e r e r et al., 1981; Lepper a n d C o r n e r , 1983; Auer, 1987). I n f o r m a t i o n , however, is s c a n t y on the e x t e n t of this problem. I n e v a l u a t i n g d i a g n o s t i c tests, i n a p p r o p r i a t e s a m p l i n g of negative c o n t r o l g r o u p s m a y lead to a b e r r a t i o n s in test i n t e r p r e t a t i o n s . B a c k g r o u n d k n o w l e d g e of t h e p r e v a l e n c e a n d degree of i m m u n e r e s p o n s e s to m y c o b a c t e r i a o t h e r t h a n M y c o b a c t e r i u m boris ( h e r e a f t e r referred to as o t h e r m y c o b a c t e r i a ) is t h e r e f o r e of value in the a s s e s s m e n t a n d a p p l i c a t i o n of dia g n o s t i c tests. In h u m a n t u b e r c u l o s i s a n d leprosy, a s p e c t r u m of i m m u n e r e s p o n s e s has
0378-1135/88/$03.50
© 1988 Elsevier Science Publishers B.V.
52
demonstrated that low cell-mediated responses are generally associated with high serum antibody levels (Lenzini et al., 1977). The enzyme-linked immunosorbent assay (ELISA) offers the potential of a sensitive and convenient method to detect those infected cattle with negative cell-mediated immune responses (intradermal tuberculin test). An indirect E L I S A using a whole cell sonicate of M. bovis as the antigen was recently evaluated as a serodiagnostic aid for bovine tuberculosis (Auer, 1987). Sensitivity was 88.7% and 63% for infected cattle which gave positive intradermal tuberculin tests and tbr cattle which had never been tuberculin tested, respectively. Specificity, however, was low at 52.6%. The present study was undertaken to determine the prevalence of mycobacterial antibodies in uninfected cattle and to investigate and identify factors which may be implicated in false positive reactions. The role of other mycobacteria, the presence of pathological conditions and the influence of tuberculin testing were examined. MATERIALS AND METHODS
Antigen All cultures used in this study were obtained from lesions found in cattle at slaughter. Organisms were classified by cultural and drug sensitivity characteristics. M. avium complex strains were serotyped by the method of Reznikov and Leggo ( 1972 ). Strains other than M. bovis were grown in Dubos broth and plated onto Middlebrook 7 H l l agar. M. bovis was grown on a modified Middlebrook 7 H l l broth and plated onto a modified Middlebrook 7 H l l agar, according to the method of Gallagher and Horwill (1977). Growth was scraped o f f w i t h a pipette either at 3-4 weeks (M. bovis) or 5-10 days (other mycobacteria ), washed and suspended in saline at 2 mg m l - 1 (antigens for inoculation ) and 10 mg m1-1 (ELISA antigen) (wet weight). Mycobacteria used for the inoculations were killed by ultraviolet irradiation (M. bovis, 60 min; other mycobacteria, 30 min). For the ELISA, the mycobacteria were heat killed (120 rain, 60 ° C, 10 mg m l - 1). The antigen used in the ELISA was sonicated (2 min, continuous mode, repeated 10-12 times in a Sonifier Cell Disruptor, Branson Sonic Power Company ) and dispensed into i ml aliquots for storage at - 20 ° C.
E L I S A procedure The procedure used was an indirect anti-bovine IgG ELISA for M. bovis (Auer, 1987 ). Briefly, M. bovis sonicated whole cells ( 10 mg m l - 1) were diluted 1 in 50 and 50 pl transferred to each of the 96 wells of a flat-bottomed microtitre plate ( A / S Nunc, Kamstrup, DK-4000 Roskilde, Denmark). After incubation at 37°C (1 h) and washing in phosphate-buffered saline, 0.05% Tween 20 ( P B S / T 2 0 ) , test serum (50 ~1, 1/100 dilution in P B S / T 2 0 / E D T A ) was
53 added to each well. Affinity-purified goat anti-bovine IgG (H and L chains) conjugated to alkaline phosphatase (Kirkegaard and Perry Laboratories Inc., Gaithersburg, MD) was added (50 zl at 1 zg m1-1) after further incubation and washing steps. Finally, the enzyme was detected by the substrate P-nitrophenylphosphate ( P N P P , 100 zl, 1 mg m1-1) after 30 min incubation. The absorbance values were read at 405 nm and 630 nm on a dual wavelength spectrophotometric microplate reader (Dynatech MR600, Dynatech Laboratories Inc., Alexandria, VA).
Mathematical analyses The ELISA absorbance data were converted to a percentage of t h a t of a positive serum (from an infected animal) replicated 8 times across and down each plate. This score was more reproducible t h a n a percentage of a negative serum also replicated 8 times, because the coefficient of variation of the posirive control (0.095) was less t h a n t h a t of the negative control (0.165) measured over 50 plates. The negative serum came from an animal in a confirmed free herd which had never been tuberculin tested. Group data was analysed by one way analysis of variance to determine significant differences between the mean values of the different groups.
Experimental procedure The sera used were from 6 groups of cattle, designated 1, 2, 3, 4, 5 and 6. These animals were negative for M. bovis either by negative bacteriological culture or confirmed free herd status. The serum samples were tested with the ELISA using the M. bovis antigen. From previous work (Auer, 1987 ) scores of greater t h a n 35% of a known positive M. bovis serum (No. 1495) were considered positive and those below 35%, negative. Group I consisted of 8 animals, 7 of which were immunized 2 or 3 times with different mycobacteria at approximately 50-day intervals. These cattle were from cont~irmed free herds which had never been tuberculin tested. The cattle were inoculated subcutaneously with 5 ml of a 2 mg m l - 1 suspension of ultraviolet killed whole mycobacterial cells. Blood samples were taken at approximately weekly intervals over a 120-150-day period. The mycobacteria inoculated were M. bovis, Mycobacterium avium-intraceUulare-scrofulaceum (MAIS) serovars 2, 8, 9, 14, 18 and M. flavescens. One animal (control) was not inoculated. All sera were collected and stored at - 7 0 ° C and were tested together after the experimental period was over. Group 2 consisted of 38 animals which had been tuberculin tested positive and which had other mycobacteria, but no M. bovis isolated from tissues submitted for bacteriological culture. Only 9 cattle showed lesions at slaughter. The lymph nodes were collected as
54
cleanly as possible from field necropsies (32) and abattoirs (6). The abattoir lymph nodes were not incised. Group 3 comprised 78 cattle which had never been tuberculin tested and which exhibited a variety of other pathological changes (not M. boris-related) at necropsy, ranging from Nocardia granulomas to purulent lymphadenitis (see Table 3). No M. bovis and no other mycobacteria were isolated from tissues submitted from Group 3. Groups 4, 5 and 6 were groups of cattle with different tuberculin testing histories, but showing no pathological changes at slaughter (time of blood collection). Group 4 animals ( 78 cattle) were tuberculin-tested negative 56 days prior to blood collection. Group 5 comprised 87 animals tuberculin-tested positive with blood samples taken 3-7 days later. Tissue samples submitted from Group 5 cattle were negative on histological examination and
A.
1 O0
M.
boris
,$
Control
i-
--
....
m
80
60 0_ ~e 40
20
I
I
10
20
I
I
I
I
30
40
50
60
I
I
I
I
t
70
I
80
90
100
1 10
120
I 70
I 80
I 90
I 100
I 110
I !20
Days
B. I O0
MAIS MAIS
9 14
m- . . . .
,~ -n
--
80
--
60
--
40
--
20
--
,t',, o_ ~e
I 10
I 20
I 30
I 40
I 50
I 60 Days
Fig. 1. Graphs of M. boris ELISA scores against time of sera from: (A) one animal inoculated with killed M. bovis whole cells (5 ml, 2 mg ml i s.c.) and one control animal which was not inoculated; (B) two animals inoculated with killed whole cells of MAIS serovar 9 and M A I S serovar 14, respectively (5 ml, 2 mg ml 1 s.c.). Arrows indicate times of inoculation.
55
mycobacteriological culture. Group 6 was a control group of 66 cattle with no pathological conditions and never tuberculin tested. RESULTS
Cattle injected with other mycobacteria (Group 1) The ELISA scores of sera taken from cattle injected with other mycobacteria (Group 1) are graphed in Figs. 1 and 2. Following inoculation, increases in antibody titre to M. bovis antigen occurred in all animals. The mean antibody score of the 6 animals injected with other mycobacteria rose from 34.7 to 44.1% at 14 days post-injection. In comparison, the M. bovis-injected animal rose from 30.9 to 59.4% and the control animal 25.9 to 31.0%. Sera from the unA. 100
MAIS MAIS
2 8
¢ ~ ....
-
80 60
/
.,--_...
":
7---....,_ .... -.
40
20
I I0
I 20
L 30
L 40
I 50
I 60
I 70
L 80
I 90
I 100
I II0
I I 120 130
Days B.
100]
MAIS 18 M. f l a v e s c e n s
~ I- ....
41
80 r~
i r ~
/"
60 I
"~I
2O
l I0
I 20
I 30
I 40
I 50
I 60
I 70
I 80
] 90
I 100
I II0
I 120
I 130
Days
Fig. 2. Graphs of M. bovis E L I S A scores against time of sera from: (A) two animals inoculated with killed whole cells of M A I S serovar 2 a n d M A I S serovar 8, respectively (5 ml, 2 m g m l - 1 s.c. ); ( B ) two animals inoculated with killed whole cells of M A I S serovar 18 a n d M. flavescens, respectively (5 ml, 2 m g m l - ] s.c.). Arrows indicate times of inoculation.
56
inoculated control animal showed only minor fluctuations in titre over the period of the experiment. E L I S A scores in this animal ranged from 25.9 to 32.4%. The percentage increase in E L I S A scores rose upon subsequent injections in the inoculated cattle, except for the 2 cattle injected with MAIS serovars 9 and 14. The time taken for peak antibody responses to be reached varied from 3 to 45 days. There was no indication that antibody scores from second or third injections peaked earlier than those from initial injections. Some fluctuations occurred unrelated to the inoculations but these were generally smaller in magnitude than the other responses. Because of the small numbers in this group, statistical analysis was not appropriate for within-group differences. For between-group means (Groups 1-6) one way analysis of variance showed significant differences (F ratio 14.138, P < 0.001 ). However, there was no significant difference in the 14-day Group 1 mean (M. bovis and control animal excluded) and Group 6 animals never tuberculin tested with no pathological conditions (Table 1).
TABLE 1
M. bovis E L I S A a n t i b o d y scores of groups of cattle n o t i n f e c t e d w i t h M. boris Mean ELISA score (% P 0 S ) 1+ S E
% Cattle w i t h s c o r e s > 35% (positives)
6
44.1 _+5.2
83.3
Cattle w i t h o t h e r m y c o b a c t e r i a isolated T T P O S :~
38
36.6 + 2.5
39.5
Cattle w i t h o t h e r pathological c o n d i t i o n s N T T :~
78
44.04 _+2.5
56.4
Cattle w i t h no pathological c o n d i t i o n s T T N E G :~
78
27.44 + 2.1
15.4
Cattle w i t h no pathological c o n d i t i o n s ~ T T P O S :~
87
54.14 + 3.2
73.6
Cattle w i t h no pathological c o n d i t i o n s N T T :~
66
35.2 _+ 1.8
42.4
Group
1. 2. 3. 4. 5. 6.
No.
Cattle injected w i t h o t h e r mycobacteria'-' N T T :~
~%POS = % of positive control s e r u m No. 1495. SSamples t a k e n 14 days after first inoculation, M. bovis a n d control a n i m a l s excluded. :~NTT = never tuberculin tested; T T N E G = tuberculin-tested negative; T T P O S = tuberculin-tested positive. ~Mean scores significantly different ( P < 0.01 ) to G r o u p 6 m e a n . '-'Histological e x a m i n a t i o n a n d mycobacteriological culture negative.
57 TABLE2 Distribution of M. bovis E L I S A positive scores found in cattle from which other mycobacteria were isolated Mycobacteria isolated
Number of times Organisms isolated
M. m a r i n u m M. phlei M. flavescens MAIS unknown serovar M. terrae, MAIS unknown serovar M. terrae MAIS serovar 3 MAIS serovar 8 M. terrae, Gp. IV M. terrae, M. flavescens M. flavescens, Gp. IV M. flavescens, MAIS serovar 15 M. simmiae, MAIS serovar 10
1 1 6 8 2 12 1 1 1 2 1 1 1
Total numbers
38
ELISA positive1
1 1 4 4 1 4 0 0 0 0 0 0 0 15 (39.5)
Walues greater t h a n 35% of a known positive serum (Auer, 1987 ).
Cattle with other mycobacteria isolated (Group 2) Fifteen of 38 animals from which other mycobacteria were isolated (39.5 % ), including 4 of 9 animals with suspect M. boris lesions at slaughter (later cultured negative for M. boris), were positive in the M. boris E L I S A (Table 2). The mean antibody titre was 36.6% (Table 1). There was no significant difference between the mean of Group 2 cattle and the mean of Group 6 cattle (no pathological conditions, never tuberculin tested).
Cattle with other pathological conditions (Group 3) The mean E L I S A score of Group 3 was 44.0% with 56.4% of scores being positive (Table 1 ). The mean of this group was significantly higher ( P < 0.01 ) than the mean of Group 6 (no pathological conditions, never tuberculin tested). The distribution of negative and false positive results in animals with lesions such as club forming foreign body granulomas was similar to that found in animals from which Rhodococcus was isolated (Table 3). The 2 animals from which Streptomyces/Nocardia were diagnosed were negative while sera from animals with granulomatous lesions of undetermined cause gave 9 out of 12 positive E L I S A results.
58 TABLE 3 Positive M. boris E L I S A scores found in cattle with other pathological conditions, culture negative for mycobacteria and never t u b e r c u l i n - tested Pathology ~
N u m b e r of animals Total No.
Parasitic granulomas Granulomatous lesion - u n d e t e r m i n e d cause Squamous cell carcinoma Rhodococcus equi granulomas Club-forming granulomas (actinobacilloses/ actinomycoses ) Foreign body granuloma Fibrosis of lymph node Hydatid cyst Other neoplasms Streptomyces/Nocardia granuloma Purulent lymphadenitis Abscess Total numbers
E L I S A positive ~
2 12 4 8
2 9 3 5
28 9 2 5 4 2 1
16 5 1 2 1 0 0
1
0
78
44 (56.4%)
~Histological diagnosis of tissues with gross lesions s u b m i t t e d for examination for tuberculosis. ~Values greater t h a n 35% of a known positive serum (Auer, 1987).
Cattle with no lesions and different tuberculin testing histories (Groups 4, 5 and 6) The mean antibody E L I S A scores (Table 1 ) were 27.4, 54.1 and 35.2%, respectively, for cattle tuberculin-tested negative (Group 4), tuberculin-tested positive (Group 5 ) and never tuberculin tested (Group 6). The corresponding percentages of positive scores in each group were 15.4, 73.6 and 42.4%. There were significant differences between each of Groups 4, 5 and 6 ( P < 0.01 ). DISCUSSION
In this investigation, nonspecific antibodies to M. bovis were frequently found in cattle uninfected with this bacteria. The results indicated that exposure to other mycobacteria and the presence of other pathological conditions would increase this nonspecific response. In contrast, the prevalence of false positive results was significantly lower in cattle which had been tuberculin tested several times in eradication surveys. The variation seen in this study indicates the importance of using large numbers of randomly-selected negative controls in the evaluation of serological tests for bovine tuberculosis. The development of an accurate serological test for tuberculosis has been
59 pursued for many decades. With the advent of highly-sensitive and practical technology such as the ELISA, there has been renewed interest in this area. ELISAs to detect M. bovis antibodies in pigs and cattle (Thoen et al., 1981; Auer, 1987) have been developed but lack specificity even when a purified protein derivative ( P P D ) is used as the antigen (Thoen et al., 1981 ). A whole cell sonicate was used as the antigen in the E L I S A in the present study in order to detect the full extent of antibody responses to shared epitopes. If cattle infected with M. boris react predominantly to shared epitopes as has been reported to occur in human tuberculosis and leprosy (Stanford, 1983 ), the search for a unique antigen of M. boris to improve specificity in serological tests may become less relevant. It would then become more important to target and identify those factors which are most implicated in false positive reactions. Harboe and Nagai (1984) have reported an antigen M P B 7 0 in high concentrations in M. bovis BCG and in lesser concentrations in M. tuberculosis H 3 7 R U and in Nocardia asteroides. More specific antigens such as these could be used to determine their potential in diagnostic tests. Exposure to other mycobacteria was found in this study to result in markedly increased M. boris E L I S A scores but at different times post-inoculation in individual animals. These responses may not reflect the natural situation due to the artificial nature of the exposure. Nevertheless, the antigens of the other mycobacteria evoked M. bovis antibody responses as measured by the ELISA. However, in animals from which other mycobacteria were isolated, only 39.5% had false positive M. bovis E L I S A scores compared to 73.6% of tuberculin test positive, culture negative cattle with no pathological conditions. The lack of response by 60.5% of Group 2 could have been due to the cattle actually being unexposed to other mycobacteria, the environmental contamination during collection of the lymph node causing the positive culture results. Against this, all cattle in the group were false positive reactors to the tuberculin test. A possible explanation is that the antigen in the tuberculin injection mopped up circulating antibody and decreased titres. The high percentage of false positive results (56.4%) in cattle with a variety of pathological conditions (Table 3) indicates the possible existence of other nonspecific stimuli. The significantly higher mean E L I S A score of this group compared to the group with no pathological conditions and never tuberculin tested (Group 6) supports this idea. Minden et al. (1972) showed that sera from rabbits immunized with M. boris strain BCG bound not only radiolabelled homologous test antigens b u t also radiolabelled antigens from many other unrelated bacteria. Alternatively, high levels of anti-mycobacterial antibodies occurring in other infectious diseases may be due to a nonspecific elevation of antibody synthesis as occurs in some chronic diseases such as malaria (Grange, 1984 ). Significant differences were found between the groups of cattle with no pathological conditions and different tuberculin testing histories. The group
60 which had been tuberculin-tested negative 56 days prior to slaughter showed the lowest mean antibody levels while those cattle tuberculin-tested positive 3-7 days prior to blood sampling showed the highest scores (P<0.01). One explanation for this could be the culling of false positive tuberculin reactor cattle in repeated herd tests. The tuberculin test negative group had a long history of tuberculin testing and this may have decreased the percentage of cattle exposed to agents mimicking M. bovis immune responses. A beneficial consequence, if ELISA was applied for diagnosis, might be greatly improved specificity in extensively-tested herds. Another factor boosting the mean score in tuberculin-tested positive animals with no pathological conditions (Group 5) could be the immunogenic effect of the tuberculin. Previous work with infected animals tuberculin-tested positive 3-7 days earlier showed higher mean ELISA scores than infected animals never tuberculin tested (Auer, 1987). However, these results are at variance with the tuberculin-positive cattle with other mycobacteria isolated (Group 2), which had unexpectedly low antibody scores. A fall in antibody titre occurred in active human tuberculosis patients 2-4 days after a tuberculin test. Such a drop did not occur in patients with inactive tuberculosis or with other diseases (Algeorge and Stoian, 1973). In cattle, it has been shown that tuberculin does not increase antibody titres from 34 days to 1 year post-inoculation (Auer, 1987). If, however, there is a shortterm fluctuation in titres, the application of ELISAs to recently tuberculintested cattle should be closely examined. In the development ofa serodiagnostic test for M. bovis, accuracy will depend upon the prevalences in sera of specific and nonspecific antibodies to M. bovis. This paper reports on various factors influencing these prevalences. By studying the factors affecting nonspecific responses, strategies for developing and applying accurate tests based on this knowledge will be of benefit in eradicating bovine tuberculosis. ACKNOWLEDGEMENTS We thank T. Veerman, G. Smith, J. Pike and K. Mason of the Animal Research Institute (ARI), Yeerongpilly, for the preparation of M. bovis antigen and bacteriological diagnosis of lesions, G. Storie and R. Pierce (ARI), for histological diagnosis of lesions and G. Tudor (ARI) and staff for care of the experimental cattle. We also thank Dr. G. Simmons, J. Walthall and staff of Veterinary Services Branch, Queensland Department of Primary Industries ( QDPI ) and J. Anderson and staff, Veterinary Public Health Branch (QDPI), for their support in the establishment of a serum bank of samples from infected and non-infected animals. Funds for this work were provided by the Australian National Brucellosis and Tuberculosis Eradication Scheme.
61 REFERENCES Auer, L.A., 1987. Assessment of an enzyme linked immunosorbent assay for the detection of cattle infected with Mycobacterium bovis. Aust. Vet. J., 64:172 176. Algeorge, G. and Stoian, M., 1973. Serological diagnosis of tuberculosis by fluorescent antibody test and kaolin agglutination test. Probl. Tuberk., 12:99 113. Corner, L.A. and Pearson, C.W., 1978. Pathogenicity for cattle of atypical mycobacteria isolated from feral pigs and cattle and the correlation of lesions with tuberculin sensitivity. Aust. Vet. J., 54: 280-286. Gallagher, J. and Horwill, D.M., 1977. A selective oleic acid albumin agar medium for the cultivation of Mycobacterium bovis. J. Hyg., Camb., 79:155 160. Grange, J.M., 1984. The humoral immune response in tuberculosis: its nature, biological role and diagnostic usefulness. Adv. Tuberc. Res., 21: 1-78. Harboe, M. and Nagai, S., 1984. MPB70, a unique antigen of Mycobacterium boris BCG. Am. Rev. Resp. Dis., 129: 444-452. Ketterer, P.J., Rogers, R.J. and Donald, B., 1981. Pathology and tuberculin sensitivity in cattle inoculated with Myobacterium avium complex serotypes 6, 14 and 18. Aust. Vet. J., 57: 61-65. Lenzini, L., Rottoli, P. and Rottoli, L., 1977. The spectrum of human tuberculosis. Clin. Exp. hnmunol., 27: 230-237. Lepper, A.W.D. and Corner, L.A., 1983. Serological tests. In: C. Ratledge and J. Stanford (Editors), The Biology of the Mycobacteria, Vol. 2. Academic Press, Sydney, pp. 490-495. Minden, P., McClatchy, J.K. and Farr, R.S., 1972. Shared antigens between heterologous bacterial species. Infect. Immun., 6:574 582. Reznikov, M. and Leggo, J.H., 1972. Modification of Schaefer's procedure for serotyping of organisms of the Mycobacterium avium-M, intracellulare and M. scrofulaceum complex. Appl. Microbiol., 23: 819-823. Stanfl)rd, J.L., 1983. Immunologically important constituents of mycobacteria: antigens. In: C. Ratledge and J. Stanf'ord (Editors), The Biology of the Mycobacteria, Vol. 2. Academic Press, Sydney, pp. 85-127. Thoen, C.O., Ivey, C.R., Hines, E.M. and Harrington, Jr. R., 1981. Application of an enzyme linked immunosorbent assay fl)r detecting antibodies in sera of feral swine naturally exposed to Mycobacteriura bovis. Proceedings of the 85th Annual Meeting of the U.S. Animal Health Association, pp. 500 504.