J. Vet. Med. B 37, 581-586 (1990) 0 1990 Paul Parey Scientific Publishers, Berlin and Hamburg ISSN 093 1 - 1793
Institut fur Veterinarmedizin des Bundesgesundheitsamtes
Serological Reactions of Leptospirosis-Positive (MAR and CFT) Bovine Sera in ELISA C. STAAK,M. MEKAPRATEEP, U. KAMPEand A. SCHONBERG Address of authors: Institut fur Veterinarmedizin des Bundesgesundheitsamtes, Robert-von-Ostertag-Institut, Postfach 33 00 13, D-1000 Berlin 33
With 5 tables
(Received for publication January 3, 1990)
Summary The collection of test sera for measuring ELISA results was composed of bovine sera with MAT 1 : 200 in the leptospirosis MAT and of 3 1 : 5 in the C F T together with sera from a titres of serologically negative and clinically non-suspicious cattle herd. To establish cut-off ODs, the geometric mean net-extinction of the negative serum collection plus 1, 2, and 3 standard deviations were calculated. By comparison of 3 different conjugates from rabbits, it was demonstrated that results from anti-total bovine Ig were superior to anti-IgG and anti-IgM conjugates. Considerations regarding sensitivity and specificity led to the recommendation to use a test serum dilution of 1 : 160, to apply anti-total bovine Ig conjugates, and to establish the cut-off OD at the geometric mean netextinction of negative sera plus 3 standard deviations. Under such conditions, agreement between leptospirosis MAT/CFT positivity on the one side and ELISA positivity on the other was reached in 74 %. This recommendation is made for cross-sectional studies but not for examinations of clinically suspicious cattle herds.
Introduction The agglutination test is the standard technique for serological diagnosis of leptospirosis. The tube agglutination test has been widely adapted to microtitre system (MAT), so that the reading of results can be performed directly in the wells of a microtitre plate by a specially equipped darkfield microscope ( D G H M Standards, 1985). Other modifications included the addition of formalin to the diluent, and the screening of test sera by pooled antigens with subsequent identification of serovar specific antibodies by monospecific antigens (HATHAWAY et al., 1982). Sensitive agglutination tests for leptospirosis require freshly cultured organisms. Thus, serological examination for leptospirosis has been restricted to laboratories which are in a position to keep a collection of leptospira strains. Inactivated antigens have been used for the slide agglutination test (BREMand WEBER,1984), complement fixation test et al., 1985; (HODGESand WEDELL,1977), and ELISA (ADLERet al., 1982; TERPSTRA COUSINS et a]., 1985). Because of low antigen consumption, the application of fractionated antigens, and the possibility to differentiate between IgG and IgM antibody responses, ELISA may become the most suitable test system for non-specialized routine laboratories which d o not keep a U.S. Copyright Clearance Center Codc Statemeor: 0931 - 1793/90/3708-0581$02.50/0
582
STAAK,MEKAPRATEEP, KAMPEand SCHONBERG
battery of leptospira strains. This applies for European laboratories but much more to laboratories in the developing countries. During a field survey on the incidence of leptospirosis in Germany (SCHONBEKG et al., 1987), sera with positive reactions to MAT and C F T have been re-tested in the ELISA. The results of such testing are presented in the following.
Material and Methods Bovine sera collected during a leucosis eradication programme have been tested for leptospira antibodies by microscopic agglutination test (MAT) and complement fixation test (CFT). Negative sera were taken from the experimental farm of the Federal Health Office. For checkerboard titration of antigens, sera from artificially infected rabbits were used. All sera from the field survey were stored at - 80 "C. The MAT was performed according to the D G H M standard technique (DGHM, Standard, 1985). The following leptospira serovars have been used as antigens: autumnalis, austrulzs, batawiue, canicola, grippotyphosu, copenhageni, pomona, hardjo, saxkoebing, tarussovi. A positive reaction up to 1 : 200 final serum dilution was regarded as a doubtful, and such of 1 : 400 and more as a positive result. The CFT with serial serum double dilutions of 1 : 5 to 1 : 160 was conducted according to the standard technique for brucellosis (ALTONet al., 1975). For the production of C F and ELISA antigens, 500 ml of leptospira cultures were centrifuged, the pellet was suspended in normal saline, centrifuged again, re-suspended in 5 ml of saline and heated to 100 "C for 30 min. This was followed by agitation with Ballotini No. 12 in the cold for 72 h. The extract was sonicated and after centrifugation the supernatant was collected as antigen. Antigens were checkerboard titrated against homologous and heterologous rabbit hyperimmune sera. For the general diagnosis of leptospirosis regardless of the serovar involved in infection, antigens were mixed according to their individual antigenic titres (pooled antigen). The pooled antigen used throughout this investigation contained the following serovars: L. Sao puulo, hardjo, grippotyphosa, tarussovi,pomonu, and copenhageni. A serum-CFT-titre of 1 : 5 and more was taken as a positive result. The ELISA technique used in this investigation has been described in detail by WALTHERand SANITZ(1979). The following peroxidase conjugates were used: 1 . Rabbit-anti-bovine IgG (H + L), 2. Rabbit-anti-bovine IgM (Fc), both from Biogenzia Lemania (Nordic, Tilburg, The Netherlands), 3. Rabbit-anti-bovine total Ig from Dakopatts (Denmark). Reactions were recorded by a photometer (Multiskan, Flow). Different cut-off OD's were calculated by the addition of standard deviations (SD) to the means of OD of negative sera. OD's were represented by net-extinctions, i. e. the differences in OD between antigen containing and antigen free wells for all sera set up in parallel.
Results The cut-off OD's for the three different conjugates were established by testing a herd of cattle with negative results in the MAT and CFT. One hundred and seventy-three sera were tested in the ELISA at dilutions from 1 : 20 to 1 : 160. Standard deviations (SD) for serum dilutions of 1 : 40, 1 : 80, and 1 : 160 were calculated for all conjugates used in the test. Cut-off extinctions were established by the addition of 1,2, and 3 SD to the mean netextinction of all sera tested. Results are presented in T a b l e 2 Sera from the field survey (see Material and Methods) with positive reactions in the MAT and/or CFT were classified as follows: Table 1. Classification of leptospirosis positive field sera according to results in the MAT and CF:T MAT
CFT
total (%)
111
1 : 400/800 over 1 : 800 1 : 200 neg.
IV
pos.'"
pos. pos. pos. pos. neg.
39 19 15 28 88
Group Ia Ib
I1
(20.6) (10.1) (7.9) (14.8) (46.6)
189 (100) Including 1 : 200 reactions.
583
Serological Reactions of Leptospirosis-Positive Bovine Sera
Table 2. ELISA (3 different conjugates) results from bovine sera negative for leptospirosis in the MAT and CFT: Cut-off OD's for 1, 2, and 3 SD and percentage of false-positive reactors An ti-bovine PO-conjugate directed against
SD
1:40 cut-off 'Xi react.
Serum dilutions 1:xo cut-off % react.
cut-off
1 : 160 % react.
total Ig
1 2 3
0.470 0.637 0.803
18.7 1.9 0
0.351 0.484 0.617
18.7 1.9 0
0.253 0.359 0.464
15.5 1.9 0
IgG
1 2 3
0.201 0.255 0.309
14.5 5.8 0.6
0.159 0.223 0.287
13.9 1.7 0
0.116 0.164 0.212
9.8 2.3 0
IgM
1 2 3
0.310 0.422 0.534
19.1 0.6 0
0.256 0.347 0.438
12.7 3.5 0
0.210 0.288 0.366
9.8 2.9 0
The results obtained when applying the different cut-offs from Table 2 for M A T K F T positive sera, are presented in Table 3 for serum dilutions of 1 : 40, 1 : 80, and 1 : 160. Figures are given as percentage of M A T K F T positive reactors. The postulate, not to produce false-positives was best fulfilled by a cut-off OD as calculated from the mean OD of MAT and/or CFT negative sera plus 3 SD (Table 2). This criterion has been applied to the collection of positive sera and the coincidence of MAT and/or C F T results with ELISA results has been summarized in Table 4.
Table 3. ELISA results (%) from bovine sera positive for leptospirosis in the MAT and CFT: Total of ELISA reactor according to 1, 2, and 3 SD added to the mean OD of a negative population Group
SI)
Anti-Ig conjugate Serum dilution 1:40 1:XO 1:160
Anti-IgG conjugate Serum dilution 1:40 1 : 8 0 1:160
Anti-IgM conjugate Serum dilution 1:40 1 : 8 0 1:160
~~
Ia
1 2 3
100 97 95
97 97 97
100 97 97
97 92 87
97 82 72
95 82 72
92 41 10
95 46 15
92 28 15
Ib
1 2 3
100 100 100
100 100 95
100 100 95
84 79 79
95 79 74
95 79 68
100 42 0
89 42 0
100 42 0
I1
1 2 3
100 93 93
100 93 93
93 93 80
87 87 80
87 80 80
87 80 80
73 13 0
87 27 0
67 13
111
1 2 3
89 54 32
79 50 32
79 46 18
86 75 54
75 50 29
75 46 25
68 25 0
79 36 0
68 29 4
IV
1 2 3
98 86 74
98 89 74
98 89 76
75 66 55
81 63 47
77 64 48
92 64 20
93 64 30
92 67 28
0
STAAK,MEKAPKAT~EP, KAMPEand SCHONBERG
584
Table 4. Leptospirosis-positiveELISA reactions (%) (Cut-off: Mean negative OD plus 3 SD in MAT and/or CFT positive bovine sera) MAT
CFT
Total
t
+
96
96
93
84
74
73
5
8
8
t
-
74
74
76
55
47
48
20
30
28
-
+
73 (38.6) 88 (46.6) 28 (14.8)
32
32
19
54
29
25
0
0
0
189
76
76
74
66
55
54
12
17
17
Total
Anti-Ig conjugate Serum dilution 1:40 1:XO 1:160
Anti-IgG conjugate Serum dilution 1:40 1:XO 1:160
Anti-IgM conjugate Serum dilution 1:40 1 : 8 0 1:160
Sequential sera from a heifer artificially infected with L. hardjo and monitored bacteriologically (HAHN,1987) were tested in MAT, CFT, and ELISA using rabbit-antibovine Ig PO as conjugate (Table 5 ) . Table 5. Serological reactions in a L. hardjo infected heifer Day p. i. of sampling shedding of L. hardjo" MAT CFT ELISA'"s;"lSD 2s 3s
4
_ _ _ + _
10
11
14
17
21
24
28
31
_ _
_ _
t
t
t
+
-
t
-
_
t
t
t
+
_ +
+
_
-
t
-
_
_
_
_
+
+
+
-
+
35
t
t
+
+
+
-
t
+
+
+
_
+
+
-
38
t i
+ t t -
42
t
45
-
+
+
-
+
t + t t t -
49 51".'>
t
+ t +
+
-
+
+
t t
-
t t t t
'i Positive L. hardjo cultures from urine. ::+ Animal slaughtered, cultures from kidneys. '+'s':. Rab. antibovine Ig-PO conjugate.
Discussion The interpretation of MAT and CFT as t o leptospirosis results is still under discussion, mainly, because differences between infection or non-infection and serological results differ to a larger degree as compared to many other diseases. The majority of authors (LEWIS,1978; HATHAWAY and LITTLE,1983; WHITE et al., 1982; COUSINS et al., 1985) claim a 50 YO agglutination at a serum dilution of 2 1 : 100 to be a positive result. ELLISet al. (1982) regarded serum titres of 1 : 10 in L. hardjo infected cattle as an indication of infection. CORDES et al. (1982) recorded results as from 2 1 : 200 serum dilutions with 50 YO agglutination as positive. Bacteriological investigation of kidneys from slaughtered cattle and serological testing of matching serum samples (MAT, positive 2 1 : 100) revealed the following (WHITEet al., 1982): Serology Bacteriology %,
+ -
+
-
+ +
19
-
52 28
-
1
Serological Reactions of Leptospirosis-PositiveBovine Sera
585
In another investigation (HATHAWAY and LITTLE, 1983), agreement between the detection of leptospires in urine samples and serological positivity was found in 14 cases while in 8 cases serology produced positive and bacteriology negative results. The answer to the question of serological reactions in confirmed non-infected cattle populations seems to be even more difficult, and relevant information has not been available in literature. From the absence of clinical signs such as abortions, still births or sudden drop in milk yield together with negative or only low level antibody reactions it can only be concluded that leptospirosis does not exist in a given herd. Therefore, it may be justified to apply two different systems to the classification of serological anti-leptospirosis reactions: 1. If a herd is suspected of being infected with leptospirosis, repeated serum samples should be taken and even low antibody reactions are et to be recorded in addition to attempts to isolate the causative organism (ELLINGHAUSEN al., 1977). 2. For cross-sectional studies more importance should be attributed to specificity rather than sensitivity (PRITCHARD, 1984) by accepting only higher titre levels as a positive result. According to ELLISet al. (1982), a positive titre of 1 : 10 in confirmed L. hurdjo infections of cattle produced a sensitivity of 0.67 and a specificity of 0.86. When the classification of serological reactions was based on titre levels of 3 I : 100, the sensitivity dropped to 0.41 but the specificity was I . From this, it is concluded that serological results referring to leptospirosis should be interpreted on a herd basis and not for individual animals. The collection of positive and negative sera against which ELISA results have been compared were obtained from a cross-sectional study and from a clinically non-suspicious farm. A system of low sensitivity but high specificity has been applied for classification by negative, doubtful and positive reactions. CFT results were negative in 47 % of all positive reactions encountered. However, this test detected another 15 % positives in MAT negative reactors. This disagreement between the results of the MAT and C F T had been observed by other authors (BATZAand ZETTL,1985), and was reflected as well in the case of the L. hardjo infected heifer presented in this paper: Agreement between MAT and CFT: 7x, MAT+/CFT-: 2x, MAT-/CFT+: 2 ~ . The comparison between the 3 different conjugates has revealed that anti-bovine Ig conjugates are t o be preferred to anti-bovine IgG or IgM conjugates. This finding in bovine infections is in contrast to those in human leptospirosis where the infection is best detected by anti-IgM conjugates (TERPSTRA et al., 1985). For establishing ELISA cut-off ODs, the mean OD plus standard deviations from the leptospirosis negative cattle herd were and GARRET (1982) used absorbance calculated. This is an accepted procedure: THIERMANN readings which were at least 2 times higher than the average value of negative reference sera for cut-off. ADLERet al. (1982) applied mean OD readings from negative animals plus 3 standard deviations as a baseline for positive results. This system has been applied to other diseases as well (FORSCHNER, personal communication). References ADLER,B.S., D.V. COUSINS, S.FAINE,and G.M. ROBERTSON, 1982: Bovine IgM and IgG response to Leptospiru interrogans seTovar hardjo as measured by enzyme immunoassay. Vet. Microbiol. 7, 577-585. ALTON,G.G., L.M. JONES, R.D. ANGUS, and J.M. VERGER, 1975: Techniques for the brucellosis laboratory. INRA, Paris 1988, ISBN 2-7380-0042-8. BATZA,H. J., und K. ZETTL,1985: Nachweis von Leptospiren-Antikorpern in Schafseren rnittels verschiedener serologischerMethoden. Proc. 16th Congress of German Veterinary Ass. (DVG), Bad Nauheirn, 17.--20.4.1985. BREM,S., und A. WEBER,1984: Zum serologischen Nacliweis van Leptospiren-Infektionen bei Rindern und Pferden mit Hilfe eines im Handel erhaltlichen Objekttrageragglutinationstestes. Der praktische Tierarzt 10, 835-841. K. G. TOWNSEND, S. F. LEWIS,and J.T. S. HOLLAND, 1982: LeptoCORDES, D.O., M.E. CARTER, spirosis: I. Clinical investigation of the infection of dairy cattle in the Waikato district of New Zealand. N.Z. Vet. J. 30, 122-124.
586
STAAK,MEKAPKATEEP, KAMPEand SCHONBERG
COUSINS,D. V., G. M. ROBERTSON, and L. HUSTAS, 1985: The use of enzyme-linked immunosorbent assay (ELISA) to detect the IgM and IgG antibody response to Leptospira interroguns serovars hardjo, pomona, and turussovi in cattle. Vet. Microbiol. 10, 439-450. DGHM-Verfahrensrichtlinien, Mai 1985: Diagnostik der Leptospirosen. Gustav Fischer Verlag, Stuttgart, 1985. ELLINGHAUSEN, H . C., 11. L. DEYOE,and R. M. NERWIG, 1977: Leptospirosis in perspective. Proc. An. Meeting US An. H. Ass. 81, 161-182. ELLIS, W. A,, J. J. O’BRIEN,S. D. NEIIJ., and J. HANNA, 1982: Bovine leptospirosis: Serological findings in aborting cows. Vet. Record 110, 178-180. HAHN,B., 1987: Ein Beitrag zur Verbesserung des kulturellen Nachweises von Leptospiren durch Selektivnahrbtiden. Vet. med. Dissertation, FU Berlin, Journal Nr. 1354. S.C., T. W. A. LITTLE, and A. E. STEVENS,1982: Serological survey of leptospiral HATHAWAY, antibodies in sheep from England and Wales. Vet. Record 110, 99-101. HATHAWAY, S. C., and T. W. A. LITTLE,1983: Epidemiological study of Leptospira hurdjo infection in second calf dairy cows. Vet. Record 112, 215-218. HODGES,R. T., and W. WEDELL,1977: Adaptation of a complement fixation test for large scale serological diagnosis of bovine leptospirosis. N . Z. Vet. J. 25, 261 -262. LEWIS,J., 1978: The routine application of a microtechnique for the demonstration of leptospiral antibodies. J. Med. Microbiol. 11, 355-358. D. G., 1984: Some examples of the use of epidemiological techniques in the study of PRITCHARD, leptospirosis. Proc. C E C Seminary: The present state of leptospirosis diagnosis and control. Belfast, Oct. 10- 14, 1984. Martinus Nijhoff Publ., Dordrecht. S C H ~ N B X KA., G ,C. STAAKund U. KAMPE,1987: Leptospirose in der Bundesrepublik Deutschland. J. Vet. med. B 34, 98-108. TERPSTRA, W.J., G. S. LIGHTHART, and G. J. SCHOONE, 1985: ELISA for the detection of specific IgM and IgG in human leptospirosis. J. Gen. Microbiol. 131, 377-385. THIERMANN, A. B., and L. A. GARRET,1983: Enzyme-linked immunosorbent assay for the detection of antibodies to Leptospiru interrogarrs serovar hardjo and pomona in cattle. Am. J. Vet. Res. 44, 884 -887. WALTER,M., und W. SANITZ,1979: Untersuchungen zur serologischen Feststellung der Zystizerkose des Rindes mit Hilfe des Enzyme-linked immunosorbent assay (ELISA). BMTW 92, 131 -135. WHITE, F.H., K.R. SULZER,and R. W . ENGEL,1982: Isolations of Leptospiru intewogans serovars hardjo, balcanzca, and pornonu from cattle at slaughter. Am. J. Vet. Res. 43 (7), 1172-1173.
Institut fur Veterinarmedizin des Bundesgesundheitsamtes
Serological Reactions of Leptospirosis-Positive (MAR and CFT) Bovine Sera in ELISA C. STAAK,M. MEKAPRATEEP, U. KAMPEand A. SCHONBERG Address of authors: Institut fur Veterinarmedizin des Bundesgesundheitsamtes, Robert-von-Ostertag-Institut, Postfach 33 00 13, D-1000 Berlin 33
With 5 tables
(Received for publication January 3, 1990)
Summary The collection of test sera for measuring ELISA results was composed of bovine sera with MAT 1 : 200 in the leptospirosis MAT and of 3 1 : 5 in the C F T together with sera from a titres of serologically negative and clinically non-suspicious cattle herd. To establish cut-off ODs, the geometric mean net-extinction of the negative serum collection plus 1, 2, and 3 standard deviations were calculated. By comparison of 3 different conjugates from rabbits, it was demonstrated that results from anti-total bovine Ig were superior to anti-IgG and anti-IgM conjugates. Considerations regarding sensitivity and specificity led to the recommendation to use a test serum dilution of 1 : 160, to apply anti-total bovine Ig conjugates, and to establish the cut-off OD at the geometric mean netextinction of negative sera plus 3 standard deviations. Under such conditions, agreement between leptospirosis MAT/CFT positivity on the one side and ELISA positivity on the other was reached in 74 %. This recommendation is made for cross-sectional studies but not for examinations of clinically suspicious cattle herds.
Introduction The agglutination test is the standard technique for serological diagnosis of leptospirosis. The tube agglutination test has been widely adapted to microtitre system (MAT), so that the reading of results can be performed directly in the wells of a microtitre plate by a specially equipped darkfield microscope ( D G H M Standards, 1985). Other modifications included the addition of formalin to the diluent, and the screening of test sera by pooled antigens with subsequent identification of serovar specific antibodies by monospecific antigens (HATHAWAY et al., 1982). Sensitive agglutination tests for leptospirosis require freshly cultured organisms. Thus, serological examination for leptospirosis has been restricted to laboratories which are in a position to keep a collection of leptospira strains. Inactivated antigens have been used for the slide agglutination test (BREMand WEBER,1984), complement fixation test et al., 1985; (HODGESand WEDELL,1977), and ELISA (ADLERet al., 1982; TERPSTRA COUSINS et a]., 1985). Because of low antigen consumption, the application of fractionated antigens, and the possibility to differentiate between IgG and IgM antibody responses, ELISA may become the most suitable test system for non-specialized routine laboratories which d o not keep a U.S. Copyright Clearance Center Codc Statemeor: 0931 - 1793/90/3708-0581$02.50/0
582
STAAK,MEKAPRATEEP, KAMPEand SCHONBERG
battery of leptospira strains. This applies for European laboratories but much more to laboratories in the developing countries. During a field survey on the incidence of leptospirosis in Germany (SCHONBEKG et al., 1987), sera with positive reactions to MAT and C F T have been re-tested in the ELISA. The results of such testing are presented in the following.
Material and Methods Bovine sera collected during a leucosis eradication programme have been tested for leptospira antibodies by microscopic agglutination test (MAT) and complement fixation test (CFT). Negative sera were taken from the experimental farm of the Federal Health Office. For checkerboard titration of antigens, sera from artificially infected rabbits were used. All sera from the field survey were stored at - 80 "C. The MAT was performed according to the D G H M standard technique (DGHM, Standard, 1985). The following leptospira serovars have been used as antigens: autumnalis, austrulzs, batawiue, canicola, grippotyphosu, copenhageni, pomona, hardjo, saxkoebing, tarussovi. A positive reaction up to 1 : 200 final serum dilution was regarded as a doubtful, and such of 1 : 400 and more as a positive result. The CFT with serial serum double dilutions of 1 : 5 to 1 : 160 was conducted according to the standard technique for brucellosis (ALTONet al., 1975). For the production of C F and ELISA antigens, 500 ml of leptospira cultures were centrifuged, the pellet was suspended in normal saline, centrifuged again, re-suspended in 5 ml of saline and heated to 100 "C for 30 min. This was followed by agitation with Ballotini No. 12 in the cold for 72 h. The extract was sonicated and after centrifugation the supernatant was collected as antigen. Antigens were checkerboard titrated against homologous and heterologous rabbit hyperimmune sera. For the general diagnosis of leptospirosis regardless of the serovar involved in infection, antigens were mixed according to their individual antigenic titres (pooled antigen). The pooled antigen used throughout this investigation contained the following serovars: L. Sao puulo, hardjo, grippotyphosa, tarussovi,pomonu, and copenhageni. A serum-CFT-titre of 1 : 5 and more was taken as a positive result. The ELISA technique used in this investigation has been described in detail by WALTHERand SANITZ(1979). The following peroxidase conjugates were used: 1 . Rabbit-anti-bovine IgG (H + L), 2. Rabbit-anti-bovine IgM (Fc), both from Biogenzia Lemania (Nordic, Tilburg, The Netherlands), 3. Rabbit-anti-bovine total Ig from Dakopatts (Denmark). Reactions were recorded by a photometer (Multiskan, Flow). Different cut-off OD's were calculated by the addition of standard deviations (SD) to the means of OD of negative sera. OD's were represented by net-extinctions, i. e. the differences in OD between antigen containing and antigen free wells for all sera set up in parallel.
Results The cut-off OD's for the three different conjugates were established by testing a herd of cattle with negative results in the MAT and CFT. One hundred and seventy-three sera were tested in the ELISA at dilutions from 1 : 20 to 1 : 160. Standard deviations (SD) for serum dilutions of 1 : 40, 1 : 80, and 1 : 160 were calculated for all conjugates used in the test. Cut-off extinctions were established by the addition of 1,2, and 3 SD to the mean netextinction of all sera tested. Results are presented in T a b l e 2 Sera from the field survey (see Material and Methods) with positive reactions in the MAT and/or CFT were classified as follows: Table 1. Classification of leptospirosis positive field sera according to results in the MAT and CF:T MAT
CFT
total (%)
111
1 : 400/800 over 1 : 800 1 : 200 neg.
IV
pos.'"
pos. pos. pos. pos. neg.
39 19 15 28 88
Group Ia Ib
I1
(20.6) (10.1) (7.9) (14.8) (46.6)
189 (100) Including 1 : 200 reactions.
583
Serological Reactions of Leptospirosis-Positive Bovine Sera
Table 2. ELISA (3 different conjugates) results from bovine sera negative for leptospirosis in the MAT and CFT: Cut-off OD's for 1, 2, and 3 SD and percentage of false-positive reactors An ti-bovine PO-conjugate directed against
SD
1:40 cut-off 'Xi react.
Serum dilutions 1:xo cut-off % react.
cut-off
1 : 160 % react.
total Ig
1 2 3
0.470 0.637 0.803
18.7 1.9 0
0.351 0.484 0.617
18.7 1.9 0
0.253 0.359 0.464
15.5 1.9 0
IgG
1 2 3
0.201 0.255 0.309
14.5 5.8 0.6
0.159 0.223 0.287
13.9 1.7 0
0.116 0.164 0.212
9.8 2.3 0
IgM
1 2 3
0.310 0.422 0.534
19.1 0.6 0
0.256 0.347 0.438
12.7 3.5 0
0.210 0.288 0.366
9.8 2.9 0
The results obtained when applying the different cut-offs from Table 2 for M A T K F T positive sera, are presented in Table 3 for serum dilutions of 1 : 40, 1 : 80, and 1 : 160. Figures are given as percentage of M A T K F T positive reactors. The postulate, not to produce false-positives was best fulfilled by a cut-off OD as calculated from the mean OD of MAT and/or CFT negative sera plus 3 SD (Table 2). This criterion has been applied to the collection of positive sera and the coincidence of MAT and/or C F T results with ELISA results has been summarized in Table 4.
Table 3. ELISA results (%) from bovine sera positive for leptospirosis in the MAT and CFT: Total of ELISA reactor according to 1, 2, and 3 SD added to the mean OD of a negative population Group
SI)
Anti-Ig conjugate Serum dilution 1:40 1:XO 1:160
Anti-IgG conjugate Serum dilution 1:40 1 : 8 0 1:160
Anti-IgM conjugate Serum dilution 1:40 1 : 8 0 1:160
~~
Ia
1 2 3
100 97 95
97 97 97
100 97 97
97 92 87
97 82 72
95 82 72
92 41 10
95 46 15
92 28 15
Ib
1 2 3
100 100 100
100 100 95
100 100 95
84 79 79
95 79 74
95 79 68
100 42 0
89 42 0
100 42 0
I1
1 2 3
100 93 93
100 93 93
93 93 80
87 87 80
87 80 80
87 80 80
73 13 0
87 27 0
67 13
111
1 2 3
89 54 32
79 50 32
79 46 18
86 75 54
75 50 29
75 46 25
68 25 0
79 36 0
68 29 4
IV
1 2 3
98 86 74
98 89 74
98 89 76
75 66 55
81 63 47
77 64 48
92 64 20
93 64 30
92 67 28
0
STAAK,MEKAPKAT~EP, KAMPEand SCHONBERG
584
Table 4. Leptospirosis-positiveELISA reactions (%) (Cut-off: Mean negative OD plus 3 SD in MAT and/or CFT positive bovine sera) MAT
CFT
Total
t
+
96
96
93
84
74
73
5
8
8
t
-
74
74
76
55
47
48
20
30
28
-
+
73 (38.6) 88 (46.6) 28 (14.8)
32
32
19
54
29
25
0
0
0
189
76
76
74
66
55
54
12
17
17
Total
Anti-Ig conjugate Serum dilution 1:40 1:XO 1:160
Anti-IgG conjugate Serum dilution 1:40 1:XO 1:160
Anti-IgM conjugate Serum dilution 1:40 1 : 8 0 1:160
Sequential sera from a heifer artificially infected with L. hardjo and monitored bacteriologically (HAHN,1987) were tested in MAT, CFT, and ELISA using rabbit-antibovine Ig PO as conjugate (Table 5 ) . Table 5. Serological reactions in a L. hardjo infected heifer Day p. i. of sampling shedding of L. hardjo" MAT CFT ELISA'"s;"lSD 2s 3s
4
_ _ _ + _
10
11
14
17
21
24
28
31
_ _
_ _
t
t
t
+
-
t
-
_
t
t
t
+
_ +
+
_
-
t
-
_
_
_
_
+
+
+
-
+
35
t
t
+
+
+
-
t
+
+
+
_
+
+
-
38
t i
+ t t -
42
t
45
-
+
+
-
+
t + t t t -
49 51".'>
t
+ t +
+
-
+
+
t t
-
t t t t
'i Positive L. hardjo cultures from urine. ::+ Animal slaughtered, cultures from kidneys. '+'s':. Rab. antibovine Ig-PO conjugate.
Discussion The interpretation of MAT and CFT as t o leptospirosis results is still under discussion, mainly, because differences between infection or non-infection and serological results differ to a larger degree as compared to many other diseases. The majority of authors (LEWIS,1978; HATHAWAY and LITTLE,1983; WHITE et al., 1982; COUSINS et al., 1985) claim a 50 YO agglutination at a serum dilution of 2 1 : 100 to be a positive result. ELLISet al. (1982) regarded serum titres of 1 : 10 in L. hardjo infected cattle as an indication of infection. CORDES et al. (1982) recorded results as from 2 1 : 200 serum dilutions with 50 YO agglutination as positive. Bacteriological investigation of kidneys from slaughtered cattle and serological testing of matching serum samples (MAT, positive 2 1 : 100) revealed the following (WHITEet al., 1982): Serology Bacteriology %,
+ -
+
-
+ +
19
-
52 28
-
1
Serological Reactions of Leptospirosis-PositiveBovine Sera
585
In another investigation (HATHAWAY and LITTLE, 1983), agreement between the detection of leptospires in urine samples and serological positivity was found in 14 cases while in 8 cases serology produced positive and bacteriology negative results. The answer to the question of serological reactions in confirmed non-infected cattle populations seems to be even more difficult, and relevant information has not been available in literature. From the absence of clinical signs such as abortions, still births or sudden drop in milk yield together with negative or only low level antibody reactions it can only be concluded that leptospirosis does not exist in a given herd. Therefore, it may be justified to apply two different systems to the classification of serological anti-leptospirosis reactions: 1. If a herd is suspected of being infected with leptospirosis, repeated serum samples should be taken and even low antibody reactions are et to be recorded in addition to attempts to isolate the causative organism (ELLINGHAUSEN al., 1977). 2. For cross-sectional studies more importance should be attributed to specificity rather than sensitivity (PRITCHARD, 1984) by accepting only higher titre levels as a positive result. According to ELLISet al. (1982), a positive titre of 1 : 10 in confirmed L. hurdjo infections of cattle produced a sensitivity of 0.67 and a specificity of 0.86. When the classification of serological reactions was based on titre levels of 3 I : 100, the sensitivity dropped to 0.41 but the specificity was I . From this, it is concluded that serological results referring to leptospirosis should be interpreted on a herd basis and not for individual animals. The collection of positive and negative sera against which ELISA results have been compared were obtained from a cross-sectional study and from a clinically non-suspicious farm. A system of low sensitivity but high specificity has been applied for classification by negative, doubtful and positive reactions. CFT results were negative in 47 % of all positive reactions encountered. However, this test detected another 15 % positives in MAT negative reactors. This disagreement between the results of the MAT and C F T had been observed by other authors (BATZAand ZETTL,1985), and was reflected as well in the case of the L. hardjo infected heifer presented in this paper: Agreement between MAT and CFT: 7x, MAT+/CFT-: 2x, MAT-/CFT+: 2 ~ . The comparison between the 3 different conjugates has revealed that anti-bovine Ig conjugates are t o be preferred to anti-bovine IgG or IgM conjugates. This finding in bovine infections is in contrast to those in human leptospirosis where the infection is best detected by anti-IgM conjugates (TERPSTRA et al., 1985). For establishing ELISA cut-off ODs, the mean OD plus standard deviations from the leptospirosis negative cattle herd were and GARRET (1982) used absorbance calculated. This is an accepted procedure: THIERMANN readings which were at least 2 times higher than the average value of negative reference sera for cut-off. ADLERet al. (1982) applied mean OD readings from negative animals plus 3 standard deviations as a baseline for positive results. This system has been applied to other diseases as well (FORSCHNER, personal communication). References ADLER,B.S., D.V. COUSINS, S.FAINE,and G.M. ROBERTSON, 1982: Bovine IgM and IgG response to Leptospiru interrogans seTovar hardjo as measured by enzyme immunoassay. Vet. Microbiol. 7, 577-585. ALTON,G.G., L.M. JONES, R.D. ANGUS, and J.M. VERGER, 1975: Techniques for the brucellosis laboratory. INRA, Paris 1988, ISBN 2-7380-0042-8. BATZA,H. J., und K. ZETTL,1985: Nachweis von Leptospiren-Antikorpern in Schafseren rnittels verschiedener serologischerMethoden. Proc. 16th Congress of German Veterinary Ass. (DVG), Bad Nauheirn, 17.--20.4.1985. BREM,S., und A. WEBER,1984: Zum serologischen Nacliweis van Leptospiren-Infektionen bei Rindern und Pferden mit Hilfe eines im Handel erhaltlichen Objekttrageragglutinationstestes. Der praktische Tierarzt 10, 835-841. K. G. TOWNSEND, S. F. LEWIS,and J.T. S. HOLLAND, 1982: LeptoCORDES, D.O., M.E. CARTER, spirosis: I. Clinical investigation of the infection of dairy cattle in the Waikato district of New Zealand. N.Z. Vet. J. 30, 122-124.
586
STAAK,MEKAPKATEEP, KAMPEand SCHONBERG
COUSINS,D. V., G. M. ROBERTSON, and L. HUSTAS, 1985: The use of enzyme-linked immunosorbent assay (ELISA) to detect the IgM and IgG antibody response to Leptospira interroguns serovars hardjo, pomona, and turussovi in cattle. Vet. Microbiol. 10, 439-450. DGHM-Verfahrensrichtlinien, Mai 1985: Diagnostik der Leptospirosen. Gustav Fischer Verlag, Stuttgart, 1985. ELLINGHAUSEN, H . C., 11. L. DEYOE,and R. M. NERWIG, 1977: Leptospirosis in perspective. Proc. An. Meeting US An. H. Ass. 81, 161-182. ELLIS, W. A,, J. J. O’BRIEN,S. D. NEIIJ., and J. HANNA, 1982: Bovine leptospirosis: Serological findings in aborting cows. Vet. Record 110, 178-180. HAHN,B., 1987: Ein Beitrag zur Verbesserung des kulturellen Nachweises von Leptospiren durch Selektivnahrbtiden. Vet. med. Dissertation, FU Berlin, Journal Nr. 1354. S.C., T. W. A. LITTLE, and A. E. STEVENS,1982: Serological survey of leptospiral HATHAWAY, antibodies in sheep from England and Wales. Vet. Record 110, 99-101. HATHAWAY, S. C., and T. W. A. LITTLE,1983: Epidemiological study of Leptospira hurdjo infection in second calf dairy cows. Vet. Record 112, 215-218. HODGES,R. T., and W. WEDELL,1977: Adaptation of a complement fixation test for large scale serological diagnosis of bovine leptospirosis. N . Z. Vet. J. 25, 261 -262. LEWIS,J., 1978: The routine application of a microtechnique for the demonstration of leptospiral antibodies. J. Med. Microbiol. 11, 355-358. D. G., 1984: Some examples of the use of epidemiological techniques in the study of PRITCHARD, leptospirosis. Proc. C E C Seminary: The present state of leptospirosis diagnosis and control. Belfast, Oct. 10- 14, 1984. Martinus Nijhoff Publ., Dordrecht. S C H ~ N B X KA., G ,C. STAAKund U. KAMPE,1987: Leptospirose in der Bundesrepublik Deutschland. J. Vet. med. B 34, 98-108. TERPSTRA, W.J., G. S. LIGHTHART, and G. J. SCHOONE, 1985: ELISA for the detection of specific IgM and IgG in human leptospirosis. J. Gen. Microbiol. 131, 377-385. THIERMANN, A. B., and L. A. GARRET,1983: Enzyme-linked immunosorbent assay for the detection of antibodies to Leptospiru interrogarrs serovar hardjo and pomona in cattle. Am. J. Vet. Res. 44, 884 -887. WALTER,M., und W. SANITZ,1979: Untersuchungen zur serologischen Feststellung der Zystizerkose des Rindes mit Hilfe des Enzyme-linked immunosorbent assay (ELISA). BMTW 92, 131 -135. WHITE, F.H., K.R. SULZER,and R. W . ENGEL,1982: Isolations of Leptospiru intewogans serovars hardjo, balcanzca, and pornonu from cattle at slaughter. Am. J. Vet. Res. 43 (7), 1172-1173.
