Diagnosis of brucellosis by serology

Transcription

1 Veterinary Microbiology 90 (2002) 447±459 Diagnosis of brucellosis by serology Klaus Nielsen * Animal Diseases Research Institute, 3851 Fallow eld Road, Nepean, Ont., Canada K2H 8P9 Abstract Serological diagnosis of brucellosis began more than 100 years ago with a simple agglutination test. It was realized that this type of test was susceptible to false positive reactions resulting from,for instance,exposure to cross reacting microorganisms. It was also realized that this test format was inexpensive,simple and could be rapid,although results were subjectively scored. Therefore,a number of modi cations were developed along with other types of tests. This served two purposes: one was to establish a rapid screening test with high sensitivity and perhaps less speci city and a con rmatory test,usually more complicated but also more speci c,to be used on sera that reacted positively in screening tests. This led to another problem: if a panel of tests were performed and they did not all agree,which interpretation was correct? This problem was further compounded by the extensive use of a vaccine which gave rise to an antibody response similar to that resulting from eld infection. This led to the development of an assay that could distinguish vaccinal antibody,starting with precipitin tests. These tests did not perform well,giving rise to the development of primary binding assays. These assays,including the competitive enzyme immunoassay and the uorescence polarization assay are at the apex of current development,providing high sensitivity and speci city as well as speed and mobility in the case of the uorescence polarization assay. Crown Copyright # 2002 Published by Elsevier Science B.V. All rights reserved. Keywords: Serology; Brucellosis; Diagnostic; Agglutination tests; Complement xation; Precipitation tests; Binding assay; Antigen 1. Introduction Brucellosis may be diagnosed de nitively by isolation and identi cation of the causative organism. This was rst reported by Bruce and co-workers in 1887 when they isolated Brucella melitensis from military personnel in Malta (Bruce,1887,1893). The disease was derived from infected goats. Brucella abortus was isolated from aborting cattle by Bang (Bang,1897) while B. suis was rst described by Taum (1914). The above three species of * Tel.: ; fax: address: nielsenk@inspection.gc.ca (K. Nielsen) /02/$ ± see front matter. Crown Copyright # 2002 Published by Elsevier Science B.V. All rights reserved. PII: S (02)

2 448 K. Nielsen / Veterinary Microbiology 90 (2002) 447±459 Brucella comprise the most important species which contain O-polysaccharide (OPS) on the cell surface. The OPS which is part of the lipopolysaccharide (LPS) molecule has been chemically de ned: homopolymer of 4,6-dideoxy-4-formamide-alpha-D-mannose,linked via 1,2-glycosidic linkages (Bundle et al.,1987). Other important species which contain no measurable OPS on the cell surface are B. ovis,isolated by MacFarlane et al. (1952) and B. canis described by Kimberling et al. (1966). The three main species containing OPS are diagnosed serologically using either a whole cell antigen or smooth lipopolysaccharide (SLPS) prepared by chemical extraction while B. ovis and B. canis are mainly diagnosed using rough lipopolysaccharide (RLPS) or protein antigens (Blasco,1990). Of other Brucella species, B. neotoma,found in the desert wood rat,although diagnosed infrequently and Brucella species of marine mammals (without of cial species designation) may be detected serologically using B. abortus antigen (Nielsen et al.,2001). Because common epitopes are present in B. abortus, B. melitensis and B. suis SLPS,virtually all serological tests for antibody to these bacteria utilize B. abortus antigen (OIE Manual, 2000a,b) while the rough Brucella species are diagnosed using antigens from each species (Blasco,1990; Carmichael,1990). Since the majority of diagnostic tests use B. abortus antigens,these tests will be the main focus. Serology for human brucellosis is not discussed although in general,the same tests are applicable to animal and human serology. Before describing the various serological tests,it is useful to examine what is being measured. Again,the antibody response to B. abortus in cattle will be used as the example as it has been studied in most detail. The antibody response to B. abortus in cattle consists of an early IgM isotype response,the timing of which depends on the route of exposure,the dose of bacteria and the health status of the animal (Beh,1973,1974; Allan et al.,1976). The IgM response is followed almost immediately by production of IgG1 antibody and later by small amounts of IgG2 and IgA (Corbel,1972; Beh,1974; Allan et al.,1976; Levieux,1978a; Nielsen et al.,1984). Most cross reacting antibody,that is antibody resulting from exposure to microorganisms other than Brucella sp. or environmental antigens,consists mainly of IgM (Corbel,1985). Serological tests that measure IgM are therefore not desirable as false positive results occur,leading to low assay speci city. Since IgG2 and IgA antibodies accumulate later after exposure and are usually present in small and inconsistent amounts,the main isotype for serological testing is IgG1 (Allan et al., 1976; Lamb et al.,1979; Nielsen et al.,1984; Butler et al.,1986). Therefore,assays that predominately measure IgG1 are the most useful. Serological diagnosis was somewhat confounded by the development of a vaccine, B. abortus S19,which was serologically indistinguishable from virulent strains (Buck, 1930). This vaccine was and is an important aspect of the control and eradication programs for brucellosis and it was and is therefore widely used,however,before the development of primary binding assays,serological distinction was not easy,leading to allowances for higher antibody levels and a common practice of vaccinating calves before the age of 10 months (allowing for a reduction in antibody level prior to sexual maturity). These problems were largely overcome with the development of competitive enzyme immunoassays,(celisa) and a uorescence polarization assay (FPA),both of which eliminated nearly all residual vaccinal antibody (Nielsen et al.,1995,1996),and the development of a vaccine, B. abortus RB51,which contains no (or minute amounts of) OPS on its cell surface (Schurig et al.,1991). Acceptance of these serological tests has been slow and the

3 K. Nielsen / Veterinary Microbiology 90 (2002) 447± use of RB51 is only starting to become widespread. This has led to innumerable diagnostic dilemmas and the quarantine or slaughter of many non-infected animals. 2. Overview of serological tests 2.1. Common in-use tests Since 1897,a considerable number of serological tests have been developed. A number of these tests were modi ed in various ways to increase performance. The list below is not all inclusive,however,it includes most of the in-use tests and some that were not widely used but interesting. Each test will be discussed brie y. Agglutination tests: standard tube (SAT); acidified antigen (RBT,card,BPAT); high salt; EDTA; reducing agents (2ME,DTT); rivanol (RIV); heat; antiglobulin tests; milk ring test (MRT). Complement fixation test: warm (CFT); cold; indirect hemolysis test (IHLT); hemolysis in gel (HIG). Precipitation tests: agar gel immunodiffusion test (AGID); single radial immunodiffusion test (SRD). Primary binding assays: radioimmunoassay (RIA); particle concentration fluorescence immunoassay (PCFIA); indirect enzyme immunoassay (IELISA); competitive enzyme immunoassay (CELISA); fluorescence polarization assay (FPA) Agglutination tests Wright and Smith (1897) described the rst serological test for brucellosis. This test consisted of adding Brucella cells to diluted serum and observing the pattern of the cell pellet after incubation. This is very similar to the standard tube agglutination test still in use in some countries. IgM isotype of antibody is the most active agglutinin at a neutral or slightly below neutral ph (Rice and Boyes,1971; Corbel,1972; Nielsen et al.,1984). Therefore the tube agglutination test is susceptible to false positive reactions by cross

4 450 K. Nielsen / Veterinary Microbiology 90 (2002) 447±459 reacting antibody. As a result,the tube agglutination test should not be used as a diagnostic test and its discontinuation is recommended by the OIE (OIE Manual,2000a). Because of the speci city problems associated with the original tube agglutination test,a large number of modi cations were made to destroy or inactivate IgM agglutinins. Of these modi cations,the acidi ed antigen,rivanol precipitation and the use of 2-mercaptoethanol are in common use in various laboratories,while most of the other modi cations (heating the serum,addition of antiglobulin,antigen with increased sodium chloride,addition of chelating agents) were of local interest and are not discussed Acidified antigen modifications There are two commonly used tests,the rose-bengal/card test (RBT) (Nicoletti,1967; Morgan et al.,1969; Davis,1971) and the buffered antigen plate agglutination test (BPAT) (Angus and Barton,1984). The RBT uses B. abortus S99 or S whole cells,stained with rose-bengal while the BPAT uses B. abortus S whole cells,stained with crystal violet and brilliant green dyes. Both antigens are used at a ph of The low ph prevents some agglutination by IgM and encourages agglutination by IgG 1 thereby reducing nonspeci c interactions (Corbel,1972,1973; Allan et al.,1976). Both tests are standardized, simple to perform and inexpensive. Both tests are considered suitable for screening individual animals,however,false negative reactions occur,mostly due to prozoning. Antibody resulting from B. abortus S19 vaccination and some cross reacting antibodies are detected by these tests and it is necessary to use other test(s) to con rm reactor animals as infected. The BPAT is an OIE prescribed test for bovine,porcine,ovine and caprine brucellosis (OIE Manual, 2000a,b,d) Rivanol precipitation Rivanol,2-ethoxy-6,9-diaminoacridine lactate,is added to serum causing high molecular weight glycoproteins to precipitate (Nicoletti,1969; Huber and Nicoletti,1986). The precipitate is then removed by centrifugation and a rapid agglutination test using serum diluted 1:25,1:50,1:100 and 1:200 is performed. This test is fairly labor intensive and is mostly used as a con rmatory test Reducing agents Dithiothreitol (Klein and Behan,1981)and 2 mercaptoethanol (Rose and Roepke,1964; Anderson et al.,1964) are reducing agents that reduce disul de bridges. Both reduce IgM into monomeric units thereby reducing its ability to agglutinate. The reducing agents are added to the serum diluent. Serum is usually tested at the same dilutions as the rivanol test but using the standard tube agglutination antigen with incubation at 37 8C for at least 18 h. While this treatment increases speci city to a certain extent,it may also cause false negative reactions as IgG molecules contain disul de bridges as well and may be reduced to the point where they are unable to agglutinate. In addition,2-mercaptoethanol in particular is toxic and should only be used in a well-ventilated area. These tests are used mainly as con rmatory tests (Nicoletti,1969).

5 The rivanol precipitation and the reducing agent modi cations to the agglutination test do not eliminate vaccinal antibody and neither test is recommended by the OIE for international trade,however,both are extensively used in various national control/eradication programs Milk ring test An adaptation of the agglutination test uses hematoxylin stained whole cell antigen added to milk (Hunter and Allan,1972; Huber and Nicoletti,1986; Sutra et al.,1986). If antibody is present in the milk,a portion will be attached to the milk fat globules via the Fc portion of the antibody. These antibodies will agglutinate with the antigen and as the fat globules rise in the milk,a purple band will appear at the top of the milk. If no antibody is present,the fat band will remain buff colored. The test may be applied to pools of milk by increasing the amount of milk tested. While this is a relatively insensitive test,subject to wrong interpretation caused by various milk conditions such as mastitis,colostrum and milk at the end of the lactating cycle,it is recommended by the OIE as a screening test for bovine brucellosis (OIE Manual,2000a) Complement fixation tests K. Nielsen / Veterinary Microbiology 90 (2002) 447± The basis for the complement xation test (CFT) is that dilutions of serum from cattle, sheep or goats (swine sera may be used as well but require addition of normal porcine serum to supplement as a modifying factor),antigen (usually whole cells) and a pretitrated amount of complement (guinea pig serum is used as a source of complement while indigenous complement in the test serum is inactivated by heating for 30±60 min at 56 8C) are added together. If antibody is present in the serum,it will bind to the antigen and providing the antibody is IgG 1 isotype,complement will be activated. An indicator system is then added. It consists of sheep erythrocytes sensitized with rabbit antibody. If the test serum contained antibody,complement is not available and lysis of the erythrocytes will not take place. Alternately,if no antibody was present,the available complement activated by interaction with receptors on the Fc portion of the rabbit antibody will lyse the erythrocytes,releasing hemoglobin,which is assessed visually,or using a spectrophotometer. There are two main variations to the CFT standardized by Hill (1963) and MacKinnon (1963). The procedures are essentially identical varying only in the primary incubation temperature and time,either 37 8C for 1 h or 4 8C for 18 h. It is obvious that the CFT is technically challenging in that a large number of different reagents are necessary,a large number of various controls are required,a number of reagent titrations are needed each time the assay is set up and the results are subjective. Other problems are its inability to distinguish vaccinal from infection antibody and the occasional occurrence of serum samples that activate complement in the absence of antigen. Such sera cannot be used for diagnosis. In spite of these shortcomings,the CFT has been a valuable asset in control/eradication programs as a con rmatory test (although an automated version has been used as a screening test as well) and it is recommended by the OIE as a test prescribed for international trade. There are several other variations of the CFT,most notable being the indirect hemolysis test (Plackett et al.,1976; Nicoletti and Carlsen,1981; Tedder and Hoffmann,1981;

6 452 K. Nielsen / Veterinary Microbiology 90 (2002) 447±459 Chappel et al.,1982a,b; Sutherland et al.,1982; Hayes and Chappel,1982) and the hemolysis in gel test (Ruckerbauer et al.,1981; Nielsen et al.,1983),however,these modi cations were used only locally and not accepted as diagnostic tests. The CFT has also been used for the diagnosis of B. ovis infection in sheep (Myers et al., 1972; Burgess and Norris,1982; Searson,1982; Marin et al.,1989). This assay uses a hot saline extracted antigen,which contains rough lipopolysaccharide (RLPS) and a variety of proteins,mostly from the cell wall. The assay is similar to that used for bovine brucellosis. Although it has not been standardized internationally,it is an OIE prescribed test for trade (OIE Manual,2000c) Precipitation tests Agar gel immunodiffusion (AGID) and single radial immunodiffusion (SRD) tests were the rst tests to distinguish vaccinal antibody from that resulting from eld strain infection with B. abortus (Diaz et al.,1968,1979). An antigen,polysaccharide B (composed of a cyclic glucose molecule and a small amount of OPS,the active part of the preparation), derived from B. melitensis was used,either incorporated into the agar matrix in SRD, followed by addition of test serum to a well cut in the agar matrix or in the AGID,added to a well in the agar matrix adjacent to another containing test serum. If antibody was present,a ring of precipitate would appear within a couple of hours or after a longer incubation period with sera containing less antibody. Similarly,in the AGID test,opposing wells in the agar matrix, lled with antigen and test serum would produce a precipitin band if the serum contained antibody. While these tests provided data not available by other means,their sensitivity was insuf cient for wide scale diagnostic use (Jones et al.,1980) and neither test is recommended by the OIE for bovine brucellosis. Similar tests have been described for the diagnosis of B. ovis infection in sheep using a RLPS antigen or a hot saline extract also containing RLPS. The AGID in particular is currently in wide use for B. ovis diagnosis and is an alternate diagnostic test according to the OIE (OIE Manual,2000c) Radioimmunoassays A number of radioimmunoassays for the serodiagnosis of brucellosis were developed (Parratt et al.,1977; Wilson et al.,1977; Levieux,1978a,b; Tedder and Hoffmann,1981; Hayes and Chappel,1982; Chappel et al.,1982a,b; Chappel and Hayes,1983; Lawman et al.,1984). However,because of the inherent problems with usage of large quantities of radioisotopes,these assays have not found widespread use as diagnostic tools. Therefore, these assays will not be discussed Indirect enzyme immunoassay The IELISA was rst developed by Carlsson et al. (1976) for the diagnosis of human brucellosis. Since then,a large number of variations have been described (reviewed by Nielsen and Gall,1994),however,the most common format uses SLPS antigen coated passively onto a polystyrene matrix. Diluted serum is added (the diluting buffer usually

7 contains a detergent such as Tween 20 and divalent cation chelating agents,edta and EGTA,to reduce non speci c binding of serum proteins (Nielsen et al.,1994)) followed by an antiglobulin reagents speci c for or cross reacting with the test species immunoglobulin, conjugated with an enzyme,usually horseradish peroxidase or alkaline phosphatase. By including a strong positive,a weak positive and a negative serum control,assay performance and quality control are easily assessed. Data is frequently expressed as a percent of the reactivity of the strongly positive serum control. The OIE approved version of this test uses puri ed SLPS as the antigen,serum diluted 1:50 and a mouse monoclonal antibody speci c for bovine IgG 1 conjugated with horseradish peroxidase (OIE Manual,2000a). The same assay format may be used for testing milk (Vanzini et al.,1997,2001). This is a very sensitive test because milk may be tested at a dilution as low as 1:1. This is an OIE prescribed test for international trade,however,it has not been standardized (OIE Manual, 2000a). The IELISA assay format may also be used for detection of caprine and ovine antibody to B. melitensis. One disadvantage of the IELISA is its inability to differentiate vaccinal antibody resulting from B. abortus S19 or B. melitensis Rev1 vaccination from antibody induced by pathogenic strains (Nielsen and Gall,1994). Since pigs are not normally vaccinated against brucellosis,the IELISA is a suitable test for the diagnosis of porcine brucellosis. This format uses SLPS antigen,diluted serum and an antiglobulin reagent speci c for porcine immunoglobulin,conjugated with an enzyme. This particular test is an OIE alternate test but it has not been standardized (OIE Manual, 2000d). Brucella ovis may also be diagnosed using IELISA. The antigen used is RLPS or hot saline (HS) extract which may be coated passively onto a polystyrene matrix,followed by diluted serum and an anti-sheep immunoglobulin reagent (a monoclonal antibody to bovine IgG 1 has also been found to be suitable) conjugated with enzyme (Vigliocco et al., 1997). This is an alternate OIE test,which has not been standardized but is generally thought to be more sensitive and speci c than the CFT and AGID (Ris et al.,1984; Spencer and Burgess,1984; Worthington et al.,1984; Marin et al.,1989) Competitive enzyme immunoassays K. Nielsen / Veterinary Microbiology 90 (2002) 447± Since both the conventional serological tests and the IELISA cannot distinguish vaccinal antibody,competitive enzyme immunoassays were developed. The main rationale for these assays was that vaccines induced antibody of lower af nity due to the shorter exposure to antigen due to immune elimination compared to eld infection in which antigen persisted, resulting in increased antibody af nity (Nielsen et al.,1989; MacMillan et al.,1990). Thus a competing antibody could be selected to inhibit binding of vaccinal but not eld strain induced antibody. Because of their inherent supply and uniformity advantages,monoclonal antibodies were selected as competing antibodies,however,a similar assay,the particle concentration uorescence immunoassay (PCFIA) used a polyclonal antibody for competition. The latter assay was used exclusively in the USA as a rapid screening procedure (Snyder et al.,1990; Nicoletti and Tanya,1993). The selected monoclonal antibody should be speci c for a common epitope of the OPS molecule,allowing its use for B. abortus, B. melitensis and B. suis serology (Nielsen et al.,1995; Silva et al.,2000; Bianchi ori et al.,2000).

8 454 K. Nielsen / Veterinary Microbiology 90 (2002) 447±459 The most commonly used format of the CELISA utilizes SLPS from B. abortus as antigen, passively attached to a polystyrene matrix,followed by incubation with competing antibody and appropriately diluted test serum. After mixing and incubation,a reagent for detecting bound monoclonal antibody,labeled with an enzyme,usually horseradish peroxidase or alkaline phosphatase,is added followed by substrate/chromogen after a suitable incubation period. A wash procedure is performed between each step. A series of controls,including a strongly positive,a weakly positive,a negative serum as well as a buffer (no serum) controls must be included. Results are calculated as percent inhibition against the buffer control (0% inhibition). The CELISA is a prescribed test by the OIE for international cattle trade and an alternate test for swine brucellosis (OIE Manual,2000a,d) Fluorescence polarization assay The uorescence polarization assay (FPA) was developed as a test that could be performed outside the diagnostic laboratory,allowing for rapid and accurate diagnosis (reviewed by Nielsen and Gall,2001). The basis of the test is that a molecule in solution rotates randomly at a rate inversely proportional to its size. If the molecule is labeled with a uorescent marker and is examined by plane polarized light,a small molecule will rotate through a given angle faster than a larger molecule. The time of rotation may be measured using horizontal and vertical measurements (Nielsen et al.,2000). For diagnosis of brucellosis,a uorescence polarization analyzer is used to obtain a background measurement of uorescence of diluted serum. Antigen consisting of an OPS fragment,approximately 22 kda in size,labeled with uorescein isothiocyanate is added and incubated for 2 min,followed by a nal reading in the analyzer which automatically subtracts the background reading. The net result is presented in millipolarization units. The FPA can be performed almost anywhere using a portable analyzer which receives power from a laptop computer,using serum,milk or EDTA anticoagulated blood (Nielsen and Gall,2001). The test is rugged,relatively inexpensive,simple and very rapid. The FPA has been validated for a large number of species,including cattle,swine,bison and a number of cervids (Nielsen and Gall,2001) and is an alternate OIE test for bovine and swine brucellosis diagnosis (OIE Manual,2000a,d). 3. Summary To summarize the serological diagnosis of brucellosis, Table 1 compiles the sensitivity and speci city values for the various tests described. The data represents a survey of the literature and unpublished data. It is clear from the table that ef ciency of serological diagnosis of brucellosis has increased since its inception in 1897 (Wright and Smith,1897). It is also clear that no individual test is perfect. In an outbred population of animals frequently vaccinated with antigens that interfere with serology,diagnostic errors occur,however,these errors can be minimized by using the more reliable test,especially those that do not require subjective interpretation of results. As the incidence of brucellosis appears not to be declining on a worldwide basis,rapid and accurate diagnosis is imperative in order to control/eradicate

9 K. Nielsen / Veterinary Microbiology 90 (2002) 447± Table 1 Sensitivity,specificity,and performance index of the serological tests for brucellosis Test Sensitivity a Specificity a Performance index a References SAT 29.1± ± ±200 Van Aert et al.,1984; Lord et al.,1989 RBT 21.0± ± ±193.9 Van Aert et al.,1984; Samartino et al.,1999 Card 74.3± ± ±187.5 Stemshorn et al.,1985; Huber and Nicoletti, 1986; Lord et al.,1989 BPAT 75.4± ± ±199.7 Stemshorn et al.,1985; Uzal et al.,1996; Samartino et al.,1999 RIV 50.5± ± ±200 Huber and Nicoletti,1986; Lord et al.,1989; Dajer et al.,1999 2ME 56.2± ± ±200 Lord et al.,1989; Stemshorn et al.,1985; Saravi et al.,1995 CFT 23.0± ± ±197.5 Huber and Nicoletti,1986; Van Aert et al.,1984; Saravi et al.,1995 PCFIA 92.0± ± ±168.0 Nicoletti and Tanya,1993; Nielsen et al.,1998 IELISA 92.5± ± ±200 Dohoo et al.,1986; Rojas and Alonso,1994 CELISA 97.5± ± ±199.8 Samartino et al.,1999; Nielsen and Gall,2001 FPA 99.0± ± ±199.3 Dajer et al.,1999; Nielsen and Gall,2001 a Sensitivity and specificity values are presented as percent ranges of the worst and best results. The literature cites for each test with very low results used sensitivity and specificity calculations relative to other serological tests while the higher values were obtained using sera from cattle from which B. abortus was cultured or experimentally infected animals. The performance index is the sum of the % sensitivity and specificity. this disease from man,domestic animals and wildlife. This communication has presented a brief summary of the data currently available,hopefully in a fashion that will allow easy assessment and intelligent decision-making. Acknowledgements The contributions of David Gall are gratefully acknowledged. A recent scan of the literature revealed more than 2000 publications on serological diagnosis of brucellosis. As a consequence,it is not possible to include all. Therefore,this cannot be an exhaustive review and work not cited is not indicative of lack of quality. References Allan,G.,Chappel,R.,Williamson,P.,McNaught,D.,1976. A quantitative comparison of the sensitivity of serological tests for bovine brucellosis to different antibody classes. J. Hyg. 76,287±298. Anderson,R.,Jenness,R.,Brumfield,H.,Gough,P.,1964. Brucella agglutinating antibodies. Relation of mercaptoethanol stability to complement fixation. Science 143,1334±1335. Angus,R.,Barton,C.,1984. The production and evaluation of a buffered plate antigen for use in the presumptive test for brucellosis. Dev. Biol. Std. 56,349±358.

10 456 K. Nielsen / Veterinary Microbiology 90 (2002) 447±459 Bang,B.,1897. Die Aetiologie des Seuchenhaften (Infectiosen) Verwerfens. Zeit. Tiermed. 1,241±278. Beh,K.J.,1973. Distribution of Brucella antibody among immunoglobulin classes and a low molecular weight antibody fraction in serum and whey of cattle. Res. Vet. Sci. 14,381±384. Beh,K.J.,1974. Quantitative distribution of Brucella antibody amongst immunoglobulin classes in vaccinated and infected cattle. Res. Vet. Sci. 17,1±4. Bianchifiori,F.,Garrido,F.,Nielsen,K.,Moscati,L.,Duran,M.,Gall,D.,2000. Assessment of a monoclonal antibody-based competitive enzyme linked immunosorbent assay (CELISA) for diagnosis of brucellosis in infected and Rev1 vaccinated sheep and goats. Microbiologia 23,399±406. Blasco,J.M.,1990. Brucella ovis. In: Nielsen,K.,Duncan,J.R. (Eds.),Animal Brucellosis,CRC Press,Boca Raton Fl. pp. 351±378. Bruce,D.,1887. Note on the discovery of a microorganism in Malta fever. Practitioner 39,161±170. Bruce,D.,1893. Sur une nouvelle forme de fievre. Ann. Inst. Pasteur 4,289±304. Buck,J.M.,1930. Studies of vaccination during calfhood to prevent bovine infectious abortion. J. Agric. Res. 41, 667±685. Bundle,D.,Cherwonogrodsky,J.,Caroff,M.,Perry,M.,1987. The lipopolysaccharides of Brucella abortus and B. Melitensis. Ann. Inst. Pasteur Microbiol. 138,92±101. Burgess,G.,Norris,M.,1982. Evaluation of the cold complement fixation test for diagnosis of ovine brucellosis. Aust. Vet. J. 57,479±480. Butler,J.,Seawright,G.,McGivern,P.,Gilsdorf,M.,1986. Preliminary evidence for a diagnostic immunoglobulin G1 antibody response among culture-positive cows vaccinated with Brucella abortus strain 19 and challenge exposed with strain Am. J. Vet. Res. 47,1258±1264. Carlsson,H.,Hurvell,B.,Lindberg,A.,1976. Enzyme linked immunosorbent assay (ELISA) for titration of antibodies against Brucella abortus and Yersinia enterocolitica. Acta Pathol. Microbiol. Scand. 84C,168± 176. Carmichael,L.E.,1990. Brucella canis. In: Nielsen,K.,Duncan,J.R. (Eds.),Animal Brucellosis. CRC Press, Boca Raton,FL,pp. 335±350. Chappel,R.,Hayes,J.,Brain,G.,McNaught,D.,1982a. A modified radioimmunoassay for antibodies against Brucella abortus. J. Hyg. 88,1±9. Chappel,R.,Hayes,J.,1983. Comparison of radioimmunoassay with the complement fixation test and the indirect hemolysis test in the field diagnosis of bovine brucellosis. J. Hyg. 90,67±70. Chappel,R.,Hayes,J.,Rogerson,B.,Shenfield,L.,1982b. The serological response of cattle to vaccines against brucellosis,as measured by the brucellosis radioimmunoassay and other tests. J. Hyg. 88,11±19. Corbel,M.J.,1972. Characterization of antibodies active in the Rose Bengal plate test for bovine brucellosis. Vet. Rec. 88,447±449. Corbel,M.J.,1973. Identification of the immunoglobulin class active in the rose bengal plate test for bovine brucellosis. J. Hyg. 70,779±795. Corbel,M.J.,1985. Recent advances in the study of Brucella antigens and their serological cross-reactions. Vet. Bull. 55,927±942. Dajer,A.,Luna-Martinez,E.,Zapata,D.,Villegas,S.,Guteirrez,E.,Pena,G.,Gurria,F.,Nielsen,K.,Gall,D., Evaluation of a fluorescence polarization assay for the diagnosis of bovine brucellosis in Mexico. Prev. Vet. Med. 14,67±73. Davis,G.,1971. The Rose Bengal test. Vet. Rec. 88,447±449. Diaz,R.,Jones,L.,Leong,D.,Wilson,J.,1968. Surface antigens of smooth Brucellae. J. Bacteriol. 96,893±900. Diaz,R.,Garatea,P.,Jones,L.,Moriyon,I.,1979. Radial immunodiffusion test with Brucella polysaccharide antigen for differentiating infected from vaccinated cattle. J. Clin. Microbiol. 10,37±46. Dohoo,I.,Wright,P.,Ruckerbauer,G.,Samagh,B.,Robertson,F.,Forbes,L.,1986. A comparison of five serological tests for bovine brucellosis. Can. J. Vet. Res. 50,485±493. Hayes,J.,Chappel,R.,1982. A comparison of the results of the brucellosis radioimmunoassay and other serological tests in experimentally infected cattle. J. Hyg. 88,21±28. Hill,W.,1963. Standardization of the complement fixation test for brucellosis. Bull. OIE 60,401±410. Huber,J.,Nicoletti,P.,1986. Comparison of the results of card,rivanol,complement fixation and milk ring test with the isolation rate of Brucella abortus from cattle. Am. J. Vet. Res. 47,1529±1531. Hunter,D.,Allan,J.,1972. An evaluation of milk and blood tests used to diagnose brucellosis. Vet. Rec. 91, 310±312.

11 K. Nielsen / Veterinary Microbiology 90 (2002) 447± Jones,L.,Berman,D.,Moreno,E.,Deyoe,B.,Gilsdorf,M.,Huber,J.,Nicoletti,P.,1980. Evaluation of a radial immunodiffusion with polysaccharide B antigen for diagnosis of bovine brucellosis. J. Clin. Microbiol. 12, 753±758. Kimberling,C.V.,Luchsinger,D.W.,Anderson,R.K.,1966. Three cases of canine brucellosis. J. Am. Vet. Med. Assoc. 148,900±903. Klein,G.,Behan,K.,1981. Determination of Brucella immunoglobulin G agglutinating antibody titer with dithiothreitol. J. Clin. Microbiol. 14,24±25. Lamb,V.,Jones,L.,Schurig,G.,Berman,D.,1979. Enzyme linked immunosorbent assay for bovine immunoglobulin subclass-specific response to Brucella abortus lipopolysaccharides. Infect. Immun. 26, 240±247. Lawman,M.,Thurmond,M.,Reis,K.,Gauntlett,D.,Boyle,M.,1984. Solid phase radioimmunoassay for the detection of immunoglobulins against Brucella abortus. Vet. Immunol. Immunopathol. 6,291±305. Levieux,D.,1978a. A solid phase radioimmunoassay for the determination of bacterial specific antibodies within different immunoglobulin classes: application to bovine Brucella abortus antibodies. Ann. Rec. Vet. 9,523±530. Levieux,D.,1978b. Bovine immunoglobulins and brucellosis. Activity of IgG 1,IgG 2 and IgM versus different batches of Rose Bengal antigen. Ann. Rec. Vet. 9,489±593. Lord,V.,Rolo,M.,Cherwonogrodsky,J.,1989. Evaluation of humoral immunity to Brucella sp. in cattle by use of an agar gel immunodiffusion test containing polysaccharide antigen. Am. J. Vet. Res. 50,1813±1816. MacKinnon,D.,1963. The complement fixation test in brucellosis. Bull. OIE 60,383±400. MacMillan,A.,Greiser-Wilde,I.,Moenning,V.,Mathias,L.,1990. A competitive enzyme immunoassay for brucellosis diagnosis. Dtsch. Tierarztl. Wchnschr. 97,83±85. Marin,C.,Jiminez de Bagues,M.,Blanco,J.,Gamazo,C.,Moriyon,I.,Diaz,R.,1989. Comparison of three serological tests for Brucella ovis infection of rams using different antigenic extracts. Vet. Rec. 125,504±508. MacFarlane,D.,Salisbury,R.M.,Osborne,H.G.,Jebson,J.L.,1952. Investigations into sheep abortion in New Zealand during the 1950 lambing season. Aust. Vet. J. 28,221±226. Morgan,W.,MacKinnon,D.,Lawson,J.,Cullen,G.,1969. The Rose Bengal plate agglutination test in the diagnosis of brucellosis. Vet. Rec. 85,636±637. Myers,D.,Jones,L.,Varela-Diaz,V.,1972. Studies of antigens for complement fixation and gel diffusion tests in the diagnosis of infections caused by Brucella ovis and other Brucella. Appl. Microbiol. 23,894±902. Nicoletti,P.,1967. Utilization of the card test in brucellosis eradication. J. Am. Vet. Med. Assoc. 151, 1778±1781. Nicoletti,P.,1969. Further evaluation of serologic procedures used to diagnose brucellosis. Am. J. Vet. Res. 30, 1811±1821. Nicoletti,P.,Carlsen,W.,1981. Indirect hemolysis test in the serodiagnosis of bovine brucellosis. Am. J. Vet. Res. 42,1494±1497. Nielsen,O.,Stewart,R.,Nielsen,K.,Measures,L.,Duigan,P.,2001. Serologic survey of Brucella spp.: antibodies in some mammals of North America. J. Wildlife Dis. 37,89±100. Nicoletti,P.,Tanya,V.,1993. Comparison of enzyme labelled immunosorbent assay and particle concentration fluorescence immunoassay with standard serologic methods and bacteriologic culture for detection of Brucella sp. infected cows in herds with brucellosis. J. Am. Vet. Med. Assoc. 202,1975±1977. Nielsen,K.,Heck,F.,Stiller,J.,Rosenbaum,B.,1983. Interaction of specifically purified isotypes of bovine antibody to Brucella abortus in the hemolysis in gel test and enzyme linked immunosorbent assay. Res. Vet. Sci. 35,14±18. Nielsen,K.,Heck,F.,Wagner,G.,Stiller,J.,Rosenbaum,B.,Pugh,R.,Flores,E.,1984. Comparative assessment of antibody isotypes to Brucella abortus by primary and secondary binding assays. Prev. Vet. Med. 2, 197±204. Nielsen,K.,Cherwonogrodsky,J.,Duncan,R.,Bundle,D.,1989. Enzyme immunoassay for the differentiation of antibody response of Brucella abortus infected and vaccinated cattle. Am. J. Vet. Res. 50,5±9. Nielsen,K.,Kelly,L.,Gall,D.,Smith,P.,Bosse,J.,Nicoletti,P.,Kelly,W.,1994. The use of divalent cation chelating agents (EDTA/EGTA) to reduce non-specific interaction in enzyme immunoassay. Vet. Res. Commun. 18,433±437. Nielsen,K.,Gall,D.,1994. Advances in the diagnosis of bovine brucellosis: use of enzyme immunoassays. Gen. Eng. Biotech. 14,25±39.

12 458 K. Nielsen / Veterinary Microbiology 90 (2002) 447±459 Nielsen,K.,Kelly,L.,Gall,D.,Nicoletti,P.,Kelly,W.,1995. Improved CELISA for the diagnosis of bovine brucellosis. Vet. Immunol. Immunopathol. 48,285±295. Nielsen,K.,Gall,D.,Jolley,M.,Leishman,G.,Balsevicius,S.,Smith,P.,Nicoletti,P.,Thomas,F.,1996. Fluorescence polarization assay for the detection of antibody to Brucella abortus. J. Immunol. Meth. 195, 161±168. Nielsen,K.,Gall,D.,Lin,M.,Massangill,G.,Samartino,L.,Perez,B.,Coats,M.,Hennager,S.,Dajer,A., Nicoletti,P.,Thomas,F.,1998. Diagnosis of bovine brucellosis using a homogeneous fluorescence polarization assay. Vet. Immunol. Immunopathol. 66,321±329. Nielsen,K.,Lin,M.,Gall,D.,Jolley,M.,2000. Fluorescence polarization immunoassay: detection of antibody to Brucella abortus. Methods 22,71±76. Nielsen,K.,Gall,D.,2001. Fluorescence polarization assay for the diagnosis of brucellosis: a review. J. Immunoassay 22,183±201. OIE Manual of Standards for Diagnostic Tests and Vaccines. 2000a. Bovine brucellosis. OIE,Paris,pp. 328± 345. OIE Manual of Standards for Diagnostic Tests and Vaccines. 2000b. Caprine and ovine brucellosis. OIE,Paris, pp. 475±489. OIE Manual of Standards for Diagnostic Tests and Vaccines. 2000c.Ovine epididymitis (Brucella ovis). OIE, Paris,pp. 467±474. OIE Manual of Standards for Diagnostic Tests and Vaccines. 2000d. Porcine brucellosis. OIE,Paris,pp. 623±629. Parratt,D.,Nielsen,K.,White,R.,1977. Radioimmunoassay of IgM,IgG and IgA Brucella antibodies. Lancet 1, 1075±1078. Plackett,P.,Cottew,G.,Best,S.,1976. The indirect hemolysis test (IHLT) for bovine brucellosis. Aust. Vet. J. 52,136±140. Rice,C.,Boyes,B.,1971. Serum immunoglobulins in bovine brucellosis. New Zealand Vet. J. 19,146±154. Ris,D.,Hamel,K.,Long,D.,1984. Comparison of an enzyme linked immunosorbent assay (ELISA) with the cold complement fixation test for the serodiagnosis of Brucella ovis infection. New Zealand Vet. J. 32, 18±20. Rojas,X.,Alonso,O.,1994. ELISA for the diagnosis and epidemiology of Brucella abortus infection in cattle in Chile. IAEA TECDOC 1055,Vienna,pp. 77±82. Rose,J.,Roepke,M.,1964. Physiochemical properties of nonspecific bovine seroagglutinins for Brucella abortus. Am. J. Vet. Res. 25,325±328. Ruckerbauer,G.,Stemshorn,B.,Nielsen,K.,1981. An hemolysis in gel test for anti-brucella antibody in cattle serum. In: Butler,J. (Ed.),The Ruminant Immune System. Plenum Press,NY,pp. 782±783. Samartino,L.,Gregoret,R.,Gall,D.,Nielsen,K.,1999. Fluorescence polarization assay: application to the diagnosis of bovine brucellosis in Argentina. J. Immunoassay 20,115±120. Saravi,M.,Wright,P.,Gregoret,R.,Gall,D.,1995. Comparative performance of the enzyme linked immunosorbent assay (ELISA) and conventional assays in the diagnosis of bovine brucellosis in Argentina. Vet. Immunol. Immunopathol. 47,93±99. Schurig,G.G.,Roop,R.M.,Buhrman,D.,Boyle,S.,Bagchi,T.,Sriranganathan,N.,1991. Biological properties of RB51,a stable,o-chain deficient mutant of Brucella abortus. Vet. Microbiol. 28,171±188. Searson,J.,1982. Sensitivity and specificity of two microtitre complement fixation tests for the diagnosis of Brucella ovis infection in rams. Aust. Vet. J. 36,194±198. Silva,P.,Vigliocco,A.,Ramondino,R.,Marticorena,D.,Bissi,E.,Briones,G.,Gorchs,C.,Gall,D.,Nielsen,K., Evaluation of primary binding assays for presumptive serodiagnosis of swine brucellosis in Argentina. Clin. Diag. Lab. Immunol. 7,828±831. Snyder,M.,McMahon,P.,Workman,E.,1990. An automated fluorescence based brucellosis test system with a proven track record. In: Nielsen,K.,Duncan,J. (Eds.),Animal Brucellosis. CRC Press,Boca Raton,FL, pp. 237±282. Spencer,T.,Burgess,G.,1984. Enzyme linked immunosorbent assay for Brucella ovis specific antibody in ram sera. Res. Vet. Sci. 36,194±198. Stemshorn,B.,Forbes,L.,Eaglesome,M.,Nielsen,K.,Robertson,F.,Samagh,B.,1985. A comparison of the standard serological tests for diagnosis of bovine brucellosis. Can. J. Comp. Med. 49,391±394. Sutherland,S.,Le Cras,D.,Evans,R.,1982. Comparison of the complement fixation test and the indirect hemolysis test for cattle vaccinated and infected with Brucella abortus. J. Clin. Microbiol. 16,599±603.

13 K. Nielsen / Veterinary Microbiology 90 (2002) 447± Sutra,L.,Caffin,J.,Dubray,G.,1986. Role of milk immunoglobulins in the Brucella milk ring test. Vet. Microbiol. 12,359±368. Taum,J.,1914. Report of the Chief of the Bureau of Animal Industry. USDA. Washington,DC,pp. 30±35. Tedder,T.,Hoffmann,E.,1981. Immunoradiometric assay for examination and quantization of Brucella abortus specific antibodies reactive with the antigen(s) used in the indirect hemolysin test. J. Clin. Microbiol. 14, 415±426. Uzal,F.,Carrasco,A.,Nielsen,K.,Echaide,S.,Cabrera,R.,1996. An indirect ELISA using monoclonal anti IgG1 enzyme conjugate for the diagnosis of bovine brucellosis. Vet. Microbiol. 52,175±180. Van Aert,A.,Brioen,P.,Dekeyser,P.,Uytterhaegen,L.,Sijens,R.,Boeye,V.,1984. A comparative study of ELISA and other methods for detection of Brucella antibodies in bovine sera. Vet. Microbiol. 10,13±21. Vanzini,V.,Aguirre,N.,Lugaresi,C.,Echaide,S.,de,T.,Canavesio,V.,Guglielmone,A.,Marchesino,M., Nielsen,K.,1997. Evaluation of an indirect ELISA for the diagnosis of bovine brucellosis in milk and serum samples. Prev. Vet. Med. 36,211±217. Vanzini,V.,Aguirre,N.,Valentini,B.,Torione de Echaide,S.,Lugaresi,C.,Merchesino,M.,Nielsen,K.,2001. Comparison of an indirect ELISA with the Brucella milk ring test for detection of antibodies to Brucella abortus in bulk milk samples. Vet. Microbiol. 82,55±60. Vigliocco,A.,Silva,P.,Mestre,J.,Briones,G.,Draghi,G.,Tossi,M.,Nielsen,K.,1997. Development and validation of an indirect enzyme immunoassay for detection of ovine antibody to Brucella ovis. Vet. Microbiol. 54,357±368. Wilson,D.,Thornley,M.,Coombs,R.,1977. A solid phase assay with radioactively labelled antibody for the detection of Brucella abortus. J. Med. Microbiol. 10,281±292. Worthington,R.,Weddell,W.,Penrose,M.,1984. A comparison of three serological tests for the diagnosis of Brucella ovis infection in rams. New Zealand Vet. J. 32,58±60. Wright,A.E.,Smith,F.,1897. On the application of the serum test to the differential diagnosis of typhoid and Malta fever. Lancet 1,656±659.