www.elsevier.com/locate/vetmic

Veterinary Microbiology 124 (2007) 173–177

Short communication

Second generation competitive enzyme immunoassay

for detection of bovine antibody to Brucella abortus

K. Nielsen a,*, P. Smith a, W.L. Yu a, C. Elmgren a, P. Nicoletti b,B. Perez c, R. Bermudez d, T. Renteria d

a Canadian Food Inspection Agency, Ottawa Laboratory Fallowfield, 3851 Fallowfield Road, Ottawa, Ontario,

Canada K2H 8P9b College of Veterinary Medicine, University of Florida, Gainesville, FL, USA

c SAG, Santiago, Chiled College of Veterinary Medicine, Universidad Autonoma de Baja California, Mexicali, Baja California, Mexico

Received 20 February 2007; received in revised form 19 March 2007; accepted 22 March 2007

Abstract

A second generation competitive enzyme immunoassay (CELISA) for detection of bovine antibody to Brucella abortus was

developed. This assay was different from previously developed CELISAs in that the detection reagent used was a recombinant

combination of the receptor portions of protein A and protein G, labelled with horseradish peroxidase. This eliminates the need

for polyclonal anti-mouse-enzyme conjugate reagents for detection thus allowing for true standardization.

The assay utilized a monoclonal antibody specific for a common epitope of the O-polysaccharide (OPS) of smooth

lipopolysaccharide (SLPS) derived from B. abortus S1119.3 but which did not react with protein A/G. This monoclonal antibody

was used to compete with antibody in the bovine test serum. Binding of bovine antibody to the smooth lipopolysaccharide

antigen was then measured directly with the protein A/G enzyme conjugate. In this case, development of colour in the reaction

was indicative of the presence of bovine antibody.

The performance characteristics, sensitivity, specificity and exclusion of B. abortus S19 vaccinated animals, of the assay

were very similar to those of the classical CELISA.

Crown Copyright # 2007 Published by Elsevier B.V. All rights reserved.

Keywords: Brucellosis; CELISA; Serology; Protein A/G; Diagnosis; Standardization; Harmonization

* Corresponding author. Tel.: +1 613 228 6698;

fax: +1 613 228 6669.

E-mail address: nielsenk@inspection.gc.ca (K. Nielsen).

0378-1135/$ – see front matter. Crown Copyright # 2007 Published by

doi:10.1016/j.vetmic.2007.03.023

1. Introduction

Competitive enzyme immunoassay is a multi-

species diagnostic test which was developed to reduce

the impact of vaccinal antibody and antibody derived

from exposure to cross-reacting antigens on the

serological diagnosis of brucellosis. This type of

Elsevier B.V. All rights reserved.

K. Nielsen et al. / Veterinary Microbiology 124 (2007) 173–177174

assay utilized a competing antibody of higher affinity

than cross reacting or vaccinal antibody thereby

eliminating some antibodies that confuse diagnosis,

unlike other serological tests such as indirect enzyme

immunoassay and the classical tests, all of which

provide global antibody measurements. This increase

in specificity was obtained with a minimal loss of

sensitivity (Gall and Nielsen, 1994; McGivern et al.,

2003; Aguirre et al., 2002; Biancifiori et al., 2000;

Samartino et al., 1999; Gall et al., 1998).

Early versions of the CELISA used OPS specific

monoclonal antibody labelled directly with enzyme

(Sutherland, 1985; Sutherland and den Hollander, 1986;

Nielsen et al., 1989; Chin et al., 1989; MacMillan et al.,

1990; Weynants et al., 1996). However, it was

recognized that minor adjustments in assay parameters

impacted on assay performance and as a result, a second

step involving the use of an enzyme labelled anti-mouse

reagent was included (Nielsen et al., 1995). This step

allowed for minor adjustments in performance but

unfortunately introduced a variable that was difficult to

control as the anti-mouse reagent was polyclonally der-

ived and therefore virtually impossible to standardize.

A number of antigens other than SLPS epitopes

have been tested. Most were outer membrane proteins

and while some showed promise as diagnostic

reagents, it appeared that combinations of antigens

were necessary, thus complicating the assay (Deb-

barah et al., 1996; Cloeckaert et al., 1992).

The described assay attempts to eliminate uncontrol-

lable variables from the CELISA in that it makes use of

SLPS antigen which can be standardized (Nielsen et al.,

1998), an anti-OPS monoclonal antibody which does

not react with protein A/G and a recombinant protein A/

G enzyme conjugate for detection of bound bovine

antibody. This format was compared to the classical

assay (OIE Manual, 2004) for detection of antibody

in sera from B. abortus infected animals, B. abortus

S19 vaccinated animals and a brucellosis free popula-

tion.

2. Materials and methods

2.1. Antigen

Smooth lipopolysaccharide was extracted from B.

abortus S1119.3 cells by the technique of Baker and

Wilson (1965) and standardized (Nielsen et al.,

1998).

2.2. Monoclonal antibody

Monoclonal antibodies were generated according

to standard protocols (Liddell and Cryer, 1991) using

spleen cells from mice immunized with B. abortus

SLPS. Antibody production was assessed by indirect

enzyme immunoassay using SLPS antigen and

commercially available goat anti-mouse IgG (H and

L chain) and recombinant protein A/G, both labelled

with horseradish peroxidase for detection. Cell lines

producing antibody which reacted with the goat anti-

mouse conjugate but not with the protein A/G

conjugate were selected, cloned twice, isotyped and

grown in bulk.

2.3. Protein A/G construction

Staphylococcus aureus (ATCC #12598) and the

group G Streptococcus sp. (ATCC #12394) were

purchased from ATCC and chromosomal DNA from

each bacterium containing IgG binding repeats (Uhlen

et al., 1984; Kihlberg et al., 1992) were generated and

amplified by polymerase chain reaction (PCR). The

recombinant protein A/G was generated as described

previously (Nielsen et al., 2005). Protein A/G was

purified using a nickel–nitrilotriacetic acid column

and labelled with horseradish peroxidase (Henning

and Nielsen, 1987). Protein A/G labelled with

horseradish peroxidase is commercially available

from several sources.

2.4. CELISA test protocol

The assay protocol was described previously (OIE

Manual, 2004) with the exception that protein A/G–

enzyme conjugate, appropriately diluted, was used for

detection.

2.5. Other serological tests

Serological tests for bovine antibody to B. abortus

were performed as outlined in the OIE Manual (2004).

Tests included the buffered antigen plate agglutination

test (BPAT), the complement fixation test (CFT), the

indirect ELISA using a monoclonal antibody specific

K. Nielsen et al. / Veterinary Microbiology 124 (2007) 173–177 175

for bovine IgG1 labelled with horseradish peroxidase

as a detection reagent (IELISA-1), classical CELISA

and fluorescence polarization assay (FPA). In addition,

a second indirect ELISA (IELISA-2), employing

protein A/G–horseradish peroxidase as the detection

reagent, was included.

2.6. Serum samples

Sera for 238 cattle from which B. abortus was

isolated either from tissues or from milk were used as

positive samples. Sera from randomly selected

Canadian cattle, n = 1033, were included as a

negative population. Canada has been free from

brucellosis in domestic animals since 1984. Sera

from 110 B. abortus vaccinated animals were tested.

These sera included 37 samples from field vaccinated

calves, bled at a point in time and 73 samples from

calves vaccinated and bled periodically for 6 months

post vaccination. All calves were approximately 7

months of age and were vaccinated with 3 � 1010 cfu

B. abortus S19. All sera from vaccinated animals

were selected based on a positive reaction in the

IELISA-1.

Control bovine samples included sera that gave a

strong positive, a weak positive and a negative

serological reaction, tested in duplicate in each test.

Duplicate buffer controls were included.

2.7. Data

Results of the IELISAs were calculated as percent

positivity relative to the strongly positive control

serum using the formula:

%P ¼ optical densitytest sample=

optical densitystrongly positive control average � 100

Results of the CELISAs were calculated as percent

inhibition based on the uninhibited control (buffer

control) using the formula:

%I ¼ 100� ðoptical densitytest sample=

optical densitybuffer control average � 100Þ

The results of the FPA were expressed as millipolar-

ization units (mP).

Cut-off values between positive and negative

samples were predetermined as follows:

Test

Cut-off Detection

BPAT

+ or � CFT >1/5

IELISA-1

>46% P Maba

IELISA-2

>20% P PAGb

CELISA-1

>30% I GAMc

CELISA-2

>11% I PAGb

FPA

>90 mP

a Monoclonal antibody specific for bovine IgG1 enzyme con-

jugate.b Protein A/G enzyme conjugate.c Commercial goat anti-mouse IgG enzyme conjugate was used

for detection.

Sensitivity and specificity estimates were calcu-

lated based on these cut-off values.

3. Results

The new version of the CELISA was found to

provide the best discrimination between brucellosis

positive and negative cattle when monoclonal anti-

body M2952, a mouse isotype IgA, was used. When

compared to the old version of the CELISA, the results

were identical, 100% sensitivity and 100% specificity

each when using cut-off values of 11 and 30% I,

respectively. The FPA gave a similar result as did both

versions of the IELISA.

Sera from 110 B. abortus S19 vaccinated animals

were selected on the basis of positivity in the IELISA-

1 and all sera were also positive in the BPAT and CFT.

The sera were collected over a period of time starting

on the day of vaccination and continuing for

approximately 6 months. Both versions of the

CELISA, the new version IELISA and the FPA

reduced the number of reactor sera in this group by

approximately 50%. These data are presented in

Fig. 1. The results of the various serological tests are

summarized in Table 1.

4. Discussion

The second generation CELISA described was

compared to other serological tests commonly used

for the diagnosis of bovine brucellosis. Using defined

K. Nielsen et al. / Veterinary Microbiology 124 (2007) 173–177176

Fig. 1. Top: antibody levels in B. abortus S19 vaccinated calves

bled approximately 1 week apart. As determined by FPA (in mP

units, >90 is positive), CFT (in reciprocal titers, 5 is positive) and

IELISA-1 using a monoclonal antibody specific for bovine IgG1

enzyme conjugate as the detection reagent (in %P, with 46% P being

positive). Bottom: IELISA-2 using protein A/G enzyme conjugate

for detection (in %P, using 20% P as positive), CELISA-1 using goat

anti-mouse IgG enzyme conjugate (in %I, using 30% I as positive)

and CELISA-2 using protein A/G enzyme conjugate for detection

(in %I, using 11% I as positive).

Table 1

Percent sensitivity and specificity of serological tests using defined

sera from cattle infected with B. abortus, unexposed or vaccinated

with B. abortus S19

Test Sensitivity

(%)

Specificity

(%)

Vaccinates

specificity

BPAT NA NA 0

CFT NA NA 0

IELISA-1 NA NA 0

CELISA-1 100 100 47.2

FPA 100 100 51.5

IELISA-2 100 100 50.0

CELISA-2 100 100 53.8

The specificity of the tests using sera from B. abortus S19 vaccinated

animals indicates the number of sera that did not give a positive

result in the assays. Data based on 238 sera from B. abortus infected

animals, 1033 Canadian cattle (negative) and 110 sera from B

abortus S19 vaccinated cattle.

serum samples, this test performed as well as the FPA

and the earlier version of the CELISA, each giving

100% sensitivity and 100% specificity as did, the

IELISA-2, the version using protein A/G enzyme

conjugate for detection. In examining sera from

vaccinated animals, the new version compared well

with the old version CELISA and the FPA. The

IELISA-2 performed as well as the CELISAs and FPA

in discriminating vaccinal antibody. The specificity of

all the assays were approximately 50%, however, it

.

should be realized that the sample set contained sera

collected throughout the antibody response, resulting

in a number of sera from the early response that were

positive in all tests. All the sera from vaccinated

animals were positive when tested in the original

IELISA-1, the BPAT and the CFT. The data is

summarized in Fig. 1.

Standardization and harmonization of serological

tests used for the presumptive diagnosis of infectious

diseases has always been difficult due to variability in

reagents and subjective assessment. Standardization

of serological diagnosis of bovine brucellosis is

currently based on the use of a strongly positive, a

weakly positive and a negative standard serum (issued

for the OIE by VLA, Weybridge, UK) and on reaction

patterns in the serological test. In the classical

serological tests, agglutination and complement

fixation tests, the patterns are dependent on visual

examination of antibody increments of 100% and

therefore highly subjective. In the primary binding

assays, especially the ELISAs, electronic assessment

of results is available, however, no attempts have been

made to standardize the various reagents used.

Therefore, the only test at this moment that can be

truly standardized is the FPA. Unfortunately, because

of its requirement for new equipment and the cost per

test, it is not available to most laboratories.

The CELISA described in this communication can

be standardized because all the reagents used are

produced in vitro. The antigen can be prepared by

chemical extraction of Brucella cells and standardized

K. Nielsen et al. / Veterinary Microbiology 124 (2007) 173–177 177

based on weight and biological function (Nielsen

et al., 1998). Monoclonal antibody has all the inherent

advantages of continuous production of a defined

product. With the elimination of the use of a

polyclonal anti-mouse enzyme conjugate in favour

of a recombinant protein A/G enzyme conjugate, all

the parameters can be controlled and test harmoniza-

tion is possible.

Acknowledgements

The authors gratefully acknowledge the expert

contributions of the Diagnostic Serology and Mono-

clonal Antibody Units of OLF.

References

Aguirre, N.P., Vanzini, V.R., Torioni de Echaide, S., Valentini, B.S.,

De Lucca, G., Aufranc, C., Canal, A., Vigliocco, A., Nielsen, K.,

2002. Antibody dynamics in Holstein Friesian heifers vacci-

nated with Brucella abortus stain 19 using seven serological

tests. J. Immunoassay Immunochem. 23, 471–478.

Baker, P.J., Wilson, J.B., 1965. Chemical characterization and

biological properties of endotoxin of Brucella abortus. J. Bac-

teriol. 90, 895–902.

Biancifiori, F., Garrido, F., Nielsen, K., Moscati, L., Duran, M., Gall,

D., 2000. Assessment of a moncolonal antibody-based compe-

titive enzyme linked immunosorbent assay (CELISA) for the

diagnosis of brucellosis in infected and Rev.1 vaccinated sheep

and goats. New Microbiol. 23, 399–406.

Chin, J., Daniels, J., Bundesen, P., 1989. Bovine brucellosis: evalua-

tion of field sera by a competitive and superimposable ELISA

utilising monoclonal antibody against Brucella abortus lipopo-

lysaccharide. Vet. Immunol. Immunopathol. 20, 109–118.

Cloeckaert, A., Kerkhofs, P., Limet, J.N., 1992. Antibody response

to Brucella outer membrane proteins in bovine brucellosis:

immunoblot analysis and competitive enzyme-linked immuno-

sorbent assay using monoclonal antibodies. J. Clin. Microbiol.

30, 3168–3174.

Debbarah, H.S., Zygmunr, M.S., Dubray, G., Cloeckaert, A., 1996.

Competitive enzyme-linked immunosorbent assay using mono-

clonal antibodies to the Brucella melitensis BP26 protein to

evaluate antibody responses in infected and B. melitensis Rev.1

vaccinated sheep. Vet. Microbiol. 53, 325–337.

Gall, D., Colling, A., Marion, O., Moreno, E., Nielsen, K., Perez, B.,

Samartino, L., 1998. Enzyme immunoassays for serological

diagnosis of bovine brucellosis: a trial in Latin America. Clin.

Diagn. Lab. Immunol. 5, 654–661.

Gall, D., Nielsen, K., 1994. Improvements to the competitive ELISA

for detection of antibodies to Brucella abortus in cattle sera. J.

Immunoassay 15, 277–291.

Henning, D., Nielsen, K., 1987. Peroxidase labelled monoclonal

antibodies for use in enzyme immunoassay. J. Immunoassay 8,

297–307.

Kihlberg, B.M., Sjobring, U., Kastern, W., Bjorck, L., Protein, L.G.,

1992. a hybrid molecule with unique immunoglobulin binding

properties. J. Biol. Chem. 267, 25583–25588.

Liddell, J.E., Cryer, A., 1991. A Practical Guide to Monoclonal

Antibodies. John Wiley and Sons, Inc., New York, pp. 1–180.

McGivern, J.A., Tucker, J.D., Perrett, L.L., Stack, J.A., Brew, S.D.,

MacMillan, A.P., 2003. Validation of FPA and CELISA for the

detection of antibodies to Brucella abortus in cattle sera and

comparison to SAT, CFT and IELISA. J. Immunol. Meth. 278,

171–178.

MacMillan, A.P., Greiser-Wilke, I., Moennig, V., Mathias, L.A.,

1990. A competitive enzyme immunoassay for brucellosis diag-

nosis. Dtsch. Tierartzl. Wochenschr. 97, 83–85.

Nielsen, K., Smith, P., Yu, W., Nicoletti, P., Elzer, P., Robles, C.,

Bermudez, R., Renteria, T., Moreno, F., Ruiz, A., Massangill, C.,

Muenks, Q., Jurgersen, G., Tollersrud, T., Samartino, L., Conde,

S., Forbes, L., Gall, D., Perez, B., Rojas, X., Minas, A., 2005.

Towards single screening tests for brucellosis. Rev. Sci. Tech.

Off. Int. Epiz. 24, 1027–1038.

Nielsen, K., Kelly, L., Mallory, M., 1998. Standardization of smooth

lipopolysaccharide preparations for use in diagnostic serological

tests for bovine antibody to Brucella abortus. J. Immunoassay

19, 239–250.

Nielsen, K., Kelly, L., Gall, D., Nicolettui, P., Kelly, W., 1995.

Improved competitive enzyme immunoassay for the diagnosis of

bovine brucellosis. Vet. Immunol. Immunopathol. 46, 285–

291.

Nielsen, K., Cherwonogrodsky, J., Duncan, J.R., Bundle, D.R., 1989.

Enzyme-linked immunosorbent assay for differentiation of the

antibody response of cattle naturally infected with Brucella

abortus or vaccinated with strain 19. Am. J. Vet. Res. 50, 5–9.

OIE Manual of Diagnostic Tests and Vaccines for Terrestrial

Animals. Chapter 2.3.1.Bovine Brucellosis. OIE, Paris, 2004,

pp 409–438.

Samartino, L., Gall, D., Gregoret, R., Nielsen, K., 1999. Validation

of enzyme-linked immunosorbent assays for the diagnosis of

bovine brucellosis. Vet. Microbiol. 70, 193–200.

Sutherland, S., 1985. An enzyme-linked immunosorbent assay for

detection of Brucella abortus in cattle using monoclonal anti-

body. Aust. Vet. J. 62, 264–268.

Sutherland, S., den Hollander, L., 1986. Comparison of an enzyme-

linked immunosorbent assay using monoclonal antibodies and a

complement fixation test for cattle vaccinated and infected with

Brucella abortus. Vet. Microbiol. 12, 55–64.

Uhlen, M., Guss, B., Nilsson, B., Gatenbeck, S., Philipson, J.,

Lindberg, M., 1984. Complete sequence of the staphylococcal

gene encoding protein A. A gene evolved through multiple

duplications. J. Biol. Chem. 259, 1695–1702.

Weynants, V., Gilson, D., Cloeckaert, A., Denoel, P.A., Tibor, A.,

Thiange, P., Limet, J.N., Letesson, J.J., 1996. Characterization of

a monoclonal antibody specific for Brucella smooth lipopoly-

saccharide and development of a competitive enzyme-linked

immunosorbent assay to improve the serological diagnosis of

brucellosis. Clin. Diagn. Lab. Immunol. 3, 309–314.