Towards single screening tests for brucellosis.

This paper describes an indirect enzyme-linked immunosorbent assay (I-ELISA) and a fluorescence polarisation assay (FPA), each capable of detecting antibody in several species of hosts to smooth and rough members of the genus Brucella. The I-ELISA uses a mixture of smooth lipopolysaccharide (SLPS) and rough lipopolysaccharide (RLPS) as the antigen, and a recombinant protein A/G conjugated with horseradish peroxidase as the detection reagent. When using individually determined cutoff values, the SLPS/RLPS combined-antigen I-ELISA detected antibody in slightly more animals exposed to SLPS or to RLPS than did I-ELISA procedures using each individual antigen separately. Similarly, the assay using combined antigens detected antibody in slightly fewer animals not exposed to Brucella sp. When a universal cutoff of 10% positivity was used (relative to strongly positive control sera of each species), the overall performance index (percentage sensitivity plus percentage specificity) value decreased by 1.0 (from 199.4 to 198.4). In the FPA, it was not possible to use a universal cutoff without significant loss of performance. The overall sensitivity value for the FPA using the combined FPA antigen was 1.0% lower than using the O-polysaccharide (OPS) from SLPS and 9.1% higher than using the core antigen (CORE) from RLPS. When the combined antigen was used, the FPA specificity was slightly higher (1.2%) than from only the OPS, and considerably higher (12.6%) than the CORE. Overall, both the I-ELISA and the FPA with combined antigens were suitable as screening tests for all species of Brucella in the animal species tested.