A PCR assay with primers derived from the 16S rRNA sequence of Brucella abortus was developed. Nine different combinations between six primers were tested. One pair of primers, which amplified a 905-bp fragment, was selected. As little as 80 fg of Brucella DNA was detected by this method. DNAs from all of the representative strains of the species and biovars of Brucella and from 23 different Brucella isolates were analyzed and yielded exclusively the 905-bp fragment. No amplification was detected with DNAs from 10 strains phylogenetically related to Brucella spp., 5 gram-negative bacteria showing serological cross-reactions with Brucella spp., and 36 different clinical isolates of non-Brucella species. Only Ochrobactrum anthropi biotype D yielded a PCR product of 905 bp, suggesting a closer relationship between Brucella spp. and O. anthropi biotype D. The specificity and high sensitivity of the PCR assay may provide a valuable tool for the diagnosis of brucellosis.

Five serological assays were evaluated for the diagnosis of brucellosis in goats: the rose bengal test (RBT), complement fixation test (CFIT), radial immunodiffusion (RID) with Brucella and Yersinia enterocolitica 0:9 polysaccharides, counterimmunoelectrophoresis (CIEP) with cytosol, and enzyme-linked immunosorbent assay (ELISA) with polyclonal and protein G conjugates and smooth lipopolysaccharide (S-LPS), native hapten polysaccharide (NH), or cytosol antigens. For optimal sensitivity, RBT had to be used with sera-antigen at a 3:1 dilution. In the RID test, Brucella melitensis biotype 1 NH could not be replaced by Brucella abortus biotype 1 or Y. enterocolitica 0:9 polysaccharides. In the ELISA, S-LPS and NH gave similar results and the protein G conjugate increased the specificity. With the sera from 55 B. melitensis culture-positive goats, the sensitivity was 100% for RBT, CFT (titer -4), and ELISA with S-LPS or NH; 94% for RID; and 93% for CIEP. All tests were negative (100% specific) when testing the sera from 127 brucella-free goats. Larger discrepancies among the results of the serological tests were obtained with sera from goats of areas where brucellosis is endemic. When the sera of 20 young goats vaccinated subcutaneously (109 CFU of B. melitensis Rev 1) and bled 6 months later were examined, the specificities were as follows: NH ELISA, 60%; CFT and S-LPS ELISA, 75%;RBT, 80%; CIEP, 90%; and RID, 94%. With the sera from 10 young goats vaccinated conjunctivally (109 CFU of B. melitensis Rev 1) all tests were 100% specific 4 months after vaccination. The proportion of goats giving a positive reaction after vaccination decreased faster in RID than in other tests.

A study was performed to evaluate the previously described PCR (C. Romero, C. Gamazo, M. Pardo, and I. López-Goñi, J. Clin. Microbiol. 33:615-617, 1995) for the diagnosis of brucellosis in dairy cattle. Milk samples from 56 Brucella milk culture-positive cattle and from 37 cattle from Brucella-free herds were examined for Brucella DNA by PCR and for specific antibodies by an indirect enzyme-linked immunosorbent assay (ELISA). The specificities of both tests were 100% when testing the milk samples from Brucella-free cattle. The milk samples from 49 infected cattle were positive by PCR (87.5% sensitivity), and 55 were positive by ELISA (98.2% sensitivity). A PCR-positive sample was negative by ELISA, and 7 ELISA-positive samples were PCR negative, yielding an observed proportion of agreement of 0.91 for the two tests. Although the results suggest that ELISA is a better screening test than PCR, the combined sensitivity of the two assays was 100%, and their simultaneous application could be more useful than one test alone for a rapid screening of brucellosis in dairy cattle.