Indirect ELISA for the diagnosis of brucellosis in water buffaloes (Bubalus bubalis) in Italy

BRUCELLOSIS is of great concern worldwide, with zoonotic implications and great economic losses for breeders. Buffalo breeding, in particular, has a long tradition in Italy, mainly in southern Italy where the consumption of buffalo mozzarella cheese has steadily increased. The importance of buffalo breeding has resulted in greater attention being paid to buffalo diseases, especially brucellosis. In 1989, national measures to control and eradicate the disease in herds were introduced by veterinary services and, after the diagnosis of buffalo brucellosis using serological tests, any infected animals were slaughtered. Due to the high level of positive buffaloes detected by applying this strategy, an attenuated live vaccine (Brucella abortus strain 19) was used between 1991 and 1998 to rid infected herds of buffalo brucellosis. Nowadays, the use of vaccine is prohibited and all buffaloes found infected by conventional serological tests, that is, the rose bengal test (RBT) for screening and the complement fixation test (CFT) for confirmation, are slaughtered. The complete eradication of buffalo brucellosis from herds is difficult due to the management systems and environmental conditions employed in buffalo breeding. Most farms have open stalls with a hardened soil paddock. The animals are free to roam within a covered structure and have free access to mangers and outside fenced-off areas. Brucellosis is transmitted via environmental contamination as a result of the animals' secretions and excrements (Morgan and others 1960, Mayfield and others 1988, Garin-Bastuji 1993) and by the use of natural breeding. In addition, disease surveillance measures have many problems because it is difficult to discriminate between the humoral response in naturally infected buffaloes and vaccinated buffaloes. The aim of this study was to develop a homologous indirect ELISA for the detection of antibodies to Brucella species in buffalo sera. An indirect ELISA would be a useful alternative test for the diagnosis of brucellosis in water buffalo (Bubalus bubalis), and may also be useful in prevention programmes. Although there is much information available about the application of ELISA methods for the diagnosis of brucellosis in other species, especially cattle and sheep (Nielsen and others 1996, Uzal and others 1996, Biancifiori and others 1997), this is the first study concerning buffalo. In this study, 796 buffalo sera from two geographically distinct areas were tested. Two hundred and ninety-eight samples were collected from six farms in the province of Latina in the Lazio region, and 498 buffalo sera were taken from six farms in the province of Caserta in the Campania region. While the farms from Latina province were serologically negative to buffalo brucellosis by conventional tests, and no history of abortions or other clinical signs of the disease had been recorded, those from Caserta province were serologically positive by both the RBT and CFT tests. One serum sample from a naturally infected buffalo that had aborted a fetus in which B abortus biotype 1 infection had been diagnosed by bacteriological isolation and characterisation, was set as the positive reference standard. This serum showed a high titre of 851 2 CFT units. Specific tissue and fluid fetal buffalo samples were tested for the presence ofB abortus as described byAlton and others (1988), using tryptose agar with 5 per cent bovine serum, cyclohexamide (30 [.g/ml), bacitracin (7.5 U/ml), polymyxin B sulphate (1-8 U/ml) and ethyl violet. Brucella species colonies were identified by morphological and growth characteristics, and biochemical testing. The ELISA test was standardised as described by Wright and others (1993). The laboratory positive reference standard serum was diluted at several scalar dilutions between 1:25 and 1:25,600 using phosphate-buffered saline plus 0 5 per cent w/v ovalbumin, and analysed by ELISA. The results for absorbance at 405 nm were plotted against the logarithm of the dilution factor, and the corresponding titration curve was calculated (Fig 1). Buffalo immunoglobulin G (IgG) was purified from serum by protein G affinity chromatography using fast protein liquid chromatography (FPLC) (Pharmacia-Biotech), equipped with a uv-MII detector set at 280 nm. Chromatography was run at a 1 ml/minute flow rate on a protein G sepharose 10/2 column (Pharmacia-Biotech) equilibrated with a 100mM TrisHCI buffer (pH 7.5); 20 ml of buffalo serum were diluted 1:1 with the same buffer and loaded onto the column by a 50 ml injection Superloop (Pharmacia Biotech). After washing the column using this buffer, IgG was eluted by changing the pH using a 100mM glycine-HCl buffer (pH 2.8). The eluted fractions were immediately collected and neutralised using IM Tris-HCl (pH 8.0). Approximately 10 mg of IgG were purified and its purity was tested by sodium dodecyl-sulphate polyacrylamide gel electrophoresis (SDS-PAGE) and the uv-visible absorbance spectrum. Purified buffalo IgG (2 mg) was mixed with incomplete Freund's adjuvant and used to immunise three male fourmonth-old New Zealand white rabbits by intramuscular and subcutaneous injections. The rabbits were boosted three times, with the procedure repeated every six weeks. Ten days after the final boost, 10 ml blood samples were taken from each animal, and the serum was separated after sample clotting at 30°C for 30 minutes and centrifuging for 10 minutes at 10,000 g. For purification, an affinity column was prepared by conjugating approximately 13 mg of buffalo IgG to a 1 ml HiTrapNHs-activated support (Pharmacia-Biotech) according to the manufacturer's instructions; the conjugation yield was 94 per cent. The affinity column was used to purify rabbit IgG antibuffalo IgG from the sera collected from the immunised rabbits. Serum (approximately 15 ml), diluted 1:1 with a 75mM Tris-HCl buffer (pH 8.0), was loaded onto the affinity column, and the same buffer was used to equilibrate and wash the column, then elution was carried out using 100mM glycine-HCl, 500mM sodium chloride buffer (pH 2.7). The eluted fractions were immediately collected and neutralised with 1M Tris-HCl (pH 8.0). After purification, 9 0 mg of rabbit IgG anti-buffalo IgG were found to be pure both by SDS-PAGE and uv-visible absorbance spectrum analysis. To conjugate the IgG to horseradish peroxidase (HRP) a kit (Pierce) was used. The purified polyclonal rabbit IgG antibuffalo IgG (9.0 mg) was extensively dialysed against 0.1M sodium bicarbonate, 0 9 per cent sodium chloride conjugation buffer (pH 9.5), concentrated to 3 mg/ml, then incubated Veterinary Record (2001) 149, 88-90