Development of a single dilution ELISA to detect antibody to Dermatophilus congolensis in goat and cattle sera

A solid phase immunosorbent assay to detect antibodies to Dermatophilus congolensis in ruminant sera was developed to be used as a single dilution ELISA in large epidemiological surveys. Optimal conditions for the test are described. The use of blocking proteins to reduce non specific binding was necessary. Non fat dry cow milk and fetal calf serum were the only two efficient blocking agents out of six tested. Comparison of 4 antigenic fractions obtained after sonication and differential centrifugations of D. congolensis cultures showed that cell-wall (CW) or membrane (M) enriched preparations were more specific than a crude extract (CR) or a soluble (S) antigen. Whole spores and filaments performed poorly as antigens. The best sensitivity and specificity of the ELISA were obtained when the cut-off point of positivity was fixed at mean absorbance of negative sera + 2.58 sd. The specificity was then 97.6% either with M or CR antigen. The sensitivity was improved from 93.4% with CR to 98.2% with M antigen. Threshold values for a positive test varied between the 3 geographical areas tested. CW and M were also the most efficient antigens for discerning between serotypes of D. congolensis. The precision of the test was evaluated with CR antigen and expressed in residual expressed in residual coefficient of variation (CV). The precision was CV = 5.1% when each serum was titrated in duplicate and the antibody levels were expressed in absorbances. The expression of antibody levels in arbitrary standard units estimated from calibration curves reduced the precision (CV = 13.8%). Several methods were tested to decrease between plate variability but these did not greatly improve the reproducibility since it was shown that the main source of variation was within the plate.