Technical Note: Using enzyme-linked immunosorbent assays to evaluate humoral responses to vaccination against respiratory viruses in beef cattle

Abstract

Abstract This experiment evaluated humoral responses in beef calves vaccinated against parainfluenza-3 virus (PI3), bovine respiratory syncytial virus (BRSV), and bovine herpesvirus-1 (BHV-1) using serum neutralization (SN) tests or enzyme-linked immunosorbent assays (ELISA). Blood samples were collected from 50 overtly healthy Angus-influenced steers (183 ± 3 kg of body weight, 212 ± 2 d of age) on days 0, 21, 35, and 49 of the experiment. Steers were vaccinated against respiratory viruses on days 0 and 21. Blood was processed for serum collection and frozen in duplicates. One of the duplicates was analyzed for antibodies against BRSV, PI3, and BHV-1 using commercially available ELISA (IDEXX Switzerland AG, Liebefeld-Bern, Switzerland), and results reported as sample:positive control (S/P, %) ratio. The other duplicate was analyzed for antibodies against the same vaccine antigens via SN. This method reports results as titers, the greatest dilution that provides complete protection of the cells, which were transformed with base 2 log for statistical analyses. Samples were classified as positive for the presence of antibodies by SN if log-transformed titer ≥ 2 for all viruses, and by ELISA if S/P ratio ≥ 50% for BHV-1 or ≥ 20% for PI3 and BRSV. Day effects were detected (P < 0.01) for SN and ELISA across all vaccine antigens, as antibody levels increased after vaccine administration. Linear fits were detected (P < 0.01) across all vaccine antigens when regressing the SN and ELISA results; as SN titer increased, the ELISA S/P ratio linearly increased (P < 0.01). Kendall (τ) and Spearman’s rank (ρ) correlations were also detected (P < 0.01) between SN and ELISA results across all vaccine antigens. The SN and ELISA were very strongly correlated (ρ ≥ 0.83) for BHV-1 and PI3 and strongly correlated (ρ = 0.66) for BRSV. Cohen’s kappa coefficient for diagnosis agreement between methods was strong for BHV-1 and PI3 (κ ≥ 0.88), but weak (κ = 0.47) for BRSV. The sensitivity of the ELISA in yielding true positive results approached 100% across all antigens. The specificity of the ELISA in yielding negative results was satisfactory for BHV-1 and PI3 assays (84.0% and 88.5%, respectively) but not for BRSV (34.4%). Despite limitations in detecting true BRSV negatives, results from this experiment indicate that the commercial ELISAs tested herein can be used as surrogate for SN tests in quantifying humoral responses to vaccination against BHV-1, PI3, and BRSV in beef cattle.