Field evaluation of a novel differential diagnostic reagent for detection of Mycobacterium bovis in cattle

Abstract

In the search for improved tools with which to control bovine tuberculosis, the development of enhanced immunodiagnostic reagents is a high priority. Such reagents are required to improve the performance of tuberculin-based reagents and allow the discrimination of vaccinated cattle from those infected with Mycobacterium bovis. In this study, we identified the immunodominant, frequently recognized peptides from Rv3873, Rv3879c, Rv0288, and Rv3019c, which, together with peptides comprising the current lead diagnostic antigens, ESAT-6 and CFP-10, were formulated into a peptide cocktail. In a test of naturally infected cattle, this cocktail was significantly better than tuberculin was for identifying skin test-negative animals with confirmed bovine tuberculosis. In addition, the specificity of this cocktail was not compromised by Mycobacterium bovis BCG vaccination. In summary, our results prioritize this peptide-based, fully synthetic reagent for assessment in larger trials.

FIG. 1.

Identification of frequently recognized peptides. Positive response was ΔOD₄₅₀ (OD₄₅₀ with peptide minus OD₄₅₀ medium control) of >0.1. Results are expressed as responder frequencies (proportion of animals tested responding to a particular peptide). Calves were infected with a field strain of M. bovis, and blood was sampled 16 to 21 weeks postinfection. (A) IFN-γ responses induced by individual peptides from the second pool of Rv3873 (representing amino acid residues 89 to 188) tested with Rv3873-responsive M. bovis-infected cattle (n = 14). (B) IFN-γ responses induced by individual peptides from the first pool of Rv3879c (representing amino acid residues 1 to 92) tested with Rv3879c-responsive M. bovis-infected cattle (n = 16). (C) IFN-γ responses induced by individual peptides from Rv0288 tested with Rv0288-responsive M. bovis-infected cattle (n = 24). (D) IFN-γ responses induced by individual peptides from Rv3019c tested with Rv3019c-responsive M. bovis-infected cattle (n = 24).

FIG. 2.

Sensitivity of diagnostic reagents in relation to disease status. Bars represent the area under the ROC curve, demonstrating the ability of ESAT-6/CFP-10 peptides (darkly shaded), cocktail 1 (black), and the tuberculins (lightly shaded) to identify M. bovis-infected animals grouped by their skin test statuses: SICTT +ve, SICTT -ve, and SICTT IR. All animals were visibly lesioned and culture positive for M. bovis. ROC curve analysis and statistical calculations were performed using Analyse-It software (Leeds, United Kingdom). *, area under ROC curve generated using cocktail 1 is significantly different from that using PPD-B and PPD-A (P = 0.0372).