Detection of latent forms of Mycobacterium avium subsp. paratuberculosis infection using host biomarker-based ELISAs greatly improves paratuberculosis diagnostic sensitivity

Bovine paratuberculosis (PTB) is a chronic granulomatous enteritis, caused by Mycobacterium avium subsp. paratuberculosis (MAP), responsible for important economic losses in the dairy industry. Current diagnostic methods have low sensitivities for detection of latent forms of MAP infection, defined by focal granulomatous lesions and scarce humoral response or MAP presence. In contrast, patent infections correspond to multifocal and diffuse types of enteritis where there is increased antibody production, and substantial mycobacterial load. Our previous RNA-Seq analysis allowed the selection of five candidate biomarkers overexpressed in peripheral blood of MAP infected Holstein cows with focal (ABCA13 and MMP8) and diffuse (FAM84A, SPARC and DES) lesions vs. control animals with no detectable PTB-associated lesions in intestine and regional lymph nodes. The aim of the current study was to assess the PTB diagnostic potential of commercial ELISAs designed for the specific detection of these biomarkers. The ability of these ELISAs to identify animals with latent and/or patent forms of MAP infection was investigated using serum from naturally infected cattle (n = 88) and non-infected control animals (n = 67). ROC analysis revealed that the ABCA13-based ELISA showed the highest diagnostic accuracy for the detection of infected animals with focal lesions (AUC 0.837, sensitivity 79.25% and specificity 88.06%) and with any type of histological lesion (AUC 0.793, sensitivity 69.41% and specificity 86.57%) improving on the diagnostic performance of the popular IDEXX ELISA and other conventional diagnostic methods. SPARC and MMP8 showed the highest diagnostic accuracy for the detection of animals with multifocal (AUC 0.852) and diffuse lesions (AUC 0.831), respectively. In conclusion, our results suggest that quantification of ABCA13, SPARC and MMP8 by ELISA has the potential for implementation as a diagnostic tool to reliably identify MAP infection, greatly improving early detection of MAP latent infections when antibody responses and fecal shedding are undetectable using conventional diagnostic methods.