Porcine epidemic diarrhea virus (PEDV), an enteric coronavirus related to transmissible gastroenteritis virus (TGEV), appeared suddenly in the United States in April 2013. Epidemic sow herd outbreaks, characterized by severe diarrhea, vomiting, and high mortality in nursing pigs for several weeks, continue to spread the disease. Virus can be detected in feces using a real-time PCR assay. Antibodies to the virus can be detected using an ELISA assay based on the nucleocapsid (N) protein. In house PEDV ELISAs are available through diagnostic laboratories to detect anti-PEDV antibodies and, recently, a commercial ELISA kit has become available (Biovet). In house ELISAs are able to detect antibodies to PEDV, but the sensitivity and specificity need to be improved. Our objective with this proposal was to create an improved ELISA with better sensitivity and specificity against PEDV. We also optimized the ELISA for use with milk, colostrum, and fecal samples. Previously, we were able to express and purify 4 PDCoV antigens; nucleocapsid (N), matrix (M), and the spike protein subunits (S1 and S2) and ELISA assays were developed for each antigen. Colostrum, milk, and fecal samples were evaluated on all 4 ELISAs to determine antibody reactivity. To improve the ELISA, we expressed large amounts of protein, refolded the protein, and coupled the protein to an HRP substrate. We then examined the refolded protein for increased reactivity in the original ELISA format and we evaluated an alternative sandwich ELISA format. ELISA analysis of colostrum and milk samples using all 4 antigens and both IgG and IgA revealed that for IgA, antibodies to the spike antigens (S1, S2) were much more prevalent than anti-N antibodies. For IgG, mainly anti-N and S2 antibodies were present. Few fecal samples were antibody positive and when looking over a time course the data was un-interpretable. However, the key antibodies to look for in feces seem to be anti-N and anti-S2 of the IgA isotype. Further examination into sampling of feces at different times after infection needs to be performed for further optimization of this ELISA. Examination of antibody reactivity to refolded proteins compared to non-refolded protein preparations showed slight increases in ELISA OD values, but no statistical improvement in the ELISA sensitivity or specificity. HRP labeling was only achieved with N protein, but it did not specifically detect antibodies for unknown reasons. Other proteins could not be labeled since they were not soluble in the labeling reaction. However, a new antigen, which contains the putative antigenic regions of the spike protein is being expressed and purified for use in the ELISA and the alternative sandwich ELISA. Overall, an optimized ELISA was produced for use with milk and colostrum samples. A fecal ELISA was produced and with a time course of fecal sampling, this ELISA can be further evaluated for analysis of antibodies in fecal samples. Lastly, the alternative sandwich ELISA was not successful in improving sensitivity and specificity of the PEDV ELISA, but further examination into this problem is underway.

Dr. Cheryl Dvorak and Dr. Michael Murtaugh
dvora013@umn.edu, murta001@umn.edu