A New Enzyme-Linked Immunosorbent Assay (ELISA) for Vi Antigen as a Potential Diagnostic for Infections by Salmonella Typhi

BACKGROUND: Salmonella Typhi causes an estimated 22 million infections and 200,000 deaths worldwide annually. The lack of reliable diagnostic tests has significantly hindered disease control efforts. Detecting S. Typhi Vi antigen in urine using polyclonal or monoclonal antibodies in an ELISA has been previously tested, with some promising results. The establishment of a typhoid human challenge model allows screening of novel diagnostics using longitudinally collected samples. Therefore, a new ELISA has been developed and validated at the Oxford Vaccine Group to detect Vi antigen with a view to testing it on clinical samples from our challenge model. METHODS: A sandwich ELISA design was conceived using monoclonal anti-Vi IgM, polyclonal anti-Vi IgG and HRP-conjugated IgG antibodies. A series of optimisation experiments was performed to optimise each component using phosphate buffered solution spiked with purified Vi antigen derived from Citrobacter freundii as a substrate. The optimised assay was validated using Vi antigen spiked urine from healthy volunteers. Stored urine samples from participants from a previous typhoid challenge study at the point of typhoid diagnosis (S. Typhi bacteraemia and/or persistent fever ≥38°C) were also tested. These stored samples had been previously centrifuged and filtered to remove debris prior to freezing. RESULTS: The final optimised ELISA assay produced a reliable standard curve replicable across multiple assays with a lower Vi antigen detection limit of 0.24 ng/mL. Validation of the assay demonstrated accurate quantification of Vi antigen in urine samples spiked with known concentrations of Vi antigen. In contrast, Vi antigen was not detectable in stored urine samples from participants diagnosed with typhoid fever collected in a previous challenge study (n = 21). CONCLUSION: These experiments demonstrate an ELISA for Vi antigen that has potential to be utilised as a diagnostic tool for typhoid using urine specimens. The absence of detectable Vi antigen in urine samples could be due to the processing methods, low bacterial load in the challenge setting or technical limitations of the assay. The assay will be used prospectively on fresh urine from future challenge studies to better characterise assay performance. DISCLOSURES: All authors: No reported disclosures.