Purpose: The objective was to assess the utility of PTS as a Salmonella serotype identification tool in comparison to traditional serotyping by the KW scheme.
Methods: 112 Salmonella isolates from the Silliker culture collection, including 38 isolates from food products, were used in this study. Samples were prepared for PTS analysis following manufacturer protocol. Twenty blinded samples from routine testing were also screened with the PTS assay; identity of these isolates was determined by traditional serotyping.
Results: The PTS and classical serotyping results matched for 68 isolates (61%); 6 additional isolates were correctly matched after being checked manually. Similarly, 12 of 20 blinded samples matched the traditional serotyping result (60%). Sixteen of 46 isolates that did not match were isolated from a variety of foods (e.g., peanuts, pasta, cheese powder, M&B meal, liquid egg). Six serotypes (Bovismorbificans, Kentucky, Newport, Ohio, Reading, Worthington) showed new pattern codes despite being represented among 100 serotypes in the validated PTS database. Considering the higher discriminatory power of PTS, two or more spot patterns can occur within a conventional serotype.
Significance: Currently, PTS can detect >300 Salmonella serotypes and has validated identification of >100 serotypes. However, the majority of patterns in the PTS database have been limited to isolates of European origin. Efforts to expand the database by analyzing isolates from different geographic locations, including strains isolated from food products, will improve robustness and performance of the methodology. An advantage of PTS is that extension of serotypes does not necessarily require a change to the DNA markers on the microarray, but that the score can be improved by simply adding more isolates to the database.