Purpose: This study evaluated GSS performance as a molecular serotyping and sub-typing tool for Salmonella.
Methods: Samples of bacterial strains were prepared by an automated sample preparation module that carried out extraction, restriction digestion and sparse labeling of genomic DNA with sequence-specific, fluorescent tags. The individual DNA fragments were microfluidically linearized and their fluorescence measured in an automated detection module to generate sequence-specific patterns, “traces”. Traces of all strains were then compared using proprietary algorithms that compute a similarity parameter, which was used to cluster strains on a GSS-generated phylogenetic tree. Sequence data from whole-genome sequenced strains were processed in-silico and added to the tree.
Results: A collection of more than 230 Salmonella strains kindly provided by FDA, representing the most frequently encountered serovars associated with human illness as well as isolated from food products, was used to test the strain typing capability of GSS. In general, strains clustered into serovar-specific branches on the GSS-phylogenetic tree, indicating a clonal lineage origin for most serovars. Polyphyletic lineage serovars like Newport formed more than one distinctly separated branch on the tree.
Significance: The results show that GSS could be used to reliably differentiate and type Salmonella serovars significant for public health and food safety and aid in epidemiological investigations with a time-to result of less than 5 h. Serovar-specific cladal organization of Salmonella strains on the GSS tree would help to allocate serovar designation to unknown strains based on their position and identity of close neighbors on the GSS tree.