Whole genome sequencing and analysis permits improved subtyping of Salmonella enterica during foodborne outbreaks

Each year, Salmonella causes an estimated 93 million gastroenteritis cases and 150,000 deaths globally. The lack of genetic discrimination among Salmonella strains, when analyzed by pulsed field gel electrophoresis (PFGE), is a major problem during foodborne outbreak investigations, usually leading to failure to promptly identify the source of exposure and to control outbreaks. Thus, tracing the source of Salmonella during foodborne outbreaks requires more specialized genomic approaches capable of typing specific strains within Salmonella serovars. We evaluated the use of whole genome sequencing (WGS) and bioinformatics analysis as an alternative to PFGE for analysis of Salmonella enterica during foodborne outbreaks. PFGE and WGS were performed on 162 Salmonella enterica serovar Heidelberg isolates. S. Heidelberg isolates were sequenced using the Illumina MiSeq platform with 300 base pair paired end sequencing. De novo assembly of the raw sequence reads was accomplished, and the draft sequences were analyzed using whole genome and core genome multi locus sequence typing (wg/cg MLST), core genome single nucleotide polymorphism (cgSNP) analysis, prophage sequence typing, and CRISPR analysis. cgSNP and wgMLST analysis proved to be the most discriminatory, and resolved multiple isolates of S. Heidelberg, originally grouped into a single PFGE type into at least four separate clusters, each representing a single foodborne outbreak. cgSNP and wgMLST seem to be the most promising approaches to use during outbreak analysis of S. Hedeilberg. Ongoing standardization of these methods in the context of PulseNet surveillance, will lead to routine adoption during foodborne outbreak investigations.