Whole genome sequencing (WGS) of foodborne pathogens such as Listeria monocytogenes is globally on the rise in the food industry. It provides an improvement for proactive surveillance and source-tracking and allows in-depth genetic characterization of the pathogen. In the present study, the virulence gene profile including 99 virulence genes of 767 L. monocytogenes isolates from the Norwegian meat and salmon processing industry was characterized. The isolate collection comprised 28 clonal complexes (CCs) that occur globally. We additionally determined the in vitro virulence potential for 13 major CCs in human intestinal epithelial Caco2 cells using cocktails of three to six representative isolates. Our aim was ... More
Whole genome sequencing (WGS) of foodborne pathogens such as Listeria monocytogenes is globally on the rise in the food industry. It provides an improvement for proactive surveillance and source-tracking and allows in-depth genetic characterization of the pathogen. In the present study, the virulence gene profile including 99 virulence genes of 767 L. monocytogenes isolates from the Norwegian meat and salmon processing industry was characterized. The isolate collection comprised 28 clonal complexes (CCs) that occur globally. We additionally determined the in vitro virulence potential for 13 major CCs in human intestinal epithelial Caco2 cells using cocktails of three to six representative isolates. Our aim was to test whether the virulence potential could be predicted from the virulence gene profiles to estimate the application potential of WGS in risk assessment in the food industry. The virulence gene profiles were highly conserved within the individual CCs and similar among phylogenetically closely related CCs. We observed a CC-associated distribution of accessory virulence genes in addition to different length polymorphisms. Furthermore, we detected different premature stop codons (PMSC) in the inlA gene, which were mainly present in CC9, CC121 and CC5 isolates. Accordingly, CC9 and CC5 were unable to invade Caco2 cells, whereas CC121 showed moderate virulence potential due to the presence of an isolate harboring full-length inlA. The highest invasion was observed for CC403 and CC415, potentially due to the presence of accessory virulence genes. We demonstrated that CC14, which harbored full-length inlA, was unable to invade Caco2 cells due to a low inlA gene expression. Reconstruction of inlA in CC9 and CC121 isolates showed that without the presence of InlA on the cell wall (as detected in the CC9 isolates), invasion into host cells failed. Our study showed that predicting the virulence potential based on genetic virulence profiles provides valuable information for risk assessment in the food industry but also has its limitations. The mere presence of a full-length inlA gene is not sufficient for virulence, but gene expression and the presence of the protein on the cell wall is required for the successful invasion of L. monocytogenes into host cells. Moreover, hypovirulent CCs like CC121 were among the most abundant human clinical isolates in Norway despite harboring a PMSC mutation in the inlA gene. In conclusion, our study highlights that combining genotypic and phenotypic data is of great importance to improve the informative value of applying WGS in the food industry.
