Whole Genome Sequencing: Bridging One-Health Surveillance of Foodborne Diseases

Abstract

Infections caused by pathogens commonly acquired from consumption of food are not always transmitted by that route. They may also be transmitted through contact to animals, other humans or the environment. Additionally, many outbreaks are associated with food contaminated from these non-food sources. For this reason, such presumed foodborne outbreaks are best investigated through a One Health approach working across human, animal and environmental sectors and disciplines. Outbreak strains or clones that have propagated and continue to evolve in non-human sources and environments often show more sequence variation than observed in typical monoclonal point-source outbreaks. This represents a challenge when using whole genome sequencing (WGS), the new gold standard for molecular surveillance of foodborne pathogens, for outbreak detection and investigation. In this review, using recent examples from outbreaks investigated in the United States (US) some aspects of One Health approaches that have been used successfully to solve such outbreaks are presented. These include using different combinations of flexible WGS based case definition, efficient epidemiological follow-up, traceback, surveillance, and testing of potential food and environmental sources and animal hosts.

Keywords: animals; environment; food; investigation; one health; outbreak; whole genome sequencing (WGS); zoonotic.

Figure 1

cgMLST UPGMA tree of Lineage I isolate sequences the Listeria outbreak linked to ice cream. All clinical isolates and a representative sample of non-clinical product and production environment isolates are included in the tree. The range of allele differences are indicated at the branches of the tree and for clusters to the right of the tree.

Figure 2

cgMLST UPGMA tree of a representative sample of sequences of Salmonella ser. Heidelberg isolates displaying the full WGS diversity and representing the seven PFGE patterns from the outbreak associated with chicken produced by Company A. The range of allele differences are indicated at the branches of the tree and subclusters to the right of the tree.

Figure 3

cgMLST UPGMA trees of representative samples of sequences of Salmonella ser. I 4,[5],12:b:- (A) and Okatie (B). Isolates of ser. I 4,[5],12:b:- all displayed the same PFGE pattern. Isolates of ser. Okatie displayed three PFGE as indicated on the figure. Product isolate identifiers are indicated with rectangles around them. The range of allele differences are indicated at the branches of the tree and subclusters to the right of the tree.

Figure 4

cgMLST UPGMA tree of a representative sample of sequences of clinical and animal isolates of Campylobacter jejuni from an outbreak associated with contact to puppies sold in a specific pet store chain. Puppy isolates are marked with gray squares. The range of allele differences are indicated at the branches of the tree and clusters to the right of the tree.

Figure 5

cgMLST UPGMA trees of representative samples of sequences from one of the Salmonella outbreaks associated with turtle exposure. (A) shows the relationships between isolates of ser. Poona; isolates from the environment, from Luxemburg or of a PFGE pattern different from that displayed by all other isolates in the figure are indicated next to the strain identifier. (B) shows the relationship between isolates of ser. Pomona; the source of the isolates is indicated next to the strain identifier. Isolates 2493-2014 and 2492-2014 are from Chile. The range of allele differences are indicated at the branches of the tree and subclusters to the right of the tree.