Foodborne Illness Part II: How Do Scientists Track Outbreaks?

Feb. 8, 2019

If you’re like 90% of Americans searching frantically for any scientific excuse to avoid eating vegetables, you might refer to last year’s multitude of produce-related outbreaks of foodborne illness. 2018 was a particularly dangerous year for leafy greens, featuring 2 large, multi-state E. coli outbreaks associated with romaine lettuce, which sickened nearly 300 people. Additionally, 2 separate, unrelated Cyclospora outbreaks—one linked to McDonald’s salad mix and the other associated with vegetable trays—collectively affected more than 750 people. Though the romaine lettuce outbreaks were among the most highly publicized, vegetables were not the only offenders. Chicken, turkey, beef, ham, and even Honey Smacks Cereal, also made the Center for Disease Control and Prevention (CDC)’s Multi-State Foodborne Illness Outbreak list, which reached an unfortunate 10-year high last year.

As the frequency of outbreaks seems to be rising, you may be wondering what the United States government plans to do to correct the dangers that lurk within our food supply. In response to the first 2018 romaine lettuce outbreak in June, FDA commissioner Dr. Scott Gottlieb wrote in a June 28 press release, “Our ability to identify outbreaks has dramatically improved due to new information technologies and laboratory techniques. This is a view that we share with other experts, including the CDC, who have indicated that as these new methods are employed to protect the public from outbreaks, paradoxically the number of outbreaks may increase since we are now able to identify problems that had previously been invisible to us.“ In other words, contrary to how it seems, the number of outbreaks may not actually be increasing each year, and could actually be decreasing (especially following the transformative introduction of the 2011 Food Safety Modernization Act). Instead, because we have better tools for detecting outbreaks, we are finding more.

Currently, pulsed field gel electrophoresis (PFGE) is considered the “gold standard” microbial identification method used by the CDC during outbreaks. Here, bacteria are encased in an agarose plug and lysed, releasing their DNA. Then, DNA fragments are separated based upon size using an electric field and run on a gel to visualize the DNA fragmentation pattern (Figure 1). The pattern produced, referred to as a “fingerprint”, is typically very specific to each strain, though banding patterns can be difficult to interpret and not objective. PFGE is also slow and labor-intensive, so the agency is working to replace PFGE with whole genome sequencing (WGS), upon which the advanced technologies that Dr. Gottlieb’s memo mentions above mostly rely.

Figure 1 shows an example of different “DNA fingerprints” obtained using PFGE. The banding patterns are typically unique amongst different species of foodborne-illness causing bacteria.
Figure 1 shows an example of different “DNA fingerprints” obtained using PFGE. The banding patterns are typically unique amongst different species of foodborne-illness causing bacteria.

 
WGS, as the name implies, reveals the entire genome of an organism, allowing all of its genomic DNA to be analyzed. WGS has become significantly less expensive, easier, and more accessible with the development of benchtop sequencing instruments and kits. In the case of an outbreak, after microbes are isolated from the suspected source, sick patients and potentially the food supplier, their genomes can be compared for similarity.  The organism’s DNA may contain unique mutations, polymorphisms, and genomic rearrangements, which can even cause variability in the microbial genomes of microbes from the same species. Sequences can be compared among isolates from human patients and potentially contaminated foods, including the individual ingredients that comprise complex foods like pasta salad and tahini. This makes WGS a highly specific tool for identifying pathogens and their sources.
 
In a June 2018 press conference, Dr. Gottlieb stated that, “During the 2015 listeriosis outbreak associated with ice cream, clinical product, and environmental samples, [the] isolates yielded many different pulsed-field gel electrophoresis (PFGE) patterns. WGS analyses demonstrated that isolates from different manufacturing locations were unrelated to one another.” In other words, while PFGE can discern unique patterns in bacterial DNA, WGS helps to clarify inconclusive PFGE results.
 
The FDA has also demonstrated commitment to furthering WGS for tracking outbreaks through the development of its GenomeTrackr network, established in 2012. GenomeTrackr is an open-source database of WGS data obtained from common foodborne pathogens, which assists in the rapid identification of organisms most likely to be implicated in a foodborne illness outbreak. While it consisted of only 10 federal and 4 state labs at its inception, 63 labs from government, academia, and clinical institutions across and outside of the United States now contribute to expanding GenomeTrakr’s database. Within the network, more than 5,000 isolates undergo WGS each month, which allows the FDA to publicly distribute large amounts of WGS data across government, private, and public entities to aid in rapid identification of outbreaks (Figure 2). In combination with the CDC’s PulseNet database, which collects, analyzes, and aims to link related groups of patient isolates, the FDA’s GenomeTrackr can help to match isolates from sick patients with contaminated foods and consumer products.

Figure 2 shows the number of reference sequences per pathogen that have been uploaded to the GenomeTrackr database. The organisms represented in the database are among the most common (Salmonella, Campylobacter, E. coli, Listeria, Vibrio) and/or most life-threatening (Cronobacter) causes of foodborne illness. Salmonella strains are the most widely sequenced strains, though the database is also acquiring an expanding collection of E.coli/Shigella sequences.
Figure 2 shows the number of reference sequences per pathogen that have been uploaded to the GenomeTrackr database. The organisms represented in the database are among the most common (Salmonella, Campylobacter, E. coli, Listeria, Vibrio) and/or most life-threatening (Cronobacter) causes of foodborne illness. Salmonella strains are the most widely sequenced strains, though the database is also acquiring an expanding collection of E.coli/Shigella sequences.

Though WGS technology is still in its salad days, and its full potential is still being realized, it has a good chance of helping us to get our daily servings of greens without fear in the future.  It’s important to note that despite their association with microbial contamination, the risk of contracting foodborne illness is still not high enough to warrant forgoing vegetable consumption altogether, as long as food storage and handling precautions, such as washing produce before consumption and temperature control, are properly exercised and there is not an active, ongoing outbreak. Sorry about that, supertasters.

Further Reading:

Foodborne Illness Part 3: How Does Salmonella Make Us Sick? ASM Blog.


Author: Rita Algorri

Rita Algorri
Rita Algorri is a freelance writer, Ph.D. candidate in Clinical and Experimental Therapeutics and master's student in Regulatory Science at the University of Southern California.