Advancements in food genomics, particularly high-throughput or next-generation sequencing, are allowing scientists and regulators to detect and identify foodborne pathogens with unparalleled speed and accuracy. By the end of this year, laboratories at FDA, USDA’s Food Safety and Inspection Service, and the CDC will rely almost exclusively on whole genome sequencing (WGS) as their main surveillance tool to differentiate strains of bacteria and identify related clusters of infections.
You Might Also Like
Explore This IssueAugust/September 2018
Also By This Author
But as the world’s food supply becomes increasingly interconnected, there is also a growing recognition that WGS and related technologies must become more widespread, particularly in economically developing food-producing countries. The U.S., for example, imports foods from about 200 different countries, including 90 percent of our seafood and at least half of our fresh fruit, depending on the season.
“This is why we’ve focused on developing and using advanced technologies and science to enhance our efforts in preventing food safety problems and improve our response time when incidents occur,” says FDA Commissioner Scott Gottlieb, MD. “We need to invest even more in these efforts, and in the tools to track and trace contaminated food in the supply chain.”
Today, high-throughput sequencing “has become progressively faster and cheaper, providing higher quality and longer and larger number of reads, resulting in better resolution and reproducibility,” says Behzad Imanian, PhD, research assistant professor at the Institute for Food Safety and Health (IFSH).
“The food industry is showing great interest in this technology and many of its members have already invested in, experimented with, or even implemented it in their research and development procedures,” Dr. Imanian tells Food Quality & Safety magazine.
But expanding food genomics will require overcoming numerous challenges, including cost, training, and data handling and storage. There are also unresolved legal, privacy, and technical standards issues, some of which still persist in the U.S. Nevertheless, governments and private researchers are working to expand WGS and related technologies domestically and internationally.
“In the future, most likely a global WGS system will enable very specific and almost real-time identification of all microorganisms,” says Jorgen Schlundt, PhD, steering committee head of Global Microbiological Identifier (GMI), a consortium of private scientists and clinicians from more than 40 countries. GMI advocates for a worldwide, interconnected platform of genomic databases to create “one harmonized and revolutionary tool supportive of international health regulations and global public health.”
Researchers at FDA’s Center for Food Safety and Applied Nutrition (CFSAN) are making WGS bioinformatics tools and data freely available through an open-source, cloud-based platform called GalaxyTrakr. This new platform “provides a user-friendly and cost-effective solution for industry and other partners to address their bioinformatic needs,” says James Pettengill, PhD, a biostatistics and bioinformatics geneticist at CFSAN. Currently about 140 researchers in 42 locations worldwide are using GalaxyTrakr, with 15 new users signing on each week.
WGS can map the genetic sequence of pathogens and other organisms with such precision that researchers can distinguish between different strains of a bacterium or even slight variations by geography within the same strain. “Just as the Hubble telescope revealed previously unsuspected star clusters in the darkest areas of the sky, so public health surveillance using methods with better resolution can identify clusters of infection that were previously missed,” says Robert Tauxe, MD, director of CDC’s Division of Foodborne, Waterborne, and Environmental Diseases.
Prior to WGS, scientists used tools such as polymerase chain reaction (PCR) and pulsed-field electrophoresis (PFGE) to genotype microorganisms for diagnostic subtyping. In addition to difficulties in standardization, “these pre-WGS techniques were often laborious and time consuming, required highly trained personnel, and expensive equipment,” says David J. Lipman, MD, former director of the National Center for Biotechnology Information (NCBI) at the National Institutes of Health. “WGS overcomes many of these old problems.”
In PulseNet, the CDC-run network that connects public health and food regulatory agency laboratories nationwide, WGS this year has replaced PFGE as the primary method for detecting and investigating Listeria outbreaks. In the coming months, WGS will also be used for Salmonella, E. coli, and Campylobacter, says Heather A. Carleton, PhD, leader of the bioinformatics team at CDC’s Enteric Diseases Laboratory Branch.
Partially driving this change are the local and state clinical labs that increasingly supply data to PulseNet from culture-independent diagnostic tests (CIDTs), such as immunoassays and nucleic-acid amplified tests. While CIDTs are cheaper, faster, and easier to use than WGS to detect bacteria in sick patients, they are unable to determine the DNA subtype (“fingerprint”) or other characteristics necessary for PulseNet to detect outbreaks, track antibiotic resistance, or monitor disease trends.
Because of this, “PulseNet is preparing for a future without isolate culture,” Dr. Carleton told an IFSH symposium on food safety and high-throughput sequencing in May. First, researchers will employ a technique called core genome multilocus sequence typing, which can determine a bacterial isolate from the internal fragments of a small number of “housekeeping” genes. Second, they will use shotgun sequencing to identify and subtype both known and unrecognized pathogens. “The latter approach will leapfrog pathogen discovery and likely the identification of known and novel pathogens causing outbreaks of unknown etiology,” Dr. Carleton explained.
The U.S. government has also been upgrading public health laboratories that contribute data to PulseNet by equipping facilities with WGS equipment and training personnel. So far, more than 100 scientists in 46 states have been PulseNet trained and certified, Dr. Carleton said. “By the end of 2018, we anticipate that WGS will be the main PulseNet surveillance tool for detecting dispersed outbreaks caused by Listeria monocytogenes, Shiga toxin-producing E. coli, and Salmonella,” Dr. Tauxe said.
Further, the analytic methods have been harmonized with those used in FDA’s GenomeTrakr food testing network, Dr. Tauxe added. GenomeTrakr currently contains more than 200,000 pathogen genome entries in its open-source portal, housed at NIH’s NCBI, with more than 5,000 isolates being sequenced and added monthly. FDA is currently working to expand GenomeTrakr’s distributed network of laboratories internationally and make its reference database more widely available in other countries.
Other current food genomics efforts include characterizing pathogens from CIDTs using metagenomics (cataloging all the species in an environmental sample) and using rapid WGS platforms to mine pathogen adaptations that directly contribute to preventive controls requirements for industry. “Taken together, it is apparent that the role for WGS in microbiological food safety continues to grow as it integrates more and more into pathogen analytic workflow,” says Eric Brown, PhD, director of FDA’s Division of Food Microbiology.
WGS Limits and Challenges
The extent to which WGS and similar technologies will truly mitigate foodborne illnesses remains to be seen. As the ability to identify outbreaks has improved due to new technologies, “paradoxically, the number of outbreaks may increase since we are now able to identify problems that had previously been invisible to us,” Dr. Gottlieb said in a recent statement.
Indeed, improvements in pathogen and risk detection technologies are partially responsible for the more than doubling in the number of food recalls during 2004-08 compared to 2009-13, according to an April 2018 report from USDA’s Economic Research Service. As CDC’s Dr. Tauxe puts it, as WGS matures, “more dispersed outbreaks will be detected and investigated, and that on average, each will involve fewer cases.”
But improvements in WGS and genetic testing are not substitutes for traceability, as illustrated by this year’s E. coli outbreak linked to romaine lettuce. Researchers used WGS to link the strain of E. coli O157:H7 that sickened at least 210 people and killed five in 36 states to lettuce from the Yuma Valley region of Arizona. The bacterium, however, was never actually found on lettuce in fields or in commerce. Indeed, only after the outbreak was declared over in late June did Dr. Gottlieb announce that canal water used for irrigation appears to have been the source of contamination. How E. coli got in the water remains a mystery.
“The genetics does not help in determining the source, or which field it came from, or when it happened. That requires old-fashioned traceability and epidemiology,” says David Acheson, MD, former FDA associate commissioner for foods. “If you don’t have the bacteria on the commodity, the genetics doesn’t help you,” he tells Food Quality & Safety.
Also mystifying is that the contamination went totally undetected. “Given the amount of testing that occurs in the industry, both pre-harvest and throughout the supply chain, how could something have such a wide public health impact but not have been detected through testing?” asks Jennifer McEntire, PhD, vice president for food safety and technology at the United Fresh Produce Association.
“In the quest for more rapid tests that often have shorter enrichment times, are we limiting our ability to detect microorganisms? I don’t know, but I feel like the question needs to be asked,” Dr. McEntire tells Food Quality & Safety.
WGS Going Global
The World Health Organization and other international groups have been encouraging wide adoption of WGS technology to help manage infectious diseases, including foodborne ailments. While industrialized countries are using WGS for food safety management, its application in developing and transitional countries has been limited. According to a 2016 report by the United Nations Food and Agriculture Organization, barriers include cost, data storage, infrastructure requirements (such as high-speed Internet), legal issues, data ownership, sharing, and intellectual property rights, and sustainability.
Several of these issues are not unique to developing countries, and present challenges to the U.S. and other developed nations. “Food producers, researchers, and regulators are each affected by the current absence in standards around laboratory preparation and bioinformatics methods,” explains Kristen Beck, PhD, technical lead for the Consortium for Sequencing the Food Supply Chain and a research staff member at IBM Research-Almaden. “This makes standardization and comparative analysis very challenging,” she tells Food Quality & Safety.
An additional limiting factor is the cost of culturing microbial samples. “This will need to be orders of magnitude less than today’s per-sample cost to be usable at the scale of current routine testing,” Dr. Beck says. But as costs decline, this will become less of an issue and “will further enable technologies, such as culture-free sequencing of food microbiomes for hazard detection.”
Another challenge to adoption lies in data sharing—who can access potentially sensitive information derived from WGS, including information on virulence factors and antibiotic resistance. “How will these data be used and interpreted, and who will have access?” asks Claudia Narvaez, PhD, professor of food safety at the University of Manitoba. “And how will regulatory agencies interpret the findings, and how could this affect current regulations?”
While policymakers and regulators wrestle with these issues, researchers and equipment manufacturers continue to make advancements in genomics instrumentation. For example, researchers at the University of Georgia Center for Food Safety have developed a portable device that can shorten pathogen sequencing time from one or two days to one or two hours. Instead of culturing the sample, which can take 24-48 hours, the USB drive-sized device uses tiny magnetic beads coated with antibodies to separate pathogen DNA from the sample. It then amplifies the DNA and sequences it in real time.
In another example, researchers at Pittsburgh State University have created hybrid nanosensors composed of special iron oxide particles blended with optical dye, plus antibodies that specifically latch onto E. coli O157:H7 cells. By combining magnetic resonance imaging technology with fluorescence emission, they can quickly detect the pathogen in fluids, such as milk or lake water.
Even PCR technology, long a mainstay of food industry testing, is being upgraded. Bio-Rad Laboratories’ Droplet Digital PCR (ddPCR) technology allows for sample partitioning. “In traditional PCR, a single measurement is performed on a single sample. In ddPCR, a single sample is partitioned into thousands of nanosized droplets, allowing thousands of independent, single amplification events within that sample,” explains Mike Clark, the company’s international PCR group manager. This can help determine whether a single E. coli cell contains the virulence genes making it pathogenic or if those genes are just present in different cells within a food sample.
Advances are also occurring in metagenomics and food microbiome sequencing. “From sequence data of a single microbiome sample, you are provided with pathogen information, a microbial community snapshot, antimicrobial resistance, and host/food matrix composition,” explains IBM’s Dr. Beck. “When you compare multiple microbiomes, you are able to observe deviations in your supply chain that can indicate a hazard before it is allowed to become an issue.”
Expanding this concept, FDA’s One Health framework “holds that a connection exists between the environment and animal and human health,” says Eric Stevens, PhD, an FDA staff fellow. One Health encourages the sharing of WGS data across the various sectors related to food. “WGS can be used within supply chain management .