Whole Genome Sequencing Updates from the USDA Public Meeting in Washington, D.C.
On October 26 and 27, the USDA hosted a public meeting in Washington, D.C. that brought together domestic and international stakeholders from industry, government, and academia to discuss the current landscape of Whole Genome Sequencing (WGS) in relation to food safety. A variety of topics were presented at the meeting including:
- A general overview of WGS and recent technological advances in the field.
- The use of WGS by the CDC, FDA, USDA, and state health departments.
- Perspectives on the use of WGS for antimicrobial resistance monitoring.
- Building capacity and standard methods for performing WGS.
- Perspectives on WGS from international governments, the produce industry, the meat industry, consumers, and third-party laboratories.
- USDA policy development for WGS.
Further information about the public meeting, links to audio recordings, and slide show presentations will be made available on the meeting website. For more articles from FSNS on how WGS works and possible uses, visit the FSNS news website.
One major topic discussed at the public meeting was the shift from Pulsed-Field Gel Electrophoresis (PFGE) to WGS as the gold standard method for foodborne pathogen subtyping. PulseNet, a national laboratory network that connects foodborne illness cases to detect outbreaks, is switching to a new database that will use WGS information as an alternative to PFGE fingerprints. At the present time, both PFGE and WGS are run in parallel on each isolate that is entered into PulseNet. However, the CDC is expected to phase out PFGE and use WGS exclusively by no later than 2019. The key advantage that WGS has over PFGE is that it essentially reveals the entire DNA sequence of a microorganism. Access to WGS data allows better discrimination between the different strains of a particular species, especially strains that are historically difficult to distinguish by PFGE. Better strain discrimination generally shortens the time to find the source of the outbreak strain and consequently helps prevent further illnesses from occurring.
The robust data obtained from WGS can also be used to detect genes that are responsible for antimicrobial resistance (AMR). Thus, the use of WGS to replace the traditional methods used by the National Antimicrobial Resistance Monitoring System (NARMS) was another major topic of the meeting. A NARMS public meeting was also held in Washington, D.C. on October 24 and 25, so the discussion on the relationship between AMR and WGS was timely. At present, NARMS uses traditional broth dilution techniques to assess the resistance of an isolate to different antimicrobials. This traditional susceptibility data is important to assess whether an isolate is a public health concern based on whether it can be treated with certain antibiotics. Research was presented at the public meeting that demonstrated that WGS data had a high correlation (> 90%) with these traditional broth dilution techniques for Salmonella. Thus, the plan for the regulatory agencies to transition traditional phenotypic methods for AMR monitoring to WGS data appears to hold merit and will likely be pursued.
Another major topic of discussion was the development of the GenomeTrakr database. The FDA is spearheading the effort to organize WGS genomic data along with isolate metadata (i.e. geographic location, time information, sample type, etc.) into the GenomeTrakr database, which is hosted by the National Center for Biotechnology Information (NCBI). At present, the GenomeTrakr database has over 176,000 isolates, and this number continues to increase. Over time, GenomeTrakr will be referenced to a greater extent in order to differentiate sources of contamination and observe patterns in the metadata that are associated with foodborne illness outbreaks. However, a recurring theme from many of the speakers was that the metadata must be of high quality to be useful. Poor quality metadata will only convolute outbreak investigations and slow down response times. Thus, it was proposed that universal standards should be adopted across the globe to ensure that isolate metadata is collected and entered into databases properly. Organizations such as the International Organization for Standardization (ISO) will play a critical role in this effort. To view and sort the isolates that have already been sequenced and uploaded to GenomeTrakr, along with their associated metadata, NCBI has created a browser known as NCBI Pathogen Detection that is located at this website.
Historically, the idea that two isolates may be arbitrarily linked together based solely on WGS matches without epidemiological data to support the linkage has been a major concern for food industry members. It was repeatedly stated by regulatory and public health officials that epidemiological information must be present in order to establish a link between two isolates that are matched by WGS. Thus, even though a food or environmental isolate may closely match a clinical isolate by WGS, it was reiterated that there should be no automatic linkage made between the two isolates without epidemiological data to reinforce that linkage.
However, it still remains unclear how the regulatory agencies will view time as an element for epidemiological linkages. There were several discussions on whether statutes of limitations should apply to foodborne illness investigations, since linkages could be discovered between clinical isolates collected fifteen years ago and environmental isolates that are currently being collected from a food production facility. Although there was no consensus from the regulatory agencies on this topic, it is one that will likely receive a great deal of attention going forward.
Another concern voiced by several food industry representatives involves access to WGS data, and other data generated from Next Generation Sequencing (NGS) technologies, during inspection activities. If a food industry member performs WGS on isolates collected from their products and processes, and this reveals a problem, such as persistence of a Listeria monocytogenes strain in the facility, industry members feel that they should have the latitude to use this information in a productive way to proactively fix the problem rather than having it used against them when undergoing inspection. Also, other forms of NGS data, such as 16S metagenomics data, are helpful for identifying causes of spoilage and mapping the flow of microorganisms through a production process. However, generating this NGS 16S metagenomics data may also reveal the presence of other questionable microorganisms. Rather than having regulatory agencies scrutinize this information during inspections, industry members would like to be able to appropriately address the risk of these questionable microorganisms on their own and still use this information for the intended purposes of correcting spoilage issues and preventing introduction of certain types of microorganisms into the process.
In order to do this, regulatory officials noted that safe harbor strategies may need to be established so that industry members do not feel discouraged from using these NGS technologies to proactively improve their processes. It was mentioned that there are some safe harbors, such as VoluntaryNet at the University of Georgia, that allow food industry members to upload their WGS information and compare their own isolates to one another and to other isolates in databases. VoluntaryNet is anonymous and has firewalls that block regulatory agency access. However, it was agreed that there are still many legal ramification questions about VoluntaryNet and other proposed safe harbors that remain to be answered, and that further input from legal experts is needed.
Overall the main take home messages from the public meeting were that:
- Agency collaboration and sharing of WGS data must continue.
- There is a need for standardization and harmonization of data across all WGS networks. In addition, web-based, easy-to-use tools are needed to process and interpret the data.
- WGS data is powerful, but it has to be used in context with other metadata, and this metadata must be of good quality.
- Databases need to be representative of the natural distributions of food and clinical isolates. Currently there may be some bias in the databases.
- Safe harbors need to be created to allow the food industry to advance their food safety systems without the fear of being penalized.
- Policies need to be established around retrospective investigations.
- There are numerous policy questions that remain to be answered, and that input from legal experts is greatly needed.