The Gold Standard in Biothreat Detection for Biodefense and Biosurveillance is Real-Time PCR
For decades real-time polymerase chain reaction (PCR) has been the gold standard for biothreat detection across the enterprise engaged in biodefense. For example, the Department of Defense (DoD) Joint Program Executive Office for Chemical, Biological, Radiological and Nuclear Defense (JPEO-CBRND), Defense Biological Product Assurance Office (DBPAO) maintains a portfolio of 46 real-time PCR assays targeting 21 biological warfare agent organisms. These assays are produced under ISO 17034 Quality Assurance/Quality Control system and have been tested across different PCR platforms. These standardized assays are available to the US Government and their partners for research and development efforts or operational use (e.g., biosurveillance) for detection of these agents. Due to use of standardized assays and protocols the data are comparable temporally across times and spatially across labs and testing locations.
Next-Generation Sequencing
Since its advent in 2005, Next-generation Sequencing (NGS) technologies have made significant strides in genomics, marching toward adoption into various realms of biology and medicine including diagnostics/detection. However, the ubiquitous use of NGS in diagnostics (Dx) and detection/identification (Di) applications, as does PCR, has not been fully realized. The recent COVID-19 pandemic really accelerated the advanced development and maturity of NGS technologies. Adoption by numerous labs in all corners of the world was evident by the more than 15 million SARS CoV-2 genomic sequences deposited in public domain to date. This revolution has also spawned the widespread use of sequence data to unique applications such as detecting emergence of new variants, tracking pathogen mobility (genomic epidemiology) and in silico monitoring of diagnostic, therapeutic, and vaccine failures. It should be noted that NGS played an instrumental role in clinical surveillance, but Dx still remains a challenge.
One of the premier technologies that accelerated this democratization is Oxford Nanopore Technology (ONT) due to its portability, ease of use, minimal infrastructure needed to establish a sequencing capability, and potential for use in far forward and austere environments. The phrase “sequence anything, anywhere, anytime, by anyone” has become reality with ONT. This technology has been successfully demonstrated for sequencing samples in austere environments such as space (International Space Station), Antarctica, moving vehicles and by students in the field/resource limited settings.
Amplicon Sequencing Can Be a Bridge Between PCR and Metagenome Sequencing
In thinking about NGS, there are primarily 2 approaches: targeted and untargeted sequencing. The former (aka Amplicon Sequencing or Amp Seq) is relatively faster, less analytically intensive and easy to adopt in labs already doing PCR-based surveillance. The latter (Metagenome Sequencing) has analytical challenges and is not ready for widespread Detection applications in the field yet although that is the desired end goal. In addition, cost and time for metagenome sequencing can be prohibitive for routine (e.g., daily) surveillance.
Targeted amplicon sequencing is focused on sequencing unique parts of a list of pathogens. How is it different or better than the highly multiplexed PCR platforms such as BioFire® FilmArray® Panels? With Amp Seq we can do high-throughput (multiple samples processed simultaneously), high-multiplex threat detection with additional sequence information (useful in addressing false positive results from environmental near-neighbors), in a short time frame and analyze data in real time (e.g., ONT) to make actionable calls. In essence, PCR can address known known (list-based pathogens) whereas Amp Seq can potentially address known unknown threats (SARS-CoV-2 variants, antimicrobial resistant bacteria).
Metagenome Sequencing is Like Finding a Needle in a Haystack
Unbiased, untargeted sequencing of samples for biothreat detection is known as metagenome sequencing or Meta Seq. Other terms include: Threat agnostic sequencing; Agnostic Diagnostics; and Agent Agnostic Detection. Meta Seq changes the paradigm and can be exploited for identifying unknown unknowns: newly emergent threats natural or intentional/genetically modified threats that we have not seen before. However, it is like finding a needle in the haystack and comes with challenges and there is no highlighter solution.
Sagan Standards a.k.a. ECREE: Extraordinary Claims Require Extraordinary Evidence
One of the NGS technologies that has come to the forefront of far forward biothreat detection is Oxford Nanopore Technologies (ONT). Before embarking on metagenome sequencing as a routine approach for sample testing (e.g., biosurveillance scenarios), some of the myths about ONT and the ground truth should be understood.
Some of the claims surrounding ONT include:
- ONT fully democratized genome sequencing;
- ONT enables sequencing anything, anywhere, by anyone;
- Unbiased sequencing reveals everything in a sample; hence, it is far better than PCR or targeted sequencing and yields more “bang for the buck”.
- One can sequence a sample with ONT and identify the pathogen in minutes because of real-time sequence analyses capability.
- One can sequence a sample with ONT and identify genetically modified threats quickly because long reads may span the entire region of genetic modifications.
While the first two claims are supported by extensive evidence, the latter three claims are only true under certain defined circumstances.
What is the Hurdle? Caveats to Keep in Mind
As alluded to above, Meta Seq to identify a threat is akin to finding a needle in the haystack. An additional challenge is defining which part of the needle one wants to find. In this analogy, the needle is the threat organism, and the haystack is everything else in a sample. Complex samples can contain millions of innocuous organisms and host, plant, and other nucleic acid materials.
Finding the biothreat in this mix depends on a lot of factors. 1) How much of the threat organism is in the sample – the more of the threat organism present, the easier it is to identify it at deeper taxonomic level. 2) At what taxonomic level (which part of the needle) does one want to detect: PCR most often detects at genus/species level discrimination; one can also detect known genetic changes such as antimicrobial resistance (AMR) using a targeted PCR assay. If one wants to drill further down into lower taxonomic levels below genus/species, sequencing requires more initial biomass, longer sequencing runs, and more complex post-sequencing analyses.
Genetic modifications at single-nucleotide levels can have profound implications on phenotype such as virulence and AMR. However, sequencing to drill down to identify nucleotide-level variations is more complex and requires higher biomass of threat agent, some form of amplification of low-density organisms, or depletion of other genetic materials. Amplification/enrichment or depletion of background genetic material are considered targeted approaches, i.e., one needs a priori knowledge of what organisms to amplify/enrich, and for depletion of the background, what is not the organisms of interest.
Nucleotide-level variant detection requires more time to sequence (not happening in minutes or a few hours especially if the biomass is low). This longer time is needed to generate sequence data of better quality, with more breadth and depth of coverage and to process the data. It also entails having bioinformatic expertise on the back end and will likely be more expensive and may be prohibitive for routine daily use for surveillance. Nonetheless, Meta Seq may be useful in initial identification of a novel threat (discovery phase) and trigger early outbreak responses such as developing a traditional real time PCR or an Amp Seq assay for routine specimen screening and for implementing public health measures (e.g., SARS-CoV-2 pandemic response).
The Needs for a Metagenome Assay Implementation
What do we need for developing, validating and sustaining after successful deployment of a Meta Seq assay?
- Define Concept of Operations (CONOPS) and Requirements – what does one want to detect? In other words, which part of the “needle” does one want to detect?
- Understand the limitations in terms of time, cost, limit of detection, quality and quantity of data
- Develop the assay, workflow (protocols and SOPs) and validate across testing sites
- Develop data Standards in every step of the end-to-end process to generate reproducible sequencing results for making actionable calls with high confidence
- Monitor data trends
What to Watch for Actionable Calls?
Data Trends (Coincidence vs Causation)
How does one make an actionable call from very few sequences of read data of a suspect pathogen?
The point to address is whether the presence of reads is a mere coincidence or cause for concern. This is critical due to the need to suppress false alarms based on few reads. That is where “trends” come into play. Temporal and spatial trends of normal microbial flora and threat agents and near neighbors (similar to lessons learned from wastewater surveillance data during the COVID-19 pandemic) need to be captured over time and space so that workflows can be developed for depleting background sequences in vitro or in silico.
Data Context vs Content
Data without context can lead to false positives. Not only is the abundance of threat specific sequence read data important, but its context also matters (sample metadata). Here, data context refers to sample source, collection sites, and other metadata associated with the samples to include additional, situational intelligence if collected and available.
Data and process standards
Irrespective of which approach (Amp Seq or Meta Seq) one takes, the major gap is in “data standards/thresholds” for making actionable calls. In other words, how does one set thresholds with respect to specific outputs of a sequencing assay?
In real-time PCR it is a simple metric based on PCR cycle threshold (Ct) value. For each PCR platform and assay, a Ct threshold is set based on performance data, and true positive and true negative calls are made. Is there a threshold for metagenome assay based on read counts for a given pathogen? Also, will the standards change with technology, workflow and use case?
Are We There Yet?
As of now, there is no estimate on the limit of detection of metagenome sequencing to detect a threat at different taxonomic levels and at nucleotide levels to decipher genetic engineering. For any assay, analytical testing data such as sensitivity (limit of detection) and specificity are critical – they inform on assay efficacy. One had come to appreciate such data in COVID-19 tests – it is understood that rapid antigen tests are less sensitive than PCR tests. There is no comparison data for metagenome assay to gold standard PCR or any other test. Until such information is available, it is a futile exercise to deploy untargeted metagenome sequencing for routine biosurveillance or claim it is ready for generalized, operational use.
Article graphics by: Justin Ritmiller, Booz Allen Hamilton
ABOUT THE AUTHOR
Dr. Shanmuga Sozhamannan is the technical coordinator for the Defense Biological Product Assurance Office (DBPAO) at the JPEO, JPL CBRND Enabling Biotechnologies. In this role, he provides SME support for the development, fielding, and sustainment of reagents, assays, and biothreat detection systems used in national biodefense. He serves as the key interface between service laboratories, quality assurance test laboratories, and product end-users while maintaining positive government, industry, and interagency partnerships. He brings in diverse and extensive experience in molecular genetics, molecular biology and genomics working in academia, private industry, and government laboratories. Shanmuga obtained his Ph. D degree in Bacterial molecular genetics and has published over 80 peer-reviewed, scientific publications and several book chapters and holds several patents. Over the years, his research work focused on bacterial and bacteriophage genetics, molecular aspects of pathogenesis, and creating phage-based and molecular diagnostic assays. He has also used next-generation sequencing technologies for various applications, ranging from pathogen detection and characterization to development of genomics tools for diagnostics. He can be reached on LinkedIn or via email at Shanmuga.Sozhamannan.ctr@army.mil
REFERENCES
We are addressing the question on setting data standards for Meta Seq in a Working Group comprised of Inter Agency/academia/private partners via the AOAC SPADA program. Some documents pertaining to Amp Seq standards and publications on Amp Seq for biodefense application are:
Player R, et al. Comparison of the performance of an amplicon sequencing assay based on Oxford Nanopore technology to real-time PCR assays for detecting bacterial biodefense pathogens. BMC Genomics. 2020 Feb 17;21(1):166. doi: 10.1186/s12864-020-6557-5. PMID: 32066372; PMCID: PMC7026984.
Player R, et al. Optimization of Oxford Nanopore Technology Sequencing Workflow for Detection of Amplicons in Real Time Using ONT-DART Tool. Genes (Basel). 2022 Oct 3;13(10):1785. doi: 10.3390/genes13101785. PMID: 36292670; PMCID: PMC9602318.
Author disclaimer: References to non-federal entities or their products do not constitute or imply Department of Defense or Army endorsement of any company, product or organization.