NEW YORK (GenomeWeb) – A team of researchers from the Lawrence Livermore National Laboratory has developed a next-generation sequencing panel to screen for pathogens.
The scientists recently published a proof of principle study in the journal BMC Research Notes, demonstrating that the panel can detect genomic material present in as little as 10 copies per milliliter in a complex sample.
Tom Slezak, who leads the informatics group at Lawrence Livermore, told GenomeWeb that the goal was to design a panel that would be cost effective and could identify a pathogen quickly and accurately even when at low levels. Now that the group has demonstrated the proof of principle, he said the next step is to expand the panel to either include more pathogens or more information, like antimicrobial resistance genes.
Slezak said the team is looking for sponsors within public health, biodefense, or infectious disease diagnostics, and depending on the partner, the panel could be tweaked to suit specific needs. For instance, he said, a doctor treating a patient in the intensive care unit would care about what antimicrobial resistance genes are present, but wouldn't be as interested in, say, the evolutionary history of the pathogen. On the other hand, an epidemiologist investigating an outbreak would want that information, he said.
The use of NGS-based panels targeting pathogens is on the rise. Researchers are finding that panels can identify relevant bugs quickly and cost effectively and are well suited for picking out small amounts of viral or bacterial DNA from a sample that is mostly human DNA.
For instance, separate groups from Columbia University and Washington University in St. Louis are developing viral capture panels designed to be more cost effective and faster than metagenomic sequencing for identifying viral infections in clinical samples. The panels they developed were also used for proof of concept and the teams are now working on refining them and potentially developing syndrome-specific panels.
In the Lawrence Livermore group's study, the researchers designed their panel using Thermo Fisher Scientific's AmpliSeq technology for sequencing on the Ion Torrent PGM. The team decided to focus on four microbes that the US Centers for Disease Control and Prevention include on their bioterrorism tier 1 and tier 2 lists: Bacillus anthracis, Yersinia pestis, Francisella tularensis, and Burkholderia pseudomallei.
To design the panel, the researchers created 9,799 organism-specific amplicons, which they submitted to Thermo Fisher for primer design. The amplicons were from specific genomic regions of the organisms and allowed for species-level identification. After optimization, 467 primer pairs were chosen for the panel.
The researchers constructed AmpliSeq libraries, using the One Touch 2 system.
Next, they created mock clinical samples — spiking in various concentrations of the pathogens into a representative human clinical sample — and compared results of their panel with shotgun sequencing. All samples were sequenced on the Ion PGM, using the 318 chip.
In every case, the panel resulted in a "sizable increase in the number of pathogen-specific reads with a concomitant reduction in human reads as compared to fragment libraries prepared from the same sample," the authors wrote.
For instance, from shotgun sequencing, as many as 90 percent of the reads were human, compared to less than 10 percent from the panel. At the same time, the proportion of pathogen-specific reads increased in the panel, compared to the shotgun sequencing strategy.
Slezak added that the panel resulted in not only an increase in the number of reads, but in the quality of information contained in those reads. The reads were "highly informative," he said. For instance, when analyzing a mixture that had about 100 pathogen genomes per milliliter, the unbiased sequencing approach resulted in some reads that could be mapped back to the pathogen, he said, "but many of those reads were not from informative regions and wouldn't be enough to give you species identification." That's the difference between being able to say the sample contains anthrax versus just being able to say that there is some kind of Bacillus in the sample, he said.
Using the panel, the team was able to get enough informative reads to identify the species when present at 10 genomes per milliliter. For unbiased sequencing, the team did not recover any pathogen reads below a level of 100 genomes per milliliter.
Slezak acknowledged that a targeted approach would not be able to identify any unknown pathogens. However, he said, the different methods could be complementary. "Look first for the nasties we know about in a cost-effective way using targeted amplification. If that doesn't yield any of the usual suspects, then we can go in and do deep unbiased sequencing," he said.
Since the first panel, he said the team has also developed a panel that analyzes 530 antimicrobial resistance genes. That panel would be geared more toward clinical use, rather than bioterror or epidemiology, he said, because it assesses resistance genes that are common across many species or that are carried on plasmids, so identifying the presence of specific genes wouldn't necessarily tell you the species.
"Some infectious disease doctors we've been working with basically said that when they're in the ICU and have a really sick patient, they don't necessarily care what the pathogen is, but they want to know what the resistance genes are in order to guide their antibiotic use," he said.
The 530-gene panel could potentially be run in conjunction with a species identification panel, or the two could be combined into one panel, Slezak said. There is just the trade-off with cost, which increases the more information is included in a panel, he said.
Slezak said he is now talking to potential sponsors in the biodefense and public health spaces, and depending on their needs, his team could tailor the panel. A commercial panel, particularly one that would be used for diagnostics, would also have to be fast. Slezak said it would be possible to perform the test in less than 24 hours from sample to answer.
Such a panel could also be designed for any sequencing technology, Slezak said. At the time the group began its work, the AmpliSeq technology was the only commercial amplicon-based method that allowed for customized panels, he said, but now there are other options, such as Illumina's TruSeq custom amplicon library kits. In addition, such panels could be run on microarrays or even by multiplexed PCR. "We'll see an interesting cost-performance arms race," he said. "It will boil down to how the samples are prepared, the targets identified, and the cost per sample."