BALTIMORE - Molecular inversion probe (MIP) technology can be coupled with next-generation sequencing for targeted, multiplexed pathogen detection, according to researchers at the United States Army Medical Research Institute of Infectious Disease (USAMRIID).
In a study published last month in The Journal of Molecular Diagnostics, Timothy Minogue, deputy division chief of USAMRIID’s diagnostic systems division, and collaborators described a MIP panel that is compatible with both Illumina and Oxford Nanopore Technologies (ONT) sequencing workflows. The team reported comparable results between the two platforms for detecting bacterial, viral, and parasitic pathogens in clinical samples.
Developed nearly 20 years ago, MIPs are single-stranded DNA probes that can bind to target sequences in a genome, followed by "gap-filling" with DNA polymerase and ligation to form circular DNA. The DNA elements can then be amplified with a single set of universal primers.
According to Minogue, the main advantage of using MIPs is the ability to multiplex. Compared with multiplex PCR, which requires different sets of primers for different target organisms, MIP enrichment uses universal primers for all amplifications, removing the PCR competition, he explained.
Meanwhile, unlike metagenomic sequencing, MIP-based sequencing can concentrate the reads on the desired regions of the genome, enhancing the assay’s sensitivity, specificity, and efficacy, according to Minogue. The technology is "taking the best of sequencing and the best of real-time PCR and finding a middle ground between them," he said.
The team designed a pathogen detection panel consisting of 94 probes targeting 17 viral pathogens and one parasitic organism, eight probes targeting variable 16S rRNA gene regions for bacterial taxonomic classification, as well as 71 probes targeting antibiotic resistance elements. After applying the panel to DNA derived from positive bacterial and viral blood cultures, the researchers sequenced the amplicons on an Illumina MiSeq sequencer and an Oxford Nanopore MinIon sequencer.
"What we were trying to do was to devise a sample prep methodology that would enhance the sensitivity and specificity of sequencing and reduce the overall bioinformatics burden so that it can be readily usable in the field setting," said Minogue, whose team has been working with MIPs for over a decade.
Overall, he said it was "sort of surprising" to find out that "both platforms performed relatively similarly" for this study. He said the initial hypothesis was that because Illumina sequencing tends to generate more reads, that platform would achieve "a higher or more sensitive detection" when paired with MIPs.
Indeed, Illumina sequencing did produce more reads than nanopore sequencing – the average number of reads per sample was approximately 500,000 for Illumina and 50,000 for ONT in the study. But in the end, both platforms performed "relatively similar in their sensitivity, specificity, and general statistics," Minogue said.
Nevertheless, the authors did discover some nuances between the two platforms, specifically when it came to 16S taxonomic classification. Of the 31 bacterial pathogens targeted with the MIP panel and subsequently sequenced, the authors found Illumina sequencing achieved a genus-level concordance of 96.7 percent compared to 90.3 percent with nanopore sequencing. While both sequencing platforms misclassified Klebsiella oxytoca as Enterobacter, nanopore sequencing also misclassified Burkholderia cepecia and Enterobacter aerogenes.
With regard to parsing out closely-related organisms, "where single nucleotide polymorphisms were the discriminating factor," Minogue pointed out, MIP-based Illumina sequencing delivered higher resolution than ONT sequencing.
Even so, Minogue said the team is planning to proceed further with the MinIon platform since the technology is "much more portable" and better suited for "less technology-centric" scenarios such as biosurveillance in the field. The work environment that Minogue's team is targeting is more austere. Therefore, he said, "the ONT platform is a much better fit for that."
However, the current turnaround time of the assay is around six to eight hours, not fast enough for rapid biosurveillance scenarios, where time-sensitive results are required from in a couple of hours to immediate, according to Minogue. Therefore, moving forward, his team is targeting to condense the overall workflow to under four hours.
ONT sequencing has a "distinct time advantage," he noted, because the sequencing data can be read in real time during the run, whereas with Illumina sequencing, it is hard to stop mid-run and do the analysis.
Shanmuga Sozhamannan, who works on biothreat detection systems for the Logistics Management Institute, a government consultancy, said in an email that he is "pleased that we are moving one step closer to making NGS-based infectious disease diagnostics a reality."
A major advantage of combining MIP with NGS is the "unparalleled multiplexing capability that is not achievable by the conventional real-time PCRs," he noted. In addition, "the MIP approach strikes a great balance of the best of amplicon sequencing with even more multiplex capability, while at the same time avoiding the complexity of metagenome sequence data handling in a regular clinical microbiology lab."
But there are also downsides to MIP-based sequencing. For one, he said, "it is targeted sequencing, so you need [to] know what you are looking for and [it] can be challenging or not well-suited when looking for unknown or engineered threats." Additionally, the technology may not be able to differentiate species or strains, he said.
A number of companies, such as Arc Bio, already offer commercial NGS-based assays for pathogen detection. However, microbial identification using MIPs with NGS appears to be novel.
While the authors of the current study said they want to explore the technology further in the biosurveillance context, Sozhamannan thinks it has potential for commercial use in diagnostics. "Although this assay is not there yet," Sozhamannan said about commercialization, "it’s the first step to achieving that goal."
To commercially compete with real-time PCR assays, which he said are currently the gold standard in infectious disease diagnostics, the MIP-based NGS approach still needs to improve in terms of cost effectiveness, limits of detection, preventing sample cross contamination within a sequencing run, turnaround time, portability, and insurance coverage, he noted.
Although the authors did not conduct a cost analysis for their study, they reported that sequencing reagents for both Illumina and ONT were "the most significant expense" for the workflow. Minogue said he sees the cost for MIP-based NGS will "fit in between" the cost of real-time PCR and metagenomic sequencing.
He said his team is currently not planning to commercialize the technology described in the paper. According to an USAMRIID spokesperson, the scientists at the agency "take the technology as far as we can and then hand it off for advanced development." Additionally, USAMRIID’s parent organization, the U.S. Army Medical Research and Development Command, "coordinates intellectual property licensing on behalf of all the subordinate laboratories."
"If successfully taken through the FDA approval process, this work opens up a new avenue of clinical diagnostics," Sozhamannan said. "This can be a trend setter for the development of other panels, such as a panel for biothreat detection. I can also imagine other areas of relevance for infectious diseases, such as for veterinary diagnostics and beyond."