Novel NGS-Based Assay Has Seemingly Limitless Virus Detection Potential

February 13, 2024

By Deborah Borfitz

February 13, 2024 | Scientists at the Center for Infection and Immunity (CII) at Columbia University Mailman School of Public Health have created a next-generation sequencing (NGS) technology that employs oligonucleotides (oligos)—short single strands of synthetic DNA—as bait for relevant targets of detection. The new agnostic diagnostic, known as virome capture sequencing platform for vertebrate viruses (VirCapSeq-VERT), outperforms existing NGS platforms with respect to speed, sensitivity, and cost in its ability to identify any virus that can possibly infect humans, including the unknown ones with pandemic-causing potential, according to Vishal Kapoor, the CII’s deputy director for laboratory medicine. 

The impressive performance of the public health screening and surveillance system was highlighted by a validation study published recently in the Journal of Clinical Microbiology (DOI: 10.1128/jcm.00612-23). VirCapSeq-VERT demonstrated near-perfect accuracy and could diagnose co-infections that went unreported by comparator assays, including single-agent polymerase chain reaction (PCR) and the BioFire RP FilmArray.  

This is all highly relevant to public health needs in the aftermath of the COVID-19 pandemic, which laid bare the inadequacies of targeted diagnostic tests limited to detecting the presence of specific and known pathogens, says Kapoor. PCR and antigen tests became widely used but long after the outbreak had begun, and many people had died. 

PCR tests work by amplifying (copying) small, isolated pieces of DNA whereas NGS amplifies and sequences every bit of nucleic acid regardless of whether that is a long chain of DNA or a tiny fragment of RNA. The big news here is that VirCapSeq-VERT can do that full-swoop detection with an enrichment process giving it a sensitivity on par with gold standard PCR assays. 

VirCapSeq-VERT targets all known vertebrate viruses, including those that may not yet have been recognized as threats to human health, domestic animals, and wildlife, Kapoor says.  

The validation study was conducted as part of the CII’s regulatory application with the Clinical Laboratory Evaluation Program (CLEP) enabling laboratories in New York State to be exempt from requirements of the federal Clinical Laboratory Improvement Amendments of 1988 (CLIA). VirCapSeq-VERT is now CLEP-authorized for diagnostic testing with plasma and respiratory specimens, and the CII has submission packages under active review for spinal fluid and multiple tissue types (brain, kidney, liver, spleen, and heart) with plans to expand the list with additional pathogens and specimen types including oral swabs, urine, and feces, says Kapoor.  

The CII is separately pursuing regulatory approval by the Food and Drug Administration (FDA) so that VirCapSeq-VERT might be deployed more universally for pandemic preparedness, he adds. In the past, testing was done as a favor to physicians at New York-Presbyterian Hospital (affiliated with Columbia University) to identify mysterious pathogens critically infecting their patients. 

Until VirCapSeq-VERT had regulatory approval in New York State, test results didn’t get included in patients’ medical chart at the hospital and the lab didn’t have a way to get reimbursed for the specimen testing, says Kapoor. Moreover, tests weren’t accepted in other states using New York accreditation as their benchmark. 

Pioneering Work

The idea to incorporate oligos into existing NGS technology occurred a decade ago to CII Director W. Ian Lipkin, M.D., professor of epidemiology, and Thomas Briese, P.D., associate professor of epidemiology. They then spent the next five years figuring out how to optimize assay output.  

Lipkin’s pioneering work in the world of infectious disease research began in the 1980s during the HIV/AIDS epidemic, says Kapoor. He identified the AIDS-associated immunological abnormalities and inflammatory neuropathy and showed that early life exposure to viral infections affects neurotransmitter function. 

Detection of pathogenic agents at the time was being done primarily through cumbersome methods, notably microbial culture and microscopy, he continues. Although the molecular structure of DNA was discovered decades earlier by James Watson and Francis Crick, “no one had really looked at genetic fingerprinting as a way to identify a disease-causing pathogen.” 

Lipkin was one of the early voices pointing to the limitations of Koch's postulates, developed in the 19th century, which require the presence of an infectious agent in cases with disease and its absence in those without and the isolation of the agent in pure cultures, says Kapoor. Many organisms cannot be readily grown in cultures and, even if they can, the association between infectious agents and disease is not always clearcut since infections can be present without causing disease and may require host risk factors. 

With the advent of PCR, microbial DNA sequences could be isolated and amplified for comparative analysis with known classes of microorganisms. PCR was successfully used by Lipkin, then at the University of California, Irvine, to link the outbreak of encephalitis along the Eastern seaboard in the late 1990s to the West Nile Virus originating in Uganda, Kapoor notes. Lipkin was subsequently recruited to Columba University where he established the CII in 2002. 

This is when Kapoor first worked at the CII. For three years, he was also a senior scientist for Qiagen, a multinational provider of sample and assay technologies for molecular diagnostics, which gave him the skillset he would later use to usher VirCapSeq-VERT through the rigors of the regulatory approval process. 

Public health and regulatory agencies are interested in the platform’s ability to detect novel viruses, and the first proof point came during the pandemic when the research team iteratively arrived at the conclusion that an emerging virus can be up to 40% different from any known virus and the assay would still pick it up, reports Kapoor. This was creatively demonstrated by an unpublished study where clinical isolates of SARS-CoV-2 were compared to an earlier coronavirus using nucleic acid probes along the entire length of the genome and provided enough information for a differential diagnosis. 

Pivoting From PCR

The platform initially doing multiplex testing at the CII was Lipkin’s MassTag PCR assay that modified the standard PCR approach by using mass spectrometry to enlarge the potential number of detectable targets from five to 25, Kapoor says. Given the limited number of fluorescent channels for detection, and the fact that signals were largely overlapping, physicians had been finding it difficult to decide on the handful of pathogens they might want to look for, he explains. 

With the multiplex PCR assay, DNA primers on the panel would get tagged with molecules of known masses (MassCodes) that bind to sequences in a sample and get amplified. After unbound primers and other impurities were removed, the PCR products were subjected to ultraviolet light to liberate the MassCodes for detection by a mass spectrometer. 

Word quickly spread about the assay’s development to scientists across the continental U.S., Europe, Africa, Asia, and Canada, who subsequently came to the CII to train with Lipkin and his team. It fell out of the limelight only when NGS technology came on the scene to make signal detection truly agnostic by allowing “unbiased massive parallel sequencing of any nucleic acid in a tube,” says Kapoor.   

Solexa (acquired by Illumina in 2007) was the first company to come up with a next-generation viral sequencing instrument, which the CII validated before it became commercially available. In the meantime, Kapoor says, honeybee colonies were collapsing, threatening to cause significant economic losses since many agricultural crops worldwide depend on their pollination services. 

PCR technology and DNA microarrays shed no light on the cause, so Lipkin decided to try the viral sequencing approach. In that study, researchers compared healthy and diseased U.S. colonies that revealed the presence of an obscure but lethal bee bug known as Israeli acute paralysis virus as well as multiple other viruses, bacteria, and fungi. These findings indicated that bees with honeybee colony collapse disorder were immunodeficient (Science, DOI: 10.1126/science.1146498). 

Thereafter, the CII team began sequencing to resolve multiple other mystery outbreaks or look for reservoirs from which pathogens might be transmitted, Kapoor continues. The approach provides clues to possible triggering causes that then require follow-up investigation. 

At the time, the biggest challenge facing NGS technology was deciphering the huge amount of data that it generated, he adds. Open-source bioinformatics software tools have since been developed to streamline the analysis and eliminate the painstaking manual labor. 

The other challenge was that the nucleic acids being sequenced far outnumbered the molecules available from a causative pathogen. This is what set Lipkin and his colleagues on the path of developing VirCapSeq-VERT with the creation of oligos. Although the first iteration of the platform involved manufacturing four million such molecules, the number more than halved by the time the first paper on VirCapSeq-VERT published in 2015 (mBio, DOI: 10.1128/mbio.01491-15).  

Oligos are a “gamechanger” in that they enable quick decision-making when patients are critically ill and especially if they are infected with a microbial-resistant bacteria where it is vital to pick the right antibiotic—including some that when taken early can prevent full-blown disease, says Kapoor. In a potential pandemic scenario, healthcare workers also need to know when patients should be isolated to prevent spread. 

Regulatory Hurdles

Both the New York State Department of Health and other regulatory agencies are looking for the same sort of performance characteristics from VirCapSeq-VERT, but federal agencies set the higher bar, says Kapoor. For example, the FDA requires that platforms be tested in combination with potential “interfering substances,” a list that includes throat lozenges that would seemingly be irrelevant when nasal swabbing for a respiratory virus. Figuring out how to appropriately spike a sample will require a degree of creativity.  

The good news, he adds, is that the research team has to date seen no interference from substances such as nasal decongestants on test results rendered by VirCapSeq-VERT from respiratory samples. 

FDA-required analytical validation has been completed, and the CII team is now negotiating with the agency over how many samples will be needed for clinical validation studies, says Kapoor. Unlike CLEP, the FDA won’t accept banked frozen samples but wants prospectively collected ones.

The agency may need to tamp down its expectations of NGS-based diagnostics, Kapoor says, given that the workflow is much lengthier and the data output vaster than with PCR and other targeted diagnostics around which regulatory policies were conceived. This is particularly true when one considers the virtually limitless capability of a tool like VirCapSeq-VERT, which has only once failed to detect a sought culprit—the Tilapia lake virus that has been affecting an important fish species but poses no threat to human health. 

Gaining Speed

NGS assays in general have a turnaround time of 24 to 48 hours, considerably slower than the one- to two-hour window with PCR tests, Kapoor says. With the current setup at the CII, either 20 or 40 samples can be run at a time on Illumina NextSeq instruments, depending on the type of flow cell used. 

But the team has started testing VirCapSeq-VERT on the smaller, faster Illumina MiniSeq machine with a run time of four hours, says Kapoor. It has also done some work on Oxford Nanopore’s portable unit that does long-read sequencing, and preliminary data suggests provides more data from fewer molecules. 

Although the output is one-tenth that of the larger instruments in terms of total bases read, speed to answers is often the more important goal in critical care and outbreak response situations. In addition to helping patients and public health, it can save the considerable cost of hospitalizations and treatment for an unknown disease etiology. 

Other NGS-based pathogen panels exist, including ones developed by Twist Biosciences whose library preparation kits and reagents were used for analysis of the respiratory samples in the latest validation study on VirCapSeq-VERT, Kapoor notes. But Twist’s interest is purely in research, not marketable diagnostics, based on his conversations with its CEO Emily Leproust. 

The composition of the nucleic acid probes used with VirCapSeq-VERT is what gives the platform its edge, he says. No other enrichment method is as comprehensive. Lipkin and Briese have spent a lifetime researching viral infections, “so they know what a relevant target is and what is not.”  

Separately, the group has developed a correlate system known as BacCapSeq-VERT for the detection and identification of bacteria and antimicrobial resistance markers, reports Kapoor. “We’re in the optimization process and will then proceed to seek regulatory approvals for the bacterial as well as a combined bacterial and viral panel. Our ultimate goal is to establish a pan-pathogen capture panel that will differentially diagnose all disease-causing pathogens including bacteria, fungi, parasites and viruses.”

Editor's note: This article has been slightly modified from the originally published version.