COVID-19 and what else? Researchers use metagenomics to find out

In a small study reported last June in Clinical Chemistry, the researchers demonstrated that whole genome sequencing of nasal swab samples from a handful of early COVID-19 patients as well as  controls could not only flag the presence of the new coronavirus, but could find other germs such as human parainfluenzavirus, a bacterial infection Moraxella catarrhalis, and rhinoviruses — which cause the common cold.

Since that study was conducted, researchers have sequenced more than 2,000 more, and are continuing to evaluate the results.

Metagenomic sequencing is very different from the COVID-19 tests now taken by millions of Americans. That now common “PCR” test has, in a sense, prior knowledge of what it is looking for. It uses sensitive genetic probes that can spot a few telltale sections of the coronavirus’ genes, which can be read like barcodes to identify the virus as SARS-CoV-2, the pandemic strain. This allows caregivers and public health officials to diagnose and treat the disease — and take steps to stop its spread.

Screening entire genomes of patients and pathogens

By contrast, the metagenomic screen used in this new research was designed as if there was nothing known about the pandemic strain. It sequences all the genetic material found in the sample. By sifting through mountains of raw genetic data and comparing it to the genetic profiles of diseases that have already been identified prior to the pandemic, the program was able to uncover a “new” virus.

Within 36 hours it found and fully sequenced SARS-CoV-2 in six of six samples from COVID-19 patients who had previously tested positive. When the analysis compared an unusual gene sequence it had unearthed to those in an electronic library of viral genes known in 2019, it correctly described it as “a novel human Betacoronavirus” related to known strains found in bats.

Chinese researchers used whole genome sequencing in late December 2019 as part of their effort to identify the yet-to-be named novel coronavirus, but first used PCR tests to search through lists of know respiratory viruses. The small UW/Fred Hutch study is a proof of principle that metagenomic screening and rapid analysis could be used in future disease surveillance.

While these early studies are essentially dry runs for the use of whole genome sequencing tools to study co-infections of COVID-19 with other bugs, they also demonstrate a way that could quickly spot the presence of a new respiratory disease so the world will not be caught so-flatfooted as it was with SARS-CoV-2.

“These results demonstrate the value of metagenomic analysis in the monitoring and response to this and future viral pandemics,” wrote the researchers, led by Fred Hutch virologist Dr. Keith Jerome, who also heads the Virology Division at UW, and Dr. Alex Greninger, assistant director of the UW Medicine Clinical Virology Laboratory.

Greninger noted that one reason these screens are so computationally intensive is that the genetic information pulled from the nasal swabs, which are capable of generating a billion snippets of code, must be matched against a vast database of more than 200 million known gene sequences from the vast range of microbial life archived by global researchers to date.

Services such as AWS can help researchers queue up their data and run analysis in a few hours that might otherwise take weeks.

“We’re starting to get to the point where we can almost sample all the molecules in a given sample,” he said. “The sequences are getting bigger and bigger, and the databases are getting bigger and bigger. That’s why cloud computing is so important.”

Putting clinical, viral and genetic data together

Elsewhere at Fred Hutch, researchers are using metagenomic sequencing to study how protein structures on the surfaces of infection-fighting blood cells might determine why some 20% of COVID-19 patients experience severe disease, while the rest do not.

“We have already completed the collection of sequences from 1,032 patients, half of them with severe COVID-19, and half without severe disease,” said Dr. Lue Ping Zhao, who is leading the project.

The focus of Zhao’s study is on a set of immune system genes called HLA Class II on the surfaces of T cells — the same gene that's profiled and matched so that donated immune cells are compatible with the tissue-type of transplant patients. Researchers want to know if HLA genes may recognize and present viral antigens for the immune system to respond to, and if an overactive immune response might lead to  the powerful inflammatory response, known as a “cytokine storm,” that is a common cause of death in severely ill patients.

 “Through this interdisciplinary collaboration, we are now putting clinical, viral and genetic data together through a statistical analysis to see which HLA genetic elements might be predictive of severe COVID-19,” Zhao said.

These early experiments are demonstrating how advances in metagenomic sequencing empower researchers by giving them a clearer, comparative view of the many mechanisms that control our biology and provide new ways to detect and treat disease.