Sequencing Tested as Infection Early Warning

Scientists are coopting the power of cheap, fast DNA sequencing as an environmental sensor for infectious pathogens that can spread through a community. They hope that the approach can provide earlier warning of disease outbreak.

In a project dubbed the Disease Weather Map, Eric Schadt and colleagues at Pacific Biosciences, a sequencing startup in Menlo Park, California, are using the technology to monitor viruses from a number of locations around the Bay Area, including sewage stations, toilet handles, and the mouths of its own employees.

The pilot project, which Schadt presented last week at the Personal Genomes conference in Cold Spring Harbor, New York, is still in its early days–it’s not yet clear how broadly or densely scientists would need to sample the environment to identify warning signs earlier than existing monitoring methods, such as doctor’s reports and tests that involve growing the pathogens in the lab. But they have shown that it’s feasible to collect, sequence, and analyze genomic samples collected from the environment in a single day.

“The idea is to build a real-time weather map of disease by sampling different locations, like airports, BART, or emergency rooms, and use it to measure pathogen flux over time,” says Schadt. “If we can identify the influx of something like H1N1 into a community very early, then maybe we can mobilize resources to react, like getting supplies of Tamiflu.” He hopes ultimately to be able to combine the virus data with mapping software to create a picture of the pathogens in a particular neighborhood.

Schadt, who became chief scientific officer at Pacific Biosciences last year, envisioned the project as a way to harness the speed of the company’s novel sequencing technology. For scientists to act fast enough to protect public health, they need to be able to sequence and analyze virus data extremely quickly. Last spring, Pacific Biosciences collaborated with the New York Department of Health to show that the company’s sequencing technology could be used to sequence and analyze different strains of influenza A virus in a single day.

Public health agencies, such as state health departments and the Centers for Disease Control, currently use a combination of approaches to detect outbreaks, relying on doctors’ reports, molecular testing, and a new generation of Internet-based tools. Google Flu Trends, for example, predicts flu outbreaks based on the number of people searching for information on the flu. But none of these approaches monitor the environment, rather than people, for pathogens. Scientists hope this approach will be able to detect the signs of outbreak earlier than existing methods–before a significant number of people get sick–allowing cities and states to better prepare for potential outbreaks or even prevent them. “This is probably the wave of the future in terms of where we want surveillance to go,” says John Brownstein, a researcher at Children’s Hospital Boston and creator of HealthMap, an Internet-based disease-surveillance tool.

For the pilot project, researchers at Pacific Biosciences took samples from common areas at the company headquarters, such as handles from doors, toilets, and refrigerators, once a week for a month. They also took cheek swabs from employee volunteers twice a month for 2.5 months.

The researchers found multiple influenza strains in a number of people and on surfaces, including the H1N1 virus linked to last year’s swine flu pandemic. It’s not yet clear how predictive the approach can be, but researchers did see hints of its potential. One week into the study, Schadt says, a number of volunteers failed to show up for testing because they were out sick. Almost all of those who did show up turned out to test positive for H1N1, and the virus was present all over the office surfaces.

Researchers were also able to track the emergence of a new variant of an influenza virus–one of the advantages of using sequencing to monitor viruses and other pathogens. Very small changes in DNA, even just single base pairs, can make microbes more dangerous or drug-resistant, says George Church, a geneticist at Harvard Medical School who is working on a similar project. With the Pacific Biosciences testing, researchers identified two novel mutations in a particular influenza virus, and then tracked how the new variant became more common than the original one.

As part of the same project, collaborators at the University of California, San Francisco, collected samples from local sewage treatment facilities, with the aim of sampling the viruses in a community on a much broader level. While the researchers are still working on the technical details of isolating viral DNA from bacteria-laden sewage, they did find some interesting strains. “When you zoom in, you not only find pathogens that cause illnesses in people, such as enterovirus and other viruses of both the respiratory and intestinal tract, but we’re also finding lots of other viruses specific to different foods you eat,” says Schadt. For example, they detected a virus common in tomato and pepper plants. “We can start thinking about epidemiological studies to identify patterns of food consumed in different areas, which would be interesting to correlate with disease.”

Rob Knight, a biologist at the University of Colorado at Boulder, and his collaborators are working on a similar project, called BioWeatherMap. They are sequencing genetic material recovered from dollar bills collected from all over the country and beyond to determine how accurately money reflects the microbial communities on our hands, as well as to search for differences by source, comparing emergency rooms with college cafeterias, for example.

Schadt aims to use the pilot data to garner funding from Google or other sources to expand the effort. Obviously, predictive power will be crucial to the project’s success. It’s not clear how much sequencing they’ll have to do to have an impact on public health. “We’re not ready to analyze hundreds of samples at once,” says Schadt, adding that, by 2014, the company expects the PacBio machine to be able to sequence gigabytes of DNA in 15 minutes. At that rate, it could sequence viruses such as influenza A, which has only about 13,000 letters of DNA, at the rate of approximately 300,000 per hour.