Putting the viruses of the world into a panopticon is no longer impossible

TO KEEP AN eye on how a single pathogen navigates, mutation by mutation, through a worldwide landscape of adaptation would have been unthinkable more than ten years ago. In the face of a pandemic, it has proved possible. Now, to forestall future pandemics, some researchers want to go much further. There are 264 virus species known to infect humans. Why not track them all?

And why stop there? Scientists estimate that among all the viruses that infect all the animals of the world there may be around 800,000 which could, in the right circumstances, jump from their habitual hosts into humans and start spreading. Could all of them be identified, too, and sentry posts set up that would provide news of their incursions?

Before covid-19 such ideas seemed both expensive and outlandish. As the pandemic has progressed, however, expense has come to seem less of a barrier: a covid-pandemic bill in the trillions of dollars puts up-front costs of $1bn or so into a new context. As to outlandishness: RNA vaccination seemed like a wild idea less than a decade ago, and impractical to implement much more recently. Now it has shifted the boundaries of the possible. It cannot be the only elegant biotechnological notion lurking in relative obscurity that has the potential to be realised.

The Global Immunological Observatory (GIO), an idea conceived by Jessica Metcalf, a disease ecologist at Princeton University, and Michael Mina, an epidemiologist at Harvard University, aim to use modern lab techniques to spot incipient health crises before they take off by providing near-real-time insight into what infections are where. It would do so by looking at hundreds of thousands of blood samples every day.

The blood stores a history of the immune system written in antibodies. Seeing what antibodies someone’s blood contains can show, for instance, whether they had measles as a child, which flu virus laid them low last winter or which cold virus they sniffed their way through with mild irritation three weeks ago. Even if a pathogen never made them sick, the memory of its attempt to do so may still circulate.

The GIO would ideally test every tiny sample of blood for hundreds of thousands of distinct antibodies. Considering how hard large-scale testing for just SARS-COV-2 has proved, this might seem impossibly ambitious. But there are tools that might be up to the job. One is VirScan, a platform developed by scientists at Harvard Medical School which looks for antibodies to more than 1,000 strains of 206 different viruses in a single tiny sample of blood using a canny combination of DNA synthesis and DNA sequencing.

First, the VirScan team synthesised genes describing the proteins found on the surfaces of all those viral strains—the sort of proteins which, like SARS-COV-2’s spike, are most likely to elicit antibodies. It then used those genes to create thousands of slightly different “phages”—simple and easily mass-produced viruses that infect bacteria. Each phage in this library expresses one of those tell-tale proteins on its surface and the relevant gene in its DNA.

Expose a copy of this library, containing billions of phage particles, to someone’s blood and the antibodies which recognise a particular viral protein will bind to the appropriate bacteriophage. Rinse away the phages that are not attached to antibodies—those which have gone unrecognised—then amplify the DNA from all the others and sequence it. The sequencing data reveal what the antibodies recognised. “[It] allows us to get a readout simultaneously of hundreds of thousands of different antibodies that somebody might have,” says Dr Mina.

A system to provide hundreds of thousands of blood samples, even tiny ones, would require a lot of planning and broad goodwill. But Dr Mina thinks blood already taken for other reasons could provide a worthwhile starting point for the project. Blood and plasma banks already exist in every city in the world. Blood samples that are collected by doctors and nurses for any number of medical reasons would also be useful—women coming into hospitals to give birth, for example, give several blood samples, as do new-born babies who routinely have samples taken from their heels to check for genetic abnormalities. Most of these samples are discarded after their primary use. Anonymising the samples and providing them to the GIO could be a simple additional task for any laboratory. Dr Mina thinks primary-care doctors could even offer their patients the option of directly enrolling in the GIO in return for information about their immune status.

One of the advantages of such a system, if deployed widely enough, is that it would detect the true prevalence of pathogens, rather than just seeing them when they cause symptoms. It could also reveal the presence of a new virus not represented in the phage library. Some of the antibodies which recognise a given virus will also recognise others to which that virus bears a family resemblance, though not as well. Thus some SARS-CoV-2 antibodies recognise the viruses behind SARS, MERS and some forms of the common cold. If GIO had been operating in late 2019 and early 2020, it would have seen antibodies to a range of coronaviruses starting to turn up in various places, a molecular shadow alerting it to a new pathogen it could not yet see directly, but which, after a little sequencing, it would understand intimately.

Such anonymised antibody surveillance could be supplemented by sequencing not just of blood. Scientists have shown that the SARS-CoV-2 virus can be detected in waste water and sewage. Using such systems to see where a virus is, neighbourhood by neighbourhood, would give public-health officials useful almost-real-time insights into pockets of infection.

A scheme likeGIO would provide something like the synoptic maps in weather forecasts, showing where the disease fronts are, which way the viral winds are blowing, and alerting people to coming health storms. For extreme events, though, other sorts of data may be needed, from across the animal kingdom.

Pathogens have crossed between species and people for as long as humans and animals have been in close contact. Measles emerged in people during the domestication of livestock. HIV originated in chimpanzees, influenza in birds and pigs, ebola in bats, which also teem with coronaviruses.

Around 1,500 viral species which belong to families known to infect both humans and non-human animals have been discovered to date. The rate of their discovery suggests that mammals and birds together may harbour 1.67m more. Jonna Mazet, an epidemiologist at the University of California, Davis, and Dennis Carroll, an infectious-disease expert formerly with America’s international development agency, estimate that up to 827,000 of them could cross over into humans.

Bats, birds and billions of dollars

Dr Mazet and Dr Carroll arrived at that daunting number when analysing what it would take to make all those unknowns known. Their PREDICT project, which ran in 35 countries from 2009 to 2020, found almost 1,000 new viruses and allowed Dr Mazet to enumerate 42 risk factors governing the likelihood of their spillover and subsequent spread, including how likely the virus was to be encountered by people using the environment in new ways (such as cutting down a forest) or whether a virus has the ability to infect more than one species. A threat assessment based on these risk factors put coronaviruses near the top of the list.

In 2018 Dr Mazet and Dr Carroll sketched out a Global Virome Project (GVP) which would build on the procedures and protocols used in PREDICT: for $4bn, they reckoned, they could find and characterise millions of viruses, massively reducing the number of unknown pathogens on Earth. A scaled-back version that would target only the highest-risk countries, communities and species—such as mammals and water birds—would allow the mapping of around 70% of the viral dark matter for a mere $1.2bn.

Though covid-19 has not brought the GVP any windfalls so far, Dr Carroll says its ideas are catalysing conversations on pandemic preparedness around the world. America’s National Academy of Medicine has asked him to create a detailed strategy for viral surveillance. Some questions are obvious, such as where and how to look. Perhaps the most important thing, though, is what might sometimes be an afterthought: standardised data management.

“It's inverting what we would typically think,” says Dr Carroll. If every country can see that the set-up is “open, transparent, and [protected] against misuse and abuse, that…will enable everything else to happen.” With the right information infrastructure, the next pandemic could be spotted almost before it has begun. ■