Login to Your Account

After Nearly a Year's Delay, Details of Airborne Bird Flu

By Anette Breindl
Science Editor

The American Association for the Advancement of Science published full details of a paper by Dutch researchers that demonstrates H5N1 avian influenza, a highly lethal but to date not very contagious strain of the flu, can acquire mutations that allow it to be spread via aerosol transmission.

The paper, which appeared in the June 22, 2012, issue of Science, and a related paper by U.S. scientists, had first entered public consciousness late last summer, when concerns about their possible nefarious use of the findings cropped up during the peer review process.

The ensuing discussions among scientists, security experts, and the general public prompted Science Deputy Editor Barbara Jasny to quip that "these may have been the most widely discussed unpublished papers of all time" at a BIO 2012 session on biosecurity on Thursday.

Indeed, Jasny said, one of the lessons of the episode is that at some point, secrecy becomes a fool's errand. Despite the fact that the papers themselves were not published, their discussion quickly "reached a point where it became very easy for someone to go to the public literature and the blogosphere and piece together what had been done."

What had been done by senior author Ron Fouchier and his colleagues was to first mutate specific amino acids in the H5N1 virus, and then let the mutated virus replicate for several rounds in ferrets.

The bad news is that under such conditions, H5N1 can indeed learn to fly. Senior author Ron Fouchier told reporters at a press conference that "our main conclusion is that the H5N1 bird flu virus can acquire the ability of aerosol transmission between mammals. . . . As little as five mutations, but certainly less than 10, are sufficient to make H5N1 virus airborne."

But the airborne virus was far less lethal than currently circulating strains. The reason is that in order to become easily transmissible via droplets, the flu needs to infect the upper respiratory tract. But to be highly lethal, it needs to infect the lower respiratory tract, where it is much more likely to lead to pneumonia.

The tradeoff is a general one for flu viruses, and has held up in the case of H5N1. None of the ferrets infected via droplets died. The virus can still kill if it is delivered directly to the lower respiratory tract, but as Jasny noted, "no one in the public is going to get flu because someone is shoving a Q-tip down their trachea."

Nor does the mutated flu virus make a particularly good bioweapon. Indeed, although the possibility of Armageddon or Doomsday terrorists does exist, most groups which might try to weaponize bird flu are more likely to work in the opposite direction, making a virus that is less contagious – making it easier to control – and more lethal, rather than the other way around.

And by showing how easy transmissibility can evolve, the authors have given new insight for viral surveillance, which is an important part of pandemic surveillance.

Still unanswered is the practical question of how likely the same strains could evolve in the wild. Several of the mutations that Fouchier and his colleagues engineered into the virus are already circulating in some wild-type strains.

In a second paper also published in Science this week, British scientists, using mathematical methods, came to the conclusion that it is possible the virus could evolve to easy transmissibility in the wild. But too many unknowns about viral transmission remain to make an accurate estimate of how likely such an evolution is.

Senior author Derek Smith said the situation "is a little like . . . living on a fault line. What we have discovered in this working collaboration with Drs. Fouchier and Kawaoka is that it's an active fault line. . . . It really could do something. And now what we need to know is, how likely is that? We find that it is possible for such a virus to evolve three mutations within a single host. . . . But we can't yet say, because of some of these unknowns, exactly what that risk is. And it makes a big difference whether that risk is 1 in 1 ,000 or whether it's 1 in 100 million. . . . This is about getting a much better accurate assessment of what the actual risk is."