danieldrezner.com :: Daniel W. Drezner :: It's not easy to catch avian flu

Thursday, March 23, 2006

previous entry | main | next entry | TrackBack (0)

It's not easy to catch avian flu

Scientific American's David Biello reports on a important finding: why, so far, the H5N1 avian flu virus has not passed from human to human yet:

A virus's ability to spread is the key to its ability to create a pandemic. New research shows that this bird flu currently lacks the protein key to unlock certain cells in the human upper respiratory tract, preventing it from spreading via a sneeze or a cough.

Virologist Yoshihiro Kawaoka of the University of Wisconsin and University of Tokyo and his colleagues tested strains of H5N1 isolated from respiratory tissue in the noses, throats and lungs of infected humans. Although regular human flu viruses bound easily with the receptors found in the nose and throat cells, H5N1 strains attached only to those receptors on cells found in the deepest regions of the lungs.

"Deep in the respiratory system, receptors for avian viruses, including avian H5N1 viruses, are present," Kawaoka explains. "But these receptors are rare in the upper portion of the respiratory system. For the viruses to be transmitted efficiently, they have to multiply in the upper portion of the respiratory system so that they can be transmitted by coughing and sneezing."

....the findings suggest one way in which H5N1 must mutate if it is to become a highly contagious virus, the researchers argue in their paper in today's Nature. It also reveals a way to monitor for the emergence of such a strain. "Identification of the H5N1 viruses with the ability to recognize human receptors would bring us one step closer to a pandemic strain," Kawaoka says. "Recognition of human receptors can serve as molecular markers for the pandemic potential of the [isolated strains]."

Here's the link to the article abstract in Nature.

posted by Dan on 03.23.06 at 04:59 PM

Comments:

I wonder if knowing how the virus must mutate can give researchers any help in anticipating what the virus will be when it becomes H2H. Can researchers narrow their search area from something like an infinite range of combinations to something in the realm of predictability? I also wonder how to express these variables numerically so we can employ the neural network software with genetic algorithms to crunch quickly and get us in the direction of a more potent anti-viral. The problem, as I understand it, is that we can't make the vaccine before the flu has mutated to a form causing the pandemic. Is that true? Can we make intelligent guesses given what we know to have, say, 100 experimental vaccines on hand when it becomes H2H so that maybe one will work and we can immediately start producing it?

Or, is there some hope in this administration that the flu might just go away if humans avoid getting it? Those birds that aren't resistant will die off and there won't be the concern?

posted by: jf on 03.23.06 at 04:59 PM [permalink]

I'm stunned, stunned, that Bird Flu hasn't killed us all by now. I mean, the few vestiges of humanity left who managed to survive SARS.

Or maybe I'm just the Omega Man.

posted by: Don Mynack on 03.23.06 at 04:59 PM [permalink]

jf -

Probably not- that there isn't just one mutation that would enable this (or other) virus to become Human to Human. It could have a (un)lucky point mutation that allows it, or recombine with any number of human/animal influenza strains (most likely in my opinion to have the undesired effect). Figuring out the probabilities of any of the millions of different things that could cause it would be quite the statistics problem -- we have a hard enough time guessing which of the wild flu viruses will hit it big and should be protected against by the vaccine.

Since you seem CS literate, it'd be like trying to solve an O(e^n) problem, but worse than NP as there's no way to check if you have the right answer short of creating the live virus ;).

In addition to that, vaccine creation isn't a "build from bottom up" type of affair. As far as I'm aware, there have only been a very small number of viruses that have had their genome "built."
Vaccines usually developed from having the live virus, replicating it/allowing it to develop its proteins in animals, and then denaturing it in a specific way and combining with other factors that make it so the acquired immune system develops memory cells that recognize the live virus or its products key to replication. Needless to say - hard to do without a live virus.

The good news is we probably don't need it. If we can get a good vaccine against the form that's out there now, there's a decent chance it'll be effective against a one off virus that can infect more efficiently.

*By live, I mean active and able to infect. I'm not confering any nature of "life" on viruses due to the debate that causes.

posted by: mozillauser on 03.23.06 at 04:59 PM [permalink]