The Avian Flu: Transparency vs. Public Safety

May 3, 2012 7:25 AM

"Experimental adaptation of an influenza H5 HA confers respiratory droplet transmission to a reassortant H5 HA/H1N1 virus in ferrets." After an extensive, months-long debate, one of two controversial papers showing ways the H5N1 "avian" influenza virus could potentially become transmissible in mammals with only 3 or 4 mutations was published in Nature today. The journal included an editorial on the merits and drawbacks of "publishing risky research" with regard to biosafety. The debate included an unprecedented recommendation by The US National Science Advisory Board for Biosecurity (NSABB) to block publication -- a decision they later reversed. (Via: 1, 2) Nature's special report has additional articles, including interviews with the teams behind both papers.
posted by zarq (37 comments total)

For anyone who doesn't want to read the paper:

It appears that the number of mutations needed for the virus go from being avian-specific to mammal-adapted (and specifically one that is dangerous to humans,) is fewer than was previously suspected. Which would mean the risk of pandemic is higher than was thought.

There are several concerns being raised, including that someone might take these findings and engineer a virus that could be distributed as bioterrorism.
posted by zarq at 7:32 AM on May 3, 2012 [1 favorite]

In case anybody's wondering why we should worry about influenza in ferrets, I offer the wheel of cross-species influenza transfer.
posted by Mercaptan at 7:55 AM on May 3, 2012 [3 favorites]

It is good that we now have a better understanding of what it takes for influenza to become a dangerous plague. I am having a really hard time thinking that it was worth doing what these researchers did in order to figure this out, though. As far as the bioterrorism issue, I have a hard time figuring out whether it's really a potential problem or whether it's just a bunch of fear-based sensationalism.

Granted I'm not that much more thrilled about "legitimate" world powers having superflu weapons either. That's the kind of thing that should just never be made by anyone ever, as far as I'm concerned there's no legitimate reason for anybody to ever grant themselves the power to unleash plagues upon the world.
posted by Scientist at 7:57 AM on May 3, 2012

Mercaptan, you might want to explain that graphic. There aren't actually any ferrets on it. Are you saying that it's an issue because ferrets can get the same types of influenza that humans can get? I would agree with that, but I'm not sure the graphic shows that very clearly.
posted by Scientist at 7:59 AM on May 3, 2012

Are you saying that it's an issue because ferrets can get the same types of influenza that humans can get?

Ferrets in disease research are a pretty standard thing these days, especially for influenza.
posted by hippybear at 8:07 AM on May 3, 2012

Scientist: " I am having a really hard time thinking that it was worth doing what these researchers did in order to figure this out, though."

Why?
posted by zarq at 8:09 AM on May 3, 2012

I am torn about this issue. We collaborate with the lab that published these, and while the press releases are scary, I don't see this as being all that useful as a blueprint for making a killer pandemic. The reason the 1917 flu was so dangerous was multiple factors, in addition to the cross species leap. In addition, this is really a confirmation in many ways to data that was anecdotally known, refer to the wheel mercaptan linked.
posted by oshburghor at 8:11 AM on May 3, 2012

wheel of cross-species influenza transfer.

Worst game show ever.
posted by zamboni at 8:13 AM on May 3, 2012 [27 favorites]

*gives up playing Ferret Legging for good*
posted by marienbad at 8:15 AM on May 3, 2012

zarq: Why?

Because it involved sickening a lot of perfectly good ferrets in order to demonstrate something that was pretty well already known and which doesn't seem to have a lot of particularly useful applications outside of the production of bioweapon abominations. It's hard for me to see how this research makes the world a better place.
posted by Scientist at 8:19 AM on May 3, 2012 [1 favorite]

Because it involved sickening a lot of perfectly good ferrets in order to demonstrate something that was pretty well already known

A theory you don't test empirically is functionally indistinguishable from superstition.
posted by mhoye at 8:21 AM on May 3, 2012 [8 favorites]

I am having a really hard time thinking that it was worth doing what these researchers did in order to figure this out, though. As far as the bioterrorism issue, I have a hard time figuring out whether it's really a potential problem or whether it's just a bunch of fear-based sensationalism.

These two questions are not unrelated.
posted by bonehead at 8:23 AM on May 3, 2012 [1 favorite]

Scientist: " Because it involved sickening a lot of perfectly good ferrets in order to demonstrate something that was pretty well already known and which doesn't seem to have a lot of particularly useful applications outside of the production of bioweapon abominations. It's hard for me to see how this research makes the world a better place."

Because mere conjecture isn't really conclusive evidence? Gaining deeper knowledge of how the virus operates and what sort of mutations might allow it to become transmissible to humans (especially since we don't have immunity to that particular strain) can't be a bad thing, can it?
posted by zarq at 8:24 AM on May 3, 2012

I mean, if someone who knows more about this stuff than me (like maybe oshburghor?) can help explain to me how this research really helps us combat flu pandemics in more detail than just "we might be able to look for these mutations during influenza monitoring" then I might concede that there's an argument for doing this research. Is this kind of monitoring something that is likely to actually improve influenza mortality rates or help us avert major epidemics?
posted by Scientist at 8:24 AM on May 3, 2012 [1 favorite]

mhoye: Genetic analysis is not the only form of empirical evidence.
posted by Scientist at 8:26 AM on May 3, 2012

I dunno, I just have this radical idea that a lot of animal research could really be done just as effectively in less invasive and exploitative ways if we were willing to be a bit more creative and patient and were willing to put in more effort. I also feel that a lot of it should probably just be skipped entirely if we can't come up with a way to do it that involves less suffering.

I realize this is an unpopular opinion in the sciences as well as here on MetaFilter so rather than make the whole thread into an argument about that I'm going to go do something else now.
posted by Scientist at 8:30 AM on May 3, 2012 [2 favorites]

I know that I've written about this before, but we just had a journal club on this paper last week (presented by my PI, who has written text book chapters on orthomyxoviruses) so I feel like I should chime in.

Some quick background that's necessary for understanding flu infection. Influenza is an enveloped virus that uses a protein called HA (the Hemagglutinin protein) to recognize a specific glycosylation (sugar chain) that is found on proteins on the cell membrane of cells that influenza wants to infect. Specifically, avian influenza (H5N1) has an HA type (HA type 5) that recognizes glycans that are only found on the surface of avian cells (α2,3-linked sialic acid, for those that are interested). Influenza strains that infect humans, such as H1N1, have an HA that specifically recognizes glycans that are found on mammalian cells (α2,6-linked sialic acid). This receptor binding difference is the primary reason that bird flu is not transmissible between mammals (and vice versa with human flu). Now, what can sometimes happen in nature is that a strain of the virus will acquire the ability to transmit itself between a new host species. There are a couple ways influenza can do that, but this paper is going to focus on mutations to the HA protein that change its binding specificity from avian cells to mammalian cells.

What's really interesting, and somewhat depressing from a scientific standpoint, is that this paper started off trying to use modern, site-directed mutagenesis to introduce mutations at points in the HA protein that are known to be important for the receptor binding function. It's a little like filing down the grooves on a key to make it fit a new lock. If you know the second tumbler is catching, you make a change there and see if it fixes the problem. So that's what they started out doing, and that's where the first two mutations (N224K and Q226L) came from. That's what they're showing in Figure 2: that those two mutations confer binding specificity to α2,6-sialic acid. However, that experiment was only looking at binding strength in vitro, and wasn't looking at biological effects.

So they probably got excited at that point and figured that they had found the mutations that would be necessary for bird flu to mutate to human flu. And maybe if this was all they found, and it had worked, then we'd really have to be having a conversation about bioterrorism and science. But look at Figure 4c: the ferrets on the left side were innoculated with the virus that contains the N224K and Q226L mutations that are supposed to confer transmissibility between mammals. And look at the right side: no transmission to healthy ferrets. Which must have been a pretty bad day for the grad student doing the work (negative results are never fun to take to the boss). So all of the structure-based, directed mutagenesis that used cutting edge receptor binding assays to predict which mutations were going to allow bird flu to jump to mammals? A bust.

So that's probably when they designed the transmission study. Let nature do the work. (Boys and girls, this is also a classic example of natural selection at work on the micro scale) Putting a sick ferret next to a healthy ferret, waiting until ferret#2 gets sick, putting him next to healthy ferret#3, and then see what kind of virus you end up with. And that's where they discovered a couple new mutations (N158D and T318I) that, when combined with their previous two, actually DID allow for transmission between ferrets. And even though they say that this doesn't mean that this mutant virus could spread between humans, there's no way I'd go anywhere near that sneezing ferret. They're good animals for modeling the airways of humans. Good enough that I'd put money on someone being able to get sick from it.

So that's a really long scientific summary of the work. But where does it leave us with respect to the ethics of publishing it? The real problem here is that the high tech approach to determining the mutations didn't do a whole lot. So listing out the specific mutations in a paper only helps other researchers who are trying to understand how these mutations function, for the sake of understanding the virus better or designing better drugs or whatever. What this paper shows is something that's been known for a long time in term of organism adaptation: we can't out-science Nature (yet). So if you were a bio-terrorist, and all you cared about was making the virus deadly, and didn't care anything about how it works, then all you'd have to do is buy some ferrets and a sick chicken. The techniques for serially passaging virus in animals to adapt the virus to a new host is really, really, really well documented in hundreds of other papers stretching back to the early 1900s. The cat's (ferret's?) out of the bag on that one. So as far as developing a bio-weapon goes, I don't think paper adds a whole lot to the already large body of knowledge. But to address Scientist's question, knowing which specific mutations arise in the natural selection process allows labs like my own to study the particular residues to determine exactly how they confer new function. And understanding how these proteins function is what really will ultimately allow us to (temporarily, probably) out-science Nature in terms of new drug development.
posted by Osrinith at 8:30 AM on May 3, 2012 [67 favorites]

showing ways the H5N1 "avian" influenza virus could potentially become transmissible in mammals with only 3 or 4 mutations

See! We're totally safe, because evolution is not real and can't happen.
posted by Pogo_Fuzzybutt at 8:37 AM on May 3, 2012

I think osrinth is on the mark. This gives one tools for treatment and perhaps some means of prediction of problems. The last avian flu in asia was found due to analysis after the shift(people getting sick and or dying), this gives the potential to maybe identify and intercede prior to cross species capability.
posted by oshburghor at 8:37 AM on May 3, 2012

Osrinith, that's exceptionally helpful. Thank you.
posted by zarq at 8:44 AM on May 3, 2012

Yea, put me in the camp of those that want less of what I call 'frivolous' animal testing of the type that is for cosmetics and hair gels testing. That said, the sort of thing I'm seeing in this post is exactly what I thought animal testing was supposed to be used for. I suppose it is a bit of a line in the sand situation and different people are going to have different definitions of 'reasons that are good enough' to do this type of thing but I can't help but think that verifying/proving things of this nature are pretty valid reasons.

Not to mention that more people should be more cognizant and use resources that allow people to find and support cruelty free companies. /plug
posted by RolandOfEld at 9:02 AM on May 3, 2012 [1 favorite]

They should have published in an Elsevier journal to make sure no one could get access to this information.
posted by snofoam at 9:05 AM on May 3, 2012 [11 favorites]

Madagascar closes her ports...
posted by Renoroc at 9:29 AM on May 3, 2012 [1 favorite]

Now if we can only get ferrets into Madagascar....
posted by Fister Roboto at 9:31 AM on May 3, 2012

Ack!
posted by Fister Roboto at 9:32 AM on May 3, 2012

Scientist: "I mean, if someone who knows more about this stuff than me (like maybe oshburghor?) can help explain to me how this research really helps us combat flu pandemics in more detail than just "we might be able to look for these mutations during influenza monitoring" then I might concede that there's an argument for doing this research. Is this kind of monitoring something that is likely to actually improve influenza mortality rates or help us avert major epidemics?"

As Osrinith mentioned, the applications of this research to drug development are compelling. If we are going to be able to fight H5N1 influenza if/when we acquire an epidemic strain, we will need to have treatment strategies that address the actual virus, not the virus we imagine it might be.

In addition to that though, MONITORING IS INCREDIBLY FUCKING IMPORTANT. When the next great lethal pandemic comes, the sooner we know how bad it is the sooner we can address it with the tools we have. In the case of influenza, the spread could be slowed through global quarantine buying valuable time, vaccine development against the novel strain could be jumpstarted with the appropriate amount of desperate urgency, schools and non-essential offices shut down, stockpiles dispensed, emergency personnel placed where they need to be, and just as importantly false alarms avoided.

When thinking about the value of a ferrets life, I think it is worth remembering how INCREDIBLY FUCKING TERRIFYING H5N1 has the potential to be. If the virus ever does in us what it routinely does in pigs, it could easily be the greatest disaster in all of human history, dwarfing all of the great wars of the 20th century by an order of magnitude.
posted by Blasdelb at 9:34 AM on May 3, 2012 [1 favorite]

If you want to learn more about what is going on with the paper from people who actually know, there was a session at the ASM Biodefense and Emerging Diseases Research Meeting with the four most involved people on the planet. The moderator, who is incidentally the chair of NSABB, jokes that it was the most attended 7:15am session he had ever seen, and I'm sure he wasn't kidding.

There is,

-Ron Fouchier, the actual Principal Investigator of the work, talking about his perspective at the beginning,

-Bruce Alberts, the Editor-in-Chief of Science, talking about Science’s response to the situation,

-Michael Osterholm, Director of the Center for Infectious Disease Research and Policy (CIDRAP), talking about the NSABB recommendations,

and

-Anthony S. Fauci, Director of the National Institute of Allergy and Infectious Diseases (NIAID), talking about government response to the recommendations.
posted by Blasdelb at 9:46 AM on May 3, 2012 [2 favorites]

All I can say is thanks for the explanations. Incredible.
posted by Scientist at 10:19 AM on May 3, 2012

Blasdelb: " -Ron Fouchier, the actual Principal Investigator of the work, talking about his perspective at the beginning,"

Just a small correction (that shouldn't detract from your link, thank you very much for posting it): the paper that was printed in Nature this morning was not Mr. Fouchier's. He is one of the authors of the paper that has not yet been published. He had been awaiting approval from the Dutch government to release his paper to the journal Science, which was only granted six days ago. The paper that was published today in Nature is from virologist Yoshihiro Kawaoka at the University of Wisconsin. Their work was conducted separately, with different research teams.

In addition, while the NSABB had approved the full release of Dr. Kawaoka's paper, it had suggested that only Dr. Fouchier's data, methods and conclusions be released, rather than the full paper -- which I believe is notable only in that they apparently considered it more controversial.
posted by zarq at 10:19 AM on May 3, 2012 [1 favorite]

zarq, that is absolutely right, and it is important to keep the two papers and labs from getting conflated.
posted by Blasdelb at 11:47 AM on May 3, 2012

What strikes me as most interesting about this whole debacle is how the only real harm done to public health was committed entirely by the NSABB in their efforts. There is only one aspect of this research that is in anyway potentially useful to anyone interested in producing an influenza pandemic, and that is the serial passage techniques that have already been around for more than 100 years.

It has occurred to me that I probably shouldn't on a public forum, but I could easily describe a method for producing a novel virulent strain of of zoonotic flu that anyone with a few thousand dollars and a lot of time on their hands could do in a large garage with a few paragraphs. These techniques are spread throughout in the old literature, accessible, and reasonably intuitive. If you think about it, those techniques were forgotten by, and more importantly boring to, everyone except those to whom they were appropriately useful until the NSABB stuck its nose into the whole business.
posted by Blasdelb at 12:18 PM on May 3, 2012

Great explanation Osrinth. Sadly, the only thing I'll remember about it is "Beware the Sneezing Ferret!"
posted by benito.strauss at 3:33 PM on May 3, 2012

We've already had a massive H1N1 pandemic—the Spanish flu in 1918—so at least we know what to expect. It killed somewhere between 50 and 130 million people, which was roughly 5% of the world's population. That was more deaths than World War I and World War II combined. If a pandemic with a similar virulence and fatality rate hit today, it would cause somewhere between 180 and 480 million deaths, mostly children and the elderly, out of a staggering 1.8 billion people infected.

Of course, the world is a much different place than it was in 1918. On the one hand, we now have much better hygiene standards, several antivirals with some degree of efficacy, and advanced life support techniques. On the other hand, we also have much greater urbanization (~50% of the world's population now live in cities, compared to ~20% a century ago), as well as rapid, widespread international travel, both of which makes it possible for viruses to spread much faster than previously possible.
posted by dephlogisticated at 4:33 PM on May 3, 2012

dephlogisticated, H5N1 outbreaks in pigs are also regularly more virulent and more transmissible than the 1918 flu was in us.

There are also a couple of other differences that will likely be important. During the 1918 pandemic the world was in the middle of WWI and so governed by almost uniformly totalitarian governments that were able to efficiently stamp down the spread of negative news as well as panic in general. The bulk of young men and bachelors were also drafted and so not around or free to riot or do stupid shit. Travel in most of the world was not only slow, but already restricted due to the war, even within many countries. The scale of the epidemic will also be larger with the world's population being so much larger, which will benefit the virus.
posted by Blasdelb at 4:56 PM on May 3, 2012

The techniques for serially passaging virus in animals to adapt the virus to a new host is really, really, really well documented in hundreds of other papers stretching back to the early 1900s. The cat's (ferret's?) out of the bag on that one. So as far as developing a bio-weapon goes, I don't think paper adds a whole lot to the already large body of knowledge.

Yeah this is the key point. The technique they used to breed this is not new, it's been known for hundreds of years. It's something that someone with even a rudimentary understanding of evolution could figure out to me.

It's shocking to me that people are actually worried about this.
posted by delmoi at 5:10 PM on May 3, 2012

If they could breed a Flu virus that would only kill "the enemy", it would have happened hundreds of times over by now. I suppose we should be thankful for the dearth of biodiversity shown in our species, at least in this case.
posted by seanmpuckett at 5:22 PM on May 3, 2012

it would cause somewhere between 180 and 480 million deaths, mostly children and the elderly

Don't forget pregnant women -- one of the largest groups of deaths both in 1918-19 and in the very mild swine flu recently were women who were pregnant (or had very recently been pregnant), as a result of a woman's immune system being wired differently while gestating a fetus.

Another fun feature of pandemic flu for pregnant women is that it can cause premature labor -- it's not just the airway and GI tract that can hemorrhage, it's everything, and whoosh, out comes baby. *shudder*
posted by Asparagirl at 5:52 PM on May 3, 2012 [2 favorites]