Resurrecting a Killer Flu
A scientist explains why he re-created the lethal virus that killed millions in 1918 and what it can teach us about today’s avian flu.

http://www.msnbc.msn.com/id/9623695/site/newsweek/

By Anne Underwood
Updated: 6:06 p.m. ET Oct. 7, 2005

Oct. 7, 2005 - Scientists have long puzzled over the exceptional lethality of the 1918 flu, which killed between 20 million and 50 million people worldwide. What features of the viral genome enabled it to become both highly transmissible and lethal at the same time? Some of those questions were answered this week, with the publication of twin papers in the journals Nature and Science. In Nature, Jeffery Taubenberger of the Pentagon's Armed Forces Institute of Pathology announced that he had completed sequencing the genome of the 1918 flu. At the same time, Terrence Tumpey, senior microbiologist at the Influenza Branch of the Centers for Disease Control and Prevention, reported in Science that he and his colleagues had used Taubenberger’s sequence to reconstruct the actual 1918 virus, a living copy of the germ that killed millions.

Fears that it could escape into the environment or be appropriated by bioterrorists made it a controversial move. But Tumpey says the risk was worth the trade off because of the information we stand to learn from the virus. What was particularly chilling about the last killer flu was that it appeared to come, with only minimal changes, from an avian virus--bringing a new urgency to the current flu sweeping Southeast Asia. Tumpey spoke with NEWSWEEK's Anne Underwood about his findings. Excerpts:

NEWSWEEK: What is the significance of these twin papers?
Terrence Tumpey: For the first time we have a truly avian pandemic influenza virus that we can study. Not only did we want to rescue the virus, but also characterize some of the important viral proteins that made it so exceptionally virulent.

And we know for sure now that this was a purely avian virus, not a hybrid. Was it really changes of just 25-30 amino acids out of 4,400 in the viral RNA that transformed the virus into a killer?
That’s Jeff Taubenberger’s work, but it appears that way. The dogma until [the current bird flu struck in] 1997 was that pandemics were caused by shuffling of genes between avian and mammalian viruses. But both the current bird flu outbreak and the 1918 virus appear not to be a human/avian reassortant virus, but an avian virus that made minimal changes to infect humans directly. Thankfully, bird flu virus hasn’t figured out how to spread yet. The 1918 virus did.

Does this confirm our worst fears about current bird flu?
It’s hard to know whether or not the current flu will emerge into the human population and spread efficiently. But it’s a good guess that with enough time, it will figure out how to transmit human to human. If so, it will fit the three criteria of a pandemic: a novel subtype, a subtype to which the population has no immunity and high transmissibility.

Why was it important to create a living virus from the sequence?
There is little information in the sequence itself that tells us why it would be so deadly. We see from the sequence that it is avianlike, but there are not any obvious molecular smoking-gun features that we can point to and say, "That is the reason why it killed so many people." Reconstructing the virus helps us do that and identify targets for vaccines and antiviral drugs. The knowledge we’re gaining to protect public health far outweighs the hypothetical risk of working with this strain or providing the information to the public.

When you tested the reconstructed 1918 virus in mice, chicken embryos and human lung cells, what did you find was different about it?
We demonstrated in mice that the hemagglutinin (HA) protein on the viral coat was essential for development of severe pulmonary disease. The 1918 HA seemed to target deeper areas of the lungs than standard viruses. It targeted the alveoli--delicate tissues where the exchange of oxygen and carbon dioxide takes place. In animals, we saw lots of inflammation deep in the lungs, blocking the airways. When you remove the HA protein from 1918 and replace it with the HA from a contemporary virus, you don’t see the virus target deeper areas of lung. That appears to be a unique function of the 1918 virus.

What’s more, it can propagate in culture without the enzyme trypsin. Normally, with garden-variety influenza, in order to grow the virus in culture, you have to add the enzyme, which enables the hemagglutinin to function properly. These enzymes are found in lung tissue, too, or else influenza wouldn’t grow in our lungs.

What does the trypsin do?
It cleaves the HA, breaking it into two pieces. One of the pieces is used to get inside cells [for the virus to replicate]. If the virus doesn’t cleave, it binds to the cell surface, but can’t get inside. But the 1918 virus doesn’t require trypsin. We demonstrated that the other protein on the viral coat, referred to as neuraminidase, was able to mediate that function by cleaving the hemagglutinin, although it’s not clear yet how it does that.

Targeting the alveoli and functioning without trypsin would help explain why the virus was so lethal. What accounted for its ease of transmission among humans?
We don’t have a clue about transmissibility of the 1918 virus. But we showed in the paper that the polymerase genes of the 1918 flu were essential for maximal replication of the virus. Polymerase is one of the most important genes of the virus. Think of polymerase as the engine of the virus, driving the replication machinery. A good polymerase makes more and more copies. A bad one won’t. The polymerase described in the Nature paper is almost all avian, with minimal changes. The current bird flu also appears to have a very good polymerase.

What kind of precautions do you have to take to work with the virus in the lab?
Before we even began this work, we had to undergo multiple safety approvals in order to protect ourselves and public. The 1918 virus was reconstructed in a high-containment, BSL-3-enhanced [biosafety-level-3-enhanced] laboratory. Of the different biosafety levels, the highest is 4. That’s for agents such as Marburg and Ebola that have no treatment at all and are deadly. The next level down is our level--BSL-3-enhanced--which is used for agents that are susceptible to current FDA-approved antiviral drugs and for which the public has some cross-immunity from other viruses.

The population already has some immunity against the 1918 flu?
Yes. H1N1 subtype viruses still exist, although they’re declining. Older individuals especially have good cross reactivity to the 1918 virus.

What other precautions do you take?
All viruses are handled inside a biosafety cabinet. In addition, we have a half-body suit with a respirator, so all the air is filtered before it reaches the head piece. We have to wear double gloves and shower out when we’re finished. And there are heightened security requirements mandated by the select agent program [to make sure the virus doesn’t fall into the wrong hands]. Fingerprinting pads are used for entering the high-containment lab. And to open the locked freezers, we use retina scanning for identification.

Once the sequence is out there, what’s to stop someone from whipping up a batch?
It takes a lot of skilled work by microbiologists. It can’t be done overnight in a less than ideal lab. And there are other complicated issues. You can’t grow it by conventional means because the 1918 flu kills eggs. Standard flu does not.

How did this project get started?
In 1995, Jeff Taubenberger started systematically going through the eight viral genes and sequencing them individually. The initial genetic material he used came from the Armed Forces Institute of Pathology, which was started as an Army medical museum in 1862 by Abraham Lincoln to help fight diseases of the battlefield. Jeff found tissues from two soldiers who had succumbed to infection in 1918. But they were just little tissue blocks the size of a nickel or a postage stamp. He didn’t have enough to sequence the whole virus.

Then a retired pathologist named Johan Hultin saw Jeff’s work and offered to help him get more tissue from victims whose bodies were preserved in permafrost. This pathologist, as part of his graduate studies in 1952, had gone to an Alaskan burial site to exhume bodies, after receiving permission from the city council. The poor village probably got the virus in the fall of 1918 from the mailman, who delivered the mail by dogsled. He delivered the mail and the 1918 flu. In November, 72 people died--about 85 percent of the adult population there. All were buried in this mass grave, below the permafrost layer. Hultin got permission from the town of Brevig Mission, and got lung tissue from a female that had died. He sent it to Jeff, who had enough genetic material then to work with.

So what’s the next step for you?
We are planning on looking at the other four of the 1918 influenza genes that were not addressed in the current work. Now that we know the full virulence of the pandemic virus, we can replace individual genes with genes from contemporary viruses and determine their importance in replication and virulence.

© 2005 Newsweek, Inc.