Go to Main Story.

Go to National Section

Go to Home Page

Labs Struggle With Mysteries of Avian Virus

By David Brown
Washington Post Staff Writer
Sunday, January 11, 1998; Page A16

ATLANTA—On the outside of the laboratory's glass wall someone has taped a hand-lettered sign reading "HOT."

On the other side of the glass, a technician sits at a lab bench. His head is encased in a plastic-and-fabric respirator hood, its air supply delivered by a large hose. The man's gloved hands hold a plastic pipette. He rhythmically deposits human blood serum into tiny tubes that hold samples of a potentially lethal influenza virus.

Nearby, but out of sight, a machine is decoding the genes of another sample of the microbe, molecular letter by molecular letter. It's a process that, start to finish, will take four days.

In another building, a dozen ferrets -- for more than 60 years the preferred species for influenza research -- are quietly incubating the virus.

It would be nice if scientists in the Influenza Branch of the Centers for Disease Control and Prevention (CDC) could study at their leisure this new virus, which has caused 17 cases of illness, and four deaths, in Hong Kong. There are so many questions to be answered.

How did it get into the human population when it previously had been seen only in birds? How has it changed -- if it has changed -- in its trip over this "species barrier"? What makes it so deadly, for some kinds of birds and occasionally for people? Can it be made less dangerous so that it can be safely used in the production of a vaccine against it, should one become necessary?

Unfortunately, there is no leisure -- not in this lab, or in a handful of others like it around the world. So scientists are trying to answer all the questions at once.

One of the earliest things CDC researchers did when they got virus samples from the Hong Kong patients with "bird flu" was to see, in fact, how much similarity it bore to influenza viruses that infect people.

To infect any living animal -- duck, pig, horse, seal or human being -- an influenza virus must first attach itself to a cell. It then burrows in, commandeers some of the biochemical machinery, replicates itself, and eventually kills the cell. Attachment, however, is the crucial first step, and it happens with the help of a viral protein called hemagglutinin.

When hemagglutinin attaches itself to a cell, it does so by sticking to something called a sialic acid receptor on the cell's surface. As it happens, the biochemical makeup of the sialic acid receptor in birds is slightly different from that in human beings. The difference between the two, in fact, is one piece of the "barrier" between the species. It's one of the things that helps keep avian influenza viruses in bird populations and out of human populations.

In at least 17 cases, however, avian virus had gotten over the barrier. There are two ways that could have happened.

The first way is that mutations might have crept into the hemagglutinin gene in an avian virus, changing the protein's structure so that it had greater "affinity" for the human sialic acid receptor.

A second possibility is that an avian influenza virus could have picked up a whole hemagglutinin gene from a human influenza virus. Such an event is unlikely in the extreme -- in fact, nearly impossible. And yet, given the trillions of reproduction cycles influenza viruses undergo every day around the world, it eventually happens.

Virologists believe such events -- they are called "reassortments" -- occur mostly in pigs. Pigs have both bird-like and human-like sialic acid receptors. Consequently, they can be simultaneously infected with both avian and human strains of virus. When that occurs, the mixing of bird and human genes is possible. The result is wholly new, hybrid influenza virus.

It is just such a reassortment, scientists believe, that produces pandemic strains of flu. The newness and strangeness of these microbes is part -- but only part -- of the reason that pandemic strains tend to cause higher mortality than ones that have been circulating for years.

When the CDC virologists looked at the hemagglutinin from the Hong Kong flu patients, what they found surprised them. It was clearly an avian version of the protein. Furthermore, there were no changes in it -- or at least in the part that binds to the receptor -- that would explain an affinity for human cells.

People in the past have been exposed to high concentrations of virus quite similar to that seen in the Hong Kong patients, and not gotten sick. So what happened this time?

"From a virologic point of view, that's one of the most interesting questions," said Nancy J. Cox, the head of CDC's Influenza Branch. "We are having to look further to see why these viruses are able to infect humans."

Influenza virus carries eight genes. Ideally, the researchers would like to decode all eight in every sample they get from a Hong Kong flu patient. They could then compare, in maximal detail, each "case" virus to the others, and to samples of avian virus.

There isn't time for that, however. Instead, the researchers are decoding the entire genes for the two most critical proteins, and are only sampling the six other genes.

"Two things seem to be clear so far," said Kanta Subbarao, the head of the molecular genetics lab in the Influenza Branch, as she sat outside the "hot" lab last week. "The first is that the genes are still all avian. There is no evidence of reassortment. And second, we are not seeing cumulative changes -- changes that would suggest the virus is adapting to a human host."

Nothing, in short, has been found that would explain how 17 people in Hong Kong came down with bird flu.

Even though the virus's route over the species barrier remains a mystery, the two observations made by Subbarao are reassuring. They both suggest that bird flu is not well-suited to humans. Which is to say, it's unlikely to become an epidemic strain unless it picks up some more "human" characteristics.

That could happen, however. After all, it's flu season in Hong Kong. It's remotely possible a person could come down with bird flu and human flu, and in a pig-like fashion act as mixing vessel for the two microbes.

For that reason, virologists at CDC, and elsewhere, are trying to come up with a version of the Hong Kong virus, or find a close relative of it, that could be used to make a vaccine to help stem a pandemic.

Making flu vaccine involves industrial production. Hundreds of people inject tens of thousands of chicken embryos with live virus, and then harvest, purify and inactivate large batches of the microbe. The Hong Kong bird flu virus is far too dangerous to be handled in such a way.

In truth, any virus worthy of being put in a vaccine is too dangerous to handle outside a lab. The solution to this problem is to do what the virus does. It is to make a reassorted hybrid -- a virus with genes from the strain you want to be protected against, and the other genes from a harmless strain.

CDC scientists are now trying to do this. But in the case of the bird flu virus, the job is more complicated than usual.

That's because its hemagglutinin contains a tiny genetic change, seen only in some H5 strains, that makes it extremely deadly for birds. (The fact that it contains this change suggests strongly that the Hong Kong patients caught the same virus that caused an epidemic of fatal influenza among chickens at three farms in Hong Kong last spring.)

To make a vaccine-acceptable version of the bird virus, virologists must first remove the hemagglutinin gene from a sample of it, snip out the lethal change using techniques of genetic engineering, and then reinsert the gene into the virus. They hope to try to do this in the next few weeks.

In the meantime, however, the people in the Influenza Branch are looking for a stand-in for the Hong Kong virus. In the last few weeks, they have been infecting ferrets with a similar but harmless avian virus, one found in Singapore last year. They want to see if the animals develop antibodies that are protective against the Hong Kong virus. If they do, the Singapore virus could be used as the basis for a vaccine.

In recent years, influenza vaccine has contained three inactivated viruses -- two against influenza A viruses, and one against influenza B virus. Which exact strains are used depends on what's circulating in the world at the time. That is the reason influenza surveillance -- isolating and identifying virus from ill people -- is essential for the production of vaccines.

Every year in late January, CDC virologists, academic researchers, and scientists from the U.S. Food and Drug Administration meet in Bethesda and decide on the precise viral "recipe" for the next year's vaccine. Scientists with the World Health Organization meet in Geneva in early February for the same purpose. (The recommendations of the two groups are rarely different). Drug companies then produce the vaccine through spring and summer, for use in the fall.

Unless there's clear evidence of person-to-person transmission, there's virtually no chance the bird flu virus will be included in next year's vaccine. That is the lesson from the world's last pandemic influenza scare -- the "swine flu" of 1976.

That year, an outbreak of influenza at Fort Dix in New Jersey killed several soldiers. The virus was similar to one found in pigs. The disease it caused bore similarities to the influenza of 1918, which was caused by a pandemic strain that killed 30 million people around the world.

No samples of that virus exist, so an exact comparison couldn't be made. However, fearing an epidemic with high mortality, the federal government urged a campaign of mass immunization against swine flu.

The campaign was halted a few months later when numerous people developed rare, severe reactions to the vaccine. More important, it was clear by then that the virus had petered out. It had never spread beyond the crowded conditions of the military post, and consequently posed no threat to the broader population.