Tiny Mutations Can Limit Influenza Spread, Study Finds

ATLANTA, Feb. 1 -- Two small mutations are enough to turn off the ability of the highly lethal, wildly contagious "Spanish influenza" virus to pass from person to person, a new study suggests.

The experiments, done in a high-security lab here using a reconstructed version of the microbe that killed about 50 million people worldwide in 1918 and 1919, show that small changes may have huge consequences in flu virus evolution.

The new finding, published in Friday's issue of the journal Science, does not shed light directly on the risk of the H5N1 "bird flu" virus, currently active in Indonesia, becoming a pandemic threat to humans. But it may help researchers know what to look for as they monitor the evolution of flu viruses in hundreds of bird and mammal species.

Genetic fingerprinting of the Spanish flu virus -- taken from the frozen body of an Inuit woman in Alaska and autopsy tissue from two American soldiers -- has revealed that it was a bird virus in the years before it entered the human population, probably someplace in the Midwest in the winter of 1917-1918.

It presumably acquired several mutations that allowed it to make that leap. The new research suggests that one or two were essential -- but probably not sufficient -- to bring about the fateful changes in the virus's behavior.

In the experiment, Terrence M. Tumpey, a virologist at the Centers for Disease Control and Prevention, constructed a Spanish flu virus from scratch. He then made changes to the hemagglutinin (H) protein, a key part of the virus's outer shell that binds to cells before invading them, and tested the results on ferrets.

Ferrets are used as stand-ins for people in influenza experiments because they can be infected with the same strain of flu viruses and suffer similar symptoms, including fever, sneezing and runny noses. Researchers test the transmissibility of the virus by putting two ferret cages next to each other, one containing an infected animal and the other an uninfected one.

When Tumpey and his colleagues made two changes in the H protein (at "letters" No. 190 and 225 in a protein that is 566 letters long), the virus killed the infected ferrets rapidly, but it did not pass through the air and infect the animals in the neighboring cages. Curiously, the animals infected with that mutant strain did not sneeze -- unlike the animals that were infected with unaltered Spanish flu virus.

The changes Tumpey made turned the 1918 microbe into one that attached much more readily to bird cells than to human cells. When he made only one change (in letter 225), the virus showed mixed behavior, attaching to both human and bird cells, but not very well to either.

One of the Spanish flu viruses found in the autopsy specimens from 1918 had that mixed form of H protein. That suggests that even though it was not as fully adapted to human hosts as the viruses in the other specimens (which showed both changes), it was presumably still being passed from person to person -- and was lethal.

It is likely that influenza viruses must undergo many changes, and some changes must occur before others, in order for a strain to make the evolutionary journey from bird microbe to human microbe with pandemic potential.