WASHINGTON - Making a small change in the outer coating of the lethal 1918 flu virus was enough to stop it from spreading, a discovery that may help scientists monitor today's bird flu and other influenza strains for signs of the next pandemic.

Chicken traders sell their stock to customers unseen at a market in Lagos, Nigeria, Tuesday, Feb. 1 2007. [AP]

The 1918 pandemic was triggered by a bird virus that mutated into one that could attack humans, going on to kill a staggering 50 million people worldwide in a matter of months.

To learn what caused that catastrophic bird-to-human transformation, scientists with the Centers for Disease Control and Prevention turned back the clock: They worked with recreated batches of the actual H1N1 flu strain that spawned the 1918 pandemic, but they altered two spots in a key protein to make that virus a little more birdlike again.

Then the researchers dripped the altered virus into the noses of ferrets, who catch and spread influenza like humans do.

The infected ferrets still sickened and died as the flu ravaged their lungs. But remarkably, they didn't infect healthy ferrets caged right next to them, the CDC team reports in Friday's edition of the journal Science.

Why not? Most flu spreads when an infected person coughs or sneezes out droplets of virus. Ferrets infected with the altered 1918 virus didn't sneeze at all.

The research suggests that for a new flu strain to become a pandemic threat, a protein called hemagglutinin that coats the virus' surface must prefer attaching to cells found mostly in the human nose and windpipe, where it can be sneezed easily.

That's good news when it comes to the notorious Asian bird flu that scientists are watching anxiously today. That strain known as H5N1 bears hemagglutinin - the H in its name - that still prefers cells mostly found in the gastrointestinal tracts of birds. While it has killed at least 164 people worldwide and killed or prompted the slaughter of millions of birds across Asia since late 2003, the H5N1 virus can't yet spread easily from person to person.

"This is very, very elegant work," said Dr. William Schaffner, an infectious disease specialist at Vanderbilt University who advises the federal government on flu issues.

"It may not be exactly the same mutations that would change an H5 virus," Schaffner cautioned. Still, he said, "We appear to be narrowing down our understanding of the kinds of mutations it might take to change a bird-specific virus to one that could be transmitted readily among humans."

The CDC's next step, in fact, is to study these same changes in hemagglutinin amino acids, the protein's building blocks, in H5N1.

But it will almost certainly take additional genetic changes to turn the H5N1 bird flu into a major human killer, changes that probably involve other proteins than just hemagglutinin, contends lead researcher Dr. Terrence Tumpey, a CDC microbiologist.

"I think that researchers may discover that the combination of genes needed is maybe unlikely to occur in nature," he said.