We should assume that an outbreak of any new strain or subtype is potentially dangerous to humans

Dr Yoshihiro Kawaoka

The discovery helps to explain why a flu outbreak four years ago in Hong Kong killed an unusually high proportion of those that it infected.

Just 18 people were known to have been infected, but six of those died.

Lead researcher Dr Yoshihiro Kawaoka, of the University of Wisconsin-Madison, US, said: "We have found that a limited number of very tiny genetic changes in a specific gene, one called PB2, can have a big effect on how potent the influenza virus is.

"Because the influenza virus constantly mutates, and because only a few changes can make a non-pathogenic virus highly pathogenic, we should assume that an outbreak of any new strain or subtype is potentially dangerous to humans."

Natural reservoirs

Wild waterfowl are natural reservoirs for the influenza virus. These birds transmit the virus to pigs or chickens, which then pass it on to people.

The deadly outbreak of influenza virus subtype H5N1 in Hong Kong in 1997 was the first documented case of an influenza virus jumping directly from chickens to people.

Dr Yoshihiro Kawaoka made the discovery

Public health authorities responded by ordering the slaughter of more than one million live poultry to prevent further spread of the virus to humans.

Dr Kawaoka and his team obtained samples of the H5N1 viruses that had infected Hong Kong residents. Testing these viruses in laboratory mice, the researchers found that they had the same impact on mice.

Just two particles of the deadly form of the virus isolated from human victims were enough to kill a mouse.

But the less virulent form of the virus, isolated from patients who suffered only mild respiratory systems, also produced only the mildest symptoms in mice.

The researchers divided the H5N1 strains into two groups: one that caused the mice to die, and one that was relatively benign.

Flu potency

They then used gene technology to pinpoint PB2 as the gene that gives the lethal strains their potency.

The key appears to be a tiny change in just one component of the gene.

The function of the PB2 gene is not completely understood, but scientists believe it controls production of chemical that helps the virus to reproduce itself in the cells of its host.

Professor Ron Eccles, director of the Common Cold Centre in Cardiff, told BBC News Online that scientists did not know a great deal about what made a virus virulent.

"This is interesting research because at present we have no real knowledge to help us predict what a virus is going to do. We just have to look at the results."

The research is published in the journal Science.