Like fashion devotees, influenza viruses change their looks every year. Mutations in the viral surface help the virus elude the human immune system, leading to yearly epidemics. Now, researchers have developed a mathematical tool that helps analyze and visualize this "antigenic drift." The study, applicable to other viruses as well, may help make the yearly choice of the best flu vaccine easier.
The study focused on H3N2, a flu strain that first afflicted humans in 1968. To keep track of its evolutionary tricks, researchers worldwide routinely infect animals with new H3N2 strains, harvest antiserum, and test how well the antiserum neutralizes other, known varieties of H3N2. When antiserum triggered by one strain neutralizes another strain, the two are presumed to be close, and people exposed to one virus may expect to be immune to the other; if they're a bad match, the two strains are more distant. But lacking a visual representation, the vast collection of "tables, lines, and numbers" was difficult to handle, says flu researcher Richard Webby of St. Jude Children's Research Hospital in Memphis, Tennessee.
A map was needed, says computer scientist Derek Smith of Erasmus University in Rotterdam (EUR), the Netherlands. Just like the distance table in the back of a road atlas can be used to produce a rough map of a country, the "distances" between viruses gleaned from antisera tests can be used to plot each strain on a map. In a study published online this week by Science, Smith and colleagues--including EUR virologist Ron Fouchier and physicist Alan Lapedes of Los Alamos National Laboratory in New Mexico--present such a map, based on data going back to 1968. It shows the virus gradually moving, in a series of 11 clusters of closely related viruses, to escape human immunity.
To test the map's reliability, the team predicted how well antisera and virus strains that had never been paired before would match up, based on the distance between them on the map, and then carried out the lab test to verify their prediction. They found that the predictions were consistently reliable.
"That gives me comfort that what they're doing is real," says Webby. Being able to visualize antigenic drift and make predictions about how well serum against one strain protects against another should also help the annual process of selecting strains for the flu vaccine, he says.
The World Health Organization's influenza site