Accessibility links

Breaking News

Researcher Tries to Stay One Step Ahead of Flu


Like all organisms, viruses evolve over time. Scientists are looking for ways to predict changes in flu viruses that could allow vaccine manufacturers to anticipate flu outbreaks, rather than scrambling to respond after people have gotten sick from a new strain.

Now scientists are one step closer to more accurate predictions of viral changes, thanks to Pennsylvania State University researcher Katia Koelle, who is working on a computer program for predicting changes to influenza viruses. She says mathematical models need to take into account how our immune systems produce antibodies against proteins on the surface of the virus.

"The antibodies that our immune system makes recognize the structure of these proteins on the virus' surface," she explains. "One of these proteins is the hemogluttenin protein, which is really the dominant protein that our immune system creates antibodies against."

Koelle says hemogluttenin-surface proteins are constantly evolving, even if their shape isn't changing. These small changes don't fool our antibodies, which continue to recognize the evolving virus, and continue to provide immunity. "

"All these mutations are accumulating in the protein, but there's no significant change in the protein shape," Koelle explains. But eventually, she says, there's one more mutation that tips the scales. "That's the kind of mutation that 'breaks the camel's back.' It's the one that causes a very large shape change in the protein."

Because of the large change on its surface protein, the flu virus has literally changed beyond recognition. And that's when vaccine manufacturers need to update their vaccines. The virus looks different enough to the immune system that immunity disappears and we can get sick again with this 'new' flu.

Koelle says her model is the first to account for this evolutionary process on the viral surface and it predicts some of the changes that can occur. But she says scientists are still a long way from predicting what new shapes proteins will adopt when they change. "The model indicates that it's still incredibly difficult -- but not impossible -- to create vaccines for upcoming strains." Part of the problem is that there are so many shapes that the hemagluttenin can change into. Koelle concludes, "it seems like [creating vaccines] still very much has to be reactive, and actually predicting what the flu is going to do is incredibly difficult."

Koelle says increased flu surveillance could help improve the predictions. She says that the more information she can put into the model, the closer she may be able to come to predicting what surface proteins on flu viruses might look like in the future. Her research appeared in the journal Science.