In a virological time warp, researchers have determined the crystal structure of a key protein of the flu strain that swept the world in 1918 and 1919. The study helps explain why the virus was so devastating, says lead researcher John Skehel of the Medical Research Council National Institute for Medical Research in London.
A handful of flu pandemics have occurred in the 20th century; the 1918 one, also called the Spanish flu, was the most deadly, killing between 20 million and 40 million people. No one knows where the virus originated. Some evidence suggests that it came from birds, but bird influenza viruses usually don't infect people. (The H5N1 strain currently sweeping across Asia is the exception, not the rule.) That left researchers to puzzle over what made the 1918 virus so virulent.
During the 1990s, a team led by Jeffery Taubenberger at the Armed Forces Institute of Pathology in Washington, D.C., sequenced key gene fragments of the 1918 flu strain, recovered from frozen victims found in the Alaskan permafrost and in archived autopsy material. Now, Skehel, along with colleagues at Harvard, Yale, and Emory University in Atlanta, Georgia, has used that sequence to build the virus's hemagglutinin (HA)--a protein that latches onto receptors on the host cell surface--and determine its structure. Another team, which included Taubenberger and was led by structural biologist Ian Wilson of the Scripps Research Institute in La Jolla, California, determined the structure of the same protein's precursor. Both teams reported their findings online in Science on 5 February.
From its sequence, researchers already knew that the 1918 HA gene resembled that of avian flu viruses. But the new protein structure suggests that the 1918 strain, despite its avianlike sequence, had a shape that helped it bind tightly to the human receptor and infect cells, Skehel says--which helps explain why the 1918 strain spread so well in the human population.
That hypothesis does not quite convince Graeme Laver, a retired flu researcher at Australian National University in Canberra. The fact that a virus binds well does not mean it has to be particularly virulent, he says. Still, Laver says, determining how a protein from a long-extinct virus looks "is really a remarkable piece of work."
Unfortunately, says Skehel, the study cannot answer questions about the likelihood of H5N1 causing an epidemic, because the 1918 virus's H1 hemagglutinin was very different from the H5 type found in the virus now circulating.