By David Brown
Washington Post Staff Writer
Thursday, March 23, 2006
Two research teams have independently discovered explanations for the chief features of the H5N1 bird flu virus -- its difficulty infecting humans, and the deadly effects when it does.
Unlike influenza viruses that are passed easily between people, H5N1 has a hard time attaching to cells in the nose, throat and upper airways. But it readily attaches to cells deep in the lungs.
This suggests that people need close and heavy exposure to the H5N1 virus for it to get into the lungs. But once it takes hold, it causes extensive damage to the machinery of respiration -- the cells and air spaces where oxygen is exchanged for carbon dioxide.
That scenario mimics the clinical experience of many of the 184 human cases of bird flu that have been officially recorded since late 2003. More than half have been fatal.
"This is information that we need to know -- fundamental and exciting," said Robert G. Webster, a leading flu virologist at St. Jude Children's Research Hospital in Memphis. He was not a member of either of the research groups, one of which published its results in yesterday's issue of the journal Nature, and the other online in Science.
Influenza infection occurs through a series of steps, with a key early one involving hemagglutinin, a protein that coats the virus's outer envelope. That protein attaches to sugar molecules that lie on the surface of many cells.
Virologists have known for years that human flu viruses attach through a sugar "linkage" designated alpha 2,6. The vast family of avian flu viruses favors a linkage with a different shape, designated alpha 2,3.
A research team led by Kyoko Shinya at the University of Wisconsin and the University of Tokyo looked at what cells in the human respiratory tract contain which linkage. They found that alpha 2,6 -- the receptor for human flu -- predominated in the nose and down the airways to the microscopic passageways that lead to the air sacs, or alveoli. At that point, cells with the alpha 2,3 linkage -- the receptor for bird viruses -- become common. Human viruses attached to the upper airways, while avian viruses attached to cells deep in the lungs.
The Japanese researchers, who published their findings in Nature, also tested an H5N1 virus taken from a person who died of bird flu in Hong Kong in 2003. That microbe's hemagglutinin recognized both the human and bird linkages, and it attached to cells from nose to lung. Most H5N1 samples isolated from people do not generally follow that pattern. The ability to bind to alpha 2,6 human receptors is one of several features that would have to become dominant for the H5N1 virus to become easily transmitted from human to human.
In the other paper, Thijs Kuiken and colleagues at Erasmus Medical Center in the Netherlands exposed H5N1 to blocks of tissue taken from all along the respiratory tree. They found the bird virus attached predominantly to cells at the entrance to the air sacs and in them, as well as to immune-system cells called alveolar macrophages that patrol the area.
This same pattern was seen in cats and ferrets. (Ferrets have traditionally been considered the species whose flu infections most closely mimic those of humans.) In mice, however, the pattern was reversed. H5N1 attached best to cells in the windpipe, not the lung -- suggesting mice are not the best animal model for research on this strain.
In many of the fatal H5N1 infections, people's lungs were filled with fluid, the result of an out-of-control inflammatory reaction. Kuiken speculated that bird flu's affinity for macrophages -- which can release inflammatory chemicals -- may be part of the reason.