Published online 16 April 2008 | Nature | doi:10.1038/news.2008.759


Flu hotspot found in Asia

Influenza strains then radiate worldwide.

The breeding ground for new influenza strains is centered in East and Southeast Asia, new research suggests. The virus bounces around countries in the region before emerging worldwide to cause seasonal epidemics.

Each year, officials struggle to predict which strains of flu will spread globally, killing up to 500,000. Pinpointing a reservoir for the disease could help focus surveillance efforts on countries where new, harmful strains get their start six to nine months before appearing elsewhere.

The influenza virus evolves in local reservoirs before spreading worldwide.CAMR / A.B. DOWSETT / SCIENCE PHOTO LIBRARY

The conclusions come from two independent studies of the influenza virus. One, released today by Science, locates the virus hotspot by mapping global changes in hemagglutinin, the influenza protein that can trigger or bypass the human immune response1. The other paper, appearing today in Nature, examines the virus's full genome to show that a reservoir must exist for new flu strains2. New variants can lurk in this reservoir for years before a rapid series of critical rearrangements allows the virus to emerge.

The flu hotspot theory is not new, but the results seem to provide a more comprehensive picture of how influenza emerges and makes its way around the world. “This is not a new hypothesis, but this is new evidence,” says computational biologist Steven Salzberg of the University of Maryland in College Park. “There’s a lot more data now, and it makes it a lot more clear that’s what's going on.”

Sink or swim

To get a detailed picture of how influenza circles the globe, the Science group analyzed World Health Organization data on roughly 13,000 human influenza A viruses (of the subtype H3N2) from six continents, collected between 2002 and 2007. H3N2 is largely responsible for today’s flu epidemics.

A trail of changes in the viruses’ hemagglutinins show that viruses each year seem to bear the most resemblance to strains collected in East and Southeast Asia. Cities that are near each other in the region may play host to epidemics at different times of the year, allowing the viruses to overlap and continually evolve. A ‘winning’ strain may then emerge from this reservoir to spread around the world some months later.

Having escaped the reservoir, a virus will spread to other continents before fizzling out at the end of a season. “Areas outside Southeast Asia are essentially the evolutionary graveyards of influenza virus,” says co-author Colin Russell, a biologist at the University of Cambridge in the UK.

A reservoir is only one idea of how flu virus is thought to spread. Some epidemiologists have speculated that the virus may linger in a region, waiting for the next cold season. Others have suggested that flu may move back and forth between the northern and southern hemisphere, an idea that would explain why flu epidemics seem to pop up in winter months. This new study does not suggest an explanation for seasonal variation in flu incidence.

Still, pinning down the location where these viruses emerge could help vaccine officials predict which strains will emerge in the coming year, say study authors. “This now tells us where to look to find out what the evolutionary pressures are in detail,” says co-author Derek Smith, a bioinformaticist with the University of Cambridge.

Shuffling the deck

Understanding how new strains of influenza are selected could be the key to forecasting epidemics farther in advance. Authors of the Nature paper say that although some hemagglutinin variants seem perfectly suited to crash the immune system, sometimes the protein is configured unfavourably to work with the rest of the virus.

“It’s not just the hemagglutinin that’s evolving that’s critical. It’s how the hemagglutinin interacts with the rest of the genome,” says biologist Edward Holmes of Pennsylvania State University in University Park.

Holmes and colleagues used 1,300 publicly available genome samples from New York state and New Zealand to look at viral diversity over 13 seasons of flu. The team found the virus does not linger in either area from year to year. Instead, slightly or dramatically altered versions seem to emerge each year from some reservoir.

Some flu strains seem to undergo rapid genetic rearrangements and, after lingering for a number of years in the background, emerge as a new, epidemic variety. The data show this occurred in the 2003–2004 season, one period when the flu vaccine failed.

Analyzing both H3N2 and H1N1, the subtype that caused the 1918 outbreak, the team also confirmed an alternating relationship between the two flu strains3. As a particular variant of H3N2 weakens, sometimes a virulent strain of H1N1 will emerge, as it did last year.

The path to better flu surveillance is unclear. While Russell and Smith call for increased surveillance in East and Southeast Asia, more focused monitoring may not help predict which strains emerge.


Although it may be time to re-examine the balance of surveillance, “I don’t think they’ve made the argument that increased surveillance in the tropics is necessary,” says Salzberg. “What we really care about is not which viruses are going around in the tropics, but which ones are fit enough to escape.” 

  • References

    1. Russell, C.A. et al. Science 320, 340–346 (2008).
    2. Rambaut, A. et al. Nature doi:10.1038/nature06945 (2008).
    3. Wolf, Y. I., et al. Biol. Direct. 1, 34 (2006).
Commenting is now closed.