New paper on H5N1 virulence gene

The headlines are exciting: Chinese scientists identify deadly gene in H5N1. The story is also upbeat:

Chinese scientists have identified a gene in the H5N1 bird flu virus which they say is responsible for its virulence in poultry, opening the way for new vaccines.

[snip]

"We can now understand how this virus becomes lethal and the molecular basis for its pathogenicity," Bu Zhigao at the Harbin Veterinary Research Institute told Reuters.

The science also turns out to be interesting, but on its face not the breakthrough the story implies. Maybe as we learn more we will find it is the key to everything, but it is more likely it is just another piece of a big puzzle.

Here is the gist. A major research agenda is to understand the relationship between viral genetics, as revealed by gene sequences, and the biology of host range and virulence. We already have some scattered data points. We know that the presence of extra amino acids adjacent to the cleavage site in the HA glycoprotein (see our series on glycoproteins), often referred to as the polybasic cleavage site, affects the number of tissues the virus can infect in poultry and is a necessary element in high pathogenicity. The E627K mutation on one of the virus's gene segments, PB2, is another factor that influences virulence of bird viruses in mice. Similarly mutations at position 701 makes duck viruses lethal in mice. A number of studies have also shown that changes in the NS1 segment of the eight segment viral genome plays a part in virulence in pigs. In other words, we have been looking at this problem in various ways and no some things. The new paper adds to this still meager store of knowledge by identifying a single genetic change in a goose H5N1 virus that makes it lethal to chickens.

In 1996 a highly pathogenic H5N1 virus caused an outbreak on a goose farm in southern China, Guangdong province. Two H5N1 viruses were isolated from the geese, who suffered a 40% mortality. One, designated A/Goose/Guandong/1/96 (GS/GD/1/96; see our entry in the Flu Wiki on naming influenza viruses) has been much studied; the other, designated GS/GD/2/96, less so. The reason for the lesser interest in the second virus was it was of low pathogenicity for geese and chickens. Its isolation was more of an incidental finding in the 1996 outbreak. The Chinese scientists have compared the two viruses and obtained interesting information about what makes one virulent to chickens and the other not.

Both viruses have the polybasic amino acid sequence adjacent to their cleavage site. Since one is high path for chickens and geese and the other not, it is clear this is not a sufficient condition (something which we've known). When the sequences of all the gene segments in the two viruses were compared, they were identical in the PB2, HA, NA, M2 and NS2 genes at the nucleotide level (the finest level of detail). In four other gene segments there were a total of eight nucleotide differences, of which three were "silent" mutations, that is, they were differences in the nucleotide sequences that didn't produce a difference in the amino acid sequence of the proteins. The other five changes produced differences in the PA, NP, M1 and NS1 genes. Thus, the differences in virulence to chickens in these two viruses were due to one or more changes in these four genes.

The Harbin team then constructed recombinant variations combining the genes in different ways. It turned out that only the change in position 149 of the NS1 gene determined whether the virus was virulent for chickens. It didn't seem to affect virulence in geese. Since NS1 expression in mammals is related to antagonism of a non-specific innate antiviral response, the release of interferons alpha and beta, they went on to show that in birds, too, the difference in NS1 in the two viruses is related to whether it can inhibit interferon production.

What have they shown? For starters, they have not discovered the virulence gene that turns H5N1 into a killer. They have shown that a specific change at one position of the NS1 gene segment in the Guangdong 1996 goose virus can affect host range and virulence of the virus by affecting its ability to interfere with interferon production. The host range change is from goose to chicken and the virulence increase is in chickens. So far there is no relation to virulence or infectivity in humans. Nor is there a showing that this is the only way H5N1 can become highly pathogenic in chickens. It appears to be one way.

Still, this is very interesting and elegant work. It directs our attention to NS1 and its ability to inhibit interferon production as a characteristic sufficient to turn an avirulent virus into a virulent one. It is like a big sign in the ground that has the letters, Dig Here, written on it.