Dave Roberts alerts me to his new bird flu blog. The site contains lots of interesting information, including links to two studies from the Journal of Virology. The first study looks at how H5N1 infects mice:
We recently analyzed a series of H5N1 viruses isolated from healthy ducks in southern China since 1999 and found that these viruses had progressively acquired the ability to replicate and cause disease in mice. In the present study, we explored the genetic basis of this change in host range by comparing two of the viruses that are genetically similar but differ in their ability to infect mice and have different pathogenicity in mice. A/duck/Guangxi/22/2001 (DKGX/22) is nonpathogenic in mice, whereas A/duck/Guangxi/35/2001 (DKGX/35) is highly pathogenic. We used reverse genetics to create a series of single-gene recombinants that contained one gene from DKGX/22 and the remaining seven gene segments from DKGX/35. We find that the PA, NA, and NS genes of DKGX/22 could attenuate DKGX/35 virus to some extent, but PB2 of DKGX/22 virus attenuated the DKGX/35 virus dramatically, and an Asn-to-Asp substitution at position 701 of PB2 plays a key role in this function. Conversely, of the recombinant viruses in the DKGX/22 background, only the one that contains the PB2 gene of DKGX/35 was able to replicate in mice. A single amino acid substitution (Asp to Asn) at position 701 of PB2 enabled DKGX/22 to infect and become lethal for mice. These results demonstrate that amino acid Asn 701 of PB2 is one of the important determinants for this avian influenza virus to cross the host species barrier and infect mice, though the replication and lethality of H5N1 influenza viruses involve multiple genes and may result from a constellation of genes. [My emphasis.]
A second paper brings more worrisome news: the 2004 strand of bird flu is more virulent for mammals than its 1997 version.