No. 8: Avian Influenza Virus (H5N1)—A Few Amino-Acid Substitutions Shy of a Pandemic

As of December 18, the World Health Organization reported 77 cases of avian influenza worldwide, most of which occurred in Indonesia (40 cases). While the case number of infection this year is lower than that tallied in 2006 (115 cases), the mortality rate of human disease, at 60%-70%, remains high, and small outbreaks in wild and commercial poultry continue to be reported in Europe, raising imminent concerns about the eruption of bird-to-human illness beyond Southeast Asia. In the United Kingdom, H5N1 was detected at a turkey farm in January 2007—the first-ever report of the disease in commercial poultry in that country, and in November, the virus was identified in a flock of free-range* turkeys in England.

 

The reason we continue to care about avian influenza, beyond its documented and would-be effects on the commercial poultry industry (among other things), is the enduring potential of the virus to mutate, thereby enabling person-to-person transmission and blowing the human race to kingdom come. Veterinary investigators in Japan, Vietnam, and Wisconsin gave us an idea of the necessary steps of this transmutation process in a neat, attention-grabbing study, published in PloS Pathogens in October.

 

Hatta et al assessed the growth and replication of 2 highly virulent H5N1 isolates, VN1203 and VN1204, which were cultivated from the same 10-year-old Vietnamese boy who died of avian influenza in 2004 (Maines TR et al. J Virol. 2005; 79:11788–11800.). The 2 isolates differ in their coding for a total of 6 amino acids, but the authors elected to concentrate on the difference at position 627 of the cap-binding transcriptase (PB2) protein, because the host specificity of influenza A viruses (that is, human vs avian) appears to be determined, at least in part, at this site.^† VN1203 codes for lysine (Lys) at the 627 position, and VN1204 codes of glutamate (Glu). As a rough generalization, most human subtypes have Lys at 627 (PB2-627Lys), and avian subtypes typically have Glu (PB2-627Glu) at this same location.

 

The authors observed that Lys-encoded VN1203 and an engineered mutant of VN1204 with Lys at position 627 demonstrated greater replicative capacity in mice than wild-type VN1204 or an engineered version of VN1203 containing Glu at 627. Consequently, the Lys residue at position 627 of the PB2 viral protein appears to elevate viral-replication capabilities. This finding notably held true in mammalian cells at relatively lower temperatures—namely, temperatures simulating those of the upper-repiratory tract.^‡

 

It is therefore proposed that several H5N1 mutations, including the PB2 mutation, are necessary to confer efficient person-to-person transmission of avian influenza. Another cited mutation of importance consists of amino-acid substitutions in the viral-surface glycoprotein hemmaglutinin (HA), which detemines host-cell recognition. Mutated variants of HA bind to saccharide-containing cell receptors with either a terminating sialic acid α-2,3-galactose (SAα2,3Gal) on avian cells or sialic acid α-2,6-galactose (SAα2,6Gal) on human cells. The authors speculate that PB2-627Lys (which broadens susceptible cell types and enables low-temperature replication), SAα2,6Gal recognition (which enables human-cell binding), and other as-yet-unknown mutations are necessary for a human pandemic.

 

* I’m not sure why WHO makes a point of identifying these commercial birds as “free-range,” unless they were exceptionally free ranging.

† In addition, the PB2-627Lys variant may be associated with higher virulence and heartier replication capabilities. Specifically, the activity of its RNA polymerase is efficient at relatively low temperatures.

‡ A curious aside is that, in mice, titers of the engineered VN1204PB2-627Lys were appreciably higher than those of VN1203 (which also has PB2–627Lys), suggesting that coding differences beyond the 627 position determine growth efficiency of the virus.