It is interesting that the A/Hong Kong/213/03 (H5N1) virus, which was isolated from a human and recognises both SA2,3Gal and SA2,6Gal (ref. 6), bound extensively to both bronchial and alveolar cells.
The above comments in tomorrow’s Nature may explain the large cluster in Turkey associated with S227N detected in the index case. S227N had previously been shown to have reduced affinity for avian receptors (SA2,3Gal) and increased affinity for human receptors (SA2.6Gal). Although the recent paper in Science failed to find increased affinity for SA2,6Gal, the S227N was on a genetic background from an isolate in Vietnam which was distinct from the Hong Kong isolate listed above. In Turkey, the S227N was on a Qinghai strain background, which is different than both the Vietnam or Hong Kong strains used in the lab experiments.
However, the size of the cluster in Turkey suggests that H5N1 was more efficiently transmitted human-to-human, which would be consistent with increased affinity for SA2.6Gal receptors. The H5N1 from the index case was said to be a mixture of S227N and wild type S227. Only wild type was detected in the sister of the index case, in contrast to the Hong Kong isolates which had S227N from both the father and son. However, the Hong Kong isolates were grown on a canine cell line, MDCK, as was the H5N1 used in the Nature paper. It is unclear what was used to isolate H5N1 from the index case and his sister. Isolation of H5N1 on chicken eggs can reduce or eliminate S227N.
However, changes in the levels of S227N and wild type could have been modulated in the patients. H5N1 was isolated from lung tissue, which may have selected against S227N in vivo, since the above data suggests the wild type would grow more efficiently in lung tissue.
Thus, the S227N could have facilitated growth in the upper respiratory tract, while S227 could have been more involved in growth in the lower respiratory tract. Genes encoding S227 could have entered the upper respiratory tract in a pseudo-typed H5N1.
These data could also explain the cluster of cases in Azerbaijan which were fatal, yet efficient transmitted between the index case, her two cousins, and a close friend.
More sequence information on the Azerbaijan cluster would be useful.