Time For Sitting On H5N1 Bird Flu Data Is Over
Recombinomics Commentary
June 29, 2006
Indonesia has become the hot spot of avian flu, with the virus spreading quickly in animal populations, and human cases occurring more often there than elsewhere. Yet from 51 reported human cases so far - 39 of them fatal - the genetic sequence of only one flu virus strain has been deposited in GenBank, the publicly accessible database for such information.
The above comments from an editorial in today's Nature are welcome. Declan Butler offered some additional comments in his blog and provided a link to letter by Dennis J Kuchinich and Wayne T Gilcrest to appealing to colleagues to support a request to secretary Michael Levitt to make deposit of H5N1 sequences to a public data within 24 hours of verification as a condition for NIH support. The letter also supports transparency of these and related H5N1 bird flu issues.
The situation in Indonesia is particularly acute. It is now almost one year since the first confirmed human case in Indonesia was reported, yet sequences from that index case are the only human H5N1 sequences publicly available. The sequences were deposited in the WHO private database on August 1, 2005 but were not made public until March 25, 2006 and sequences from six of the eight gene segments have still not been released. Malik Peiris declined comment on direct questions on mutations in the sequences because the sequences were done under contract with WHO. WHO said they couldn't release the sequences without approval from the Indonesian government, and the Indonesian government said that no one had requested release.
The sequences are at Los Alamos and can be released to the public by simply removing the password protect, as was done with the index case. Data released at the Jakarta meeting also raised significant questions about what data is being hidden in the database
WHO's statement on the sequences in the Karo cluster indicated there were no "significant mutations" and no reassortment with human and avian genes. They also indicated that the sequences indicated sensitivity to Tamiflu. However, data presented at the closed door meeting indicated the sequences in the cluster had the S31N polymorphism in the M2 protein, which indicated the H5N1 was amantadine resistant.
Similarly, media was told that the nephew of the index case and his father shared a "minor mutation" indicating the son infected his father. However, data at the closed door meeting indicated the father had 13 additional differences with his son's H5N1 sequence strongly suggesting that the father had been infected by two H5N1 viruses, one from his son and one from an unknown source. The sequence of the H5N1 from the father had evidence for both reassortment and recombination.
WHO consultants have indicated the recombination in H5N1 was rare. However, analysis by WHO fails to detect recombination, which is common in H5N1. Obvious recombination is present in H5N1 generated by the very WHO consultants who say that can't find homologous recombination. These obvious cases were in sequences from Hong Kong in 2003 and 2003 and the sequences were generated via a collaboration between Hong Kong and St Jude labs. The recombination was not described in the peer reviewed publication of these sequences. Obvious recombination was also present in 2003 H9N2 sequences from Korea. These sequences were also generated by St Jude, yet they were also not reported in the peer reviewed paper,
H5N1 bird sequences from Indonesia have acquired Qinghai sequences, including the Qinghai HA cleavage site in at least one isolate from Bali. Similarly, Qinghai sequences are in the H5N1 from the father who was infected by his son. These Qinghai sequences on an Indonesian genetic background are additional evidence fro recombination in H5N1 sequences in Indonesia.
The sequence data is not being shared. The human sequences from Turkey were finally released. However, a request for identification of the WHO confirmed cases was not answered. Nor was a simple question about the cell line or chicken eggs used to isolate the H5N1 from Turkey because the pattern of S227N sequences was unusual.
The time for secrecy regarding these sequences has passed. Release of the sequences and providing minimal information on the source of the sequences should be made immediately. Many of the Indonesian human sequences have bee held for almost a year. The sequences from Turkey were held for six months.
H5N1 is rapidly evolving and this evolution should be matched by researchers who are hoarding sequences for publication. The time for release of the sequences has long since passed.
Media Sources
Phylogenetic Trees