Highly Likely H5N1 On South Korean Farms
Recombinomics Commentary
November 23, 2006
According to the ministry, 6,000 of 13,000 chickens at the farm died between Sunday and yesterday. After an initial investigation, the quarantine service said bird flu was the suspected cause. The final test results will be announced on Saturday, but experts at the service said the virus was "highly likely" to be a virulent strain transmissible to humans.
The above comments support the outbreak of the H5N1 Qinghai strain in South Korea. The last reported outbreak of H5N1 in South Korea was in late 2003. That was followed by an outbreak of closely related H5N1 in Japan in early 2004. Sequences from these isolates were subsequently found in the Qinghai isolates from Qinghai Lake in May of 2005, supporting acquisition of these multiple polymorphisms via recombination.
The sequences in the Qinghai strain could be traced to the H5N1 in Korea/Japan, as well as low path H5 in Europe and southeast Asia and high path H5N1 in Shantou in southeastern China.
The fixing of the Qinghai H5N1 strain in long range migratory birds at Qinghai Lake was followed by H5N1 outbreaks in southern Siberia, Mongolia, Kazakhstan in the summer of 2005. The Qinghai strain subsequently spread to India, Afghanistan, Europe, the Middle East, and Africa. Although the majority of H5N1 isolated in southern China was the Fujian strain, the detection of only one Qinghai isolate was largely due to selective surveillance, which was focused on H5N1 in feces in live poultry markets. The predictions of a wave of the Fujian strain displacing the Qinghai strain is unlikely, although such a scenario was mention in the recent PNAS paper on the spread of the Fujian strain in southern China.
2006 isolates of H5N1 in Afghanistan, Mongolia, and Tyva in southern Siberia have been Qinghai, signaling the continued dominance of Qinghai in migratory birds, which are the likely source of the outbreak in South Korea. Similarly, the sequence of the H5N1 from the recent fatal infection in Egypt was also the Qinghai strain, signaling more Qinghai outbreaks this season in Europe, the Middle East and Africa in birds and people
H5N1 evolves via recombination, so the sequence database provides a history of transmission and dual infections involving H5N1. This database predicts that the isolates in South Korea will share many polymorphisms with the 2006 isolates in Mongolia and Tyva. However, many additional serotypes (H3N2, H6N1, H9N2) were found in live markets in South Korea, so sharing of polymorphisms with these isolates is also possible.
These predictions, however, are dependent on a complete and robust sequence database. The current public database is heavily biased by partial and withheld sequences, as well as poor surveillance. The detection of just one Qinghai isolate in eastern China is but one of many examples of poor surveillance. This is compounded by the multiple partial H5N1 sequences that have been deposited for Asian isolates collected between 1999 and 2006. Although the sequences have been published, and a free sequencing program is available through NIAID, the labs publishing the data have elected to withhold the samples and sequences.
This behavior should cease. These hoarding labs act as consultants to the WHO, and therefore are mailed samples from around the world for confirmation. The confirmed samples are used to isolate and sequence H5N1, but the sequencing labs control the data and contribute to a WHO sponsored private database at Los Alamos that can be accessed by a handful of WHO consulting labs. Since most of the sequences have been released in partial form or discussed in peer review publications, the excuses used to hoard the data are no longer operative.
St Jude and the University of Hong Kong have collected most of the published H5N1 sequences. The partial sequences include isolates from 1999 through 2006 and have been discussed in multiple publication. Moreover, St Jude has used the NIAID program to generate and release over 1000 flu sequences, but all released data are from sequences other than H5N1. Similarly, the WHO affiliate lab in Weybridge is hoarding over 1000 of the sequences from the H5N1 outbreaks in late 2005 and early 2006 in Europe and the Middle East, including human sequences with important receptor binding domain changes, some of which have also been discussed in peer reviewed publications.
The continued hoarding of these sequences prevents independent confirmation of data presented in the peer reviewed papers in high profile journals such as Nature, Science, and PNAS as well as more specialized journals such as Emerging Infectious Diseases, Virology, and Journal of Virology.
These WHO affiliated labs have maintained that the changes in H5N1 sequences are due to random mutations, and future vaccine targets cannot be predicted. However, the public sequence data includes obvious examples of recombination, and most of the "random mutations" are readily found in prior sequences, supporting acquisitions of the "point mutations" by recombination.
The hoarding of these sequences, including those that were described years ago, undermines the underlying concept of peer review, which demands independent confirmation. On a more short term, practical level, the hoarded data limits the generation of current PCR primers or selection of appropriate vaccine targets.
The hoarded sequences were largely generated by public funds from samples collected by public health agencies. The withholding of the sequences for years, as H5N1 rapidly evolves, should not be tolerated by the funding agencies, or the public, which funds the funding agencies.
Media sources
Phylogenetic Trees