Homologous Recombination in PB2 in Korean Chicken
Recombinomics Commentary
March 4, 2005
>> They won't embrace a new idea like Niman's until it's published in a peer-reviewed journal -- a scientific publication that sends research papers out to various experts for comments and approval before it will print the study.
"Without this kind of quality control it's very difficult to find meaningful results and make conclusions," Hoffman said. "It's not a perfect system, but it's better than just spreading information (without) quality control." <<
Although acceptance of the swine paper by the Korean group is waiting for confirmation by an outside party, the Korean and St Jude labs have just published a paper together in the peer reviewed journal, Virology, entitled "Avian influenza viruses in Korean live poultry markets and their pathogenic potential".. In this case the sequencing was done in Memphis. The Korean lab just supplied the samples. Remarkably, the same PB2 gene that has recombined between Korean avian H9N2 and human WSN/33 H1N1 in 2004 swine has also recombined in at least one of 2003 chicken isolates. However in this case the recombination is between H9N2 PB2 Korean genes and H9N2 PB2 Hong Kong genes. The cross over is again near the middle of the gene, and again the amount of the foreign (Hong Kong) gene present in Korea cannot be determined because only a partial sequence has been submitted to the gene depositories, and like the swine sequences, additional information is missing at strategic positions.
The swine data at GenBank and Los Alamos can be graphically viewed on the computer screen. The top panel shows the polymorphisms in the full swine sequences (A/swine/Korea/S10/2004, A/swine/Korea/S175/2004, A/swine/Korea/S452/2004) which are in gold and on the bottom is the full A/WSN/33 sequence in gray. In between are two partial swine sequences, A/swine/S81/2004 and A/swine/Korea/S109/2004. The available sequence for S81 is all avian while S109 has WSN/33 sequences at the 5' end of the submitted sequence.
In the lower panel, the missing information is filled in showing that all of the missing information was WSN/33 and both PB2 genes are recombinants, although the cross over positions are slightly different. However, both genes are chimeras with the 5' half coming from WSN/33 and the 3' half coming from Korean avian sequences. However, the swine data has not been published in a peer reviewed journal because the issue of contamination was raised in e-mails between St Jude and WHO. Although it has been over 4 months since the sequences were submitted to GenBank, and 3 months since the e-mail exchanges, a definitive conclusion on the sequences has not been reached.
The sequences used for the recent Virology paper have also been deposited at GenBank and Los Alamos, but those sequences contain no WSN/33 sequences. Instead the H9N2 sequences are all avian. However avian sequences in Korea are quite unique and readily distinguished from avian sequences in Hong Kong. Because of the large number of differences, it is easy to see recombination between viruses isolated from the two regions.
In the PB2 gene, three of the 2003 H9N2 chicken isolates (A/chicken/Korea/S18/03, A/chicken/Korea/S16/03, A/chicken/Korea/S4/03) are closely related to each other and they are also related to the 2004 swine isolates that are all avian in PB2 (S10, S175, S452). All of these Korean sequences are quite different than a 2003 Hong Kong H9N2 isolate, A/chicken/Hong Kong/NT366/03 or a 2000 H9N2 isolate from adjacent Guangdong, A/chicken/Guangdong/10/00. Complete sequences for the China chicken and Korea swine sequences have been deposited, but only the 5' half of the 2003 Korean chickens were submitted. The S4 is Korean-like up to position 1077, where it abruptly changes to the China sequence. It is China-like until the end of the submitted sequence at position 1221.
Like the swine sequences on the computer screen in the top panel, it is clear that recombination has happened, but it is unclear how far the recombination goes. For the other two Korean sequences, the submitted portion is all Korean. However, the two sequences stop shortly after the cross over, so although it is clear that they would have different cross over points, it is not clear if these two genes are all Korean, or if they too switched to the China sequences.
However, it is clear that the publicly available sequences of Korean isolates from swine in 2004 and chickens in 2003 show recombination of the PB2 gene and the cross over was near the middle of each gene.
Although the recombination is quite clear in the deposited sequence, there is no mention of the recombination in the published paper, and additional information on the missing regions has been unavailable.
Thus, recombination in PB2 in Korean isolates is clearly present in the controversial swine sequences as well as the peer reviewed published chicken sequences.
Media link