Spontaneous Generation of WSN/33 in Korean Swine?
Recombinomics Commentary
February 22, 2005
>> Someone in the South Korean veterinary lab may have innocently pulled the wrong computer file of genetic sequence data into an e-mailed transmission to GenBank, resulting in the display of this potentially terrible viral code. The lab in question may have contaminated its research samples <<
The mysterious origin of the WSN/33 swine sequences at GenBank remains unsolved. Although the sequences were deposited on Oct 24, 2004 and were publicly available at the beginning of December, the fact that the existence of the associated viruses remains unresolved is truly remarkable. If the sequences are real, and no credible evidence has been presented to show that they are not, then at a minimum there was a major laboratory lapse that allowed a dangerous human virus to escape and infect swine in conjunction with avian flu viruses. As Julie Gerberding said today, people, pigs, and poultry are a dangerous combination when it comes to avian flu. That defines the unresolved situation in Korea.
It is hard to imagine someone suggesting the sequences at GenBank are a wrong computer file. The data deposited are extensive. There are eight genes per isolate and there are six isolates with human sequences, so there are 48 genes involved. The six swine isolates have 7,7,5,4,4, or 3 human genes, which means there were 30 human and 18 avian gene sequences submitted. The genes were not random. The two isolates with 7 human genes are both H1N1 and have a PB2 avian gene. The two isolates with 4 human and 4 avian had the same combination of genes. Moreover, although all of the human genes clearly came from WSN/33, virtually all were different than each other and different than WSN/33 (although all homologies were greater than 99%).
Thus, simply looking at the sequences alone would have eliminated the "wrong file" nonsense. Most of the participants exchanging e-mails were well aware of the fact that the lab depositing these sequences knew that these were novel human / avian reassortants. They knew this long before the sequences were made public.
The issue of contamination is always a difficult one. However, in this case it seems unlikely. The lab says it does not have WSN/33 in the lab or in the facility. Moreover, the viruses were not isolated in cell cultures, but were isolated in chicken eggs. Thus, it is hard to see how so many different WSN/33 sequences could have gotten into all six isolates in a non-random manner. Moreover, the polymorphisms followed the rules of influenza evolution, so the polymorphisms were not created by sequencing errors.
However, the above difficulties did not prevent South Korea from telling WHO that the sequences were a lab error, using yet another human virus as evidence. It is extremely difficult to even come up with an improbable scenario to implicate a different human sequence. The data at GenBank are public, and the sequences do not have this other human sequence. Moreover, no one has been able to confirm these other sequences because the material has not been released. Even if confirmed, there would still be the questions of where did 30 different WSN/33 sequences originate, and how did they get deposited at GenBank as sequences from six swine?
The introduction of the new human genes certainly does not provide answers. They simply raise more questions about the origin of the 30 WSN/33 genes, and the ability of industrialized nations to determine when or if they have been attacked by a bioterrorist.
Media link