Bird Flu Reassortment Experiments Under Natural Conditions
Recombinomics Commentary
January 26, 2005
>> The only way to answer the question is to produce reassortment in a secure laboratory, and one laboratory is doing that, Stohr writes. He doesn't name the lab, but the US Centers for Disease Control and Prevention recently revealed that it would conduct reassortment experiments. Stohr's article says this research may not be completed before the end of this year. <<
There is more than one way to get answers about reassortment between human and avian genes commingled in the same host, and the "experiment" is ongoing in swine on farms in Korea. The environment is far from a "secure lab', and the potential for a serious outbreak appears to be great.
The secure lab experiment by the CDC has a few drawbacks. It is focused on reassortment, but dual infections can produce both reassortment and recombination. Thus, the planned experiments using reassorted virus fall well short of modeling a "natural environment". A second set of experiments which infect the same cell with H5N1 and a human virus is more "natural", but many combinations of reassortants and recombinants are theoretically possible. Sorting out those that will emerge in a natural environment may require more selection factors than might be present under in vitro laboratory conditions.
The experiment that is being done in Korea has already produced results. The "experiment" uses WSN/33 as the human virus and a series of related H9N2 Korean viruses as the avian source. Since it is not clear how the WSN/33 made its way into swine in farms, it is not clear that the starting virus was WSN/33. It could have been some altered version because none of the six publicly available sequences have a full human PB2 gene. If all 8 WSN/33 genes were initially used, then there does seem to be some selection against this gene. It is possible that WSN/33 with all 8 genes is lethal and those swine have not been identified.
WSN/33 does have a number of genetic alterations that allow it to grow well in labs and kill mice. One of those alterations is at position 627 in PB2 which changes a glutamic acid to a lysine, which increases virulence. The virus is also missing a glycosylation site on NA so it is able to sequester plasminogen to facilitate viral entry into cells. This allows the virus to grow well in tissue culture, but also contributes to its ability to grow in a variety of tissue types including neurological tissue.
The six isolated viruses are clearly reassortants, but four versions have been found. The number of human genes is 7 for the two H1N1 isolates, and 5,4, or 3 for the 4 H9N2 isolates. However, in addition to a variety of reassortments, the viruses also have recombined PB2 in some instances to put human sequences at the 5' end of the gene and avian sequences at the 3' end. Similarly, there is recombination in the NA gene in the H9N2 isolates.
Thus the "natural" experiments show that many combinations are possible involving reshuffling of whole genes (reassortment) and creation of new genes (recombination).
This ongoing experiment still has not been acknowledged by the WHO even though the sequences have been publicly available for almost 2 months. Likewise, there has been no discussion about issuing alerts for virulent H1N1. The H1N1 isolates have 7 of 8 WSN/33 so they have an NA missing the glycosylation site. They also have a mutation at position 31 making the virus resistant to the anti-viral drugs amantadine and rimantadine. Since the H and N are very similar to the 1933 version, most people born after 1933 will have limited immunity.
Thus, the "experiment" above offers some answers about dual infections in hosts, but since the new isolates are spread out among many farms in Korea, it is not clear if other versions are in swine, birds, or people.
Media link