H5N1 Evolution in Action via Recombination
Recombinomics Commentary
May 4, 2005
The latest concerns voiced by officials at WHO and CDC are driven by the fact that the human-to-human transmissions, along with animal transmissions, in northern Vietnam are increasing. At the same time, the sequences of the northern H5N1 are changing. These changes include the loss of one amino acid. In the past the WHO has offered assurances that the H5N1 contained no reassorted human genes. Six years of H5N1 human infections in the absence of reassortment with human genes, has begun to have an effect on some of the comments regarding bird flu and looming pandemics. Reassortment is still cited on occasion, but now comments on mutations have become more frequent.
However, H5N1 has been evolving and expanding its host range via recombination, which is how all viruses rapidly evolve. The two pillars of influenza genetics, drifts by mutation and shifts by reassortment, will soon fall. These pillars are largely supported by a good deal of hand waving, and H5N1 isn't watching or listening. H5N1 and all influenza viruses are just like all of their cousins and distant cousins. They evolve via recombination, and because of their genetic diversity and frequent dual infections, they evolve more quickly than most.
The recombination in action can be easily described using the recent events in Vietnam. The missing amino acid is almost certainly in the poly-basic region at the HA cleavage site. To get into a cell H5N1 and all influenza viruses have to cleave HA. A string of basic amino acids at the cleavage site increases the number of proteases that can do the cleaving. H5N1 has had a long string of basic amino acids at the cleavage site since 1997, when H5N1 was first isolated from humans. The basic amino acids are well known, and now HPAI (highly pathogenic avian influenza) is molecularly defined by a string of basic amino acids at the cleavage site.
In 2003, several isolates in China were found to be missing one of the basic amino acids. These isolates were the Z genotype, as were the isolates from Vietnam, but the Vietnam and Thailand isolates had some unique amino acid changes (polymorphisms) not found in other H5N1 isolates, including the ones in China. It now sounds like the H5N1 in northern Vietnam is a recombinant containing the changes associated with infections in humans, and the missing amino acid associated with isolates from China.
Creating a recombinant with those two properties would be fairly straightforward. The 2004 H5N1 in Vietnam could kill humans, yet thrive in the guts of asymptomatic ducks. An infected duck in Vietnam in 2004 could fly north for the summer and get infected with a 2004 H5N1 from a duck in China. The dual infection would set the stage for recombination. The genetic material of H5N1 is copied by a polymerase complex encoded by H5N1. Contrary to influenza dogma, this polymerase rarely makes errors, or if it does, the errors are corrected, because the fidelity in isolates is extremely high. The identical sequence of some genes can be found in new isolates, in new locations, in new hosts, after years of replication.
Mistakes do happen after decades of replication, and H5N1 isolates with a missing amino acid at the cleavage site were found in 2003. However, acquisition of this change by the H5N1 now in circulation in northern Vietnam did not require this mistake to be made again, because it was already flying around China. Thus, when the polymerase in the dually infected duck began copying the HA from Vietnam, it switched over to the HA gene from China and copied the region missing the one amino acid. It could have then continued copying the rest of the China gene or switch back to Vietnam. In fact, it could have switched back and forth several times and could have done similar switching for all 8 genes in the dually infected host. Thus, it could have made many new recombinant genes in that one dually infected duck.
In the fall of 2004 the duck with the recombinant H5N1 could then have flown back to Vietnam where H5N1 started infecting people again. However, this time the cleavage wasn't as efficient because of the dropped amino acid at the cleavage site, so the cases were milder. When the WHO and CDC saw several examples of the missing amino acid associated with milder cases and larger clusters, they became concerned.
Such concern was easy to understand. Increased efficiency in human-to-human transmission of a recombinant virus that produces milder disease is the formula for a pandemic. It creates more hosts for more dual infections and more recombinants. H5N1 in humans offers more chances of a human being infected with H5N1 and H3N2, which could undergo a recombination similar to the one described earlier. However, this time the human receptor binding domain in H3N2 could be acquired. The recombinant H5N1 could achieve the type of efficiency seen in human flu viruses, which are almost impossible to control.
These changes of course set the stage for more recombination, when more humans are infected, and the H5N1 that has a 100% case fatality rate in southern Vietnam and Cambodia is still circulating. The greater the number of dual infections, the greater the chance of more recombination.
This leads to a greater chance of a recombinant with efficient transmission combined with a high case fatality rate to create a pandemic rivaling 1918.