The swine flu virus had three parents from two continents and appeared suddenly without warning, evading all routine flu surveillance and quarantine; sequence data suggest it may have been created from a faulty vaccine given to pigs in North America.Several papers reporting phylogenetic analyses of the gene sequences of the new pandemic swine-origin H1N1 virus (S-OIV) have been published. All show that S-OIV inherited its genes from parents that came from two well-known groups of swine flus. Flu viruses have 8 different genes and in mixed infections sometimes shuffle those genes to form ‘reassortants’ with new combinations of the 8 genes chosen from those of the parents. Several reports have shown that six of the genes of S-OIV came from a ‘triple-reassortant’ influenza virus (or viruses). These viruses have been common in North American pigs for more than a decade, and have never been found in Europe. The other two genes (NA and MP) came from Eurasian ‘avian-like’ viruses common in Europe for longer, but never found in North America. Both groups of viruses have however been found recently in pigs in SE Asia.
The analyses reveal an unusual feature of the S-OIV genes, which is that none of the genes have been found recently in swine influenzas collected during routine flu surveillance. The NA gene had not been sampled (identified in samples) for 17 years before it reappeared in S-OIV, and the others, including the MP gene, for around 11 years. Thus the NA and MP genes were most likely acquired by S-OIV from Eurasian ‘avian-like’ viruses on separate occasions, and therefore S-OIV probably had at least three parents.
We have done further specific analyses to find out which isolates contain the genes that are closest to those of S-OIV. We find that all are viruses of pigs. The NA gene of S-OIV is closest to that of European ‘avian-like’ H1N1 influenza viruses sampled in 1991/1993, its MP gene closest to that of H3N2 Asian ‘avian-like’ viruses sampled in 1999, and its other six genes are closest to those of North American H1N2 ‘triple-reassortant’ viruses sampled in 1999/2000. Note that the dates those isolates were collected agree with the calculated ‘time-line’ leading to S-OIV; NA 17 years ago, all the others 11 years.
In summary, S-OIV is a reassortant with at least three parents. The parents were sampled over a decade ago, all found in pigs, and in three very distant parts of the world; North America, Europe and SE Asia!!!
Where were S-OIV’s genes between the time that they were last sampled and 2008 when they all reappeared in S-OIV? In which virus or viruses, which host, and when and how did they get together? It is important for us to try to answer these questions as it might help us avoid similar pandemics in future.
There are two theories regarding the origin of S-OIV, as described below.
The “unsampled pig theory”, was published by Gavin Smith and his colleagues in Nature . They suggest that “the progenitor of the S-OIV epidemic originated in pigs”. The “long unsampled history observed for every segment” of the S-OIV genome “suggests that the reassortment of Eurasian and North American swine lineages may not have occurred recently, and it is possible that this single reassortant lineage has been cryptically circulating rather than two distinct lineages of swine flu”, and that “Movement of live pigs between Eurasia and North America seems to have facilitated the mixing of diverse swine influenzas, leading to the multiple reassortment events associated with the genesis of the S-OIV strain.” Note that Smith and his colleagues had not realized that their results showed that S-OIV had at least three, not two, parents.
Their theory has three parts. First, S-OIV’s parents reassorted to produce S-OIV at least a decade ago; second, the reassortant has been circulating in untested pigs since then, and third, it entered North America in the live pig trade from Eurasia.
The sequence databases show that during the past decade only the swine influenzas of Europe, North America and SE Asia were collected (and gene sequenced), so Smith and colleagues are correct in concluding that S-OIV, or its parents, could have been circulating undetected in, for example, South America or Africa or much of Asia. However the parent(s) of S-OIV would require at least two trans-continental trips to get together in one place, and hence this theory requires at least two quarantine failures, probably three. Is this likely? Given that quarantine has kept the widespread Eurasian avian-like swine viruses out of North America, and the ‘triple reassortant’ swine viruses of North America out of Europe. Furthermore S-OIV is very infectious for both human beings and pigs, a characteristic that is probably an ‘emergent property’ of the reassortant, not of its parents, therefore it is more likely that the reassortment event producing S-OIV occurred immediately before it first appeared in humans in late 2008, than over a decade ago. If S-OIV had originated earlier then most likely it would have spread to the human and/or pig populations earlier, and would probably have been found.
The “vaccine theory” was contrasted with the “unsampled pig theory” in a paper we submitted for publication in an academic journal. We note that influenza viruses survive well in virus laboratories, that laboratories are not subject to routine surveillance, and that there are probably many laboratories in the world where a range of swine influenzas from different sources and continents are kept. These viruses are used for research, diagnostic tests and for making vaccines. Thus if laboratory activity was involved in the genesis of S-OIV, this would explain most simply why S-OIV’s genes had escaped surveillance for over a decade, and how viruses last sampled in North America, Europe and Asia could have got together. Note too that there have been several reports, especially by Deborah Mackenzie in the ‘New Scientist’, about swine influenzas in agribusiness piggeries, and the uncontrolled production of veterinary vaccines, mostly “multivalent” (i.e. containing several different viruses).
So how could a laboratory mistake produce S-OIV? The simplest scenario is that a multivalent vaccine was not fully sterilized. Multivalent ‘killed’ vaccines are mixtures of virus particles that have been grown in hen’s eggs and then chemically sterilized. Such vaccines are widely used in North American piggeries to control influenzas. Thus S-OIV might have been produced if insufficient sterilant, usually formaldehyde or propiolactone, had been added to the particle mixture. The live mixture would then infect pigs ‘vaccinated’ with it, and the growing viruses could then reassort to produce S-OIV.
It is significant that one of the North American H1N2 ‘triple reassortants’ closest to S-OIV is probably used in commercial multivalent pig vaccines in North America . This may be the reason why, as the S-OIV pandemic started, there has been no report of an outbreak of S-OIV in a pig farm in the USA, whereas it has been reported from unvaccinated pig herds in other parts of the world; two each in Canada and Australia, and one in Argentina. It is also relevant that the three likely parents of S-OIV are those one would choose if one were designing a multivalent swine influenza vaccine for international use.
There are historical precedents for laboratories being involved in virus outbreaks, not just foot-and-mouth disease and polio, but also influenza. For example there was the H1N1 influenza lineage that circulated in the human population for four decades after the 1918 Spanish influenza epidemic but disappeared during the 1957 Asian influenza pandemic and reappeared in 1977. The H1N1 that reappeared was found to be genetically very close to an H1N1 isolate collected in 1950, indicating that it had probably been held in a laboratory freezer between 1950 and 1977. There is also the recent incident of a commercial human test vaccine being found to contain live H5N1 virus.
So what can be done to distinguish between these two theories, and any others that are proposed? The phylogenetic patterns in the present gene sequence data do not distinguish between them - more data are needed.
First, samples of all isolates of influenza used in swine vaccines in North America in 2008 must be collected, and their gene sequences determined. This would immediately check the vaccine theory. Second, the refrigerators of vets throughout the world, especially those of South and Central America, Asia and Africa, should be checked for samples of swine influenzas collected over the past decade, and the gene sequences of the isolates viruses determined. Third, the quarantine authorities of North America should discuss, if they have not already done so, whether it is likely that pigs infected with S-OIV or its parents could evade quarantine measures when coming from other parts of the world.
It is important for the relevant authorities to obtain this evidence while the ‘scent’ is still warm - the search for the source of S-OIV must not be relegated to the ‘too hard’ basket, some possibilities can still be checked. Influenza is a significant and very costly cause of mortality and morbidity in the human population. If we wish to avoid new outbreaks rather than just minimizing the damage they cause, we must better understand what conditions produce them.
Jean Downie and Adrian Gibbs are viologists, they study the evolution of viruses using their gene sequences.
Prof. Adrian Gibbs and Dr. Jean Downie
This report has been submitted to the US FDA and CDC, and to Sir Liam Donaldson, UK’s Chief Medical Officer
1.Smith GJD, Vijaykrishna D, Bahl J, Lycett SJ, Worobey M et al. Origins and evolutionary genomics of the 2009 swine-origin H1N1 influenza A epidemic. Nature 2009, 459, 1123-6.
2.H1N1 Influenza Veterinary Talking Points - 5/5/09, http://www.aasp.org/public/H1N1InfluenzaVeterinaryTalkingPoints-05-05-09.doc