H5N1 in Scotland Matches Denmark and Northern Germany
Recombinomics Commentary 14:43
December 28, 2007
The sequences of three H5N1 gene segments (PB2, PB1, PA) from a whooper swan, A/whooper swan/Scotland/1430/2006, that washed up on the shores of Scotland in March, 2006 has been released. All three sequences are closely related to wild bird sequences from northern Germany and Denmark (implied earlier in a phylogenetic tree). Also closely related to Germany and Denmark are sequences from another gene (NS) from a series of wild bird isolates from Sweden, also collected in early in 2006 and released this month. Although these sequences have been released almost two years after collection, they paint a very clear picture of a cluster of isolates sharing common regional markers found in northern German, Denmark, Sweden, and Scotland.
Although this relationship would not surprise students in a grade school geography class, the results are at odds with the long standing explanation of why H5N1 was not detected in an EU country in 2005. The widely circulated explanation suggested H5N1 migrated into eastern Europe in the fall of 2005 and the flew due west in the dead of winter in early 2006 in search of open water, and when they died in early 2006, tests of these birds identified H5N1.
However, the yarn falls apart when the sequence data is released, because the H5N1 in this cluster do not match H5N1 sequences from eastern Europe. Some of the markers match western Africa isolates in Nigeria and Niger, but don’t match sequences from eastern Europe. The same can be said for another series in southern Germany, which match sequences in Switzerland
The H5N1 sequences are quite reveal when released, but most have been hidden away in WHO’s private database, giving the yarn an extended lifespan.
Sequence data also identifies additional surveillance failures in the fall of 2006 and beginning of 2007. The reported cases of H5N1 in EU countries were again minimal. Hungary reported one outbreak, which appears to have spread to England via poor bio-security of the turkey processing company with facilities in both countries. However, in the summer of 2007, H5N1 reports from the Czech Republic, Germany, and France began to come out.
The most detailed reports were from FLI in Germany, which included details on the sequences. The summer 2007 sequences from wild bird isolates in multiple locations in central and southern Germany matched sequences from a massive Uva Lake outbreak in southern Siberia and northern Mongolia (Uva Lake is the largest lake in Mongolia and is adjacent to the Russian boarder (see satellite map). The outbreak in June, 2006 was similar in size to the Qinghai Lake outbreak a year earlier (Qinghai Lake is the largest lake in China). The Qinghai Lake outbreak was followed by outbreaks in Siberia and Mongolia in the summer of 2005, followed by the introduction of H5N1 into Europe, the Middle East, and Africa. However, after Uva Lake the reports of the strain were limited. Recently, sequences from South Korea infections at the end of 2006 were released and these were Uva Lake sequences. Similarly, reports indicate the outbreaks in Kuwait at the beginning of 2007 were also related to Uva Lake (the Kuwait sequences are still being hoarded at the WHO private repository).
The presence of the Uva Lake sequences in EU countries in the summer however, was most easily explained by an earlier migration of the sequences into the region in the fall of 2006 and detection failures until the summer of 2007. These detection failures were the basis of the announcement in Toronto in June, 2007 indicating that H5N1 had essentially disappeared from wild birds in Europe. This “disappearance” is also at the heart of the upcoming FAO report on reduced H5N1 activity in 2007.
The “invisible” H5N1 in wild birds however, has had a major resurgence. In addition to the outbreaks in Europe in the summer of 2007, Krasnodar reported an outbreak in September. The full sequence set of sequences were promptly released, and the sequences were clearly Uva Lake. Those sequences were followed by sequences from three of the summer outbreaks in Germany, and all three were Uva Lake. That outbreak was followed by an outbreak in free range turkeys in England, and those sequences were said to be like the summer Uva Lake sequences from Europe. The sequence of H5N1 from a whooper swan sequence in Krasnodar was released and it was 99.95% identical to the chicken sequence from Krasnodar (and the HA sequence was an exact match). More recently there have been multiple outbreaks in Poland as well as adjacent regions in Germany. The sequences from Germany were like the summer Uva Lake sequences in Europe.
Thus, the Uva Lake sequences appears to have replaced the series various clade 2.2 sub-clades detected in early 2006, but like those sequences they flew into Europe in the preceding fall, and went undetected until wild birds started to die in large numbers. To date, there the has been virtually no detection of H5N1 in live wild birds in Europe, and additional factors like a harsh winter, or possibly a co-infection in the summer, lead to detectable levels of H5N1 in sick or dying birds.
Media Links
Recombinomics Presentations
Recombinomics Publications
Recombinomics Paper at Nature Precedings