H5N1 Is Diverse and Widespread in Wild Birds in England
Recombinomics Commentary 20:58
January 31, 2008
Genetic analysis of four of the viruses recovered places these isolates in
a cluster with others recovered in mid to late 2007 from the Czech Republic, Romania and Poland. Nucleotide sequence comparisons identify a Czech Republic isolate with the closest similarities with these isolates, ranging from 98.7% to 99.4%.
The genetic relatedness at the nucleotide level between the four UK mute swan isolates analysed fully to date is 99.3% to 99.8%. Similarities between these isolates from the outbreak of HPAI H5N1 in turkeys in England in November 2007 is 99.2% to 99.3%. Analysis of the remaining two virus isolates continues and will be reported separately.
In summary the viruses from the swans are closely related to each other, and more closely related to viruses from Europe than to those recovered from the recent outbreak in turkeys in Suffolk.
The above comments are from a DEFRA report on H5N1 in wild swans in Dorset, England. Although the report concludes that there is no evidence for widespread infections of wild birds by H5N1, the above data on the sequences from four of the H5N1 positive dead swans clearly shows that H5N1 is widespread in the swan and wild bird populations in England.
If the swans were infected by a common source, which would be other wild birds, then the sequences from the four swans reported above would be virtually identical. Instead the relatedness between the four swan isolates is 99.3% to 99.8%.
There have been many prior examples of sequence data from H5N1 infections by a common source. In the DEFRA analysis of the infection of domestic turkeys early last year, the sequences between the H5N1 in turkeys in England and H5N1 from waterfowl in Hungary was 99.96%. This analysis used all eight gene segments from H5N1 from the two sources and found few differences.
A similar analysis was done for H5N1 isolated from an outbreak in Krasnodar. Full H5N1 sequences were available from a chicken at the farm as well as a whooper swan near the farm. The two sets of sequences were 99.95% identical.
However, full sequences are not required to identify a common source. This can be done with the HA sequence. For the chicken and whooper swan from Krasnodar, the two HA sequences were 100% identical.
Similarly, HA sequences from a chicken, duck, and cat linked to the recent outbreak in Romania were 100% identical. The same was true for the recent HA sequences from an outbreak in Ghana. The HA sequence from the lab in Italy was an exact match with three sequences generated by NAMRU-3 in Egypt, so even when separate samples are sequenced by independent labs, the four HA sequences exactly matched each other.
In contrast, the four sequences described for the four swans in Dorset are related to each other (greater than 99% identity), but not from a common source, which would generate identities greater than 99.9%, as was seen in the examples above.
Instead, the sequences in the swans have significant differences, indicating they were independently infected by a variety of distinct sources. Thus, H5N1 is diverse and widespread in the wild bird population in England.
This is supported further by the timing of the discovery of the positive birds. The number of H5N1 positive swans is now up to seven and these birds were collected over a one month period.
The number of positive birds is low because the assay lacks the sensitivity to detect H5N1 in live healthy birds, which is why all of the healthy birds are negative and why the only positive dead birds are mute swans, which tend to have a higher level of H5N1 than other species.
Thus, the H5N1 in wild birds in England is diverse, and the failure to detect these infections lies in the assays and procedures used, which have a low sensitivity and can only detect H5N1 in the birds with the highest levels.
These surveillance flaws account for the failure to detect H7 in wild birds prior to the H7N3 and H7N2 outbreaks in England in 2006 and 2007, as well as the source of the H5N1 in the free range turkeys in Suffolk.
Thus, the “lack of evidence” cited in the report is due the lack of an assay that can detect H5N1 in live healthy wild birds, which clearly have H5N1, which has produced H5N1 positives in dead wild birds for the past month.
Media Links
Recombinomics Presentations
Recombinomics Publications
Recombinomics Paper at Nature Precedings