Dual H5N1 Bird Flu Infections in North America
Recombinomics Commentary
September 3, 2006
It is possible that these birds were not infected with an H5N1 strain, but instead with two separate avian influenza viruses, one containing H5 and the other containing N1. The confirmatory testing underway at USDA's National Veterinary Services Laboratories will clarify whether one or more strains of the virus are present, the specific subtype, as well as confirm the pathogenicity.
The above comments raise the possibility of dual or multiple infections in the H5N1 positive mallards, which is a distinct possibility. However, the cloned H5N1 from a mute swan in Michigan was H5N1 and sequence data from mallards in Maryland and Pennsylvania indicate that the H5 positive samples in those states are similar to the H5N1 detected in Michigan. Thus, although the isolates from Maryland and Pennsylvania may be mixtures, both samples likely include H5N1.
The finding of the H5N1 serotypes in all three states is unexpected based on results from southern Canada in 2005. The Canadian tests found H5 in 218 birds, and the vast majority, 187, was in mallards. However, most of these isolates were not H5N1. Most of the H5 positives were in British Columbia, where H5N2 and H5N9 serotypes were identified. The second highest total was in Quebec, where H5N3 was found. The initial reports indicated H5N1 was only detected in Manitoba. Only three birds in Ontario were H5 positives, even though three of the collection sites were near the northern shores of Lake Erie.
However this year, the Michigan isolates were at the western end of Lake Erie and the Pennsylvania isolates were in Crawford County, near the southern shores of Lake Erie. This year Canada has not reported H5 in wild birds, but H5 was detected in a dead goose on a farm on Prince Edward Island. Three other geese also died after showing symptoms found in waterfowl infected with the high path Qinghai strain of H5N1 bird flu. The H5 test was a confirmatory PCR test, which not only identifies samples that produce a positive reaction, built also generate an insert that can be sized and sequenced. Although the insert was not sequenced, it was scanned for intensity and its size should distinguish between high and low path since the Qinghai strain of H5N1 has 4 additional amino acids at the HA cleavage site which would generate an insert that was 12 BP larger than low path H5N1.
The insert size could be used to identify mixtures of H5N1. The HA insert that covered the HA cleavage site would be 12 BP larger than low path. Similarly, the H1 from the Qinghai strain has a 20 amino acid deletion, which would generate an insert that would be 60 BP shorter, as seen in lane 3 of the linked elctrophoregram.
Last year the H5 detected in wild birds in Canada was frequently in a mixture of serotypes. Similarly, wild bird isolates collected by Ohio State University frequently had mixtures of two or more serotypes. However, detection of H5N1 in North America is rare. At Genbank there are only three North American isolates of H5N1 and the most recent was in 1986. More isolates are in the process of being sequenced and are listed below. However, none of these isolates are clearly H5N1 Maryland 851 has a mixture of H5 and H6 with N1), including mallards in Maryland in 2002, which were H5N2. The samples being sequenced extend through 2005 collections in Alaska, but none are H5.
The sudden increase in H5N1 in the United States is cause for concern. Since the low path is the same serotype as the high path in migratory birds, mixtures will be more difficult to detect. The H5 from the dead goose on Prince Edward Island was in low abundance which could be obscured by high levels of low path H5N1. Moreover, low path H5N1 can acquire mammalian sequences, as was seen in the H5N2 from British Columbia collected last year. Low path H5 and easily recombine with high path H5 because of extensive regions of identity in the H5 gene, as well as the other 7 gene segments.
Although the 2006 surveillance plan for Canada calls for increasing the numer of tested birds to over 12,000, the program has only released one sequence from the 208 postives birds from 2005. Moreover, there have been no reports from 2006, even though testing from the northern shores of Lake Erie would almost certainly detect H5N1 last month. Thus far the H5N1 positives in the United States are in the Atlantic Americas flyway, which links to the East Atlantic flyway, which coverns western Europe and Africa as well as Siberia, the source of many recent Qinghai H5N1 infections..
Release of the H5 low path sequences may offer clues as to why H5N1 has become the dominant H5 sequence detected In the United States and Canada in 2006.
Isolates by OSU which are being sequenced by TIGR and soon to be public at GenBank
A/mallard/Ohio/184/1986(H5N1)
A/mallard/Maryland/470/2002(H5N2)
A/mallard/Maryland/786/2002(H5N2)
A/mallard/Maryland/851/2002(H5,6N1)
A/mallard/Maryland/792/2002(H5N2)
A/mallard/Maryland/890/2002(H5N2)
A/mallard/Maryland/789/2002(H5N2)
A/mallard/Maryland/795/2002(H5N2)
A/mallard/Ohio/556/1987(H5N9)
A/mallard/Maryland/302/2001(H5N2)
A/mallard/Maryland/252/2001(H5N2)
A/mallard/Maryland/791/2002(H5N2)
Media Sources
Phylogenetic Trees