Message in a bottle: detecting bird flu

Genome project gathers samples of bird feces, wetland sediment and water to test for viruses and provide early warning

SUMMERLAND, B.C. — Scientists hope that a new system of detecting and identifying avian influenza viruses will lead to a new warning system that better protects poultry farmers.

A pilot project that explored molecular and genomic methods to detect and identify virus types during the avian influenza outbreak of 2014-15 found traditional testing methods fall well short of what genomic technology could offer.

Researchers hope it eventually leads to an avian influenza warning system similar to a homeland security threat level, Chelsea Himsworth told the B.C. Agrifood and Seafood Conference in Kelowna, B.C.

“The idea if you tell people like the poultry industry, ‘you should be on high alert,’ every six months of the year, every year (during bird migratory periods), it leads to complacency,” she said.

“That’s well-accepted, no matter what sector you’re in.”

Himsworth, who is head of veterinary science and diagnostics with the provincial agriculture ministry’s Animal Health Centre, said she hopes the new research will ultimately provide an annual inventory of what viruses are being brought into North America by waterfowl, which will then enable a surveillance system to be set up to provide more effective warnings for producers.

She said British Columbia’s Fraser Valley was the first region hit by the 2014-15 pandemic with 13 farms affected and more than 250,000 birds killed.

“Most of you are probably familiar with the bird flu hysteria because there were ongoing concerns that avian influenza could be the source of the next major global pandemic in people,” she said.

“But I would argue, even now, it is a global pandemic in poultry.”

Specific numbers are not available for Canada, but Himsworth said the 2014-15 outbreak in the United States caused estimated damages of US$3.3 billion, although that figure does not account for the food shortages and price increases that also occurred.

ADVERTISMENT

“This was incredibly frustrating to us when this outbreak occurred because despite ongoing investment in wild bird surveillance, we had failed to detect the incursion of this virus into North America at all in advance of the poultry outbreak,” she said.

Himsworth and her team realized that existing testing systems were ill equipped to detect this type of outbreak.

The traditional system relied mainly on testing individual birds, which involved rushing out to collect dead waterfowl when they were discovered or convincing hunters to turn in birds for testing.

However, those methods had drawbacks in getting representative samples, particularly when considering that the H5 avian influenza strain, which was behind the outbreak, affected only wild birds.

Himsworth said testing only dead birds was akin to monitoring for human diseases by testing only human car crash victims.

“If that sounds like it doesn’t make a lot of sense, it’s because it doesn’t.”

So under a project called Genomic Analysis of Wetland Sediment as a Tool for Avian Influenza Virus Surveillance in Wild Waterfowl, the team explored ways to gather samples from the environment: bird feces, wetland sediment and water.

Sediment was seen as theoretically ideal, but researchers had to overcome the problem of how to find influenza viruses in the massive amounts of organic material found in wetland sediment.

“If you can imagine trying to sort through what is at the bottom of a wetland and detect what is really a needle, or the head of a needle, in a haystack, that’s what sediment is.”

ADVERTISMENT

This is where genomic technology proved its worth. It gave re-searchers the ability to detect and characterize viruses in wetland sediment.

With the 2014-15 outbreak underway, Himsworth quickly managed to put a team together to put the new system to the test.

The results surprised everybody.

Researchers gathered 300 samples from 15 major wetlands near infected farms in the Fraser Valley. They also collected 45 samples from mini wetlands on infected farms.

Samples were put through a targeted resequencing genomic platform, which enabled researchers to detect and characterize specific flu viruses in 102 samples, including the outbreak H5 virus in 16 samples.

The team found that 29.6 percent of samples were positive for any type of influenza, which includes 22.7 percent of wetland samples and 75.6 percent of the farm samples.

Himsworth said the H5 type, which attracts the most attention because it tends to be a dangerous subtype, is not commonly detected in wild birds, but this test showed that in 2014-15 the H5 type was one of the highest occurring types, which is probably why the outbreak occurred.

The virus was not detected using the traditional surveillance technique of testing individual birds, which was in place before the outbreak.

The national detection rates for the H5 type were previously at one percent.

ADVERTISMENT

However, cost is one possible limiting factor of this new testing system. Himsworth said it is far more expensive than traditional methods, but a two-year study is now underway to explore cost efficiencies and further develop and refine the techniques.