Genetic analyses of the bird flu virus unveil its evolution and potential

A flurry of reports about the genetics of the bird flu currently infecting U.S. cattle are offering insight into how the virus has and continues to spread. Since it first emerged in late 2020, this particular type of bird flu has infected a dizzying array of bird species, about 20 mammal species and some people (SN: 3/6/23; SN: 4/3/24). But transmission from cow to cow and from cow to person and other animals is new.

Now researchers tracing the family tree of the H5N1 avian influenza virus say that the outbreak in cattle, first reported in late March, probably started in late December 2023 (SN: 4/25/24). Cases of low milk production — a symptom of infection in dairy cows — in the Texas panhandle were reported in late January and early February. The U.S. Department of Agriculture confirmed the H5N1 diagnosis on March 25.

Since then, the virus has spread to dairy cows in at least nine U.S. states. A dairy farm worker in Texas got an eye infection, presumably from contact with sick cows. And genetic remains of the virus have been found in grocery store milk, suggesting the outbreak is widespread.

Mia Kim Torchetti, a veterinarian who directs the USDA’s Diagnostic Virology Laboratory at the National Veterinary Services Laboratories in Ames, Iowa, says she had hoped this incursion could be stamped out quickly, but as detections in birds and mammals pile up, “I have rapidly lost hope.”

Though all public health agencies consider the risk of the bird flu spreading widely in people to be low, the outbreak is still reminiscent of the early days of the COVID-19 pandemic.

In that case too, researchers had used genetic analyses to determine that the outbreak had started long before cases were recognized (SN: 1/29/20). And as with COVID, preliminary data are coming out in press releases and preprints without first going through peer review. That doesn’t mean the results aren’t trustworthy, but it does suggest we are in the early days and conclusions may change. The early data also point to myriad versions of influenza viruses preceding the cattle outbreak, just as many waves of SARS-CoV-2 variants caused peak after peak of COVID cases.

We often call the avian influenza virus currently infecting cattle by its nickname, H5N1 bird flu. But its full name is “highly pathogenic avian influenza A H5N1 clade 2.3.4.4b genotype B3.13.”

That specificity denotes the virus’ place in its family tree. Highly pathogenic avian influenza A H5N1 viruses — which are deadly to chickens and related birds — are a huge family tree of bird flu viruses. They all have the H5 form of hemagglutinin, a protein that latches onto host cells so the virus can infect them. The first highly pathogenic avian influenza H5N1 virus was found in 1996. Since then, scientists have documented the tree’s expansion, with some limbs dying off and others making it big. One successful limb of the tree is clade 2.3.4.4b. It has sprouted branches of its own, including genotype B3.13.

Various H5N1s have winged their way around the world after infecting wild birds. A different version crossed the Atlantic in 2014 and caused an outbreak in North American poultry in 2015, but it didn’t take hold, Torchetti says. This time is different.

Clade 2.3.4.4 viruses have been infecting poultry and wild birds for several years. But the limb of the tree we’re dealing with now — H5N1 clade 2.3.4.4b — emerged in Europe in October 2020 when two bird flu viruses swapped parts. It came to the Americas in 2021. It has killed more than 90 million birds in the United States since January 2022, including wild birds and commercial poultry and backyard and hobbyist flocks that were culled when the virus was detected.

How the current bird flu emerged

Influenza viruses are all about the swap meet.

Instead of one long novel, the genetic instruction books of influenza A viruses are more like a series of eight novellas, known as gene segments. Each segment carries one or more of the 11 genes that the virus needs to infect host cells and copy itself. When people, birds or other animals are simultaneously infected with more than one type of influenza virus, the viruses may exchange segments and thus create a new type of virus. This process — called reassortment — has resulted in pandemic strains of flu, including the 1918 influenza pandemic and 2009’s swine flu (SN: 5/22/09).

Viruses can’t swap parts willy-nilly. Not all combinations are compatible with each other. But what’s unusual about this clade of H5N1s is that it undergoes reassortment far more often than earlier relatives, Torchetti says.

In wild birds in the Americas, “this interchange of genes has been occurring for the last almost 24 months” among H5N1 and other bird flus, says Rafael Medina, a virologist at Emory University School of Medicine in Atlanta.

Torchetti and colleagues have found more than 100 genotypes in clade 2.3.4.4b, mostly generated by reassortment. About 20 of those genotypes managed to spread among wild birds, poultry and the occasional other wild animal, the researchers reported May 1 in a preprint posted at bioRxiv.org.

One such reassortment happened shortly before the start of the cattle outbreak, scientists reported May 3 at Virological.org. Genotype B3.13 is a mix of four gene segments from the H5N1 that arrived from Europe in 2021 and four gene segments from a low pathogenicity bird flu from North America. (Low pathogenicity viruses aren’t usually deadly and may not produce any symptoms in infected birds.) It shows up relatively rarely among the viruses sampled in birds, Torchetti says. “The B3.13 genotype is actually not common. The cattle have made it common.” In fact, if predicting which virus might spillover into cattle based on prevalence in wild birds, “this one was a little bit of an underdog,” she says.

All the dairy cattle that have tested positive for H5N1 bird flu have this genotype, suggesting that the virus made the leap from birds to cows just one time. That probably happened in Texas toward the end of last year, Torchetti and colleagues as well as the team posting to Virological.org conclude.

Of the four gene segments the B3.13 genotype picked up, one produces an enzyme that helps copy the virus and the other makes a protein that encases the virus’ RNA. “These specific gene segments have a role in the efficiency of virus replication,” but scientists don’t yet know whether that swap or other changes allowed the virus to more easily infect cattle or grow in mammalian cells, says Tavis Anderson, a research biologist at the USDA Agricultural Research Service’s National Animal Disease Center in Ames. 

With COVID-19 variants, specific genetic changes led to new properties of the virus that made it more contagious or helped it evade the immune system (SN: 12/16/21; SN: 3/1/22). But there’s no obvious indicator of that happening with the H5N1 currently spreading, Anderson says. In other words, B3.13 has been successful at replicating in cattle, but that may be more happenstance than thanks to any special properties of the virus.

What we know about the virus’ spread

No one knows exactly where, when and how the virus passed from wild birds into cattle.

Cows may have grazed on grass that wild birds carrying the virus pooped on, or the cows may have picked it up through contaminated feed or other livestock-bird interactions, Medina says. “It’s present at such a high levels in nature [that] the potential of spilling over into domestic animals is something that shouldn’t surprise us anymore,” he adds.

Once in cows, the virus started spreading from cow to cow. There’s now concern that cows could serve as mixing vessels for new varieties, much the way that pigs have been crucibles for the reassortment of avian, human and swine influenza viruses (SN: 5/14/24; SN: 2/12/10). USDA monitors influenza viruses in domestic swine and wild hogs but hasn’t detected any H5 viruses in those animals, Anderson says.

Genetic signals suggest that cattle carrying the virus spread it from Texas to Kansas, Michigan and New Mexico. There’s also a genetic link suggesting that the cows from Michigan spread the virus to North Carolina, but the USDA researchers have found no record of cows moving between those states. More likely, these researchers say, cows that were moved from Texas to North Carolina spread the virus there.

Since getting into cattle, the virus has jumped into other species including cats in Kansas and Texas that drank infected raw milk. More than half of infected cats from one north Texas dairy died within a few days of having the milk, probably because the virus went to the cats’ brains and nervous systems, researchers reported April 29 in Emerging Infectious Diseases. 

The B3.13 genotype virus has also spilled from cows to raccoons, poultry and wild birds including blackbirds and grackle, researchers said in the Virological.org report. There were as many as five spillbacks from cattle to poultry and three from cows to wild birds, Torchetti and colleagues found. More spillbacks create more possibilities for swapping gene segments and thus more opportunities for a lethal or transmissible virus to emerge, possibly even one that could spread in people. Though agencies agree that risk to people is low, they have warned that human cases from exposure to an infected animal should be expected to pop up from time to time. And though there may be limited spread between people, such as family members, experts don’t expect the virus as is to spread easily from person to person.

The Texas dairy farm worker who got an H5N1 eye infection was carrying a slightly different but closely related version of B3.13 from the one found in cows, researchers from the U.S. Centers for Disease Control and Prevention and colleagues reported May 3 in the New England Journal of Medicine. The person had contact with some dairy cows showing signs of H5N1 infection that were similar to symptoms in cattle at nearby farms with confirmed cases.

A veterinarian who works at multiple dairies in the Texas panhandle encouraged the worker to get tested for H5N1. They went to a Texas Department of Health field site but didn’t reveal where they worked.

The worker is one of 13 people worldwide confirmed to have been infected with a clade 2.3.4.4b virus, and the only one thought to have been infected by a mammal. Some, including the dairy farm worker and a poultry farm worker diagnosed in 2022, have had no symptoms or mild ones. Others have developed severe or critical illnesses. One person in China died in 2022. The dairy farm worker got an antiviral medication called oseltamivir and has fully recovered.

A close look at the genetic makeup of the virus that infected the dairy farm worker revealed that it carried a mutation in a gene known to help the virus replicate better in mammalian cells. But without samples from cattle or other people on the farm, researchers have little information about the evolution of the virus in cattle and whether it can pass from person to person.

It’s possible that the worker’s virus is a slightly earlier version of the one from cows, the CDC researchers say. That suggests that after first jumping from a wild bird into a cow, the virus spread more widely in cattle than previously thought. One twig of the B3.13 branch moved from Texas to other states. That’s the one that has been identified in cows and milk. Meanwhile, close cousins may have continued to quietly infect cows including ones at the worker’s farm.

Let’s hope any callbacks to the early days of 2020 will end here. New regulations that went into effect April 29 governing the movement of dairy cattle and other measures may help contain the virus’ spread in cattle. So far, it hasn’t turned up in the 30 samples of ground beef the USDA has tested. And the U.S. Food and Drug Administration announced May 10 that the latest round of nearly 300 samples of dairy products it tested did not contain the virus.

A wider outbreak in cattle might allow the virus to adapt to spread easily in mammals, including humans. One big thing coronavirus taught us is to never underestimate a virus, especially one that can change quickly.