PB2 E627K mutation found in dairy cows infected with H5N1: why it matters

You may have read some recent news articles about a “concerning new mutation” noted in some infected dairy cattle herds recently reported.  Sequence data reportedly shows the PB2 E627K mutation is present in four herds, with the data being reported on Tuesday, March 11, 2025.

While the scientific community has been on alert for this mutation, this is the first time* that it has been documented in dairy cattle herds.  

*Note: In May 2024, this mutation was found to be present “as a minor variant” in one H5N1 virus sequenced from one cow in Kansas. This indicates the sampling was done at the time the gene was undergoing within-host mutations. What we have now however, is the report of the PB2 E627K mutation in viruses sequenced from multiple cows in multiple dairy herds.

What is the PB2 E627K mutation? 

Let’s break this down.

H5N1, or bird flu, like all flu viruses, are made up of different genes joined together like cars on a train.  Each gene does different things to help the virus succeed.  The PB2 gene is one of those genes and is associated with viral replication.  

For a bird flu virus to be able to replicate efficiently in mammals, including humans, the genetic components of that bird virus has to become more suited to a mammal or human host.  So, to some extent, the virus’ genes, including the PB2 gene, need to change and adapt to the new environment.  That is where the E627K mutation comes in.

Without getting too much into the ABCs of amino acids and genetics, E627K refers to a switch in amino acids found at position 627 of the PB2 gene.  Generally, bird flu viruses carry a 627E, and mammalian-adapted viruses carry a 627K.  Thus, the switch from E to K is a signal that the virus is expanding its host range and adapting to mammals, at least with regards to replication in host cells.  Of note, the PB2 E627K mutation is also associated with more severe disease, as discussed further below.

We have seen the PB2 E627K mutation before:

So the reason PB2 E627K gets a lot of attention is its association with mammalian adaptation and with severe disease. Indeed, PB2 E627K has been on the bird flu “short list” of concerning mutations for a while.  While it is concerning that it has most recently been found in newly detected dairy cattle infections, we’ve seen this mutation numerous times before.

For example:

• The H5N1 virus from the first human case connected to dairy cattle, the farm worker in Texas (April 2024), carried the PB2E627K mutation.  This is the virus that was shown to be transmissible between ferrets by respiratory droplets in a study published in October 2024.
• The H5N1 virus from the recent human case in Wyoming (February 2025), who was hospitalized and recovered, carried the PB2 E627K mutation.
• The PB2 E627K mutation was partially detected in viral samples from the severe human case in British Columbia, Canada, (November 2024), indicative of rapid mutation upon infection.
• The H5N1 viruses from human cases in Cambodia (October 2023) and Vietnam (November 2023) both carried the PB2 E627K mutation.

To be sure, the PB2 E627K mutation is a fairly common one in mammalian or human cases, because it is a mutation known to be associated with host adaptation and increased replication of H5N1 in mammals.  But there is still some basis to be concerned here.

Several studies conducted with H5N1 viruses of the current clade 2.3.4.4b have shown that PB2 E627K is heavily involved in replication of these viruses in mammalian or human cells and is associated with increased severity of disease.

For example:

• In November 2024, the CDC conducted a ferret study with the human H5N1 virus from the first Texas case, finding that all infected ferrets experienced severe disease and died.  The virus also spread to 3/3 direct contact ferrets and spread by respiratory droplets to 4/6 ferrets.  This H5N1 virus carried the PB2 E627K gene.  The CDC noted that “[c]hanges like those seen in the A/Texas/37/2024 virus may help the virus replicate better in mammals and humans.”
• In an August 2024 study that compared replication between different H5N1 viruses, the authors noted that the Texas human H5N1 virus, which carried the PB2 E627K mutation, replicated more efficiently in human lung cells than an H5N1 virus from a cow in Ohio, which did not have this mutation.  The authors noted that the PB2 mutation could account for this increase in viral replication, noting that such “may explain the increased replication observed for the A/Texas/37/2024 isolate compared to the bovine isolate.”
• A preprint study published in April 2023, found that a highly virulent H5N1 clade 2.3.4.4b virus transmitted between mammals in a lab.  The H5N1 virus was isolated from a hawk that had been naturally infected in Canada.  This H5N1 hawk virus was found to be “efficiently transmitted by direct contact between ferrets, resulting in lethal outcomes.” Even the authors of the study were surprised at this outcome.  In analyzing the genetic components of the hawk H5N1 virus, the PB2 E627k mutation was identified.

Why are bird flu mutations like PB2 E627K important?

As the above cases and studies indicate, PB2 E627K promotes viral replication, and anything that helps a bird flu virus adapt to mammals is something of particular concern to the scientists.

PB2 E627K is one of the key mutations associated with avian flu viruses adapting to mammals.  That doesn’t mean it’s the only one out there.  

Moreover, this mutation isn’t the only concerning mutation out there.  Numerous publications provide long lists of potentially concerning mutations, although PB2 E627K is definitely one of the big ones.  

But as we’ve already seen, there is more than one road avian flu viruses can take in their quest for adaptation to mammals, and humans.  For example:

• The first H5N1 human case from Michigan did not carry the PB2 E627K mutation.  Like the H5N1 viruses isolated from the ongoing dairy cattle outbreaks, this Michigan H5N1 virus carried the PB2 M631L mutation.
• A human H5N1 case in Chile, involving a lengthy hospitalization before recovery, contained a different mutation in the same gene, PB2 D701N, also associated with increased replication in mammals.
• A March 2023 study of H5N1 infections in wild mammals in Finland noted the PB2 E627K mutation in an otter and a red fox,  the PB2 D701N mutation in another red fox, and both mutations in the virus from a lynx.  The dual PB2 E627K + D701N “is rarely observed in nature.”

Thus, H5N1 continues to do things we didn’t expect.  What matters is whether the current clade of H5N1 is actively adapting to mammals and, if it is, what we can do about that.  

The implications of PB2 E627K in dairy cattle infected with H5N1: 

Up until this point, an interesting finding regarding the cattle H5N1 viruses was the absence of PB2 E627K.  Rather, the cattle viruses contained a different PB2 mutation, M631L, that was thought to perform a similar function.  Indeed, a simple summary of the mutation situation was provided in one of the periodic H5N1 updates put out by the  UK Health Security Agency:

The finding of PB2 E627K in dairy cattle is clearly concerning, as it indicates further adaptation of H5N1 to mammals as it continues to spread through dairy cattle.  But it is only one piece of the puzzle that H5N1 must complete before it has gained the ability to efficiently infect humans and transmit between them, which may or may not ever happen. We really don’t know.

Nor do we know if this will continue to be a trend, leading to more sequenced cattle viruses with this mutation. 

What we do know is that every time H5N1 infects a new host, like a dairy cow, the virus will undergo within-host adaptations, like PB2 E627K.  Further, every infection of a non-avian host is an additional opportunity for this bird virus to adapt to mammals, and potentially humans.  So we need to keep an eye on this new development and as well as any further detections of mammalian adaptation markers.

 Until next time.