Editor's note: I've been offline for much of the past 36 hours due to an internet problem, which was restored late yesterday afternoon.
#18,747
It obviously isn't easy for a novel avian influenza virus to jump - and fully adapt - to a mammalian species, otherwise we'd be hip-deep in pandemic viruses all of the time. Instead we usually see sporadic and tentative spillovers - often to what turn out to be dead-end hosts - and only rarely do we see ongoing transmission of the virus.
But we need look no further than avian H3N2's successful spillover into dogs in South Korea in 2007 - the marine mammal mass mortality events of the past two years from H5N1 - or the multi-state spread of H5N1 to dairy cows - to see examples of avian viruses getting a solid foothold in a mammalian species.
Recent studies have turned up H5N1 in wild rats in Egypt - and H5N6 in Shews in China - raising uneasy questions about what other inroads these viruses might be making around the globe outside of our view.
The assumption has long been that it would likely take long-chains of infections in a mammalian species for the virus to successfully adapt (see serial passage graphic below). One-off, or dead-end, infections were viewed as posing a lower risk.
Adaptive mutations following long chains of Infection
In these experiments, subsequent infections are done artificially, in order to overcome early transmission barriers. In the wild, most of the time these chains are unable to sustain transmission.PB2-627K, which is associated with enhanced replication and pathogenicity in mammals, is one of the most important mutations that H5Nx is thought to need in order to spread more efficiently in mammals (see A rapid review of the avian influenza PB2 E627K mutation in human infection studies).Unless, of course, a permissive mutation - like PB2-E627K - were to emerge in the index infection.
There are others, of course (PB2 D701N, PB2 Q591K, HA Q226L, etc.) - each providing the virus with unique advantages - but if you wanted to kickstart transmission, PB2-E627K would be at or near the top of your list.
All of which brings us to a study - published yesterday in the Virology Journal - which finds that H5N1 lab infected rats quickly developed the PB2-627K mutation and by day 6, the mutation was nearly fixed (60%) in the index animal (see graphic below).
Furthermore, all challenged mice succumbed to the (now nearly 100% fixed) infection, with significant viral loads in both their lungs and brains. The authors wrote:
Notably, the PB2-E627K variant, initially present at 4% in the virus stock, was selected and reached near-fixation (~ 100%) in the lungs and brains by 6 days post-challenge and was subsequently transmitted. No other mammalian-adaptive mutations were identified, emphasizing the pivotal role of PB2-E627K in early stages of mammalian adaptation.
This is a fascinating (and detailed) open access study, and I've only posted the link, and a few excerpts. You'll want to follow the link to read it in its entirety.
Brief Report
Published: 05 June 2025
Volume 22, article number 183, (2025)
Deok-Hwan Kim, Dong-Yeop Lee, Yeram Seo, Chang-Seon Song & Dong-Hun Lee
Abstract
The highly pathogenic avian influenza virus (HPAIV) H5N1 clade 2.3.4.4b has rapidly disseminated globally, with mammalian infections reported in multiple species. Recent evidence of mammal-to-mammal transmission has heightened concerns about the virus’s potential adaptation to mammals. The polymerase basic 2 (PB2) protein E627K mutation appears to be of key importance for mammalian adaptation.
We isolated an HPAI H5N1 clade 2.3.4.4b virus from wild birds in Korea with 96% E and 4% K at amino acid position 627 of PB2. To investigate the genomic characteristics of this clade regarding mammalian adaptation, we studied the replication and transmission of the H5N1 virus in mice. Two experiments with different challenge-to-contact ratios were conducted to assess transmission dynamics and mutation development.
In experiment 1, a 4:1 challenge-to-contact ratio resulted in 100% transmission among direct-contact mice, with all mice succumbing to the infection. In experiment 2, a 1:1 ratio yielded 50% transmission, with all challenged mice also succumbing. High viral loads were observed in the lungs and brains in both experiments, with viral titers increasing over time.
Notably, the PB2-E627K variant, initially present at 4% in the virus stock, was selected and reached near-fixation (~ 100%) in the lungs and brains by 6 days post-challenge and was subsequently transmitted. No other mammalian-adaptive mutations were identified, emphasizing the pivotal role of PB2-E627K in early stages of mammalian adaptation. These findings highlight the need for continuous genomic monitoring to detect mammalian adaptation markers and assess interspecies transmission risks.
(SNIP)
Since the widespread dissemination of the clade 2.3.4.4b H5N1 virus after 2020, mammalian-adapted mutations in the PB2 protein have been reported in both mammals [19] and wild birds [20]. Studies conducted between 2020 and 2023 identified PB2-E627K or PB2-D701N mutations in 10 of 48 documented H5N1 infections in mammals [19].
Furthermore, from 2023 to 2024, seventeen cases of clade 2.3.4.4b H5N1 virus carrying PB2-E627K, K526R, or D701N mutations were detected in wild and domestic birds across eight European countries, indicating that birds may play a role in the dissemination of HPAI viruses with enhanced mammalian infectivity [20]. Among the 5,311 genome sequences of clade 2.3.4.4b viruses identified between 2003 and 2023, only 53 (1.0%) viruses had the PB2-E627K mutation. Notably, 48 of these 53 viruses were identified since 2021.
Of these viruses, twenty-three 627K variants were from avian species, suggesting potential spillover of mammalian-adapted viruses to avian species [21]. Furthermore, a previous study demonstrated that the PB2-E627K mutation did not affect pathogenesis or transmission in ducks, indicating that the 627K variant can persist in avian species [22]. Mutations acquired within infected mammals, rather than direct transmission from wild ducks, could facilitate mammal-to-mammal transmission [23].
Although this is obviously a significant finding, it must be tempered with the knowledge that PB2-E627K has been circulating (albeit at low levels) in HPAI in birds, and spilling over into other species, for years and it has yet to spark a pandemic.
The notion that we are `one mutation away' from a pandemic is more media hype than science. But anything that permits or extends the transmission of HPAI in mammals is a concern, particularly since so much of H5Nx's evolution occurs outside of our view.While a key component, PB2-E627K would likely need to team up with several other advantageous mutations before H5Nx could pose a global health threat.
