Preprint: Amino Acid Residues Responsible for Differential Replication and Pathogenicity of HPAI H5N1 Isolated from Human and Cattle in Texas, US

Flu Virus binding to Receptor Cells – Credit CDC

#18,352

Eleven months ago, in CDC Summary Analyses of Genetic Sequences From HPAI Human Case In Texas, we learned that the first human infection with the `bovine' B3.13 H5N1 virus carried a well known mammalian adaptation; PB2-E627K.

Avian influenza in birds is predominantly a gastrointestinal malady, and is spread mostly via infected droppings. Birds run `hotter’ than mammals, which means avian flu viruses must adapt to lower temperatures found in the respiratory tract if they are to succeed in human or mammalian hosts.

PB2-E627K describes the swapping out of Glutamic acid (E) for Lysine (K) at position 627 in the PB2 protein, which - among other things - allows the virus to replicate at lower temperatures found in the human respiratory tract.

While concerning, additional mammalian adaptations are thought needed in order to make an avian virus a genuine pandemic threat (some we know about, while others we may not), and reassuringly this mutation wasn't found in any of the other cattle or human isolates over the remainder of 2024.

Last week, however, the CDC announced the detection of this same E627K mutation in a recent human case exposed to the D1.1 genotype carried by wild and domesticated birds:

CDC's analysis of the genetic sequence of the virus from the patient in Wyoming identified an avian influenza A(H5N1) virus from clade 2.3.4.4.b (genotype D1.1). The virus had a genetic mutation in its PB2 protein that has previously been associated with more efficient virus replication in people and other mammals (i.e., change of PB2 E627K). This change was previously identified in a human case in Texas during 2024.

While the following preprint was submitted prior to this announcement, it warns on the importance of active surveillance to eliminate the H5N1 virus from cattle, and to prevent further human infections.  Most importantly, it confirms the importance of the E627K mutation in increasing replication and pathogenicity in mice. 

The full preprint runs 44 pages, and parts tend to be highly technical, so I've only posted the abstract and snippet from the conclusion. Follow the link to read it in its entirety. 

I'll return with a postscript after the break. 

Identification of Amino Acid Residues Responsible for Differential Replication and Pathogenicity of Avian Influenza Virus H5N1 Isolated from Human and Cattle in Texas, US

Mahmoud Bayoumi, Ramya S Barre, Ruby A Escobedo, Vinay Shivanna, Nathaniel Jackson, Chengjin Ye, Adolfo Garcia-Sastre, Ahmed Mostafa, Luis Martinez-Sobrido

doi: https://doi.org/10.1101/2025.03.01.640810
Preview PDF

Abstract

Highly pathogenic avian influenza viruses (HPAIV) pose a serious public health concern. In March 2024, a first-time outbreak of HPAIV H5N1 in dairy cattle herds was reported in the United States (US). Since then, the virus has continued to spread in cattle herds and spilt over into humans. 

We recently showed that the first human isolate reported in the US in Texas (HPhTX) from a dairy worker in an affected cattle farm has enhanced replication kinetics and pathogenicity in mice compared to a closely related bovine isolate (HPbTX). 

However, the molecular determinants of differential pathogenicity have not yet been identified. 

Herein, we show that HPhTX has enhanced polymerase activity, compared with HPbTX, in human cells and that the polymerase basic 2 (PB2) protein is the main factor responsible for this difference. Through single and combined site-directed mutagenesis and swapping the three amino acids different between HPhTX and HPbTX, we found that PB2 mutation E627K is the major contributor to the enhanced polymerase activity of HPhTX. 

E362G substitution in HPhTX PB2 affected the polymerase, although to a lesser extent than E627K. Moreover, M631L mutation in HPhTX PB2 enhanced polymerase activity.

Rescue of a loss-of-function recombinant HPhTX (rHPhTX) containing mutations at residues 627 and 362, alone or in combination, revealed a contribution of PB2 E362G and K627E in morbidity, mortality, and viral replication as compared to rHPhTX wild-type (WT), and significantly reduced viral pathogenicity to levels comparable to rHPbTX WT. 

These findings indicate that HPAIV H5N1 of cattle origin isolated from the first human case has post-transmission amino acid changes that increase viral replication in human cells and pathogenicity in mice.

(SNIP)

Analysis of HPAIV H5N1 sequences identified in humans during the recent outbreak in the US reveals that the mutation 627K observed in the PB2 gene is distinctive to the HPhTX isolate, not present in other bovine or human B3.13 or D1.1 strains (Fig. 10C). It is also important to highlight that the recently reported human isolates of D1.1 strains, which exhibit enhanced pathogenicity, possess the 362E mutation (and lack 627K) in the PB2 gene. Nevertheless, the molecular factors contributing to their pathogenicity require further investigations 39 .

This data suggests the possibility that these mutations rapidly emerged after human infection with the bovine HPAIV H5N1 for efficient viral replication. Notably, it has been recently found that the HPhTX can also efficiently infect and transmit in ferrets 27,34 , pointing out that a few amino acid mutations would allow efficient viral replication and transmission among mammals. Furthermore, these results also demonstrate the importance of monitoring HPAIV H5N1 infecting humans for the presence of these or similar mutations that allow them to better replicate and possibly transmit.

Finally, these results suggest the importance of active measures in place to eliminate HPAIV H5N1 from cattle to prevent further human infections and potential transmissions of HPAIV H5N1.

         (Continue . . . )
 

Last week's update from the CDC also identified another H5N1 mutation of concern - D701N - in a human infection in Nevada. This time, a D1.1 genotype which has only recently been identified in cattle.  They write:

CDC's analysis of the genetic sequence of the virus isolated from the patient in Nevada identified the virus as an avian influenza A(H5N1) virus from clade 2.3.4.4.b (genotype D1.1). The nucleotide sequence was nearly identical to that of the viruses that USDA reported from dairy cows in Nevada that the person worked with. The virus had a genetic mutation in its polymerase basic 2 (PB2) protein that has previously been associated with more efficient virus replication in mammalian cells (i.e., change of PB2 D701N). This change was previously identified in a human case in Chile in 2023.

This PB2-D701N mutation - the swapping out of aspartic acid (D) with asparagine (N) at position 701 of the PB2 protein - has been linked to increased severity of infection (see here and here), and has been previously seen in both human H5N1 and H5N6 infections.

While human infections with the B3.13 `bovine' genotype have been reassuringly mild (including the Texas case with E627K), the recently emerged D1.1 genotype has produced several severe (and 1 fatal) human infections. 

A reminder that H5Nx continues along an unpredictable evolutionary path, and that anything we might have said about its threat yesterday could easily change tomorrow.