🦠 microbiology

H5N1 2.3.4.4b HA E190D and Q226H mutations, picked up as minority variants in a patient, result in an inability to bind sialic acid.

This study demonstrates that the E190D and Q226H mutations in the hemagglutinin receptor-binding site of clade 2.3.4.4b H5N1, identified as minority variants in a Canadian patient, completely abolish the virus's ability to bind sialic acid receptors across multiple H5 backgrounds, indicating that such single mutations disrupt rather than facilitate adaptation to human-type receptors.

Original authors: Kovacs, E., Rios Carrasco, M., Guerreiro Cabana, M. F., de Vries, R. P.

Published 2026-03-08

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Kovacs, E., Rios Carrasco, M., Guerreiro Cabana, M. F., de Vries, R. P.

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ⚕️ This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

Imagine the influenza virus as a burglar trying to break into a house. The "house" is a human cell, and the "lock" on the door is a specific molecule on the cell's surface called a receptor (specifically, a type of sugar called sialic acid).

The virus has a special tool to pick this lock: a protein on its surface called Hemagglutinin (HA). Think of HA as the burglar's master key.

The Story So Far

Most bird flu viruses (like the H5N1 strain discussed here) are excellent at picking the locks found on bird cells. These locks are shaped differently than the ones on human cells. Usually, for a bird flu virus to infect a human, it needs to change its master key (the HA protein) so it can fit the human lock instead.

Recently, a 13-year-old girl in Canada got very sick with a bird flu virus. When scientists looked closely at the virus inside her, they found something interesting: the virus had started to mutate. It had two tiny changes in its master key (at positions 190 and 226).

Scientists were worried. They thought, "Maybe these changes are the first step toward the virus learning how to pick human locks!"

The Experiment: Testing the New Keys

The researchers in this paper decided to play the role of the locksmith. They wanted to see if these new "mutated" keys actually worked better on human locks, or if they were just broken.

They created three versions of the virus's master key in the lab:

Key A: Had the first mutation (E190D).
Key B: Had the second mutation (Q226H).
Key C: Had both mutations combined.

They then tried to use these keys on a test set of locks. These locks were:

Bird locks (the original target).
Human locks (the new potential target).

The Big Surprise: The Keys Broke

The result was shocking. Instead of the virus getting better at infecting humans, the mutations completely broke the key.

The single mutations made the key useless.
The double mutation made the key completely flat and unable to turn in any lock.

It was as if the burglar tried to modify his key to fit a new door, but in doing so, he accidentally filed the teeth off the key entirely. Now, the key couldn't open the bird door or the human door.

The researchers tested this on different "versions" of the virus (from different years and locations), and the result was the same: The mutations destroyed the virus's ability to grab onto cells at all.

Why Does This Matter?

This is actually good news, but it teaches us a valuable lesson about how viruses evolve.

It's not a straight line: We used to think that if a virus changed just a few letters in its code, it would instantly become a human virus. This paper shows that's not true. In fact, changing just one or two spots often breaks the virus's ability to infect anyone.
Adaptation is hard: For a bird flu virus to successfully infect humans, it likely needs a very specific, complex series of changes. It can't just happen by accident with a single mutation. The virus has to walk a very narrow tightrope; if it steps off the path too early, it falls off the cliff (loses its ability to bind).
Surveillance is working: The fact that we found these "broken" mutations in a sick patient is a sign that our surveillance systems are working. We caught a virus that was trying to change but failed. It didn't become a super-spread human virus; it just became a virus that couldn't hold on.

The Bottom Line

Think of the virus's evolution like trying to learn a new language. You might think that changing one word in a sentence will make it sound like a different language. But in reality, changing just one or two words often makes the sentence gibberish that no one understands.

This paper confirms that the specific mutations found in the Canadian patient didn't help the virus speak "Human." Instead, they made the virus mute. It lost its ability to grab onto cells entirely. This suggests that for H5N1 to become a true human pandemic threat, it will need a much more complicated and difficult set of changes than we previously feared.

Problem Statement

In November 2024, a critical respiratory infection in a 13-year-old girl in Canada was caused by a clade 2.3.4.4b H5N1 influenza A virus. Sequencing of the patient's sample revealed the presence of minority variants (subpopulations within the viral quasispecies) carrying two specific mutations in the Hemagglutinin (HA) receptor-binding site (RBS): E190D and Q226H.

While mutations at positions 190 and 226 are historically associated with the switch from avian-type ( $\alpha2,3$ -linked sialic acid) to human-type ( $\alpha2,6$ -linked sialic acid) receptor specificity in H1, H2, and H3 subtypes, their impact on H5 subtypes is complex and not fully understood. There was a critical need to determine if these specific patient-derived mutations represented a step toward human adaptation (increased binding to human receptors) or if they had a different phenotypic effect.

Methodology

The researchers employed a recombinant protein approach combined with a quantitative ELISA-based glycan-binding assay to assess receptor specificity.

Protein Engineering:
- They constructed recombinant HA proteins from the A/Texas/37/24 (H5TX) strain (a common human-derived H5 background).
- They generated single mutants (E190D, Q226H) and a double mutant (E190D/Q226H).
- To test for background dependency, the Q226H mutation was also introduced into two other distinct H5 backgrounds:
  - H5FR: A/duck/France/161108h/16 (representing early European 2.3.4.4b strains).
  - H5AK: A/black swan/Akita/1/16 (representing clade 2.3.4.4e).
- Proteins were expressed in HEK293S GnTI(−) cells, engineered with an mOrange2 tag, a GCN4 trimerization domain, and a Twin-Strep tag for purification.
Binding Assay (Glycan ELISA):
- Streptavidin-coated plates were used to immobilize biotinylated glycans.
- Test Ligands:
  - Avian-type: Linear L-3SLN3 ( $\alpha2,3$ -linked sialic acid).
  - Human-type: Linear L-6SLN3 and a symmetrical bi-antennary N-glycan (N-6SLN3) with $\alpha2,6$ -linked sialic acids.
- Recombinant HA proteins were incubated with the glycans, and binding was quantified via optical density (OD485) after developing with an OPD substrate.
- Controls: A human H1N1 strain (A/Kentucky/UR06-0258/07) was used as a positive control for human-type receptor binding.

Key Contributions

Independent Validation: The study provides independent experimental validation (using ELISA) of recent findings from a glycan microarray study (preprint reference 10) regarding the H5BC E190D/Q226H mutations.
Background Independence: By testing the Q226H mutation across three different H5 genetic backgrounds (2.3.4.4b Texas, 2.3.4.4b France, and 2.3.4.4e Akita), the authors demonstrated that the loss of binding is a robust phenotype not limited to a single viral strain.
Clarification of H5 Adaptation: The work challenges the simplified model that single mutations at 190/226 automatically confer human receptor specificity in H5 viruses, highlighting the complexity of H5 RBS adaptation.

Results

Loss of Binding in H5TX:
- The wild-type (WT) H5TX bound strongly to the avian-type ligand (L-3SLN3).
- All mutants (E190D, Q226H, and E190D/Q226H) showed a complete loss of detectable binding to both avian-type ( $\alpha2,3$ ) and human-type ( $\alpha2,6$ ) sialosides.
- The human H1N1 control successfully bound the human-type ligands, confirming the assay's functionality.
Loss of Binding in Other Backgrounds:
- In the H5FR background, the WT bound L-3SLN3, but the Q226H mutant abolished all binding.
- In the H5AK background (which previously showed some potential for human receptor binding with a Q226L mutation), the Q226H mutant similarly resulted in no detectable binding to any tested glycans.
Phenotypic Outcome:
- Contrary to the hypothesis that these mutations might shift specificity toward human receptors, the data indicates that these specific substitutions severely impair or completely abolish the receptor-binding capacity of the H5 HA protein.

Significance

Mechanism of Adaptation: The findings suggest that the mutational pathway for H5N1 adaptation to human receptors is not a simple "switch" via single amino acid changes at positions 190 or 226. Instead, single mutations in these key RBS residues often disrupt the structural integrity required for any receptor binding.
Viral Evolution: The presence of E190D and Q226H as minority variants in a patient may represent a "dead-end" evolutionary step or a precursor state that requires additional compensatory mutations to regain binding affinity while potentially shifting specificity.
Surveillance Implications: The study underscores the importance of analyzing minority variants not just for their potential to gain human specificity, but also for their potential to lose binding capacity entirely. It highlights that the H5 RBS network is highly sensitive, and "hallmark" mutations from other subtypes do not translate directly to H5.
Scientific Rigor: By confirming these results across multiple backgrounds and using a complementary assay (ELISA vs. microarray), the paper strengthens the consensus that E190D and Q226H are detrimental to H5 receptor binding in their current context.