Pseudoparticles carrying three H3N2 hemagglutinins
The study asks whether the hemagglutinin (HA) of H3N2 subclade K changes how the virus enters cells and how well antibodies block it. HA is the viral surface protein that binds receptors on a target cell and then drives fusion of viral and cellular membranes; antibodies against HA can interfere with that process.
To isolate HA behaviour, the authors built pseudotype particles: vesicular stomatitis virus particles that display chosen influenza HA and neuraminidase (NA) proteins and carry a luciferase reporter. Entry was scored as luciferase activity in cell lysates 16 to 18 hours after inoculation, normalised to particles bearing no viral glycoprotein (set as 1) and log10-transformed. Three HA/NA pairs were compared: an Early-2025 strain, the 2025 egg-based vaccine strain (VAC-2025), and Late-2025, the subclade K strain. Pseudovirus titres measured by quantitative PCR were comparable across the three, so entry differences are not explained by unequal particle input.
Subclade K entry gains appear in specific cell lines
Entry mediated by H3N2 HA was cell line- and strain-dependent, unlike control particles bearing VSV-G or Machupo virus GPC, which entered all tested lines efficiently. The panel covered nine human cell lines plus dog kidney MDCK cells.
Early-2025 particles generally entered more efficiently than either VAC-2025 or Late-2025 particles. The informative comparison for subclade K is against VAC-2025: differences between those two appeared in HEK-293T (kidney-derived), NCI-H1299 (lung-derived) and RPMI-2650 (nasal epithelium-derived) cells. The authors read this pattern as a sign that receptor binding differs between the strains, but the assay measures reporter output after entry, not receptor affinity directly.
A186D links faster entry to weaker neutralisation
Mutation analysis pointed to a single change. Introducing A186D into VAC-2025 HA strongly enhanced cell entry, and the same mutation was the main cause of reduced neutralisation. Two other subclade K changes, N158D and K189R, also contributed to neutralisation escape, while T135K and I160K increased neutralisation sensitivity in some groups.
The location helps explain the effect: residue 186 sits next to the receptor-binding pocket and inside immunodominant antigenic site B, a region the authors describe as a known determinant of receptor choice and antibody evasion. That makes a dual entry-and-escape phenotype plausible rather than surprising, but the paper's evidence is phenotypic linkage, not a structural demonstration of how A186D reshapes the binding pocket.
Reversing the mutation does not behave symmetrically
The authors also tested the reverse change, D186A, in the HA backbones of both Early-2025 and Late-2025 viruses. In both contexts it severely reduced entry into HEK-293T and Caco-2 cells, which prevented the team from testing whether that reversal also restored neutralisation sensitivity.
This asymmetry matters for interpretation. The effect of the residue at position 186 depends on the surrounding HA sequence, so A186D should not be read as a portable switch that produces the same entry gain in every H3N2 background. The paper explicitly calls the impact of D186/N186 context-dependent.
Vaccination raises titres yet the subclade K gap persists
Neutralisation was measured in two cohorts vaccinated in November 2025: 59 healthcare professionals and 56 haemodialysis patients (median age 72, 57% male). Plasma was collected before and at least 14 days after seasonal vaccination with Flucelvax, Influvac or Efluelda 2025/2026. Titres are reported as NT50 values, the plasma dilution giving 50% neutralisation, with entry normalised against samples containing no human plasma.
Before vaccination, both cohorts neutralised all strains, but neutralisation of Late-2025 particles was about 3- to 4.5-fold lower than for Early-2025 or VAC-2025 particles. Vaccination increased neutralising titres in both cohorts, yet activity against Late-2025 remained about 3.5- to 4.5-fold lower than against VAC-2025. The paired analysis covered 17 healthcare professionals and 56 dialysis patients, so the fold-change estimates rest on different sample sizes for the two cohorts.
Pseudovirus results need live-virus confirmation
All entry and neutralisation data come from immortalised cell lines and pseudovirus particles, so the authors state that confirmation with authentic influenza A virus and primary cell cultures is still needed. The study also cannot say whether the entry differences change spread within an infected host or disease severity.
Two further boundaries affect the vaccination result. Post-vaccination plasma was collected 14 to 28 days after immunisation, so durability beyond that window is untested, and the H3N2 infection history of both cohorts is unknown. The authors also note that cohort sizes were too small to examine age or biological sex effects, and that reference sera with defined immune backgrounds would be needed before generalising.
The A186D link is the finding to watch
The paper's most decision-relevant claim is not that subclade K escapes antibodies in general, but that one HA change, A186D, ties together two phenotypes that are usually assessed separately: how efficiently a particle enters a cell line and how much plasma antibody is needed to block it. That coupling is what makes the mutation a useful marker for functional monitoring of emerging H3N2 variants.
The vaccination data add a counterweight. Even with the 3.5- to 4.5-fold gap relative to VAC-2025, both cohorts showed increased neutralising activity after vaccination, which the authors interpret as suggesting current vaccines may provide considerable protection. The authors also note that published live-virus studies report reduced neutralisation of authentic subclade K viruses, while one study of matched human sera found only a minor difference, pointing to cohort-dependent escape. The practical priority is therefore to track A186D alongside real-world vaccine effectiveness rather than to treat the pseudovirus fold-change as a stand-alone protection estimate.
Sources used in this answer
Host cell entry and neutralisation sensitivity of H3N2 influenza A virus subclade K
Subclade K HA mutation A186D enhanced entry into certain cell lines and was the main cause of reduced neutralisation, while vaccination still boosted neutralising titres against subclade K.