The potential of H5N1 viruses to adapt to bovine cells varies throughout evolution

Abstract

Avian influenza H5N1 clade 2.3.4.4b viruses caused a global panzootic and, unexpectedly, widespread outbreaks in dairy cattle, therefore representing a pandemic threat. To inform control strategies, it is critical to determine whether the potential to adapt to bovine cells is a general feature of H5N1 viruses, is specific to viruses of clade 2.3.4.4b, or narrowly restricted to some genotypes within this clade. Using a large panel of recombinant viruses representing >60 years of H5N1 history and other IAVs for comparison, we demonstrate replicative fitness in bovine cells is: (i) highly variable across 2.3.4.4b genotypes, (ii) limited in viruses predating the global expansion of this clade, (iii) determined by the internal gene cassette, and (iv) not restricted to udder epithelial cells. Mutations in the PB2 polymerase subunit emerge as key determinants of adaptation, although their phenotypic effects are context dependent. Bovine B3.13 and some avian genotypes exhibit enhanced modulation of bovine interferon-induced antiviral responses, determined by at least PB2, nucleoprotein, and the non-structural protein NS1. Our results highlight the polygenic nature of IAV host range, and reveal that the replication fitness in bovine cells, and likely their potential to adapt to cattle, varies greatly during the evolutionary trajectory of H5N1 viruses.

Fig. 1. Global phylogenies of IAV sequences used in this study.

Global maximum likelihood phylogenetic trees of nucleotide sequences from the internal segments of IAVs used in this study. Tips are annotated by strain names and correspond to viruses evaluated experimentally. Viruses of interest belonging to the European and North American 2.3.4.4b clade are coloured blue and red, respectively.

Fig. 2. Replication of 2.3.4.4b and other IAV reassortant viruses in bovine cells and udder tissue.

a Viral RNA in supernatants of bovine skin fibroblasts (BSF) infected with 0.0005 genome copies/cell, quantified by RT-qPCR. b–d Infectious titres at 24 hpi in BSF (b) and chicken embryonic fibroblasts (DF1) (c) infected with 0.001 PFU/cell, measured by plaque assay on MDCK cells (b, c) or TCID50 on DF1 cells (d). e, f Viral titres from bovine udder explants infected with 1000 PFU/explant, quantified by plaque assay on MDCK cells. g Immunohistochemistry of r-Bovine-B3.13- or mock-infected udder explants showing IAV nucleoprotein-positive cells (brown) in the epithelial layer. In (a–d), data are mean ± s.d. from three biological replicates (n = 3), except in b and c, where r-Bovine-B3.13 (n = 6), r-EA-2020-C (n = 9 in (b), n = 6 in (c)) and r-Avian-B1.1 (n = 6); in (a) data points are mean of two technical duplicates. Data were Log-transformed (a) or normalised to r-Bovine-B3.13 (b–d). Statistics were determined by one-way ANOVA with Tukey’s post-hoc test (two-tailed; α = 0.05). Comparisons are shown relative to r-Bovine-B3.13(H5N1) in all panels, with additional comparisons to r-EA-2020-A(H5N8) or r-CHK/Scot/59 in b–d. In e, f, each data point represents one independently infected tissue explant (n = 18). Data were log-transformed and statistical comparisons were determined using Kruskal-Wallis test with Dunn’s multiple comparisons (two-tailed; α = 0.05). e Shows data medians with interquartile range (error bars). In f, box-and-whisker plots show the median (central line), interquartile range (box) and full data range (whiskers). Individual data points are overlaid. Significance: ns = p > 0.05; * = p ≤ 0.05; ** = p ≤ 0.01; *** = p ≤ 0.001; **** = p ≤ 0.0001. H5N1 viruses are shown in black, non-H5N1 IAVs in grey. Data in g are representative of one experiment, validated by five optimisation runs with varying primary antibody dilutions showing the same result. (a–e) and (g) images created in BioRender. Bakshi, S. (2025) https://BioRender.com/vtz04mr.

Fig. 3. Polymerase activity of diverse IAV strains in human cells.

a Minireplicon assays in HEK293T cells at 24 hpt to assess polymerase activity of a broad panel of Influenza A viruses. Data are mean +/- s.d. of three biological replicates (n = 3), except for Bovine-B3.13 (n = 15), Goose-B3.13 (n = 6) and EA-2020-C (n = 12). Each data point corresponds to the mean of two technical duplicates per biological replicate. All values were normalised to r-Bovine-B3.13, log-transformed, and statistical comparisons between all data were determined using one-way ANOVA with Dunnett’s multiple comparison test (two-tailed; α = 0.05), with comparisons shown relative to r-Bovine-B3.13. Statistical symbols: ns = p > 0.05; * = p ≤ 0.05; ** = p ≤ 0.01; *** = p ≤ 0.001; **** = p ≤ 0.0001. H5N1 virus polymerases are shown in black, non-H5N1 IAVs in grey. b Table of PB2 residues previously implicated in increased polymerase activity. Black boxes indicate residues associated with increased polymerase activity described in earlier studies^,–,,–.

Fig. 4. Segment-specific contribution of B3.13 internal genes to virus replication.

a Schematic of 4:4 reassortant virus containing segments 1,2,3 and 5 from an H5N1 virus in a PR8 backbone. b, c Replication of 4:4 recombinant viruses in BSF (b) or DF1 (c) cells for 24 hpi. d Schematic of 2:6 viruses used in e–h, with internal gene segments reciprocally exchanged between r-Bovine-B3.13 and either r-EA-2020-C or r-Avian-B1.1. e–h Replication of 2:6 reciprocal viruses in BSF (e, g) and DF1 (f, h) cells at 24 hpi. i Amino acid changes between r-Bovine-B3.13 and r-Goose-B3.13. j, Replication of 2:6 reciprocal viruses between r-Bovine-B3.13 and r-Goose-B3.13 in BSF cells. An MOI of 0.001 PFU/cell was used in all infections; virus released into the supernatant at 24 hpi was titrated by plaque assay on MDCK cells. Data are mean +/- s.d. of three biological replicates (n = 3); Values were log-transformed and multiple comparisons between all viruses were determined by one-way ANOVA with Tukey’s test (two-tailed; α = 0.05); only comparisons relative to controls are shown. P-values in bold indicate statistical significance. a–j Created in BioRender. Bakshi, S. (2025) https://BioRender.com/bolea9d.

Fig. 5. Replication of avian 2:6 viruses bearing M631L and other r-Bovine-B3.13 associated PB2 substitutions.

a Alignment of PB2 residues against r-Bovine-B3.13 (bold). b BSF or DF1 cells infected with r-Bovine-B3.13 or r-Goose-B3.13 WT or mutant viruses carrying PB2 amino acid substitutions from the reciprocal virus. An MOI of 0.001 PFU/cell was used and virus in the supernatant at 24 hpi was titrated by plaque assay on MDCK cells. c Minireplicon assays in HEK293T cells at 24 hpt assessing the effects of PB2 amino acid substitutions in b on polymerase activity; data normalised to r-Bovine-B3.13. d, e Replication of WT or mutant r-Bovine-B3.13 and avian genotype representatives bearing PB2-M631L in BSF cells (d), or udder explants (e). f Minireplicon assays (as in c), comparing polymerase activity of r-Bovine-B3.13 and WT or PB2-E627K-bearing r-Avian-EuDG. g Replication of r-Bovine-B3.13 and WT or PB2-E627K-bearing r-Avian-EuDG in BSF and DF1 cells at 24 hpi (as in b). In b–d and f–g, data are mean +/- s.d. of three biological replicates (n = 3), except in b (n = 6 for r-bovine-B3.13, r-Goose-B3.13 and r-Goose-B3.13-M631L); Data were log-transformed and statistical differences between all groups were determined by one-way ANOVA with Tukey’s test (two-tailed; α = 0.05). In e, each data point represents one independently infected tissue explant (n = 12). Data were log-transformed and did not meet the assumption of normality. Box-and-whisker plots show the median (central line), interquartile range (box), and full data range (whiskers). Individual data points are overlaid. Statistical significance between virus groups was determined using the Kruskal–Wallis test with Dunn’s multiple comparisons (two-tailed; α = 0.05). P-values in bold indicate significance. (b, d, e and g) Created in BioRender. Bakshi, S. (2025) https://BioRender.com/732qna2.

Fig. 6. Effect of Mx1 and BTN3A3 on 2.3.4.4b reassortant viruses.

a Identification of virus and Mx1 expressing cells by immunohistochemistry. Udder slices infected with r-Tx/37-B3.13. NP (brown) is visible in the epithelial layer lining the duct, while Mx1 is localised in the cytoplasm of ductal epithelial cells as shown by immunohistochemistry. Images are representative of experiments carried out in 8 different sections from two different animals. Scale bars, 100 μm. b Image showing detection of virus NP and Mx1 by immunofluorescence. NP and Mx1 show miminal colocalization in the duct epithelium. Scale bars, 20 μm. Immunofluorescence was performed on sections from the tissue in (a) as a final validation. cWestern blot of lysates from A549 and BSF cells stably expressing human (MxA), bovine, chicken and rat Mx1. d Single-cycle infection of ZsGreen-expressing reporter viruses on modified A549 cells at 7 hpi. Titres normalised to Chicken Mx1 control cells. e, f A549 (e) or BSF (f) cells overexpressing Mx1 were infected at an MOI of 0.001 and infectious titres at 48 (e) or 24 (f) hpi were determined by plaque assay on MDCK cells. g As described in (e, f) except recombinant 4:4 viruses (Table S2) were used and peak titres of each virus (24 h or 48 h) are shown. h Western blot of lysates from MDCK cells overexpressing human BTN3A3. i Virus titres on modified MDCK cells overexpressing human BTN3A3 or transduced with an empty vector control. j Alignment of NP residues at positions associated with Mx1 or BTN3A3 resistance. Residues are aligned to r-Bovine-B3.13 (bold). Changes associated with resistance are boxed (bold). The data in (d–f) and (g, i) are mean +/- s.d. three biological repeats (n = 3), except for (f) where n = 4. Data in (e–g and i) were log-transformed and statistical significance was determined by two-way ANOVA with Tukey’s (d–f) or Šídák (g, i) multiple comparisons tests (two-tailed; α = 0.05). Only comparisons relative to untreated or chicken Mx1 controls are shown. P-values in bold indicate statistical significance. Gating data for (d) can be found in Fig. S7.

Fig. 7. Sensitivity of 2.3.4.4b reassortant viruses to type I IFN in bovine cells.

Immortalised BSF were pre-treated for 24 h with increasing concentrations of type I IFN prior to low MOI infection (0.001 PFU/cell). a Western blot of interferon-stimulated gene expression (pSTAT1, Mx1, RSAD2) at 24 h post IFN-treatment. b Virus titres in supernatants were determined by plaque assay on MDCK cells at indicated times post-infection. c, d As in (a) and (b) but including r-Goose-B3.13. e, f Replication of r-Bovine-B3.13, r-Avian-B1.1 and reciprocal segment swaps in the absence or presence of universal IFN. BSF were pre- or mock-treated with 2 U/mL IFN for 24 h prior to infection at MOI of 0.001. Western blot for Mx1 at 24 hours post-treatment (e), viral titres at 24 hpi were determined by plaque assay on MDCK cells (f). g, Immortalised udder skin fibroblasts expressing an ISRE-firefly luciferase reporter infected with 2.3.4.4b 2:6 viruses at an MOI of 5 PFU/cell. Luminescence was measured at 24 hpi and normalised to mock controls. h, Western blot of viral proteins and ISGs in udder fibroblasts infected at an MOI of 5 for 24 hours. Data are mean +/- s.d. from three biological replicates (n = 3); each point is an average of technical duplicates. In g, n = 6, error bars show s.d. For b, d and f, data were log-transformed; statistical significances in b, d and f between IFN treated and untreated groups was assessed by two-way ANOVA with Dunnett’s (b and d) or Šídák (f) multiple comparisons tests (two-tailed; α = 0.05). For g, comparisons were determined by one-way ANOVA with Tukey’s test (two-tailed; α = 0.05). For b, d and f, P values in bold indicate statistical significance. For g, only significant values are shown.

Fig. 8. Host-shutoff activity of 2.3.4.4b reassortant viruses in bovine cells.

Immortalised BSF were infected at an MOI of 5 PFU/cell for 16 h or 24 h and pulsed with puromycin for 1 hour. Cell lysates were assessed by western blot for total cellular proteins, phosphorylated STAT1 and the viral proteins PB2, NP and NS1 following puromycin-labelling. β-actin was assessed as a loading control. Western blot data are representative of two independent experiments.

Fig. 9. Phenotypes of the reassortant viruses observed in this study.

Schematic diagram summarising the phenotypes of the reassortant viruses described in this study using either bovine or human cells and restriction factors as indicated. Note that summary of “IFN/host translational modulation” derives from results shown in Figs. 7 and 8. Created in BioRender. Bakshi, S. (2025) https://BioRender.com/90353mv.