Enhanced immunogenicity and dose-sparing efficacy of self-amplifying RNA vaccines against seasonal influenza across subtypes

First of all, to investigate whether sequence optimization can enhance HA expression and overcome the suboptimal efficacy of IBV vaccines, we used the HA sequence from A/Puerto Rico/8/1934 (H1N1) (PR8) as a template to assess the impact of different sequence optimization strategies on mRNA-based protein expression. A total of 6 sequence design tools were used, focusing on codon optimization and RNA stability, with a systematic analysis of key parameters like the codon adaptation index (CAI) and minimal free energy (MFE). The results showed that the wild-type PR8-HA sequence (PR8-HA-WT) had a CAI of 0.7 and an MFE of −419 kcal/mol. After optimizing for maximum CAI, the CAI increased to 1.0, while the free energy dropped to −657.2 kcal/mol. In comparison, an optimization strategy focused on MFE and RNA secondary structure stability led to a CAI of 0.75 and further decreased the free energy to −1103.8 kcal/mol. Additional sequences with varying CAI and MFE parameters were also tested for expression analysis (Table S1). After transfecting BHK21 cells with mRNAs designed using various optimization methods, Western blot (WB) analysis revealed that both the CAI- and MFE-optimized PR8-HA sequences increased HA protein expression efficiency by approximately 40% (Figure S1A).

We then applied the sequence optimization strategies to design mRNAs for seasonal influenza viruses. mRNA sequences encoding HA genes from the WHO-recommended 2023–2024 influenza strains-A/Wisconsin/67/2022 (H1N1, W67), A/Darwin/6/2021 (H3N2, D6), and B/Austria/1359417/2021 (Victoria, AUT21)-were optimized using the maximal CAI and MFE strategies (Figure S1B). Agarose gel electrophoresis confirmed the uniformity of all three subtype-specific mRNA transcripts (Figure S1C, Table S2). Capillary electrophoresis further verified the integrity and expected size distribution of the optimized mRNA transcripts, showing a predominant full-length product with minimal detectable degradation or aberrant species across all three HA constructs (Figure S1D). Expression validation in BHK21 cells revealed that MFE-optimized mRNAs exhibited the highest protein expression for W67 and D6, whereas AUT21 showed comparable expression levels across optimization strategies (Figures S1E and S1F). Based on these results, we selected MFE-optimized constructs for subsequent vaccine production. Quantitative Western blot analysis, which calibrated with recombinant HA proteins (Standard curve, W67-HA: Y = 0.007625X–0.07734, R² = 0.9359; D6-HA: Y = 0.02268X + 0.09294, R² = 0.9795; AUT21-HA: Y = 0.006237X-0.01886, R² = 0.9951), confirmed efficient expression of all three HA subtypes by both mono- and trivalent (TRV) mRNA vaccines in BHK21 cells. Absolute quantification revealed higher expression levels of H3-D6-HA (0.47–1.69 ng/μL) and IBV-AUT21HA (0.26–0.60 ng/μL) compared to H1-W67-HA (0.21–0.41 ng/μL), a consistent observation across all transfection doses (250–1000 ng) (Figures S1H-J). These results demonstrate the effectiveness of sequence optimization and indicate sufficient IBV HA expression for vaccine development.

Next, the optimized mRNAs were encapsulated into SM102 lipid nanoparticles (LNPs), with subsequent physicochemical characterization demonstrating excellent formulation properties: all LNPs exhibited a particle size <100 nm, polydispersity index (PDI) < 0.1, and mRNA encapsulation efficiency >95% (Figure S1G), fully meeting the quality standards for vaccine delivery systems.

WHO global influenza surveillance data (2022–2024) showed that H1N1 and BV strains maintained stable antigenic characteristics, while H3N2 exhibited significant antigenic drift, leading to the update of the H3N2 component in the 2024–2025 vaccine recommendations to A/Massachusetts/18/2022 (H3N2) (MC22). In a parallel evaluation, immunization with 10 μg mRNA-MC22 induced comparable HA-specific IgG levels to those achieved by mRNA-D6 (Figures S3A-D), demonstrating consistent immunogenicity across different H3N2 variants in our mRNA platform. Meanwhile, we also assessed the neutralizing antibodies in sera from mRNA-MC22 immunized mice against the HK14, HK19, and D6 strains (Figures S3E-G). We found that MC22 was also able to induce effective neutralizing antibodies against the D6 strain. Moreover, the mRNA-MC22 vaccine at 10 μg dose conferred complete protection against HK68 challenge (Figures S3H and S3I).

Neutralization assays demonstrated that the 1 μg and 10 μg dose groups of mRNA-TRV produced significantly higher neutralizing titres against matched strains (W67 and D6) for influenza A than the QIV group. Additionally, only the 10 μg dose group showed enhanced neutralizing titres against the IBV strain (AUT21) (Figure 3(E–G)). Crossneutralization evaluation against mismatched strains revealed a broad-spectrum neutralizing activity of the 10 μg mRNA-TRV dose against MCG15, HK19, and WST19, with particularly superior neutralization potency against the mismatched B strain compared to the QIV group (Figure 3(H–J)). However, no detectable neutralizing antibodies against the B/Yamagata strain were induced by mRNA-TRV (Figure S4A).

To further assess the safety profile of mRNA-TRV, an independent safety evaluation was performed using a simplified immunization schedule (Figure S5A). No obvious abnormal body weight changes were observed during the post-immunization monitoring period, indicating that mRNA-TRV was well tolerated across all tested doses (Figure 3(K)). Consistently, serum biochemical analysis showed no evident treatment-related changes in aspartate aminotransferase (AST) or creatine kinase (CK) levels after immunization (Figure 3(L,M)). Moreover, additional serum biochemistry parameters, including alanine aminotransferase (ALT), alkaline phosphatase (ALP), total protein (TP), albumin (ALB), creatinine (Crea), and lactate dehydrogenase (LDH), remained comparable between placebo and mRNA-TRV immunized groups at all tested doses (Figures S5B-G), further supporting the favourable safety profile of the trivalent mRNA vaccine.

To address the challenge of high dosage requirements in conventional mRNA vaccines, particularly for IBV strains, we successfully synthesized seasonal influenza saRNA vaccines (Figure S6A) and circRNA vaccines (Figure S6E). These platforms are designed to overcome dosage limitations by leveraging their intrinsic capacity for low-dose efficacy and prolonged antigen expression. Subsequent Western blot analysis demonstrated relatively stable protein expression in cells over time from saRNA vaccines. Sequence optimization strategies were systematically evaluated for influenza HA antigens expressed via saRNA. Expression analysis identified CAI as the optimal method for W67 and AUT21 strains, while MFE optimization proved to be most effective for the D6 variant (Figures S6B and S6C). CircRNAs were co-circularized during in vitro transcription and purified via high-performance size-exclusion chromatography. E-gel EX electrophoresis systems (Thermo Fisher), which are able to distinguish circRNAs from other RNA species, showed relatively low circRNA formation rates and substantial premature termination products with the MFE optimization method (Figure S6F), suggesting its incompatibility with the circRNA platform. Western blot analysis subsequently revealed superior expression from CAI-optimized circRNA-W67 and circRNA-D6, while circRNA-AUT21 performed better with wild-type (WT) sequence (Figure S6G). Encapsulation of saRNA or circRNA using SM102 lipid yielded vaccines with favourable physicochemical properties: all formulations exhibited particle sizes <100 nm, PDI values <0.1, and encapsulation efficiencies >95%, confirming excellent nanoparticle uniformity (Figures S6D and S6H).

Challenge studies showed 100% protection against homologous IBV strain (GZ12) in the 0.1 μg saRNA-TRV group with stable body weight, whereas all circRNA-TRV immunized mice succumbed by day 10 and mRNA-TRV achieved only 14% survival (Figure 5(E)). At the 0.6 μg dose, both saRNA-TRV and circRNA-TRV achieved complete protection, while mRNA-TRV provided partial protection with 71% survival (Figure 5(F)). At 1 μg, both saRNA-TRV and circRNA-TRV provided full protection, while mRNA-TRV showed 86% survival (Figure 5(G)).

To further evaluate the functional relevance of binding antibody responses, we assessed neutralizing antibody titres against matched viral strains. Consistent with the binding antibody data, saRNA-TRV at a 0.1 μg dose produced neutralization titres similar to those of mRNA-TRV against the matched strains (W67), with no significant differences observed at the 1 μg dose (Figure 7(E)). For the H3 matched strains (D6), neutralization titres across platforms showed only marginal differences (Figure 7(F)). Notably, saRNA-TRV maintained its immunogenicity advantage against IBV (AUT21), with neutralization titres surpassing those of mRNA at both dose levels (Figure 7(G)).

We next implemented a booster immunization strategy, and splenocytes were collected on day 12 post-boost for the same stimulation and downstream analyses (Figure 7(H)). Under H1-W67 stimulation, all platforms robustly produced IFN-γ, TNF-α, and IL-2, with no significant differences among groups (Figures S9A-C). Under H3-D6 stimulation, the induction of IFN-γ by mRNA-TRV was superior to that of saRNA-TRV and circRNA-TRV (Figures S9D-F). Notably, under BV-AUT21 stimulation, saRNA-TRV induced significantly higher ELISpot responses than mRNA-TRV across all three cytokines (Figure 7(I–K)).

These results collectively demonstrate that saRNA-TRV, particularly at lower doses, induces superior neutralizing antibody responses and T cell immunity, further supporting its potential as a versatile platform for IBV vaccine development.

The binding antibody responses to the D6 antigen largely mirrored those observed for W67. At both 0.1 μg and 1 μg, single-dose immunization with mRNA, saRNA, and circRNA led to increased antibody levels by week 9 followed by a gradual decline by week 20, whereas under prime-boost conditions, antibody levels declined continuously from week 3 onward (Figure 8(D–F)). Additionally, at 0.1 μg, single-dose saRNA induced significantly higher antibody levels at weeks 9 and 20 than circRNA. At 1 μg, there were no statistical differences among the three RNA vaccines under either singledose or prime–boost regimens; however, in the prime–boost condition, the saRNA-TRV group exhibited significantly higher antibody levels than its single-dose counterpart (Figures S10D-F).

For the influenza BV-AUT21-HA antigen, the pattern differed markedly from W67 and D6. At 0.1 μg, neither mRNA-TRV nor circRNA-TRV induced detectable antibody responses, whereas saRNA-TRV elicited significantly higher antibody levels than both mRNA-TRV and circRNA-TRV under single-dose and prime–boost conditions (Figure 8(G–I)). At 1 μg, there were no significant differences among the three RNA vaccines in the single-dose groups; however, in the prime–boost regimen, saRNA-TRV induced significantly higher antibody levels than mRNA-TRV at week 9, and by week 20, both saRNA-TRV and circRNA-TRV were significantly higher than mRNA-TRV (Figures S10G-I).

In summary, at a low dose (0.1 μg), the saRNA platform consistently induces the most durable and highest antibody responses across most antigens—particularly the conserved influenza HA antigen—highlighting the unique advantage of its self-amplifying mechanism in sustaining long-term immunity.

E. coli DH5α cells (AlpalifeBio, Guangzhou, China) were used for plasmid cloning and cultured in LB broth (Sango) supplemented with or without ampicillin (100 μg/mL) or kanamycin (50 μg/mL). BHK-21, HEK293 T, MDCK and MDCK-STAT1KO cells were purchased from ATCC and cultured in Dulbecco’s modified Eagle’s medium (DMEM; BasalMedia, Gibco) supplemented with 10% heat-inactivated fetal bovine serum (FBS; Sigma-Aldrich) and 10 U/mL penicillin–streptomycin (Gibco) at 37 ℃ and 5% CO₂.

The A/Hong Kong/68 (X-31) (H3N2) and B/Lee/1940 strains were propagated in 9- to 10-day-old specific pathogen-free (SPF) embryonated chicken eggs (Wens). Briefly, eggs were candled to confirm embryo viability before inoculation. Virus stock (1,000 FFU/egg) was injected into the allantoic cavity using a sterile needle. Inoculated eggs were incubated at 37℃ for 48–72 h, then chilled at 4℃ for ≥4 h to facilitate harvesting. The allantoic fluid was collected, clarified by centrifugation (3,000 ×g, 10 min, 4℃), aliquoted, and stored at −80℃.

The following strains were propagated in MDCK-STAT1 KO cells to enhance viral yield: A/Wisconsin/67/2022 (H1N2), A/Michigan/45/2015 (H1N1), A/Puerto Rico/8/1934 (H1N1), A/California/07/2009 (H1N1pdm09), A/Darwin/6/2021 (H3N2), A/Hong Kong/45/2019 (H3N2), B/Austria/1359417/2021(Victoria), B/Washington/02/2019 (Victoria), B/Guangzhou/0215/2012 (Victoria). Cells were infected at an MOI of 0.1 and incubated at 37℃, 5% CO₂ for 48–72 h. Viral supernatants were harvested, centrifuged (3,000 g, 10 min, 4℃), aliquoted, and stored at −80℃. Viral titres were measured using a Fluorescent Focus Assay with MDCK cells (as detailed in the Fluorescent Focus Assay section). All viruses were sequenced to confirm that the correct sequences were retained during propagation.

MDCK cells were seeded in 96-well plates at a density of 2.5 × 10⁴ cells/well using DMEM supplemented with 5% fetal bovine serum, and incubated overnight at 37℃ with 5% CO₂ until reaching approximately 90% confluence. Virus stocks were serially diluted 10-fold in infection medium (DMEM containing 2 μg/mL TPCK-trypsin (Sigma)). Cell monolayers were washed once with PBS prior to inoculation with 100 μL of diluted virus per well. Following 1 h incubation at 37℃ with gentle rocking every 15 min to ensure uniform viral adsorption, the inoculum was removed and replaced with 100 μL of 1% CMC in MEM mixture. Plates were then incubated for 2842 h at 37℃ with 5% CO₂ to allow viral replication. After incubation, cells were fixed with 4% paraformaldehyde in PBS for 1 h at room temperature, followed by permeabilization with 0.1% Triton X-100 in PBS for 30 min. After three washes with PBS, cells were blocked with 1% BSA in PBS for 1 h at room temperature. Immunostaining was performed using mouse anti-NP monoclonal antibody (4R5) (1:4000 dilution in 1% BSA/PBS) for 1 h, followed by three washes with PBST (PBS containing 0.05% Tween-20). HRP-conjugated goat anti-mouse IgG (1:2000 dilution in 1% BSA/PBS) was then applied and incubated for 1 h protected from light. Following three final washes with PBST, TMB One Solution(HRP-based WB) (YEASEN) was added for 10 min, and quantification was performed using an ELISpot reader (AT-Stop1100/CTL).

The HA genes were derived from the WHO-recommended 2023–2024 Northern Hemisphere vaccine strains: H1N1 A/Wisconsin/67/2022 (GISAID: EPI_ISL_15928563), H3N2 A/Darwin/6/2021 (GISAID: EPI_ISL_3534319), and B/Victoria lineage B/Austria/1359417/2021 (GISAID: EPI_ISL_983345). Target sequences were optimized using mammalian codon bias adaptation and RNA secondary structure minimization algorithms to enhance translational efficiency and stability in mammalian cells. At the same time, we sought to improve mRNA secondary structure stability by maximizing the minimum free energy (MFE) while maintaining the original amino acid sequence. This was achieved by exploring synonymous substitutions within the coding region, leveraging the degeneracy of the genetic code. We employed the CDSfold algorithm to guide the identification of synonymous sequence. The software is available at https://github.com/gterai/CDSfold↗ [35]. The mRNA was synthesized via in vitro transcription (IVT) from linearized DNA templates. These templates, which contained the human β-globin UTRs and a 100-nt poly(A) tail, were generated by PCR and purified (Promega gel extraction kit). The IVT setup included T7 RNA polymerase (APExBIO), pseudouridine (Ψ; APExBIO) in place of UTP to reduce immunogenicity, and a Cap1 analog (APExBIO) for 5’ capping. Post-transcription, the mixture was treated with DNase I (APExBIO) to remove the template DNA, followed by mRNA purification using oligo-dT chromatography [36]. The saRNA constructs contain the Venezuelan Equine Encephalitis Virus strain TC-83 5’ UTR, 3’ UTR, non-structural protein genes, and the influenza virus HA gene. The saRNA was synthesized by IVT as previously described for mRNA. circRNA was synthesized using the permuted intronexon (PIE) splicing strategy. The circularization protocol was performed as follows:

First, IVT was conducted at 47℃ for 2 h using thermostable T7 polymerase (APExBIO) to generate pre-circRNA. Following initial purification using conventional methods (which proved insufficient for removing precursor and nicked byproducts), the circularized products underwent additional purification through HPLC (Sepax) with elution in 1×TE buffer, followed by a final purification step using an RNA clean-up kit (Magen Biotechnology Co., Ltd., Guangzhou, China). The integrity and purity of mRNA and saRNA were assessed by TBE-urea gel electrophoresis, and mRNA was further characterized by capillary electrophoresis using the Agilent DNF-471 RNA Kit. The circularization efficiency of circRNA was assessed using the E-Gel system (Thermo Fisher).

To assess protein expression, BHK-21 and HEK293T cells were cultured in 12-well plates until reaching 50–60% confluency. These cultures were then transfected with varying amounts (250, 500, or 1000 ng) of monovalent (MRV) or trivalent (TRV) mRNA vaccines, complexed with Lipo8000 transfection reagent (Beyotime) in OptiMEM reduced serum medium (Gibco). At 24 h after transfection, the cells were lysed using RIPA buffer supplemented with a protease inhibitor cocktail. The protein concentration in the resulting lysates was determined by a bicinchoninic acid (BCA) assay (Thermo Scientific). To enable precise quantification of HA expression, standard curves were generated using known concentrations of recombinant HA proteins (W67HA, D6-HA, and AUT21-HA; South China Vaccines). All cell lysate samples were subsequently normalized to specific loading concentrations based on these standardsW67 (20 µg/well), D6 (10 µg/well), and AUT21 (20 µg/well)-before proceeding to immunoblotting. The normalized protein samples were electrophoresed on 4–20% polyacrylamide gradient gels (ACE Biotechnology) and subsequently electrotransferred onto PVDF membranes (Millipore). After blocking with a 5% solution of non-fat dry milk, the membranes were incubated with specific primary antibodies targeting H1, H3, or BV HA proteins (all from Sino Biological) for one hour at ambient temperature. Following extensive washes, the blots were exposed to a horseradish peroxidase (HRP)-conjugated goat anti-human secondary antibody (Sangon Biotech; 1:5000) for another hour. Immunoreactive bands were detected by applying an enhanced chemiluminescence substrate (CoWin Biotech), and digital images were captured using an iBright CL750 system (Thermo Fisher). The band intensities were finally quantified through greyscale analysis with ImageJ software [36].

Lipid nanoparticle (LNP) formulation was prepared as follows: SM102 (Shenzhou Pharmaceutical), DSPC (Nippon Fine Chemical), cholesterol (Nippon Fine Chemical), and PEG2000-DMG (JenKem Technology) in absolute ethanol at a molar ratio of 50:10:38.5:1.5. This lipid mixture was then combined with RNA dissolved in 25 mM sodium acetate buffer (pH 5.0) using a Nano Precision microfluidic system (Shanghai RNACure Biopharma) with controlled flow parameters to ensure consistent nanoparticle formation. Immediately following mixing, the formulations were subjected to a two-fold dilution with the same sodium acetate buffer to optimize particle stability, then dialysed overnight at 4℃ against 20 mM Tris-acetate buffer (pH 7.5) using Slide-A-Lyzer dialysis cassettes with a 10 kDa molecular weight cutoff. After dialysis, the formulations were processed through sequential steps including concentration using Amicon Ultra centrifugal filters (Millipore), sterile filtration through 0.22 µm PVDF membranes (Merck), and comprehensive characterization. The nanoparticle size distribution and polydispersity index were determined by dynamic light scattering, while surface charge was measured via phase analysis light scattering using a Zetasizer Nano ZA system (Malvern Instruments). RNA encapsulation efficiency was quantitatively assessed using the Quant-iT RiboGreen assay (Thermo Fisher Scientific), with measurements taken both before and after Triton X-100 treatment to accurately determine the percentage of encapsulated RNA.

The vaccination procedures were conducted in strict accordance with the approved animal protocol (GZLAB-AUCP-2023-10-A7) from the Animal Care and Use Committee of Guangzhou Laboratory, China. The rodent housing facility maintained an ambient temperature of 22℃ ± 3℃ with relative humidity controlled between 30% and 70%. An automated 12-hour light/12-hour dark cycle was implemented throughout the study period.

For Figure 2, female BALB/c mice (7-week-old) were randomly allocated into experimental groups and immunized intramuscularly in the quadriceps with mRNALNP formulations at graded doses (0.1, 1, or 10 μg). Age-matched control mice received either 2 μg of 2023–2024 seasonal QIV. All immunizations followed a primeboost schedule with identical doses administered at 3-week intervals. Serial blood collections were performed via retro-orbital bleeding at 3 weeks post-prime and 3 weeks post-boost (week 6) for humoral immunity evaluation. Four weeks after the boost immunization, mice were challenged with a 5×LD₅₀ dose of virus.

For the safety evaluation shown in Figure 3(K–M) and Figure S5, 7-week-old female BALB/c mice were intramuscularly immunized with a single dose of mRNA-TRV (1 μg, 10 μg, or 30 μg), with placebo-treated mice as controls. Body weight was monitored for 14 days after immunization, and blood samples were collected at 0 h and 14 d postimmunization for serum biochemical analysis, including AST, CK, ALT, ALP, TP, ALB, CREA, and LDH.

For Figure 6: Seven-week-old female BALB/c mice were intramuscularly immunized with either saRNA-LNP or circRNA-LNP (0.1 μg, 0.6 μg, or 1 μg), alongside control groups receiving mRNA-LNP at matched doses (0.1 μg, 0.6 μg, or 1 μg). Blood samples were collected at week 3 for immunological assessment. Four weeks after a single immunization, mice were challenged with a 5×LD₅₀ dose of virus. Following infection, body weight was monitored daily for 14 days. Mice reaching the predefined humane endpoint (≥25% weight loss from baseline) were euthanized.

Serum levels of antigen-specific IgG antibodies were measured using a standardized ELISA. Briefly, 96-well microplates were coated with 100 μL/well of recombinant HA protein (0.5 μg/mL in coating buffer) and incubated overnight at 4℃. After three washes with PBS, nonspecific binding sites were blocked with 200 μL/well of 5% nonfat milk in PBS for 1 h at 37℃. Serum samples and monoclonal antibodies were subjected to three-fold serial dilutions (initial dilution 1:300) in PBS. Following incubation and three additional washes, 100 μL/well of horseradish peroxidase (HRP)conjugated goat anti-mouse IgG secondary antibody (Sangon Biotech; 1:2000 dilution in PBS) was added and incubated for 1 h at 37℃. Colour development was initiated by adding TMB substrate (Beyotime), and the reaction was terminated with 1 M sulphuric acid. Optical density was measured at 450 nm using an Ensight Multimode Plate Reader (PerkinElmer). The endpoint titre was defined as the reciprocal of the highest serum dilution that yielded an absorbance value ≥2-fold above the blank control.

The focus reduction neutralization test (FRNT) was employed to quantify the neutralizing antibodies present in immune sera. Briefly, MDCK cell monolayers were established by seeding cells into 96-well flat-bottom plates (Costar 3599, Corning) at a density of 3 × 10⁵ cells per well and culturing for 24 h under standard conditions (37°C, 5% CO₂). Sera obtained from immunized mice underwent pretreatment with receptor-destroying enzyme (RDE; Denka-Seiken 340122). This involved combining serum with RDE at a 1:3 ratio, incubating the mixture overnight (18 h) at 37°C, and subsequently heat-inactivating it at 56°C for 30 min. The treated sera were then subjected to three-fold serial dilutions in virus diluent (DMEM containing 25 mM HEPES and 2 μg/mL TPCK-trypsin), commencing at a 1:25 dilution. An aliquot of each serum dilution was combined with a challenge dose of 100 FFU of virus in MEM and allowed to interact for one hour at 37°C. These serum-virus complexes were then applied to the pre-formed MDCK monolayers. After a one-hour adsorption period, the inoculum was aspirated, and the cell layers were rinsed twice with phosphate-buffered saline (PBS). To restrict viral spread and facilitate focus formation, the cells were overlaid with 100 μL of a semi-solid medium consisting of MEM, 1% carboxymethyl cellulose (CMC), 1 μg/mL TPCK-trypsin, and 25 mM HEPES, followed by a 24-hour incubation at 37°C. Post-incubation, the cells were fixed with 4% paraformaldehyde. They were then permeabilized and blocked in a single step using a solution of 0.2% Triton X-100 and 1% bovine serum albumin (BSA). For immunostaining, the fixed cells were probed for one hour at 37°C with specific primary antibodies: either a mouse anti-nucleoprotein (NP) monoclonal antibody (4R5) or a rabbit anti-influenza B virus hemagglutinin (BV HA) antibody (Sino Biological). This was followed by a one-hour incubation with the corresponding horseradish peroxidase (HRP)-conjugated secondary antibodies: goat anti-mouse IgG or goat anti-rabbit IgG (Sangon Biotech; both at 1:2000 dilution). Viral foci were developed by applying a TMB-based substrate (TMB One Solution, YEASEN) for 10 min. The resulting foci were enumerated using an ELISpot reader (AT-Stop1100/CTL). The neutralization titre, reported as the FRNT₅₀, was calculated as the reciprocal of the maximum serum dilution that achieved a ≥ 50% reduction in focus count relative to the virus-only control wells [36].

Antigen-specific T-cell responses were evaluated by enzyme-linked immunospot (ELISpot) assays measuring IFN-γ, IL-2, and TNF-α. Splenocytes were isolated from mice following either a single immunization or a prime-boost regimen, with assays conducted 12 days after the respective immunization. 96-well ELISpot plates (MabTech, Sweden) were coated overnight with capture antibodies against mouse IFN-γ (#33214HST-2), IL-2 (#3441-4HPW-2), and TNF-α (#3511-4HPW-2). Prior to cell plating, the coated plates were washed once with PBS and subsequently blocked for 30 min at room temperature with complete RPMI-1640 medium supplemented with 10% fetal bovine serum (FBS). Single-cell suspensions of splenocytes were seeded into the plates at a density of 2.5 × 10⁵ cells per well. The cells were stimulated with 2 μg/mL of peptide pools spanning the HA sequences of the relevant influenza strains. These pools included: W67-H1 (H1 strain), D6-H3 (H3 strain), and AUT21-BV (influenza B strain). All peptides were 15-mers with an 11-amino-acid overlap and synthesized to ≥90% purity (GenScript). Unstimulated wells served as negative controls. The plates were then incubated for 36 h at 37°C under 5% CO₂ to allow for cytokine secretion and capture. Following the incubation period, the cells were decanted, and the plates were subjected to five rigorous washes with 200 μL of PBS per well. The captured cytokines were detected by incubating the wells for 2 h at room temperature with biotinylated anti-cytokine antibodies (MabTech; 1 μg/mL in PBS containing 0.5% FBS). After another series of washes, streptavidin-conjugated horseradish peroxidase (HRP) was added to the wells and incubated for 1 h at room temperature. Spot development was achieved by adding a TMB substrate solution. The resulting cytokine spots were quantified using an automated ELISpot plate reader (PerkinElmer). The frequency of antigen-specific T cells was expressed as the number of spot-forming units (SFU) per million splenocytes after subtracting the background from unstimulated controls [36].

Antigen-specific T-cell responses were evaluated by intracellular cytokine staining following in vitro peptide restimulation. Splenocytes were isolated from mice following either a single immunization or a prime-boost regimen, and assays were performed 12 days after the respective immunization. Splenocytes (1 × 10⁶ cells per well) were cultured in RPMI-1640 medium supplemented with 10% fetal bovine serum (FBS) and stimulated with 5 μg/mL of peptide pools spanning the HA sequences of the relevant influenza strains, including W67-H1 (H1 strain), D6-H3 (H3 strain), and AUT21-BV (influenza B strain). All peptides were 15-mers with an 11-amino-acid overlap and synthesized to ≥90% purity (GenScript). Unstimulated cells served as negative controls, whereas Concanavalin A (10 μg/mL; InvivoGen, #inh-cona) was used as a positive control. GolgiPlug™ (BD Biosciences, #555029) and GolgiStop™ (BD Biosciences, #554724) were added at a 1:1,000 dilution, and the cells were incubated for 6 h at 37°C.

Following stimulation, cells were washed with PBS and surface-stained for 30 min at 4°C in the dark using anti-CD3e FITC (BD Biosciences, #553061), anti-CD4 PE-Cy7 (BD Biosciences, #552775), anti-CD8a BV510 (BD Biosciences, #563068), and Fixable Viability Stain 780 (BD Biosciences, #565388), diluted in MACS buffer (PBS containing 2% BSA and 1 mM EDTA). Cells were then fixed and permeabilized with Cytofix/Cytoperm solution (BD Biosciences, #554722) for 20 min at 4°C, followed by intracellular staining with anti-IFN-γ BV711 (BD Biosciences, #564336), anti-TNF-α BB700 (BD Biosciences, #566510), and anti-IL-2 PE (BD Biosciences, #554428), each diluted 1:100 in Perm/Wash buffer. After staining, cells were resuspended in MACS buffer and acquired on a NovoCyte Advanteon flow cytometer (Agilent Technologies). Data were analysed using Novo Express software, and antigen-specific responses were expressed after subtraction of background responses in unstimulated controls [36].

Data were analysed using GraphPad Prism version 10.2.3. The difference between two groups was calculated using a t-test. Statistically significant results were determined when p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Data are presented as the mean with standard deviations as indicated.