Adjuvant-dependent protection of SARS-CoV-2 spike vaccines: comparative immunogenicity of human-applicable formulations

HEK293T and HEK293 cells were obtained from the American Type Culture Collection (ATCC). The HEK293 cells stably expressing human ACE2 (HEK293-ACE2) were generated by infecting HEK293 cells with a human ACE2 lentivirus, followed by puromycin selection. Vero E6 cells were kindly provided by Dr. Rong Zhang of Fudan University. HEK293T, HEK293-ACE2, and Vero E6 cells were cultured in Dulbecco's Modified Eagle's Medium (DMEM; Gibco, USA) supplemented with 10% fetal bovine serum (FBS; Gibco, USA), 1% penicillin-streptomycin (Thermo Fisher Scientific), 1% L-glutamine (Thermo Fisher Scientific), and 1% non-essential amino acids (Thermo Fisher Scientific) at 37°C with 5% CO₂. FreeStyle CHO-S suspension cells (Invitrogen, USA) were cultured in EX-CELL 325 PF CHO Serum-Free Medium (Sigma-Aldrich) supplemented with 1% L-glutamine and 1% penicillin-streptomycin. Cells were maintained in nonpyrogenic, vented polycarbonate Erlenmeyer flasks (Thermo Fisher Scientific) at 37°C with 5% CO₂ and constant shaking at 130 rpm in an incubator shaker. All SARS-CoV-2 viruses were isolated from nasopharyngeal swab samples collected from nucleic acid-positive cases and subsequently propagated and titrated in Vero E6 cells as previously described (36). The viral strains included the following: SARS-CoV-2 ancestral strain (GenBank: MT622319.1↗), Omicron BA.2 (GenBank: MT627325.1↗), Omicron BA.5 (GenBank: PQ213095.1↗), and Omicron BF.7 (GenBank: PQ351184.1↗).

The SARS-CoV-2 S1 protein, S2 protein, and RBD protein corresponding to the ancestral, Delta, and Omicron variants were obtained from Sino Biological (Beijing, China).

The mammalian cell codon-optimized ectodomain of ancestral SARS-CoV-2 spike trimer cDNA was synthesized by Generay Biotech Co., Ltd (Shanghai, China), in which a trimeric foldon from T4 phage fibritin (Tfd) was fused to the C-terminus of SARS-CoV-2 spike ectodomain (residues 1–1,211; GenBank: NC_045512.2), with the furin cleavage site RRAR replaced by GSAS and two proline mutations (K986P/V987P) introduced to stabilize the prefusion conformation, along with a C-terminal polyhistidine tag for purification (1). This cDNA was cloned into a modified pCI-GS CHO expression vector and electroporated into CHO cells, followed by selection of stable spike trimer-expressing clones using glutamine-deficient medium containing 25 µM methionine sulfoximine (Sigma). A selected clone was gradually scaled up to 200 mL culture volume, after which supernatants were harvested and purified sequentially through ultrafiltration, nickel affinity chromatography, and desalting. The purified recombinant protein (designated S-2P) was quantified by BCA assay (Thermo Fisher Scientific), with purity verified by Coomassie blue staining and S1-domain-specific western blotting under both denaturing and non-denaturing conditions.

The binding of S-2P to human ACE2 was assessed by enzyme-linked immunosorbent assay (ELISA). Briefly, S-2P protein (2 µg/mL) was coated onto high-binding 96-well plates (Thermo Fisher Scientific) and incubated overnight at 4°C. After washing with PBST (PBS containing 1% Tween-20), plates were blocked with blocking buffer (PBST containing 1% BSA) for 2 hours at room temperature. Following another PBST wash, serially diluted human ACE2 protein (mFc Tag; Sino Biological, Beijing, China) was added to the plates and incubated for 2 hours at room temperature. After washing, HRP-conjugated goat anti-mouse IgG1 cross-adsorbed secondary antibody (Invitrogen; 1:1,000 dilution) was added and incubated for 2 hours at room temperature. Following final PBST washes, the reaction was developed using 1-Step Slow TMB solution (Thermo Fisher Scientific) and stopped with ELISA Stop Solution (Solarbio, China). Absorbance was measured at 450 nm.

C57BL/6 mice were obtained from Shanghai Jihui Laboratory Animal Care Co., Ltd. (Shanghai, China), and CAG-hACE2 transgenic mice on the C57BL/6 background were acquired from Shanghai Model Organisms Center, Inc. (Shanghai, China). Unless otherwise specified, the experimental mice were 6–8 weeks of age. For the aged group, 18-month-old mice were used, as the standard laboratory mouse has an average lifespan of about 2 years, with 18–24 months generally considered old age (37). All animals were housed in individually ventilated cages under specific pathogen-free conditions. All animal experiments were reviewed and approved by the Institutional Committee on Ethics of Medicine of Navy Medical University and conducted in accordance with China's Regulations for the Administration of Affairs Concerning Experimental Animals.

The recombinant S-2P protein was formulated with alum (Croda, Denmark), Montanide ISA 720 VG, Montanide ISA 51 VG, or Sepivac SWE (all from Seppic, France), designated as S-2P:Al, S-2P:ISA720, S-2P:ISA51, and S-2P:SWE, respectively. Each mouse was intramuscularly injected with 100 µL of vaccine formulation containing 10 µg of S-2P protein, with adjuvant volume ratios of 70%, 50%, and 50% for S-2P:ISA720, S-2P:ISA51, and S-2P:SWE, respectively, while S-2P:Al contained 100 µg of aluminum hydroxide. Unless otherwise specified, C57BL/6J or hACE2 mice were immunized twice at 3 week intervals with S-2P:Al, S-2P:ISA720, S-2P:SWE, or S-2P:ISA51, while control animals received PBS alone.

Mouse serum antibodies against SARS-CoV-2 were evaluated by ELISA as described (38). Briefly, recombinant SARS-CoV-2 S1, S2, or RBD protein (1 µg/mL) was coated onto high-binding 96-well plates overnight at 4°C, followed by blocking with blocking buffer. Serially diluted serum samples were then added to the plates and incubated at room temperature for 2 hours. After extensive washing, HRP-conjugated secondary antibodies (goat anti-mouse IgG, goat anti-mouse IgG1, or goat anti-mouse IgG2b; Thermo Fisher Scientific) were added. Endpoint titers were determined as the reciprocal of the highest dilution showing an absorbance at 450 nm ≥2.1fold higher than control mouse sera.

For the blocking ELISA assessing antibody inhibition of RBD-hACE2 binding, RBD-coated (1 µg/mL) 96-well plates were first incubated with serially diluted serum samples for 2 hours at room temperature before adding hACE2 protein (0.25 µg/mL). After washing, rabbit anti-hACE2 antibody (Sino Biological) was added for 2 hours at room temperature, followed by HRP-conjugated goat anti-rabbit IgG (Thermo Fisher Scientific). The remaining steps were performed as described above.

Vero E6 cells were seeded in 96-well plates (1 × 10⁴ Vero E6 cells per well) and cultured overnight. Serially diluted serum samples were mixed with either 0.1 MOI (multiplicity of infection) of SARS-CoV-2 ancestral strain or 0.01 MOI of Omicron BA.2 strain at 37°C for 30 minutes, then the virus-serum mixtures were added to Vero E6 cells in the presence of 2 µg/mL TPCK-treated trypsin. At 24 hours post-infection, cells were fixed with methanol at −20°C for 30 minutes and blocked with 3% BSA. Rabbit polyclonal antibodies against SARS-CoV-2 nucleocapsid protein (NP; Sino Biological, Beijing, China) were added and incubated overnight at 4°C, followed by PBS washing and incubation with Alexa Fluor 488-conjugated goat anti-rabbit IgG (Thermo Fisher Scientific) at room temperature for 1.5 hours. After DAPI staining (Sigma-Aldrich), cells were imaged using a Cytation 5 system (BioTek, USA). Infected cell numbers were quantified using Gen5 3.10 software, with neutralization percentage calculated relative to virus-only controls. The 50% inhibitory concentration (IC₅₀) was determined as the reciprocal serum dilution showing 50% infection inhibition and calculated using GraphPad Prism 8.0 (GraphPad Software, Inc.).

HEK293T cells were co-transfected using Lipofectamine 2000 reagent (Thermo Fisher Scientific) with HIV capsid packaging plasmids, a transfer plasmid harboring the EGFP reporter gene, and a plasmid harboring the spike gene of SARS-CoV-2 or SARS-CoV. In the SARS-CoV-2 spike expression plasmid, the fragment encoding a 19-amino-acid at the C-terminal of spike was truncated. Following 48 hours of incubation, pseudovirus-containing supernatants were harvested, centrifuged at 1,000 × g for 10 minutes, and stored at −80°C until use. For neutralization assays, serially diluted samples of the test sera were pre-incubated with pseudoviruses at 37°C with 5% CO₂ for 30 minutes, then the mixtures were added to 293T-ACE2 target cells (seeded at 1 × 10⁴ cells/well). After 4 hours, the medium was replaced with fresh DMEM supplemented with 2% FBS, and EGFP-positive cells were quantified at 48 hours post-infection using a Cytation 5 cell imaging multimode reader (BioTek). Neutralization titers (ID50) were calculated using nonlinear regression analysis in GraphPad Prism 8.0 (GraphPad Software).

T-cell-mediated immune responses in mice were evaluated using an interferon gamma (IFN-γ) ELISPOT kit (Dakewe, China) as described (38). Spleen and lung cells were isolated 14 days post-secondary immunization. Briefly, IFN-γ monoclonal antibody-precoated ELISpot plates were blocked with 200 µL/well of serum-free ELISPOT medium (Dakewe, China) for 5–10 minutes at room temperature. Spleen or lung cells (2 × 10⁵/well) were stimulated with ancestral SARS-CoV-2 S1 or S2 protein at 400 µg/mL for 20 hours at 37°C with 5% CO₂. IFN-γ-producing cells were detected using biotinylated detection antibody and streptavidin-HRP, followed by spot visualization with an S6 Ultra ImmunoSpot Reader (Cellular Technology Ltd.) and quantification using ImmunoSpot 5.1.36 software (Cellular Technology Ltd.). Spot-forming cells were enumerated as a measure of antigen-specific T-cell responses.

Under isoflurane anesthesia, hACE2 transgenic mice were intranasally inoculated with 50 µL of viral suspension containing SARS-CoV-2 ancestral strain (1 × 10² PFU), Omicron BA.5 (1 × 10⁴ PFU), or Omicron BF.7 (1 × 10⁴ PFU) as described (38). Mice were monitored daily for body weight changes and survival rates. At 4 days post-infection (dpi), a subset of mice was humanely euthanized for tissue collection (nasal turbinates, lungs, and brains) to assess viral load and inflammatory factors, while the remaining mice were maintained for longitudinal survival analysis.

Viral RNA levels and inflammatory cytokine gene expression were analyzed by quantitative reverse transcription PCR (qRT-PCR). Tissue samples (brain, lung, and nasal turbinates) from mice were homogenized in TRIzol Reagent (Invitrogen) for total RNA extraction according to the manufacturer's protocol. First-strand cDNA synthesis was performed using PrimeScript RT Master Mix (TaKaRa, Japan), followed by quantitative PCR amplification with TB Green Premix Ex Taq II (TaKaRa, Japan) and the following gene-specific primers (5′→3′): SARS-CoV-2 nucleocapsid (N) gene (forward:, reverse:); mouse β-actin (forward:, reverse:); IL-6 (forward:, reverse:); CCL2 (forward:, reverse:); and CXCL10 (forward:, reverse:). All data were normalized to β-actin expression levels. AAGGCGTTCCAATTAACACCA TGCCGTCTTTGTTAGCACCA GGCTGTATTCCCCTCCATCG GCACAGGGTGCTCCTCAG TCGGAGGCTTAATTACACA TCATACAATCAGAATTGCCAT TCGGAACCAAATGAGATCAGA TAGCTTCAGATTTACGGGTCA AATTTAATGAAAGCGTTTAGCC ATTAGGACTAGCCATCCAC

Passive immunization was performed through adoptive transfer of immune sera to evaluate protective efficacy against SARS-CoV-2 challenge in recipient hACE2-transgenic mice, as previously described (39). Briefly, hACE2 transgenic mice were intravenously administered 0.1 mL of pooled sera collected 14 days post-booster immunization (from vaccinated mice) or PBS-treated control mice. Twenty-four hours post-serum transfer, mice were intranasally challenged with live SARS-CoV-2 ancestral strain. Body weights were monitored daily until day 4 post-infection, when mice were humanely euthanized for virological and immunological analyses. Viral RNA loads and inflammatory factor expression profiles in brain tissues were quantified by qRT-PCR.

Statistical analyses were conducted using GraphPad Prism 8.0 (GraphPad Software). Between-group comparisons were performed using either unpaired or paired two-tailed Student's t-tests as appropriate, while multiple group comparisons were analyzed by one-way ANOVA followed by Tukey's post hoc test. Statistical significance was defined as P < 0.05, with the following notation: *P < 0.05, **P < 0.01, and ***P < 0.001; ns (not significant) for P ≥ 0.05.

The prefusion-stabilized spike trimer of ancestral SARS-CoV-2 was expressed in CHO cells and purified via affinity chromatography. The purified protein, designated S-2P, was confirmed by native and denaturing SDS-PAGE and western blot using a monoclonal antibody (mAb) targeting the S1 subunit (). ELISA analysis demonstrated dose-dependent binding of soluble human ACE2 to immobilized S-2P (). Furthermore, S-2P effectively inhibited SARS-CoV-2 infection in Vero-E6 cells (). Fig. S1A Fig. S1B Fig. S1C

Aluminum hydroxide (Croda, Denmark) and Seppic's Montanide ISA 720, ISA 51, and Sepivac SWE (Seppic, France) were evaluated for their ability to enhance immune responses to recombinant S-2P protein in C57BL/6 mice. Mice were immunized intramuscularly twice at a 3-week interval with 10 µg S-2P protein formulated with either aluminum hydroxide or Seppic's adjuvants (denoted as S-2P:Al, S-2P:ISA720, S-2P:ISA51, and S-2P:SWE, respectively). Serum samples were collected 2 weeks after each immunization and analyzed for antigen-specific IgG and neutralizing antibodies (Fig. 1A). Comparative analysis showed significant differences in immunogenicity depending on the adjuvant used. Both S-2P:ISA720 and S-2P:ISA51 induced stronger antibody responses compared to S-2P:Al or S-2P:SWE. Sera from S-2P:ISA720-immunized mice showed the highest levels of S1- and S2-specific IgG (Fig. 1B and C), the strongest inhibition of RBD binding to hACE2 (Fig. 1D) and the most potent neutralizing activity against the ancestral SARS-CoV-2 (4.3-, 5.1- and 2.2-fold higher than S-2P:Al, S-2P:SWE and S-2P:ISA51 groups at week 5, respectively) (Fig. 1E).

To assess antigen-specific T-cell responses, immunized mice were sacrificed 3 weeks after the booster dose. Single-cell suspensions were prepared from both spleens and lungs, then stimulated ex vivo with ancestral S1 or S2 proteins. IFN-γ-producing cells were quantified using enzyme-linked immunospot (ELISpot) assay. The results showed that mice immunized with S-2P:ISA720 or S-2P:ISA51 exhibited significantly stronger antigen-specific T-cell responses in both splenic and pulmonary compartments, whereas S-2P:Al and S-2P:SWE failed to induce detectable antigen-specific IFN-γ-producing T cells (Fig. 1E and F; Fig. S1D and E).

The CAG-hACE2 transgenic mouse represents a valuable small animal model for evaluating candidate vaccine efficacy. Although SARS-CoV-2 infection affects both upper and lower respiratory tracts and induces tissue damage, viral invasion into the central nervous system emerges as the primary cause of mortality in this model (40). We immunized CAG-hACE2 transgenic mice (C57BL/6 background) twice at 3-week intervals with 10 µg S-2P protein formulated with different adjuvants (14 mice/group), along with an additional S-2P:ISA720 single-dose group (10 mice) (Fig. 2A). Consistent with observations in wild-type C57BL/6 mice, all formulations efficiently induced anti-RBD IgG (Fig. 2B and C) and virus-neutralizing antibodies in hACE2 mice (Fig. 2D and E), with S-2P:ISA720 showing superior immunogenicity, exhibiting 10.8-, 7.8-, and 3.9-fold higher neutralizing titers versus S-2P:Al, S-2P:SWE, and S-2P:ISA51, respectively at week 5. Remarkably, single-dose S-2P:ISA720 outperformed two-dose S-2P:Al and S-2P:SWE in neutralizing antibody induction, while S-2P:Al showed consistently weaker responses than ISA51- or ISA720-formulated vaccines (Fig. 2C and E). Moreover, S-2P:ISA51 and S-2P:ISA720 induced Th1-skewed immunity (evidenced by elevated IgG2b:IgG1 ratios; Fig. S2A and B) that correlated with a robust IFN-γ-producing T-cell response observed in wild-type C57BL/6 mice.

Three weeks after booster immunization, mice were intranasally challenged with 100 PFU of ancestral SARS-CoV-2. Unvaccinated control mice exhibited rapid disease progression, with severe weight loss and 100% mortality within 5 days post-challenge (Fig. 2F and G). All vaccinated groups showed varying degrees of protection: the S-2P:ISA720 group demonstrated complete protection (100% survival, 14/14), while S-2P:ISA51 showed 85.7% survival (12/14) (Fig. 2F and G). Comparatively, S-2P:SWE and S-2P:Al groups exhibited relatively lower protection rates of 42.9% (6/14) and 35.7% (5/14), respectively. Notably, even a single dose of S-2P:ISA720 conferred 90% protection (9/10 survivors), highlighting its superior efficacy.

To elucidate the protective role of vaccine-induced neutralizing antibodies, we analyzed the correlation between humoral immune responses and survival rates in vaccinated mice. The S-2P:Al and S-2P:ISA720 groups demonstrated strong correlations between anti-RBD IgG and neutralizing antibody titers, while the S-2P:SWE and S-2P:ISA51 group showed a relatively weaker correlation (Fig. S2C through G), suggesting adjuvants critically influence both the magnitude and functional quality of antibody responses. Protection analysis revealed that neutralizing antibody titers ≧3 × 10³ conferred complete protection (5/5) in S-2P:Al-immunized mice and 71% protection (5/7) in S-2P:SWE-immunized mice (Fig. S2C through G). Since neither S-2P:Al nor S-2P:SWE immunization induced detectable T-cell response, the observed protection can be primarily attributed to neutralizing antibodies.

We further confirmed the protective efficacy of vaccine-induced antibodies through passive immunization in naïve hACE2 mice. The mice (n = 10 per group) were administered 100 µL of pooled sera from immunized or control mice 1 day prior to SARS-CoV-2 challenge. Five mice per group were sacrificed at 4 days post-infection for analysis of brain viral loads and cytokine profiles, while the remaining five mice were monitored for clinical outcomes. All mice that received control sera or sera from S-2P:Al-immunized mice exhibited significant weight loss and mortality. In contrast, Seppic-adjuvanted vaccine immune sera conferred partial protection, with 1, 3, and 3 mice receiving sera from mice immunized with SWE-, ISA51-, and ISA720-adjuvanted vaccine surviving, respectively (Fig. 2H). The observed reduction in viral titers and pro-inflammatory cytokine levels in murine brain tissue demonstrates that vaccine-induced antibodies effectively inhibit SARS-CoV-2 neuroinvasion and mitigate associated neuropathological damage (Fig. 2I and J).

Pseudovirus-based neutralization assays were employed to quantify the cross-neutralizing antibody induced by various adjuvanted S-2P vaccines against SARS-CoV-2 variants of concern (VOCs) in murine models. While Delta variant neutralization showed modest (0.8- to 1.3-fold) titer reduction versus ancestral strain, Omicron BA.1 neutralization was substantially impaired across sera of all groups (). The majority of sera from S-2P:Al- and S-2P:SWE-immunized mice showed no detectable neutralization against BA.1 at 100-fold dilution, while 84.6% (11/13) of S-2P:ISA51 and 100% (11/11) of S-2P:ISA720 immune sera maintained neutralizing activity, despite a 17-fold decrease in geometric mean neutralization titers compared to neutralization against ancestral strain for the S-2P:ISA720 group. Importantly, both ISA-adjuvanted vaccine groups demonstrated cross-neutralization capacity against SARS-CoV-1 pseudovirus at low serum dilutions (). Fig. S3A Fig. S3B

Next, hACE2-transgenic mice were immunized with either S-2P:Al or S-2P:ISA720 to evaluate the in vivo protective efficacy against Omicron variants BA.5 and BF.7. Consistent with the findings in Results 1 and 2, immunization with S-2P:ISA720 elicited significantly higher titers of anti-RBD antibody and neutralizing antibody against the ancestral strain compared to S-2P:Al (Fig. 3A and B). Notably, in authentic virus neutralization assays, sera from the majority of S-2P:ISA720-immunized mice potently neutralized Omicron subvariants BA.2, BA.5, and BF.7. In contrast, only minimal neutralizing activity was detected in sera from S-2P:Al-immunized mice, with significantly lower titers (Fig. 3B). Following viral challenge, all of the S-2P:Al-immunized mice exhibited progressive weight loss and succumbed (Fig. 3C and D; Fig. S3C and D). Conversely, all of the S-2P:ISA720-immunized mice maintained normal body weight and survived (Fig. 3C and D; Fig. S3C and D). Quantitative analysis demonstrated that S-2P:ISA720 vaccination significantly reduced viral loads in lungs, brains, and nasal turbinates (Fig. 3E and F). Correspondingly, cytokine mRNA profiling in brain tissues revealed marked inflammatory responses in PBS-control and S-2P:Al groups, but minimal response in S-2P:ISA720-immunized mice (Fig. 3G and H).

In summary, ISA720-adjuvanted S-2P not only broadened the neutralization capacity against diverse SARS-CoV-2 variants but, more significantly, conferred robust cross-protection against antigenically distinct VOCs.

Booster immunization, whether homologous or heterologous, is a widely adopted strategy to enhance vaccine efficacy against rapidly evolving SARS-CoV-2 variants (20). To assess the impact of booster doses on antibody responses elicited by differently adjuvanted spike proteins, we immunized hACE2 mice three times (weeks 0, 3, and 11) with either S-2P:Al or S-2P:ISA720. Serum samples were collected at 2 and 8 weeks after the second and third immunizations (weeks 5, 11, 13, and 19), enabling a longitudinal assessment of antibody response kinetics (Fig. 4A).

Consistent with Results 1 and 2, S-2P:ISA720 induced significantly higher RBD-specific IgG and neutralizing antibody titers against the ancestral strain than S-2P:Al at all indicated timepoints (Fig. 4B). After the second and third immunizations, anti-RBD IgG antibody titers in S-2P:Al-immunized mice showed faster decay kinetics compared to the S-2P:ISA720 group, indicating superior durability of the S-2P:ISA720-induced humoral response. Notably, while the S-2P:Al group showed a significant increase in IgG titers after the third dose, the S-2P:ISA720 group reached peak IgG levels after two doses, with no further increase after the third dose (Fig. 4B). Similarly, the third dose of S-2P:Al markedly enhanced neutralizing antibody titers against the ancestral strain, whereas the third dose S-2P:ISA720 showed no significant improvement in serum neutralization post-booster (Fig. 4C). These results suggest that antibody responses to S-2P:ISA720 vaccination reached a ceiling effect after prime-boost immunization. Moreover, the S-2P:Al group showed a modest (1.6-fold) increase in IgG antibodies against the Omicron BA.2 RBD after the third dose, while S-2P:ISA720 failed to enhance RBD-specific IgG titers post-booster (Fig. 4D). Neutralization of Omicron BA.2 improved slightly with the third S-2P:Al dose, with the number of sera exhibiting neutralizing titers > 100 increasing from 2 to 3. In contrast, sera from S-2P:ISA720-immunized mice displayed stronger neutralization against Omicron BA.2 after the second dose, but the third dose provided no additional benefit (Fig. 4E).

To evaluate long-term immunity, we compared antibody persistence and protective efficacy between the S-2P:Al and S-2P:ISA720 in hACE2 mice immunized at weeks 0 and 3. The S-2P:Al group exhibited a 6.3-fold decline in anti-RBD IgG titers between weeks 5 and 20, compared to only a 2.4-fold decline in the S-2P:ISA720 group (Fig. S4A). Even worse, the S-2P:Al group showed a more than 7.6-fold decline in neutralizing titers, whereas the S-2P:ISA720 group exhibited only a 2.7-fold decline (Fig. S4B). Following the challenge at week 21 with ancestral SARS-CoV-2, all S-2P:Al-immunized mice ultimately succumbed (albeit with delayed mortality versus PBS controls), whereas the S-2P:ISA720 group achieved complete protection (Fig. 4G and H).

The elderly population exhibits heightened susceptibility to severe COVID-19 outcomes (17, 18). To evaluate vaccine protection in aging populations, we compared immune responses in young (3-month-old) and aged (18-month-old) hACE2 mice immunized with either S-2P:Al or S-2P:ISA720. Serum antibody responses were measured at 2 weeks after prime and booster vaccination. Following prime immunization, aged mice exhibited significantly lower anti-RBD IgG levels compared to young adults (1.7-fold lower for S-2P:Al and 1.9-fold lower for S-2P:ISA720). This age-dependent reduction persisted after booster immunization, with aged mice showing 2.1-fold (S-2P:Al) and 1.8-fold (S-2P:ISA720) lower antibody levels than their younger counterparts (Fig. 5A). Neutralizing antibody levels after the booster immunization exhibited a similar pattern, being 3.6-fold lower in aged S-2P:Al recipients and 2.1-fold lower in S-2P:ISA720 recipients compared to young adults (Fig. 5B). Consistent with observations in young mice, S-2P:ISA720 similarly elicited a Th1-skewed immune response in aged animals, as demonstrated by significantly elevated IgG2b:IgG1 ratios (Fig. 5A and B). Following SARS-CoV-2 challenge, control and S-2P:Al-immunized aged mice showed rapid weight loss and 100% mortality. In contrast, aged mice immunized with S-2P:ISA720 demonstrated prolonged survival and a 40% survival rate (Fig. 5C and D). These findings demonstrate that while both vaccines displayed substantially reduced efficacy in aged mice, S-2P:ISA720 provided significantly better protection compared with S-2P:Al.

Interestingly, although some aged mice immunized with S-2P:ISA720 developed higher anti-RBD IgG titers than young survivors receiving S-2P:Al immunization, they nevertheless ultimately succumbed to infection (Fig. 5E). To investigate this apparent disconnect between antibody levels and protection in aged animals, we performed passive immunization experiments. Naïve young and aged hACE2 mice received 100 µL of pooled sera from PBS-, S-2P:Al-, or S-2P:ISA720-immunized young donors prior to SARS-CoV-2 challenge. The mice were sacrificed at 4 days post-infection for analysis of brain viral loads and cytokine profiles. Compared to serum from S-2P:Al-vaccinated mice, serum from S-2P:ISA720 vaccinated mice demonstrated superior effect in reducing viral load in both aged and young mouse brains (Fig. 5F). Notably, the antiviral efficacy of serum from either S-2P:Al- or S-2P:ISA720-vaccinated mice was significantly attenuated in aged recipients compared to young recipients, indicating age-dependent impairment of serum-mediated protection. We also observed that young naïve mice exhibited significantly stronger neuroinflammatory responses post-challenge than aged mice, with 14.2-fold higher CXCL10 (P = 0.043), 10.3-fold higher CXCL2 (P = 0.047), 2.9-fold higher TNF-α (P = 0.049), and 9.9-fold elevated IL-6 (P = 0.019) levels in brains (Fig. 5G and H). IFN-γ levels showed comparable changes (P = 0.126) between age groups. This suggests that the attenuated innate immune response in aged animals may contribute to impaired antibody-mediated protection, ultimately leading to poorer outcomes despite comparable or higher antibody titers.