AddaVax, AddaS03, and Alum Effectively Enhance Cross-Reactive and Cross-Neutralizing Antibody Responses Against SARS-CoV-2 Induced by the Inactivated NDV-HXP-S Vaccine in Mice

TL;DR

This study shows that adding specific adjuvants to a SARS-CoV-2 vaccine boosts antibody responses in mice, potentially improving vaccine effectiveness and reducing costs.

Contribution

The study identifies AddaVax, AddaS03, and Alum as effective adjuvants for enhancing immune responses to a SARS-CoV-2 vaccine in mice.

Findings

AddaVax, AddaS03, and Alum induced strong IgG responses to the ancestral spike protein.

These adjuvants boosted cross-reactive antibodies against S1 and S2 subunits and elicited high cross-neutralizing titers.

Abstract

Background/Objectives: We previously developed a low-cost vaccine based on Newcastle disease virus expressing a stabilized pre-fusion spike of SARS-CoV-2 (NDV-HXP-S), which has shown safety and immunogenicity in pre-clinical and clinical studies. Due to the emergence of immune-evasive variants and the need to protect vulnerable populations, we evaluated adjuvanted NDV-HXP-S vaccine formulations to enhance and broaden immune responses. Methods: We tested the antibody responses of mice immunized intramuscularly with an inactivated NDV-HXP-S vaccine adjuvanted with AddaVax, AddaS03, Alhydrogel adjuvant 2% (Alum), or Quil-A. Results: AddaVax, AddaS03, and Alum induced the strongest IgG responses to the ancestral spike protein, boosted cross-reactive antibodies against both S1 and S2 subunits, and elicited high cross-neutralizing titers. Conclusions: The present results highlight the critical role of adjuvant selection in shaping both the magnitude and breadth of the immune response induced by the NDV-HXP-S vaccine. AddaVax, AddaS03, and Alum stand out as promising candidates to enhance NDV-HXP-S vaccine immunogenicity, with potential applications in booster strategies against SARS-CoV-2, enabling dose sparing and reducing costs.

Figures

Full text

1. Introduction

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) emerged at the end of 2019 and rapidly spread worldwide, causing a pandemic with great health, social, and economic impacts [1,2,3]. In response to this challenge, multiple vaccines were developed targeting the viral spike glycoprotein located on the surface of the virus, which mediates entry into host cells. The spike consists of two subunits, S1 and S2, separated by a polybasic S1/S2 cleavage site [4,5] (Figure 1). Within S1, two major antigenic regions are critical for viral infection and vaccine immune recognition: the receptor-binding domain (RBD) and the N-terminal domain (NTD). RBD binds to the cellular angiotensin-converting enzyme 2 (ACE2) receptor during viral entry [5,6] and is the primary target for neutralizing antibodies that prevent the virus attachment to the host cell [7,8,9,10], whereas NTD contributes to spike protein stability and is also an important target of neutralizing antibodies [5,7,9,11] (Figure 1).

We previously developed a SARS-CoV-2 vaccine using Newcastle disease virus (NDV) as a vector to express a stabilized pre-fusion spike protein. Stabilization was achieved by removing the S1/S2 cleavage site and introducing six proline substitutions in S2 (F812P, A892P, A899P, A942P, K986P, and V987P) (Figure 1) [12]. This hexapro-spike (HXP-S) is anchored in the membrane of the NDV virion by replacing the SARS-CoV-2 spike transmembrane domain and cytoplasmic tail with those from the NDV fusion (F) protein. Finally, the transgene is inserted between the phosphoprotein and the matrix genes of NDV [12]. Pre-clinical studies have shown that the NDV-HXP-S vaccine, either live or inactivated, elicits robust humoral and cellular immunity and provides protection in various animal models [12,13,14,15,16,17,18]. Clinical trials conducted in the US (NCT05181709), Thailand (NCT04764422), Vietnam (NCT04830800), Brazil (NCT04993209), and Mexico (NCT04871737) have demonstrated that this vaccine is safe and highly immunogenic in humans [19,20,21,22], leading to emergency approval of an inactivated NDV-HXP-S vaccine in Thailand (HXP-GPOVac, produced by Thailand’s government Pharmaceutical Organization (GPO)) [23] and a live NDV-HXP-S vaccine in Mexico (Patria, developed by Avimex) [24].

Although effective, the NDV-HXP-S vaccine still must be adapted to emerging SARS-CoV-2 variants of concern (VOCs), which are capable of escaping vaccine-elicited immune responses. Immune escape by VOCs is largely driven by mutations in the immunodominant RBD of the spike, which have led to substantial reductions in neutralization by both vaccine-induced and infection-induced antibodies [25,26,27,28,29].

Given the strong immune pressure on the RBD, immunization strategies that redirect antibody responses toward epitopes less affected by RBD mutations are needed to achieve durable and broad protection. In this context, the NTD, a target of potent neutralizing antibodies [30,31,32], and the more conserved S2 subunit, which elicits broadly cross-reactive antibodies across divergent variants [33,34], represent promising targets for enhancing cross-protective immunity.

To broaden vaccine protection, we previously demonstrated that a multivalent vaccine formulation incorporating NDV-HXP-S viruses expressing the spike from the ancestral, Beta, and Delta variants increased cross-neutralization against phylogenetically distant VOCs [13,17], thereby improving the breadth of protection conferred by the NDV-HXP-S vaccine. Another strategy to improve cross-protection is the use of adjuvants. However, it remains unclear which adjuvant best enhances and broadens the immune response elicited by the NDV-HXP-S vaccine. In this study, we evaluated the antibody responses elicited by an inactivated NDV-HXP-S vaccine administered with different adjuvants.

We evaluated two oil-in-water nanoemulsions: AddaVax, which has a similar formulation to that of MF59, and AddaS03, with a composition resembling the adjuvant system 03 (AS03). MF59 and AS03 have both been employed in Europe as adjuvants in influenza vaccines [35,36,37,38,39,40]. Additionally, we tested Alhydrogel adjuvant 2% (Alum), an aluminum hydroxide wet gel suspension widely used in human vaccines [41,42], and Quil-A, a plant-derived extract from the tree Quillaja saponaria, composed of a mixture of over 100 different saponins [43,44,45]. Currently, Quil-A is restricted to veterinary vaccines [46,47], but a purified fraction from this adjuvant, designated as QS-21, is a component of commercial vaccines against shingles, malaria, and SARS-CoV-2 [48].

Vaccine adjuvants represent a powerful strategy to modulate both the magnitude and epitope specificity of antibody responses. Oil-in-water emulsions such as MF59 and AS03 (and their analogs AddaVax and AddaS03) enhance immune responses primarily by promoting local innate immune activation, increasing antigen uptake, and facilitating recruitment and differentiation of antigen-presenting cells at the injection site and draining lymph nodes [35,40,49]. Alum acts through the formation of an antigen depot, enabling sustained release, inflammasome activation, and has been associated with strong Th2-biased humoral responses [50,51,52]. In contrast, saponin-based adjuvants such as Quil-A and its purified derivative QS-21 stimulate immunity through inflammasome activation and enhanced cross-presentation, often favoring a sustained Th1-type activation [47,53]. However, the extent to which these adjuvant classes differentially shape epitope targeting and cross-variant antibody breadth in the context of NDV-based SARS-CoV-2 vaccines is unknown.

We found that AddaVax, AddaS03, and Alum significantly enhanced cross-reactive IgG responses to both the S1 and S2 subunits of different SARS-CoV-2 variants when combined with the inactivated NDV-HXP-S vaccine, with Alum being the most effective at boosting antibodies against S2 and the NTD. Importantly, AddaVax, AddaS03, and Alum increased cross-neutralizing antibody titers against diverse VOCs. In contrast, Quil-A also enhanced some antibody responses against the spike, but to a lesser extent.

2. Materials and Methods

2.1. Cells, Proteins, and Adjuvants

Baby hamster kidney cells stably expressing human ACE2 (BHK-hACE2) were cultured at 37 °C with 5% CO_2_ in Dulbecco’s modified Eagle’s medium (DMEM) (Thermo Fisher Scientific, Waltham, MA, USA; Cat# 11995-065) supplemented with 10% (v/v) of heat inactivated fetal bovine serum (FBS) (Gibco, Thermo Fisher Scientific, Waltham, MA, USA; Cat# 10437-028), and 100 units/mL of penicillin plus 100 μg/mL of streptomycin (Pen Strep) (Thermo Fisher Scientific, Waltham, MA, USA; Cat# 15140122).

Full-length SARS-CoV-2 ancestral spike and the RBD domains for the ancestral, Beta, Delta, and BA.1 variants were produced as described previously [54,55]. The S1 subunit, S2 subunit, and the NTD proteins of the different variants were obtained from ACROBiosystems (Newark, DE, USA) (Table 1).

AddaVax (Cat# vac-adx), AddaS03 (Cat# vac-as03), Alhydrogel adjuvant 2% (Cat# vac-alu), and Quil-A (Cat# vac-quil) were obtained from InvivoGen (San Diego, CA, USA).

2.2. NDV-HXP-S Virus Propagation

The NDV-HXP-S ancestral virus was previously rescued and characterized [12]. The virus was propagated in 10-day-old specific pathogen-free embryonated chicken eggs (Charles River Laboratories, Wilmington, MA, USA). Eggs were incubated at 37 °C for 72 h and subsequently cooled at 4 °C overnight. Allantoic fluid was harvested and clarified by centrifugation at 2000× g for 10 min at 4 °C using a Sorvall Legend RT Plus Refrigerated Benchtop Centrifuge (Thermo Fisher Scientific, Waltham, MA, USA). Viral replication was confirmed by hemagglutination (HA) assays with turkey red blood cells (LAMPIRE Biological Laboratories, Pipersville, PA, USA). HA-positive allantoic fluids were pooled for virus inactivation and purification.

2.3. Inactivation of NDV-HXP-S Virus

As described previously [16], clarified allantoic fluid was mixed with 0.5 M disodium phosphate at a ratio of 38:1 (v/v). Ice-cold 2% β-propiolactone (BPL) (Sigma-Aldrich, Saint Louis, MO, USA; Cat# P5648) was added dropwise with continuous shaking to a final concentration of 0.05% BPL. The mix was incubated on ice for 30 min and then transferred to 37 °C. After a 2 h incubation, the allantoic fluid was clarified at 2000× g for 30 min at 4 °C using a Sorvall Legend RT Plus Refrigerated Benchtop Centrifuge (Thermo Fisher Scientific, Waltham, MA, USA).

To confirm inactivation, 10-day-old embryonated chicken eggs were inoculated with the inactivated allantoic fluid, incubated at 37 °C for 72 h, and then tested by an HA assay to confirm the absence of viral replication.

2.4. Purification of NDV-HXP-S Virus for Vaccination

The inactivated ancestral NDV-HXP-S virus present in clarified allantoic fluid was purified through a 20% sucrose cushion in 1× phosphate-buffered saline (PBS) (pH 7.4) by ultracentrifugation in a Beckman L7-65 ultracentrifuge at 25,000 rpm for 2 h at 4 °C using a Beckman SW28 rotor (Beckman Coulter, Brea, CA, USA). Supernatants were aspirated, and the resulting pellet was resuspended in 1× PBS (pH 7.4). The total protein concentration of the preparation was determined by the Pierce bicinchoninic acid protein assay kit (Thermo Fisher Scientific, Waltham, MA, USA, Cat# 23227).

2.5. Ethics Statement

Animal experiments were conducted under protocols approved by the Institutional Animal Care and Use Committee (IACUC) of the Icahn School of Medicine at Mount Sinai (PROTO202000098, latest approval date 23 July 2025).

2.6. Mouse Immunization Experiments

Eight-week-old female BALB/c mice (Jackson Laboratories, Bar Harbor, ME, USA) were housed in a temperature-controlled biosafety level 2 facility under a 12 h light/dark cycle. Mice were vaccinated intramuscularly (IM) with 100 μL of purified inactivated NDV-HXP-S vaccine (50 μL per hind leg), with or without adjuvant. Three vaccine doses were tested: 1 μg, 0.1 μg, or 0.01 μg (total protein). All vaccinations followed a prime-boost regimen with a four-week interval. For adjuvanted formulations, AddaVax, AddaS03, or Alum were mixed with each vaccine dose at a 1:1 (v/v) ratio, and 10 μg of Quil-A was mixed with each vaccine dose.

Blood was collected 4 weeks after the prime dose via submandibular vein bleeding. Four weeks after the boost, mice were sedated by intraperitoneal (IP) injection with ~100 mg/kg ketamine hydrochloride (Dechra Veterinary Products, Overland Park, KS, USA) and ~20 mg/kg xylazine (AnaSed Injection, Akorn Animal Health, Lake Forest, IL, USA). Terminal blood was collected by cardiac puncture, followed by euthanasia in accordance with IACUC protocols. Serum was isolated from blood samples by centrifugation at 2800× g for 30 min at 4 °C using an Eppendorf centrifuge 5430R (Eppendorf, Enfield, CT, USA) and stored at −80 °C until processing.

2.7. Enzyme-Linked Immunosorbent Assays (ELISAs)

Immulon 4 HBX clear flat-bottom immuno nonsterile 96-well plates (Thermo Fisher Scientific, Waltham, MA, USA, Cat# 3855) were coated overnight at 4 °C with 50 μL/well of the indicated recombinant protein at a concentration of 2 µg/mL in 1× coating buffer (SeraCare Life Sciences, Milford, MA, USA, Cat# 5150-0014). Plates were washed with 1× PBS (pH 7.4) containing 0.1% (v/v) Tween-20 (PBST) and incubated with blocking solution (3% goat serum, 0.5% nonfat dried milk powder, 96.5% PBST) for 1 h at room temperature (RT). Then, three-fold serial dilutions of individual serum samples prepared in blocking solution were added to the plates and incubated for 2 h at RT. Plates were washed with PBST and incubated with sheep anti-mouse IgG horseradish peroxidase (HRP) secondary antibody (Cytiva, Marlborough, MA, USA, Cat# NA931), which was diluted in blocking solution. After 1 h incubation at RT, plates were washed with PBST and developed using SigmaFast o-phenylenediamine dihydrochloride substrate (Sigma-Aldrich, Saint Louis, MO, USA, Cat# P9187) for 10 min. The reaction was stopped with 3 M hydrochloric acid. Absorbance at 492 nm was measured using a FilterMax F3 multi-mode microplate reader (Molecular Devices, San Jose, CA, USA).

For each ELISA plate, the cutoff for endpoint titer was defined as the mean absorbance of blank wells plus three standard deviations. Endpoint titers were calculated using GraphPad Prism version 10.0.2 for Mac OS X (GraphPad Software, Boston, MA, USA; www.graphpad.com).

2.8. Microneutralization Assay

Replication-competent vesicular stomatitis virus (rcVSV) carrying an enhanced green fluorescent protein (eGFP) reporter gene and expressing the SARS-CoV-2 spike from either the ancestral strain, or important VOCs (Alpha, Beta, Delta, BA.1, BA.5, BQ.1.1, or XBB.1.5) in place of VSV glycoprotein (G protein) were produced as described previously [56]. Flat-bottom 96-well plates (Corning Inc., Corning, NY, USA, Ca# 353072) were seeded with 2 × 10^4^ BHK-hACE2 cells/well and incubated overnight. The next day, pooled serum samples from each group were heat-inactivated at 56 °C for 30 min prior to use. The different viruses were pre-mixed with four-fold serially diluted serum (starting at 1:5) in DMEM supplemented with 10% FBS and 1% Pen Strep, and then incubated for 15 min at RT. The virus-serum mixtures were transferred onto the BHK-hACE2 cell plates. Cells infected in the absence of serum (“no serum” control) were included as a reference. At 12 h post-infection (hpi), plates were scanned using the Nexcelom Celigo S Image Cytometer BFFL 5C-AV (Nexcelom Biosciences, Lawrence, MA, USA), and infected eGFP-positive cells were quantified using Celigo software (version 5.5.0.0). EGFP counts for each serum-treated condition were normalized to the “no serum” control, which was set as 100% infection. The inhibitory dilution at which 50% neutralization is achieved (ID_50_) was determined by analyzing the normalized values using GraphPad Prism (version 10.0.2 for Mac OS X; GraphPad Software, Boston, MA, USA; www.graphpad.com) with the “log(inhibitor) vs. normalized response-variable slope” model.

2.9. Statistics

Statistical significance between the unadjuvanted group and each adjuvanted group was assessed using the Mann–Whitney–Wilcoxon test in GraphPad Prism (version 10.0.2 for Mac OS X; GraphPad Software, Boston, MA, USA; www.graphpad.com). All statistical tests were two-sided, and a p-value ≤ 0.05 was considered statistically significant.

3. Results

3.1. AddaVax, AddaS03, and Alum Significantly Enhance IgG Titers Against the Ancestral SARS-CoV-2 Spike

To assess the magnitude and breadth of antibody responses elicited by the different adjuvanted inactivated NDV-HXP-S vaccines, we performed a dose-escalation study with or without adjuvants. An NDV-HXP-S vaccine expressing the full-length ancestral SARS-CoV-2 spike was used as a reference to evaluate cross-reactive immune responses.

The inactivated purified NDV-HXP-S vaccine was administered IM to BALB/c mice in a prime-boost regimen with a 4-week interval (Figure 2A). As shown in Figure 2B, mice were immunized with 1, 0.1, or 0.01 µg of vaccine, either alone (groups 1–3) or formulated with AddaVax (groups 4–6), AddaS03 (groups 7–9), Alum (groups 10–12), or Quil-A (groups 13–15). An unvaccinated control group received PBS only (group 16). Mice were bled 4 weeks post-prime and terminally bled 4 weeks post-boost (Figure 2A). IgG responses against the full-length ancestral spike protein were measured by ELISA.

After a single immunization, the unadjuvanted vaccine induced a geometric mean IgG endpoint titer of ~2 × 10^4^ at 1 µg. The reduction in unadjuvanted vaccine to a 0.1 µg dose modestly reduced antibody titers, while a decrease to a 0.01 µg dose resulted in a significant reduction (Figure 2C, solid white bars). Compared to the unadjuvanted vaccine, AddaVax did not enhance IgG responses at 1 µg but significantly increased titers at lower doses (Figure 2C, solid blue vs. solid white bars). In contrast, AddaS03 significantly boosted titers at 1 µg but showed no effect at lower doses (Figure 2C, solid red vs. solid white bars). Alum was the most effective adjuvant after priming, substantially increasing IgG titers across all doses tested (Figure 2C, solid green vs. solid white bars).

Following the booster dose, the unadjuvanted vaccine elicited significantly higher IgG titers compared to post-prime levels (Figure 2C solid white bars vs. hatched white bars). After the boost, AddaVax increased titers by ~10-fold across all vaccine doses tested relative to the unadjuvanted vaccine (Figure 2C, hatched blue vs. hatched white bars). AddaS03 (Figure 2C, hatched red bars) and Alum (Figure 2C, hatched green bars) also significantly improved antibody responses at 1 µg and 0.1 µg as compared to the unadjuvanted vaccine, though their effects were moderate at 0.01 µg (Figure 2C).

In contrast, formulations with Quil-A did not further enhance IgG titers at either timepoint (Figure 2C, solid and hatched purple bars).

These results suggest that AddaVax, AddaS03, and Alum are promising adjuvants for improving the IgG response against the ancestral spike protein of SARS-CoV-2 after priming or boosting with the inactivated NDV-HXP-S.

3.2. The Unadjuvanted Inactivated NDV-HXP-S Vaccine Induces a Predominantly RBD-Specific Antibody Response

We next characterized the antibody responses against different spike domains in the absence of adjuvants. Using post-boost sera, we measured IgG titers against S1, S2, the NTD, and the RBD of the ancestral spike protein, as well as the corresponding domains from three phylogenetically different VOCs (Beta, Delta, and BA.1) (Figure 3A).

Reducing the vaccine dose of the unadjuvanted NDV-HXP-S vaccine revealed that S1 and RBD are the most dominant domains when antibody responses to the ancestral spike are measured (Figure 3B). Similar results were observed when the immune responses against different domains of VOCs were measured (Figure 3C–E).

Together, these findings indicate that the unadjuvanted NDV-HXP-S vaccine primarily induces antibodies targeting S1, particularly the RBD, while responses to S2 and NTD are subdominant.

3.3. Adjuvant Selection Differentially Influenced Domain-Specific Antibody Responses Induced by the Inactivated NDV-HXP-S Vaccine

Next, we evaluated how adjuvants influenced antibody responses against different domains of the spike. At the highest vaccine dose (1 µg), AddaVax, AddaS03, and Alum significantly increased S1-specific titers across all variants, whereas Quil-A was less effective (Figure 4A). At lower vaccine doses, the boosting effect of AddaS03 and Alum on S1 responses was modest for most viruses tested. Notably, only AddaVax consistently induced a significant enhancement of cross-reactive S1 antibodies against all three VOCs (Beta, Delta, and BA.1) at low doses. In contrast, Quil-A tended to reduce titers towards S1 in most conditions as the vaccine dose decreased (Figure 4A).

For the S2 subunit, Alum was the only adjuvant to consistently induce a strong boost in antibody responses across all tested conditions. However, AddaVax, AddaS03, and Quil-A showed variable effects, significantly boosting antibody responses under select conditions depending on dose and the VOC tested (Figure 4B).

For the RBD, AddaS03 was most effective at 1 µg, while AddaVax performed best at 0.01 µg. On the other hand, Alum and Quil-A had mostly moderate effects on the RBD response (Figure 4C). For the NTD, Alum was the most consistent enhancer across all tested conditions, while AddaS03 and Quil-A boosted responses only at doses ≥ 0.1 µg. AddaVax effectively enhanced NTD responses at the highest vaccine dose (1 µg), but at lower doses consistently boosted antibody titers only against the Delta and BA.1 NTDs (Figure 4D).

In summary, these results indicate that AddaVax, AddaS03, and Alum were the most effective adjuvants for boosting IgG responses against different spike domains of SARS-CoV-2, with their potency varying depending on both the domain and the vaccine dose. Interestingly, Alum was the most effective for enhancing S2- and NTD-specific antibody responses.

3.4. AddaVax, AddaS03, and Alum Are the Most Effective Adjuvants for Boosting Cross-Neutralizing Antibodies

Neutralizing antibodies are a key correlate of protection against SARS-CoV-2. We therefore evaluated the neutralizing activity of vaccine-induced antibodies in post-boost sera using an rcVSV engineered to express eGFP and the SARS-CoV-2 spike (rcVSV-eGFP-CoV2-S). For this assay, we tested viruses carrying the ancestral spike or those from major variants of concern (Alpha, Beta, Delta, BA.1, BA.5, BQ.1.1, and XBB.1.5). Viruses were incubated with pooled, heat-inactivated sera, and neutralization was measured in BHK-hACE2 cells. At 12 h post-infection, ID_50_ values were determined. Pooled sera from each group were selected for this assay to evaluate group-level neutralization estimates as an initial screening of the adjuvant candidates.

At a 1 µg dose, the unadjuvanted vaccine induced comparable neutralization titers against the ancestral, Alpha, Beta, and Delta spikes, but titers against BA.1 and BA.5 were ~10-fold lower (Figure 5). As the dose decreased, neutralization elicited by the unadjuvanted vaccine declined, and at 0.01 µg activity, it was largely restricted to the ancestral and Alpha variants. Cross-neutralization of BQ.1.1 and XBB.1.5 was undetectable at all doses (Figure 5).

Adjuvants markedly improved neutralizing responses. At 1 and 0.1 µg, AddaVax, AddaS03, Alum, and Quil-A significantly enhanced neutralization against most variants. At the lowest vaccine dose (0.01 µg), only AddaVax, AddaS03, and Alum sustained enhanced neutralization across several variants, while Quil-A had little effect. Similar to the unadjuvanted vaccine, cross-neutralization of BQ.1.1 and XBB.1.5 remained minimal across all adjuvant conditions (Figure 5).

In summary, AddaVax, AddaS03, and Alum were the most effective adjuvants for enhancing cross-neutralizing antibodies, although they did not broaden responses to all immune-evasive VOCs tested.

4. Discussion

In this study, we performed an initial systematic evaluation of the breadth of antibody responses elicited by the inactivated NDV-HXP-S vaccine formulated with various adjuvants to identify candidates that best broaden antibody responses. We specifically assessed responses to an NDV-HXP-S vaccine expressing the ancestral SARS-CoV-2 spike against antigenically distinct VOCs.

Consistent with previous reports, a prime-boost regimen with the unadjuvanted vaccine induced high serum antibody levels in mice, even at a low antigen dose of 0.01 µg, confirming the strong intrinsic immunogenicity of the NDV-HXP-S vaccine. The antibodies elicited by the NDV-HXP-S vaccine were primarily targeted to the RBD located in the S1 subunit. Cross-reactive responses to Beta, Delta, and BA.1 domains were also dominated by RBD.

This aligns with prior work in humans, where inactivated NDV-HXP-S vaccination elicited a strongly RBD-focused antibody profile with limited S2 reactivity [57]. This bias has been suggested to result from limited accessibility to NTD and S2 epitopes on HXP-S, either due to (1) steric hindrance from the densely packed NDV hemagglutinin–neuraminidase (HN) and F proteins, which favor B cell receptor access to the RBD located at the top of the HXP-S trimer, or (2) the pre-fusion-stabilized HXP-S design masking key S2 and NTD epitopes that are only exposed in the post-fusion conformation. A combination of these two mechanisms is also plausible [57].

Adjuvant formulation revealed that AddaVax, AddaS03, and Alum were the most effective compounds, each with specific strengths depending on vaccine dose and epitope specificity. For many conditions, the enhancing effects of the adjuvants were moderate, likely due to the high intrinsic immunogenicity of NDV-HXP-S. However, we found that AddaVax markedly increased IgG titers against the full-length spike at all doses after boosting, with particularly strong effects at 0.01 µg. AddaS03 and Alum boosted responses mainly at higher doses, suggesting a particular value for high-dose formulations such as those used in populations with weakened immune systems (e.g., older adults).

All three adjuvants, AddaVax, AddaS03, and Alum, enhanced antibody responses to the different spike domains of the ancestral, Beta, Delta, and BA.1 variants. Interestingly, Alum most strongly boosted S2- and NTD- specific antibodies, an attribute that is important for sustaining protection against emerging variants harboring extensive mutations in the RBD. These differences may result from the distinct adjuvant-antigen interactions, as well as the different innate immune activation and antigen presentation pathways promoted by each adjuvant [37,40,50]. This distinction suggests that different adjuvants may skew epitope targeting within the HXP-S protein and highlights the importance of adjuvant selection to improve the antibody response elicited by the vaccine.

In addition, neutralization assays further support the advantages of AddaVax, AddaS03, and Alum as promising components for adjuvanted formulations. Notably, these three adjuvants maintained neutralizing activity even at the lowest vaccine dose (0.01 µg) to a subset of phylogenetically different strains like Omicron BA.1, highlighting their capacity to extend protective responses under conditions of limited antigen availability. However, given that neutralization activity was not observed with highly immune-evasive VOCs such as BQ.1.1 and XBB.1.5, the combination of these adjuvants to NDV-HXP-S vaccines expressing the circulating VOCs [14,18] or multivalent formulations [13,17] might be needed to further extend the breadth of protection to more recent VOCs.

In contrast, Quil-A was less promising as a candidate adjuvant for the inactivated NDV-HXP-S vaccine. Quil-A improved binding and neutralization under select conditions but generally was less effective than AddaVax, AddaS03, and Alum. It is important to note that Quil-A was administered at the manufacturer-recommended mouse dose (≤15 µg/dose), which did not correspond to the composition or dose of human-use saponin-based adjuvants such as QS-21. Our findings, therefore, reflect mouse-optimized formulations only and are not intended to imply direct comparability to human vaccine adjuvant systems.

Study limitations include the use of pooled sera for neutralization, which obscures individual variability and the ability to perform formal statistical inference, the short follow-up period preventing assessment of humoral response durability, and the lack of cellular immunity analysis. Future work should address these limitations, assess adjuvant effects in multivalent NDV-HXP-S formulations and updated strains, and evaluate responses in pre-immune models. Additionally, the quantification of antibody quality measures such as affinity/dissociation kinetics, IgG subclasses, or Fc-mediated effector activity of the antibodies elicited by the adjuvanted vaccines would strengthen mechanistic interpretation and the relevance of adjuvant differences.

Taken together, our findings provide new guidance for future NDV-HXP-S vaccine design aimed at improving protection against continuously evolving VOCs. First, the strong RBD-focused immunodominance observed with the ancestral HXP-S highlights the need for adjuvant formulations to redirect antibody responses toward subdominant but more conserved epitopes, including the NTD and S2 subunit. Moreover, the preferential boosting of NTD- and S2-specific antibodies by Alum suggests that specific adjuvants can modulate epitope targeting of NDV-based vaccines and may complement next-generation HXP-S designs that improve accessibility of these regions. Second, our neutralization data against immune-evasive VOCs indicate that adjuvants alone are not sufficient to overcome extensive antigenic drift, supporting the incorporation of variant-specific spikes or multivalent antigen formulations expressing spikes from antigenically distinct lineages. Combining optimized adjuvants with multivalent or updated NDV-HXP-S antigens may provide synergistic benefits by simultaneously enhancing antibody magnitude, breadth, and epitope diversity, thereby increasing cross-protection against future variants. These insights emphasize that integrating antigen selection, multivalency, and adjuvant choice will be critical for sustaining the effectiveness of NDV-HXP-S–based vaccines in the context of ongoing SARS-CoV-2 evolution.

5. Conclusions

In summary, NDV-HXP-S vaccination induces strong RBD-focused antibody responses that can be broadened with selected adjuvants. Among those tested, AddaVax, AddaS03, and Alum markedly enhanced immunogenicity, with Alum most effective at boosting S2-specific responses associated with broad cross-neutralization. Quil-A was less beneficial or, in some instances, detrimental. These findings support further evaluation of Alum, AddaVax, and AddaS03 in next-generation vaccine formulations. Finally, future studies in humans will be critical to establish the clinical potential of these adjuvants and guide rational adjuvant selection for adjuvanted NDV-HXP-S vaccines with broader protection.