Immunogenicity and protective efﬁcacy of SARS-CoV-2 recombinant S-protein vaccine S-268019-b in cynomolgus monkeys

The vaccine S-268019-b is a severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) spike (S)- protein vaccine consisting of full-length recombinant SARS-CoV-2 S-protein (S-910823) as antigen, mixed with the squalene-based adjuvant A-910823. The current study evaluated the immunogenicity of S- 268019-b using various doses of S-910823 and its vaccine efﬁcacy against SARS-CoV-2 challenge in cynomolgus monkeys. The different doses of S-910823 combined with A-910823 were intramuscularly administered twice at a 3-week interval. Two weeks after the second dosing, dose-dependent humoral immune responses were observed with neutralizing antibody titers being comparable to that of human convalescent plasma. Pseudoviruses harboring S proteins from Beta and Gamma SARS-CoV-2 variants displayed approximately 3- to 4-fold reduced sensitivity to neutralizing antibodies induced after two vac- cine doses compared with that against ancestral viruses, whereas neutralizing antibody titers were reduced >14-fold against the Omicron variant. Cellular immunity was also induced with a relative Th1 polarized response. No adverse clinical signs or weight loss associated with the vaccine were observed, suggesting safety of the vaccine in cynomolgus monkeys. Immunization with 10 l g of S-910823 with A-910823 demonstrated protective efﬁcacy against SARS-CoV-2 challenge according to genomic and subgenomic viral RNA transcript levels in nasopharyngeal, throat, and rectal swab specimens. Pathological analysis revealed no detectable vaccine-dependent enhancement of disease in the lungs of challenged vaccinated monkeys. The current ﬁndings provide fundamental information regarding vaccine doses for human trials and support the development of S-268019-b as a safe and effective vaccine for controlling the current pandemic, as well as general protection against SARS-CoV-2 moving forward.


Introduction
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the etiological agent of coronavirus disease 2019 (COVID- 19), was first identified in Wuhan, China in December 2019. The World Health Organization (WHO) subsequently declared the outbreak a Public Health Emergency of International Concern on January 30, 2020 and a pandemic on March 11, 2020. COVID-19 remains a global health crisis, causing a critical need for effective intervention and prevention of SARS-CoV-2 infections [1]. This demonstrates the critical need for effective interventions and prevention strategies in order to control and resolve this global pandemic.
The development of efficacious vaccines against SARS-CoV-2 is the best approach for controlling its spread and curbing the pandemic [2]. As reviewed by Harrison et al. [3], SARS-CoV-2 is a member of the Coronaviridae family with its virion containing four structural proteins, including the spike (S) protein. The S protein forms a trimer anchored to the viral membrane and directly contacts the cellular receptor angiotensin-converting enzyme 2 (ACE2) on the host cell through its receptor-binding domain (RBD) [2,4]. As neutralizing epitopes have been identified not only on RBD, but also on the N-terminal domain (NTD), full-length S protein is consider a preferred target antigen of the immune system [5,6]. The characteristics and biological function of the S protein makes it an obvious vaccine target. Not surprisingly, S protein-based vaccines are under active development and are some of the most widely approved vaccines for COVID-19. As of April 1, 2022, WHO reports there are 153 COVID-19 vaccines in clinical development and 196 vaccines in pre-clinical development, with 51 protein subunit vacccine candidates being in clinical phase [7].
We are actively developing S-268019-b, an adjuvanted vaccine for COVID-19 consisting of antigen S-910823 mixed with the squalene-based, oil-in-water emulsion adjuvant A-910823. S-910823 is a recombinant full-length SARS-CoV-2 S protein based on amino acid sequenes of the Wuhan-Hu-1 isolate (Genbank MN908947) and has mutations in the furin cleavage site to inhibit S-protein cleavage into S1 and S2 subunits, as well as two substituted proline residues to improve prefusion conformaton stability of the S-protein trimers [8][9][10]. Although most relevant cell lines for insect cell-based recombinant protein expression are known to be persistently infected with various adventitious viruses, the S-910823 protein is generated using a baculovirus expression system (BEVS) with rhabdovirus-free insect cells. The use of BEVS technology is a feasible platform to manufacture recombinant antigens and is currently employed for production of several commercial vaccines against influenza and human papilloma viruses [11,12]. Previous studies suggest the balance between Th1 and Th2 immune responses may correlate with potential risks of vaccine-associated disease enhancement (VDE) [13][14][15][16]. Thus, the oil-in-water emulsion-based adjuvant A-910823 was screened through exploratory studies and selected in part based on profiles of immunogenicity and Th1/Th2 balance. The current study evaluated the immunogenicity, protective efficacy, and S-910823 dose dependency of the novel S-268019-b vaccine in cynomolgus monkeys. The findings will provide critical fundamental information on vaccine dosing for human trials and insight toward the goal of developing a safe and efficient vaccine against SARS-CoV-2.

Vaccine (S-268019-b) preparation
The S-268019-b vaccine (Shionogi & Co., Ltd., Osaka, Japan) consists of the recombinant protein S-910823 (Shionogi & Co., Ltd.) and the squalene-based, oil-in-water emulsion adjuvant A-910823 (Shionogi & Co., Ltd.). The recombinant S-910823 protein was expressed using BEVS as previously described [11,12]. Briefly, baculovirus-infected cells were suspended in a buffer containing a nonionic detergent in order to extract the S-910823 protein. The cell extract was centrifugated and the supernatant filtered using a depth filter. The S-910823 protein was purified by affinity chromatography and hydrophobic interaction chromatography. Host cell DNA and baculoviruses were removed by Q membrane chromatography. Tangential flow filtration was used for protein concentration and buffer exchange. The S-910823 protein was prepared at 0, 4, 10, 20, 40, and 100 lg/mL in vehicle consisting of 0.02% polysorbate 20 (PS20) buffer (15 mM phosphate buffer, 150 mM sodium chloride, 0.02% w/v PS20, pH 7.5).

Vaccination of cynomolgus monkeys
Twenty-eight female Cambodian cynomolgus monkeys (Macaca fascicularis), 3 to 7 years of age and weighing 2.1 to 8.0 kg at the begining of a 3-d acclimation period, were obtained from the stock colony of Shin Nippon Biomedical Laboratories, Ltd. Drug Safety Research Laboratories (Kagoshima, Japan). The overall study design is shown in Fig. 1A, with full details provided in the Supplementary Material. Briefly, the monkeys were randomly assigned to seven experimental groups (n = 4/group). Vaccines were administered intramuscularly in a volume of 500 lL. The day prior to the first dosing was day À1 with the first day of dosing being day 1. The second vaccine dose was administered 3 weeks later (day 22). According to the experimental group, the animals received either vehicle (PS20 buffer) only, S-910823 (1, 2.5, 5, 10, or 25 lg) in vehicle plus A-910823 (S+A-910823), or 10 lg S-910823 in vehicle without A-910823 (S-910823-only) for each vaccination. The animals were observed for 5 weeks (through day 36). The dosing schedule and route were in accordance with the intended clinical application. Blood samples were drawn from the femoral veins at the volumes and time points specified for the various analyses. Sample preparation is described in the Supplementary Material.

Anti-S and anti-RBD IgG analysis
Antibody titers were measured using enzyme-linked immunosorbent assays (ELISA) as described in the Supplementary Material. Sera collected on days À1, 21, and 36 were evaluated for anti-S-protein and anti-RBD IgG. All samples were analyzed in duplicate. Absorbance values of A 405 minus A 490 were calculated and the mean and coefficient of variation (CV) of absorbance of the duplicate samples were determined. The mean value was regarded as the analysis value of each sample. The highest dilution factor with an analysis value !the cut point was considered the antibody titer.

Neutralizing antibody titer testing
Neutralizing antibody levels in the monkey sera collected at days À1, 21, and 36 were analyzed against live SARS-CoV-2 WK-521 and the Omicron strain using VeroE6/TMPRSS2 cells as described in the Supplementary Material. Briefly, 50 lL of each serum sample was mixed with an equal volume of WK-521 or Omicron virus suspension, and each 100 lL mixture was then added to individual culture-plate wells. The plates were incubated for approximately 1 h at room temperature for neutralization. After incubation, 3 Â 10 4 VeroE6/TMPRSS2 cells in 100 lL of assay medium were added to each well (100 and 10,000 TCID 50 /well for WK-521 and Omicron, respectively). The plates were then incubated at 37°C with 5% CO 2 for 3 days. The assay was performed in duplicate for each sample. After the 3 day incubation, living cells were examined using CellTiter-Glo 2.0 (G9243; Promega, Madison, WI, USA) according to the manufacturer's instructions. WHO International Reference Panel of human pooled convalescent plasma, High [20/150: 1,473 international unit (IU)/mL)] and Negative [20/142], were used as positive and negative controls, respectively [19]. The WHO International Reference Panel High [20/150] is prepared from UK patients that were mainly infected with SARS-CoV- 2 Wuhan strain in June 2020 [20]. The International Standard was determined by neutralizing assays using live viruses and pseudoviruses of the Wuhan-Hu-1 isolate (Genbank accession number MN908947). The neutralizing antibody titers in the serum samples were calculated as the reciprocal of the highest dilution at which viable cell percentage was !50%. Geometric mean titer (GMT) of the serum samples was also calculated for each group. Details regarding the preparation, titration, and use of the lentivirusbased SARS-CoV-2 pseudovirus variants are provided in the Supplementary Material.

Preparation of peripheral blood mononuclear cells (PBMCs)
Blood was collected (3 mL) via the femoral vein from all animals on days À1, 21, and 36 using a syringe containing heparin sodium. The PBMCs were isolated using Ficoll-Paque Plus (GE Healthcare Life Sciences, USA) and a LeucoSep tube (Grainer Bio-One, Frickenhausen, Germany). The number of cells were counted using trypan blue staining and their density adjusted to 2.0 Â 10 6 cells/mL with Cellbanker 1 (Nippon Zenyaku Kogyo Co., Fukushima, Japan). The cells were prepared as 1 mL aliquots and stored at À70°C until use.

Analysis of cytokine production from PBMCs
Frozen PBMCs samples from individual cynomolgus monkeys were thawed and prepared as 1 Â 10 6 living cells/mL in RPMI 1640 medium supplemented with L-glutamine, fetal calf serum, penicillin-streptomycin, HEPES, sodium pyruvate, MEM Non-Essential Amino Acids Solution, GlutaMAX TM Supplement, and 2mercaptoethanol. The PBMCs were cultured with or without SARS -CoV 2 S-protein peptides and cytokine production from the cells were detected using a Monkey IFN-c/IL-4/IL-5 FluoroSpotFLEX Kit (Mabtech AB, Nacka Strand, Sweden). The S-protein peptides used were PepTivator Ò SARS-CoV-2 Prot_S, PepTivator Ò SARS-CoV-2 Prot_S1, and PepTivator Ò SARS-CoV-2 Prot_S+ (Miltenyi Biotec, Bergisch Gladbach, Germany) at 0.15 lg/mL. These peptide cocktails almost completely span the S-protein. The assays were performed in triplicate for each sample. The plates were scanned and imaged using an ImmunoSpot Analyzer with ImmunoSpot 7.0 software (Cellular Technology Ltd., Cleveland, OH, USA). Image analysis was performed using ImmunoSpot 7.0 software (Cellular Technology Ltd.). The number of spots observed were averaged for wells of individual animals under the same conditions (e.g., sampling date, cytokine, with/without antigen peptide). The number of antigen-specific spots for each cytokine at each sampling day was also determined for individual animals. The results are presented as spot-forming cells (SFC)/1 Â 10 6 PBMCs.

Infection of vaccinated cynomolgus monkeys
To monitor body temperature throughout the virus challenge study, the monkeys were anesthetized 3 or 4 weeks before virus infection with a mixture of 50 mg/mL ketamine and 20 mg/mL xylazine (2:1; 0.2 mL/kg) and intraperitoneally implanted with small temperature probes that measure and record temperature (DST micro-T: 8.3 Â 25.4 mm; Star-oddi, Gardabaer, Iceland). The probes were retrieved at necropsy. Eight or 13 weeks after the second immunization, the monkeys were infected under anesthesia with 10 6.5 TCID 50 SARS-CoV-2 in 1.8 mL of culture medium through a combination of intranasal and intratracheal inoculations (1.5 mL intranasally and 0.3 mL intratracheally).

Sample collection
Approximately 7 days before the challenge infection, the monkeys were anesthetized with ketamine and were transferred from the quarantine room to the biosafety level 3 (BSL3) animal facility. On days 0, 1, 4, and 7 post experimental infection, clinical samples, including nasopharyngeal swabs, throat swabs, rectal swabs, and blood, and the weight of the monkeys were collected under anesthesia. Sera were prepared and stored at À80°C until use. Ethylenediaminetetraacetic acid (EDTA) and heparinized blood samples were used immediately for hematologic tests. The monkeys were sacrificed by exsanguination under excess ketamine anesthesia at 7 days post infection and then necropsied. Clinical and tissue samples were collected for viral titration, viral RNA analysis, and histopathological analysis. The tissue samples included the lungs, trachea, hilar lymph nodes, tonsils, cervical lymph nodes, mesenteric lymph nodes, small intestine, and large intestine.

Clinical scores of infected monkeys
The monkeys were observed once daily for clinical signs and scored for dietary intake of food pellets (0-5) and fruits (0-5), drinking water intake (0-5), behavior in front of regular observers, such as standing up, showing interest in activity outside their cage and receivng attention, intimidation, up and down movement (0-5), and stool consistency, including color, stiffness, form, volume, and frequency (0-5).

Measurement of SARS-CoV-2 genomic and subgenomic RNA levels
Total RNA was extracted from swab samples and tissue homogenates using TRIzol LS or TRIzol reagents (Thermo Fisher Scientific) and a Maxwell RSC 48 automated nucleic acid purification system (Promega, Madison, WI). The RNA was then used to quantify the SARS-CoV-2 genome. The copy number of viral genomic RNA and subgenomic RNA transcripts in the samples were estimated by real-time reverse transcription polymerase chain reaction (RT-PCR) using a QuantiTect Probe RT-PCR Kit (Qiagen, Venlo, Netherlands) and LightCycler 480 system (Roche, Basal, Switzerland). Specific details regarding the RT-PCR primers and probes are provided in the Supplementary Material.

Statistical analysis
Tukey's multiple comparison test was performed for analysis of the clinical scores of infected monkeys and the SARS-CoV-2 genome and subgenomic RNA transcript analyses.

Immunogenicity of S-268019-b in cynomolgus monkeys
To assess the immunogenicity of S-268019-b, cynomolgus monkeys were administered S-268019-b twice, at a 3-week interval (Fig. 1A), and anti-spike IgG serum titers before and after vaccinations were measured (Fig. 1B). The titers two weeks after the second vaccination increased dose-proportionally in animals receiving S+A-910823, with 2.5 lg S-910823 inducing near peak titers. Meanwhile, 10 lg S-910823 without A-910823 induced IgG titers approximately 1,000-fold less than those of 10 lg S+A-910823, indicating that A-910823 enhanced the immunogenicity of 10 lg S-910823. Similarly, anti-RBD IgG titers also peaked following administration of 2.5 lg S+A-910823 (Fig. 1C). To evaluate the functional quality of the antigen-specific antibody induced by S-268019-b, we next evaluated the levels of neutralizing antibodies (Fig. 1D). Consistent with the anti-S-protein and anti-RBD IgG levels, a dose-dependent increase in neutralizing antibody against ancestoral SARS-CoV-2 WK-521 was observed. The GMT for 5 lg and 10 lg S+A-910823 demonstrated the highest neutralizing antibody titers of approximately 2,560. This was comparable with that of the titer for the WHO ''High" reference sample of pooled convalescent plasma prepared from UK patients infected with wild-type SARS-CoV-2 [19]. Although the ages of the monkeys ranged from three to seven years, no correlation was seen between immune response and age.
Obvious changes were not observed throughout the study regarding clinical signs, body weight, immunological cell populations, or hematological values (Supplementary Table S1, Supplementary Fig. S1-S4). Transient elevation of C-reactive protein (CRP) was noted on days 2 and 23 post-dosing in the groups adminstered S+A-910823; however, this was not considered to be dose-dependent and all had recovered by day 29 (Supplementary Fig. S2). Although eosinophil counts increased on day 23 after the second dosing in the group administered 1 lg S+A-910823, this change reverted to basal levels by day 36 (Supplementary Fig. S3A). This transient increase in eosinophils was possibly due to temporal production of cytokines and the migration of eosinophils at the injection site. Overall, these results suggested that S+A-910823 vaccination displayed safe profiles for non-human primates.
To evaluate neutralizing antibody titers elicited by S-268019-b against variant SARS-CoV-2 viruses, we employed recombinant lentiviruses pseudotyped with coronavirus S protein derived from D614G, Alpha, Beta, Gamma, Delta, and Omicron strains ( Fig. 2A). Analysis of neutralizing antibody levels of day 36 sera from monkeys (n = 8) twice vaccinated with 10 lg S+A-910823 at a 3- week dosing interval revealed that 50% pseudovirus neutralization titers (pVNT 50 ) against the parent D614G virus, and Alpha and Delta variants were comparable, while those against Beta, Gamma, and Omicron variants were decreased 3-fold to 14-fold ( Fig. 2A). We also evaluated neutralizing antibody titers against live WK-521 and Omicron viruses (Fig. 2B). Consistent with the pVNT 50 results, neutralizing antibody titers against the Omicron variant was decreased approximately 25-fold compared with that against WK-521.

Analysis of PBMC cytokines
To investigate induction of a cellular immune response and the potential effect of S-268019-b vaccination on Th1/Th2 polarity, PBMCs derived from monkeys vaccinated with S+A-910823, S-910823-only, or vehicle were pre-cultured with SARS-CoV-2 Sprotein peptides and the production of Th1 (IFN-c) and Th2 (IL-4, IL-5) cytokines by the PBMCs were then assessed using FluoroSpot assays. Representative images of the IFN-c/IL-4/IL-5 FluoroSpot assay are shown in Fig. 3A, and quantitative analysis results are shown in Fig. 3B-3D. None of the cytokines were detectable in cultures of PBMCs from animals vaccinated with vehicle or S-910823only. However, antigen-specific responses for each of the cytokines demonstrated higher trends on day 36 compared to those on day 21 in all S+A-910823-vaccinated groups. Th1 cytokine (IFN-c)producing cells were observed in all S+A-910823-vaccinated groups with the cytokine level being dose-dependent on S-910823 (Fig. 3B). Individuals with PBMCs that produced Th2 cytokines (IL-4, IL-5) were also observed in all the S+A-910823vaccinated groups. However, SFC values for the Th2 cytokines in the S+A-910823-vaccinated animals tended to be lower than those for the Th1 cytokine ( Fig. 3C and 3D). In the animals vaccinated with 5 lg or 10 lg S+A-910823, the SFC values for IFN-c were 4.4-fold to 4.8-fold higher than those for IL-4 and 1.9-fold to 2.2fold higher than those for IL-5 (Fig. 3E). These results suggested that S-268019-b enhanced both Th1 and Th2 responses in the primates and did not induce a Th2-dominant phenotype.

SARS-CoV-2 challenge of vaccinated cynomolgus monkeys
To assess the protective efficacy of S-268019-b from SARS-CoV-2 infection and the potential of VDE, we performed a challenge study using monkeys previoulsly immunized intramuscularly with two doses of vehicle, two doses of 10 lg S+A-910823, or two doses of 10 lg S-910823-only at a 21-day interval (Fig. 4A). The ratios of clinical score, body weight, and respiratory rate of monkeys after challenge infection to those before infection were evaluated (Fig. 4B). The clinical score ratio dropped during the initial 2 days after infection with SARS-CoV-2 in the vehicle group and S-910823-only group, and then began to recover. The monkeys immunized with S+A-910823 showed a slight decrease in the clinical score ratio, but the extent was smaller compared with that for the vehicle and S-910823-only vaccinated monkeys. An obvious difference was seen between the clinical score ratios of the S+A-910823 group and the S-910823-only, but not S+A-910823 group and the vehicle, at 1 day post-infection (p < 0.001). The body weight and respiratory rate ratios did not show any remarkable changes among the three groups after challenge infection (Fig. 4B).
Blood biochemistry values of the challenge infected monkeys are shown in Supplementary Fig. S5. Overall, the monkeys immunized with S+A-910823 did not show any meaningful changes blood biochemistry, although a temporal elevation of CRP value was observed on day 1, which was lower than that of the vehicle group. After infection with SARS-CoV-2, all the monkeys showed temporary higher body temperatures during the night following inoculation. Then, two monkeys from the vehicle group and three monkeys from the S-910823-only group had 0.5℃ higher temperatures within 2 days post infection, while none of the S+A-910823 immunized monkeys exhibited an elevated temperature beyond the first night ( Supplementary Fig. S6). One of the monkeys from the vehicle group showed hypothermia. After SARS-CoV-2 inoculation, lymphocyte counts transiently decreased and then recovered in all three experimental groups, but there were no obvious differences among the groups (Supplementary Fig. S7).
We next evaluated the neutralizing antibody titer in serum samples from the monkeys administrated S+A-910823 and subse-quently challenged with SARS-CoV-2 after 8 to 13 weeks post second vaccination (Fig. 4C). In the challenged vehicle group, neutralizing antibody titers in the serum samples were below the lower limit of detection in all samples. Substantial levels of neutralizing antibody titer were detected before infection, and the levels were unchanged on days 1, 4, and 7 post infection in the S+A-910823administered group.
The copy numbers of virus genomic and subgenomic RNAs in swab samples are shown in Fig. 4D. The average amount of genomic RNA in nasopharyngeal swab samples from monkeys of the vehicle and S-910823-only groups were approximately 1 Â 10 7 copy/mL on day 1 and 1 Â 10 5 copy/mL on day 7, which was higher than that for the throat swab samples and rectal swab samples. Meanwhile, the average copy number of genomic RNA in nasopharyngeal swab samples from monkeys of the of S+A-910823 group was approximately 1 Â 10 6 copy/mL on day 1 and 4 Â 10 3 copy/ mL on day 7, which were both lower than those of the vehicle and S-910823-only groups. In comparison, the average copy number of subgenomic RNA transcripts in the nasopharyngeal swab samples of the S+A-910823 group was approximately 1 Â 10 4 copy/mL on day 1, which was again lower than that of the vehicle group (approximately 1 Â 10 6 copy/mL). The copy number of virus genomic and subgenomic RNAs in tissue homogenates on day 7 are shown in (Supplementary Fig. S8). The copy numbers for the S+A-910823 group were lower than those for the vehicle group in each sample. In addition, the copy number of viral genomic RNA and subgenomic RNA of the S-910823-only group were lower than those of the vehicle group.
Inflammatory cytokines and chemokines levels in the serum of the three groups of SARS-CoV-2 challenged animals are shown in (Supplementary Fig. S9). While the concentrations of IL-6, IL-1ra, and MCP-1 were increase on day 1 in the challenged vehicle group, they were not increased in the challenged S+A-910823 group.
Histopathological examination revealed mild interstitial pneumonia in the lungs of all the monkeys of the vehicle group following challenge infection (Fig. 5A, upper panels). All of the monkeys showed viral antigen positive cells in the lesions; however, only slight and focal intrestitial pneumoneia was observed in the lungs of the S+A-910823 group with no viral antigen positive cells (Fig. 5A, middle panels). For some animals, mild or slight mononuclear infiltration was present around blood vessels in the challenged S+A-910823 group. One monkey in the S-910823-only group exhibited severe pulmonary edema associated virus infection (Fig. 5A, lower panels). Other monkeys exhibited a few viral antigen-positive cells associated interstitial pneumonea with eosinophil and/or neutrophil infiltraion. The average of scores of thrombus and eosinophilic pneumonia of the challenged S+A-910823 group were lower than that of the challenged vehicle and S-910823 groups (Fig. 5B).

Discussion
In this study, we demonstrated that a new COVID-19 vaccine, S-268019-b, induced anti-S and anti-RBD IgG expression, and the production of neutralizing antibodies against SARS-CoV-2 and its variants in monkeys. S-268019-b also induced PBMCs to produce cytokines upon re-stimulation with S-protein peptides. Immunization with S-268019-b was also shown to protect cynomolgus monkeys against SARS-CoV-2 challenge, without any safety concerns based on hematological analysis, cytokine and chemokine profiles, and histopathological analysis. These results support the efficacy of S-268019-b as a novel vaccine against SARS-CoV-2 for eventual use in clinical settings.
In evaluating vaccines, appropriate animal models are essential. As thoroughly reviewed by Muñoz-Fontela and colleagues [21], non-human primate models have been evaluated as models of SARS-CoV-2 infection and pathogenesis and shown to demonstrate beneficial traits. For instance, they support SARS-CoV-2 replication in the upper respiratory tract, lower respiratory tract, and other organs; develop pneumonia with bilateral involvement, ground glass opacities, focal edema, and inflammation; immunologically seroconvert, develop neutralizing antibodies, induce T-cell immu- nity and pro-inflammatory cytokines, and exhibit more severe disease in older individuals [22][23][24][25][26]. This supports our choice to use cynomolgus monkeys as an in vivo model to evaluate the COVID-19 vaccine S-268019-b, which is under clinical development (Phase 1/2 Study: jRCT2031210269; Phase 2/3 Studies: jRCT2031210383 and jRCT2031210470).
To evaluate immunogenicity and antigen-dose dependency, the high-end amount of antigen was set at 25 lg/dose, which is a suf- In previous reports, inactivated SARS-CoV and S-protein vaccines induced not only neutralizing antibodies, but also eosinophilic  immunopathology in the lungs of mice upon challenge infection, which was related to insufficient induction of antibodies against SARS-CoV and a skewed immune reponse toward a Th2 response [13,15,27,28]. Given that the relatively greater induction of Th2 cytokines by 25 lg S-910823 may elevate risks for VDE, we concluded that the appropriate dose of S-910823 in the cynomolgus monkey model was 5 or 10 lg/dose.
S-268019-b induced potent neutralizing antibody titers against SARS-CoV-2, which were comparable to titers of the ''High" (20/150) WHO International Reference Panel. According to WHO, the titer of the 20/150 reference panel is higher than that of the WHO International Standard (20/136) [19]. The 20/136 standard has recently been used to estimate correlates of protection against COVID-19 in clinical trials, and the neutralizing antibody titer of the 20/136 standard has been shown to be higher than that induced by viral vector-based or mRNA-based vaccines [29][30][31]. Given the neutralizing antibody titer induced by S-268019-b was comparable to that of 20/150, S-268019-b is expected to have the potential to provide substantial protection against COVID-19. In addition, we showed that immunization with S-268019-b containing 10 lg of S-910823 provided protective efficacy against challange SARS-CoV-2 infection. As the neutralizing antibody titer declined to approximately 320 in immunized monkeys at 8 to 13 weeks post second vaccination, this basal level of neutralizing antibody titer may be adequate for preventing SARS-CoV-2 infection as the challenged monkeys were protected. This suggests the protective efficacy elicited by S-268019-b was retained for several months. However, further studies correlating neutralizing antibody titer and protecton against SARS-CoV-2 infection are still required.
Although S-268019-b elicited potent neutralizing antibody against an original SARS-CoV-2 virus, reaching levels comparable to those of the WHO International Reference Panel High 20/150 (1473 international unit/mL), Beta and Gamma variants were approximately 3-fold more resistant compared to the original viruses, and the Omicron variant was approximately 14-fold more resistant. The reduced neutralizing antibody titers were comparable to those of currently marketed vaccines, such as BNT162b2 and mRNA-1273 [32,33]. Given the antigen of S-268019-b is based on full-length S protein and anti-RBD antibodies are well-correlated with neutralizing antibody titer, RBD mutations may have been responsible for the reduced neutralizing antibody titers. Therefore, continuous monitoring of mutations that affect neutralizing antibody titer is important.
Previous studies suggest the balance between Th1 and Th2 immune responses may correlate with potential VDE risk [13,15,27,28]; therefore, we aimed to develop a vaccine that induces a Th1 > Th2 balanced immune response. We confirmed antigen-specific IFN-c production from PBMCs of S-268019-b immunized monkeys, whereas Th2-related IL-4 and IL-5 production were lower than that of Th1-related IFN-c. These results suggest that S-268019-b induced a Th1/Th2 balanced immune response, and did not predominantly enhance a Th2 type immune response. Accordingly, we concluded the risk of S-268019-b vaccination inducing VDE attributed to allergic inflammation-related cytokine production would be limitted. As VDE is a potential complication hindering the successful development of vaccines against viruses, the COVID-19 pandemic has fueled concerns that VDE may hinder the safety of SARS-CoV-2 vaccines [13,15,27,28]. However, previous studies using non-human primate models to evaluate adjuvanted SARS-CoV-2 vaccines with S-protein as antigen have reported no indication of VDE upon virus challenge [23,24]. Regardless, careful monitoring will still be needed to evaluate the potential risk of VDE in clinical trials of S-268019-b in humans.
The current study had limitations. First, potential speciesrelated differences exist between SARS-CoV-2 infections in humans and non-human primates, such as cynomolgus monkeys [26]. Second, the number of cynomolgus monkeys in each group was small, limiting some quantitative analyses. Third, since a model to estimate the potential risk of VDE has not been established in cynomolgus monkeys, it is possible that the current results underestimated the actual potential risks [16].
In conclusion, this study is the first evaluation of immunogenicity of the COVID-19 vaccine S-268019-b in non-human primates. The S-268019-b vaccine, which is composed of SARS-CoV-2 S-protein and A-910823 adjuvant, induced neutralizing antibody and cellular immunity in cynomolgus monkeys after the second vaccine administration. The findings also confirmed the safety of the vaccine in cynomolgus monkeys. Overall, the results support the clinical evaluation of S-268019-b in effort to provide a safe and efficient tool in combatting the worldwide health threat of COVID-19.

Ethical Approval/Informed Consent
All procedures using SARS-CoV-2 were conducted in a BSL3 laboratory. Experiments using recombinant DNA and pathogens were approved by the Committee for Experiments using Recombinant DNA and Pathogens at Shionogi & Co., Ltd. and NIID, Japan. For the immunogenicity studies, the animal protocols were approved by the Institutional Animal Care and Use Committee (IACUC 055-870) of Shin Nippon Biomedical Laboratories in Japan, which is accredited by AAALAC International. All animal protocols were in accordance with the animal welfare bylaws of Shin Nippon Biomedical Laboratories. The SARS-CoV-2 challenge studies were carried out in strict accordance with the Guidelines for Proper Conduct of Animal Experiments of the Science Council of Japan. All animal experiments were conducted in strict compliance with animal husbandry and welfare regulations and were approved by the Committee on Experimental Animals at NIID, Japan (approval no. 521006). All experimental animals in the BSL3 animal facilities were handled according to the guidelines of the biorisk management of NIID, Japan (approval no. 21-43).

Unblinded statements
All animal experiments were approved by the SNBL Institutional Animal Care and Use Committee (Approval No. IACUC055-870) and performed in accordance with the animal welfare bylaws of SNBL. The animal care and manipulations are detailed in the Supplementary Material. The SARS-CoV-2 challenge studies were carried out in strict accordance with the Guidelines for Proper Conduct of Animal Experiments of the Science Council of Japan. All animal experiments were conducted in strict compliance with animal husbandry and welfare regulations and were approved by the Committee on Experimental Animals at NIID, Japan (approval no. 521006).