Specific pathogen-free 10 week-old BALB/c mice were purchased from SLC (Hamamatsu, Japan). The mice were housed in a room with a 12-h:12-h light:dark cycle (lights on, 8:00 am; lights off, 8:00 pm) and unrestricted access to food and water. All animal experiments with or without SARS-CoV-2 were conducted in accordance with the guidelines of Osaka University for the ethical treatment of animals and were approved by the Animal Care and Use Committee of the Research Institute for Microbial Diseases, Osaka University, Japan (approval numbers: BIKEN-AP-R02-09-0).

The Gamma strain (hCoV-19/Japan/TY7-503/2021) of SARS-CoV-2 was obtained from the National Institute of Infectious Diseases (Tokyo, Japan). The virus was expanded in VeroE6/TMPRSS2 cells (JCRB Cell Bank, Osaka, Japan) and stored at −80 °C until use. VeroE6/TMPRSS2 cells were cultured in DMEM (Nacalai Tesque, Kyoto, Japan) containing 10% fetal calf serum (FCS) and 1% penicillin/streptomycin. Mice were intranasally challenged with MA10, a mouse-adapted SARS-CoV-2. MA10 was generated using a circular polymerase extension reaction as described previously (53). The SARS-CoV-2 NIID strain (2019-nCoV_Japan_TY_WK-5212020) served as the backbone of MA10. MA10 contains seven mutations, which were introduced adaptively into SARS-CoV-2 during serial passages in BALB/c mice (54). Each experiment with live SARS-CoV-2 was performed within a biosafety level 3 facility at Osaka University, adhering to stringent guidelines.

VeroE6/TMPRSS2 cells in 225 cm² flasks (Corning, Corning, NY, USA) were infected with the Gamma strain of SARS-CoV-2 at 0.01 multiplicity of infection (MOI) at 37 °C. After three days, the cell culture medium was collected via centrifugation at 3000 rpm for 10 min to remove debris. Live SARS-CoV-2 in the supernatant was inactivated using β-propiolactone (Wako, Osaka, Japan) at a concentration of 1:4000 (v/v) at 4 °C for 24 h, then incubated at 37 °C for 2 h to hydrolyze the remaining β-propiolactone. To confirm viral inactivation, we performed a plaque assay before purification. Briefly, VeroE6/TMPRSS2 cell monolayers (1 × 10⁶ cells/well) in 6-well plates (Corning) were incubated with medium as a negative control, live SARS-CoV-2 (diluted 10⁴-fold with medium) as a positive control, or inactivated WV (undiluted) at 37 °C for 2 h. The cells were then rinsed twice with phosphate-buffered saline (PBS) and covered with agar (final 1%; SeaPlaque TM Agarose, Lonza, Basel, Switzerland) in a medium containing 2% FCS and 0.5% penicillin/streptomycin. The plates were then incubated at 37 °C. After three days, the cells were fixed with 10% formalin (Nacalai Tesque) and the agar was removed. The cells were then stained with 1% crystal violet (Tokyo Chemical Industry, Tokyo, Japan) and the plaques were counted. Next, the inactivated WV was purified using sucrose density gradient centrifugation. Briefly, using 20% sucrose, inactivated WV pellets were collected by centrifugation at 32,000 rpm at 4 °C for 2 h. The pellets were resuspended in PBS, and to exclude artefacts, purified through a 15–60% sucrose gradient at 30,000 rpm at 4 °C for 2 h. After centrifugation, 23 fractions (1.5 mL each) were collected from the top to bottom, and the expression of S protein (16–19 fractions) was confirmed by western blotting. Subsequently, to exclude sucrose, 16-19 fractions were diluted with PBS and centrifuged at 30,000 rpm at 4 °C for 3 h. Finally, the pellets were resuspended in PBS and the amount of inactivated WV was quantified using a Pierce protein assay kit (Thermo Fisher Scientific, Hampton, NH, USA) with a bovine serum albumin standard.

To confirm viral inactivation in vivo, mice were treated with SARS-CoV-2 (Gamma strain; 1.7 × 10⁵ Plaque-forming unit; PFU), SARS-CoV-2 MA10 (2 × 10⁵ PFU), or inactivated WV (3 μg or 10 μg) in 20 μL of PBS under anesthesia. The nasal turbinate was harvested and homogenized in DMEM containing 2% FCS and 1% penicillin/streptomycin 3 days after treatment, followed by centrifugation and supernatant collection. To titrate the infectious virus, the supernatant of the nasal turbinate was serially diluted in DMEM containing 2% FCS and 1% penicillin/streptomycin. The plaque assay was performed as described above.

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed in accordance with previous publications (32). Briefly, purified proteins were mixed in sample buffer solution (Nacalai Tesque) containing 2-mercaptoethanol (Sigma-Aldrich, St. Louis, MO, USA) at 1:1 (v/v) ratio and heated at 95 °C for 5 min before being loaded onto a 10% Mini-PROTEAN TGX Precast Protein Gel (Bio-Rad, Hercules, CA, USA). Following electrophoresis, the gels were stained with Coomassie Brilliant Blue (Nacalai Tesque). After gel electrophoresis, protein bands were transferred from the SDS-PAGE gel onto polyvinylidene fluoride membranes (Bio-Rad). After blocking in 5% skim milk (w/v) diluted in distilled water, the transferred membrane was washed thrice with PBST (0.05% Tween-20 in PBS) and incubated with mouse anti-SARS-CoV-2 (2019-nCoV) spike antibody (clone: #42; Sino Biological, Beijing, China) or rabbit anti-SARS membrane protein antibody (polyclonal; Novus Biologicals, Littleton, CO, USA). The membrane was then washed thrice with PBST, and incubated with a rabbit anti-mouse horseradish peroxidase-conjugated secondary antibody (Abcam Ltd., Cambridge, UK) or goat anti-rabbit horseradish peroxidase-conjugated secondary antibody (Abcam Ltd.), and detected using a ChemiDoc Touch Imaging System (Bio-Rad). The amount of S protein in inactivated WV was determined using western blotting ( Supplementary Figure 1B ). We used recombinant S protein as standard (from 4.3 ng/well to 69.3 ng/well). After SDS-PAGE, the proteins were transferred onto the membrane for western blotting. The transferred membrane was incubated with rabbit anti-SARS-CoV-2 (2019-nCoV) spike antibody (Polyclonal IgG; Sino Biological). The membrane was then washed thrice with PBST and incubated with a goat anti-rabbit horseradish peroxidase-conjugated secondary antibody (Abcam Ltd.). The three points (34 ng/well, 68 ng/well, and 136 ng/well) for inactivated WV were averaged. Immunoreactive bands were quantified using Image J software (NIH Image, Bethesda, MD, USA).

The inactivated WV was directly analyzed using a Tecnai G2 20 twin transmission electron microscopy (TEM; FEI Company, Hillsboro, OR, USA) at 200 kV. Carbon-coated copper grids with 250 mesh (STEM Co. Ltd, Tokyo, Japan) were glow-discharged using DII-29020HD (JEOL, Tokyo, Japan). The inactivated WV was fixed in 2% glutaraldehyde overnight at 4 °C. Samples were diluted in PBS to 20 μg/mL immediately before measurement, and 10 μL of the sample was applied to the grids for 5 min. Four-diluted EM stainer (Nisshin EM, Tokyo, Japan) was then used to stain the sample on the grid for 30 min. The specimen was finally gently blotted from the side with filter paper and air-dried before imaging. Images were recorded on an Eagle 2 K × 2 K CCD camera with a defocus of -2 μm at a nominal magnification of 25,000× or 50,000×.

The cDNA of the S protein ectodomain from the Gamma strain of SARS-CoV-2 with a C-terminal hexahistidine tag (His-tag) was cloned into the pcDNA3.1 expression plasmid (Thermo Fisher Scientific). The foldon sequence (GYIPEAPRDGQAYVRKDGEWVLLSTFL) derived from the fibritin of bacteriophage T4 was integrated at the C terminus of the S protein. S protein production was performed in accordance with previous publications (32).

For the transcription of mRNA expressing the S protein of SARS-CoV-2, template pDNA encoding the S protein (Wuhan-Hu-1) was prepared using polymerase chain reaction. The pDNA-spike was linearized using restriction enzymes. After phenol-chloroform extraction and ethanol precipitation, the linearized pDNA was transcribed into mRNA using the MEGAscript™ T7 transcription kit (Thermo Fisher Scientific) according to the manufacturer’s instructions. N1-methylpseudouridine was incorporated into the mRNA by replacing uridine. Residual dsRNA was removed as previously described (55). The 5′ cap was added according to the ScriptCap Cap 1 Capping System protocol (Madison, WI, USA). The 3′ poly(A) tail was added according to the protocol provided with the poly(A) Tailing Kit (Thermo Fisher Scientific). A lipid mixture containing SM-102 (Cayman Chemical Company, Ann Arbor, MI, USA), DMG-PEG2000 (NOF Corporation, Tokyo, Japan), DSPC (NOF Corporation), and cholesterol (Sigma-Aldrich) in an ethanol solution (SM-102/DSPC/cholesterol/DMG-PEG2000 = 50/10/38.5/1.5) were prepared. The mRNA-lipid nanoparticles (LNPs) were produced by mixing the mRNA dissolved in acetic acid/NaOH buffer (pH 5.0) and the lipid mixture in ethanol at an N/P ratio of 5.5 using a NanoAssemblr^® instrument (Precision Nanosystems, Vancouver, Canada) at a flow rate ratio of 3:1. The external solution of the mRNA-LNPs was replaced with PBS (Nacalai Tesque) by ultrafiltration using Amicon Ultra-4-100K centrifugal units. The size and zeta potential of mRNA-LNPs were measured using a Zetasizer Nano ZS (Malvern Instruments, Malvern, UK). The encapsulation efficiency of mRNA was measured using the RibogreenTM reagent (Invitrogen, Carlsbad, CA, USA).

For nasal vaccination, BALB/c mice were administered with inactivated WV (3 μg/mouse) in a total volume of 5 μL, divided equally between both nostrils. Subcutaneous vaccination was achieved via administration of the BALB/c mice at the base of the tail with inactivated WV (3 μg/mouse) combined with Alhydrogel (alum; InvivoGen, San Diego, CA, USA) as an adjuvant (50 μg/mouse) in a total volume of 50 μL. For both routes, the mice were administered on days 0 and 21. Unimmunized mice were used as controls. On day 32, serum, nasal wash, and bronchoalveolar lavage fluid (BALF) samples were collected and stored at -80 °C. The nasal cavity was gently flushed using 400 μL of PBS to obtain nasal wash samples. BALF was obtained via lung lavage with 1.2 mL PBS. In some experiments, mRNA expressing the SARS-CoV-2 Spike (1 μg/mouse) was administered intramuscularly to mice in a total volume of 50 μL. On day 21 after mRNA vaccination, the mice were intranasally administered with inactivated WV (3 μg/mouse) in a total volume of 5 μL, divided equally between both nostrils. On day 28, serum and nasal wash samples were collected.

We determined the levels of S-specific antibodies in the serum, nasal wash, and BALF using enzyme-linked immunosorbent assay (ELISA). To detect SARS-CoV-2-specific total IgG and IgA, ELISA plates (Corning, Corning, NY, USA) were coated overnight at 4°C with S protein of the Gamma strain in PBS (1 μg/mL or 10 μg/mL). ELISA was performed in accordance with previous publications (32). Endpoint titers were determined using the following procedure: The background value was subtracted from the OD value. A subtracted OD value of 0.1 or more was regarded as positive, and the maximum dilution to give a positive result was used as the endpoint titer.

For the neutralization assay, VeroE6/TMPRSS2 cells (1.2 × 10⁴ cells/well) were added to a 96-well half-white plate (Greiner BIO-ONE, Kremsmunster, Austria) and incubated in DMEM containing 10% FCS and 1% penicillin/streptomycin at 37°C for 24 h. On the day after incubation, serum samples and nasal wash samples were evaluated using 50-819,200-fold, or 4-65,536-fold serial dilutions of undiluted solutions. The serial dilutions were mixed 1:1 with pseudotyped viruses, replication-deficient vesicular stomatitis virus (VSV) bearing Gamma spike of SARS-CoV-2, and incubated at 37°C for 1 h. After removing the growth medium from the cells, 50 μL of serum/virus or nasal wash/virus mixture was added to the cells and then incubated at 37°C for 48 h. Following 48 h post-infection, 50 μL of ONE-Glo™ EX Reagent (Promega, Madison, WI, USA) was pipetted into each well and mixed. Luminescence was measured using a microplate reader (CORONA electric, SH-9000, Hitachi High-Tech Corporation, Tokyo, Japan).

To evaluate the percentage of germinal center (GC) B cells in the nasal passage, lymphocytes were collected seven days after the last immunization and analyzed via flow cytometry. The nasal passage lymphocytes were added to a 96-well plate, and incubated with an anti-mouse CD16/CD32 antibody (1:400 dilution; clone:93; BioLegend, San Diego, CA, USA), PE anti-mouse CD45 antibody (1:200 dilution; clone:30-F11; BioLegend), BV421 anti-mouse B220 antibody (1:200 dilution; clone: RA3-6B2; BioLegend), AF647 anti-mouse GL7 antibody (1:200 dilution; clone: GL7; BioLegend), and PE/Cy7 anti-mouse CD95 (Fas) antibody (1:200 dilution; clone: SA367H8; BioLegend) in PBS with 2% FCS, 1 mM EDTA (DOJINDO, Kumamoto, Japan), and 0.05% azide (Wako) for 30 min at 4°C. GC B cells were defined as CD45+ B220+ Fas+ GL7+ cells. Flow cytometry was performed using the CytoFLEX Flow Cytometer (Beckman Coulter, Brea, CA, USA). Kaluza software (Beckman Coulter) was used for analyzing the flow cytometry data.

On day 32 after vaccination, splenocytes (1 × 10⁶ cells/well) were added to a 96-well plate. Subsequently, the cells were stimulated for three days at 37 °C with S protein or left unstimulated (final concentration: 20 μg/mL). Post-incubation, the levels of interferon (IFN)-γ and interleukin (IL)-13 in the supernatants were analyzed using commercial ELISA kits (BioLegend for IFN-γ; R&D Systems, Minneapolis, MN, USA, for IL-13), as per manufacturer’s instructions. Standards supplied with the kits were used to quantify the cytokine levels.

Fourteen days after the last immunization, the mice were challenged intranasally with SARS-CoV-2 MA10 (5 × 10⁴ PFU) in 5 μL of PBS under anesthesia. Unimmunized mice were used as controls. After challenge, the nasal turbinates were harvested and homogenized in 500 μL of DMEM containing 2% FCS and 1% penicillin/streptomycin. Next, 500 μL of DMEM containing 2% FCS and 1% penicillin/streptomycin was added, followed by centrifugation and collection of the supernatant. To titrate the infectious virus, the supernatant of the nasal turbinate was serially diluted in DMEM containing 2% FCS and 1% penicillin/streptomycin. The plaque assay was performed as previously mentioned.

Fourteen days after the last immunization, mice were infected intranasally with SARS-CoV-2 MA10 (5 × 10⁴ PFU or 2 × 10⁵ PFU) in 20 μL of PBS under anesthesia. Unimmunized mice were used as controls. After the infection, the body weight and survival of the mice were observed for eight days. The day of death was defined as the specific day on which the body weight of the mice fell below 75% of their initial weight.

Statistical analyses were performed using Prism software (GraphPad Software, San Diego, CA, USA). All data are presented as mean ± standard deviation (SD). One-way ANOVA was used to determine significance differences, followed by Tukey’s test. P < 0.05 was considered to indicate statistical significance.

To generate inactivated WV, we used a Gamma strain (hCoV-19/Japan/TY7-503/2021) as the seed virus. The virus was propagated in VeroE6 cells expressing transmembrane serine protease 2 (TMPRSS2) and was inactivated using β-propiolactone ( Figure 1A ). Using a plaque formation assay, we confirmed that β-propiolactone completely inactivated SARS-CoV-2 ( Figure 1B ). In addition, no live virus was detected in the nasal turbinates of mice intranasally administrated inactivated WV using plaque formation assay ( Supplementary Figure 1A ). Inactivated WV was purified using sucrose density gradient centrifugation and the purity of the inactivated WV was confirmed using SDS-PAGE ( Figure 1C ) and western blotting ( Figure 1D ). SDS-PAGE analysis revealed three major bands. Bands of approximately 150, 50, and 15 kDa indicated S protein, nucleocapsid (N) protein, and membrane (M) protein, respectively (56). Western blot analysis also showed bands for the S and M proteins ( Figure 1D ). Next, we evaluated the amount of S protein in inactivated WV using western blotting. As a result, we found that 3 μg of inactivated WV contained approximately 2.1 μg of S protein ( Supplementary Figure 1B ). Furthermore, using transmission electron microscopy (TEM; Figure 1E ), we demonstrated that the inactivated WV retained the structure of the virus. These results suggest that inactivated WV could be used as a vaccine antigen.

To evaluate the functionality of nasal vaccines, mice were immunized intranasally (nasal vaccine) with inactivated WV without an adjuvant or subcutaneously (subcutaneous vaccine) with inactivated WV and alum as an adjuvant. The endpoint titers of S protein-specific IgG and IgA in the serum, bronchoalveolar lavage fluid (BALF), and nasal washes after vaccination were analyzed using ELISA. The endpoint titers of S-specific IgG in the serum of the nasal vaccine group were significantly higher than those in the unimmunized control group but lower than those in the subcutaneous vaccine group ( Figure 2A ). Subcutaneous vaccines induced significantly higher levels of S-specific IgG in the serum, BALF, and nasal washes than those with nasal vaccines ( Figures 2A–C ). In contrast, the endpoint titers of S-specific IgA in the serum ( Figure 2D ) and nasal wash ( Figure 2F ) were significantly higher in the nasal vaccine group than in the subcutaneous vaccine and unimmunized control groups, although no change was observed in BALF ( Figure 2E ). To examine neutralizing antibody titer in serum and nasal wash from mice vaccinated with inactivated WV, we used a pseudotyped virus displaying a Gamma spike of SARS-CoV-2. The serum from intranasally and subcutaneously vaccinated mice neutralized the pseudotyped virus, whereas the serum from non-vaccinated mice was ineffective ( Figure 2G ). In contrast, the nasal wash from intranasally vaccinated mice tended to neutralize compared to that from subcutaneously vaccinated mice and non-vaccinated mice ( Figure 2H ). Next, to evaluate germinal center (GC) B cells in the nasal passage induced by vaccination, we examined the percentage of GC B cells among total B cells using flow cytometry ( Figure 3A ; Supplementary Figure 2 ). Seven days after the last immunization, the percentage of GC B cells among the total B cells was significantly higher in the nasal vaccine group than in the subcutaneous vaccine and unimmunized control groups ( Figure 3A ). These results indicated that the inactivated WV nasal vaccine strongly induced S-specific IgA in the upper respiratory tract.

To examine the SARS-CoV-2-specific CD4+ T cell response induced by vaccination, we examined cytokine production in splenocytes. Splenocytes from vaccinated mice were stimulated with S protein, and the amount of cytokines in the supernatant was analyzed. The levels of IFN-γ and IL-13 (cytokines associated with T helper type 1 and 2 cells, respectively) produced by splenocytes after S protein stimulation were significantly higher in the subcutaneous vaccination group than in the nasal vaccination and unimmunized control groups ( Figures 3B, C ). In contrast, there was no difference in the level of IFN-γ between the unimmunized control and nasal vaccine groups, whereas the level of IL-13 in the nasal vaccine group was significantly higher than that in the unimmunized controls ( Figures 3B, C ).

To evaluate the defensive effects of nasal vaccination on the upper respiratory tract, vaccinated mice were challenged intranasally with a mouse-adapted strain of SARS-CoV-2 (SARS-CoV-2 MA10; 5 × 10⁴ PFU) (54) in a total volume of 5 μL. In this model, a plaque assay was used to measure the viral titer in the nasal turbinates as an indicator of vaccination effectiveness. After challenge, the viral titer in the nasal turbinates of the nasal vaccine group was significantly lower than that in the subcutaneous vaccine and unimmunized control groups ( Figure 4A ). In addition, the virus titer in the nasal turbinates of the subcutaneous vaccine group was significantly lower than that in the unimmunized control group ( Figure 4A ). Furthermore, to assess the defensive effects of nasal vaccination on the lower respiratory tract, vaccinated mice were challenged intranasally with SARS-CoV-2 MA10 (5 × 10⁴ PFU or 2 × 10⁵ PFU) in a total volume of 20 μL. In this model, body weight loss and survival were measured as indicators of vaccination effectiveness and these parameters were monitored for eight days after infection. At a low-titer challenge (5 × 10⁴ PFU), the unimmunized controls lost weight and some of them died after the challenge, whereas mice in the nasal vaccine and subcutaneous vaccine groups did not ( Figures 4B, C ). In addition, at a high titer challenge (2 × 10⁵ PFU), the subcutaneous vaccine group completely protected from weight loss and death ( Figures 4D, E ). The nasal vaccine group was also completely protected from death, even though it showed a body weight decrease ( Figures 4D, E ). These results suggest that vaccination with intranasally inactivated WV can protect against upper and lower respiratory tract infections caused by SARS-CoV-2.

Finally, we examined whether systemic priming with an mRNA vaccine, followed by intranasal boosting with the inactivated WV, could enhance antibody responses in the nasal cavity. The mice were divided into four groups ( Figure 5A ). Group 1 was vaccinated intranasally with inactivated WV for priming with no boosting (prime: i.n., boost: none). Group 2 was vaccinated intranasally with the inactivated WV for both priming and boosting (prime: i.n., boost: i.n.). Group 3 was intramuscularly vaccinated with an mRNA vaccine encoding the S protein for priming and intranasally vaccinated with the inactivated WV for boosting (prime: i.m., boost: i.n.). Group 4 was intramuscularly vaccinated with the mRNA vaccine for both priming and boosting (prime: i.m., boost: i.m.). Seven days after final vaccination, the endpoint titers of S-specific IgA in the nasal mucosa of group 2 were found to be significantly higher than those in groups 1 and 4 ( Figure 5B ). Further, the endpoint titers of S-specific IgA in nasal mucosa were significantly higher in group 3 than in groups 1 and 4 ( Figure 5B ). The endpoint titers of S-specific IgG in the serum of group 4 were significantly higher than that of groups 1, 2, and 3 ( Figure 5C ). Further, the endpoint titers of S-specific IgG in the serum of group 3 were significantly higher than those in group 1 but comparable to those in group 2 ( Figure 5C ). Our data suggest that inactivated WV can be used as a booster for individuals already administered the mRNA vaccine.