Polypeptide antigens of spike protein were cloned, expressed, and purified as described earlier (17, 36). Briefly, spike cDNA corresponding to polypeptide antigens were cloned into a pET21a vector (Takara, Shiga, Japan), followed by PCR, then PCR products were ligated into an expression vector using suitable restriction enzymes (Takara, Shiga, Japan). The positive clone containing the polypeptide cDNA insert was confirmed by analysis of their respective DNA sequencing (Cosmogen, Seoul, Korea). Next, expression vectors were transformed into BL21-CodonPlus (Stratagene, San Diego, CA, USA) through a heat-shock technique. After collecting the expressed polypeptides, they were purified using their 6 × his-tag at the C-terminus by TALON^® Magnetic Beads (Takara) followed by HPLC purification. Their concentrations were verified via silver staining and Bradford assay.

Purified spike antigens were assessed for their neutralizing ability using serum samples from SARS-CoV-2-vaccinated and naturally infected people, which were approved by the Institutional Review Board of Yeungnam University Medical Center, Korea (approval no. 2020-07-063) (17). Homemade enzyme-linked immunosorbent assay (ELISA) was used to detect neutralizing antibodies (anti-SARS-CoV-2 antibodies) within human sera against a list of purified spike antigens, which was used to coat max-flat-bottom 96-well plates at a final concentration of 1 µg/ml and kept at 4°C 1 day before the assay. The next day (the day of the assay), the plates were washed three times with phosphate-buffered saline and 0.1% Tween (PBS-T) and blocked with 200 μl/well 2% BSA for 1 h at room temperature (RT), followed by washing with PBS-T three times, and then incubated with serially diluted serum samples for 2 h at RT. Next, the plates were washed three times with PBS-T, incubated on a rocker for 0.5 h at RT with antibody-HRP, washed three times with PBS-T, and incubated with TMP-substrate 100 μl/well for 20 min at RT followed by 100 μl/well of stop solution. The ELISA plate was read at 450 nm on a microplate reader. The same ELISA steps were used for the titration assay of mice serum.

All the animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) at Konkuk University. Two vaccines, vaccine A and vaccine B, were designated among the purified spike antigens. Both were injected twice subcutaneously with 2-week intervals; the first injection (on day 21) was formulated with a complete adjuvant, and the second (on day 7) was formulated with an incomplete adjuvant. Antigens and adjuvants (Freund’s adjuvant, a known solution of antigen emulsified in mineral oil used as an immunopotentiation; both complete and incomplete adjuvants were used: the complete adjuvant is made of inactivated and dried mycobacteria, and the incomplete adjuvant lacks the mycobacterial components) were mixed in a 1:1 ratio to the mentioned final concentration. Male mice of K18-hACE2 TG at 7 weeks of age were used; each group was injected with a dose of 2 μg per mouse, and all mice were preserved with food and water and weighted and monitored daily. The first group was vaccinated with vaccine A followed by those infected/challenged with SARS-CoV-2 virus (n = 5); the second group was vaccinated with vaccine B followed by those infected/challenged with SARS-CoV-2 (n = 5); and the control group was only infected/challenged with SARS-CoV-2 (n = 5), 1 × 10⁵. The median tissue culture infectious dose (TCID₅₀) of SARS-CoV-2 virus (NCCP 43326) was given intranasally at day 0. The viral infection of SARS-CoV-2 by real-time RT-PCR was performed according to the guidelines of Korea Centers for Disease Control & Prevention (KCDC&P). Next, sera were collected on day 6 or 7, and all sera were kept at 4°C until use.

The following elements were checked to evaluate changes and compare the three groups of mice: mouse weight, mouse activity, and survival rate until 7 dpi. Moreover, lung, spleen, and small intestine tissue excisions from sacrificed mice on day 7 were used for histopathological score measurements. Virus titer was measured for lung tissues, and tissue weight/body weight was measured for lung tissues.

Lung, spleen, and small intestine organs from sacrificed mice were collected on day 7 after infection. The collected tissues were fixed using 4% paraformaldehyde, paraffin-embedded, cut into sections equally, and stained with hematoxylin–eosin (H&E) staining for detection of histopathological changes. Inflammation, edema, and bronchiolitis lesions were measured for lung tissues. Spleen atrophy of the white pulp was measured in the spleen. The number of goblet cells was measured in the small intestine.

Statistical analysis was completed using Prism 8.0 (GraphPad Software). One-way ANOVA or two-way ANOVA, followed by Tukey’s post-hoc correction, was used; P-value <0.05 was considered significant.

Various antigens/polypeptides of the spike protein were constructed and purified to be evaluated as a vaccine candidate. These antigens were checked for their ability to bind to neutralizing antibodies within the human sera of either vaccinated people or patients naturally infected with the SARS-CoV-2 virus. Among them, four antigens summarized in Table 1 were selected for further experiments. These antigens were used as a SARS-CoV-2 vaccine and to compare whether one antigen is sufficient to offer protection similar to multiple antigens. We used two vaccines as follows: vaccine A is composed of a single antigen-1 (Ag1), whereas vaccine B is a mixture of four antigens (Ag1, Ag2, Ag3, and Ag4). The scheme of the designed experiments is summarized in Figure 1. Briefly, mice were immunized with two different vaccines (two groups each with either vaccine A or vaccine B) subcutaneously in a final dose of 2 μg and scheduled within 3 weeks prior to infection as the first injection was at day 21 and the second injection was at day 7 (2-week interval), and then mice were infected with 10⁵ PFU of SARS-CoV-2 virus at day 0, sera were collected, and mice were necropsied at day 7.

Two-step purified recombinant antigens were visualized by 10% SDS-PAGE and silver staining (Supplementary Figure 1). These four antigens were selected to be examined as vaccine candidates because of their highest binding ability (higher OD₄₅₀) with vaccinated or naturally infected human serum samples. COVID-19 patient 2’s resulting titer against all four antigens was high and tightly associated with the COVID-19 neutralizing index 640 (Figure 2A). However, the vaccinated human sera exhibited very low titers compared to naturally infected patients’ sera. In addition, the vaccinated mouse sera were examined for their titers using the four antigens (Figure 2B). In general, the multiple spike polypeptide (vaccine B)-immunized group exhibited higher titers compared with the single polypeptide (vaccine A)-immunized group.

After the vaccination schedule was performed as shown in Figure 1, the three groups of K18-hACE2 TG mice were intranasally infected and on day 7 sacrificed for collection of organ tissue and blood samples. Prior to and postinfection, the weight of mice was monitored for the two groups that have been vaccinated, that is, those vaccinated with vaccine A and those vaccinated with vaccine B, till the day of infection in which there was no difference in the body weight of these two groups. From day 0 up to day 7 postinfection, the three groups—the control group (infected only) was added—were compared for their body weight changes as percent change (Figure 3A, upper panel). At 6 dpi, both the SARS-CoV-2-infected group and the vaccine A-administrated group showed a body weight loss of nearly 15%, whereas the vaccine B-administrated group revealed an 11% loss of body weight. From 4 to 7 dpi, three mice of the vaccine-administered groups (one from the vaccine-A group and two from the vaccine-B group) recovered their weight after weight loss (Figure 3A, bottom panel). Although there was no statistical significance between the three groups in body weight changes, the vaccine B-administrated group demonstrated the lowest body weight loss and improved recovery (Supplementary Figure 2).

In addition, the activity of the three groups was observed and the changes were summarized as percentages (Figure 3B, upper panel). The activity in the SARS-CoV-2-infected group started to decrease at 5 dpi and continuously worsened leading to a moribund state at 7 dpi, whereas in the vaccine-administered group, the activity decreased from 6 dpi and then continued to decrease leading to a moribund status in some mice at 7 dpi. Moreover, the activity of two mice in the vaccine-A group as well as two mice in the vaccine-B group remained 100% up to 7 dpi, and a moribund status was observed in two mice of the vaccine-A group but one mouse of the vaccine-B group (Figure 3B, bottom panel). As indicated also in the survival outcomes that were monitored up to 7 dpi, all mice were sacrificed on day 7 (Figure 3C). There was no significant difference between the three groups in the tested measurements; body weight changes, activity, and survival; however, the mice of group vaccine-B showed the most promising results.

The three groups of mice were sacrificed at 7 dpi, and the viral titers and histopathologic changes were evaluated. In the lung, both viral titers and histopathologic changes are as shown in Figure 4. The vaccine-B group had a markedly lower viral titer (2.1 × 10⁴) compared with those of SARS-CoV-2 infection only (3.7 × 10⁷). Also, the virus titer of the vaccine-A group (9.2 × 10⁵) showed a decrease compared with the SARS-CoV-2-infected group (Figure 4A), although there was no statistically significant difference. On the other hand, histopathological changes in the lungs of mice of the vaccine B-administered group were not greatly improved compared with the group of SARS-CoV-2-infected mice but was not worsened as in the vaccine A-administered group (Figures 4B–E). The total histopathological score showed a non-significant difference between the three groups (Figure 4D). However, for lung edema (Figure 4B), the vaccine A-administered mice showed higher edema compared with SARS-CoV-2-infected mice and vaccine B-administered mice (lowest edema score). In addition, out of five mice, two mice of the vaccine B-administered group revealed the lowest histopathological score when comparing all 15 mice of the three groups (Figure 4D, right panel). These quantitative histopathological scores were confirmed with the histopathological analysis of the mice’s lungs (Figure 4E). Lung granulomas are highly formed in the mice of SARS-CoV-2 infection, followed by the vaccine A-administered group, and the least granuloma formation was in the vaccine B-administered group. Thus, taking all together, this suggests that vaccine B confers a promising vaccine nomination as it showed lower viral titers and histopathological scores.

We investigated the histopathological changes in the spleen and small intestine, other than the lung, to evaluate the degree of spreading of the infection and how vaccines impact these organs (Figures 5, 6). The changes in spleen pathology of the three groups are shown in Figure 5, and the changes in other tissue weight over body weight (liver, spleen, right and left kidneys, and lung) are shown in Supplementary Figure 3. The spleen-to-body weight ratio showed significantly higher percentages in the vaccinated groups (vaccine A; *p = 0.025 and vaccine-B; *p = 0.042) compared with the SARS-CoV-2-infected group (Figure 5A). The spleen showed to be enlarged under infectious conditions due to the increased immune cell proliferation and differentiation against infected pathogens such as SARS-CoV-2. Thus far, the severity of the damage after infection or injury is mainly measured by the white pulp atrophy rather than spleen enlargement, whereas the atrophy lesion in the splenic white pulp was shown to be increased in the vaccine A-administered group with no significant difference in the vaccine B-and SARS-CoV-2-infected groups (Figures 5B, C). Furthermore, small intestine goblet cells are known for their participation in the immune response; however, increasing the number of these cells indicated worsening of the case and mucus hypersecretion will result in goblet cell hyperplasia (37). In Figure 6, the histopathological changes in terms of goblet cell number of the three mice were evaluated. As the results showed, the vaccine B-administered group represents a statistically significant low number of goblet cell in villi (*p = 0.027), representing a healthier intestine among the three groups.