FMD structural protein expressed in Escherichia coli was purified by nickel ion affinity chromatography, and assembled into FMD VLPs in vitro after tag removal (Guo et al., 2013). The target protein was detected by SDS-PAGE and Western blot analysis.

The biomineralization of FMD VLPs was completed by a well-established method in the laboratory (Du et al., 2019) and the morphology of CaP-VLPs was examined by transmission electron microscopy (TEM; HT7700; Hitachi, Tokyo, Japan). CaP-VLPs solution was dropped onto a carbon-coated 300-mesh copper grid. The excess sample solution was removed with filter paper and dried at room temperature prior to observation. For VLPs, it is necessary to drop the phosphotungstic acid solution for negative staining for 30 s after the VLPs solution has dried; excess solution was removed with filter paper before observation. The particle size of VLPs and CaP-VLPs was measured by dynamic light scattering using a Zetasizer Nano (Malvern Zetasize Nano ZS90; Worcestershire, UK).

In this study, MONTANIDE™ GEL 02 PR (Number: 36084X) and MONTANIDE™ IMS 1313 VG NST (Number: 36071H; SEPPIC company), and STR products (patent number: ZL201310018498.X, China Biotech Co. Ltd.) were selected as adjuvants to mix with FMD CaP-VLPs. The above adjuvant products are abbreviated as G, I, and S, respectively.

All experimental procedures involving BALB/c mice complied with the Animal Welfare Procedures and Guidelines of the People’s Republic of China for Animal Ethics. Experiments were approved by the Animal Ethics Committee of Lanzhou Veterinary Research Institute, Chinese Academy of Agricultural Sciences. BALB/c female mice (6–8 weeks old, 16–20 g) were randomly divided into five groups (20 mice in each group). On days 0, 7, and 14, mice in the experimental group were intranasally immunized with 50 μg CaP-VLPs each time, and mice in the control group were immunized with phosphate-buffered saline (PBS). Serum and mucosal samples were collected at different time points after immunization for antibody analysis. Lung lavage fluid, small bowel lavage fluid, and fecal pellets were collected from mice at different time points after immunization and immersed in ice-cold buffer overnight. The buffer was a PBS solution containing soybean trypsin inhibitor (Gibco; 0.1 mg/ml), EDTA (50 mM), Pefabloc (Merk; 0.35 mg/ml), bovine serum albumin (BSA; 0.1 mg/ml). The buffer formula is from Christensen’s work (Christensen et al., 2019). After incubation, lung and small bowel lavage fluid, and fecal pellet suspensions were clarified and recollected by centrifugation at 12000 g for 30 min at 4°C. The supernatants were frozen and analyzed at a later date by the indirect enzyme-linked immunosorbent assay (ELISA).

Indirect ELISA: Inactivated type-O FMDV from the FMD type-O liquid-phase blocking ELISA (LPBE) antibody detection kit was used as coating antigen on ELISA plates (Costar), and the free binding sites were blocked with 5% skim milk in PBS-Tween-20 (PBST). Mucosal samples to be tested and goat anti-mouse IgA (HRP; Abcam) were added in sequence. TMB was used as substrate to detect antibody responses. After the reaction was stopped, the optical density (OD) was measured at 450 nm. The contents of different types of IgG in serum at 28 days after immunization were detected by indirect ELISA. The ELISA plate was coated with inactivated type-O FMDV overnight at 4°C, blocked with 5% skim milk in PBST, and serum after immunization for 28 days (diluted 100 times) and goat anti-mouse IgG1, IgG2a (HRP; Abcam) were added. TMB was used as substrate and the plate was read at 450 nm.

Liquid phase blocking ELISA: Orbital venous blood of immunized mice was collected and the serum was isolated by centrifugation. Specific antibodies in the serum of immunized mice were detected and analyzed using the FMD type-O LPBE antibody detection kit (Lanzhou Shouyan Biotechnology Co., Ltd.).

Double sandwich ELISA: Serum interferon (IFN)-γ and interleukin (IL)-4 on day 28 after immunization were detected by mouse IFN-γ and IL-4 ELISA kits (Shenzhen Xin Bosheng Biological Technology Co., Ltd.).

Serum of mice immunized for 28 days was inactivated at 56°C for 30 min, dilute the serum to be tested with cell culture medium containing 2% fetal bovine serum (FBS) (v/v) (inactivated) and set negative serum control and positive serum control. The FMDV strain O/China99 was diluted to 200 TCID₅₀/0.1 ml, and the same volume was mixed with the diluted serum and incubate at 37°C for 1 h. Each well was seeded with 1 × 10⁶ BHK-21 cells and incubated for 72 h in a temperature chamber containing 5% CO₂ at 37°C. Cytopathic effects were then observed under an inverted microscope. The serum dilution that protected 50% of cells against cytopathic changes was calculated by the Reed-Muench method.

On day 28 after immunization, mice were sacrificed after anesthesia with ether and spleen lymphocytes were aseptically isolated from mice. Cells were resuspended in RPMI-1640 containing 10% FBS (v/v) and 1% penicillin–streptomycin solution (v/v). The cells added to a 96-well cell culture plate (Corning, 10⁶ cells/well, 100 μl), 1 μl Concanavalin A (Con A) (Sigma, 1 mg/ml), and type-O FMD VLPs (1 mg/ml) were added to the wells as positive control and a blank control group containing only culture medium but not cells was established. After 72 h of culture, the culture supernatants before and after the stimulation of lymphocytes in each group were analyzed for IFN-γ content. The OD at 490 nm was detected, and finally, the stimulation index (SI) was calculated as follows:

Body weight changes of the mice in each group after immunization were detected. On day 28, mouse lungs were aseptically removed, observed, photographed, and weighed. The lung index was calculated as follows:

Lung tissues were removed and fixed in 4% paraformaldehyde, dehydrated in an alcohol gradient, transparent with toluene, embedded in paraffin, sectioned, and stained with hematoxylin and eosin (H&E) and Masson trichrome.

H&E: Sections were dewaxed to water, stained with hematoxylin for 10 min, differentiated with 1% hydrochloric acid alcohol for several seconds, stained with eosin for 5 min, dehydrated with gradient alcohol, transparent with toluene, and sealed with neutral gum. Histopathological changes were observed, recorded, photographed with a microscope (BA210Digital, USA), and lung inflammation was scored (National Toxicology Program, 2000, 2007).

Masson trichrome: Sections were dewaxed to water, treated with Bouin’s solution at 37°C for 2 h, stained with celestine blue droplets for 2–3 min, stained with Mayer’s hematoxylin for 3 min, differentiated with acidic ethanol for several seconds, and stained with Ponceau Fuchsin staining solution for 10 min. Phosphorous molybdic acid solution was used for color separation for 10 min followed by aniline blue staining solution for 5 min, weak acid solution for 2 min, rapid dehydration with 95% alcohol, dehydration with anhydrous ethanol, clear xylene, and sealing with neutral gum. The distribution and composition of lung fibers were observed; fibrils, mucus, and cartilage were blue, muscle fibers, cellulose, muscle, glia, and cytoplasm were red, red blood cells were orange-red, and nuclei were dark-blue. Slices were imaged under a microscope (BA210Digital, USA) at 400× magnifications. Three fields of view were randomly selected and Image-Pro Plus (Media Cybernetics, USA) was used to measure the fibrous tissue area in the collected images. Fibrous tissue was calculated as follows (de Castro et al., 2017):

GraphPad Prism 8.0 was used for graphing the data in each group. All statistical analyses were conducted with one-way analysis of variance (ANOVA) by SPSS 22.0. p < 0.05 (*) were considered statistically significant, p < 0.01 (**) were considered highly significant.

The FMD VLPs were assembled from three full-length foot-and-mouth disease structural proteins VP0 (33 kDa), VP3 (24 kDa) and VP1 (23 kDa) in vitro. SDS-PAGE and Western blot analysis (Figure 1A) showed that the apparent molecular weight of the three FMD structural proteins was consistent with the expected size. Hydration diameters of the two complexes were measured by dynamic light scattering (DLS). As shown in Figure 1B, the hydrodynamic diameter of VLPs was approximately 37 nm, while the diameter of CaP-VLPs was 92 nm. The morphology of VLPs and CaP-VLPs was observed using TEM, the ultrastructure of CaP-VLPs showed a dense nanoparticle structure. The particle size of CaP-VLPs was significantly higher than that of VLPs (Figures 1C,D), which is consistent with the results observed by DLS, indicating that calcium and phosphorus elements successfully encapsulated VLPs. The above results confirm the successful acquisition of VLPs and CaP-VLPs.

To determine whether intranasal immunization elicits an immune response at multiple mucosal sites, mice were immunized three times and local mucosal immune responses were analyzed by indirect ELISA (Figure 2A). On days 17 and 21, specific sIgA was successfully detected in fecal extract, bronchoalveolar lavage fluid, and enteral lavage fluid in all mice inoculated with CaP-VLPs (Figure 2B). All the experimental groups vaccinated with CaP-VLPs produced antibodies anti inactivated FMDV. The specific sIgA in the small intestine and stool mainly concentrated on day 17 after immunization and decreased on day 21, while the specific sIgA detected in alveolar lavage fluid could still maintain a high level on day 21.

Nasopharyngeal associated lymphoid tissue is a unique induction site for B cell response and plasma cell generation, allowing the intranasal vaccine administration to effectively induce systemic antibodies (Brandtzaeg, 2011). To evaluate the systemic humoral immune response elicited by intranasal vaccination, we measured FMDV-specific antibodies in mouse serum by LPBE. In natural hosts of FMD, such as cattle, sheep, and pigs, the serum antibody titer measured by LPBE is positively correlated with the results of animal challenge protection, LPBE is considered to be the gold standard for evaluating the immune effect of FMD vaccine (Lu et al., 2008; Ma et al., 2011). Anti-FMDV-specific serum IgG was produced in each immunization group, and the high level of IgG titer was maintained for up to 7 weeks. It should be noted that the serum of the mice already contained high levels of antibodies on day 14, suggesting that activated lymphocytes had been efficiently circulating in vivo after the first two intranasal immunizations (Figure 3A). In addition, serum IgG of most immune groups reached the highest level on day 28. The antibody titer of the CaP-VLPs-S group was significantly higher than that of the CaP-VLPs group, and there was no significant difference between the antibody titers of other adjuvant groups and the CaP-VLPs group (Figure 3B). Next, the neutralization effect of mouse serum antibodies against the specific strain of type-O FMDV was tested by virus neutralization assay. Figure 3C showed that the neutralizing antibody titers induced by the CaP-VLPs vaccine combined with S adjuvant were significantly higher than those induced by other experimental groups. The above results supported that the VLPs vaccine biomineralized by calcium phosphate can induce an effective humoral immune response after mucosal immunization. IL-4 is a Th2-type cytokine that can promote the proliferation and activation of B cells and differentiation into plasma cells, and is a key regulator of humoral and adaptive immune responses (Yokota et al., 1986). On day 28, mice immunized with CaP-VLPs in combination with adjuvant induced higher levels of serum IL-4 than CaP-VLPs alone (Figure 3D).

We further evaluated IgG subtype responses in mice immunized with CaP-VLPs with or without adjuvant to compare the strength of Th1 and Th2 responses. Levels of the IgG1 and IgG2a antibody isotypes reflect Th2 immunity and Th1 immunity, respectively. CaP-VLPs with adjuvant effectively increased the IgG1 and IgG2a responses (Figures 3E,F).

To determine whether intranasal inoculation of CaP-VLPs can induce a specific cellular immune response, spleen lymphocytes were collected on day 28 after the first dose and re-stimulated with the VLPs ex-vivo, followed by detection of cell proliferation to compare the intensity of cellular immune responses elicited by each immune group. The immunization group with S or I adjuvant effectively induced proliferation of lymphocytes under antigen stimulation (Figure 4A). IFN-γ is a marker cytokine of Th1-type cells and is considered a key regulator of cellular immune responses (Otani et al., 2021). Therefore, we detected the IFN-γ content in serum and the supernatant of spleen lymphocytes collected after antigen stimulation. The serum IFN-γ content in the immunized group was higher than that in the control group, indicating that CaP-VLPs can effectively induce the immune response. However, there was no significant difference between the experimental groups (Figure 4B). The stimulated culture supernatant of spleen lymphocytes contained a large amount of IFN-γ, further demonstrating that the CaP-VLPs could induce cellular immune responses through nasal immunization, and adjuvant enhanced the strength of this immune response (Figure 4C).

During the immunization period, the animals did not show any signs of discomfort or significant weight loss (Figure 5A). There was no significant difference in lung weight (Figure 5B) or organ index (Figure 5C) between the experimental and control group. Thus, the vaccine did not have a significant impact on the growth and development of mice after intranasal inoculation.

Gross observation revealed that the lungs of mice in each group were uniform in texture, rose-colored, elastic to touch, and showed no obvious signs of lesions (Figure 6A). H&E staining and histopathological score showed that the lungs in each group had normal and complete histological structure, with clean alveoli, alveolar ducts, and alveolar cavities. The morphology of alveolar epithelial cells was normal, no obvious degeneration or necrosis was observed, and no obvious inflammatory cell infiltration was observed in the interstitium (Figure 6B; Table 1). The connective tissue network of the lung is mainly composed of collagen fibers, which are mainly located in the loose connective tissue below the epithelium and oriented circularly or obliquely around the bronchiole. Using Masson trichrome staining to assess collagen fiber expression in lung tissue, no significant difference in the expression level of collagen fibers in the lung tissue of each group of animals was observed (Figures 6C,D).