Promotion of Cellular and Humoral Immunity against Foot-and-Mouth Disease Virus by Immunization with Virus-Like Particles Encapsulated in Monophosphoryl Lipid A and Liposomes

Virus-like particles (VLPs) have emerged as promising vaccine candidates against foot-and-mouth disease (FMD). However, such vaccines provide a relatively low level of protection against FMD virus (FMDV) because of their poor immunogenicity. Therefore, it is necessary to design effective vaccine strategies that induce more potent immunogenicity. In order to investigate the means to improve FMD VLP vaccine (VLPFMDV) immunogenicity, we encapsulated VLPs (MPL/DDA-VLPFMDV) with cationic liposomes based on dimethyldioctadecylammonium bromide (DDA) and/or monophosphoryl lipid A (MPL, TLR4 agonist) as adjuvants. Unlike inactivated whole-cell vaccines, VLPFMDV were successfully encapsulated in this MPL/DDA system. We found that MPL/DDA-VLPFMDV could induce strong cell-mediated immune responses by inducing not only VLP-specific IFN-γ+CD4+ (Th1), IL-17A+CD4+ (Th17), and IFN-γ+CD8+ (activated CD8 response) T cells, but also the development of VLP-specific multifunctional CD4+ and CD8+ memory T cells co-expressing IFN-γ, TNF-α, and IL-2. In addition, the MPL/DDA-VLPFMDV vaccine markedly induced VLP-specific antibody titers; in particular, the vaccine induced greater Th1-predominant IgG responses than VLPFMDV only and DDA-VLPFMDV. These results are expected to provide important clues for the development of an effective VLPFMDV that can induce cellular and humoral immune responses, and address the limitations seen in current VLP vaccines for various diseases.


Introduction
Foot-and-mouth disease virus (FMDV) can cause highly contagious foot-and-mouth disease (FMD) in cloven-hoofed livestock, particularly cattle, sheep, goats and pigs [1,2]. This virus belongs to a prototypical member of the Aphthovirus of the Picornaviridae family, and is classified into seven distinct serotypes (O, A, C, SAT 1 to 3, and Asia 1), as well as >65 subtypes [3,4]. Given that

Preparation of Vaccine Formulations and Immunization of Mice
C57BL/6 mice (female, 6-7 weeks of age) were purchased from Orient Bio Inc. (Seoul, Korea). Mice were acclimatized to the following controlled conditions: temperature (25 ± 2 • C), humidity (55 ± 5%), and 12 h light/dark cycle at the Central Animal Research Laboratory of the Korea Atomic Energy Research Institute (KAERI, Jeongeup, Korea). The animal experiments were approved by the Institutional Animal Care and Use Committee (KAERI-IACUC-2019-018) and were performed in strict compliance with the guidelines prescribed by the committee. For adjuvant formulation, MPL and DDA were obtained from Sigma-Aldrich (St. Louis, MO, USA). DDA can be applied to the preparation of cationic liposomes, as described previously [25]. Briefly, DDA (10 mg/mL) mixed with distilled water (DW) was heated at 80 • C until micelles formed, and then cooled to room temperature (RT). MPL mixed with DW containing 0.2% trimethylamine was heated at 70 • C for 30 seconds, then sonicated for 30 seconds, and these steps were repeated twice. MPL was subsequently mixed with DDA immediately before use (

VLP-Specific IgG Isotype ELISA
The VLP-specific IgG, IgM, IgG1 and IgG2a responses in the serum from antigen-immunized mice were analyzed by sandwich ELISA. Ninety-six-well plates were coated with 1 µg/mL VLP at 4 • C for 24 h. The plates were washed 5 times with PBS containing 0.2% Tween 20 (PBS-T), and then blocked with 5% BSA in PBS at RT for 1 h. After blocking, diluted serum was added to each well and incubated at RT for 1 h. Unbound antibodies were removed by washing with PBS-T, and then goat anti-mouse IgG-HRP (Sigma-Aldrich), goat anti-mouse IgM-HRP (Sigma-Aldrich), goat anti-mouse IgG1-HRP (Sigma-Aldrich), or IgG2a-HRP (Sigma-Aldrich) were added to the wells, and incubated for 30 min at RT. After washing the plates 5 times with PBS-T, 100 µL of TMB substrate reagent (BD Biosciences, Franklin Lakes, NJ, USA) was added. When colors developed, 50 µL of 2 N H 2 SO 4 was added, and optical absorbance was measured at 450 nm using a Victor X3 light plate reader (Perkin-Elmer, Waltham, MA, USA).

Statistical Analysis
All analyses were repeated at least twice. The level of significance for comparisons between samples was determined by unpaired Student's t-tests (between two groups), and one-way ANOVA followed by Dunnett's multiple comparison test (between three or more groups) using statistical software (GraphPad Prism, version 5; San Diego, CA, USA). Results are expressed as means ± standard deviation (S.D.). Values of * p < 0.05, ** p < 0.01 and *** p < 0.001 were considered to be statistically significant.

Higher VLP-Specific T Cell Immunity Induced by Formulating in MPL/DDA
We first performed a DLS analysis on five types (DDA only; MPL/DDA: MPL and DDA formulation; VLP FMDV only; DDA/VLP FMDV : VLP FMDV formulated with DDA; MPL/DDA/VLP FMDV : VLP FMDV : VLP FMDV formulated with DDA and MPL) of formulated VLP FMDV vaccines to confirm the particle sizes and successful formulation. As shown in Figure 1A, VLP FMDV had an average diameter of 44.6 nm. In addition, DDA ( Figure 1B) and MPL/DDA ( Figure 1C) without VLP FMDV showed average diameters of 1950 nm and 1900 nm, respectively. After formulating with VLP, the average diameters of DDA-VLP FMDV ( Figure 1D,F) and MPL/DDA/VLP FMDV ( Figure 1E,F) increased significantly to 2402 nm and 2442 nm, respectively. There was no significant effect on particle size between DDA-VLP FMDV and MPL/DDA-VLP FMDV ( Figure 1F). These results strongly suggested that VLPs were successfully encapsulated into DDA or MPL/DDA liposomes, which led to the increase in liposome particle size. blocked with 5% BSA in PBS at RT for 1 h. After blocking, diluted serum was added to each well and incubated at RT for 1 h. Unbound antibodies were removed by washing with PBS-T, and then goat anti-mouse IgG-HRP (Sigma-Aldrich), goat anti-mouse IgM-HRP (Sigma-Aldrich), goat anti-mouse IgG1-HRP (Sigma-Aldrich), or IgG2a-HRP (Sigma-Aldrich) were added to the wells, and incubated for 30 min at RT. After washing the plates 5 times with PBS-T, 100 μL of TMB substrate reagent (BD Biosciences, Franklin Lakes, NJ, USA) was added. When colors developed, 50 μL of 2 N H2SO4 was added, and optical absorbance was measured at 450 nm using a Victor X3 light plate reader (Perkin-Elmer, Waltham, MA, USA).

Statistical Analysis
All analyses were repeated at least twice. The level of significance for comparisons between samples was determined by unpaired Student's t-tests (between two groups), and one-way ANOVA followed by Dunnett's multiple comparison test (between three or more groups) using statistical software (GraphPad Prism, version 5; San Diego, CA, USA). Results are expressed as means ± standard deviation (S.D.). Values of * p < 0.05, ** p < 0.01 and *** p < 0.001 were considered to be statistically significant.

Generation of VLP-Specific Multifunctional CD8 + T Cells by MPL/DDA-VLP FMDV
We next analyzed the functional composition of VLP-specific multifunctional CD8 + T cells. Multifunctional viral-specific CD8 + T cells have been shown to be important and effective immune cells during viral infection [29,30]. Unlike single-positive CD4 + T cells, there was no increase in TNF-α or IL-2 single-positive CD8 + cells, and IFN-γ single-positive CD8 + cells were only increased in VLP FMDV -vaccinated groups (G4, G5, G6). Among these, the MPL/DDA-VLP FMDV group (G6) had the highest proportion of VLP-specific IFN-γ single-positive CD8 + T cells compared to the other groups (G1-G5). Surprisingly, the VLP FMDV only group (G4) showed no or few double and triple multifunctional CD8 + T cells, while the G5 and G6 groups had significantly higher levels of those cells than the G4 group. In addition, we found that MPL/DDA-VLP FMDV (G6) represented the highest frequency of IFN-γ + TNF-α + (G4 vs. G6; up to 86-fold, G5 vs. G6; up to 2.26-fold), IFN-γ + IL-2 + double-positive (G4 vs. G6; up to 75-fold, G5 vs. G6; up to 1.8-fold), and triple-positive CD44 + CD8 + T cells (G4 vs. G6; up to 115.6-fold, G5 vs. G6; up to 2.8-fold) compared to the other groups ( Figure 4A,B). Based on the above data, we analyzed the relative proportions of single-, double-and triple-positive CD44 + CD8 + T cells ( Figure 4C). We found only single-positive multifunctional CD8 + cells in the G1, G2, G3 and G4 groups. The proportion of double-and triple-positive CD44 + CD8 + T cells dramatically increased in the G5 and G6 groups. Together, these data show that the addition of MPL and DDA adjuvants to the VLP FMDV vaccine can effectively induce superior multifunctional T cell responses. in VLPFMDV-vaccinated groups (G4, G5, G6). Among these, the MPL/DDA-VLPFMDV group (G6) had the highest proportion of VLP-specific IFN-γ single-positive CD8 + T cells compared to the other groups (G1-G5). Surprisingly, the VLPFMDV only group (G4) showed no or few double and triple multifunctional CD8 + T cells, while the G5 and G6 groups had significantly higher levels of those cells than the G4 group. In addition, we found that MPL/DDA-VLPFMDV (G6) represented the highest frequency of IFN-γ + TNF-α + (G4 vs. G6; up to 86-fold, G5 vs. G6; up to 2.26-fold), IFN-γ + IL-2 + double-positive (G4 vs. G6; up to 75-fold, G5 vs. G6; up to 1.8-fold), and triple-positive CD44 + CD8 + T cells (G4 vs. G6; up to 115.6-fold, G5 vs. G6; up to 2.8-fold) compared to the other groups ( Figure  4A,B). Based on the above data, we analyzed the relative proportions of single-, double-and triple-positive CD44 + CD8 + T cells ( Figure 4C). We found only single-positive multifunctional CD8 + cells in the G1, G2, G3 and G4 groups. The proportion of double-and triple-positive CD44 + CD8 + T cells dramatically increased in the G5 and G6 groups. Together, these data show that the addition of MPL and DDA adjuvants to the VLPFMDV vaccine can effectively induce superior multifunctional T cell responses.   Th1-predominant IgG isotypes, such as IgG2a, IgG2b, IgG2c and IgM, are more likely to  induce stronger protective effects against viral infection, including FMDV, than Th2-predominant  IgG responses [12,31]. Thus, we evaluated whether the formulations of DDA and/or MPL could promote the generation of humoral Th1-and/or Th2-predominant immune responses. As shown in Figure 5, the MPL/DDA-VLP FMDV group (G6) had significantly higher levels of total IgG, IgM, and Th1-predominant IgG2a than the G4 and G5 groups. In contrast, the Th2-predominant IgG1 level of G6 was similar to that of G5 ( Figure 5). These findings suggest that the MPL and DDA formulations induced greater VLP-specific Th1-predominant IgG responses.

VLP-Specific Antibody Responses Elicited by Immunization with MPL/DDA-VLP FMDV
Vaccines 2020, 8, 633 9 of 14 = 5 mice/group) shown are representative of two independent experiments. * p < 0.05, ** p < 0.01, *** p < 0.001. n.s.: not significant. Th1-predominant IgG isotypes, such as IgG2a, IgG2b, IgG2c and IgM, are more likely to  induce stronger protective effects against viral infection, including FMDV, than Th2-predominant  IgG responses [12,31]. Thus, we evaluated whether the formulations of DDA and/or MPL could promote the generation of humoral Th1-and/or Th2-predominant immune responses. As shown in Figure 5, the MPL/DDA-VLPFMDV group (G6) had significantly higher levels of total IgG, IgM, and Th1-predominant IgG2a than the G4 and G5 groups. In contrast, the Th2-predominant IgG1 level of G6 was similar to that of G5 ( Figure 5). These findings suggest that the MPL and DDA formulations induced greater VLP-specific Th1-predominant IgG responses. Figure 5. Analysis of VLP-specific serum antibody isotype. Mice were immunized twice with VLPFMDV in different adjuvant combinations. Two weeks after final immunization, serum from five mice in each group was obtained, and the VLP-specific IgG, IgM, IgG1 and IgG2a were analyzed using ELISA. Data from one of two independent experiments are shown. * p < 0.05, ** p < 0.01, *** p < 0.001. n.s.: not significant.

Discussion
FMD is one of the most economically devastating veterinary diseases worldwide [32]. Although a chemically inactivated whole FMD vaccine has been widely used and is known to provide a reduction in FMD prevalence in endemic areas, there is a clear need for a new generation of FMD vaccines to improve overall immunogenicity and safety [33]. Among the several types of vaccines employed for veterinary and human needs, VLPFMDV vaccines have been proven to show reliable immunogenicity and safety in pre-clinical and clinical studies [34][35][36]. Since VLPs can be manufactured in biosafety level 1 (BSL1) facilities, they are also certainly more safe and economical than the current inactivated FMD vaccine [37]. Previous studies have shown that VLPFMDV can be stably expressed and packaged using an E. coli system, but this has not yet been commercialized because of the lack of exceptionally high immunogenicity compared to current vaccines [12,17]. To improve vaccine immunogenicity, several adjuvant systems have been formulated for use with Figure 5. Analysis of VLP-specific serum antibody isotype. Mice were immunized twice with VLP FMDV in different adjuvant combinations. Two weeks after final immunization, serum from five mice in each group was obtained, and the VLP-specific IgG, IgM, IgG1 and IgG2a were analyzed using ELISA. Data from one of two independent experiments are shown. * p < 0.05, ** p < 0.01, *** p < 0.001. n.s.: not significant.

Discussion
FMD is one of the most economically devastating veterinary diseases worldwide [32]. Although a chemically inactivated whole FMD vaccine has been widely used and is known to provide a reduction in FMD prevalence in endemic areas, there is a clear need for a new generation of FMD vaccines to improve overall immunogenicity and safety [33]. Among the several types of vaccines employed for veterinary and human needs, VLP FMDV vaccines have been proven to show reliable immunogenicity and safety in pre-clinical and clinical studies [34][35][36]. Since VLPs can be manufactured in biosafety level 1 (BSL1) facilities, they are also certainly more safe and economical than the current inactivated FMD vaccine [37]. Previous studies have shown that VLP FMDV can be stably expressed and packaged using an E. coli system, but this has not yet been commercialized because of the lack of exceptionally high immunogenicity compared to current vaccines [12,17]. To improve vaccine immunogenicity, several adjuvant systems have been formulated for use with VLP FMDV [12]. In order to enhance the cellular immunity of the VLP FMDV vaccine, we used DDA, a liposome-type adjuvant, to encapsulate VLP FMDV . In addition, MPL as an adjuvant was combined with DDA liposomes to increase VLP-specific multifunctional T cell responses.
Liposomes are an attractive delivery system that has been used to induce the effective uptake of various vaccine candidates, such as proteins, peptides, and genes (DNA and RNA), for presentation to APCs such as dendritic cells and macrophages [38][39][40]. Furthermore, recently, researchers have investigated the incorporation of synthetic TLR agonists, such as MPL and CpG oligodeoxynucleotides, into the membrane bilayer of liposomes to enhance their adjuvant effect [41][42][43]. MPL, a detoxified bacterial lipopolysaccharides, has been shown to have beneficial effects in a range of vaccines, including for allergies, cancer and pathogens [44][45][46]. For example, in a set of studies evaluating the impact of adjuvants on vaccine immunogenicity, Mycobacterium tuberculosis (Mtb) antigen ESAT-6, which has a lower immunogenicity, was able to induce strong antigen-specific Th1 responses, and high titers of antigen-specific IgG1 and IgG2b antibodies, when combined with MPL/DDA [47,48]. In addition, the MPL/DDA formulation has also been reported to activate antigen-specific cytotoxic T-lymphocytes (CTL or activated CD8 + T cells) and humoral responses to the malaria sporozoite antigen [41]. However, the virus is too large to encapsulate in liposomes, and it is very difficult to apply to the liposome system because of the irregular aggregation of inactivated viruses during the inactivation process. Although MPL-based liposome adjuvants are a promising system to enhance the cellular immunogenicity of protein-or peptide-based vaccines, we first reported here that this adjuvant system is attractive for enhancing VLP FMDV -induced humoral and cellular immune responses. In fact, VLP FMDV produced in E. coli has 44.6 nm diameter, which is much smaller than other viral VLPs, such as influenza VLPs (70-200 nm) and human papillomavirus VLPs (55-65 nm) [13,49,50]. Thus, the smaller-sized VLP FMDV is likely to be an important factor in liposome encapsulation.
To date, CD4 + and CD8 + T cells capable of co-producing IFN-γ, TNF-α and IL-2 (namely multifunctional CD4 + and CD8 + T cells) have been considered to be important for controlling various viral infections [49,50]. Darrah et al. reported that T cells co-producing IFN-γ, TNF-α and IL-2 cytokines have a greater capacity to aid cognate cells (effector and/or memory cells against pathogens) compared with single-or double-cytokine-producing T cells [28]. In particular, IFN-γ, TNF-α and IL-2 produced by T cells were recently shown to be associated with protection against FMDV infection [51]. For example, inactivated FMDV-vaccinated mice receiving IFN-γ, TNF-α and IL-2 as vaccine adjuvants exhibited a 100% survival rate, whereas inactivated FMDV-only vaccinated mice exhibited a 40% survival rate, which was correlated with FMDV-specific antibody titers, and with levels of memory B cells, memory γδ T cells, and effector/memory T cells. We also demonstrated that the frequency of multifunctional CD4 + and CD8 + T cells increased in VLP vaccines (MPL/DDA-VLP FMDV ) formulated with MPL/DDA. Upon stimulation with VLPs, MPL/DDA-VLP FMDV -immunized mice displayed significantly greater levels of VLP-specific CD4 + CD44 + and CD8 + CD44 + memory T cells capable of producing single-(in only CD4 + T cells), double-(particularly IFN-γ + TNF-α + and IFN-γ + IL-2 + T cells) and triple-positive effector cytokines in the spleen, thereby indicating that the MPL/DDA formulation can be included in interventions capable of promoting multifunctional T cells to control FMDV. Although we addressed much important evidence demonstrating the enhancement of cellular and multifunctional immune responses by the encapsulation of VLP FMDV , several additional experiments should be performed to support these results. For example, mice strain-, age-, and sex-dependent differences in MPL/DDA-VLP FMDV vaccine-induced immunity should be investigated to clarify the immunogenicity and vaccine efficacy of MPL/DDA-VLP FMDV . In addition, the efficacy of the MPL/DDA-VLP FMDV vaccine against that of a currently inactivated vaccine must be compared by using a challenge animal model, which might explain the advantages of MPL/DDA-VLP FMDV vaccine-induced immunity.
Although the protective roles of Th1 and activated CD8 + T cells against FMDV infection have been extensively studied [12,27], relatively little is known concerning the role of the Th17 response. Since IL-17-producing cells are critical for host defense through controlling innate and adaptive immunity against infectious diseases in the mucosa, it might be important when FMDV invades animals through mucosal surfaces [51,52]. For example, Lee et al. showed that inoculation with an inactivated FMDV vaccine and IL-23, which is essential for Th17 cell differentiation, induces strong memory B cell expansion, confers protective immunity, and imparts significant protection upon a murine model against FMDV infection [53]. We found that VLP FMDV alone had a lesser or no impact on Th17 activation, but one formulated with MPL/DDA could induce significant levels of VLP-specific IL-17A-producing CD4 + T cells. The protective immune response of VLP against the Asia 1 serotype used in this study was certainly high. Since VLPs have highly strain-specific immunological properties, although an adjuvant effect of MLP/DDA may somehow contribute to the activation of T cells, we should not expect pan-serotype specific effects against multiple FMDV strains other than Asia 1 serotype by MLP/DDA/VLP FMDV . However, our present study provided the proof-of-concept for developing an immunologically enhanced and effective FMDV vaccine by employing VLP and MLP plus DDA systems. The subsequent study will be focused on the development of multi-serotype effective FMDV vaccines by including VLPs from multiple prevalent FMDV serotypes.

Conclusions
Our present data demonstrated that the MPL/DDA formulation shows potential as an effective immunostimulatory adjuvant for VLP FMDV vaccines by satisfying the following criteria: (1) the ability to induce strong antigen-specific Th1 and Th17 immune responses; (2) showing remarkable generation of antigen-specific multifunctional CD4 + and CD8 + memory T cells; and (3) showing excellent increases in antigen-specific antibody titers. Thus, MPL/DDA might be an excellent adjuvant for VLP FMDV vaccines, especially considering the otherwise low immunogenicity of VLP FMDV vaccine candidates against FMDV.