Efficacy of genotype-matched vaccine against re-emerging genotype V Japanese encephalitis virus

ABSTRACT

Japanese encephalitis (JE), caused by the Japanese encephalitis virus (JEV), is a highly threatening disease with no specific treatment. Fortunately, the development of vaccines has enabled effective defense against JE. However, re-emerging genotype V (GV) JEV poses a challenge as current vaccines are genotype III (GIII)-based and provide suboptimal protection. Given the isolation of GV JEVs from Malaysia, China, and the Republic of Korea, there is a concern about the potential for a broader outbreak. Under the hypothesis that a GV-based vaccine is necessary for effective defense against GV JEV, we developed a pentameric recombinant antigen using cholera toxin B as a scaffold and mucosal adjuvant, which was conjugated with the E protein domain III of GV by genetic fusion. This GV-based vaccine antigen induced a more effective immune response in mice against GV JEV isolates compared to GIII-based antigen and efficiently protected animals from lethal challenges. Furthermore, a bivalent vaccine approach, inoculating simultaneously with GIII- and GV-based antigens, showed protective efficacy against both GIII and GV JEVs. This strategy presents a promising avenue for comprehensive protection in regions facing the threat of diverse JEV genotypes, including both prevalent GIII and GI as well as emerging GV strains.

KEYWORDS:

• Japanese encephalitis virus
• genotype V
• recombinant vaccine; bivalent vaccine
• neutralizing antibody

Introduction

The Japanese encephalitis virus (JEV), a flavivirus carried by Culex mosquitoes, is the causative agent of Japanese encephalitis (JE) in endemic countries (Table 1) [1]. JEV is a spherical, enveloped, positive-sense single-stranded RNA virus with gene lengths approximately 11 kb and a size of around 50 nm [2]. JEVs are classified into five genotypes, I to V, based on variation in genetic sequences [3]. Since the first isolation of the genotype III (GIII) Nakayama from the cerebrospinal fluid (CSF) of a Japanese patient in 1935 [4], GIII has served as a primary genotype of JEV, maintaining dominance across most of Asia [5]. More recently, genotype I (GI) has surpassed GIII as the most commonly identified genotype in several Asian countries, including China, India, Japan, Malaysia, Republic of Korea (ROK), Thailand, and Vietnam [6]. Genotype V (GV) has been reported since 1952, but its global occurrence is relatively low compared to GI and GIII. However, since 2010, it has successfully established itself as the dominant strain in the ROK [7,8]. The GV Muar was isolated from JE patient in Malaysia, and the GV K15P38 was isolated from a patient in ROK in 2015 [8]. These two strains constitute the only GV isolates reported to be isolated from the CSF of human patients so far. Studies have also demonstrated the continuous presence of GV JEVs in mosquito pools in the ROK [9–12], with the GV Sangju being isolated from a mosquito pool in 2020 [12,13].

Vaccination is crucial to prevent JE because there is no developed cure. All commercially approved JE vaccines (JE-VAX) have been developed based on the GIII strains [14]. Vaccination with JE-VAX can generate neutralizing antibodies against JE genotypes I to IV [15], but they do not effectively induce the production of neutralizing antibodies against the GV Muar and GV XZ0934, a JEV isolate from mosquito pool in China [16,17]. GI-based vaccine was also reported to generate relatively low cross-neutralizing antibodies against the GV Muar compared to JEVs belonging to other genotypes [18]. Therefore, the development of a vaccine specifically targeting GV is necessary for the effective prevention of GV JEV infection.

The envelope (E) protein covers the surface of JEV and is considered a primary target in vaccine development [19,20]. The E protein is composed of three subdomains, E domain (ED) I to III, and EDIII is known to be a major target involved in the neutralizing antibody response [20]. Indeed, several attempts at vaccine development using the EDIII of JEV have been successful in preventing experimental encephalitis [21–25]. In our earlier study, we also developed a vaccine candidate that presents five EDIII monomers derived from the GIII Nakayama using cholera toxin B subunit (CTB) as a pentameric scaffold [25]. CTB naturally forms a pentamer and possesses inherent adjuvanticity, making it a useful structural component for designing nanoparticle vaccines. When tested in a mouse model, the pentameric CTB-EDIII induced neutralizing antibodies against the antigen-matched GIII Nakayama, but it remained unclear whether CTB-EDIII could produce cross-reactive antibodies against GV strains. As previous research has revealed inefficient anti-GV JEV seroconversion with GIII-based vaccines [16,17], and sequence analysis has further demonstrated that the gene coding for the E protein of the GV Muar is phylogenetically distinct from the other four genotypic strains of JEVs [3,7,26], we anticipate that a GV-derived antigen should be included in the vaccine to achieve optimal protection against GV JEV strains.

In the current study, we utilize Escherichia coli for expression system. However, EDIII expressed in the E. coli expression system is insoluble, requiring an additional step for solubilization [22,23,25]. To achieve soluble expression of CTB-EDIII in E. coli, we employed the RNA-mediated chaperone technique called “Chaperna,” (chaperone + RNA) which can mitigate structural misfolding [27–29]. The conjugation of the Chaperna (Cha) domain with the CTB-EDIII fusion protein resulted in efficient soluble expression in E. coli, forming a self-assembled pentameric nanostructure. Overall, we confirmed that our GV-based recombinant vaccine can provide superior protection against GV JEV strains in mice. In addition, we evaluated a bivalent vaccine formulation that provides comprehensive protection against lethal challenges from both GIII and GV JEV strains.

Materials and methods

Viruses

The GIII Nakayama (GenBank accession No. EF571853) was kindly provided by International Vaccine Institute, ROK. The GV K15P38 and GV Sangju were obtained from the National Culture Collection for Pathogens (NCCP) in the ROK. Each strain, K15P38 and Sangju, is assigned the codes NCCP 43279 (GenBank accession No. RO500440) and NCCP 43413 (GenBank accession No. NP478074), respectively. They will be referred to as GV 43279 and GV 43413 in this paper (Figure 1).

Protein structure prediction

The AlphaFold structure prediction model was employed for the design of vaccine antigens, and AlphaFold2-ColabFold was utilized to predict the pentameric structure [30,31]. The reliability of structure predictions was assessed using predicted local distance difference tests (pLDDT) scores [32]. To compare the three-dimensional (3D) structures of the predicted proteins, we aligned the models using the Matchmaker tool in ChimeraX software and quantified structural similarity by calculating the root mean square deviation (RMSD) of atomic positions and measuring the average distance between overlapping protein atoms. The RMSD calculation focused exclusively on C-alpha atoms, with an iteration cutoff of 2 Å for pruning residues. The Needleman-Wunsch algorithm was employed for sequence pair alignment during RMSD calculation. An RMSD value of less than 2 Å is generally interpreted as indicative of a high structural resemblance between the predicted and experimental structures.

Recombinant vaccine antigens

Replacing the N-terminal appended domain (hRID) of human Lysyl-tRNA synthetase (hKRS) in the previous study [33,34], we employed the WHEP domain from human Glutamyl-prolyl-tRNA synthetase (hEPRS) as a Cha domain, which triggers the folding of the down-stream located EDIII-CTB chimeric antigen in an RNA-interaction dependent manner. The WHEP acronym, derived from the initial letters of tRNA synthetases such as WARS1, HARS1, EPRS, and PARS where this domain is found, is a tRNA interaction domain to facilitate tRNA-mediated folding and assembly of the viral antigen [33,34]. The cloned expression vectors were introduced to Shuffle T7 Express competent E. coli (New England Biolabs, MA, USA). For protein expression, transformed E. coli was cultured at 30°C. The culture was subsequently induced with 0.5 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) and incubated at 30°C for 6 h or at 16°C for 14 h. Cells were harvested and lysed by sonication, followed by purification using Ni-NTA resin (Qiagen, California, USA). To remove the Cha domain, TEV protease (Genscript, NJ, USA) was added at a 1:100 (v/v) ratio and incubated at 30°C for 14 h. Size exclusion chromatography (SEC) was then performed using HiLoad 16/600 Superdex 200 pg (Cytiva, MA, USA). The size of eluted proteins by SEC was estimated by comparing with co-eluted standard size markers. The final concentration of recombinant vaccine antigens was quantified using a BCA protein assay kit (Thermo Fisher Scientific, MA, USA). The purity and the size of monomeric and multimeric form of the purified antigens were analysed in SDS-PAGE, either in reducing/non-reducing conditions or in boiling/non-boiling conditions after addition of SDS buffer.

Table 1. List of abbreviation used in this study.

Abbreviations	Definition
ADE	Antibody Dependent Enhancement
Ag	Antigen
Cha	Chaperna
CTB	Cholera Toxin B Subunit
E	Envelope
G	Genotype
JE	Japanese Encephalitis
JEV	Japanese Encephalitis Virus
JE-VAX	Commercially Approved JE Vaccines
NCCP	The National Culture Collection for Pathogens
pLDDT	Predicted Local Distance Difference Tests
PRNT	Plaque Reduction Neutralization Test
ROK	Republic of Korea
TEV	Tobacco Etch Virus

GM1 binding assay

The 96-well immunoplate (SPL life science, Gyeonggi-do, ROK) was coated with 300 ng/well of ganglioside GM1 (Biosynth, Staad, Switzerland) in a carbonate–bicarbonate buffer (pH 9.6) (Sigma-Aldrich, MO, USA) and incubated at 4°C overnight. The plate was blocked with 1% bovine serum albumin (BSA) (Sigma-Aldrich) in phosphate-buffered saline with 0.05% Tween 20 (PBST) for 3 h at room temperature (RT). The recombinant protein samples were serially diluted 3-fold starting from 300 ng with PBST and then applied to the GM1-coated wells for 4 h at RT. Subsequently, primary anti-6×His Tag antibody (Thermo Fisher Scientific) and secondary HRP-conjugated anti-mouse IgG antibody (Bethyl Laboratories, TX, USA) were incubated for 1 h at RT. Plates were incubated with 3,3′,5,5′ tetramethylbenzidine (TMB) substrate reagent (Thermo Fisher Scientific). The reaction was quenched with the addition of 2 N H₂SO₄, and the optical density was measured at 450 nm using a SYNERGY H1 plate reader (BioTek Instruments, VT, USA) [22].

Mouse experiments

Five-week-old female BALB/c mice were purchased from Orient Bio (Gyeonggi-do, ROK) and kept under specific pathogen-free conditions in The Catholic University of Korea's animal facility. Rigorous examination and approval of all experimental procedures were carried out by the Institutional Animal Care and Use Committee at The Catholic University of Korea (approval No. CUMC- 2022-0279-02). In accordance with the approved protocol, a total of 399 mice were utilized for this study. Mice were immunized intramuscularly with 20 μg of purified recombinant protein or 1/100 of the adult human dose of cell culture-derived inactivated JE-VAX (Green Cross, Gyeonggi-do, ROK). All mice received booster immunization twice with the same dose at two-week intervals. Two weeks after the last immunization, mice were bled from the retro-orbital sinus for enzyme-linked immunosorbent assay (ELISA) or plaque reduction neutralization test (PRNT), and spleens were harvested to prepare splenocytes. To see the protective effect of the vaccine, mice were intravenously infected two weeks after the final immunization with one of the following strains: GIII Nakayama, GV 43279, and GV 43413 at doses of 1×10⁷ pfu, 2×10⁶ pfu, and 2×10⁷ pfu, respectively, using 125 μl for injection, and were observed for 2 weeks. These doses were chosen to induce 20–50% mortality to ensure a comparable level of disease pressure.

Statistics

GraphPad Prism 9.0 (GraphPad Software, CA, USA) was utilized for graph generation and statistical analysis. Two-group comparisons used the unpaired Student's t-test, while one-way ANOVA was applied for three or more groups. Post-hoc analysis involved Fisher's Least Significant Difference. PRNT titres were reported as geometric mean titres (GMT) with standard deviation (SD). Other results were presented as mean ± SD. Survival rates were analysed using the Mantel–Cox log-rank test. A p-value of <0.05 was considered statistically significant.

Figure 1. Flowchart illustrating the process of the design, production, and evaluation of antigens used in this study.

Results

Efficacy of GIII-based JE-VAX against GV strains

We first tested the efficacy of a GIII-based inactivated JE-VAX, approved for human use in ROK, using a mouse model. In order to compare the efficacy of the humoral immune response elicited by JE-VAX against GIII Nakayama and ROK isolates, GV 43279 and GV 43413, mice were administered with PBS or JE-VAX three times at two-week intervals. Mouse sera were obtained two weeks after the final administration, and specific antibody titres were evaluated using ELISA. From the ELISA outcomes derived from serial dilutions, 1:240 was determined as the optimal dilution factor for the presentation of immune response (Figure S1). When 1:240 sera were evaluated for the genotype-matched GIII Nakayama, sera from JE-VAX immunized mice exhibited a 7.5-fold higher absorbance compared to those from mock-immunized mice (Figure 2(A)). However, when the same sera were tested against GV 43279 and GV 43413, the increases were relatively lower, at 4.4- and 4.5-fold, respectively (Figure 2(A)). When neutralizing antibody titres were assessed using PRNT, sera from mock-immunized mice showed no inhibition against any of the JEV strains tested, while sera from JE-VAX immunized mice exhibited significant neutralizing antibody responses against GIII Nakayama (Figure 2(B)). However, JE-VAX did not induce comparable neutralizing antibodies against GV strains (Figure 2(B)), confirming inefficient cross-genotypic protection by the GIII-based vaccine. Next, mock or JE-VAX treated mice were challenged either with GIII Nakayama or GV 43279. JE-VAX immunization elicited 100% protection against GIII Nakayama infection, while mock treatment resulted in 60% mortality (Figure 2(C)). In contrast, JE-VAX failed to induce statistically significant protection against GV 43279, although the mortality rate in JE-VAX immunized mice (40%) was lower compared to mock-treated mice (67%) (Figure 2(C)). These data suggest that current GIII-based vaccines are insufficient to control GV outbreaks, highlighting the need for GV-based vaccines in a comprehensive epidemic prevention.

Figure 2. Efficacy of commercially approved JE-VAX against GIII and GV JEVs. JE-VAX or PBS was administered to mice three times at two-week intervals. Sera were collected two weeks after the final immunization. (A) Specific antibody titres against GIII Nakayama, GV 43279, and GV 43413 (n = 6-10) were determined by ELISA with a serum dilution of 1:240. (B) Neutralizing antibody titres against GIII Nakayama, GV 43279, and GV 43413 (n = 14–19) were determined by PRNT. The red dotted line indicates the detection limit (PRNT₅₀ titre = 10). Seronegative samples were arbitrarily assigned a value of 7. (C) Immunized mice were challenged with GIII Nakayama or GV 43279 (n = 15). **p < 0.01, ***p < 0.001, ****p < 0.0001. Data are pooled from three independent experiments.

Structure prediction of vaccine candidates

EDIII, which contains the receptor-binding domain, plays an essential role in flavivirus infection and is therefore used as a prominent target antigen for vaccine development [21–25,35]. We used a 104 aa-long EDIII sequence that starts from position 297 and extends to position 400 of the whole E protein as an antigen (Figure 3(A)). Although GIII and GV JEVs share E protein structures, they may have distinct microstructures that could affect their antigenic properties due to mutations accumulated during their evolution into separate phylogenetic branches [36]. Indeed, the amino acid sequences of this domain from the GIII Nakayama and GV 43279 exhibited 91.3% homology (Figure 3(A)). Additionally, GV 43413 has a D389G substitution compared to GIII Nakayama and GV 43279 (Figure 3(A)). We chose the GV 43279 sequence to design the GV vaccine candidate, as this strain is significant for being isolated from a human patient.

Figure 3. Prediction of 3D structure of vaccine antigens. (A) Sequence alignment of the EDIII amino acid residues from GIII Nakayama, GV 43279, and GV 43413 JEV strains. (B) Predicted 3D structures of antigens with model confidence represented in a colour spectrum. Model confidence was evaluated using the pLDDT score. (C) RMSD-based 3D alignment of predicted EDIII monomers with pentameric vaccine antigen structures. Superimposition is depicted in contrasting colours for clarity. (D) Comparative 3D alignment of predicted structures among vaccine antigens.

We first employed the AlphaFold-Multimer program to predict the 3D structures of pentameric vaccine antigens, specifically Ag-GIII and Ag-GV. These antigens include the CTB scaffold conjugated with the EDIII antigen sequences (CTB-EDIII) of GIII and GV, respectively (Figure 3(B)). We inserted GSGSGS linker between CTB and EDIII. CTB without the antigen (Ag-neg) was used as a negative control in this study. The pLDDT scores for our modelled vaccine antigens consistently exceeded 70, confirming the reliability of the structural analysis.

As the original antigen structure of EDIII can be altered after conjugation with the CTB scaffold, vaccine antigen candidates were aligned with the predicted EDIII monomer of the matched genotype (Figure 3(C)). The RMSD values for the comparison between the pentameric CTB-EDIII and the matched monomeric EDIII were 0.485 Å for GIII and 0.463 Å for GV, respectively. Our data suggest that, even after modifications for vaccine use, only negligible structural alterations have been introduced to EDIII.

The conjugation of an antigen with a scaffold could potentially disrupt the conformation of the scaffold protein. To further confirm that the vaccine antigens maintain the structural integrity of the CTB scaffold, we compared the predicted pentameric CTB scaffold (Ag-neg) structures with Ag-GIII and Ag-GV (Figure 3(D)). The RMSD values were 0.524 Å for Ag-GIII and 0.352 Å for Ag-GV, respectively, indicating that the structural integrity of the CTB scaffold is well-preserved even after antigen conjugation. The RMSD value for the comparison between the GIII and GV vaccine antigens was relatively high at 0.598 Å, suggesting that the structural differences between these two vaccine antigens lie in the EDIII structure, rather than in the CTB scaffold. The structural distinctions between the GIII and GV EDIII support the necessity of a bivalent JE vaccine to efficiently prevent the co-circulation of emerging GV JEVs.

Vaccine antigen production

We cloned Ag-neg, Ag-GIII, and Ag-GV coding sequence into an expression vector which includes a Cha domain (Figure 4(A)). The vector has TEV cleavage site and a 6×His tag between the Cha domain and antigen. E. coli was transformed with cloned vector and the proteins were dominantly expressed in a soluble form during the culture. After testing various conditions, optimal expression was achieved at 16°C with 0.5 mM IPTG for 14 h (Figure S2). Notably, all three protein constructs (Cha-Ag-neg, Cha-Ag-GIII and Cha-Ag-GV) were produced predominantly as soluble form (compare S(soluble) and P(pellet) fractions in Figure S2(A)), and were present as pentameric assembly (shown in dotted area in Figure S2(B)) where the size of the multimers are consistent with the expected size (Figure S2(C)). The expressed fusion proteins in the form of Cha-CTB-EDIII were designed to allow the removal of the Cha domain through treatment with TEV protease. The soluble fraction of the proteins was incubated with TEV protease, and by confirming both pentamer and monomer size bands under non-reducing and reducing conditions, respectively, we verified that the proteins were expressed to the desired sizes (Figure 4(B)). Furthermore, this process confirmed the successful self-assembly of monomeric antigens into pentamer harnessed with the pentameric scaffold of CTB. Next, the proteins were subjected to SEC-FPLC. The estimated sizes of the constructs determined by SEC closely matched the pentamer sizes observed on SDS-PAGE (Figure S2). GM1 ganglioside serves as the receptor for CTB, and the efficient binding of CTB to GM1 indicates that CTB retains its native pentameric conformation [25,37,38]. Thus, to ascertain that our CTB-conjugated vaccine antigen forms a pentameric structure, we conducted a GM1 binding assay. This assay confirmed that all our constructs effectively bound to GM1 (Figure 4(C)). Lastly, the particle sizes were determined by DLS, revealing that CTB alone measured 7.6 nm, while the Ag-GIII and Ag-GV measured 15.3 and 19.2 nm, respectively. These findings indicate that vaccine antigens successfully form nanoparticles (Figure 4(D)).

Figure 4. Expression and evaluation of vaccine antigens. (A) Diagram of the expression vector construction. Cha: Chaperna domain; TEV: TEV protease cleavage site; CTB: cholera toxin B subunit; Ag: antigen. (B) Vaccine antigen expression and self-assembly analysed by SDS-PAGE. Non-reducing conditions display pentameric forms, while reducing conditions show monomeric forms. Triangles indicate expected sizes. M: size marker. (C) Pentameric assembly confirmed by GM1 binding assay; serum samples were diluted at 1:240. (D) Vaccine antigen particle sizes determined by Dynamic Light Scattering (DLS). **p < 0.01, ***p < 0.001, ****p < 0.0001.

Immunogenicity and protection efficacy of Ag-GIII and Ag-GV

We immunized mice three times with either Ag-GIII or Ag-GV to evaluate their immunogenicity and protective efficacy against both genotype-matched and genotype-unmatched JEV strains. Immune sera were prepared after the final immunization and serially diluted to evaluate the specific antibody titre against GIII Nakayama, GV 43279, and GV 43413 (Figure S3(A), Figure 5(A)). The results demonstrated that specific antibody titre was higher against genotype-matched virus. When neutralizing antibodies were assessed using the same serum samples, the genotype-matched vaccine antigens also exhibited effective plaque reduction activity (Figure 5(B)). Notably, both Ag-GIII and Ag-GV failed to show any neutralizing antibody titres against GV 43279.

Figure 5. Efficacy of monovalent Ag-GIII and Ag-GV. Mice were immunized three times with 20 μg of monovalent vaccine antigens at two-week intervals. Sera and spleens were collected two weeks after the final immunization (n = 8). (A) Specific antibody titres against GIII Nakayama, GV 43279, and GV 43413 were determined by ELISA with a serum dilution of 1:240. (B) Neutralizing antibody titres against GIII Nakayama, GV 43279, and GV 43413 were determined by PRNT. (C) IFN-γ levels in culture supernatants from antigen-restimulated splenocytes isolated from immunized mice, measured after 72 h. (A–C) *p < 0.05, **p < 0.01, ****p < 0.0001. (D) Immunized mice were challenged with GIII Nakayama or GV 43279 (n = 15). * indicates statistical significance versus Ag-neg. # indicates a significant difference between Ag-GIII and Ag-GV: ***p < 0.001, #p < 0.05, ##p < 0.01. Data are pooled from three independent experiments.

The induction of cellular immunity is imperative to control virus infection. We isolated splenocytes from the immunized mice and restimulated them with Ag-GIII or Ag-GV (Figure 5(C)). When IFN-γ levels were measured in the culture supernatant, antigen-matched stimulation resulted in higher concentrations. As also observed in the antibody response, cross-immune reactions were detectable in genotype-unmatched cases, but their intensity was significantly lower.

Next, we challenged the Ag-GIII or Ag-GV immunized mice with either GIII Nakayama or GV 43279 (Figure 5(D)). As expected, the vaccination with a genotype-matched antigen protected all mice from the lethal challenge. However, while the mice in the genotype-unmatched group exhibited a higher survival rate compared to those in the Ag-neg immunized group, they showed significantly lower protection than the mice in the genotype-matched group.

Immunogenicity and protection efficacy of bivalent vaccine antigen

Our experiments suggest that a bivalent vaccine containing GIII and GV antigens could be effective against all genotypes. To confirm this, sera obtained from mice immunized with the bivalent vaccine antigen, Ag-GIII+GV, were tested for specific antibody responses against GIII Nakayama, GV 43279, and GV 43413 (Figure S3(B), Figure 6(A)). The same serum samples exhibited plaque reduction activity against GIII Nakayama and GV 43413, but not against GV 43279 (Figure 6(B)). Splenocytes were isolated from mice immunized with Ag-GIII+GV and subsequently restimulated with either Ag-GIII or Ag-GV. The culture supernatant of these restimulated cells exhibited equivalent concentrations of IFN-γ, indicating that the bivalent vaccine induced a consistent cell-mediated immune response against both genotypes (Figure 6(C)).

Figure 6. Efficacy of bivalent vaccine antigen. Mice were immunized three times with 20 μg of bivalent vaccine antigens at two-week intervals. Sera and spleens were collected two weeks after the final immunization. (A) Specific antibody titres against GIII Nakayama, GV 43279, and GV 43413 were determined by ELISA with a serum dilution of 1:240 (n = 10). (B) Neutralizing antibody titres against GIII Nakayama, GV 43279, and GV 43413 were determined by PRNT (n = 16–30). (C) IFN-γ levels in culture supernatants from antigen-restimulated splenocytes isolated from immunized mice, measured after 72 h (n = 4). (D) Immunized mice were challenged with either of GIII Nakayama, GV 43279, GV 43413 (n = 15). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Data are pooled from three independent experiments.

Finally, we evaluated the protective efficacy against GIII and GV JEV strains. All mice in the Ag-GIII+GV immunized group survived, while 60% to 80 of the mice succumbed to the challenge in the Ag-neg immunized group (Figure 6(D)). We also evaluated the viral titre and neuropathic cytokines in the brains of Ag-neg and Ag-GIII+GV immunized mice (Figure S4(A, B)). Although not all analyses showed statistical significance, a reduction in both viral and immunopathologic burden was observed in the group immunized with Ag-GIII+GV. Furthermore, a decrease in the infiltration of key inflammatory cells, such as neutrophils and monocytes, into the brain was noted in Ag-GIII+GV immunized mice (Figure S4(C)).

Discussion

In this study, we utilized a nanoparticle platform with a CTB scaffold to test a vaccine candidate against GV JEVs. We produced protein antigens using Chaperna technology, which is originally aimed at increasing the solubility of monomeric antigen and their subsequent assembly into highly regulated multimers expressed in E. coli [27,28]. The Cha domain exerted a chaperone function ensuring the pentameric assembly of the CTB-conjugated chimeric antigen (Figure S2(B), Figure 4(B)). This experiment demonstrated that the use of Chaperna technology overcomes the challenges of soluble expression, folding and assembly in E. coli, contributing to bacterial production of low-cost vaccines.

Immunization with Ag-GV led to the production of specific binding antibodies against both GV strains and successfully defended against lethal dose infections in a mouse model. However, while neutralizing antibodies were induced against GV 43413, they were not observed for GV 43279. Particularly, considering that Ag-GV was based on the EDIII of GV 43279, the lack of a neutralizing antibody response against GV 43279 is difficult to explain. GV 43279 and GV 43413 differ by only one amino acid in the antigenic EDIII region used and has no further amino acid differences in non-EDIII region of whole E protein. The substitution at amino acid position 389 from aspartic acid (D) in GV 43279, carrying a negative charge, to glycine (G) in GV 43413, the smallest and non-polar amino acid, is noteworthy. This alteration between the strains, despite being a single amino acid difference, could significantly affect protein binding affinity due to the distinct properties of these amino acids. Such a substitution within the EDIII region might not only introduce structural variations but also potentially lead to distortions that influence the overall envelope protein structure. Moreover, the amino acids spanning positions 387–389 form the RGD motif, strategically located at the 5- and 3-fold junctions of the envelope proteins, prominently exposed on the virion surface [39]. The structural positioning of this motif suggests its critical role in determining the virion's morphology, which, in turn, could impact the efficacy of neutralizing antibody binding. In addition, as observed in experiments with the commercialized JE-VAX, GIII antigens are also ineffective in inducing an antibody response against GV strains in a whole inactivated vaccine form. Given the current lack of extensive research on the structural properties of the GV JEV E protein, this area warrants further study.

The development of vaccines targeting flaviviruses frequently confronts challenges associated with antibody-dependent enhancement (ADE) [40]. This phenomenon, where preexisting antibodies exacerbate the severity of secondary flavivirus infections, has been extensively studied, especially in dengue virus cases [41,42]. Further, evidence from other research suggests JE vaccinations might worsen dengue virus infections in animal models [43]. Thus, in developing vaccines for different genotypes of JEV, such as GIII and GV, there was a concern for ADE due to cross-reactive antibodies. Contrary to these apprehensions, even genotypically mismatched immunizations did not worsen infections but instead offered partial protection. In addition, our recombinant antigens showed enhanced cellular immunity compared to commercial vaccines, likely a result of CTB conjugation [25,44]. Research indicates that antigen-specific T-cell responses may mitigate ADE and improve protective efficacy of JEV vaccine [45,46]. Although our findings do not completely resolve questions about ADE, they highlight the need for more research into the advantages of CTB-conjugated recombinant protein vaccines concerning ADE.

Overall, our study not only confirmed the protective efficacy in experimental animal model of pentameric sub-virion type recombinant vaccine but also emphasized the importance of introducing genotype-matched vaccines to combat the emerging GV JEV outbreak. Furthermore, we demonstrated that a bivalent vaccine formulation could effectively provide protection against both GIII and GV JEVs. In our experiments, the neutralizing antibody response was less than optimal. Future studies should aim to enhance this response, potentially by using the entire E protein for antigen development and introducing novel adjuvants. However, it is also crucial to recognize that our antigen showed sufficient efficacy despite the suboptimal neutralizing antibody response. This highlights the necessity of evaluating a broad spectrum of immunological factors for a comprehensive assessment of vaccine efficacy. With a view to provide an alternative option to the currently licensed cell-culture based vaccines, efforts are underway to navigate the vaccine through the necessary processes for clinical trial readiness, signifying a crucial step towards translating into a low-cost recombinant JE vaccines.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This research was supported by a grant of the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (grant numbers: HV22C0250 and HV22C0259).