Palabras clave: dengue, vacuna tetravalente, dominio III, cápsida, DIIIC, respuesta protectora.

INTRODUCTION

Currently, there is no effective vaccine available against Dengue virus (DENV) infection and it is associated mosquito-borne dengue disease which affects about 390 million people every year [1]. Particularly, some Dengue epidemics have occurred in Cuba since 1977, with the most severe in 1981 and 1997. But Dengue disease remains as a serious health problem in spite of the numerous and sustained efforts to eradicate both, the vector and the disease.

An effective vaccine against Dengue implies the generation of immunity against the four DENV serotypes. This is necessary due to the lack of long-lasting and cross-protective immune response among serotypes. Besides, the heterotypic immunity makes individuals susceptible to a most severe dengue subsequent infection by a heter-ologous serotype. In fact, the most advanced candidate is composed of chimeric attenuated strains showing preliminary efficacy results lower than 60 % [2-4].

Therefore, in this work, a tetravalent vaccine candidate was generated for the first time based on the fusion of the DENV envelope domain III (DIII) and capsid protein (DIIIC) of all the four DENV serotypes, formulated as an aggregate with oligodeoxynucleotides (ODN). DIII of the viral envelope protein mediates the virus attachment to its cellular receptor and bears neutralizing epitopes. On the other hand, the capsid protein was able to induce cell-mediated protective immunity in mice and monkeys [5, 6] and it was demonstrated that in the presence of ODNs it forms particulate aggregates with immunogenic potential. Previous evidences on the induction of a functional and protective immune response in mice and monkeys by administering the DIIIC protein from DENV2 formulated as an aggregate [7], supported the extension of this DIIIC design to the rest of DENV serotypes.

DIIIC chimeric variants for DENV serotypes 1, 3 and 4 were obtained by recombinant procedures, purified and subjected to antigenic characterization. Furthermore, their immunogenicity and protective efficacy was studied in mice, once formulated together with the previous DIIIC-2 chimeric protein and aggregated with ODN 39M, as a tetravalent vaccine candidate (Tetra-DIIIC). The immunogenicity of the tetra-DIIIC vaccine candidate formulation was evaluated in non-human primates using different administration routes. This research granted the 2015 Award of the Cuban National Academy of Sciences.

Recombinant chimeric molecules comprising DIIIC proteins against DENV serotypes 1, 3 and 4 were generated by cloning their respective fusion genes into the pET28a plasmid expression vector. These genetic constructions called ACDC-1, 3, 4 plasmids contain a 6-His-tag at the N-terminal sites of the fusion proteins. The recombinant proteins were expressed under the control of the T7 promoter in the BL21(DE3) strain of Escherichia coli using IPTG as inductor. SDS-PAGE revealed the presence of a band of approximately 28 kDa that was also immunoindentified with anti-DENV hyperimmune murine ascitic fluids (HMAF) [8].

The three chimeric proteins were purified and antigenically characterized by measuring their reactivity against an anti-dengue murine and human panel of sera. As shown in figure 1, DIIIC-1 and DIIIC-3 were highly recognized by both panel of sera assayed, indicating a proper folding of domain III region in the context of the homologous capsid protein. In the specific case of DIIIC-4, the reactivity was lower than that measured for the rest of the proteins. Similar results have been obtained with several tetravalent recombinant candidates against DENV where the sero-type 4 was less immunogenic [9-11].

Subsequently, the characterization of DIIIC proteins by transmission electron microscopy revealed the presence of aggregates once they were incubated with ODN39M, an oligonucleotide of proven immunostimulatory capacity [7]. The microscopy showed particle of 50-55 nm depending of ODN addition. The particle nature of these aggregates could favor their internalization by dendritic cells and their transport to lymph nodes where the induction of an effective immune response takes place.

Subsequently, the chimeric DIIIC proteins were evaluated, either separate or all combined with DIIIC2 in a tetravalent formulation, in BALB/c mice. For this purpose, the proteins were incubated with ODN 39M at the protein: ODN ratio promoting the 50 % of protein precipitation and the presence of soluble and insoluble species of protein-ODN. Afterward the DIIIC preparation was adjuvanted with alum and inoculated in mice by the intraperitoneal route on days 0, 15 and 45. Fifteen days after the third dose, the humoral immune response against each recombinant protein was determined. All animals seroconverted and the administration of the tetravalent formulation generated high anti-DIIIC IgG titers against the four recombinant proteins. The results shows a lack of antigenic competition in the tetravalent formulation with similar titers of humoral responses against each chimeric protein to that obtained with the respective monovalent formulations (Figure 2).

Furthermore, the functionality of the humoral immune response was determined by means of its neutralizing activity (Table 1), showing a similar response related to the percentage of responders for the tetravalent and monovalent formulations against DENV serotypes 1, 2 and 3. No neutralizing antibodies were detected against the DENV4 protein for any of the formulations assayed.

One month after the administration of the last dose, the cell-mediated immune response was determined by measuring the mice splenocytes secretion of interferon gamma (IFNγ) in response to the stimulation with the recombinant DIIIC proteins [12]. When DIIIC-1, 2 and 3 proteins were used for stimulation, all the animals immunized with the monovalent and tetravalent formulations showed a positive response. In the case of serotype 4 the secretion of IFNγ was lower, with only 5 out of 7 and 6 out of 7 animals showing a positive response in the monovalent and tetravalent groups, respectively. The low immunogenicity of DENV-4 has been extensively described in mouse experiments with other dengue vaccine candidates such as the recombinant DENV-4 DIII fused to maltose binding protein [8], and a vaccine candidate based on dengue E protein, expressed in Drosophila S2 cells [13]. Taken together, we postulate that DIII of DENV-4 is not an immunodominant region in BALB/c mice.

The tetravalent formulation was able to induce a protective response against the four DENV serotypes in the DENV murine encephalitis model, in spite of the lower response detected against DENV4. Animals receiving the tetravalent formulation reduced the viral loads in the brain, with statistical differences with respect to the placebo group (P < 0.05), and statistically similar to the positive control groups (P > 0.05), indicating a solid protection [12].

Subsequently, the tetravalent DIIIC formulation adjuvanted in alum was evaluated in non-human primates, using different immunization routes (subcutaneous, intradermal or intramuscular route). The schedule comprised three immunizations at two-month intervals, in groups of three animals. After immunization, the animals developed antiviral antibodies with neutralizing activity against the four DENV serotypes (Figure 3 and Table 2). The formulation administered by the intradermal route trended to generate the lowest antibody response in terms of percentage of responders, maybe due to the 10-fold lower vaccine dose used by this route in comparison with the other immunizations.

Additionally, the frequency of IFNγ-secreting cells and the cytotoxic capacity of peripheral blood mononuclear cells (PBMCs) were measured by ELISPOT and LDH release in immunized monkeys, following the in vitro stimulation with each recombinant chimeric protein [12].

RELEVANCE OF THE STUDY