Adjuvant-Mediated Epitope Specificity and Enhanced Neutralizing Activity of Antibodies Targeting Dengue Virus Envelope Protein

Abstract

The heat-labile toxins (LT) produced by enterotoxigenic Escherichia coli display adjuvant effects to coadministered antigens, leading to enhanced production of serum antibodies. Despite extensive knowledge of the adjuvant properties of LT derivatives, including in vitro-generated non-toxic mutant forms, little is known about the capacity of these adjuvants to modulate the epitope specificity of antibodies directed against antigens. This study characterizes the role of LT and its non-toxic B subunit (LTB) in the modulation of antibody responses to a coadministered antigen, the dengue virus (DENV) envelope glycoprotein domain III (EDIII), which binds to surface receptors and mediates virus entry into host cells. In contrast to non-adjuvanted or alum-adjuvanted formulations, antibodies induced in mice immunized with LT or LTB showed enhanced virus-neutralization effects that were not ascribed to a subclass shift or antigen affinity. Nonetheless, immunosignature analyses revealed that purified LT-adjuvanted EDIII-specific antibodies display distinct epitope-binding patterns with regard to antibodies raised in mice immunized with EDIII or the alum-adjuvanted vaccine. Notably, the analyses led to the identification of a specific EDIII epitope located in the EF to FG loop, which is involved in the entry of DENV into eukaryotic cells. The present results demonstrate that LT and LTB modulate the epitope specificity of antibodies generated after immunization with coadministered antigens that, in the case of EDIII, was associated with the induction of neutralizing antibody responses. These results open perspectives for the more rational development of vaccines with enhanced protective effects against DENV infections.

Keywords: adjuvants; antibodies; dengue virus; envelope protein; heat-labile toxins; immunosignature; labile toxins; vaccines.

Figure 1

Analysis of envelope glycoprotein domain III (EDIII)-specific antibody responses induced in mice immunized with different vaccine formulations. (A) Schematic representation of the vaccine regimen tested in this study. BALB/c female mice were s.c. immunized with three doses of EDIII alone (10 µg) or EDIII coadministered with one of the tested adjuvants [LT1 (1 µg), LT1-B (3.2 µg), or alum (12.5 µg)]. (B) Analyses of the antigen-specific IgG responses performed in blood samples collected 2 weeks after each vaccine dose. (C,D) Evaluation of the anti-EDIII total IgG (C) or IgG subclass (D) responses was carried out 2 weeks after the third dose. (E,F) Reactivity of serum antibodies carried out by ELISA with the native purified EDIII antigen (E) or the same antigen submitted to a heat denaturation treatment (F) used as solid phase bound antigens. Anti-EDIII titers were determined in three independent experiments with at least five animals per immunization group. Values represent means ± SD of the IgG titers. ***p < 0.001, **p < 0.01, and *p < 0.05, comparing adjuvanted EDIII-immunized mice to non-adjuvanted EDIII-immunized mice or alum-adjuvanted mice (ANOVA with Bonferroni post hoc test).

Figure 2

Characterization of purified envelope glycoprotein domain III (EDIII)-specific IgG antibodies raised in mice immunized with different adjuvants. (A) Reactivity of antibodies with the purified antigen following ELISA. (B) Affinity of purified antibodies to the purified EDIII antigen. Values are expressed by percentage of antibodies that remain bound to the solid phase-adsorbed antigen under presence of ammonium thiocyanate in regard to the reaction performed without ammonium thiocyanate. Immunization groups represented by different symbols as indicated in the insert. (C,D) Plaque reduction neutralization test of a DENV2 strain (New Guinea C) carried out with serially diluted serum samples (C) or various amounts of purified EDIII-specific IgG antibodies (D) derived from mice immunized with the different vaccine formulations, as indicated in panel (B). Results are based on two independently performed experiments carried out with duplicate samples. Values represent means ± SD. ***p < 0.001, **p < 0.01, and *p < 0.05, significant differences with regard to mice immunized with non-adjuvanted EDIII (two-way ANOVA with Bonferroni post hoc test). NS: values without statistically significant differences.

Figure 3

Immunosignature analyses of purified envelope glycoprotein domain III (EDIII)-specific IgG antibodies. (A) Heat map demonstrating the immunosignatures detected with EDIII-specific IgG antibodies purified from serum samples collected from mice submitted to the different immunization regimens and a mock mouse group (control). A Student’s t-test (p < 3.33 × 10⁻⁶) between vaccine groups was used to select the informative peptides. Hierarchical clustering using EUCLIDEAN distance was used as a measure of similarity to cluster the selected peptides (x axis) and vaccine groups (y axis). Peptides’ intensity is colored where blue corresponding to low intensity and red to high intensity. (B) Variance of immune responses among different vaccine groups. Variation among vaccine groups [pooled sera per group (n = 5)] for all significant peptides sequences in the principal component analysis plot, where the first two principal components are plotted and individuals are colored according to the vaccination groups. (C) Venn diagram between vaccine groups. Results summarize the overlapping peptides that are significantly different above 1.3-fold change of the mock group versus each vaccine group. (D) Antigenic epitope prediction for vaccine groups using random peptide arrays. Significant peptides sequences recognized by antibodies of each vaccine group were submitted to epitope prediction with the GuiTope software and using the EDIII protein sequence from DENV2 New Guinea C (AHG97599.1) as reference. Each line graph represents the GuiTope prediction score for each vaccine group.

Figure 4

Identification of a linear envelope glycoprotein domain III (EDIII)-derived epitope involved in binding of DENV2 to host cells. (A) 3D representation of DENV E protein dimer. Domains I (yellow), II (red), and III (blue) are highlighted. The insert displays EDIII structural features. The amino acid sequence marked in light blue encompasses the epitope recognized by antibodies raised in mice immunized with labile toxins (LT)-adjuvanted vaccines. (B) Amino acid sequence of DENV2 EDIII (GenBank accession number: AF038403.1). The positions of the EDIII β strands are indicated above the sequence by dark blue arrows, while the sequences of the synthetic peptide used in the competition assays are shown below. The epitope predicted by the immunosignature is marked in light blue. (C) Inhibition of DENV2 infection. Different molar concentrations of synthetic peptide 47 were mixed with DENV2 New Guinea C and subsequently added to cells. Purified EDIII and heat-denatured EDIII were used as positive and negative controls in the inhibition of virus infectivity assay, as indicated in the figure. Cells stained with primary mAb 4G2 and subsequently with Alexa 488-conjugated anti-mouse IgG were measured by cytometry, and the number of DENV-positive infected Vero cells was determined for each group. Values express the reduction in the percentage of the number of virus-infected cells compared to the control group denatured EDIII- (untreated DENV2). Dashed line indicates 50% inhibition of virus infection. Bars represent the mean values and SEM from three independent experiments. ***p < 0.001, **p < 0.001, and *p < 0.05 represent significant differences with regard to denatured EDIII-treated and untreated DENV2 (two-way ANOVA with Bonferroni post hoc test).