Virus-like particles derived from Pichia pastoris-expressed dengue virus type 1 glycoprotein elicit homotypic virus-neutralizing envelope domain III-directed antibodies

Abstract

Background: Four antigenically distinct serotypes (1-4) of dengue viruses (DENVs) cause dengue disease. Antibodies to any one DENV serotype have the potential to predispose an individual to more severe disease upon infection with a different DENV serotype. A dengue vaccine must elicit homotypic neutralizing antibodies to all four DENV serotypes to avoid the risk of such antibody-dependent enhancement in the vaccine recipient. This is a formidable challenge as evident from the lack of protective efficacy against DENV-2 by a tetravalent live attenuated dengue vaccine that has completed phase III trials recently. These trial data underscore the need to explore non-replicating subunit vaccine alternatives. Recently, using the methylotrophic yeast Pichia pastoris, we showed that DENV-2 and DENV-3 envelope (E) glycoproteins, expressed in absence of prM, implicated in causing severe dengue disease, self-assemble into virus-like particles (VLPs), which elicit predominantly virus-neutralizing antibodies and confer significant protection against lethal DENV challenge in an animal model. The current study extends this work to a third DENV serotype.

Results: We cloned and expressed DENV-1 E antigen in P. pastoris, and purified it to near homogeneity. Recombinant DENV-1 E underwent post-translational processing, namely, signal peptide cleavage and glycosylation. Purified DENV-1 E self-assembled into stable VLPs, based on electron microscopy and dynamic light scattering analysis. Epitope mapping with monoclonal antibodies revealed that the VLPs retained the overall antigenic integrity of the virion particles despite the absence of prM. Subtle changes accompanied the efficient display of E domain III (EDIII), which contains type-specific neutralizing epitopes. These VLPs were immunogenic, eliciting predominantly homotypic EDIII-directed DENV-1-specific neutralizing antibodies.

Conclusions: This work demonstrates the inherent potential of P. pastoris-expressed DENV-1 E glycoprotein to self-assemble into VLPs eliciting predominantly homotypic neutralizing antibodies. This work justifies an investigation of the last remaining serotype, namely, DENV-4, to assess if it also shares the desirable vaccine potential manifested by the remaining three DENV serotypes. Such efforts could make it possible to envisage the development of a tetravalent dengue vaccine based on VLPs of P. pastoris-expressed E glycoproteins of the four DENV serotypes.

Keywords: Antibody dependent enhancement; Dengue virus; Envelope domain III; Envelope glycoprotein; Neutralizing antibody; Pichia pastoris; Virus-like particles.

Fig. 1

Investigation of VLP formation by purified DENV-1 E protein. a EM analysis of DENV-1 E VLPs by negative staining with 1 % uranyl acetate. b DLS analysis of particle size distribution by volume

Fig. 2

VLPs stability study. DLS analysis of DENV-1 E VLP size distribution by volume after incubating them for 2 weeks at (a) 4 °C or (b) 37 °C

Fig. 3

MALDI-TOF-MS analysis of N-linked Oligosaccharides of P. pastoris-expressed DENV-1 E protein. a Tabular representation of molecular weights of sugars attached to DENV-1 E protein with their respective structures, number of mannose residues and their percent area (% area). The % area corresponds to the relative proportion of the oligosaccharide out of total oligosaccharide populations. b Graphical representation of peaks corresponding to different molecular weight sugars with their respective intensities on y-axis and corresponding mass/charge (m/z) ratios on x-axis

Fig. 4

Characterization of antibodies generated by DENV-1 E VLPs. Pooled sera from mice immunized with DENV-1 E VLPs, collected on day 37 (dashed line) and day 100 (solid line) were analysed in indirect ELISA format using DENV-1 E VLPs as the coating antigen (a). Pooled day 100 immune serum was evaluated for reactivity using DENV-E VLPs (b), recombinant EDIIIs (c), and intact DENVs (d) taken as coating antigens, corresponding to serotypes 1 (red curve), 2 (green curve), 3 (blue curve) and 4 (black curve). Grey lines (in all panels) represented serum from PBS-immunized BALB/c mice. e Indirect immunofluorescence analysis of DENV-1 infected BHK-21 cells using (i) PBS-immunized serum (ii) 4G2 mAb (iii) anti-DENV-1 E VLP antiserum taken as the source of primary antibodies. Anti-mouse IgG-FITC conjugate was used to visualize bound antibodies

Fig. 5

Determination of virus neutralization potency of antibodies induced by DENV-1 E VLPs. Immune sera were serially diluted two-fold and pre-incubated separately with DENV-1 (red curve), DENV-2 (green curve), DENV-3 (blue curve) and DENV-4 (black curve). The antibody-virus complexes were allowed to infect Vero cells and neutralizing antibody titers (FNT₅₀) were determined by using FACS analysis. The Y and X axes of the graph represent % infected cells and log of reciprocal of sera dilution, respectively. The dotted horizontal line in the graph represents 50 % infection. The table on top of the panel lists the FNT50 titers of DENV-1 E antiserum against all the four serotypes

Fig. 6

Analysis of the role of anti-EDIII antibodies in virus neutralization: DENV-1 E VLP antiserum was used without prior depletion (a) or treated as follows. It was incubated with immobilized MBP (b) or immobilized MBP-EDIII-1 (c) matrix. Residual antibody titers in these three DENV-1 E VLP antiserum treatment groups (a, b and c), were analyzed in indirect ELISA using purified recombinant MBP-EDIII-1, −2, −3, −4 protein (panel a) or DENV-1, −2, −3 E VLPs (panel b). Residual neutralizing antibody titers (FNT50) in the same three sera groups were estimated using the FACS assay (Panel c)