Rational Zika vaccine design via the modulation of antigen membrane anchors in chimpanzee adenoviral vectors

Abstract

Zika virus (ZIKV) emerged on a global scale and no licensed vaccine ensures long-lasting anti-ZIKV immunity. Here we report the design and comparative evaluation of four replication-deficient chimpanzee adenoviral (ChAdOx1) ZIKV vaccine candidates comprising the addition or deletion of precursor membrane (prM) and envelope, with or without its transmembrane domain (TM). A single, non-adjuvanted vaccination of ChAdOx1 ZIKV vaccines elicits suitable levels of protective responses in mice challenged with ZIKV. ChAdOx1 prME ∆TM encoding prM and envelope without TM provides 100% protection, as well as long-lasting anti-envelope immune responses and no evidence of in vitro antibody-dependent enhancement to dengue virus. Deletion of prM and addition of TM reduces protective efficacy and yields lower anti-envelope responses. Our finding that immunity against ZIKV can be enhanced by modulating antigen membrane anchoring highlights important parameters in the design of viral vectored ZIKV vaccines to support further clinical assessments.

Fig. 1

Zika vaccine design. a A phylogenetic tree for ZIKV genomes up to November 2015; blue and red labels represent the African and Asian lineages of ZIKV, respectively. b Conservation homology of an Asian consensus sequence versus all genomic sequences depicted in Supplementary Fig. 1. The circle represents the ZIKV-BR strain used for the challenge experiment. c Schematic representation of the Zika immunogen versions used in this study; red box represents the tPA leading sequence. d Restriction enzyme analysis of the plasmid DNA vaccine constructs that were further cloned into ChAdOx1 vector; a 3.3 kb band size represents the DNA vaccine plasmid back bone. e Expression of ZIKV immunogens by western immunoblotting using an anti-ZIKV envelope antibody. M molecular marker, C control, E empty plasmid. Uncropped images are shown in Supplementary Fig. 2

Fig. 2

Immune responses elicited by DNA versus ChAdOx1 vaccines. a For ZIKV DNA vaccines, BALB/c mice (n = 6/group) were immunised intramuscularly (i.m.) with a dose of 100 μg/mouse, followed by a DNA Boost 2 weeks thereafter. For ChAdOx1 ZIKV vaccines, a single dose of 10⁸ IU/mice was i.m. administered. Blood samples were obtained at several time points for either ELISA or ELISPOT assay. b Humoral responses elicited by DNA Prime and Prime-Boost after 2 weeks (left) and by a single immunisation of ChAdOx1 ZIKV vaccines at 10 days, and 2, 4, 16 and 36 weeks (right). Antibody responses were quantified by ELISA with plates coated with ZIKV envelope protein. Positive index was calculated as depicted by ELISA kit manufacturers. c PBMCs–IFNγ-producing cells from DNA Prime-Boost after 2 weeks (left) and ChAdOx1 ZIKV vaccines at 2 weeks and 3 months after single immunisation were quantified by ELISPOT. 20-mer peptides spanning the ZIKV envelope protein (10 μg/ml) were used for stimulation. Line colours and shapes represent mice vaccinated with each vaccine. Antibody production by ELISA and cellular responses by ELISPOT were replicated three times in the laboratory. A positive control sample was included to validate the ELISA assay

Fig. 3

Assessment of protective efficacy induced by ChAdOx1 ZIKV vaccines. a BALB/c mice (n = 5) immunised with a single i.m. injection of ChAdOx1 ZIKV vaccines and naive mice were intravenously challenged with 10⁵ VP of ZIKA-BR at week 4 after prime. Pre-challenge serum samples were collected for ELISA as shown. b Viral load in ZIKV-challenged groups was monitored for 7 days in sera to follow the onset of viraemia. Each blue line represents one mouse. c Assessment of various parameters of viraemia in groups vaccinated with ChAdOx1 ZIKV vaccines (100% = 6 out of 6 mice). d Pre-challenge reciprocal ELISA titres (Supplementary Fig. 4) were plotted against global ZIKV VL after ZIKV challenge and a comparison was made between ChAdOx1 Env ∆TM versus ChAdOx1 prME ∆TM (left) and ChAdOx1 prME versus ChAdOx1 Env (right). e ELISA OD 450 reads of serially diluted serum samples were plotted to assess the impact of ∆TM in antibody production. Antibody production by ELISA in pre-challenge sera was replicated two times. f Immunofluorescence analysis of Vero cells expressing the ZIKV immunogens (green) to assess subcellular staining and distribution. Antigen was detected using a commercial anti-flavivirus antibody. Blue (DAPI) represents the nucleus and scale bar is 20 μm. Cell transfections were performed in technical duplicates in two biological replicates. g Kinetics of ZIKV envelope antigen expression in HEK-293 cells by western blot. Antigen was detected using an anti-ZIKV monoclonal antibody in both, non-concentrated supernatant and cell extracts. Red arrows, ZIKV envelope. Actin was used as a loading control. Expression of antigen was verified in three biological replicates. Uncropped blots are shown in Supplementary Fig. 5

Fig. 4

ZIKV neutralisation assay and antibody-dependent enhancement (ADE) to DENV2 infection. a Microneutralisation titre 50 (MN50) assay in Vero cells was assessed upon incubation of ZIKV-BR in the presence of serially diluted sera from BALB/c mice vaccinated with ChAdOx1 ZIKV vaccines at 4 weeks and 4 months. Results show the average of three biological replicates, with duplicates and black lines represents the mean (n = 6). A two-way ANOVA followed by Sidak’s Multiple comparison was performed. ****p < 0.0001, ***p < 0.0003. Supplementary Table 1 shows the multiple comparison between all the groups. b ELISA cross-reactivity of DENV1-4 in sera from mice vaccinated with the highly protective ChAdOx1 prME ∆TM vaccine. Bars represent the average and error is the s.d. c Infection of U937 cells by a DENV2 strain in the presence of sera from mice vaccinated with ChAdOx1 ZIKV candidates. Infection enhancement fold was calculated as the ratio of focus-forming units (FFU) of an anti-DENV2 monoclonal antibody to that of the FFU in the presence of serial dilutions of mouse sera collected 4 weeks after ChAdOx1 ZIKV vaccination. Each coloured line and shape represents a mouse for each group. n = 6/vaccinated group. Data represent three technical replicates, with a positive control anti-DENV2 antibody