Modified mRNA Vaccines Protect against Zika Virus Infection

Abstract

The emergence of ZIKV infection has prompted a global effort to develop safe and effective vaccines. We engineered a lipid nanoparticle (LNP) encapsulated modified mRNA vaccine encoding wild-type or variant ZIKV structural genes and tested immunogenicity and protection in mice. Two doses of modified mRNA LNPs encoding prM-E genes that produced virus-like particles resulted in high neutralizing antibody titers (∼1/100,000) that protected against ZIKV infection and conferred sterilizing immunity. To offset a theoretical concern of ZIKV vaccines inducing antibodies that cross-react with the related dengue virus (DENV), we designed modified prM-E RNA encoding mutations destroying the conserved fusion-loop epitope in the E protein. This variant protected against ZIKV and diminished production of antibodies enhancing DENV infection in cells or mice. A modified mRNA vaccine can prevent ZIKV disease and be adapted to reduce the risk of sensitizing individuals to subsequent exposure to DENV, should this become a clinically relevant concern.

Keywords: Dengue virus; RNA vaccine; Zika virus; antibody neutralization; flavivirus; immunity; pathogenesis; protection.

Figure 1. ZIKV mRNA LNP vaccine testing in AG129 mice

A. ZIKV prM-E modified mRNA is packaged into LNPs for intramuscular delivery. B. Schematic of ZIKV genome (top). An mRNA encoding the ZIKV prM/M and E structural genes was engineered (bottom). In this construct, prM is directed into the ER using a heterologous IgE signal sequence (IgE_sig) at the amino-terminus. C. HeLa cells were transfected with prM-E mRNA, and SVPs in the supernatant were purified and concentrated by ultracentrifugation and then subjected to electron microscopy and negative staining. Low- and high-power images of purified SVPs are shown. Scale bar = 30 nm. One representative experiment of several is shown D. HeLa cell supernatants were collected for Western blotting under non-reducing conditions with a mAb against the ZIKV E protein. E. Scheme of immunization of AG129 mice with one (prime) or two (prime-boost) doses of 2 or 10 μg with IgE_sig prM-E or placebo mRNA LNP vaccines. F–H. Serum was collected at 6 weeks after vaccination and analyzed for neutralization of ZIKV by PRNT assay. Representative curves are shown (F), and EC50 (G) and EC90 (H) values were calculated for individual animals in each group. Each point represents the mean of two independent experiments per animal. Data are a composite of two independent experiments with sera from each of the 10 animals per group. EC50 and EC90 data were analyzed by a Kruskal-Wallis test with a multiple comparisons correction and compared to the placebo LNP vaccine (*, P < 0.05; ***, P < 0.001; ****, P < 0.0001). The dashed lines indicate the limit of detection of the assay. I–J. AG129 Mice were challenged at 6 weeks post vaccination with 10⁴ PFU of ZIKV P6-740. Animals were monitored for survival (I) and weight loss (J). Error bars indicate standard error the mean (SEM). Survival data was analyzed by the log rank test (*, P < 0.05; ***, P < 0.001). Weight loss was analyzed by two-way ANOVA (***, P < 0.001). See also related figures S1 and S2.

Figure 2. ZIKV mRNA LNP vaccine protects C57BL/6 mice

WT C57BL/6 mice (n = 10, pooled from two independent experiments) were immunized with 10 μg of placebo or IgE_sig prM-E mRNA LNPs and boosted with an equivalent dose four weeks later. A. Serum was collected at 4, 8, and 18 weeks post initial vaccination and analyzed for ZIKV neutralization activity by FRNT assay. Representative neutralization curves are shown for each group. Error bars denote the standard deviation (SD) of triplicate technical replicates. B–C. EC50 (B) and EC90 (C) values were calculated for individual animals in each group. The dashed lines indicate the limit of detection of the assay. Data were analyzed using the Mann-Whitney test and compared to the placebo LNP vaccine at each time point (**, P < 0.01; ***, P < 0.001; n.s. indicates not significant). D–F. At week 8 or 18, vaccinated C57BL/6 mice were administered 2 mg of anti-Ifnar1 blocking antibody and one day later challenged with 10⁵ FFU of mouse adapted ZIKV Dakar 41519. Animals were monitored for survival (D) and weight loss (E). At day 5 after viral challenge, serum was analyzed for levels of ZIKV RNA (F). The dashed line indicates the limit of detection of the assay. Survival data was analyzed by the log rank test (***, P < 0.001). Weight loss was analyzed by two-way ANOVA (***, P < 0.001) for surviving animals; arrows indicate days having statistically significant differences from placebo vaccine. Viremia data was analyzed by a Mann-Whitney test (*, P < 0.05; ***, P < 0.001). See also related figure S3.

Figure 3. ZIKV mRNA LNP vaccines containing WT or mutant FL sequences induce neutralizing antibody responses and protect BALB/c mice A

Immunization scheme. Female WT BALB/c mice (n = 10, pooled from two independent experiments) were immunized with 2 or 10 μg of prM-E mRNA LNP vaccines containing either IgE or JEV signal sequences at the N-terminus of prM and WT or mutant FL sequences in the E gene. Animals were boosted with the equivalent dose of the same vaccine 28 days later. B. HeLa cells were transfected with different modified mRNA (Lane a, IgE_sig-prM-E; Lane b, IgE_sig-prM-E FL mutant; Lane c, JEV_sig-prM-E; and Lane d, JEV_sig-prM-E FL mutant), and ZIKV E protein in the supernatant was detected by Western blotting with a type-specific anti-ZIKV E antibody (left panel) or a cross-reactive anti-ZIKV E antibody that binds the FL (right panel). Results are the representative of two independent experiments. E, monomer; E₂, dimer. C–F. At week 8, serum was harvested from IgE_sig-prM E (C) or JEV_sig prM-E (D) mRNA LNP vaccinated mice and analyzed for neutralization capacity using ZIKV RVPs. Representative curves for each group (with EC50 values at or near the group median) are shown in (C) and (D). Error bars indicate the range of duplicate technical replicates. EC50 (E) and EC90 (F) values were calculated for individual animals in each group. Each point represents the results from a single experiment or the mean of two independent experiments. The dotted line indicates the limit of detection of the assay. G–J. At week 8, vaccinated BALB/c mice were administered 2 mg of anti-Ifnar1 blocking antibody and one day later challenged with 10⁵ FFU of mouse adapted ZIKV Dakar 41519. At 3 (G) and 7 (H–J) days after viral challenge, serum (G), uterus (H), spleen (I), and brain (J) tissues were harvested and analyzed for ZIKV RNA. EC50, EC90, and viral titer data were analyzed by a Kruskal-Wallis test with a multiple comparisons correction and compared to the placebo LNP vaccine (*, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; n.s., not significant). K–L. Correlates of day 3 viral load (K, left panel), day 7 spleen titers (L, left panel) and protective efficacy (K–L, right panels) are shown. Data from all JEV_sig-prM-E and IgE_sig-prM-E mRNA LNP vaccines was included in this analysis. P values and R² values reflect Spearman rank-correlation tests. For correlate data, values at which the line would cross the limit of detection of the y-axis are indicated below the graphs. Red line represents the best fit linear regression. For protective efficacy data, bars indicate median values (****, P < 0.0001; Mann-Whitney test). M–N. Anamnestic neutralizing antibody response. Paired sera were collected from vaccinated animals (JEV_sig-prM-E (M) or IgE_sig-prM-E and JEV_sig-prM-E FL (N)) immediately before (Pre) or 7 days after (Post) ZIKV challenge and analyzed for neutralizing activity using ZIKV RVPs. EC50 and EC90 values were analyzed for differences by a paired t-test (n.s., not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001). Indicated at the bottom of the graphs are the number of animals showing a 4-fold increase in neutralization titer (positive anamnestic response) at 7 days after ZIKV challenge. See also related figure S4.

Figure 4. ZIKV mRNA LNP vaccines containing mutant FL sequences showed reduced ADE against DENV in cell culture and in AG129 mice

A–B. Serial dilutions of serum obtained at week 8 (see Fig 3) from BALB/c mice immunized with 2 or 10 μg of IgE_sig-prM E (A) or JEV_sig-prM-E (B) mRNA LNPs with WT or mutant FL sequences were mixed with DENV-1 RVPs and incubated with Fc-γ receptor expressing K562 cells. Infected cells were quantified by flow cytometry. Representative curves for each group (with peak enhancement titers (PET) at or near the group median) are shown in (A) and (B). Results are displayed relative to the maximum infectivity observed with the control cross-reactive WNV mAb E60 (FL-specific) run in parallel (C). The dotted line is provided as a reference for mAb E60 normalization. Error bars indicate the range of duplicate technical replicates. D. Peak enhancement titer (PET) for each mouse per group. Data were analyzed by a Kruskal-Wallis one-way ANOVA with a multiple comparisons correction and compared to the control LNP vaccine (**, P < 0.01; ***, P < 0.001). E. Magnitude of enhancement (extent of infection at the PET) for each mouse per group. F. Neutralization of DENV-2 RVPs by sera pooled from placebo or IgE_sig-prM-E (2 μg dose of WT or FL mutant) vaccinated mice. Error bars indicate the range of duplicate technical replicates. G–H. Enhancing effects of ZIKV immune serum on DENV-2 infection in AG129 mice. Recipient AG129 mice were passively transferred PBS, 1 μl (G) or 10 μl (H) of pooled serum from BALB/c mice vaccinated with WT or FL mutant IgE_sig-prM E LNPs, or 15 μg of anti-prM mAb (2H2, positive control). One day later, animals were challenged with 10⁵ FFU of DENV-2 (strain S221) and followed for mortality (left panels) and clinical score (right panels) (1 (healthy) to 7 (deceased) scale; see STAR Methods). Results are pooled from two to three independent challenge experiments (numbers of animals indicated beneath graph) with the exception of the anti-prM mAb, which was administered in only one of the two experiments as a positive control. Survival curves between serum transfers from IgE_sig-prM-E (WT and FL mutant LNPs) vaccinated mice were statistically different (****, P < 0.0001, log-rank test). See also related figures S5 and S6.