A bivalent self-amplifying RNA vaccine against yellow fever and Zika viruses

Abstract

Introduction: Yellow fever (YFV) and Zika (ZIKV) viruses cause significant morbidity and mortality, despite the existence of an approved YFV vaccine and the development of multiple ZIKV vaccine candidates to date. New technologies may improve access to vaccines against these pathogens. We previously described a nanostructured lipid carrier (NLC)-delivered self-amplifying RNA (saRNA) vaccine platform with excellent thermostability and immunogenicity, appropriate for prevention of tropical infectious diseases.

Methods: YFV and ZIKV prM-E antigen-expressing saRNA constructs were created using a TC-83 strain Venezuelan equine encephalitis virus-based replicon and complexed with NLC by simple mixing. Monovalent and bivalent vaccine formulations were injected intramuscularly into C57BL/6 mice and Syrian golden hamsters, and the magnitude, durability, and protective efficacy of the resulting immune responses were then characterized.

Results and discussion: Monovalent vaccines established durable neutralizing antibody responses to their respective flaviviral targets, with little evidence of cross-neutralization. Both vaccines additionally elicited robust antigen-reactive CD4⁺ and CD8⁺ T cell populations. Notably, humoral responses to YFV saRNA-NLC vaccination were comparable to those in YF-17D-vaccinated animals. Bivalent formulations established humoral and cellular responses against both viral targets, commensurate to those established by monovalent vaccines, without evidence of saRNA interference or immune competition. Finally, both monovalent and bivalent vaccines completely protected mice and hamsters against lethal ZIKV and YFV challenge. We present a bivalent saRNA-NLC vaccine against YFV and ZIKV capable of inducing robust and efficacious neutralizing antibody and cellular immune responses against both viruses. These data support the development of other multivalent saRNA-based vaccines against infectious diseases.

Keywords: Zika virus; bivalent vaccine; flavivirus vaccine; nanostructured lipid carrier; self-amplifying RNA; yellow fever virus.

Figure 1

Characterization of the components in AAHI’s bivalent flavivirus saRNA-NLC vaccine. (a) Construct designs for the YFV and ZIKV saRNA vaccine constructs. YFV replicon size is 9.89 kb, and ZIKV replicon size is 9.94 kb. SGP = sub-genomic promoter. (b) Schematic of the NLC RNA delivery particle. Design by Cassandra Baden. (c) Structural alignment of the YF-17D (PDB ID: 6IW4; red) and ZIKV (PDB ID: 5JHM; blue) E protein pre-fusion dimers, indicating significant structural similarity between the two antigens. (d) Western blot verifying in vitro protein expression of 54 kDa YFV and ZIKV E protein in cellular lysates after vaccine HEK293T cell transfection. YF-17D and Zika virus stocks were diluted to 10⁴ PFU/lane and run in duplicate.

Figure 2

YFV saRNA-NLC vaccination induces serum antigen-binding IgG and neutralizing antibody responses in mice comparable to those induced by the YF-17D vaccine. Mice were vaccinated with a dose range of AAHI-YFV saRNA-NLC and (a, b) serum YFV E protein-binding IgG and (c, d) YFV neutralizing antibody titers were measured (a, c) 28 days post-prime and (b, d) 28 days post-boost. The YF-17D group represents mice that were subcutaneously dosed once with 10⁴ PFU of YF-17D. Statistical analysis was conducted on log₁₀ transformed data using one-way ANOVA with Tukey’s correction for multiple comparisons. ns = non-significant (p > 0.05). Black dotted line shows the limit of detection (LOD) for the assay. Results are from a single independent experiment; n = 10 mice per group, 5 male and 5 female. (a, b) Scatter plots show geometric mean ± geometric SD. (c, d) Box plots show median and IQR ± min/max value. Red numerical values represent group PRNT₅₀ GMT.

Figure 3

Durability of YFV-specific responses from monovalent saRNA-NLC vaccination. (a) Serum YFV E protein-binding IgG and (b) YFV neutralizing antibody responses were assessed 28 days post-prime, 28 days post-boost, and 6 months post-boost with AAHI-YFV saRNA-NLC vaccines. The YF-17D group represents mice that were subcutaneously dosed once with 10⁴ PFU of YF-17D. Statistical analysis was conducted on log₁₀ transformed data using a mixed-effect analysis with Tukey’s correction for multiple comparisons. Black dotted line shows the limit of detection (LOD) of the assay. ns = non-significant (p > 0.05). Results are from a single independent experiment; n = 10 mice per group, 5 male and 5 female. Animals were removed from the 6-month time point due to length of study: 1 animal from the 10 µg group, 1 animal from the 20 µg group, 3 animals from the 30 µg group, and 5 animals from the YF-17D group. (a) Scatter plots show geometric mean ± geometric SD. (b) Box plots show median and IQR ± min/max value. Red numerical values represent group PRNT₅₀ GMT.

Figure 4

ZIKV-specific serum IgG and neutralizing antibody responses following monovalent saRNA-NLC vaccination. Mice were vaccinated with a dose range of AAHI-ZKV saRNA-NLC and (a, b) serum ZIKV E protein-binding IgG and (c, d) ZIKV neutralizing antibody titers were measured (a, c) 28 days post-prime and (b, d) 28 days post-boost. Statistical analysis was conducted on log₁₀ transformed data using one-way ANOVA with Tukey’s correction for multiple comparisons. Black dotted line shows the limit of detection (LOD) for the assay. ns = non-significant (p > 0.05). Results are from a single independent experiment; n = 10 mice per group, 5 male and 5 female. (a, b) Scatter plots show geometric mean ± geometric SD. (c, d) Box plots show median and IQR ± min/max value. Red numerical values represent group PRNT₅₀ GMT.

Figure 5

Serum YFV and ZIKV neutralizing antibody titers resulting from different bivalent vaccination strategies. Mice were vaccinated with monovalent or bivalent formulations of AAHI-YFV or AAHI-ZKV saRNA-NLC and serum neutralizing antibody titers against (a, b) YFV and (c, d) ZIKV were measured (a, c) 28 days post-prime and (b, d) 28 days post-boost. Statistical analysis was conducted on log₁₀ transformed data using one-way ANOVA with Tukey’s correction for multiple comparisons. Comparisons were made against monovalent AAHI-YFV or AAHI-ZKV and between bivalent AAHI-YFV/ZKV groups. ns = non-significant (p > 0.05). Black dotted line shows the limit of detection (LOD) for the assay. Results are from a single independent experiment; n = 10 mice per group, 5 male and 5 female. Box plots show median and IQR ± min/max value. Red numerical values represent group PRNT₅₀ GMT.

Figure 6

Monovalent and bivalent YFV and ZIKV saRNA-NLC vaccination induces polyfunctional CD4⁺ and IFN-γ⁺ CD8⁺ T cells. Mice were vaccinated with monovalent or bivalent formulations of AAHI-YFV or AAHI-ZKV saRNA-NLC, and splenocytes were isolated (a, c, e, g) 28 days post-prime or (b, d, f, h) 28 days post-boost. The antigen-specific polyfunctional (IFN-γ⁺ IL-2⁺ TNFα⁺) CD4⁺ T cell response was measured following stimulation with (a, b) YFV or (c, d) ZIKV prM-E peptide pools. The antigen-specific IFN-γ⁺ CD8⁺ T cell response was measured following stimulation with (e, f) YFV or (g, h) ZIKV prM-E peptide pools. Statistics were measured using Kruskal-Wallis with Dunn’s correction (a-e, g, h) or Brown-Forsythe and Welch ANOVA with Dunnett’s T3 correction (f) for multiple comparisons. Comparisons were made against monovalent AAHI-YFV or AAHI-ZKV and between bivalent AAHI-YFV/ZKV groups. ns= non-significant (p > 0.05). Results are from a single independent experiment; n = 10 mice per group, 5 male and 5 female. Scatter plots show mean ± SD.

Figure 7

Monovalent and bivalent ZIKV saRNA-NLC vaccination protects mice from lethal ZIKV challenge. (a) Post-prime mouse ZIKV challenge study design (–44) (b) Serum ZIKV neutralizing antibody titer 28 days post-prime. (c) Post-prime survival curves. (d) Post-prime body weight. (e) Post-boost mouse ZIKV challenge study design (–44) (f) Serum ZIKV neutralizing antibody titer 28 days post-boost. (g) Post-boost survival curves. (h) Post-boost body weight. (b, d) Statistical analysis was conducted on log₁₀ transformed data using one-way ANOVA with Dunnett’s correction for multiple comparisons. ns = non-significant (p > 0.05). Black dotted line shows the limit of detection (LOD) for the assay. Results are from a single independent experiment; n = 6 mice per group, 3 male and 3 female. Box plots show median and IQR ± min/max value. Red numerical values represent group PRNT₅₀ GMT. (c, g) Statistics were assessed by Mantel-Cox log-rank test against 10 µg SEAP-expressing vector control. *p < 0.05, **p < 0.01. Results are from a single independent experiment; n = 10 mice per vaccination group, 5 male and 5 female.

Figure 8

Monovalent and bivalent YFV saRNA-NLC vaccination protects hamsters from lethal YFV challenge. (a) Post-prime hamster YFV challenge study design (42, 44, 46) (b) Pre-challenge serum YFV neutralizing antibody titer 28 days post-prime. (c) Post-prime survival curves. (d) Post-prime body weight. (e) Post-boost hamster YFV challenge study design (42, 44, 46) (f) Pre-challenge serum YFV neutralizing antibody titer 28 days post-boost. (g) Post-boost survival curves. (h) Post-boost body weight. (b, d) Statistical analysis was conducted on log₁₀ transformed data using one-way ANOVA with Dunnett’s correction for multiple comparisons. ns = non-significant (p > 0.05). Black dotted line shows the limit of detection (LOD) for the assay. Box plots show median and IQR ± min/max value. Red numerical values represent group PRNT₅₀ GMT. (c, g) Statistics were assessed by Mantel-Cox log-rank test against 10 µg SEAP-expressing vector control. *p < 0.05, **p < 0.01, ***p < 0.001. Results are from a single independent experiment; n = 10 hamsters per group, 5 male and 5 female, or n = 5 female hamsters for unvaccinated/uninfected (Unvax/Uninf) controls.