Enhanced attenuation of chikungunya vaccines expressing antiviral cytokines

Abstract

Alphaviruses are vector-borne, medically relevant, positive-stranded RNA viruses that cause disease in animals and humans worldwide. Of this group, chikungunya virus (CHIKV) is the most significant human pathogen, responsible for generating millions of infections leading to severe febrile illness and debilitating chronic joint pain. Currently, there are limited treatments to protect against alphavirus disease; thus, there is a tremendous need to generate safe and effective vaccines. Live-attenuated vaccines (LAVs) are cost-effective and potent immunization strategies capable of generating long-term protection in a single dose. However, LAVs often produce systemic viral replication, which can lead to unwanted post-vaccination side effects and pose a risk of reversion to a pathogenic phenotype and transmission to mosquitoes. Here, we utilized a chimeric infectious clone of CHIKV engineered with the domain C of the E2 gene of Semliki Forest virus (SFV) to express IFNγ and IL-21-two potent antiviral and immunomodulatory cytokines-in order to improve the LAV's attenuation while maintaining immunogenicity. The IFNγ- and IL-21-expressing vaccine candidates were stable during passage and significantly attenuated post-vaccination, as mice experienced reduced footpad swelling with minimal systemic replication and dissemination capacity compared to the parental vaccine. Additionally, these candidates provided complete protection to mice challenged with WT CHIKV. Our dual attenuation strategy represents an innovative way to generate safe and effective alphavirus vaccines that could be applied to other viruses.

Fig. 1. Genome organization of chimeric cytokine-expressing vaccines.

Genome schematic of a (a) standard alphavirus genome, (b) the chimeric CHIKV genome with SFV domC (in red) expressing a nanoluc (nLuc) reporter gene, (c) the mouse IFNγ-expressing CHIKV/SFV chimera, and (d) the mouse IL-21-expressing CHIKV/SFV chimera. T2A refers to the Thosea asigna virus 2A self-cleaving peptide, which cleaves the foreign protein away from the viral envelope proteins.

Fig. 2. In vitro vaccine characterization.

The concentration of mouse interferon-gamma (IFNγ) or mouse interleukin-21 (IL-21) (a) with respect to CHIKV-SFV/DomC was assessed by ELISA in viral supernatants rescued from HEK293T cells. Genetic stability of each cytokine was assessed by measuring the (b) concentration of IFNγ or IL-21 in CHIKV-SFV/DomC-IFNγ or CHIKV-SFV/DomC-IL-21, respectively, after viral passaging in Vero cells. The line for CHIKV-SFV/DomC represents values for both cytokines as values were undetectable for either cytokine. Each datapoint represents an independent passage series. Viral kinetics of all vaccine strains and WT CHIKV in (c) baby hamster kidney (BHK-21), (d) 3T3 mouse fibroblasts, and (e) U4.4 mosquito cells after infection at a multiplicity of infection of 0.01 PFU/cell. Data are shown as two biological replicates conducted in triplicate for ELISAs with statistical comparisons analyzed through unpaired t-tests. Concentrations were extrapolated from standard curves for each cytokine. Growth curve data is reported from two biological replicates for mammalian cells and one biological replicate for mosquito cells conducted in triplicates per group. Statistical comparisons with respect to the parental CHIKV-SFV/DomC were done using a mixed-effects model with Dunnett’s multiple comparisons test *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001.

Fig. 3. Vaccination and challenge of immunocompetent C57BL/6 mice.

a Four-week-old male C57BL/6 mice (n = 5 per group) were inoculated via the left hind footpad with 10⁴ PFU of vaccine candidate or mock-infected with viral diluent. b Footpad swelling was measured daily and analyzed by area under the curve (AUC) analysis. c Mice were bled 3 days post-vaccination (dpv) to measure viremia by TCID50. Mice were challenged with 10³ PFU of WT CHIKV via the footpad 31 or 71 dpv with (d, f) footpad swelling monitored for each group, respectively. Mice were bled up to 3 days post-challenge to measure viremia as done previously (e, g). Post-vaccination data comprise of two independent experiments and one independent experiment for short- and longer-term challenges, respectively. Bars indicate standard deviations and dotted lines represent the limit of detection. Statistical comparisons were made by 2-way ANOVA, with Dunnett’s multiple comparisons test for footpad swelling *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001.

Fig. 4. Vaccination of transiently immunocompromised C57BL/6 mice.

a Mice (n = 10 per group) were inoculated via the footpad with 10⁴ PFU of vaccine candidate or mock-infected with viral diluent 24 h after MAR1 antibody (0.1 mg) treatment. b Weight and (c) footpad swelling were measured daily, with area under the curve (AOC) analyses presented, and (d) mice were bled daily for 3 days post-vaccination (dpv) to measure viremia by plaque assay. e Antibodies were assessed 30 dpv for percent WT CHIKV neutralization over a series of serum dilutions. f Mean PRNT₅₀ values were determined through non-linear regression analysis and compared to CHIKV-SFV/DomC. g Isotype and antibody subclass responses after 30 dpv were measured by ELISA. Bars indicate standard deviations and dotted lines represent the limit of detection or 50% neutralization. Data is representative of two biological replicates of equal numbers of male and female mice. Statistical comparisons were made compared to CHIKV-SFV/DomC by either mixed-effects analysis or 2-way ANOVA with Dunnett’s multiple comparisons test for weights and footpad swelling, and viremia, %neutralization, and antibody isotyping, respectively. PRNT₅₀ values were compared by one-way ANOVA with Dunnett’s multiple comparisons test; *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001.

Fig. 5. Wildtype CHIKV challenge of vaccinated immunocompromised C57BL/6 mice.

a Vaccinated and mock-vaccinated mice (n = 10 per group) were inoculated via the footpad with 10³ PFU of WT CHIKV SL15649 32 days after vaccination after an additional MAR1 antibody (0.1 mg) treatment. b Weight and (c) footpad swelling were measured daily, and (d) mice were bled daily for 3 dpi to measure viremia by plaque assay. e Survival over the course of the study. f Representative images of footpad and tail swelling for mock-vaccinated compared to vaccinated mice are shown. Bars indicate standard deviations and dotted lines represent the limit of detection. Data is representative of two biological replicates using both male and female mice. Statistical comparisons were made compared to CHIKV-SFV/DomC by either mixed-effects analysis or 2-way ANOVA with Dunnett’s multiple comparisons test for weights and footpad swelling, and viremia, respectively; *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001.

Fig. 6. Viral dissemination and in vivo cytokine kinetics.

Mice (n = 6–10 per group/day) were inoculated via the footpad with 10⁴ PFU of each vaccine candidate or mock-infected with viral diluent after MAR1 antibody (0.1 mg) treatment. a The inoculated footpad was dissected 1, 2, and 3 days post-vaccination (dpv) to assess viral titers by plaque assay. Viral titers were also assessed in the (b) calf tissues, (c) the popliteal lymph node [PL_LN], and (d) the inguinal lymph node [IN_LN]. Data was collected from two independent experiments. e, f Levels of interferon-gamma (IFNγ) and interleukin-21 (IL-21) were assessed in the inoculated footpads of vaccinated mice 1, 2, and 3 dpv. Bars indicate standard deviations and dotted lines represent the limit of detection. Statistical comparisons against CHIKV/SFV-DomC were performed using 2-way ANOVA with Dunnett’s or Šídák’s multiple comparisons test for viral titers and ELISAs, respectively. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001.