A self-amplifying RNA RSV prefusion-F vaccine elicits potent immunity in pre-exposed and naïve non-human primates

Abstract

Newly approved subunit and mRNA vaccines for respiratory syncytial virus (RSV) demonstrate effectiveness in preventing severe disease, with protection exceeding 80% primarily through the generation of antibodies. An alternative vaccine platform called self-amplifying RNA (saRNA) holds promise in eliciting humoral and cellular immune responses. We evaluate the immunogenicity of a lipid nanoparticle (LNP)-formulated saRNA vaccine called SMARRT.RSV.preF, encoding a stabilized form of the RSV fusion protein, in female mice and in non-human primates (NHPs) that are either RSV-naïve or previously infected. Intramuscular vaccination with SMARRT.RSV.preF vaccine induces RSV neutralizing antibodies and cellular responses in naïve mice and NHPs. Importantly, a single dose of the vaccine in RSV pre-exposed NHPs elicits a dose-dependent anamnestic humoral immune response comparable to a subunit RSV preF vaccine. Notably, SMARRT.RSV.preF immunization significantly increases polyfunctional RSV.F specific memory CD4⁺ and CD8⁺ T-cells compared to RSV.preF protein vaccine. Twenty-four hours post immunization with SMARRT.RSV.preF, there is a dose-dependent increase in the systemic levels of inflammatory and chemotactic cytokines associated with the type I interferon response in NHPs, which is not observed with the protein vaccine. We identify a cluster of analytes including IL-15, TNFα, CCL4, and CXCL10, whose levels are significantly correlated with each other after SMARRT.RSV.preF immunization. These findings suggest saRNA vaccines have the potential to be developed as a prophylactic RSV vaccine based on innate, cellular, and humoral immune profiles they elicit.

Fig. 1. SMARRT.RSV.preF design and in vitro activity.

a Design of SMARRT.RSV.preF. The conserved sequence element (CSE), needed for replication, downstream loop (DLP), to promote translation, non-structural proteins 1–4, which form the replication complex, sub genomic promoter (SG) to drive sub genomic RNA synthesis, transgene region encoding prefusion stabilized RSV F, and poly A tail (schematics not to scale). b Total and c Surface expression of RSV.F protein in BHK cells 24 h post-transfection with LNP formulated SMARRT.RSV.preF. Data for two separate batches of formulated material are presented as mean percentage of cells expressing RSV.preF protein ± SEM of n = 2 (closed circles) biological replicates tested for each concentration of LNP formulated SMARRT.RSV.preF. Source data are provided as a Source Data file.

Fig. 2. SMARRT.RSV.preF vaccination elicits humoral and cellular response in mice.

a Vaccines SMARRT.RSV.preF (0.1, 1, and 10 mcg) or Ad26.RSV.preF (10¹⁰ VPs) were administered intramuscularly to female BALB/c mice. While Ad26.RSV.preF was given as a single dose, SMARRT vaccine was administered as a 2-dose homologous regimen given 4 weeks apart. Components of this figure were sourced from Open Clip Art, under a Creative Commons Attribution 3.0 Unported License; https://creativecommons.org/licenses/by/3.0/. The serum from n = 8 biologically independent animals per group was sampled on days 0, 27 and 56. b RSV.CL57 neutralizing antibody titers and c RSV.preF specific IgG antibodies in the serum of immunized animals measured at day 27 (open circles) and day 56 (open squares). Limit of detection (LoD) is represented by the dotted line. Gray lines represents paired measurements. Statistical comparisons of day 27 to day 56 humoral responses were determined with ANOVA and adjusted for multiple comparisons with Bonferroni correction of log transformed data. d RSV.F specific T-cells measured in the spleens of 10 mcg SMARRT.RSV.preF immunized animals (n = 8 biologically independent animals), at day 56, by ICS. Different subsets of IFNγ, TNFα and IL2 positive CD8⁺ and CD4⁺ T-cells gated on Live CD45⁺ cells. Gray circles represent splenocytes stimulated with RSV.F peptide pool and black circles represent corresponding medium stimulated samples. Red horizontal lines represent median response. Source data are provided as a Source Data file.

Fig. 3. Humoral immunogenicity elicited by SMARRT.RSV.preF vaccine in RSV infected and in naïve NHPs.

a Cynomolgus NHP (n = 12) infected with RSV at week −12 were distributed into 3 groups based on RSV.postF antibody titers measured at week −8. At week 0, pre-exposed animals (n = 4/group) received an intra-muscular immunization with either 50 mcg of PRPM (group 1) or 1 mcg (group 2); 10 mcg (group 3) of SMARRT.RSV.preF. A RSV naïve group (n = 4) received 10 mcg of SMARRT.RSV.preF vaccine (group 4). PBMCs and serum samples were collected from n = 4 biologically independent animals per group at weeks 1, 2, 4, 8, and 12 in addition to nasal swabs (week 8). Components of this figure were sourced from Google Images, under a Creative Commons Attribution 2.0 Unported License; https://creativecommons.org/licenses/by/2.0/. b RSV.CL57 neutralizing antibodies; c AUC for RSV.CL57 VNA and d RSV.preF IgG antibodies determined for each animal (gray or black circles) from week 0 through week 12 of the study. The geometric mean titers (GMT) and LoD for each group is represented by the red horizontal line and the dotted line. e RSV.preF specific IgA antibodies measured in nasal swab elutes (n = 4/group). Samples were only included if they were free of blood contamination, with maximum number of uncontaminated samples available from weeks 0 and 8. Measurements were corrected for total protein content in elutes. The dotted line represents the LoD while the gray solid line represents the upper limit of 95% CI of mean response in RSV-infected animals measured at week 0 prior immunization. Statistical significance was determined with ANOVA and adjusted for multiple comparisons with Bonferroni correction between (c, e) RSV-infected and naïve animals following SMARRT.RSV.preF immunization (b, d) between responses at weeks 2 and 12 to week 0. Source data are provided as a Source Data file.

Fig. 4. Magnitude and polyfunctionality of cellular responses elicited by SMARRT.RSV.preF in NHPs.

a RSV.F specific IFNγ secreting cells in PBMCs of immunized animals (n = 4 biologically independent animals per group) represented as number of Spot Forming Units (SFU) per million cells. The dotted line represents the positivity threshold of the assay which is set at 50 SFU and the red line indicates the geometric mean response. Background subtracted response of each animal (gray circles) overtime is shown. Comparisons were made between baseline response to weeks 8 and 12. b CD4 (square) and CD8 (circle) memory T-cells gated on Live CD45⁺CD28^±D95⁺ positive for CD107a or IFNγ or TNFα or IL2 stimulated with RSV.F peptide pool were identified using intracellular cytokine staining. After subtracting the corresponding response in the medium stimulated sample of each animal, values are shown with a threshold represented by the dotted line and median response by the red line. Due to insufficient number of cells, data from an RSV pre-exposed animal immunized with 10 mcg SMARRT.RSV.preF group is not available (b–d). Polyfunctional subsets of CD4 and CD8 T-cells by Boolean gating and after background subtraction were subsequently analyzed using SPICE at week 4 post-immunization. Bar plots represent background-subtracted median frequency of cells in each subset for each animal (circles). Pie chart wedges represent the functional subsets producing different combinations of cytokines i.e., 4+, 3+, 2+ and monofunctional subsets indicated by the color coding under “Pie”, while the surrounding pie arcs represent total median level of each analyte. Statistical analysis was done on square root transformed values. b–d Comparisons between SMARRT.RSV.preF immunized groups was determined with ANOVA (TOBIT model) and adjusted for multiple comparisons with Bonferroni correction. Source data are provided as a Source Data file.

Fig. 5. Serum chemokine/cytokine profiling from NHPs immunized with SMARRT.RSV.preF.

Analytes were measured in the serum of immunized animals (n = 4 biologically independent samples per group) prior to immunization and at 24 h post immunization. a Principal component analysis of analytes by treatment and sampling timepoints. Color represents the vaccines and corresponding timepoints while shapes represent the pre-exposure status of animals to RSV. b Volcano plots showing significance (-log10 of adjusted p-value) versus log10 fold changes of analytes at 24 h post-immunization with 1 mcg and 10 mcg of SMARRT.RSV.preF vaccine compared to baseline. Statistical comparisons of analyte concentrations between treatment groups were conducted using a mixed effects model (moderated two-sided t-statistics) by the limma package. The gray dotted line shows a significant threshold of adjusted p-value = 0.05 with analytes that are significantly upregulated indicated by the closed black circles and those that are significantly downregulated represented by orange closed circles. Analytes that fall below the significance threshold are indicated by open circles below the dotted line. c Correlations of log10 analyte concentrations presented as a heatmap correlogram in animals dosed with 10 mcg of SMARRT.RSV.preF. Spearman rank correlation coefficients (r) between each pair of analytes with significant (p < 0.05) non-adjusted p-values are shown in the correlation matrix with the color intensity of the circles indicating the direction and degree of correlation. The red box identifies a cluster of analytes namely CXCL10, IL15, TNF and CCL4 that have a pairwise positive correlations with each other. Individual, two-tailed, Spearman correlation plots of these analytes in this cluster for each animal are shown with Spearman coefficient values (r) and non-adjusted p-values. Source data are provided as a Source Data file.