Generation of potent cellular and humoral immunity against SARS-CoV-2 antigens via conjugation to a polymeric glyco-adjuvant

Abstract

The SARS-CoV-2 virus has caused an unprecedented global crisis, and curtailing its spread requires an effective vaccine which elicits a diverse and robust immune response. We have previously shown that vaccines made of a polymeric glyco-adjuvant conjugated to an antigen were effective in triggering such a response in other disease models and hypothesized that the technology could be adapted to create an effective vaccine against SARS-CoV-2. The core of the vaccine platform is the copolymer p(Man-TLR7), composed of monomers with pendant mannose or a toll-like receptor 7 (TLR7) agonist. Thus, p(Man-TLR7) is designed to target relevant antigen-presenting cells (APCs) via mannose-binding receptors and then activate TLR7 upon endocytosis. The p(Man-TLR7) construct is amenable to conjugation to protein antigens such as the Spike protein of SARS-CoV-2, yielding Spike-p(Man-TLR7). Here, we demonstrate Spike-p(Man-TLR7) vaccination elicits robust antigen-specific cellular and humoral responses in mice. In adult and elderly wild-type mice, vaccination with Spike-p(Man-TLR7) generates high and long-lasting titers of anti-Spike IgGs, with neutralizing titers exceeding levels in convalescent human serum. Interestingly, adsorbing Spike-p(Man-TLR7) to the depot-forming adjuvant alum amplified the broadly neutralizing humoral responses to levels matching those in mice vaccinated with formulations based off of clinically-approved adjuvants. Additionally, we observed an increase in germinal center B cells, antigen-specific antibody secreting cells, activated T follicular helper cells, and polyfunctional Th1-cytokine producing CD4⁺ and CD8⁺ T cells. We conclude that Spike-p(Man-TLR7) is an attractive, next-generation subunit vaccine candidate, capable of inducing durable and robust antibody and T cell responses.

Keywords: Adjuvant; COVID-19 vaccine; Glycopolymers; Polymer-protein conjugates; Polymeric glyco-adjuvant; Subunit vaccine formulation.

Fig. 1

The prefusion-stabilized Spike antigen conjugated to p(Man-TLR7) is a potent activator of BMDCs. (A) Spike-p(Man-TLR7) is composed of Spike antigen (i.) conjugated, via a self-immolative linker (ii.), to a random copolymer synthesized from monomers that either activate TLR7 (red ovals) or target mannose-binding C-type lectins (blue ovals; iii.). (B) SDS-PAGE analysis of Spike before (i.) and after the two step conjugation reaction (ii., iii.). The observed band between 15 and 25 kDa in the Spike-p(Man-TLR7) conjugate (iii.) comes from free p(Man-TLR7) polymer. (C) Analysis of the binding ability of Spike-p(Man-TLR7) to human ACE2 (hACE2) via enzyme-linked immunosorbent assay (ELISA). (D) Concentration of IL-6, TNFα and IL-12p70 in the supernatant of BMDCs stimulated for 18h with either Spike or Spike-p(Man-TLR7) at the concentration corresponding to 25 μM of the adjuvant, as determined by ELISA. Dotted horizontal lines represent the assay background. In (C and D), columns and error bars indicate mean + SD; statistical comparisons are based on one-way ANOVA with Tukey's post-test: ***p < 0.001; #p < 0.001 as compared to bovine serum albumin (BSA). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 2

Spike-p(Man-TLR7) and Spike-p(Man-TLR7)+alum generate potent humoral responses in mice. (A) Mice were vaccinated with Spike-p(Man-TLR7), Spike-p(Man-TLR7)+alum, Spike + AS04-L, Spike + AS03-L, Spike + alum, or Spike at weeks 0 (prime) and 3 (boost), and their plasma was collected weekly until week 4. Plasma from naïve mice was collected at the same time points. (B) Total Spike-specific IgG antibodies over time reported as the area under the log-transformed curve (AUC) of absorbance vs. dilution. (C) Comparison of Spike-specific IgG isotypes (IgG1, IgG2b, IgG2c and IgG3) and (D) corresponding IgG2b:IgG1 ratios at one week post-boost (week 4). (E) Circulating anti-Spike IgA antibodies in the serum of vaccinated mice quantified at week 4 using AUC analysis. (F) Quantification of Spike-specific IgG⁺ antibody secreting cells by enzyme-linked immunosorbent spot (ELISpot) assay with splenocytes. (G) Neutralization assay of SARS-CoV-2 infection on Vero-E6 cells in vitro. SARS-CoV-2 was pre-incubated with plasma isolated from mice at week 4. Percent neutralization was calculated based on viability of cells that did not receive virus (100%) or virus without plasma preincubation (0%). (H) Viral neutralization titers, representing plasma dilution at which 50% of SARS-CoV-2-mediated cell death is neutralized. Shaded area represents the lower limit of detection (titer of 2.11); green dotted horizontal line represents the FDA recommendation for “high titer” classification (= 2.40). (I) Comparison of total Spike-specific IgG antibodies in the plasma of 8, 21, 47 and > 64 week old mice that received the indicated vaccines, following the same schedule as in (A). (J) Change in total Spike-specific IgG antibodies over time in plasma of mice (n = 5) that received the indicated vaccines, following the same vaccination schedule as in (A). All data are presented as mean ± SEM with n = 4–10 mice per group, unless stated otherwise. Comparisons were made using (B and I) two-way ANOVA with Tukey's multiple comparison test, (C and E) Brown-Forsythe ANOVA with Dunnet's T3 test, (D) one sample t-test (compared to the theoretical value of 1, representing an unbiased Th1/Th2 response), (F and H) Kruskal-Wallis non-parametric test with Dunn's post-test, or (J) mixed-effects analysis with Tukey's multiple comparison test: *p < 0.05, **p < 0.01, ***p < 0.001; #p < 0.05 (for comparison to both Spike and naïve groups). Additional comparisons were made in (H) using Wilcoxon signed rank test: §p < 0.05 and ‡p = 0.11 (as compared to the FDA “high titer” classification). In (B), comparisons noted on the graph are between the indicated timepoints for all groups except Spike and Naïve, while comparisons noted in the legend are between the indicated groups at week 4. In (I), comparisons indicated in the legend are true for mice at each age. In (J), comparisons noted on the graph in black are between the indicated timepoints for all groups except Spike and Naïve, and comparisons indicated in red are only for Spike-p(Man-TLR7). Comparisons indicated in the legend of (J) are true for every timepoint. In (B, I, and J), only relevant statistical comparisons are shown. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 3

Vaccination with Spike-p(Man-TLR7) and Spike-p(Man-TLR7)+alum elicits a broad humoral response targeting the receptor binding motif (RBM) of RBD and other neutralizing linear epitopes. Mice were vaccinated as in Fig. 2A. Plasma was collected at week 4, pooled by vaccination group, and analyzed for binding to linear epitopes using a peptide array. X-axis represents the sequential peptide number within the Spike amino acid sequence (overlapping 15-amino acid peptides with 5-amino acid offsets). Y-axis quantifies the level of antibody binding to each peptide, detected via luminescence (a.u.). Axis begins from the value of the background, which was set at 5 × 10⁶ a.u. Several relevant regions of the Spike protein are indicated above the graphs: S1 and S2 subunits, RBD (light orange box), and RBM (dark orange box). Regions corresponding to neutralizing Spike epitopes identified by Shrock et al. [70] and Poh et al. [71] are indicated in light green and light blue boxes, respectively. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 4

Secondary lymphoid organ-resident B cells and CD4⁺ T follicular helper (T_fh) cells are activated in mice vaccinated with Spike-p(Man-TLR7) or Spike-p(Man-TLR7)+alum. (A and B) Quantification of B cells resident within the (A) draining lymph nodes (dLNs) of the vaccination site or (B) spleen via flow cytometry at week 4 after vaccination as in Fig. 2A. From left to right, total B cells (CD19⁺ B220⁺) within live cells, germinal center (GC) B cells (IgD⁻GL7⁺CD38⁻) within live B cells, and RBD tetramer reactive GC B cells (RBD⁺) as a percentage of GC B cells. (C) Activation of CD4⁺ T_fh cells in the dLNs and the spleen was characterized by flow cytometry. T_fh cells were defined as CXCR5⁺BCL6⁺CD4⁺ and quantified within CD4⁺ T cells. ICOS^hi T_fh cells were quantified within T_fh cells. Data plotted as mean ± SEM with n = 4–5 mice per group. *p < 0.05, **p < 0.01, ***p < 0.001 by one-way ANOVA with Tukey's post-test; #p < 0.05 as compared to both Spike and naïve.

Fig. 5

Vaccination with Spike-p(Man-TLR7) and Spike-p(Man-TLR7)+alum elicits robust antigen-specific T cell responses. Splenocytes harvested one week post-boost from vaccinated mice (vaccinated as in Fig. 2A) were restimulated ex vivo with a Spike-derived peptide pool for 6h prior to flow cytometry analysis or with the full length Spike protein for 3 days prior to multiplexed cytokine analysis. (A to D) Cytokine-producing (A and B) CD4⁺ and (C and D) CD8⁺ T cell responses were detected by intracellular staining and quantified by flow cytometry. (A) IFNγ⁺, TNFα⁺ and IL2⁺ CD4⁺ T cells quantified within CD4⁺ T cells. (B) Polyfunctional CD4⁺ T cells (IFNγ⁺ TNFα⁺ IL2⁺ CD4⁺ T cells) quantified as a percentage of CD4⁺ T cells. (C) IFNγ⁺, TNFα⁺ and IL2⁺ CD8⁺ T cells, as a percentage of CD8⁺ T cells. (D) Polyfunctional CD8⁺ T cells (IFNγ⁺ TNFα⁺ IL2⁺ CD8⁺ T cells), as a percentage of CD8⁺ T cells. (E) Cytokine production by splenocytes after 3 day ex vivo restimulation with full-length Spike protein. The cytokines quantified (in pg/mL) include Th1 cytokines (IFNγ, TNFα, IL-2), Th2 cytokines (IL-4, IL-5, IL-13), and IL-6. Data presented as mean ± SEM with n = 4–5 mice per group; *p < 0.05, **p < 0.01 by Kruskal-Wallis test with Dunn's post-test.