Supporting Information
Figure S1. 6pVLPs specifically bind to human ACE2 receptor coated beads but not to recombinant IL-1β coated beads . Carboxyl modified latex beads were coated with hACE2 or recombinant IL-1β and then incubated with serially diluted CFSE labeled VLPs. Beads were washed once and analyzed by flow cytometry for bead-bound VLP-CFSE signal. Top dot plots show the gating strategy and the lower plots represent bead-bound CFSE signal (y-axis) versus forward scatter (x-axis). Histograms present VLP CFSE signal intensity (mean fluorescence intensity, MFI) of serially diluted VLPs associated with negative control IL-1β beads (top 4 histograms) or hACE2 coated beads (lower 4 histograms).
Figure S2. Stability of SARS-CoV-2 VLPs under elevated temperature (a) and following adsorption to alum and CpG adjuvantation (b) . (a ) 6p VLPs were subjected to thermal stress for the indicated time periods. Antigen identities were assessed by western blot (anti-His Tag antibodies) and compared to a sample preserved at 5ºC for one week. Positions of S, M and E proteins are indicated by arrows. (b) Transmission electron microscopy image of formulated 6p VLPs demonstrate that VLPs retain their morphology after adsorption to alum and CpG adjuvantation.
Figure S3. VLPs elicit robust antibody responses in mice.BALB/c mice (n = 12 per group) were immunized on days 0 and 14 with 0.4 ((a ) and (b ), low dose; LD) or 4 µg ((c ) and (d ) high dose; HD) 6p VLP or 2p VLPs without or with Alum (5 µg/mouse), K3 CpG ODN (20 µg/mouse) or Alum+CpG ODN. Control BALB/c mice were administered Alum or CpG ODN alone (black and gray). Sera were collected 2 weeks post- 2 weeks post-boost and assessed for SARS-CoV-2 RBD- (a and c ) or N-specific (b andd ) IgG, IgG1 and IgG2a by enzyme-linked immunosorbent assay (ELISA). Vaccinated groups were compared by one-way ANOVA with Dunnett’s multiple comparisons test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Data are presented as GMT ± geometric SEM.
Figure S4. 6p VLP+Alum+CpG induce S-specific TH1 dominated helper T cells. (a ) 1×106/250 µL splenocytes from naïve or immunized BALB/c mice (n=6) were stimulated with recombinant nucleocapsid (20 µg/ml) in the presence of 1 μg/mL anti-mouse CD28. T helper cytokine levels were assessed from 48h culture supernatants using the LEGENDplex™ MU Th Cytokine Panel (12-plex). Groups were compared by one-way ANOVA with Dunnett’s multiple comparisons test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Data are presented as mean cytokine levels ± SEM. (b ) Pie charts representing the proportions of individual secreted N-specific T helper cytokines are presented.