Fig 1.
A: Preparation protocol for Erythro-VLPs: Erythrocyte liposomes were prepared from human RBCs. 14 mg/ml erythrocyte liposomes were incubated in a 3 μM and 25 mM Triton-X100 solution. The detergent was removed by Amberlite XAD-2 and size-exclusion chromatography (SEC). B: Snapshot of a MD simulation of the S-protein in a 25 mM aqueous Triton-X 100 solution after 500 ns. The three chains of the protein are visualized as black, red and blue tubes; Triton-X 100 is represented by cyan rods (hydrophilic head group) and purple rods (hydrocarbon tails). C and D: Snapshots of the S-protein in aqueous solution after 500 ns with and without Triton-X 100, respectively. The angle Θ measures the tilt of the TMD relative to the ectodomain trimer and is plotted in E. F MD snapshot after 50 ns of the S-protein insertion process into the erythrocyte membrane. G Snapshot after 500 ns, with the S-protein fully embedded in the membrane. Triton-X 100 density maps from both simulations averaged along the y-axis are displayed in H and J, maps averaged along the z-axis in K and L.
Fig 2.
A: Size exclusion chromatogram of the Erythro-VLPs showing two signals from Erythro-VLPs and Triton-X100. B: Size distribution of Erythro-VLPs, as determined by DLS. While erythrocyte liposomes measured 102 nm (polydispersity: 0.19) an average diameter of 222 nm (polydispersity: 0.32) was determined for the S-protein carrying liposomes. C: Binding of Erythro-VLPs to human ACE-2 protein was measured by biolayer interferometry (BLI). A dose-dependent reduction in BLI signal was observed upon exposure of the ACE-2 immobilized biosensors to increasing concentrations of Erythro-VLPs, consistent with the binding of large particles to the optical biosensor. D: Association and dissociation curves for Erythro-VLPs in the absence (light purple) and presence (dark purple) of human ACE-2 immobilized onto the biosensor. The dark purple curve is reproduced from C (8×) for the purpose of comparison. E: Schematic of the BLI. Biotinylated human ACE-2 was immobilized onto the Streptavidin BLI sensor. The sensor was then exposed to Erythro-VLPs and association and dissociation was monitored.
Fig 3.
A: Protein structure of the SARS-CoV-2 S-protein. The protein is shown as ribbon diagram and cysteine is shown as sphere. The red and green color indicates solvent accessible and non-accessible cysteine residues. The Solvent Accessible Surface Area (SASA) was determined by the Getarea software and is graphed in B. C: Epifluorescent microscopy images of giant Erythro-VLPs grown on agarose gel. The membrane was stained in red using TR-DHPE; the SARS-CoV-2 S-protein was stained in green using Alexa Fluor 488 maleimide. D: CLSM images of a cluster of giant liposomes after harvesting from the agarose. E: Magnified image of one isolated giant Erythro-VLP taken with CLSM. Images in C-E show the red-, green-, and combined fluorescent channel, respectively. F: Cryo-TEM images of erythrocyte liposomes and Erythro-VLPs. Liposome sizes of ∼100 nm and ∼230 nm agree well with the results of DLS. S-proteins with their TMD anchored in the erythrocyte membranes are observed.
Fig 4.
A: Time line of the mouse study. Each mouse received 3 injections of Erythro-VLPs suspended in sterile saline buffer at 0, 5, and 10 days. The liposome concentration in each dose was approximately 30 nM containing 8 μg of the S-protein. Blood was drawn at 0 (control), 7, and 14 days. Final draw was after 28 days. B: An enzyme-linked immunosorbent assay (ELISA) was used for antibody detection. C: Optical density as function of time for the ELISA essay for all samples. D: Measured optical density ratios. Bars represent the mean optical density ratio averaged over all three dilution runs. Values above 1 ratio are considered positive in the SARS-CoV-2 antibody ELISA. A strong antibody response was observed in both mice after 14 days; no response was observed in the control.