Enhanced stability of the SARS CoV-2 spike glycoprotein following modification of an alanine cavity in the protein core

Fig 1. A, Three-dimensional structure of the SARS CoV-2 S ectodomain drawn with the coordinates PDB ID 6VSB [4]. S1 is shown in green and blue, RBM, ACE2 receptor binding motif with receptor-interacting amino acids in red; RBD, receptor binding domain in blue; NTD, N-terminal domain in teal. S2 is shown in yellow and pink: FP, fusion peptide in red; central coiled coil in pink with Ala cavity highlighted in purple. B, heptad repeat motifs within the coiled coil sequences of 3 betacoronaviruses. C, a comparison of the central coiled coil in the prefusion (PDB ID 6VSB [4]) and postfusion (PDB 6XRA [9]) conformations. Although incompletely shown here, HR1 helices (C-terminal part shown in yellow) extend the coiled coil upward toward the cellular membrane in the postfusion conformation [9]. D, Illustration of how the Ala cavity of prefusion S might be filled following substitution with hydrophobic amino acids. 16L and VI are codes for A1016L and A1016V/A1020I mutations.

Fig 2. Expression and thermostability screen of Ala cavity mutants. A, Coomassie blue stained S2P-FHA mutants purified from culture supernatants with TALON resin following SDS-PAGE under reducing conditions. The Tms of S2P-FHA mutants obtained in B are indicated below each lane. Bold type: major species; small font: minor species; nd: not determined. B, Differential scanning fluorimetry of S2P-FHA proteins shown in A using SYPRO Orange. The rate of change of fluorescence over time [–d(RFU)/dt] as a function of temperature is shown. Graphs are representative of at least two independent experiments.

Fig 3. Purification and characterization of selected Ala cavity mutants. A, Superose 6 SEC of S2P-FHA proteins following elution from TALON resin. The calibration standards are thyroglobulin (669 kDa), ferritin, (440 kDa) and IgG (150 kDa). B, Superose 6 SEC of purified trimers following a freeze (-80°C)-thaw cycle. C, Differential scanning fluorimetry of purified S2P-FHA trimers. D, SDS-PAGE and Coomassie blue staining of purified proteins under nonreducing (top) and reducing (bottom) conditions.

Fig 4. Effects of Ala cavity mutations on fusion function. A, Reducing SDS-PAGE and western blotting of SARS CoV-2 S glycoproteins expressed in 293T cells with rabbit anti-S1 polyclonal antibody. B, Cell-cell fusion activity of S glycoproteins determined in a luciferase reporter assay. Firefly luciferase relative light units obtained for each coculture was normalized according to Gaussia luciferase activity produced by the corresponding well of effector cells used for each coculture. Relative Gaussia activity was first obtained by dividing Gaussia RLU of each transfection (see S2A Fig) by Gaussia RLU produced by the WT transfection. Firefly luciferase relative light units were then divided by relative Gaussia activity produced by the corresponding effector cells. The mean ± SEM from at least 3 independent experiments shown. ns, not significant, *, P < 0.05, ** P < 0.01, versus WT, Kruskal-Wallis test. C, Representative microscopy fields at 10x magnification of cocultures prior to lysis and firefly luciferase assay. D, Cell-cell fusion activity of S glycoproteins determined in a fluorescence assay. The nuclei of S-expressing 293T effector cells stained with Hoechst 33342 (blue) were…

Fig 5. Immunogenicity of Ala cavity mutants. A, Immunization protocol. B, Pseudovirus neutralization ID50s of vaccinal sera obtained at weeks 6 and 16. The geometric mean pseudovirus neutralization ID50 of the control group of 8 animals receiving 3 doses of 50% Addavax-PBS was <1/120 at both time-points. C and D, ELISA binding titers of week-16 vaccinal sera to streptavidin-captured biotinylated RBDs or S2P-FHA trimers, respectively, derived from ancestral Hu-1, Delta and Omicron BA.1 (indicated below the graphs). The endpoint was determined as 5-times background optical density obtained in the absence of primary antibody. The horizontal bars are the geometric means. The horizontal dotted line is the geometric mean binding titer of the control group receiving 3 doses of 50% Addavax-PBS. A Kruskal-Wallis test was used to determine whether the differences in ID50s or binding titers are significantly different. ns, not significant; *, P < 0.05; **, P < 0.01.

Fig 6. Neutralization activity of immune sera. A) Pseudovirus neutralization ID50s of week-16 vaccinal sera. The S variant genotypes used in S-HIV pseudotype assays are shown below the x axis. Neutralization ID50 obtained with sera from individual animals are indicated with various symbols. The bars are geometric mean ID50s. The geometric mean pseudovirus neutralization ID50 of the control group receiving 3 doses of 50% Addavax-PBS was <1/200. B) Neutralisation assays performed in high-throughput format with authentic SARS-CoV-2 variants. The SARS-CoV-2 variants are shown below the x axis. Horizontal bars are the geometric mean ID50s for each immunogen group. The geometric mean neutralization ID50 of the control group receiving 3 doses of 50% Addavax-PBS was <1/320 for Ancestral and Delta and <1/40 for Omicron BA.1. A Kruskal-Wallis test was used to determine whether the differences in ID50s observed between variants was significant: ns, not significant; *, P < 0.05; **, P < 0.01. C) Correlations between RBD binding titer (obtained in Fig 5C) and pseudovirus neutralization ID50 (top) and authentic virus ID50 (bottom). The data points are color-coded according to variant. Spearman ρ and…

Fig 7. VI stabilizes S2P-FHA trimers derived from Omicron BA. 1. A, Superose 6 SEC profiles of S2P.OmiBA1-FHA and, S2P.OmiBA1.VI-FHA eluted from TALON or HiTRAP columns, respectively. B, Superose 6 SEC profiles of presumed trimers obtained in A. C, Differential scanning fluorimetry of purified S2P.OmiBA1-FHA and, S2P.OmiBA1.VI-FHA trimers. D, SDS-PAGE/Coomassie blue staining of purified trimers under non-reducing and reducing (1% v/v betamercaptoethanol) conditions. The samples were not boiled prior to electrophoresis. m, markers. E, SDS-PAGE of trimers under non-reducing and reducing (1% v/v betamercaptoethanol) conditions after boiling for 5 min. The position of the S2P.omicron.VI-FHA major band under the various conditions is indicated with an arrow. m, markers. F, SDS-PAGE of S2P-FHA, S2P.OmiBA1-FHA, S2P.Omi.VI-FHA trimers under non-reducing (left) and reducing (1% v/v betamercaptoethanol) (right) conditions. Sample buffer containing SDS was added to the samples with and without 1% betamercaptoethanol and the samples were either left at room temperature (22°C) or boiled (100°C) for 3 min prior to electrophoresis. Thyroglobulin that had been chemically crosslinked with 1 mM…

Fig 8. The VI mutation stabilizes the trimerization of the omicron BA.1 and BA.4/5 S ectodomain in the absence of the T4 foldon trimerization motif. A) Superose 6 size exclusion chromatography of S2P-1208.H8 proteins purified from 293F (ancestral) or Expi293F (Omicron BA.1 and BA.4/5) cells by hiTRAP affinity chromatography. B, Superose 6 size exclusion chromatography of S2P-1208.H8 glycoproteins purified in A following a freeze (-80°C)-thaw cycle. C, Differential scanning fluorimetry of purified S2P-1208.H8 proteins following a freeze (-80°C)-thaw cycle performed using SYPRO Orange. D, SDS-PAGE and Coomassie blue staining of purified proteins under reducing conditions. WT: Ala at amino positions 1016 and 1020; VI: Val and Ile at amino positions 1016 and 1020, respectively. m = markers.