ChAdOx1 COVID vaccines express RBD open prefusion SARS-CoV-2 spikes on the cell surface

Abstract

Vaccines against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) have been proven to be an effective means of decreasing COVID-19 mortality, hospitalization rates, and transmission. One of the vaccines deployed worldwide is ChAdOx1 nCoV-19, which uses an adenovirus vector to drive the expression of the original SARS-CoV-2 spike on the surface of transduced cells. Using cryo-electron tomography and subtomogram averaging, we determined the native structures of the vaccine product expressed on cell surfaces in situ. We show that ChAdOx1-vectored vaccines expressing the Beta SARS-CoV-2 variant produce abundant native prefusion spikes predominantly in one-RBD-up conformation. Furthermore, the ChAdOx1-vectored HexaPro-stabilized spike yields higher cell surface expression, enhanced RBD exposure, and reduced shedding of S1 compared to the wild type. We demonstrate in situ structure determination as a powerful means for studying antigen design options in future vaccine development against emerging novel SARS-CoV-2 variants and broadly against other infectious viruses.

Keywords: Cell biology; Virology.

Graphical abstract

Figure 1

Cytometry analysis of SARS-CoV-2 spike expression (A and B) Percentage of cells expressing SARS-CoV-2 spike after ChAdOx1 19E and 19E6 transduction as assessed by staining with Ab222, an S1-targetting monoclonal antibody (A), and an ACE2-Fc chimera (B). (C and D) Cell surface expression levels of SARS-CoV-2 spike after ChAdOx 19E and 19E6 transduction as assessed by staining with Ab222 (C) and an ACE2-Fc chimera (D). (E) S1 concentration in microvesicle-depleted cell-culture supernatant determined by quantitative S1 ELISA. (F) Viability of cells transfected with ChAdOx1 19E and 19E6 at 24 h, 48 h, and 72 h post-infection. p values from two-way ANOVA with Tukey multiple comparisons test. Error bars represent the stardard deviation (SD). Experiment was performed two times with biological triplicates. See also Figure S1.

Figure 2

Subtomogram averaging of ChAdOx1 spikes in situ (A) A tomogram slice of 19E6 infected cell showing spikes decorating the cell membrane. Red arrows point to the representative spikes. Inset box shows the spikes in the top slide of the filopodia. Scale bar = 100 nm. (B) Orthogonal views of subtomogram averaging of ChAdOx1 spike 19E6 (top) and 19E (bottom) with C3 symmetry applied. (C) Orthogonal views ChAdOx1 spike 19E6 density map overlay with a structure model of spike (PDB 6ZGG). (D) A close-up view of the boxed region in C. The density extrusions indicate good fit with the N-linked glycosylation. Several glycans are highlighted. (E) Pairwise distance distribution of spikes for ChAdOx1 19E and 19E6 samples. (F) Model of dimerized spikes (EMD-22354) has a distance of 10 nm between spikes. The density map EMD-22354 was Gaussian filtered for presentation. See also Figures S2–S4, Table S1 and Videos S1 and S2.

Figure 3

Classification of spike conformations (A) Three classes were identified in both ChAdOx1 19E and ChAdOx1 19E6 spikes: one RBD up (blue), three RBD down (brown), and two RBD down with one RBD flexible (gray). (B) Distribution of the three spike conformations in both ChAdOx1 19E and ChAdOx1 19E6 vaccines. (C) STA map of ChAdOx1 19E6 spike in one-RBD-up conformation at 9.6 Å resolution. (D) Comparison between ChAdOx1 19E6 spike (one RBD up) in situ with cryo-EM SPA study (PDB 6ZGG) with an RMSD of 0.875 Å. The ACE2 binding site is on the top of RBD domain. See also Figure S4.