Fig 1.
Structural and sequence-based comparison of S1 domains between SARSr-CoV spike proteins.
(a) Structural alignment of the spike, cCoV-SZ3 (green), cCoV-007 (blue) and bCoV-WIV1 (purple), other monomers displayed in gray as a surface representation. Black arrow indicates receptor binding motif (RBM), purple and white arrows indicate important trimer interfaces. (b) Protein alignment of RBM loop of interest between RaTG13 (pink) (PDB: 6ZGF) and bCoV-WIV1. RMSD calculated to 3.4 Å. Bond lengths measured in ChimeraX and hydrogen bonds indicated by dashed blue line. (c) Protein alignment of the glycoproteins against RaTG13 and SARS-CoV-2 WT (tan) (PDB: 7QUS) at RBM and (d) fusion peptide (FP). (e-f). Key residues at the trimer interface involved in structure guided stabilization of SARS-CoV-2 in closed conformation compared to the glycoproteins. (g) Protein alignment of the glycoproteins against RaTG13 and SARS-CoV-2 WT with D614 side chain displayed. (h) Protein alignment of four N-terminal domain (NTD) loops in the glycoproteins against RaTG13 and SARS-CoV-2 WT. (i) Geneious global amino acid sequence alignment of RBM from 6 glycoproteins, bCoV-WIV1, cCoV-SZ3, cCoV-007, human (h) SARS-CoV-2 WT, bCoV RaTG13, pangolin (p) CoV GX (PDB: 7CN8) and pCoV GD (PDB:7BBH). Scoop loop outlined in red.
Fig 2.
Biliverdin-binding pocket of SARSr-CoV glycoproteins.
The hydrophobic biliverdin binding pocket is formed by the NTD beta sandwich. The corresponding density is rendered as a mesh within the pocket. Residues within 5 Å of biliverdin are displayed. (a) The 007 structure is overlayed with a 1.8 Å reference structure (7B62, tan) known to be bound to biliverdin (lime, molecule from 7B62). Black arrows indicate solvent-exposed carboxyl groups which lack density in the electron maps of all three spikes. (b) The biliverdin density in WIV1 is continuous with the density for some of the pocket residues, namely L110 and N105. (c) The SZ3 map possesses the most isotropic biliverdin density. All SARSr-CoV spikes share conserved Y187 and N105 residues.
Fig 3.
(a) Relative location of fatty acid (9-octadecenoic acid, magenta) within the FABP of bCoV-WIV1 spike protein. (b) Overlay of cCoV-007 (green), cCoV-SZ3 (cyan) and bCoV-WIV1 acids (purple). Pocket residues within 4 Å are displayed in stick representation. (c) Water molecule coordination between the linoleic acid and chain A of the cCoV-SZ3 spike is represented by a yellow dotted line. Donor-acceptor hydrogen bond cut-offs were set to 3.4 Å. Waters are displayed as red spheres and bond lengths are reported in angstroms. (d) Water molecule densities are shown, within the FABP of cCoV-SZ3.
Fig 4.
Extensive N-glycosylation of SARSr-CoV glycoproteins.
(a) Monomer of cCoV-SZ3, cCoV-007 and bCoV-WIV1 are shown with glycans displayed in cyan. (b) The distribution of glycosites varies between the three SARSr-CoV spikes across the main architectural domains, except within the RBD. (c) Trimeric cCoV-007 spike is represented with molecular surface display and coloured by chain. All 45 N-glycans are displayed as cyan spheres regardless of carbohydrate composition. (d) Chain A is displayed nested within the Coulomb density map. (e) Carbohydrate tree compositions (based on density-directed modelling) are reported for cCoV-007. (f) Each glycosylation site is mapped to the domain organization of the cCoV-007 glycoprotein. Although there are two N-linked glycosylations within the receptor binding domain, they are both distant from the receptor binding motif, with the closest glycan approximately 9 Å from the nearest RBM residue (Ile-415). Schematic not to scale. (g) A representative carbohydrate from each type of complex N-glycosylation shown with the Coulomb density map displayed as a mesh. The N-linked asparagine and other relevant neighbouring residues are labelled.