The Spike of Concern-The Novel Variants of SARS-CoV-2

Abstract

The high sequence identity of the first SARS-CoV-2 samples collected in December 2019 at Wuhan did not foretell the emergence of novel variants in the United Kingdom, North and South America, India, or South Africa that drive the current waves of the pandemic. The viral spike receptor possesses two surface areas of high mutagenic plasticity: the supersite in its N-terminal domain (NTD) that is recognised by all anti-NTD antibodies and its receptor binding domain (RBD) where 17 residues make contact with the human Ace2 protein (angiotensin I converting enzyme 2) and many neutralising antibodies bind. While NTD mutations appear at first glance very diverse, they converge on the structure of the supersite. The mutations within the RBD, on the other hand, hone in on only a small number of key sites (K417, L452, E484, N501) that are allosteric control points enabling spike to escape neutralising antibodies while maintaining or even gaining Ace2-binding activity. The D614G mutation is the hallmark of all variants, as it promotes viral spread by increasing the number of open spike protomers in the homo-trimeric receptor complex. This review discusses the recent spike mutations as well as their evolution.

Keywords: B.1.1.28; B.1.1.298; B.1.1.7; B.1.351; B.1.427; B.1.429; B.1.617; B1.526; Marseille-4; P.1; P.2.

Figure 1

Domain Organization of the SARS-CoV-2 spike protein and structural features. (a) Domain organization of spike. Signal peptide: 1–12aa; S1-domain: 13–685aa; S2-domain: 686–1273aa; N-terminal domain (NTD): 13–305aa; Supersite loops: N1:14–20aa; N3: 140–158aa; N5: 245–264aa; Receptor binding domain (RBD): 319–541aa; Receptor binding motif (ACE2): 437–508aa; Furin cleavage sequence: 680–685aa; Fusion peptide (FP): 788–806aa; Heptad repeat region 1 (HR1): 912–984aa; Heptad repeat region 2 (HR2): 1163–1213aa; Transmembrane domain (TMD): 1214–1237aa; Cytoplasmic domain: 1238–1273aa; (b) Surface of the closed spike trimer (PDB: 6VXX); (c) Surface of the open spike trimer—one protomer open (green) (PDB: 6VYB); (d) Open protomer from (c) with the Ace2 binding motif (magenta) and the RBD both indicated; (e) Surface of the open protomer with the three loops of the N-terminal supersite (N1: yellow; N3: blue; N5: red) and the Ace2 binding site (magenta) highlighted (PDB: 7DF4); (f) Cartoon rendering of the lower S2 section. The Furin loop is not fully visible due to its high mobility (magenta). The S2 cleavage site is shown in red and the fusion peptide (FP) in blue (PDB: 7DF4). Domain annotation according to [19]. Loops N1, N3, and N5 annotation as in [10]. Visualisation with Polyview-3D [20].

Figure 2

Timeline of the SARS-CoV-2 pandemic and the emergence of the variants. See text for further details.

Figure 3

Structural impact of the D614G mutation. (a) Cartoon rendering of the open spike protomer (only top half is shown) with an aspartate at position 614 (D614, orange spheres); the disordered 620–640 loop (red circle) is absent from the structure; NTD (N-terminal S1 domain); RBD (receptor binding domain in S1) (PDB: 7KRR); (b) Cartoon rendering of the open spike protomer with a glycine at position 614 (G614, orange spheres) (PDB: 7DK3); the ordered 620–640 loop (red) is now visible; (c) Surface rendering of spike D614; (d) Surface rendering of spike G614. Visualisation with PyMOL (Molecular Graphics System, Version 2.0 Schrödinger, LLC).

Figure 4

The receptor binding motif of spike (RBD). (a) View into the binding site. The key amino acids which are mutated in the variants (Table 1) are highlighted. The binding partners in hAce2 are shown in the same colour. Please note that L452 (red) is next to Y453 (light blue) but facing away from the binding site. E484 (magenta) is positioned at the rim of the binding site. (b) The backside of the binding site. The polar link between E484 with K31 in hAce2 and L452 are now visible. The other 14 residues making contact with hAce2 (G446, Y449, L455, F456, A475, F486, N487, Y489, Q493, G496, Q498, T500, G502, Y505) are not shown. hAce2 (blue): amino acids 19–99 and 312–395 are shown. Spike (orange): amino acids 393–516 are shown. The image was rendered with Ezmol 2.1 [63] using the structure PDB 7DF4.

Figure 5

Surface structure of the hAce2 binding site of Spike and its variants. (a) The Wuhan strain. The 17 amino acids making direct contact with hAce2 are highlighted in orange, K417 in yellow, N501 in green and E484 in magenta [9]; (b) D614G (PDB: 7BNN); (c) B.1.1.7 (UK) (PDB: 7LWT); (d) P.1/B.1.1.28 (Brazil) (PDB: 7LWW); (e) B.1351 V1 (South Africa) (PDB: 7LYQ); (f) B.1.1.298 (Denmark) (PDB: 7LWO); only the Wuhan structure was modelled with Phyre2 (Protein Homology/analogy Recognition Engine V 2.0) and visualised with EzMol2.1 [63,64]; the other structures are visualisations of the published structural data sets with the complete coverage of the RBD (i.e., no unstructured regions). Visualisation with PyMOL (Molecular Graphics System, Version 2.0 Schrödinger, LLC). Wuhan spike protein sequence: UniProt P0DTC2.

Figure 6

Models of the protein surfaces of wild-type Spike and its variants with the three loops of the super site, the hAce2 binding site, and the furin loop indicated. (a) The Wuhan wild-type spike (loops N1 (14–20aa): yellow; N3 (140–158aa): blue; N5 (245–264aa): red; hAce2 binding site (magenta); furin loop (green arrow)); (b) B.1.1.7 (UK); (c) B.1.429 (USA); (d) B.1.526 (USA); (e–g) sub-variants V1, V2, and V3 of B.1.351 (South Africa); (h) B.1.617 (India) (please note the model contains only the mutations listed in Table 1); (i) B.1.1.298 (Denmark); (j) P.1 (Brazil); (k) P.2 (Brazil). Since three-D structure are not yet published for all mutated Spike proteins, models were generated with the mutated amino acid sequences of Spike using the Phyre2 (Protein Homology/analogy Recognition Engine V 2.0) server. The resulting structure files were then rendered with PyMOL (Molecular Graphics System, Version 2.0 Schrödinger, LLC). Arrows indicate the predicted structural changes in loop N3 (blue) or loop N5 (red). Please note that the furin loop is modelled differently in both USA variants compared to all other variants. The reason for this is as yet unknown.

Figure 7

The structural plasticity of the hAce2 binding site in Spike. (a) The surface of the hAce2 binding site of the Wuhan strain is shown in all subpanels. The 17 amino acids making direct contact with hAce2 are highlighted in orange; (b) the positions of the mutations K417T/N (red), E484K (magenta), and N501Y (green) present in the Brazilian P.1 and the South African B.1.351 Spike proteins are highlighted; (c) the epitope of the neutralising antibody M369 is shown in dark purple; (d) the epitope of the neutralising antibody 80R is shown in black [9]. Please note that N501 resides within both epitopes. Hence, its mutation enhances hAce2 affinity while simultaneously impairing antibody binding. Images were rendered with PyMOL (Molecular Graphics System, Version 2.0 Schrödinger, LLC); the structure file was generated with Phyre2 (Protein Homology/analogy Recognition Engine V 2.0) using the Wuhan Spike protein sequence: UniProt P0DTC2.