Overview
The spike protein (S) is a class I fusion glycoprotein: a single-pass type I transmembrane protein that forms non-covalent homotrimers on the viral surface. Each monomer carries ~20–24 N-linked glycosylation sites, and the resulting glycan shield occludes a substantial proportion of the protein surface from antibody recognition, contributing to immune evasion.
Spike mediates the two critical steps of SARS-CoV-2 entry into host cells:
- Receptor binding: The S1 subunit’s receptor-binding domain (RBD) directly contacts angiotensin-converting enzyme 2 (ACE2) on the host cell surface.
- Membrane fusion: After cleavage and conformational change, the S2 subunit drives fusion of the viral envelope with the host cell membrane, releasing the viral genome into the cytoplasm.
Spike is the target of all approved COVID-19 vaccines worldwide, whether delivered as mRNA (mRNA-1273, BNT162b2), viral vector (AZD1222), or protein subunit (NVX-CoV2373). Its high mutation rate in circulating variants makes ongoing immunological monitoring essential.
Key structural insight: Cryo-EM studies (Wrapp et al., Science 2020; PDB 6VSB) revealed the pre-fusion trimer with RBD in the “up” conformation — the immunogenic state captured by 2P-stabilized vaccine antigens. This structure drove vaccine design in early 2020.
Domain Structure
| Domain | Residues | Function & Notes |
|---|---|---|
| Signal peptide | 1–12 | Co-translational cleavage; directs S to ER for glycosylation and processing |
| NTD (N-terminal domain) | 14–305 | Antibody epitope supersite for class I/IV and NTD-supersite (class V) antibodies; contains recurrent deletion regions (Δ69–70, Δ141–145, Δ211) in variants; exact receptor interaction disputed but NTD mutations affect cell tropism |
| RBD (receptor-binding domain) | 333–527 | Directly contacts ACE2; receptor-binding motif (RBM) = residues 438–506; dominant target for class I and class II neutralizing antibodies; hotspot for immune-evasion mutations in variants |
| SD1 / SD2 | 528–685 | Subdomain 1 and 2; structural scaffolding between RBD and S1/S2 junction; SD1/SD2 mutations (H655Y, N679K) influence furin cleavage efficiency |
| S1/S2 furin cleavage site | 680–689 (PRRAR↓SV) | Polybasic insert unique to SARS-CoV-2 (absent in SARS-CoV-1 and bat coronaviruses); cleaved by furin in trans-Golgi during protein maturation → primes S for ACE2-triggered activation; P681H/P681R mutations in variants enhance furin cleavage efficiency |
| S2 | 686–1273 | Fusion subunit; contains heptad repeats HR1 and HR2 that form 6-helix bundle during membrane fusion; highly conserved across betacoronaviruses; target for pan-coronavirus antibody strategies |
| Fusion peptide (FP) | ~816–833 | Inserts into host membrane bilayer after S2′ cleavage; initiates membrane merger; hydrophobic; blocked by site-0/site-III antibodies |
| Transmembrane domain (TM) | 1213–1237 | Anchors spike in viral envelope; required for viral assembly |
| Cytoplasmic tail | 1238–1273 | Multiple cysteine palmitoylation sites; membrane raft association; interacts with M protein for virion assembly; ER retention signal |
Vaccine antigen note: The 2P stabilization mutations used in mRNA-1273 and BNT162b2 are K986P + V987P, located in the central helix of the HR1 region. These proline substitutions introduce structural rigidity that locks the trimer in the immunogenic pre-fusion conformation, preventing the large conformational change to the post-fusion state during protein expression.
ACE2 Binding Mechanism
The RBD cycles dynamically between two conformations within the trimer: a “down” state (receptor-inaccessible, largely antibody-inaccessible) and an “up” state (ACE2-accessible; exposes neutralizing epitopes). At any given time in a trimer, one or two RBDs may be in the “up” state; this equilibrium is shifted by mutations in variants.
ACE2 engages the RBM (residues 438–506) via a 17-residue contact interface. Key intermolecular contacts in the original WH-01 strain include:
- Phe486 (S) → hydrophobic pocket around Met82 (ACE2)
- Gln493 (S) → Lys31 (ACE2) — mutated to Arg in Omicron (Q493R)
- Asn501 (S) → Tyr41 (ACE2) — N501Y in Alpha, Beta, Delta, Omicron → ~10-fold affinity increase
- Tyr505 (S) → Arg393 (ACE2)
Binding affinity for the original strain: Kd ~4–35 nM (depending on assay method). Omicron RBD binds ACE2 approximately 2.5-fold more tightly due to the cumulative effects of Q493R, N501Y, and other interface mutations, contributing to its transmissibility advantage despite reduced lung tropism.
ACE2 is expressed at highest levels in type II pneumocytes (SARS-CoV-2 primary respiratory target), small intestinal enterocytes, nasal goblet/ciliated cells, cardiac pericytes, vascular endothelium, and renal tubular cells. TMPRSS2 co-expression with ACE2 defines cells permissive for efficient cell-surface entry.
Membrane Fusion Mechanism
Entry proceeds through a sequence of irreversible conformational changes triggered by receptor binding and protease cleavage:
- Furin pre-cleavage (S1/S2): During viral biogenesis in the producer cell, furin in the trans-Golgi cleaves at the PRRAR↓SV site, non-covalently separating S1 and S2. This priming step is unique to SARS-CoV-2 among SARS-like betacoronaviruses and is a key determinant of efficient human-to-human transmission.
- RBD “up” and ACE2 engagement: RBD transitions to the up conformation; ACE2 binds the RBM, stabilising S1 in the receptor-bound state and triggering S1 shedding.
- TMPRSS2 cleavage (S2′) — cell-surface pathway: TMPRSS2 (type II transmembrane serine protease) at the plasma membrane cleaves the S2′ site (~residue 815), liberating the fusion peptide. This is the predominant pathway in lung cells and is not blocked by endosomal pH-raising agents.
- Cathepsin L/B cleavage — endosomal pathway: In cells with low TMPRSS2, virus is endocytosed; cathepsin B/L in late endosomes (pH ~5.0) substitute for TMPRSS2. This pathway is blocked by lysosomotropic agents (chloroquine, hydroxychloroquine, bafilomycin A1) in vitro, but clinical trials confirmed no benefit from these agents, consistent with TMPRSS2-dependent lung entry being dominant in vivo.
- 6-helix bundle formation: After FP insertion into the host membrane, HR1 and HR2 zipper together antiparallel into a thermostable 6-helix bundle — a topology conserved across all class I fusion proteins (HIV gp41, influenza HA2, RSV F). This irreversible collapse brings the viral and host membranes into apposition and drives membrane fusion.
Variants and Key Spike Mutations
| Variant (Lineage) | Key Spike Mutations | Phenotypic Effects |
|---|---|---|
| D614G (all post-June 2020) |
D614G | Increased furin cleavage efficiency; greater proportion of RBD “up” state; higher ACE2 binding; global sweep by July 2020; no major antibody evasion |
| Alpha (B.1.1.7) |
N501Y, Δ69–70, Δ144, P681H, D614G | N501Y: ~10-fold increased ACE2 affinity; Δ69–70: NTD supersite disruption; P681H: modestly increased furin cleavage; ~50% increased transmissibility vs. WT |
| Beta (B.1.351) |
K417N, E484K, N501Y, D614G | K417N + E484K: escape from class I/II and class III neutralizing antibodies; partial immune evasion; reduced vaccine efficacy vs. symptomatic disease (~75% → ~10% for some vaccines); N501Y: ACE2 affinity gain |
| Delta (B.1.617.2) |
L452R, T478K, P681R, D614G, T19R, Δ156–157 | P681R: enhanced furin cleavage efficiency, TMPRSS2-dependent entry, lung tropism; L452R: partial immune evasion; globally dominant mid-2021; most pathogenic pre-Omicron variant |
| Omicron BA.1 (B.1.1.529) |
32 spike mutations: K417N, E484A, Q493R, N501Y, G496S, Q498R, Y505H, H655Y, N679K, P681H, Δ69–70, Δ142–144, Δ211, ins214EPE | Highest immune evasion to date (3–40× reduction in vaccine-elicited neutralisation); reduced fusogenicity (reduced TMPRSS2 usage) → upper airway tropism → less severe lower respiratory disease; 2–3× increased ACE2 affinity; ACE2 quantity in nasal epithelium > lung → enhanced transmissibility |
| XBB.1.5, JN.1, KP.2 (recombinant/sub-lineages, 2023–2024) |
Progressive additional RBD mutations; XBB.1.5: F486P (high ACE2 affinity + antibody evasion); JN.1: L455S (strong antibody evasion); KP.2: R346T, F456L | Stepwise optimization of ACE2 affinity and antibody escape; each sub-lineage drives booster antigen updates; T-cell responses (especially S2-directed) maintain protection against severe disease |
Vaccine Target Strategies
| Platform | Examples | Antigen Design | Key Points |
|---|---|---|---|
| mRNA | mRNA-1273 (Moderna), BNT162b2 (Pfizer-BioNTech) | Full-length spike with K986P + V987P (2P) proline stabilization, locking pre-fusion conformation; N1-methylpseudouridine (m1Ψ) modification; LNP delivery (SM-102 or ALC-0315) | 94.1% (mRNA-1273) and ~95% (BNT162b2) efficacy vs. original strain; induce spike-specific IgG1/IgG3 (anti-RBD dominant) and CD8+/CD4+ T cells; updated annually for variant-matched boosters (BA.4/5 bivalent, XBB.1.5, JN.1 monovalent) |
| Viral vector | AZD1222 (Oxford-AstraZeneca, ChAdOx1) | Replication-deficient chimpanzee adenovirus expressing full-length spike (without 2P stabilization in original formulation) | ~70–79% efficacy (ChAdOx1 COV001/COV002); rare VITT (vaccine-induced immune thrombocytopaenia and thrombosis) via anti-PF4 antibodies; AdS5-nCoV (Sputnik V) similar platform |
| Protein subunit | NVX-CoV2373 (Novavax) | Recombinant 2P-stabilized trimeric spike nanoparticles (baculovirus system) + Matrix-M saponin adjuvant | ~90% efficacy (PREVENT-19); preferred by those avoiding novel platforms; strong Th1 response; favourable myocarditis profile vs. mRNA vaccines in young males |
| T cell epitope conservation | All platforms | S2 subunit contains more conserved CD4+/CD8+ T cell epitopes across variants than S1/RBD | T-cell cross-reactivity against conserved S2 epitopes maintains protection against severe disease and hospitalisation even when antibody neutralisation of new variants is substantially reduced; correlate of protection for severe disease outcomes |
Neutralizing Antibody Epitopes
Spike-targeting antibodies are classified into classes based on the epitope they contact and their mechanism of neutralization:
- Class I antibodies: VH3-53/VH3-66 germline usage; contact RBM in the “up” RBD conformation; directly compete with ACE2 binding; highly potent but sensitive to RBD mutations at key contact residues (K417, E484, N501); most pre-Omicron therapeutic mAbs (e.g., CB6/LY-CoV016)
- Class II antibodies: Contact both RBM and non-RBM surfaces of RBD; compete with ACE2; accessible in “up” and “down” states; include some of the broadest mAbs (e.g., S309 cross-reacts with SARS-CoV-1)
- Class III antibodies: Target RBD epitopes outside the RBM (“cryptic face”); do not directly compete with ACE2; often cross-reactive across betacoronaviruses; accessible in both conformations
- Class IV antibodies: Cryptic epitope accessible only when RBD is in “up” state and within a specific trimer context; generally lower potency
- NTD supersite antibodies (Class V): Target a glycan-free loop cluster on the NTD (approximately residues 14–20, 140–158, 245–264); highly sensitive to NTD deletions (Δ69–70, Δ141–145, Δ211) that recur convergently across variants of concern; include 4A8, S2L28
- S2 antibodies: Target HR1/HR2 or fusion peptide; cross-reactive across distantly related coronaviruses; generally lower potency but pan-coronavirus therapeutic interest (site 0, site III antibodies targeting the fusion peptide region)
Therapeutic mAb landscape: Most original therapeutic mAbs (casirivimab, imdevimab, bamlanivimab, etesevimab, sotrovimab) lost substantial activity against Omicron and sub-variants. Bebtelovimab retained activity against BA.2/BA.5 but was later compromised. As of 2024, commercially available mAbs have limited effectiveness against circulating XBB/JN/KP variants — T-cell immunity and mucosal IgA from vaccination and infection remain the primary ongoing defences.
Connections
- part ofSARS-CoV-2 — spike is the outermost structural protein of the virion; its evolution drives variant phenotype, transmissibility, and immune escape
- bindsACE2 Receptor — primary host entry receptor; expressed in type II pneumocytes, enterocytes, nasal epithelium, cardiac pericytes; ACE2 downregulation after binding may contribute to ARDS pathophysiology
- vaccine targetmRNA-1273 (Spikevax) — encodes 2P-stabilized full-length spike in SM-102 LNP; 94.1% efficacy in COVE trial
- vaccine targetBNT162b2 (Comirnaty) — encodes 2P-stabilized full-length spike in ALC-0315 LNP; ~95% efficacy in C4591001 trial
- vaccine targetAZD1222 — ChAdOx1 expressing full-length spike; ~70–79% efficacy; VITT risk informs allocation decisions
- vaccine targetNVX-CoV2373 — recombinant 2P spike nanoparticles + Matrix-M adjuvant; ~90% efficacy; preferred protein subunit option
- immune responseImmune System — spike-specific humoral (anti-RBD IgG) and cellular (Th1, CD8+ CTL) responses are both induced by vaccination and infection and are correlates of protection
References
- Wrapp D, Wang N, Corbett KS, et al. Cryo-EM structure of the 2019-nCoV spike in the prefusion conformation. Science. 2020;367(6483):1260–1263. doi:10.1126/science.abb2507
- Lan J, Ge J, Yu J, et al. Structure of the SARS-CoV-2 spike receptor-binding domain bound to the ACE2 receptor. Nature. 2020;581(7807):215–220. doi:10.1038/s41586-020-2180-5
- Harvey WT, Carabelli AM, Jackson B, et al. SARS-CoV-2 variants, spike mutations and immune escape. Nat Rev Microbiol. 2021;19(7):409–424. doi:10.1038/s41579-021-00573-0
- Starr TN, Greaney AJ, Hilton SK, et al. Deep mutational scanning of SARS-CoV-2 receptor binding domain reveals constraints on folding and ACE2 binding. Cell. 2020;182(5):1295–1310. doi:10.1016/j.cell.2020.08.012