key: cord-0867329-gt1nlh1t
authors: Zhang, Jun; Xiao, Tianshu; Cai, Yongfei; Chen, Bing
title: Structure of SARS-CoV-2 Spike Protein
date: 2021-09-08
journal: Curr Opin Virol
DOI: 10.1016/j.coviro.2021.08.010
sha: e99ede4628ac3a6db02b4a5a3287c6b3097c93df
doc_id: 867329
cord_uid: gt1nlh1t

The COVID-19 (coronavirus disease 2019) pandemic, caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has led to loss of human life in millions and devastating socio-economic consequences worldwide. The disease has created urgent needs for intervention strategies to control the crisis and meeting these needs requires a deep understanding of the structure-function relationships of viral proteins and relevant host factors. The trimeric spike (S) protein of the virus decorates the viral surface and is an important target for development of diagnostics, therapeutics and vaccines. Rapid progress in the structural biology of SARS-CoV-2 S protein has been made since the early stage of the pandemic, advancing our knowledge on the viral entry process considerably. In this review, we summarize our latest understanding of the structure of the SARS-CoV-2 S protein and discuss the implications for vaccines and therapeutics.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is the causative agent of the COVID-19 (coronavirus disease 2019) pandemic [1] , and its infection has led to millions of lives lost and devastating socio-economic consequences throughout the globe. There are urgent needs for innovative vaccine and therapeutic strategies to control this unprecedented crisis, as well as potential future needs if it becomes seasonal with continuous emergence of new variants. A deep understanding of the structure-function relationships of viral proteins and relevant host factors will be required in order to meet these needs. Coronaviruses (CoVs) are enveloped positivestranded RNA viruses that enter a host cell by fusion of its envelope lipid bilayer with the target cell membrane. This first critical step of viral infection is catalyzed by its trimeric spike (S) protein, which decorates the virion surface as a major antigen and induces neutralizing antibody responses. The protein is therefore an important target for development of diagnostics, therapeutics and vaccines. Remarkable progress in the structural biology of SARS-CoV-2 S protein has been made since the initial outbreak of the virus [2], substantially advancing our molecular understanding of the viral entry process. Here we summarize our current knowledge on the structure of the SARS-CoV-2 S protein and discuss the implications for vaccines and therapeutics.

The SARS-CoV-2 spike glycoprotein is a type I membrane protein (Fig. 1A) , which forms a trimer, anchored to the viral membrane by its transmembrane segment, while decorating the virion surface with it large ectodomain (Fig. 1B) . It binds to the receptor angiotensin-converting enzyme 2 (ACE2) on a host cell and undergo large structural rearrangements to promote membrane fusion [1, 3] . The protein is heavily glycosylated with each protomer containing 22 Nlinked glycosylation sites [4, 5] . The full-length S protein of the Wuhan-Hu-1 strain from the initial outbreak has 1273 amino acid residues, including a N-terminus signal peptide, a receptorbinding fragment S1 and a fusion fragment S2. S1 can be further divided into N-terminal domain (NTD), receptor-binding domain (RBD) and C-terminal domains (CTD1 and CTD2), while S2 includes fusion peptide (FP), fusion-peptide proximal region (FPPR), heptad repeat 1 (HR1), central helix (CH), connector domain (CD), heptad repeat 2 (HR2), transmembrane segment (TM) and the cytoplasmic tail (CT), depicted in Fig. 1A .

Structures of S protein fragments derived from the Wuhan-Hu-1 strain, including the S ectodomain stabilized in its prefusion conformation[6,7], RBD-ACE2 complexes [8] [9] [10] [11] , and segments of S2 in the postfusion state [12], were determined within the first several months of the pandemic. Soon after, structures of detergent-solubilized, full-length S proteins in both prefusion and postfusion conformations [13, 14] , as well as those of the intact S trimer on the virion surface, studied by cryo-electron tomography [15] [16] [17] [18] , were also reported ( Fig. 1B and   1C ). Overall, the SARS-CoV-2 S structure shows many similarities to those of other coronavirus spike proteins [19] [20] [21] [22] [23] . In the prefusion structure, the S1 fragment, adopting a "V" shaped architecture with the NTD at one arm and the RBD, CTD1 and CTD2 at the other (also see Fig.   2A ), which wrap around the central helical bundle formed by the prefusion S2 fragment, projecting the N-terminal end of HR1 toward the viral membrane. Three RBDs form the apex of the S trimer, sampling two distinct conformations -"up" representing a receptor-accessible state and "down" representing a receptor-inaccessible state (Fig.1B) . The three NTDs are located at the periphery of the trimer, each making contacts with the RBD from the adjacent protomer. The CTD1 and CTD2 pack underneath the RBD against S2 and between the two neighboring NTDs, indicating they could modulate these domains and play important roles in the structural rearrangements required for membrane fusion.

In the postfusion conformation, S1 dissociates as a monomer, while S2 adopts a rigid, baseballbat-like shape (~220 Å long), and the HR1 flips over to form a continuous long helix together with the CH, which is further surrounded by short helices and β-sheets at the distal end of the membrane (Fig 1C and 1E ). The connector domain (CD), together with a segment (residues 718 to 729) in the S1/S2-S2' fragment, form a three-stranded β sheet, and residues 1127 to 1135 join the connector β sheet to expand it into four strands. Another segment (residues 737 to 769) in the S1/S2-S2' fragment makes up three helical regions locked by two disulfide bonds that pack against the groove of the CH part of the coiled coil to form a short, six-helix bundle structure (6HB-1). The N-terminal region of HR2 adopts a one-turn helical conformation and also packs against the groove of the HR1 coiled coil; the C-terminal region of HR2 forms a longer helix that makes up the second six-helix bundle structure with the rest of the HR1 coiled coil (6HB-2)

[13, 24, 25] .

At the periphery of the spike (Fig. 1B) [13], the NTD projects away from the 3-fold axis, and can be divided into the top, core and bottom regions (Fig. 2B ). The core structure has a galectin-like antiparallel β-sandwich fold, formed by one six-stranded β-sheet and the other with seven strands. The top region has two antiparallel β strands connected by a short loop, while the bottom region is primarily made up of two short β sheets and a helix. The overall structure of the NTD is decorated by eight N-linked glycans and similar to that of the S proteins from Middle East respiratory syndrome coronavirus (MERS-CoV) [26] and bovine coronavirus [27] .

The exact function of the NTD in SARS-CoV-2 S remains unknown, although NTDs of other coronaviruses have been shown to recognize sugars upon initial attachment or specific protein receptors, or play a role in the prefusion-to-postfusion transition [27] . Nonetheless, NTD- [33] ). Most antibodies target the NTD-1 region, which is thus named the NTD-1-antigenic supersite. It is located at the edge of the NTD top-core region, including five surface loops: N1 (residues 14-26), N2 (residues 67-79), N3 (residues 141-156), N4 (residues 177-186), N5 (residues 246-260) (Fig. 2C) , and a β-hairpin structure near N3, surrounded by four N-linked glycans (Asn17, Asn74, Asn122 and Asn149). These loops reconfigure upon binding to various antibodies (Fig. 2C ).

In the S-4A8 complex structure (Fig. 2D) [28], the third complementarity determining region (CDR3) of the 4A8 heavy chain inserts to a cleft formed by the N3 β-hairpin/loop and N5 loop, while the CDR1 and CDR2 interact with the tips of the two loops. Moreover, the glycan at Asn149 is very close to the interface and may also contribute to antibody binding (Fig. 2E ).

Other antibodies, such as S2M28, 4-18, DH1050, CM25, FC05, 12C9 [33] , also use their CDR1-3 to contact the N3 and N5 loops, but some interact with the nearby N1 loop or the glycan at Asn17 as well. Despite the differences in approaching angles among these antibodies, their interface with the NTD-1 is highly conserved. Up till now, NTD-2 is recognized by nonneutralizing antibodies, such as by DH1052 and 81D6 [33] . and/or deletions within the NTD-1-supersite, rendering resistance to neutralization by NTDdirected antibodies [34, 35] .

The RBD contains two subdomains -a five-stranded antiparallel β sheet connected by short helices and loops, and an extended loop, named receptor binding motif (RBM) [8, 9, 36] . In the host cell, ACE2 is an important component of renin-angiotensin system (RAS) and catalyzed the hydrolysis of angiotensin II to angiotensin 1-7 [9]. The full-length human ACE2 is also a chaperone of the amino acid transporter B 0 AT1 and forms a homodimer mediated by its neck domain in the presence of B 0 AT1 (Fig. 3A) [10]. Cryo-EM structures of the soluble uncleaved S protein in complex with monomeric ACE2 show that the S trimer can bind one, two or three J o u r n a l P r e -p r o o f ACE2s in the RBD-up conformations (Fig. 3B) [37, 38] . The crystal structure of the SARS-CoV-2 RBD-ACE2 complex reveals a similar structure to the SARS-CoV RBD-ACE2 complex [8, 9] . of concern, leading to enhanced affinity for ACE2 and immune evasion [39, 40] .

The RBD is a dominant target of nAbs elicited by either natural infection or vaccination, confirming its pivotal role during infection [41, 42] . The RBD-directed nAbs can recognize multiple distinct epitopes, showing great potencies at the pM-nM level in vitro neutralization assays (Fig. 3D ) [42] . The nAbs that target the ACE2-binding-site, such as REGN10933, C144 and S2H14, directly compete for ACE2 association [41] [42] [43] [44] . Those recognizing the non-ACE2binding-site, such as REGN10987 and C135, probably prevent ACE2 binding either by clashing with ACE2 or by blocking the transition of the RBD from the "down" to the "up" conformation [42] [43] [44] . Other nAbs against the so-called "cryptic supersite", such as CR3022 and S304, can destabilize the S trimer and induce S1 dissociation [41, 42, 45] . Although the great potency of this class of antibodies makes them promising therapeutic agents, emergence of resistant variants could limit their clinical applications for treating the COVID-19.

A recombinant human ACE2, named APN01, is under evaluation as a treatment for COVID-19 in a phase 2 clinical trial, based on the favorable results from a previous phase 1 trial [46] , and evidence that the protein blocks SARS-CoV-2 infection effectively in vitro [47] . Other ACE2based fusion inhibitors have been developed with optimized binding and potency comparable to those of the nAbs [37, 48, 49] . The ACE2 constructs with multivalency, such as the dimeric protein sACE2 2 .v2.4-IgG1 carrying the mutation T27Y/L79T/N330Y and the trimeric protein ACE2-foldon T27W, can inhibit the viral infection with a potency 1000-and 1700-fold greater than that of the monomeric soluble ACE2 with the wildtype sequence [37, 48] . Substitution of J o u r n a l P r e -p r o o f T27 with an aromatic residue appears to further stabilize the binding interface through non-polar interactions with residues Y489, F456 and Y473 of the RBD (Fig. 3C) . In addition, a series of miniproteins, created using computer-generated scaffolds to mimic the N-terminal helix of ACE2, can bind the RBD and inhibit viral infection at a concentration below the nM level [49] .

These ACE2-derived inhibitors may show even greater potency to those SARS-CoV-2 variants that have gained increased receptor binding than the Wuhan-Hu-1 virus. Nonetheless, pharmacokinetics, in vivo efficacy and safety profile of these new designs still require further validation.

The C-terminal domains (CTDs) are formed primarily by β-structures of segments from S1, as well as the N-terminal segment of S2 adjacent to the furin cleavage site (Fig. 4) . CTD1 contains two antiparallel β-sheets, with two strands and four strands, respectively. CTD2 also has two βsheets: a four-stranded one and another four-stranded one that includes a strand from the S2 

In the prefusion conformation [13], three S2 subunits tightly pack around a central three-stranded coiled-coil of ~140Å long, formed by CH (Fig. 1B) . Portion of the HR1 together with another segment of S2 (residues 758-784) adopt -helical conformation and assemble into a nine helixbundle with the central coiled-coil, forming the most rigid part of the entire S trimer. The CD region links CH and the C-terminal HR2 through a linker region (Fig. 5A) . The FP forms a short helix and tucks in a pocket formed by two neighboring S protomers. The structured FPPR J o u r n a l P r e -p r o o f clashes with the CTD1 if the RBD moves up and thus appears to help clamp the prefusion S trimer in the closed, RBD-down conformation. It has also been suggested to function as a pHdependent switch domain that modulates the RBD position [38] . The remaining HR2, TM and CT segments are disordered in the most S trimer structures, but show low-resolution density in the cryo-ET reconstructions that can be tilted away from the three-fold axis of the trimer with an angle from 17 • to 60 • [16].

In the postfusion conformation [13, 24] , the HR1 and CH form a continuous α-helix and three copies of them assemble into a long central three-stranded coiled-coil of ~180Å (Fig. 5A) . Two proline substitutions at the boundary between the HR1 and CH to prevent formation of the postfusion helix have been introduced to stabilize the prefusion conformation and such a design has been used for structural studies and the first-generation vaccines [6, 7, 51] . Part of the HR2 folds into α-helix and packs against the groove between two HR1-CH helices to form a six-helix bundle structure, reminiscent of the postfusion organization of other viral fusion proteins [52, 53] .

The CD remains unchanged from the prefusion conformation, as a three-stranded β-sheet covering the C-terminal end of HR1-CH helices. Comparison of the prefusion and postfusion conformations of S suggests that HR1 undergoes large rearrangements to form a coiled-coil, translocating its N-terminal end by a large distance to project the FP towards the target cell membrane (Fig. 5B ). In addition, the HR2 and the TM at its C-terminal end must fold back to pack along the groove of the HR1-CH coiled-coil to form the postfusion six-helical bundle (Fig.   5C ). These refolding events effectively bring the viral and target cell membranes close together, ultimately leading to membrane fusion (Fig. 5D ). Interestingly, five N-linked glycans decorate the postfusion S2 surface along the long axis with a regular spacing and may protect the S2 from the host immune responses.

The SARS-CoV-2 S protein is the key component of almost all the first-generation COVID-19 vaccines [51] . Based on structural studies, concerns have been raised that the inactivated-virus vaccines or those used the wildtype sequence of the Wuhan-Hu-1 strain may have too many postfusion spikes and induce mainly non-neutralizing antibodies [13, 18] . Indeed, these vaccines have induced lower levels of neutralizing antibody responses than other S constructs containing J o u r n a l P r e -p r o o f stabilization modifications to prevent conformational changes [54] . Additional studies have identified the G614 S trimer as a possible superior immunogen candidate [50, 55] , as it is naturally constrained in a prefusion state presenting both the RBD-down and RBD-up conformations with great stability. Moreover, the global spread of SARS-CoV-2 and the consequently vast number of replication events make emergence of new variants inevitable, and substantially increases the genetic diversity of the virus, which will bring much greater challenges for vaccine development than it was at the beginning of the pandemic. Indeed, genetic diversity is also the major hurdle for development or optimization of vaccines against several other human pathogens, such HIV-1, hepatitis C virus and influenza virus [56] [57] [58] . If SARS-CoV-2 becomes seasonal, structure-based innovative strategies will likely be needed for developing next-generation vaccines designed to elicit broadly neutralizing antibody responses.

Likewise, high-resolution structural information has been instrumental for creating peptide-and ACE2-based fusion inhibitors [12, 37, 49, 59] , it will undoubtedly continue to serve as a foundation for rational design of antiviral therapeutics to fight against the pandemic.

Tremendous progress in the structural biology of SARS-CoV-2 spike protein has been made since the initial outbreak of the virus. The structural knowledge not only fills the major gap in our understanding of the viral entry process, but also provides a solid foundation for development and optimization of vaccines and therapeutics against the current and future pandemics of coronaviruses. Comparison with the SARS-CoV RBD has revealed that the ACE2-binding site of the SARS-CoV-2 RBD has a more compact conformation and that several residue changes in the SARS-CoV-2 RBD stabilize two virus-binding hotspots at the RBD-ACE2 interface. Moreover, the bat coronavirus RaTG13 also uses human ACE2 as its receptor, providing evidence for the potential animal-to-human transmission of SARS-CoV-2. In this study, cryo-electron tomography, subtomogram averaging and molecular dynamics simulations were used to structurally analyze SARS-CoV-2 spike (S) protein in situ. It shows that the stalk region of S contains three hinges, giving the head unexpected orientational freedom.

An inactivated SARS-CoV-2 preparation of the original strain has been genetically and structurally characterized. It shows that the virus particles are roughly spherical or moderately pleiomorphic and that most spikes appear nail shaped

Cryo-electron microscopy structure of a coronavirus spike glycoprotein trimer

Pre-fusion structure of a human coronavirus spike protein

Cryo-electron microscopy structures of the SARS-CoV spike glycoprote in reveal a prerequisite conformational state for receptor binding

Cryo-EM structures of MERS-CoV and SARS-CoV spike glycoproteins reveal the dynamic receptor binding domains

The human coronavirus HCoV-229E S-protein structure and receptor binding

Cryo-EM analysis of the post-fusion structure of the SARS-CoV spike glycoprotein

Tectonic conformational changes of a coronavirus spike glycoprotein promote membrane fusion

Structural definition of a neutralization epitope on the N-terminal domain of MERS-CoV spike glycoprotein

Crystal structure of bovine coronavirus spike protein lectin domain

This paper reported isolation and characterization of monoclonal antibodies (mAbs) from 10 convalescent COVID-19 patients. The antibody 4A8 is a potent neutralizing antibody that targets the N-terminal domain (NTD) of the S protein

Potent SARS-CoV-2 neutralizing antibodies directed against spike Nterminal domain target a single supersite. Cell Host Microbe 2021. In this paper, structures for seven potent NTD-directed neutralizing antibodies in complex with SARS-CoV-2 spike or isolated NTD have been determined by either cryo-EM or x-ray crystallography. The structures revealed that all seven antibodies target a common surface

Prevalent, protective, and convergent IgG recognition of SARS-CoV-2 non-RBD spike epitopes

This work described a large number of human monoclonal Abs (mAbs) derived from memory B cells, targeting the SARS-CoV-2 S N-terminal domain (NTD) with a subset neutralizing the virus ultrapotently. A supersite recognized by all known NTD-specific neutralizing mAbs has been defined, and ehese mAbs inhibit cell-to-cell fusion

The functions of SARS-CoV-2 neutralizing and infectionenhancing antibodies in vitro and in mice and nonhuman primates

Memory B cell repertoire for recognition of evolving SARS-CoV-2 spike

Recurrent deletions in the SARS-CoV-2 spike glycoprotein drive antibody escape

Antibody resistance of SARS-CoV-2 variants B.1.351 and B.1.1.7

Structure of SARS coronavirus spike receptorbinding domain complexed with receptor

A trimeric human angiotensin-converting enzyme 2 as an anti-SARS-CoV-2 agent

Cryo-EM Structures of SARS-CoV-2 Spike without and with ACE2 Reveal a pH-Dependent Switch to Mediate Endosomal Positioning of Receptor-Binding Domains

Effect of natural mutations of SARS-CoV-2 on spike structure, conformation and antigenicity

Structural basis for enhanced infectivity and immune evasion of SARS-CoV-2 variants

Mapping Neutralizing and Immunodominant Sites on the SARS-CoV-2 Spike Receptor-Binding Domain by Structure-Guided High-Resolution Serology

Structural Analysis of Neutralizing Epitopes of the SARS-CoV-2 Spike to Guide Therapy and Vaccine Design Strategies

A series of neutralizing antibodies generated from humanized mice and convalescent patients against the SARS-CoV-2 Spike protein were characterized in binding, neutralization and structure to identify pairs of highly potent antibodies ideal for cocktail antibody therapy

SARS-CoV-2 neutralizing antibody structures inform therapeutic strategies

Neutralization of SARS-CoV-2 by Destruction of the Prefusion Spike

Pharmacokinetics and pharmacodynamics of recombinant human angiotensin-converting enzyme 2 in healthy human subjects

Inhibition of SARS-CoV-2 Infections in Engineered Human Tissues Using Clinical-Grade Soluble Human ACE2

Engineering human ACE2 to optimize binding to the spike protein of SARS coronavirus 2

De novo design of picomolar SARS-CoV-2 miniprotein inhibitors

Cryo-EM structures of a full-length G614 S trimer have been determined and the trimer adopts three distinct prefusion conformations that differ primarily by the position of one receptor-binding domain. A loop disordered in the D614 S trimer wedges between domains within a protomer in the G614 spike. This added interaction appears to prevent premature dissociation of the G614 trimer effectively increasing the number of functional spikes and enhancing infectivity and to modulate structural rearrangements for membrane fusion

Viral targets for vaccines against COVID-19

Core structure of gp41 from the HIV envelope glycoprotein

Atomic structure of the ectodomain from HIV-1 gp41

Krammer F: SARS-CoV-2 vaccines in development

SARS-CoV-2 spike-protein D614G mutation increases virion spike density and infectivity

Advancing an HIV vaccine; advancing vaccinology

Hepatitis C Virus Vaccine: Challenges and Prospects. Vaccines (Basel) 2020

Next-generation influenza vaccines: opportunities and challenges

Fusion mechanism of 2019-nCoV and fusion inhibitors targeting HR1 domain in spike protein

Distinct conformational states of the SARS-CoV-2 spike protein. (A) Schematic representation of the SARS-CoV-2 spike protein organization. Segments of S1 and S2 include: NTD, N-terminal domain

RBD, receptor-binding domain

S1/S2 cleavage site; S2', S2' cleavage site; FP, fusion peptide

FPPR, fusion peptide proximal region

CH, central helix region; CD, connector domain; HR2, heptad repeat 2

Right: cryo-EM structure of the full-length S trimer in the RBDdown conformation (PDB ID: 6XR8). (C) Left: viral SARS-CoV-2 S2 trimer in the postfusion conformation (EMD-30428; ref [15]), fitted with the structure of the purified protein (PDB ID: 6XRA; ref [13]). Right: cryo-EM structure of the full-length S2 trimer in the postfusion conformation (PDB ID: 6XRA). (D) Additional structures of coronavirus S proteins, including the full-length SARS-CoV-2 S trimer carrying G614 in the one RBD-up conformation

PDB ID: 7C2L; ref [28]) and DH1052 Fab (PDB ID: 7LAB; ref [32]), as indicated. Heavy and light chains of 4A8 are colored in red and pink, respectively, and those of DH1052 are in green and cyan, respectively. (D) The NTD (in blue) from its complex with 4A8 is superposed with the domain from the full-length S trimer in gray

Close-up view of the binding interface for the NTD-4A8 and NTD-DH1205 complexes with

We thank all the former and current members of the Chen laboratory at Boston Children's Hospital for their contributions to our research. Our work was supported by