Coronaviruses (CoVs) belong to a large family of viruses that display a "crown" (corona) of surface Spike (S) proteins on their membranes. It is these S proteins that bind to host cell receptor proteins to initiate an infection cycle. The coronaviruses are enveloped, positive-sense, single stranded RNA viruses (see Figure 1).

Most coronaviruses circulate among animal "reservoirs" (e.g. pigs, camels, bats, cats), but some can move into human hosts, causing zoonotic diseases that present as a mild to moderate upper-respiratory tract illness. Despite the relatively benign human symptoms caused by most coronaviruses, a few novel coronaviruses have emerged from animal reservoirs over the past two decades that seriously impact infected individuals.

In 2002, a coronavirus, SARS-CoV, emerged and caused Severe Acute Respiratory Syndrome (SARS). In 2012, the Middle East Respiratory Syndrome coronavirus (MERS-CoV) jumped from camels to humans and elicits severe respiratory pathology. The most recent coronavirus to zoonotically infect humans is SARS CoV-2 which has generated a global pandemic (2019). This SARS-CoV-2 virus elicits Coronovirus Disease (COVID-19) that has caused serious and widespread illness and death.

Enveloped viruses contain a lipid bilayer envelope that must achieve the energetically unfavorable fusion of the viral and host cell membranes in order to deliver the genetic material of the virion into the cytoplasm of the host during the initial stages of infection. The S Proteins of the coronaviruses are essential to this membrane fusion and the molecular "gymnastics" that these proteins conduct in order to elicit fusion are astounding!

To the left is a model of the SARS-CoV-2 S protein. This protein is a glycosylated class I viral fusion protein that binds to the ACE2 receptor protein (angiotensin converting enzyme-2), and in a series of molecular events, initiates viral entry into host cells by fusion with a host cell membrane. Fusion can occur between the viral membrane and the host plasma membrane if particular proteases are present at the cell surface, or between the viral and endososomal membrane if the virus is endocytosed after ACE-2 binding (e.g. Verma and Subbarao, 2021). In either case, the virus has gained entry and can begin its replication cycle. The dual entry mechanisms allow the virus to attack a range of host tissues.

Cleavage of the S protein into two fragments (S1 and S2) by a furin-like proteolytic enzyme exposes an S protein region that binds to a second receptor, Neuropilin-1 (NRP1). Further cleavage by a serine protease triggers dissociation of S1 and elicits extensive conformational changes in S2, bringing the viral and host membrane into close contact for membrane fusion and entry of the virus into the cytoplasm (Kielan, 2020) - see illustrations below. For a summary of the initial infection events, see Figure 2.

This exhibit provides visualizations of some of the CoV-2 S protein structure/function relationships that are important to viral infection and also illustrates important interactions with the ACE-2 receptor protein. Example mutations found in variants of CoV-2 that affect transmissabiity and virulence are also presented.

Return to Beginning

As can been seen in its pre-fusion state at left, the CoV-2 S protein is a homotrimer consisting of three protomers: protomer I, protomer II, and protomer III. Each protomer has multiple domains and extensive secondary structure. The S protein is a glycoprotein and numerous glycans are seen in the structure (PDB ID: 6XR8).

The S protein shown is oriented with the viral membrane down and the potential host membrane up. Each protomer contains a receptor binding domain (RBD). If one visualizes the protein as a tree, the RBDs lie in the canopy, ready to bind a host cell ACE-2 receptor, enhanced by a conformational change that causes an RBD to project upward (see below).

The prefusion trimer is cleaved in the early stages of host cell infection by a furin-like protease into the receptor binding fragment, S1, and the fusion fragment, S2, that mediates fusion of the virus' membrane and that of the host cell plasma membrane or endosomal membrane (see Figure 2). S1 contains the RBDs.

Some of the structural components important in launching an infection of a host cell can be visualized by examining a single protomer, followed by highlighting these in the prefusion trimer:

• S1 N Terminal Domain (NTD)
• S1 Receptor Binding Domain (RBD)
• S1 C Terminal Domain 1 (CTD1)
• S1 C Terminal Domain 2 (CTD2)
• S2 Fusion Peptide (FP) and S2 Fusion Peptide Proximal Region (FPPR)
• S2 Heptad Repeat 1 (HR1)
• S2 Central Helix (CH)
• S2 Connector Domain (CD)
• S2 Linker (L)
• Not resolved in the 6XR8 structure: S2 Heptad Repeat 2 (HR2), S2 Transmembrane Domain (TM), and S2 C Terminal (CT). These would project downward from the S2 Linker (L) in the view shown at left, with the TM domain anchoring the Spike protein in the viral membrane.

The S1 domains (NTD, RBD, CTD1, CTD2,) are seen to envelop the S2 fragment beneath.

In the trimer, the RBDs are stabilized in a tight cluster atop the Spike, in the "closed" conformation, which may hide them from host immune recognition. A conformational change elicited by proteolytic cleavage causes an upward projection of an RBD to engage in host cell receptor binding (see below). The Fusion Peptide Proximal Region (FPPR) plays an essential role in ths clamping down of the RBDs. As can be seen, a FPPR of one protomer lies on the opposite side of a CTD1 of another protomer, which in turn is adjacent to an RBD from the same protomer.

The FPPR is stabilized both by a disulfide bridge and by a salt bridge between the side chains of a lysine and an aspartate.

The CTD1 may serve as a relay between an RBD and an FPPR. Since the FPPR is directly connected to the Fusion Peptide (FP) that will be inserted into the host cell membrane after S Protein processing and profound conformational rearrangement (see section V, below), changes to the FPPR structure likely play a functional role in viral membrane-host cell membrane fusion (Cai, et al., 2020).

Further evidence of the important role of the FPPR in clamping down the RBDs is provided by the observation that mutation of an aspartate (residue 614) to a glycine (D614G) in certain CoV-2 isolates provides more efficient infection of host cells (e.g. Korber, et al., 2020). The aspartate in the CTD2 of one protomer is bridged to a lysine at the end of an adjacent protomer's FPPR, and the D614G mutation would therefore destabilize the FPPR, leading to the movement of RBSs toward the host cell membrane, thereby promoting receptor binding (see below).

As is typical for Spike glycoproteins, decoration of the CoV-2 Spike with numerous glycans can be observed. Glycans on the Spike may function to promote proper protein folding and priming by host proteases (Walls et al. 2020). Glycan-mediated shielding of key S protein residues may also allow the virus to evade surveillance by host immune responses, enhancing infection of host cells. To view an animated model of a glycosylated S protein embedded in a viral membrane, see Figure 3.

The fatty acid linoleic acid (LA) is observed to occupy a binding pocket in each protomer's RBD. The pocket is rich in phenylalanine side chains, which together with cysteine disulfide bridges and a tryptophan side chain form a "greasy tube" into which the hydrophobic tail of LA fits. The polar head group of LA is anchored by electrostatic attractions to an arginine and a glutamine from an adjacent RBD in the trimer.

The conformation of individual protomers is altered by LA binding. In this view, a gating helix containing two tyrosine residues can be observed to swing free of the LA binding pocket to accomodate LA when it is bound (LA not shown here). RBDs have been observed to be more loosely packed when LA is not bound, suggesting that LA has a role in maintaining the RBDs in the compacted, closed, conformation referenced above. Indeed, LA has been shown to augment remdesivir therapy in suppressing SARS CoV-2 replication and may do so by helping to lock RBDs in the clamped configuration (Toelzer, et al., 2020). [This morph of cryo-EM structures (6VXX to 6ZB4 - Walls, et al., 2020; Toelzer, et al., 2020) was created using the linear interpolation morpher provided by Karsten Theis: http://www.bioinformatics.org/pdbtools/morph2.]

Return to Beginning

A distinctive feature of SARS Cov-2 is the presence of a polybasic sequence motif at the S1/S2 boundary, lacking in other closely related coronavirus S proteins. This motif falls within a 13 residue loop of each protomer and provides a cleavage site for a host proprotein convertase, furin (see Figure 4). [the loop is not visible in the 6XR8 cryo-EM structure and is simulated here].

Cleavage in this loop, either during viral shedding from a host cell or on approach to the cell surface of a potential target cell, exposes a "C-endR" (C-end Rule) motif at the newly generated C terminus of S1. This RRAR (Arg-Arg-Ala-Arg) sequence is the recognition motif for the Neuropilin-1 receptor that enhances S protein binding and viral uptake (see Figure 2). Furin priming increases infectivity whereas furin inhibition lowers the capacity for SARS-CoV-2 entry (Walls, et al., 2020; Wrapp et al., 2020). Note that furin cleavage separates S1 and S2, although these remain noncovalently associated until further digestion of S2 with the serine protease TMPRSS2 (e.g Hoffman, et al., 2020) - see below.

Furin cleavage in the loop also results in a conformational change of a protomer such that an RBD moves from its clamped, downward, "closed" state to project upward toward the host cell membrane in an "open" conformation that facilitates efficient interaction with the host cell ACE-2 receptor. [This morph of cryo-EM structures (6ZGI, 6ZGG - Wrobel, et al., 2020) was created using the linear interpolation morpher provided by Karsten Theis: http://www.bioinformatics.org/pdbtools/morph2.]

Return to Beginning

At left is the structure of two RBDs of the CoV-2 S protein individually bound to a dimer of two ACE2 receptor proteins (the planes of a membrane are simulated). Note the receptors' hydrophobic membrane spanning helices embedded in the simulated planes of the bilayer. The model of the membrane planes was produced by using the PPM server at the Orientations of Proteins in Membranes (OPM) site at the University of Michigan applied to PDB 6M17.

ACE2 comprises three primary domains: the peptidase (PD), the collectrin (neck), and the linker/transmembrane/cytoplasmic domains. It has been suggested that ACE2 dimers can engage two full prefusion S proteins, provided that one of the S RBDs is the open state, since the PD of ACE2 would clash with the S trimer if an RBD in the closed conformation were bound (Yan, et al., 2020). Dimerization of the ACE2 monomers is largely based on polar interactions between residues in helices of the PD and neck domains.

Turning now to the ACE2 binding of a SARS CoV-2 S protein RBD, an exquisite fit of the molecules can be seen by examining their surface complementarity.

The Receptor Binding Domain contains a twisted, antiparallel 5-stranded β-sheet, several small β-strands, 5 main α-helices, and numerous loops.

The RBD's Receptor Binding Motif (RBM) contains most of the residues that are responsible for SARS Cov-2 binding to the the ACE2 receptor. Three pairs of disulfide bonds are observed to stabilize the RBD core and another pair connects the loops at the end of the RBM.

Numerous ACE2-RBM residues interact along the curved molecular interface in which the RBM "cradles" the N-terminal helix of ACE2 (17 RBM amino acids interact with 20 ACE2 residues).

For one example, a glutamine (Gln498) of SARS CoV-2 is observed to interact with the polar sidechains of Asp38, Tyr41, Gln42, and Lys353 of ACE2.

Another example is the two salt bridges formed between a lysine (K417) of the RBD and an aspartate (D30) of ACE2. K417 endows the surface with a positively charged patch in the groove into which the N-terminal helix of ACE2 fits. It is notable that the position occupied by K417 in the SARS CoV-2 RBD is provided by a hydrophobic valine in SARS CoV, the less infective, 2002 coronavirus. This and other subtle differences between the RBDs may contribute to the higher binding affinity of the CoV-2 Spike to ACE2, and partially account for its greater infection capabilities (Lan, et al., 2020).

Return to Beginning

V. The SARS Cov-2 Spike (S) Protein Structure (postfusion conformation)

Upon receptor binding, and after dissociation of the S1 subunit, cleavage of the S2 subunit of the trimeric Spike by proteases, either at the cell surface or within an endosome if the virus has been endocytosed, leads to dramatic molecular gymnastics that thrust the fusion peptide of the S2 subunit into the host membrane. The substantial conformational rearrangements can be seen by comparing the relative positions of the Heptad Repeat 1 (HR1), the Central Helix (CH), the Fusion Peptide Proximal Region (FPPR), and the Fusion Peptide (FP) in the pre-fusion and post-fusion S2 trimers. [The FPPR and FP were not resolved in the post-fusion, cryo-electron microscopy structure of Cai, et al, (2020) and are represented schematically here].