The SARS-CoV-2 Virus Structure Explained

SARS-CoV-2 is a roughly spherical particle about 90 to 100 nanometers across, wrapped in a stolen patch of host-cell membrane and studded with club-shaped protein spikes that give coronaviruses their crown-like name. Beneath that membrane sit four structural proteins and a single strand of RNA nearly 30,000 nucleotides long, making it one of the largest known RNA virus genomes. Each component has a specific architectural role, and understanding how they fit together explains a great deal about how the virus infects cells, evades immunity, and evolves.

Size and Shape of the Virion

Electron microscopy studies have pinned down the dimensions with good precision. One cryo-electron tomography study calculated an average virion diameter of about 90 nanometers, with some natural variation from particle to particle, across virions grown in three different cell lines.¹ A separate morphometry study measuring particles in ultrathin plastic sections found a median diameter of about 100 nanometers when spikes were excluded, with spikes extending up to 23 nanometers beyond the surface. That same study compared SARS-CoV-2 directly with the original SARS-CoV and found the two viruses virtually identical at the ultrastructural level, though SARS-CoV particles carried roughly 30 percent more spikes per virion.¹

The viral envelope itself is a lipid bilayer taken from the host cell’s internal membranes during assembly. Measurements of the bilayer show it is slightly thinner than the host plasma membrane, at about 3.6 nanometers between phospholipid layers versus 3.9 nanometers for the cell surface, even though membrane-spanning viral proteins are embedded in it.¹ Inside each particle, cryo-electron tomography has resolved an average of about 38 ribonucleoprotein complexes, each a compact cylinder roughly 14 nanometers long, composed of nucleocapsid proteins stacked in parallel pillars around segments of the RNA genome.¹

The Spike Protein

The spike, or S protein, is the most studied component of SARS-CoV-2 and the primary target of vaccines. Each spike is a trimer, meaning three identical protein chains twist together into a single unit that juts from the viral surface. Structurally, the ectodomain of each spike trimer is about 160 angstroms long with a triangular cross-section, resembling the spike of the closely related SARS-CoV.² Each chain divides into two functional halves. The upper portion, called S1, contains the receptor-binding domain (RBD) that makes first contact with human cells. The lower portion, S2, contains the machinery that drives the virus’s membrane into the host-cell membrane so the two can merge.

The spike does not sit in one rigid shape. In its resting “closed” state, the three RBDs point downward, burying the receptor-binding surfaces at the interface between the three chains. To grab onto a human cell, at least one RBD must flip upward into an “open” position, exposing the binding patch. Cryo-electron microscopy has captured both states in fine detail, revealing that the SARS-CoV-2 spike appears more sensitive to switching into the open conformation once ACE2 is nearby than the original SARS-CoV spike was, which may contribute to its higher infectivity.³

How the Spike Binds ACE2 and Enters Cells

The receptor the spike targets on human cells is angiotensin-converting enzyme 2, or ACE2, a protein found on cells in the lungs, gut, heart, and other tissues. The binding interface is surprisingly compact: about 17 residues on the RBD make direct contact with 20 residues on ACE2, using a mix of electrostatic attraction and water-repelling contacts to hold the two surfaces together.⁴ A computational study identified several specific residues on the RBD as “hotspots” responsible for tighter ACE2 binding compared with SARS-CoV, including positions that form both hydrogen bonds and salt bridges with the receptor.⁵

Comparing the SARS-CoV-2 and SARS-CoV binding interfaces reveals they share many of the same contact residues; eight positions are identical between the two viruses, and five more use chemically similar amino acids. But SARS-CoV-2 has one contact residue, Lys417, that sits outside the main binding motif and forms a salt bridge with ACE2 that the original SARS-CoV lacks, potentially giving SARS-CoV-2 a small binding advantage.⁴

The Furin Cleavage Site and Membrane Fusion

After binding ACE2, the spike still needs to be cut by host enzymes before it can drive membrane fusion. This is where one of SARS-CoV-2’s most distinctive features comes in: a furin cleavage site at the junction between S1 and S2. This site is absent in SARS-CoV and most other closely related bat coronaviruses, and its presence has been a subject of intense scientific scrutiny.⁶ Furin, a host enzyme found in many cell types, “primes” the spike by cutting it at this site, and that priming step is important for infectivity in both cell cultures and animal models.⁷ A second protease, TMPRSS2, found on the surface of certain host cells, makes an additional cut at a site called S2′, which unleashes the fusion machinery.

The fusion process involves a dramatic structural transformation. In the prefusion state, parts of the S2 subunit are folded back on themselves in short helices connected by a U-turn loop. When fusion is triggered, these segments refold into one long, continuous helix that stretches roughly 80 angstroms further toward the target membrane than the prefusion form, essentially punching the viral fusion peptide into the host-cell membrane.⁸ Recent work has captured at least two distinct intermediate states during this transformation. The first intermediate forms after ACE2 engagement alone; the second requires protease cleavage at S2′.⁸ This stepwise unfolding ensures that the fusion machinery fires only when the virus is properly docked to a host cell, not prematurely in the extracellular environment.

The Glycan Shield

The spike is not a naked protein. Sugar molecules called glycans are attached across its surface, forming a coat that helps the virus avoid recognition by antibodies. Despite contributing only about 17 percent of the spike trimer’s total molecular weight, these glycans manage to shield roughly 40 percent of the protein surface from immune detection.⁹ The shielding effect holds up across different glycan forms, meaning the specific sugar patterns do not matter much for concealment; it is the physical bulk of the sugars blocking antibody access that counts.⁹

One important exception stands out: the receptor-binding domain itself is relatively poorly shielded by glycans, which makes sense because that surface needs to be exposed to grab ACE2. This gap in the glycan armor is also why many of the most potent neutralizing antibodies target the RBD. However, independent molecular-dynamics work has argued that the SARS-CoV-2 glycan shield is not as dense or as rich in under-processed sugars as those of some other viruses, suggesting the shield has real vulnerabilities despite its extent.¹⁰

The M Protein and Viral Shape

The membrane protein, or M, is the most abundant structural protein in the virion. It is a small transmembrane glycoprotein that spans the lipid envelope and acts as a scaffold, organizing the other structural proteins and giving the particle its shape. M exists in two conformational forms, a long form and a short form, and their relative proportions influence the virion’s properties. Structural studies show that the long form is responsible for the rigidity and tight curvature of the envelope and also recruits spike proteins to the surface, while the short form adds flexibility and lowers spike density.¹¹ M proteins line up in tandem arrays along the membrane, and these oligomeric chains induce a slight curvature that, when propagated across the entire envelope, drives the formation of a round particle.¹¹

The E Protein Ion Channel

The envelope protein, E, is the smallest structural protein and is present in far fewer copies than M. Five E protein subunits assemble into a ring to form a pentameric ion channel, a type of structure sometimes called a viroporin.¹² Multiple independent methods, from analytical ultracentrifugation to solid-state NMR to cryo-electron microscopy, have confirmed the pentameric assembly; no smaller or larger oligomeric forms have been detected.¹³

These channels sit in the membranes of a compartment between the endoplasmic reticulum and the Golgi apparatus inside infected cells, where new virus particles are assembled. There, the E channel conducts calcium ions, and this calcium-channel activity has been linked to the inflammatory responses seen in COVID-19.¹⁴ Blocking the E channel is an active area of drug development, since silencing its ion-conducting activity could potentially dampen both viral assembly and the harmful inflammatory cascade.

The Nucleocapsid Protein and RNA Packaging

Inside the envelope, the roughly 30-kilobase RNA genome does not float loosely. It is wound up by thousands of copies of the nucleocapsid protein, N, into a set of compact cylindrical ribonucleoprotein complexes. N protein has two ordered domains connected by flexible, disordered linker regions. The N-terminal domain is the RNA-binding end. It grabs onto the genome and creates a high local concentration of protein, which then triggers the C-terminal domain to lock neighboring N proteins together through oligomerization, collapsing the RNA into a tightly packed complex.¹⁵

The atomic-resolution structure of the N-terminal domain, solved by combining X-ray diffraction with solid-state NMR, reveals a beta-hairpin loop that is critical for RNA binding and makes this domain a target for both antiviral drugs and vaccine design.¹⁶ From a practical standpoint, rapid antigen tests for COVID-19 typically detect the N protein rather than the spike, because N is produced in large quantities during infection and is less prone to the mutations that rapidly reshape the spike.

The RNA Genome’s Built-In Switches

The SARS-CoV-2 genome is not just a flat string of genetic code. It folds into elaborate secondary structures, including stem-loops, hairpins, and a pseudoknot, that regulate how the virus replicates. The untranslated regions at both ends of the genome contain conserved structural elements shared with other SARS-related viruses.¹⁷

One of the most functionally important RNA structures is the frameshifting stimulation element, which sits in the middle of the genome and causes the ribosome, the cell’s protein-making machinery, to slip backward by one nucleotide while reading the genetic code. This controlled “slippage” allows the virus to produce two different polyproteins from overlapping reading frames, essentially doubling its coding capacity without needing extra genome length. Structural studies have shown that this element forms a three-stemmed pseudoknot, and its shape can switch between multiple conformations including an attenuator hairpin and the pseudoknot form itself.¹⁸ Further cryo-EM work has captured the pseudoknot with an upstream multibranch loop that includes the “slippery site” itself, revealing that the frameshifting signal is structurally more complex and more context-dependent than earlier models assumed.¹⁹ Because this element is essential for replication and highly conserved across coronaviruses, it is considered an attractive drug target: a small molecule that locked the pseudoknot in one conformation could potentially cripple viral replication regardless of what mutations appear in the spike.

The Replication Complex

Once the RNA genome is released inside a host cell, the virus needs to copy it. The core copying enzyme is nsp12, a protein often called the RNA-dependent RNA polymerase. It does not work alone. Structural studies consistently show it forms a complex with two accessory proteins, nsp7 and nsp8, which position the RNA template within the active-site cleft of nsp12 and stabilize the whole assembly.²⁰ Two copies of nsp8 bind on opposite sides of the cleft, and together they help guide the RNA through more than two turns of template-product duplex during copying.²¹

The nsp12 subunit also contains a domain found only in the order of viruses that includes coronaviruses, called the NiRAN domain, which sits on the back of the main polymerase fold. Its exact role is still being clarified, but it appears involved in capping newly made RNA. Remdesivir, one of the first approved antivirals for COVID-19, works by mimicking a nucleotide building block and inserting itself into the growing RNA chain within this polymerase complex, ultimately stalling the copying process.²⁰

Before the polymerase can even be assembled, the virus must cut its large polyproteins into individual functional pieces. This job falls mainly to the main protease (Mpro), which recognizes and cuts at multiple specific sites along the polyprotein chain. Structural work capturing the protease in its resting, precleavage, and postcleavage states has shown that Mpro is selective for only about four positions within each cut site, even though it can accommodate substrates up to ten residues long.²² The antiviral nirmatrelvir (the active ingredient in Paxlovid) was designed to fit into Mpro’s active site and block this cutting process.

How Variants Reshaped the Architecture

The basic blueprint described above applies to all SARS-CoV-2 lineages, but variants of concern have introduced structural changes, mostly concentrated in the spike, that affect how the virus behaves. The Omicron variant arrived with an unprecedented number of spike mutations, many clustered in the RBD and the N-terminal domain. Cryo-EM structures of Omicron’s spike bound to ACE2 revealed that new salt bridges and hydrogen bonds formed by mutated residues partly compensate for other mutations known to weaken ACE2 binding, leaving overall receptor affinity roughly similar to that of the Delta variant.²³

Meanwhile, the remodeled surface of Omicron’s RBD looks dramatically different from earlier variants, which explains its strong ability to dodge neutralizing antibodies raised by vaccination or prior infection.²⁴ Interestingly, the same excessive mutation load that granted immune evasion also compromised the spike’s ability to drive efficient cell-cell fusion, which may be one reason Omicron tended to cause less severe lower-lung disease than Delta in many patients.²⁴ Structural analysis of Omicron’s RBD-ACE2 complex confirmed enhanced affinity for the receptor relative to the ancestral virus, achieved through remodeling of the interaction interface rather than simply adding more contact points.²⁵ The story of variant evolution, then, is one of structural trade-offs: gains in immune escape can come at the cost of fusion efficiency, and gains in receptor affinity may rely on reshuffling existing contacts rather than inventing new ones.

The Lipid Envelope

Because SARS-CoV-2 buds from internal cell membranes rather than the plasma membrane, its lipid envelope reflects the composition of the endoplasmic reticulum and Golgi apparatus. Integrative modeling work has cataloged up to six major lipid types in the viral envelope, including phosphatidylcholine, phosphatidylethanolamine, phosphatidylserine, phosphoinositol, cardiolipin, and cholesterol.²⁶ Simulations and lipid-fingerprinting analyses show that the viral membrane proteins attract negatively charged lipids, with cardiolipin especially enriched in the immediate neighborhood of M and E proteins.²⁶ This selective lipid recruitment is not merely cosmetic. The lipid environment influences how membrane proteins fold, how stiff the envelope is, and how readily the particle fuses with host membranes. The fact that the envelope is a lipid bilayer also explains why soap and alcohol-based sanitizers are so effective against SARS-CoV-2: they dissolve the lipid layer, destroying the particle’s structural integrity and rendering it noninfectious within seconds.

1
PubMed Central. Morphometry of SARS-CoV and SARS-CoV-2 particles in ultrathin plastic sections of infected Vero cell cultures
2
Cell. Structure, Function, and Antigenicity of the SARS-CoV-2 Spike Glycoprotein
3
PubMed Central. Conformational dynamics of SARS-CoV-2 trimeric spike glycoprotein in complex with receptor ACE2 revealed by cryo-EM
4
Nature. Structure of the SARS-CoV-2 spike receptor-binding domain bound to the ACE2 receptor
5
PubMed. Key Interacting Residues between RBD of SARS-CoV-2 and ACE2 Receptor: Combination of Molecular Dynamics Simulation and Density Functional Calculation
6
PubMed Central. The Emergence of the Spike Furin Cleavage Site in SARS-CoV-2
7
PubMed Central. Distinctive Roles of Furin and TMPRSS2 in SARS-CoV-2 Infectivity
8
PLoS Pathogens. Capture of fusion-intermediate conformations of SARS-CoV-2 spike requires receptor binding and cleavage at either the S1/S2 or S2′ site
9
PubMed Central. Beyond Shielding: The Roles of Glycans in the SARS-CoV-2 Spike Protein
10
PubMed Central. Vulnerabilities in coronavirus glycan shields despite extensive glycosylation
11
Nature Communications. Structure of SARS-CoV-2 membrane protein essential for virus assembly
12
The Journal of Physical Chemistry B. Pentameric Architecture of the SARS-CoV-2 Envelope Protein Revealed by SEC-MALS, Cryo-EM, and Molecular Dynamics
13
PubMed Central. SARS-CoV-2 Envelope Protein Forms Clustered Pentamers in Lipid Bilayers
14
PubMed Central. Atomic structure of the open SARS-CoV-2 E viroporin
15
Nucleic Acids Research. Structural domains of SARS-CoV-2 nucleocapsid protein coordinate to compact long nucleic acid substrates
16
PubMed Central. Atomic-Resolution Structure of SARS-CoV-2 Nucleocapsid Protein N-Terminal Domain
17
PubMed Central. RNA genome conservation and secondary structure in SARS-CoV-2 and SARS-related viruses: a first look
18
PubMed Central. The SARS-CoV-2 Programmed −1 Ribosomal Frameshifting Element Crystal Structure Solved to 2.09 Å Using Chaperone-Assisted RNA Crystallography
19
PubMed Central. Structure of the SARS-CoV-2 Frameshift Stimulatory Element with an Upstream Multibranch Loop
20
Cell. Structural Basis for the Inhibition of the SARS-CoV-2 RNA-Dependent RNA Polymerase by Remdesivir
21
Cell Reports. Structural and Biochemical Characterization of the nsp12-nsp7-nsp8 Core Polymerase Complex from SARS-CoV-2
22
PubMed Central. Structural basis for replicase polyprotein cleavage and substrate specificity of main protease from SARS-CoV-2
23
PubMed Central. SARS-CoV-2 Omicron variant: Antibody evasion and cryo-EM structure of spike protein-ACE2 complex
24
PubMed Central. Structural and functional impact by SARS-CoV-2 Omicron spike mutations
25
PubMed Central. Structural basis of SARS-CoV-2 Omicron immune evasion and receptor engagement
26
Cell. Integrative multi-scale structural modeling of the SARS-CoV-2 envelope