Poliovirus Structure: Genome, Capsid, and Function

Poliovirus is one of the simplest and best-understood human pathogens in structural biology, yet its architecture is remarkably elegant. The virus consists of just two main components: a protein shell called the capsid, built from 60 copies of four structural proteins arranged in icosahedral symmetry, and a single strand of RNA roughly 7,500 nucleotides long that encodes everything the virus needs to hijack a human cell. The first atomic-resolution structure of poliovirus was solved by X-ray crystallography in 1985, making it one of the earliest animal viruses visualized at that level of detail.¹ That early work laid the groundwork for decades of research into how this tiny particle enters cells, copies itself, assembles new virions, and evades the immune system.

The Capsid Shell

The poliovirus capsid is an icosahedron about 30 nanometers across, made up of 60 identical building blocks called protomers. Each protomer contains one copy of four viral proteins: VP1, VP2, VP3, and VP4. The first three are the major capsid proteins and form the outer surface, while VP4 is a small internal protein that lines the inside of the shell. Five protomers assemble into a pentamer, and twelve pentamers lock together to form the complete particle.

One of the capsid’s most functionally important features is a depression called the “canyon” that encircles each of the twelve fivefold axes of symmetry. This canyon is where the poliovirus receptor, a cell-surface protein known as CD155, docks during infection. Structural studies showed that the outermost domain of CD155 penetrates into the canyon and contacts all three major capsid proteins simultaneously.² The canyon’s footprint on poliovirus is similar in location to the receptor-binding site seen on related rhinoviruses, pointing to a shared evolutionary strategy among picornaviruses.³

Buried beneath the floor of the canyon, inside the VP1 protein, sits a hydrophobic pocket that normally holds a small lipid molecule sometimes called the “pocket factor.” This lipid acts like a molecular doorstop, stabilizing the capsid in its native conformation. When CD155 binds, the pocket factor can be displaced, destabilizing the capsid and setting in motion the structural rearrangements that ultimately allow the virus to release its genome into the host cell. Research has shown, however, that loss of the pocket factor alone is not enough to trigger irreversible capsid expansion; the receptor must also be present.⁴

The RNA Genome and Its Unusual Features

Inside the capsid sits a single molecule of positive-sense RNA, meaning it can serve directly as a template for protein synthesis as soon as it enters the host cell’s cytoplasm. The genome encodes a single large open reading frame flanked by untranslated regions at both ends. At the very 5ʹ end, instead of the methylguanosine “cap” found on normal cellular messenger RNAs, poliovirus RNA is covalently linked to a small viral protein called VPg. This linkage was first discovered in the mid-1970s and distinguishes poliovirus from capped mRNAs in a way that matters for both replication and translation.⁵

Because the RNA lacks a cap, the virus cannot use the standard mechanism host cells rely on to begin translating mRNAs. Instead, poliovirus evolved a workaround: its 5ʹ untranslated region contains a long, highly structured RNA element called an internal ribosome entry site, or IRES. The IRES folds into a complex series of stem-loops and domains that recruit cellular ribosomes directly to an internal position on the RNA, bypassing the need for a cap entirely.⁶ The site where ribosomes actually begin translating sits about 155 nucleotides downstream of the IRES, at a specific AUG start codon.⁷ Several conserved sequence motifs within the IRES, particularly in domain V, are shared across enteroviruses and rhinoviruses, suggesting that this translation strategy is ancient and deeply embedded in the biology of these viruses.⁸

At the opposite end of the genome, a poly(A) tail serves a critical role during replication. To begin copying its RNA into a complementary negative strand, poliovirus needs both the poly(A) tail and a cloverleaf-shaped RNA structure at the 5ʹ end. Proteins bind to each of these elements and interact with one another, effectively bringing the two ends of the genome together into a circle. This circularization through a protein bridge is required for replication to initiate.⁹

How Poliovirus Enters a Cell

Infection begins when the virus binds CD155 on the surface of a susceptible human cell. That binding event triggers a dramatic structural change: the capsid expands by about 4%, and holes open up at specific symmetry axes. Internal components that were previously hidden, including VP4 and the amino-terminal extension of VP1, become exposed on the outside of the particle. The result is an altered form of the virus known as the 135S particle, named for its sedimentation rate.¹⁰

For years researchers treated the 135S particle as a single, well-defined intermediate. More recent cryo-electron microscopy work has complicated that picture. It turns out that what was called “135S” likely encompasses at least two distinct states. In an early state, the capsid has expanded and developed openings, but most copies of the VP1 amino terminus remain inside the shell, with the holes partially blocked by a loop of VP3. In a later state, those VP1 extensions push through the openings and become fully externalized.¹¹ This distinction matters because the externalized VP1 and VP4 segments are thought to interact with the host cell membrane, helping the virus punch through it to deliver its RNA into the cytoplasm.

How exactly the RNA crosses the membrane has been a longstanding puzzle. Real-time imaging experiments showed that VP4 plays a direct role: it appears to form a protective, membrane-spanning channel through which the genome passes. Adding extra copies of wild-type VP4 to the system boosted the efficiency of genome translocation, while mutant VP4 did not help, suggesting that VP4 interacts dynamically with the RNA rather than forming a rigid, static tube.¹² Once the RNA is safely inside, the empty 80S capsid shell is left behind on the cell surface.

From One Protein to Many

Once the RNA reaches the cytoplasm and ribosomes begin translating it via the IRES, the result is a single enormous polyprotein. This polyprotein has no function on its own; it must be chopped into individual working proteins by viral proteases that are themselves embedded within it. Two proteases do most of the cutting. The 2A protease makes one of the earliest cuts, separating the structural capsid region from the rest of the polyprotein. The 3C protease handles most of the remaining cleavages, recognizing and cutting at specific amino acid pairs throughout the chain.¹³

These proteases do not limit their activity to viral proteins. They also attack host cell factors to shut down host translation, clearing the field for the virus’s own RNA. The 2A protease cleaves a key translation initiation factor called eIF4G, which host cells need to read their own capped mRNAs. But that single cut turns out not to be enough: experiments showed that cleaving eIF4G alone causes only partial translation shutoff. Full suppression of host protein synthesis requires the 3C protease to also cleave poly(A)-binding protein (PABP), stripping away a domain that interacts with several translation factors. Poliovirus thus uses a two-pronged strategy, with both proteases needed to thoroughly silence the host cell’s protein-making machinery.¹⁴

There is an ironic twist to PABP cleavage. Because the poliovirus IRES itself benefits from PABP during translation, destroying too much PABP eventually starts to inhibit viral translation as well. Studies demonstrated that 3C-mediated PABP cleavage suppresses IRES-driven translation at later time points in infection.¹⁵ This self-limiting feedback may serve as a signal for the virus to shift from translation mode to replication and assembly.

Copying the Genome

Poliovirus replicates its RNA using 3Dpol, a viral RNA-dependent RNA polymerase. This enzyme has no equivalent in the host cell, which is why it has long been a target for antiviral drug design. Before 3Dpol can begin copying a new strand, though, it needs a primer. That primer is VPg, the small protein attached to the genome’s 5ʹ end. In the first step of replication, 3Dpol attaches two uridine residues to VPg in a reaction called uridylylation, creating VPgpUpU. This modified VPg then serves as the starting point for RNA chain elongation.¹⁶

The uridylylation reaction depends not only on 3Dpol but also on 3CD, a precursor protein that combines protease and polymerase domains before being fully processed. Mutations in the “thumb” region of 3Dpol dramatically reduced the enzyme’s ability to be stimulated by 3CD during uridylylation, indicating that a specific surface on the polymerase mediates this protein-protein interaction.¹⁷ The dependence on multiple viral proteins working in concert illustrates how poliovirus, despite its tiny genome, orchestrates a surprisingly coordinated replication process.

Replication does not happen freely in the cytoplasm. Poliovirus remodels the host cell’s internal membranes to build specialized compartments called replication organelles.¹⁸ These structures, once thought to be simple vesicles, are actually convoluted, branching, tubular chambers that originate from membranes associated with the cis-Golgi. Early in infection they appear as single-walled tubes; over time they transform into double-membrane structures that enclose bits of cytoplasm.¹⁹ Clustering the replication machinery inside these organelles likely concentrates viral components and shields the RNA from cellular antiviral sensors.

Assembly and the Maturation Cleavage

After the structural proteins are cleaved from the polyprotein, they fold and assemble into pentamers. Twelve pentamers then come together around a copy of the newly synthesized genomic RNA to form a provirion. At this stage, the capsid still contains the uncleaved precursor protein VP0 rather than the mature VP4 and VP2. A final autocatalytic cleavage of VP0 into VP4 and VP2 completes virion maturation, locking in the network of internal protein contacts that makes the particle stable enough to survive outside the cell.²⁰

This cleavage is RNA-dependent: empty capsid shells that assemble without RNA do not undergo it. Mutations at the VP0 cleavage site (the asparagine-serine pair where the cut occurs) delay the cleavage and cause provirions to accumulate. These provirions can eventually mature at elevated temperature, which tells us that the cleavage site is conformationally sensitive.²¹ The requirement for RNA in the maturation cleavage likely serves as a quality-control step, ensuring that only genome-containing particles become fully infectious.

Three Serotypes and Their Antigenic Differences

There are three serotypes of poliovirus (types 1, 2, and 3), each capable of causing disease. They share the same overall architecture and life cycle, but differ in the surface-exposed loops of their capsid proteins. The structural differences cluster mainly in the loop regions of VP1 near the fivefold axes of the capsid. In type 2 poliovirus, for instance, the BC loop of VP1 is disordered, a feature not seen in the other serotypes.²²

The immune system distinguishes the three serotypes primarily through three antigenic sites (sites 1, 2, and 3), each involving specific stretches of amino acids on VP1, VP2, and VP3. Site 1, a 12-amino-acid stretch on VP1, is strongly immunodominant in types 2 and 3 but was not detected for type 1 in the original mapping studies. The other sites are more complex, built from residues contributed by two different capsid proteins.²³ These differences explain why immunity to one serotype does not protect against the others and why vaccines must include components from all three.

The fragility of these antigenic sites has practical consequences for vaccine manufacturing. When poliovirus is chemically inactivated for use in killed (Salk-type) vaccines, the inactivation process can alter epitopes, particularly at site 1 on types 2 and 3. Some monoclonal antibodies that bind the live virus fail to recognize the inactivated form, meaning the immune response elicited by the vaccine may differ subtly from what the live virus would provoke.²⁴

Drugs That Target the Capsid Pocket

The hydrophobic pocket inside VP1, the same one occupied by the pocket factor, has been a major target for antiviral drug development. Small molecules can be designed to wedge themselves into this pocket, mimicking the lipid and stabilizing the capsid against the conformational changes needed for receptor binding and uncoating. Because the amino acids lining the pocket are highly conserved across enteroviruses and rhinoviruses, a single compound can potentially work against many different viruses in these families.²⁵

Interestingly, the structural changes these drugs induce in the capsid when they bind are quite small. The prevailing model is that the drug works not by pushing the capsid into a wrong shape, but by preventing the conformational changes that would otherwise occur during entry. In other words, the drug freezes the capsid in a state that cannot respond to receptor contact.²⁶ Compounds like pleconaril, which advanced to clinical trials for rhinovirus infections, operate on this principle. Drug resistance can arise when mutations in the pocket lining reduce drug binding, and the virus can sometimes compensate by losing the need for a pocket factor altogether, which highlights the evolutionary flexibility built into the capsid.

Recombination and Vaccine-Derived Poliovirus

Poliovirus can swap genetic material with related enteroviruses during co-infection of the same cell. This recombination typically involves the nonstructural genes in the 3ʹ half of the genome rather than the capsid-coding region, meaning a recombinant virus can retain its poliovirus identity (and its ability to cause paralysis) while carrying replication machinery borrowed from another enterovirus species. Laboratory experiments have detected recombinant forms of poliovirus carrying genes from coxsackievirus A21, which mirrors what happens in the real world with circulating vaccine-derived polioviruses (cVDPVs).²⁷

cVDPVs arise when the attenuated Sabin vaccine strains, used in oral polio vaccine, replicate in the human gut long enough to revert key attenuating mutations and recombine with other enteroviruses. Most cVDPV outbreaks involve recombinant strains, though not all do. A cluster of cVDPV1 isolates from Guizhou, China, for instance, had no recombination at all: every region of their genomes remained most closely related to the Sabin 1 parent strain.²⁸ The difference between recombinant and non-recombinant cVDPVs matters for surveillance: sequencing the whole genome rather than just the capsid region is essential for tracking how these viruses evolve and spread.

How Few Mutations Separate Vaccine From Virulence

The structural simplicity of poliovirus makes the margin between a safe vaccine strain and a dangerous revertant alarmingly thin. When researchers compared the full genome sequence of the neurovirulent type 3 poliovirus (P3/Leon/37) with its attenuated Sabin vaccine derivative, they found a maximum of ten point mutations separating the two, and at least five of those were considered minor. Only three positions in the entire genome showed amino acid changes: two in structural proteins (one in VP3, one in VP1) and one in a nonstructural protein.²⁹ The authors suggested that a single base change might be responsible for the attenuated phenotype. This razor-thin genetic margin is part of why the global polio eradication effort has increasingly shifted from the live oral vaccine to the inactivated version, and why even the newest oral vaccine formulations have been engineered with additional genetic stabilizations to resist reversion.