Science Wiki

An RNA virus is a virus whose genetic material is RNA. The genome may be single-stranded (ssRNA) or double-stranded (dsRNA), and it is usually replicated by a virus-encoded RNA-dependent RNA polymerase (RdRp). In the reverse-transcribing RNA viruses, such as retroviruses, the genome is RNA but replication passes through a DNA intermediate that is made by a reverse transcriptase. RNA viruses infect bacteria, archaea, protists, fungi, plants, and animals.

The International Committee on Taxonomy of Viruses (ICTV) places viruses with RNA-directed polymerases in the realm Riboviria. This realm has two kingdoms. The kingdom Orthornavirae contains the RNA viruses that replicate through an RdRp, which correspond to Groups III, IV, and V of the Baltimore classification. The kingdom Pararnavirae contains the reverse-transcribing viruses, including the retroviruses (Baltimore Group VI) and the DNA viruses of the family Hepadnaviridae (Group VII).

In epidemiology, RNA viruses are the main source of emerging infectious diseases in humans, livestock, and wildlife. They are a minority of all described virus species, but surveys of human pathogens indicate that they account for a majority of emerging and re-emerging zoonotic infections, and several of them have caused pandemics. Human diseases caused by RNA viruses include COVID-19 (SARS-CoV-2), influenza, HIV/AIDS, measles, poliomyelitis, rabies, hepatitis A, hepatitis C, hepatitis E, rotavirus gastroenteritis, Ebola virus disease, Lassa fever, hantavirus pulmonary syndrome, and the major arthropod-borne diseases dengue, yellow fever, Zika, chikungunya, and West Nile fever.

The capacity of RNA viruses to cross species barriers and to evade immunity and drugs is linked to their replication biology. Most viral RdRps and reverse transcriptases lack 3′ to 5′ proofreading exonuclease activity. The main exception is the order Nidovirales, which includes the coronaviruses. RNA viruses therefore mutate at rates of roughly 10⁻⁶ to 10⁻⁴ substitutions per nucleotide per cell infection, which is orders of magnitude higher than the rates of cellular DNA replication. They exist as diverse populations of related variants called quasispecies, and the high mutation rate puts a limit on the size of their genomes, as described by the error threshold of Manfred Eigen. Environmental change, including climate change, land-use change, and global travel, alters the geographic range and transmission of many RNA viruses.

Origins and the monophyly of Riboviria

The origin of RNA viruses is discussed in the context of the RNA world hypothesis and of hypotheses in which viruses derive from cellular genetic elements. Comparative structural genomics indicates that the RNA-directed polymerases of all Riboviria are homologous despite extreme sequence divergence. They share the catalytic core of the RdRp (in Orthornavirae) or the reverse transcriptase (in Pararnavirae).

Viral RdRps and reverse transcriptases have a conserved "right-hand" fold with fingers, palm, and thumb subdomains. The palm contains conserved sequence motifs (A to G), and motif C commonly carries a Gly-Asp-Asp (GDD) sequence. Two catalytic aspartates in motif A and motif C coordinate two divalent metal ions (Mg²⁺ or Mn²⁺), which catalyze phosphodiester bond formation by a two-metal-ion mechanism. Cellular organisms do not encode an RdRp from this lineage. The RdRps of RNA interference in eukaryotes are not related to it and are derived from cellular DNA-dependent RNA polymerases. This supports the view that the viral RdRp lineage is ancient. Whether it predates the last universal cellular ancestor is not established. Structural proteins such as capsid proteins appear to have been acquired from cellular or other viral sources several times.

ICTV taxonomy of Orthornavirae

The ICTV established the realm Riboviria in 2018 and 2019. The kingdom Orthornavirae is divided into several phyla, which correspond to major branches of the RdRp phylogeny:

1. Lenarviricota: This phylum contains the ssRNA(+) RNA bacteriophages of the family Fiersviridae (such as MS2 and Qβ), and the capsidless mitoviruses and narnaviruses of fungi and other eukaryotes, together with the Botourmiaviridae. It is thought to be the most basal group.
2. Pisuviricota: This is a large phylum of positive-sense ssRNA viruses and some dsRNA viruses. It includes the orders Picornavirales (poliovirus, rhinoviruses, other enteroviruses, and hepatitis A virus), Nidovirales (coronaviruses and arteriviruses), Sobelivirales, Patatavirales (potyviruses), and Durnavirales (partitiviruses and picobirnaviruses). The caliciviruses (norovirus) and astroviruses are classified in this phylum as well.
3. Kitrinoviricota: This phylum contains positive-sense ssRNA viruses, including the family Flaviviridae (dengue, yellow fever, Zika, West Nile, and hepatitis C viruses) in the order Amarillovirales, the order Martellivirales (the alphaviruses of the family Togaviridae, such as chikungunya virus, together with many plant viruses), and the order Hepelivirales (hepatitis E virus). Rubella virus, which belongs to the Togaviridae, is classified in the genus Rubivirus.
4. Duplornaviricota: This phylum contains dsRNA viruses with multilayered capsids that transcribe mRNA within the capsid. It includes the order Reovirales (human rotavirus, bluetongue virus, and Colorado tick fever virus) and the Totiviridae of fungi and protists.
5. Negarnaviricota: This phylum contains the negative-sense ssRNA viruses and is divided into two subphyla. The subphylum Haploviricotina includes the order Mononegavirales (the Paramyxoviridae such as measles, mumps, Nipah, and Hendra viruses; the Pneumoviridae; the Rhabdoviridae such as rabies virus; the Filoviridae such as Ebola and Marburg viruses; and the Bornaviridae). The subphylum Polyploviricotina includes segmented viruses that initiate transcription by cap snatching, such as the Orthomyxoviridae (influenza viruses) and the order Bunyavirales (hantaviruses, Crimean-Congo hemorrhagic fever virus, Rift Valley fever virus, Oropouche virus, and the arenaviruses such as Lassa virus). Not all negative-sense viruses use cap snatching.

Other phyla, such as Ambiviricota and Taraviricota, have been proposed or established on the basis of metatranscriptomic data. Large-scale RNA sequencing of environmental samples has greatly expanded the known diversity of RNA viruses, especially in the oceans, and has discovered many lineages that do not infect any known host.

The Baltimore classification

In David Baltimore's system (1971), viruses are classified by the route that leads from the virion genome to messenger RNA (mRNA) that host ribosomes can translate. The RNA viruses are found in Groups III to VI.

Group VII viruses, such as hepatitis B virus, have DNA genomes and are described in the article on DNA viruses.

Chemical lability of RNA

The ribose of RNA carries a 2′-hydroxyl group that is absent from the deoxyribose of DNA. This group can act as an internal nucleophile and attack the adjacent phosphorus in an in-line transesterification, which cleaves the phosphodiester backbone and leaves a 2′,3′-cyclic phosphate and a 5′-hydroxyl. The reaction is accelerated by heat, alkaline pH, and some divalent metal ions. As a result, RNA is much less stable than DNA, by several orders of magnitude, and it is also vulnerable to ubiquitous ribonucleases.

RNA viruses reduce this vulnerability in several ways:

• Single-stranded genomes fold into base-paired stem-loops and pseudoknots, and structured regions are less prone to in-line cleavage.
• The genomes of negative-sense ssRNA viruses are encapsidated along their length by nucleoprotein (about six to nine nucleotides per protomer, depending on the family), and they are not found as naked RNA in infected cells. This protects them from hydrolysis and also from some innate sensors.
• Double-stranded RNA adopts an A-form helix in which the ribose has a C3′-endo pucker. This conformation holds the 2′-hydroxyl away from the geometry required for in-line attack.

Capsid structure and envelopes

Many RNA viruses have icosahedral capsids or helical nucleocapsids. Icosahedral capsids are described by the triangulation number T = h² + hk + k², where h and k are non-negative integers. Picornaviruses have T = 1 capsids with a pseudo-_T_ = 3 arrangement, 60 copies each of four proteins (VP1 to VP4). In many negative-sense RNA viruses, the nucleocapsid is helical and enclosed in a lipid envelope. The structural form of the virion strongly influences its stability in the environment and its route of transmission, as discussed below.

Sites of replication

After entry, an RNA virus must replicate its genome while concentrating the replication proteins and protecting double-stranded intermediates from cytosolic sensors such as RIG-I, MDA5, PKR, and 2'-5'-oligoadenylate synthetase (OAS). Most RNA viruses of eukaryotes replicate in the cytoplasm, and replication is organized in the following ways:

• Single-membrane compartments: Flaviviruses use the non-structural proteins NS4A and NS4B, among others, to invaginate the endoplasmic reticulum and form vesicle packets of roughly 80 nm. Alphaviruses form spherules of about 50 to 60 nm at the plasma membrane and on endolysosomes. Nodaviruses form spherules at the outer mitochondrial membrane. Each spherule has a narrow neck, about 10 nm wide, that allows the import of nucleoside triphosphates and the export of progeny RNA, and in alphaviruses a ring of nsP1 surrounds the neck.
• Double-membrane vesicles: Coronaviruses (using nsp3, nsp4, and nsp6), picornaviruses (using proteins such as 2BC and 3A), and some caliciviruses remodel host membranes into double-membrane vesicles of about 200 to 300 nm. In coronaviruses a molecular pore formed by nsp3 spans the double membrane and is proposed to export newly made RNA.
• Condensates: Many non-segmented negative-sense RNA viruses (rabies virus, which forms Negri bodies; measles virus; respiratory syncytial virus; and Ebola virus) and rotaviruses (which form viroplasms) assemble cytoplasmic inclusions. These are formed at least in part by liquid-liquid phase separation driven by interactions involving the nucleoprotein and phosphoprotein, and they concentrate the polymerase while excluding host proteins.
• Nuclear replication: Orthomyxoviruses (influenza viruses) and bornaviruses replicate in the nucleus, and retroviruses integrate their DNA in the nucleus. Incoming influenza ribonucleoproteins are imported through nuclear pores by importin-_α_/_β_. The heterotrimeric polymerase (PB1, PB2, and PA) binds the C-terminal domain of host RNA polymerase II. PB2 binds the 5′ cap of nascent host pre-mRNAs, and the endonuclease in PA cleaves the host RNA about 10 to 13 nucleotides downstream. This "cap snatching" generates primers for viral mRNA. Viral transcripts of the M and NS segments are spliced by the host spliceosome. HIV-1 reverse transcription is completed inside the capsid core as it travels along microtubules and passes through the nuclear pore, and the proviral DNA is integrated mainly into transcriptionally active chromatin, aided by the host protein LEDGF/p75.

Replication fidelity

DNA polymerases of cellular organisms and large DNA viruses have 3′ to 5′ exonuclease proofreading, and replication is followed by mismatch repair. These systems yield mutation rates of about 10⁻⁸ to 10⁻¹⁰ per nucleotide per replication. Viral RdRps and reverse transcriptases lack proofreading, except in the large nidoviruses, and fidelity depends on the selectivity of nucleotide binding and incorporation, in which conformational changes of the palm and fingers (including motifs A and D) play a part. RNA viruses typically mutate at a rate μ of about 10⁻⁶ to 10⁻⁴ substitutions per nucleotide per cell infection. For a genome of L nucleotides, the genomic mutation rate per replication is:

A virus with a 10 kb genome and μ = 10⁻⁵ therefore makes about 0.1 mutations per newly synthesized genome. For comparison, U is about 0.003 for DNA-based microbes (see Drake's rule), so RNA viruses mutate far more per genome copy.

Quasispecies and Eigen's error threshold

Manfred Eigen formulated quasispecies theory in 1971 for self-replicating molecules with high mutation rates. In a quasispecies, a population is a distribution of closely related variants that are linked by mutation around the fittest sequence, and selection acts on the whole distribution and not on the individual. Let L be the genome length, μ the per-nucleotide mutation rate, and σ > 1 the superiority of the master sequence, which is its replicative fitness relative to the average of its mutants. The probability that a genome is copied without error is:

The master sequence can be maintained only if σ Q > 1. This gives the error threshold:

or, equivalently, μ_max < ln σ / L. Above this threshold, genetic information is lost over successive rounds of replication, a process called error catastrophe. The result has several consequences:

• Genome size limit. Most RNA viruses without proofreading have genomes of about 3 to 15 kb, and the largest is about 20 kb. The size is compatible with μ of about 10⁻⁴ to 10⁻⁵ and a modest value of ln σ. Whether the error threshold is the only reason is debated. The limits may also be affected by other constraints, such as the inefficiency of replicating long RNAs, the need to protect them, and the lack of repair.
• Coronavirus exoribonuclease. Nidoviruses have genomes of about 26 to 41 kb in the coronaviruses and larger in some invertebrate nidoviruses. The coronavirus nsp14 protein, which forms a complex with nsp10, has a 3′ to 5′ exoribonuclease (ExoN) domain of the DEDDh family, and this domain provides proofreading. It improves fidelity by roughly 10-fold or more, and mutation rates of coronaviruses are about 10⁻⁷ to 10⁻⁶ per nucleotide per replication. Inactivation of ExoN makes the virus more susceptible to mutagenic nucleoside analogs.
• Fidelity and fitness. Increasing RdRp fidelity too far can be harmful. The high-fidelity G64S variant of poliovirus polymerase 3D^pol, studied by Vignuzzi, Andino, and others, makes a less diverse quasispecies and is attenuated in mice because it adapts less well to host barriers. This result indicates that virus polymerases have fidelity set by selection for a balance between adaptability and error load, and not necessarily at the point just below the threshold.

Recombination and reassortment

In addition to point mutation (which leads to antigenic drift), RNA viruses change by recombination and by reassortment of genome segments.

• Recombination. In copy-choice recombination, the RdRp or the reverse transcriptase switches from one template to another during synthesis. It is frequent in positive-sense ssRNA viruses such as coronaviruses, picornaviruses, and caliciviruses, and in retroviruses, which package two RNA copies and can switch templates several times in each replication cycle. Examples include circulating vaccine-derived polioviruses, which can recombine with other enteroviruses of species C, the recombinant origin of Western equine encephalitis virus from an ancestor related to eastern equine encephalitis virus and a Sindbis-like virus, and recombination among sarbecoviruses in horseshoe bats, which affects the spike receptor-binding domain. Recombination is rare in many negative-sense ssRNA viruses.
• Reassortment. Viruses with segmented genomes can exchange whole segments when two strains infect a single cell. Influenza A virus has eight segments, so in principle two co-infecting parents can yield 2⁸ = 256 combinations of segments. Antigenic shift arises when a novel hemagglutinin (HA) or neuraminidase (NA) segment is introduced from another host species. It contributed to the pandemics of 1957 (H2N2), 1968 (H3N2), and 2009 (H1N1pdm09). Reassortment among bunyaviruses produced Ngari virus, which contains the L and S segments of Bunyamwera virus and the M segment of Batai virus, and it has also been described in Oropouche virus.

Determinants of cross-species transmission

Compared with many DNA viruses, which tend to co-diverge with their hosts over long periods, RNA viruses account for many contemporary zoonotic spillover events. A virus must clear several barriers to emerge in humans. It must come into contact with humans, bind a compatible cellular receptor, overcome host restriction factors (such as APOBEC3G, tetherin, TRIM5α, MxA, and STAT2), and be transmitted efficiently between humans, which in a simple model requires that the basic reproduction number R₀ exceeds 1. The large quasispecies, the short generation time, and cytoplasmic replication (which avoids dependence on species-specific nuclear machinery) of RNA viruses increase the chance that a variant can use a human receptor, such as ACE2, DPP4, or α2,6-linked sialic acid. Ephrin-B2, the receptor for Nipah virus, is highly conserved among mammals, which explains in part the broad host range of that virus.

Reservoir hosts

Virion structure and transmission route

Enveloped and non-enveloped viruses tend to differ in how they are transmitted.

• Non-enveloped RNA viruses (Picornaviridae, Caliciviridae, Astroviridae, Hepeviridae, and Sedoreoviridae) have protein capsids that resist desiccation, gastric acid (pH about 1.5 to 3), bile salts, and many disinfectants, including alcohol-based hand rubs (though not all). Fecal-oral, waterborne, foodborne, and fomite transmission are therefore common. Hepatitis A virus and hepatitis E virus circulate in blood as quasi-enveloped particles, wrapped in host membranes, and are shed in feces as non-enveloped virions after the membranes are removed by bile salts. Rotavirus infectivity is enhanced by the proteolytic cleavage of VP4 by trypsin.
• Enveloped RNA viruses (Coronaviridae, Orthomyxoviridae, Paramyxoviridae, Pneumoviridae, Filoviridae, Rhabdoviridae, Retroviridae, Flaviviridae, Togaviridae, and Bunyavirales) are generally more sensitive to detergents, solvents, and drying. Most are transmitted in respiratory aerosols or droplets, by contact with blood or other body fluids, by sexual contact, or by arthropod bites. Some enveloped viruses nevertheless infect the gut, including some coronaviruses.

Reproduction numbers

The spread of directly transmitted RNA viruses is described with compartmental susceptible-exposed-infectious-recovered (SEIR) models. The basic reproduction number R₀ is the mean number of secondary cases caused by one typical case in a wholly susceptible population. In a simple model without births and deaths, R₀ = β / γ, where β is the transmission rate and 1/_γ_ is the infectious period. In a partly immune population, the effective reproduction number is:

where _S_(t) is the susceptible fraction. Setting R_e = 1 gives the herd immunity threshold:

Estimates of R₀ depend on the population, the setting, and the method, and the values below are approximate.

Climate change and environmental drivers

Climate change and environmental change affect RNA viruses through several pathways, and the strength of the evidence differs.

Arbovirus transmission and temperature

Mosquitoes, ticks, sandflies, and midges are ectotherms, so their biting rate _a_(T), development rate, and survival _μ_(T) depend on temperature T. The virus also develops faster in the vector as temperature rises, so that the extrinsic incubation period _n_(T) becomes shorter. Combined, these produce unimodal thermal response curves for transmission. Trait-based models give estimated optima of about 26 °C for Aedes albopictus with dengue and chikungunya viruses, about 25 °C for Culex pipiens and Culex quinquefasciatus with West Nile virus, and about 29 °C for Aedes aegypti with dengue, Zika, and chikungunya viruses. The mean temperature also affects vector competence. Warming is associated with the spread of dengue and chikungunya to higher elevations and latitudes, with the spread of West Nile virus in Europe and North America, with the expansion of tick-borne encephalitis and Powassan virus at higher latitudes, and with the arrival of Crimean-Congo hemorrhagic fever virus, carried by Hyalomma ticks, in southern Europe. Changes in land use, travel, trade, vector control, and human behavior also contribute, and attributing a particular outbreak to climate change is difficult.

Range shifts and viral sharing

Climate-driven redistribution of mammals may bring species that were previously separate into contact. A modeling study (Carlson and colleagues, 2022) projected that warming to 2070 could lead to thousands of first-time cross-species viral transmission events among mammals, mainly in high-elevation regions of tropical Africa and Asia, with bats expected to account for much of this change because they can disperse widely. The study was a projection based on models, and the real-world consequences for human spillover are uncertain.

Seasonality of respiratory viruses

In temperate regions, influenza, respiratory syncytial virus (RSV), human seasonal coronaviruses, and several other respiratory viruses peak in winter. Several factors are associated with this pattern, and their relative importance is debated. Studies of influenza show that low absolute humidity (not only low relative humidity) is associated with increased virus survival and transmission. Cold outdoor air holds little water vapor, and heating it indoors gives low relative humidity. Dry air can speed the evaporation of respiratory droplets into small droplet nuclei that stay airborne longer, and it can impair mucociliary clearance and some innate immune responses in the airway. Changes in contact patterns, such as indoor crowding and the school calendar, and host factors such as vitamin D status are also thought to play a role. In the tropics, seasonal peaks of influenza and RSV are often associated with rainy seasons.

Rodent-borne, flood-related, and waterborne viruses

• Hantaviruses. Heavy rainfall during El Niño events in the southwestern United States in 1991 to 1993 was followed by increased vegetation and food, which was followed by a sharp rise in the population of deer mice (Peromyscus maniculatus), the main reservoir of Sin Nombre virus. This trophic cascade is proposed to have contributed to the 1993 outbreak of hantavirus pulmonary syndrome. Rainfall and rodent seed-mast cycles have also been linked to outbreaks of Andes virus in Chile and Argentina.
• Rift Valley fever. In East Africa, extreme rainfall associated with El Niño and the positive phase of the Indian Ocean Dipole floods shallow depressions called dambos. Floodwater mosquitoes, including Aedes mcintoshi, hatch from drought-resistant eggs that may be infected through transovarial transmission. They start epizootics in livestock, and Culex and other mosquitoes amplify them.
• Enteric viruses. Floods and storms can overwhelm sewage systems and contaminate water, which increases outbreaks of non-enveloped viruses such as norovirus, rotavirus, and hepatitis A and E viruses. Hepatitis E infection during late pregnancy has a high case fatality, with reports of about 15 to 25 percent.

Panzootic H5N1 clade 2.3.4.4b

The highly pathogenic avian influenza virus H5N1 of clade 2.3.4.4b has spread through wild birds across Europe, Asia, Africa, North America, South America, and Antarctica since about 2020, and has caused mass mortality in wild birds and in some marine mammals, including sea lions and elephant seals in South America. The virus has repeatedly infected mammals, including foxes, mink, and other carnivores. In March 2024, infection was identified in dairy cattle in the United States (genotype B3.13, and later D1.1 and other genotypes), where the virus replicates to high titers in mammary tissue and is shed in raw milk. Human infections have been reported mainly in workers on dairy farms and poultry farms, and most have been mild, with conjunctivitis the most common sign, although severe and fatal cases have occurred. The risk of sustained human-to-human transmission is currently assessed as low by public health agencies, but the situation changes quickly, and current reports should be consulted.

Several changes in the virus are monitored because they are associated with adaptation to humans:

1. Receptor binding. Avian influenza HA binds α2,3-linked sialic acids, whereas human-adapted HA binds α2,6-linked sialic acids on the upper airway epithelium. Substitutions such as Q226L (H3 numbering) have been shown to shift binding of H5 HA toward α2,6 receptors in experiments, and a single change might be sufficient in some H5 backgrounds.
2. Polymerase adaptation. Avian polymerases are adapted to the warmer temperature of the avian gut (about 40 °C) and to avian ANP32A, a host protein that supports polymerase function. They function less well in the human upper airway (about 33 °C) and with human ANP32A and ANP32B, which lack a 33-amino-acid insertion found in avian ANP32A. Substitutions in PB2 such as E627K, D701N, and Q591R improve activity in mammalian cells. PB2 mutations such as M631L have been associated with the bovine B3.13 genotype.
3. HA stability. A lower pH threshold for membrane fusion by HA, which gives greater stability in the mildly acidic environment of human airway aerosols, is associated with efficient transmission through air in animal models.

Surveillance and diagnostics

• Nucleic acid detection. DNA polymerases such as Taq cannot use RNA as a template, so RNA viruses are detected by RT-PCR, in which viral RNA is first copied into complementary DNA by a reverse transcriptase. Isothermal methods such as RT-LAMP are used for point-of-care tests, and CRISPR-based diagnostic methods using Cas12 and Cas13 (such as SHERLOCK and DETECTR) have been developed.
• Sequencing and phylodynamics. Metagenomic and metatranscriptomic sequencing can identify novel viruses in clinical samples. Bayesian phylogenetic methods (for example BEAST) and platforms such as Nextstrain combine genomes with sampling dates and locations to estimate the evolutionary rate, the time to the most recent common ancestor, and the effective reproduction number, and to separate local transmission from repeated importation.
• Wastewater-based epidemiology. Many RNA viruses are shed in feces or urine, including poliovirus, SARS-CoV-2, influenza A virus (with subtyping used to monitor H5), RSV, norovirus, hepatitis A virus, and some arboviruses. Quantitative PCR and sequencing of sewage give an unbiased population-level signal that can lead clinical case counts by days to about two weeks.

Antiviral drugs

RNA viruses encode a small set of enzymes that can be targeted with little effect on host enzymes, including the RdRp, reverse transcriptase, proteases, and capping enzymes.

Combination therapy and resistance

A person with untreated chronic HIV-1 or hepatitis C virus infection produces on the order of 10¹⁰ to 10¹¹ new virions daily at a mutation rate of about 10⁻⁵ to 10⁻⁴ per nucleotide, so every single-nucleotide substitution is probably already present in the viral population before treatment starts, and many double mutants may be as well. A single drug with a low barrier to resistance therefore selects resistant variants rapidly. If resistance to each of k drugs requires a separate mutation with probability μ, the probability that a given genome already carries all k is roughly:

For μ = 10⁻⁵ and k = 3, this is about 10⁻¹⁵, a value well below the number of virions in the body in most cases. This simple estimate assumes independent mutations at fixed sites, and it ignores linkage, recombination, drug levels, and reservoirs of latent virus. Combination therapy of this kind has nevertheless made HIV-1 infection and hepatitis C effectively controllable or curable.

Vaccines

Vaccination is the main way of controlling many RNA viruses. Platforms include:

1. Live attenuated vaccines, which replicate at a limited level and induce humoral and cellular immunity after one or two doses. Examples are yellow fever 17D, measles-mumps-rubella (MMR), rotavirus vaccines (Rotarix and RotaTeq), oral poliovirus vaccine (including the genetically stabilized novel OPV2, which was designed to reduce reversion to neurovirulence), dengue vaccines (TAK-003, and the Butantan-DV candidate), the chikungunya vaccine VLA1553 (Ixchiq), and the intranasal live attenuated influenza vaccine. Safety concerns have led regulators to restrict or suspend use of Ixchiq in some older adults, and its current status should be checked.
2. Inactivated vaccines, including inactivated poliovirus vaccine, rabies, hepatitis A, Japanese encephalitis, tick-borne encephalitis, and seasonal influenza vaccines.
3. Virus-like particle vaccines, which are empty capsids that display native epitopes. These include the hepatitis E vaccine (Hecolin, licensed in China) and the chikungunya vaccine Vimkunya.
4. Viral-vectored vaccines. Recombinant vesicular stomatitis virus expressing the Zaire Ebola virus glycoprotein (rVSV-ZEBOV, Ervebo) has shown high efficacy, in the range of about 97 to 100 percent in ring vaccination trials, within 10 days of a single dose.
5. Prefusion-stabilized antigens and mRNA vaccines.

Structure-guided vaccine design, pioneered by Barney Graham, Jason McLellan, Peter Kwong, and others, built on the observation that class I viral fusion glycoproteins exist on the virion in a metastable prefusion conformation and refold irreversibly into a postfusion conformation. The F protein of the Pneumoviridae and Paramyxoviridae, the spike of coronaviruses, the glycoprotein of filoviruses and arenaviruses, and the HA of influenza viruses all work in this way. Most potent neutralizing antibodies recognize epitopes that exist only in the prefusion form, whereas the postfusion form displays decoy epitopes. Stabilizing the prefusion conformation, for example with engineered disulfide bonds and cavity-filling substitutions (the DS-Cav1 design for RSV F) and with proline substitutions in a hinge loop (the 2P and HexaPro designs for coronavirus spikes), yields more effective immunogens. This work helps explain why a formalin-inactivated RSV vaccine tested in the 1960s did not protect and was associated with enhanced disease.

Prefusion-stabilized antigens are delivered as recombinant proteins (RSVPreF3, Arexvy; and RSVpreF, Abrysvo) or as nucleoside-modified lipid-nanoparticle mRNA vaccines, such as those against SARS-CoV-2 (BNT162b2 and mRNA-1273) and RSV (mRNA-1345, mRESVIA). Arexvy and Abrysvo were approved in 2023 for older adults, and Abrysvo is also given to pregnant people at 32 to 36 weeks of gestation to protect newborns through passive transfer of antibody. The ability to design and produce an mRNA vaccine within weeks of sequencing a new virus is a major development. For the work on nucleoside modification of mRNA, Katalin Karikó and Drew Weissman received the 2023 Nobel Prize in Physiology or Medicine.

How do coronaviruses maintain 30 kb RNA genomes without suffering Eigen's error catastrophe?

Coronaviruses and related large nidoviruses overcome the ~15 to 20 kb genome size ceiling of other RNA viruses by encoding a dedicated 3′-to-5′ exoribonuclease proofreading domain (ExoN) within non-structural protein 14 (nsp14), activated in complex with nsp10. When the viral RNA-dependent RNA polymerase (nsp12) misincorporates a ribonucleotide, nsp14-nsp10 excises the mismatched 3′ terminal base, improving replication fidelity roughly 10- to 20-fold (μ ≈ 10⁻⁷ to 10⁻⁶ substitutions per nucleotide per replication).

What is the mechanistic difference between antigenic drift and antigenic shift in influenza viruses?

Antigenic drift is the gradual accumulation of point mutations in the surface hemagglutinin (HA) and neuraminidase (NA) glycoprotein genes driven by the error-prone viral RdRp, allowing seasonal influenza A and B strains to progressively escape neutralizing antibodies. Antigenic shift occurs exclusively in segmented viruses (such as influenza A, which has 8 RNA segments) when two distinct viral strains co-infect a single host cell and exchange entire HA or NA genome segments via reassortment, generating novel pandemic subtypes.

Why does nucleoside modification with N1-methylpseudouridine improve mRNA vaccine efficacy?

Unmodified exogenous single-stranded RNA activates endosomal Toll-like receptors (TLR7 and TLR8) and cytosolic sensors (RIG-I, PKR, and OAS), triggering type I interferon signaling that blocks ribosomal translation and degrades the mRNA before sufficient antigen is produced. Incorporating N¹-methylpseudouridine (m¹Ψ) in place of uridine alters secondary structure and steric interactions within innate immune sensors, suppressing premature inflammatory activation while simultaneously increasing ribosomal translation yield and antigen presentation.