Viral genomes and classification

Mutation is not an accident that happens to viruses; it is a rate that follows from the chemistry of their copying enzymes. Replication is not perfectly accurate in any organism, and the error rate depends on whether the polymerase can proofread. DNA polymerases generally have a 3′→5′ exonuclease activity that removes a mis-incorporated base immediately after it is added, which brings the error rate of DNA replication down to roughly one mistake per billion bases copied. The RNA-dependent RNA polymerases of RNA viruses have no such proofreading activity, so their error rate is on the order of one mistake per ten thousand to one per million bases. That difference of a thousandfold or more means an RNA virus does not exist as a single defined sequence: it exists as a population of slightly different variants, sometimes called a quasispecies, and selection acts on that population. A useful quantitative rule follows. There is an error threshold above which a genome accumulates so many mistakes that it can no longer encode functional proteins; the higher the mutation rate, the smaller the genome that can be maintained. This is why RNA viruses generally have small genomes, and why a large DNA virus such as a herpesvirus can afford far more genes. The consequences are visible in the clinic. Influenza viruses change their surface glycoproteins, haemagglutinin and neuraminidase, by gradual point mutation, which is antigenic drift, and by reassortment of their segmented genomes when two strains infect the same cell, which is antigenic shift and can produce a pandemic. This is why the influenza vaccine must be reformulated almost every year, and why surveillance programmes track circulating lineages instead of assuming that last year’s protection still holds. HIV mutates so rapidly that a single drug is defeated within weeks by resistant variants, which is why antiretroviral therapy always uses combinations of three drugs aimed at different steps. For hepatitis C, the same variability delayed vaccine development for decades, although the error rate is also the virus’s weakness: drugs such as sofosbuvir and ribavirin imitate the normal building blocks of RNA, and a polymerase that cannot proofread cannot remove the impostor. For nurses, the practical translation is adherence: an RNA virus left partly suppressed by missed doses keeps the population size and the mutation supply it needs to escape treatment.

Every antiviral drug must interrupt a step that the virus performs and the host cell does not, because a drug that blocks a host enzyme would poison the patient. The Baltimore group identifies exactly which non-host step is available. The clearest example is group VI, the retroviruses. HIV carries its genome as RNA, but reverse transcriptase copies it into DNA, and that DNA is then inserted into the host chromosome by integrase; only afterwards is viral messenger RNA transcribed and translated into proteins. Reverse transcriptase has no human counterpart, so it is an ideal target. Nucleoside reverse transcriptase inhibitors such as zidovudine, tenofovir and emtricitabine are chemical analogues of the normal DNA building blocks. Inside the cell they must be phosphorylated by host kinases to become active, and the viral enzyme then incorporates them into the growing DNA chain, where they act as chain terminators because they lack the 3′ hydroxyl group needed for the next link. Non-nucleoside reverse transcriptase inhibitors such as efavirenz bind a different, allosteric pocket of the same enzyme and distort it; they need no phosphorylation. Integrase inhibitors such as dolutegravir block the insertion step. Protease inhibitors block the cutting of viral polyproteins into working pieces. Because the virus mutates so readily, one drug selects resistant variants within weeks; combination therapy with three drugs from different classes makes simultaneous escape far less likely. The same logic predicts other groups. Group I herpesviruses depend on a viral thymidine kinase to activate acyclovir, so a virus that loses that kinase becomes resistant. Groups III, IV and V need their RNA-dependent RNA polymerase, which is the target of sofosbuvir in hepatitis C and of remdesivir in coronaviruses. Group VII, hepatitis B virus, uses reverse transcriptase and is treated with tenofovir or entecavir. For the nurse this is not abstract: it explains why antiretroviral drugs are never given one at a time, why adherence is a biological requirement rather than a moral one, and why a patient who stops treatment can harbour a resistant reservoir that limits later options.

A classification earns its place only if it changes what someone does, and the capsid scheme passes that test because the envelope is a lipid bilayer. A lipid membrane is held together by hydrophobic interactions, so it is dissolved by detergents and by the alcohols in hand rub, and it is damaged by drying, heat and acid. An enveloped virus therefore survives poorly outside a host: it must move quickly from one wet surface to another, which is why influenza and SARS-CoV-2 spread mainly through respiratory droplets and aerosols, and why hand hygiene and surface cleaning interrupt them. A naked capsid is a protein shell with no membrane, and protein shells are far more resistant to desiccation and to many disinfectants. Naked viruses such as hepatitis A virus and poliovirus survive for long periods on inanimate surfaces, in water and on fomites, which is why they spread by the faecal–oral route and resist the acidity of the stomach. Norovirus, a naked calicivirus, is a classic cause of explosive outbreaks in closed wards, because a very small dose of dried virus on a door handle or a shared toilet can infect the next person. The practical consequences follow directly. For enveloped viruses, soap, alcohol rub, drying and ordinary detergent disinfection are enough, and a virus left on a dry surface loses viability within hours. For naked viruses, alcohol gel is unreliable, chlorine-releasing disinfectants such as sodium hypochlorite are preferred, and handwashing with soap and water matters more than an alcohol rub. Nursing judgement therefore rests on two questions asked together: what is the genome, and does the particle carry an envelope? The first predicts the enzyme and the drug target; the second predicts how long the virus waits on a surface for the next patient.