PERSPECTIVES mammalian RNA viruses1–3 and small OPINION DNA viruses such as parvoviruses4–7. Virus sequence change occurs so quickly that Prisoners of war — host adaptation phylogenetic trees of their genes are often temporally structured — viruses from older samples show systematically less and its constraints on virus evolution divergence from the most recent common ancestor (MRCA) than those collected Peter Simmonds, Pakorn Aiewsakun and Aris Katzourakis more recently. As an early example of this phenomenon, the distance from the tree Abstract | Recent discoveries of contemporary genotypes of hepatitis B virus and root of sequences of enterovirus 70 isolates parvovirus B19 in ancient human remains demonstrate that little genetic change collected through the 1970s and 1980s has occurred in these viruses over 4,500–6,000 years. Endogenous viral elements in showed a linear relationship with collection host genomes provide separate evidence that viruses similar to many major date; the calculated nucleotide substitution contemporary groups circulated 100 million years ago or earlier. In this Opinion rate of 5 × 10−3 substitutions per site per year article, we argue that the extraordinary conservation of virus genome sequences is (SSY) allowed the start of the outbreak to be dated to 1967 (REF.8). The availability of virus best explained by a niche-filling model in which fitness optimization is rapidly samples collected over relatively wide date achieved in their specific hosts. Whereas short-term substitution rates reflect the ranges, often stretching back to the 1950s accumulation of tolerated sequence changes within adapted genomes, longer-term or 1960s, has enabled more sophisticated rates increasingly resemble those of their hosts as the evolving niche moulds and Bayesian methods (for example, Bayesian effectively imprisons the virus in co-adapted virus–host relationships. Contrastingly, evolutionary analysis by sampling trees (BEAST)9,10) to estimate dates of origins viruses that jump hosts undergo strong and stringent adaptive selection as they and substitution rates for a wide range of maximize their fit to their new niche. This adaptive capability may paradoxically viruses associated with recent outbreaks. create evolutionary stasis in long-term host relationships. While viruses can evolve Furthermore, these evolutionary timescales and adapt rapidly , their hosts may ultimately shape their longer-term evolution. have often been linked to historical events. Among many examples, a nucleotide substitution rate of 5 × 10−4 SSY in the Viruses, with their often small genomes for extreme genetic conservation of viruses hepatitis C virus (HCV) genome was used and error-prone replication mechanisms, over longer periods of evolution. Newly to calculate dates of emergence of various possess extraordinary adaptive abilities and developed methods to characterize viruses genotype 2 subtypes to 1470, a finding that can display rates of sequence change that are from ancient DNA (aDNA) samples have might explain the association of genotype 2 orders of magnitude greater than those of revealed that viruses that circulated in infections in areas where the slave trade the hosts they infect. They display evolution ancient times do not substantially differ operated several hundred years ago11. On the in real time as they acquire antiviral drug genetically from those that currently basis of its substitution rate and geography resistance, mediate persistent infection circulate in humans. Furthermore, the of currently circulating genotypes, through escape from T and B cell immune discovery of endogenous viral elements hepatitis B virus (HBV) was proposed to system responses to infection or, at the (EVEs) in the genomes of mammals, birds have originated in native South American experimental level, rapidly adapt to different and other eukaryotes shows that viruses populations and spread into Europe and cell culture conditions, new receptors and similar to contemporary virus species elsewhere after contact with the Europeans new hosts. Although biologists since the existed tens of millions of years ago. in the 1500s12. Likewise, the origin of the time of Darwin have convincingly inferred In this Opinion article, we describe a four genotypes of hepatitis E virus (HEV) the existence of natural selection from the niche-filling model of virus evolution that infecting humans was estimated to be current species distributions of animals aims to reconcile these conflicting aspects between 536 and 1,344 years ago13, and and plants, and their genetic relationships, of virus evolutionary histories over different this was suggested to be associated with this evidence is almost always indirect evolutionary timescales — a framework in the spread of pig farming; HEV strains and observational. By contrast, virologists which the host represents the primary driver of genotypes 3 and 4 in Japan apparently have access to the remarkable field of of the longer-term evolution of viruses. originated from the 1900s, when pigs were experimental evolution, such that adaptive first imported from Yorkshire in England14. processes that may occur over centuries Rapid virus sequence change Virus sequence change is often dominated or millennia in larger organisms can be A large body of literature documents by synonymous substitutions in coding observed in viruses over days or weeks. remarkably high nucleotide substitution regions that leave sequences of the encoded The paradox that this Opinion article rates in virus genomes, up to 0.1–1.0% proteins unaltered. Fixation of these changes aims to address is the increasing evidence per year in HIV-1 and a wide range of may be facilitated by repeated transmission NATURE REVIEWS | MICROBIOLOGY VOLUME 17 | MAY 2019 | 321 PERSPECTIVES bottlenecks that reduce effective population immunodeficiency virus (SIV) strains methods. These newly developed methods sizes that occur as the virus transmits that were the source of HIV-1 and HIV-2 have allowed the genomes of ancient between hosts15. Sequence change may infections in humans and of SIV variants human populations to be sequenced and be augmented by adaptive changes. For infecting various monkey species21,25. have enabled direct analyses of genetic example, influenza A virus shows rapid, Relatively rapid substitution rates, such as relationships between contemporary antibody-driven antigenic drift of the the 1.38 × 10−3 SSY (range 1.03–1.73 × 10−3) humans, Neanderthals and Denisovans and haemagglutinin gene that enables it to escape calculated for SIV strains infecting African other archaic human population groups from neutralizing antibodies16. Both HIV-1 green monkeys25, predicted time spans over the past hundred thousand years29–31. and HCV fix several amino acid changes in of hundreds of years for these divergence aDNA-based studies have also contributed to immunodominant T cell epitopes during events and strengthened concepts of investigations of the longer-term evolution primary infection that prevent antigen their relatively recent origins. However, a of viruses over historical timescales, presentation to cytotoxic T cells, contributing subsequent study of SIV strains infecting including the analysis of parvovirus B19 to their ability to replicate and transmit17,18. isolated populations of Old World monkeys (B19V) in human remains dating from These observations contribute to a general on the island of Bioko, Equatorial Guinea, the Second World War in Russia32, the perception of the ephemeral nature of RNA 32 km off the coast of Africa, was entirely pandemic 1918 influenza A virus H1N1 viruses and a broader idea that viruses incompatible with this recent origin strain from Alaskan permafrost33 and are rapidly evolving entities with perhaps hypothesis26. Although post-glacial sea level HBV and smallpox in mummified material frequent recent origins5,19,20. This appears rises separated the island from the African from the 1600s34,35. The timescales over particularly applicable to those emerging landmass over 10,000 years ago, SIV strains which aDNA sequences can be recovered viruses responsible for the numerous recent were found to be minimally divergent have now been extended by three recent and often severe disease outbreaks that from those infecting the same species in reports of the detection of viruses in have afflicted humans, animals and plants. mainland Africa monkey populations. human samples dating back to the early This impression is reinforced by what These observations lowered the minimum Neolithic (5000 bc)36–38. we know about the origins of particular substitution rates of each of the SIV strains Two recent studies report the detection viruses; the emergence of HIV-1 is indeed by over two orders of magnitude and, of HBV in several individuals in European documented to be recent, originating from extrapolated back, predicted an MRCA and Central Asian populations as early as the multiple cross-species transmissions of a for SIVs infecting different host species to Bronze Age and Neolithic (2500–3000 bc36,37). chimpanzee lentivirus into humans in the around 80,000 years before the present (bp). Viruses circulating in these prehistoric late 19th century in Gabon and the Congo21. This isolated (literally) geological times in many cases matched currently This was followed by various genomic separation event provided a single circulating HBV genotypes (genotypes A, changes associated with human adaptation opportunity to look at longer timescales for B and D) and were only