Virus%20EvolutionApril7-21.PDF

Virus Evolution Esteban Domingo, Universidad Auto´noma de Madrid, Madrid, Spain Viruses act as independent evolutionary units in continuous interaction with their host organisms. The evolutionary strategies of DNA and RNA viruses differ. Introduction Viruses are a large group of parasites of cells; they use either deoxyribonucleic acid (DNA) or ribonucleic acid (RNA)asgeneticmaterial,andaredependentonhostcell metabolism for replication. Their intracellular life cycle endsbyformationofdiscreteparticles,termedvirions,that are released from cells and can transfer the viral genetic material to other cells. Viruses vary greatly in their structural and genetic complexity, and use an amazing variety of strategies to replicate in cells. The RNA bacteriophage MS2, the ?rst virus for which the entire genomicnucleotidesequencewasdetermined,iscomposed of 3569 nucleotides, and it encodes ?ve proteins. In contrast, the poxviruses have as genetic material a covalently closed, double-stranded, linear DNA molecule of up to 375000bp, and they encode 200 to 300 proteins. Variola virus, the causative agent of smallpox, the only viral disease to have been eradicated in 1977 through human intervention,belongstothis group. Based on the type of nucleic acid present in virions and the nucleic acids used as replication intermediates, four general replication strategies can be distinguished for bacterial,animalandplantviruses(Table 1).Theexistence of such disparate replicative pathways suggests multiple origins of the viruses that we can isolate and study today: they are polyphyletic. It is not possible to derive a single phylogenetic tree that relates all known viruses, not even fortheDNAorRNAvirusesseparately.Viruseshavebeen classi?edintoanumberoffamilies(Table 2).Superfamilies ofanimalandplantRNAviruseshavebeende?nedonthe basis of some nucleotide sequence conservation, which suggeststhattheyareevolutionarilyrelated.Someviruses that follow strategy 1 (Table 1) have a genomic RNA of positivepolaritythat,byconvention,meansthattheRNA has the same polarity as the messenger RNAs used for translation of the viral proteins in infected cells (bacter- iophages Qb, and MS2, animal picornaviruses, alpha- viruses and ?aviviruses). Other RNA viruses have as a genome an RNA of negative polarity (complementary to themessengerRNAs),eitherunsegmented(rhabdoviruses or paramyxoviruses) or segmented (in?uenza viruses). Others have a segmented, double-stranded RNA genome (bacteriophagef6, reoviruses,rotaviruses). The advent of gene isolation and rapid nucleotide sequencing techniques have provided new insights into the structure, genetic organization and evolutionary strategies of viruses. Synthesis of complementary DNA (cDNA) and molecular cloning rendered amenable to analysis viruses that cannot be grown in cell culture Article Contents Secondary article . Introduction . DNA Viruses . RNA Genomes and Quasispecies . Virus Fitness . Evolution and Disease . Acknowledgements Table 1 Basic replication strategies of important viral pathogens a Nucleic acid type in replicative intermediate. Example Replication strategy a Virus Disease 1. RNA ? RNA Poliovirus Poliomyelitis Dengue virus Dengue fever; occasionally, haemorrhagic fever Hepatitis C virus Chronic hepatitis; cirrhosis Aphthoviruses Foot-and-mouth disease of cloven-hooved animals 2. RNA ? DNA ? RNA Rous sarcoma virus Connective tissue tumours in chickens Human immunodeficiency virus 1 Acquired immune deficiency syndrome (AIDS) 3. DNA ? DNA Papillomaviruses Condylomas; genital/oral carcinomas Variola virus Smallpox (eradicated) Herpes simplex virus Genital and oral infections; encephalitis 4. DNA ? RNA ? DNA Hepatitis B virus Chronic hepatitis; cirrhosis; hepatocarcinoma 1ENCYCLOPEDIA OF LIFE SCIENCES / & 2001 Nature Publishing Group / www.els.net (hepatitis viruses, papillomaviruses, etc.). Inferences on virus origins are also becoming more realistic, as they are probably linked to evolutionary mechanisms that are still inoperationatpresent;however,theoriginofvirusesisstill amatterofconsiderablespeculation.Virusesdonotleave any fossil record. Viral-coded proteins very often have cellular counter- parts. This is true of viral nucleic acid polymerases, viral proteases and a number of nucleic acid modi?cation enzymes (endonucleases, helicases, methyltransferases). Viral replicases and reverse transcriptases have overall three-dimensional structures similar to the cell poly- merases, and common catalytic motifs with polymerases andtelomerases(Sousa, 1996). Viruses use the same mechanisms of genetic variation that operate in cells: mutation, recombination and intragenomic rearrangements (duplications, inversions, additions, deletions, transpositions). A salient property ofvirusesisthecompactgeneticinformationachievedbya number of molecular mechanisms (overlapping reading frames, readthrough proteins, ambisense RNA, proteoly- tic processing, alternative RNA splicing, RNA editing, multifunctional proteins). Viruses are independent evolu- tionaryunits in continuousdependenceontheirhostcells andorganisms. Two main models for the origin of viruses have been proposed (Straussetal., 1996). One model proposes that virusesaredescendedfromprimitivegeneticelementsthat must have existed before gene assemblies could be compartmentalized as protocells (primitive forms of modern cells). The second model proposes that viruses are modern derivatives of a cellular world. In particular, they could originate from complex cellular genomes that were parasites of other cells, and that gradually degener- ated, lost their ability to synthesize a number of required metabolites,andbecamecompletelydependentonthehost cells.Virusesneedoneoriginofreplicationorapolymerase recognition signal to act as autonomous replicons in infectedcells.Somesimplegeneticelementssuchasviroids may be derivatives of primitive RNA-like replicons, like Table 2 Virus families, and superfamilies of RNA viruses a a Classification based on data compiled in Webster and Granoff (1999) and Fields (1996). It must be noted that a number of viruses are still unclassified, and that classifications are periodically updated by the International Committee on Taxonomy of Viruses. b Examples of specific viruses mentioned in the text are given in parentheses with the corresponding family name (CB3, Coxsackievirus B3; HBV, Hepatitis B virus; HCV, Hepatitis C virus; HIV, Human immunodeficiency virus; IV, Influenza virus; LCMV, Lymphocytic choriomeningitis virus; PV, Poliovirus; VSV, Vesicular stomatitis virus). Bacterial viruses b DNA Plasmaviridae Fuselloviridae Tectiviridae Lipothrixviridae Myoviridae (T4) Corticoviridae Podoviridae Siphoviridae (lambda) Inoviridae Microviridae Rudiviridae RNA Cystoviridae (?6) Leviviridae (Q?, MS2) Eukaryotic viruses b DNA Hepadnaviridae (HBV) Circoviridae Parvoviridae Papovaviridae Adenoviridae Herpesviridae Poxviridae (Variola virus) Iridoviridae RNA Picornaviridae (PV, CB3) Calciviridae Astroviridae Togaviridae Flaviviridae (HCV) Coronaviridae Paramyxoviridae Rhabdoviridae (VSV) Filoviridae Orthomyxoviridae (IV) Bunyaviridae Arenaviridae (LCMV) Reoviridae Birnaviridae Retroviridae Superfamilies of animal and plant RNA viruses Picorna-like Alpha-like Flavi-like Carmo-like Sobemo-like Corona-like Negative-stranded Double-stranded Virus Evolution 2 ENCYCLOPEDIA OF LIFE SCIENCES / & 2001 Nature Publishing Group / www.els.net those that presumably populated the earth before the advent of protein-mediated catalysis. The prevailing view is that viruses originated once some type of cellular organization hadalready beenestablished,orthat viruses coevolvedwiththeirhostcellsusingcommonassembliesof functionalmodules andreplicationorigins. DNA Viruses There are up to 100-fold di?erences in complexity among di?erentfamiliesofDNAviruses(Table 2),re?ectedinthe size of their genomic DNA. Tailed DNA bacteriophages constitute one of the oldest and most diversi?ed DNA viruses. They have been divided into three groups accordingtotail morphology:contractile(familyMyovir- idaein Table 2), long noncontractile (familySiphoviridae) and short tail (familyPodoviridae). Their genetic material is double-stranded, linear DNA of a size ranging from about19000to about725000 bp. One of the best studied tailed bacteriophages is the Escherichia coli phage T4. Several T4 genes show signi?cant homology with both prokaryotic and eukar- yotic genes (Ackermann, 1998). The T4-coded DNA polymerase, an essential enzyme for T4 DNA replication, has some homology with human and yeast DNA polymerases, but not with the DNA polymerases I and IIIofE.coli,thepresent-dayhostofT4.Enzymesinvolved inDNAreplication,byvirtueoftheessentialfunctionthat they perform, must have been established early in the evolution of DNA genomes. Horizontal gene transfers across cells and viruses are likely to have involved genes encodingDNApolymeraseandotherproteinsneededfor DNA replication. Moreover, T4 and some other bacter- iophages include group I introns (intervening sequences foundwithinageneanditsprimarytranscripts,typicalof eukaryotes). Such introns are also foundin some bacteria and they often include open reading frames promoting theirownmobility(Shub,1991).Bacteriophagesmayhave retained introns that were gradually lost in their host bacteria. Alternatively, the presence of introns in bacter- iophages could be due to horizontal gene transfer from eukaryotic cells. The molecular studies with bacterio- phages suggest an ancient origin of these viruses, in likely coevolution with prokaryotic and eukaryotic cell ances- tors. The parvoviruses of mammals and birds have a single- stranded DNA genome of about 5000 residues, and the human hepatitis B hepadnavirus has a double-stranded DNAgenomeofabout3200bp,althoughinviralparticles the DNA is partially single-stranded. In contrast, the poxvirus, iridovirus and herpesvirus double-stranded DNA genomes range from 130000 to 370000bp. Small DNA viruses (polyomaviruses, papillomaviruses, parvo- viruses) do not encode their own DNA-dependent DNA polymerase, and use cellular DNA polymerases endowed withtheproofreading-repaircorrectionmechanismsoper- atingduringcopyingofthechromosomalDNAofthecell. Some of these viruses can persist as silent episomes (unintegrated covalently closed DNAs) and their DNA replication is often linked to cellular proliferation or di?erentiation (Shadan and Villarreal, 1996). However, they encode early proteins involved in viral DNA replication thatbelong tothe family of Repproteins, also encoded by other DNA viruses. Tight molecular links of small DNA viruses with their hosts may explain their relative genetic stability, in spite of their limited genetic complexity (ShadanandVillarreal, 1996). Thereisaparallelismbetweenthephylogeneticrelation- shipsofindependentisolatesofDNAvirusesandtheirhost species. This suggests frequent viral?host cospeciations duringDNAvirusevolution.Founderevents(theinfection by a single viral genome type that will be transmitted to other members of the same host species) could contribute to the relative genetic homogeneity of the viral genomes that spread throughout the species. However, it seems unlikely that such single events would be common, and thatviralevolutionwouldremainsoconstrained.Evidence ofvirus?hostcospeciationhasbeenobtainedforsmalland large DNA viruses, such as the herpesviruses (McGeoch etal.,1995). The genetic stability of DNA viruses is not continuous nor complete. Phylogenetic trees may be considerably rami?ed in spite of branch lengths (genetic distances between genotypes or clades) being shorter than for some variable RNA viruses (Figure 1). A phylogenetic tree relating many papillomavirus isolates showed a geogra- phical clustering of isolates, but each cluster included a number of distinct, closely related isolates, reminiscent of the pattern seen in RNA virus evolution. Cytopathic parvoviruses show considerable genetic variation in their capsid genes, comparable also to the variation of some RNA viruses, with good evidence for cross-species transmission. A murine parvovirus persisting in L cells establishedaprocessofvirus-cellcoevolutionwithfeatures that are parallel with those described for some RNA viruses (reoviruses, coronaviruses and picornaviruses) (DomingoandHolland,1997). Characterization of the functions performed by viral genes, together with nucleotide and deduced amino acid sequence comparisons, have suggested that the seeming diversity of viruses may actually have emerged from a limited number of genetic elements or modules (Figure 1). Prokaryotic and eukaryotic DNA and RNA viruses may have arisen from favourable combinations of modules comprising groups of genes, individual genes or subge- nomicregionsencodingde?nedfunctionaldomains.Such combinations may have been selected to produce high- ?tnessvirusesunderaparticularbiologicalniche(Botstein, 1980). Virus Evolution 3ENCYCLOPEDIA OF LIFE SCIENCES / & 2001 Nature Publishing Group / www.els.net The likely reshu?ing of modules implies a prominent roleofrecombinationeventsandhorizontalgenetransfers inshapingthepresent-dayDNAgenomes.Thepossibility ofcapturinghostgenesbyDNAviruseswassoonrealized with the discovery of transduction. A group of bacter- iophages, the classical prototype being the E. coli bacteriophagel,arecalledtemperatebecauseratherthan killingtheirhostcellstheycanintegratetheirDNAintothe hostDNA.TheintegratedviralDNAdividesasifitwerea genecluster belongingtothe bacterialchromosome.Only uponinductionofthelyticcycledoesthevirusreplicateto producelargenumbersofprogeny,resultingincelldeath. This is conceptually parallel to integration of the proviral DNA of retroviruses into the animal cell DNA, and its activationbyanumberofexternalstimuli.Bacteriophages cantransducecellulargenesfromonehostcelltoanother, and retroviruses can transduce cellular oncogenes. Poly- omavirusesengageinexuberantgeneticexchangeswiththe DNAofcellsorofothercoinfectingviruses.Thisresultsin the formation of defective viral genomes with deletions, insertions and rearrangements, but keeping always the replication origin of the parental virus. The capture of cellulargenesbyDNAvirusesmaybeanimportantdriving forceforevolutionwhensuchacaptureresultsinaselective advantagefor the new genome combination.New clinical isolates of adenoviruses arise from recombination of existinggenomes, andalso frommutationalevents. Coexistence of a virus and its host means that the virus mustbeabletocounteractdefensiveresponsesofthehost. DNA viruses encode a number of proteins which have a cellular homologue, serving to counteract or modulate host defence responses. The adenovirus proteins E3/19K and E1a suppress surface molecules of the major histo- compatibilitycomplex(MHC)classIandclassIIthatare requiredforT-cellrecognitionofinfectedcells.Glycopro- tein C of Herpes simplex virus blocks complement activation. A number of viral genes encode homologues oftheextracellularbindingdomainsofcytokinereceptors, thereby counteracting the antiviral responses induced by the hostcytokines(Alcam?´ etal.,1998). Although a few host-interfering proteins are expressed by some RNA viruses, the majority are encoded by DNA viruses. This re?ects a fundamentally di?erent evolution- ary strategy to ensure adaptability of complex DNA viruses and of simple DNA and RNA viruses. The tolerance of a genome to accept random (undirected) mutations decreases with genome complexity (Eigen and Biebricher,1988).Maintenanceofthegeneticinformation containedinlargecomplexDNAgenomes,betheyviralor cellular, require the presence of error-correcting mechan- isms to avoid high frequencies of debilitating mutations and genome extinctions. Error correction includes a proofreading-repair 3??5? exonuclease, and a number of postreplicative repair activities that can normally act on double-stranded DNA but not on RNA. The herpes simplex virus DNA polymerase includes a 3??5? exonu- cleasedomain,butthepolioviruspolymerase,thevesicular stomatitis virus transcriptase or the retroviral reverse transcriptasedonot(Sousa,1996;DomingoandHolland, 1997). Complex DNA viruses evolved a combination of gene capture and modulation of activity of captured gene products by mutation, as a central strategy to cope with hostresponses.Theycannotusuallyexploithighmutation rates as a general adaptive strategy. RNA viruses tend to base their adaptive strategies on the production of many mutants (error copies of the parental genome), some of (a) (b) Figure 1 Some general concepts of virus evolution. (a) Genomic sequence alignments of related viruses often show conservation of functional domains (blocks) although they may vary in size, nucleotide sequence, relative position or orientation (block depicted by an arrow in the last two genomes). Sequence alignments of related viral genomes are the first data sets to be used for phylogenetic analyses. (b) Viral genomic sequences can be related by phylogenetic trees. Branches represent minimal genetic distances (nucleotide differences) among the different isolates. Clusters of related viruses are often referred to as clades, types or subtypes. Each tip of a branch is represented by a cloud to indicate the multiple variants that are often found in each viral isolate. Clouds are generally larger for RNA viruses than for complex DNA viruses. Deviant representatives of a distant cloud (arrow) may originate a new virus group that will undergo further diversification. Virus Evolution 4 ENCYCLOPEDIA OF LIFE SCIENCES / & 2001 Nature Publishing Group / www.els.net whichmaymeettherequirementsimposedonthevirusby anenvironmental change. RNA Genomes and Quasispecies The genomes of the RNA viruses have a size in the range 3000?30000 nucleotides. There are even simpler RNA geneticelements(viroids,satellites),andtheyeitherdonot formviralparticlesortheyaredependentonahelpervirus for replication. Many studies suggestthat adaptation and evolutionofthesesimplerepliconsandotherRNAgenetic elements (retroposons, retrotransposons, etc.) have many features incommon with RNAviruses. The ?rst experiments of site-directed mutagenesis, completedbeforetheadventofinvitroDNArecombinant techniques,tookadvantageoftheabilityofQbreplicaseto synthesizefull-lengthcopiesofgenomicRNAharbouring a mutation ata preselected positionof the genome. These procedures allowed the synthesis of an extracistronic mutant of bacteriophage Qb which showed a selective disadvantage relative to the wild-type virus. A mutation rateofabout10 24 substitutionspernucleotideandround of copying was calculated for the precise reversion of the extracistronic mutation to the wild type sequence. Furthermore, it was also observed that serial passage of clones of bacteriophage Qb in its E. coli host produced mutantswithahighfrequency(reviewedinDomingoetal., 1988). Studies with large numbers of RNA viruses, or DNA viruses which include RNA as a replicative inter- mediate(strategies1,2and4in Table 1)haveindicatedthat high mutability is a feature shared by RNA viruses and other RNA genetic elements (Eigen and Biebricher, 1988; Domingo and Holland, 1997). Mutation rates for RNA viruses are several orders of magnitude larger than for cellularDNA(Figure2).ItisestimatedthatanysingleRNA genome copied from a template molecule will contain an averageof0.1?1mutationrelativetothetemplate.Because many mutations (or combinations of mutations accumu- latedafterseveralreplicationrounds)willbedetrimentalto the virus (will cause a decrease in ?tness), many mutants will be eliminated (or kept at very low levels). The elimination of deleterious mutations is called negative or purifying selection. Mutation rates must be distinguished from mutation frequenciesandrateofevolution,andtheseandotherbasic terms are de?ned in Table 3. As a result of high mutation ratesandthetoleranceofgenomesinacceptingmutations, populationsofRNAvirusesexistascomplexdistributions of related but nonidentical genomes called viral quasis- pecies (Figure 3). This concept was ?rst proposed on a theoretical basis to describe self organization and evolu- tion of primitive replicons at the onset of life (Eigen and Biebricher, 1988). The early studies with bacteriophage Qb, and then with many animal and plant RNA viruses, indicated that they have evolutionary features of quasis- pecies. Virologists use an extended de?nition of quasis- pecies to represent dynamic distributions of mutant and recombinantgenomessubjectedtoacontinuousprocessof variant generation, competition and selection (Eigen and Biebricher,1988;DomingoandHolland,1997).Themain departure of quasispecies from previous models of molecular evolution is the consideration of the wild type notasasingle,de?nedgenomebutratherasadistribution ofrelatedgenomes, amutantspectrum (Figure 3). Quasispeciesaretherawmaterialonwhichselectionand randomsamplingprocessesact.Itisacriticalphaseinthe molecular evolution of simple replicons. A number of biological implications of quasispecies (Table 4) relate to the complexity of the mutant distributions, and to the unpredictable environmental changes that viruses must often face. A mutant spectrum replicating in a de?ned environmentwillbepopulatedmainlybythemostneutral Hypermutability 10 ?10 10 ?9 10 ?8 10 ?7 10 ?6 10 ?5 10 ?4 10 ?3 10 ?2 10 ?1 Cellular DNA RNA genomes Hypermutability Hypomutability Figure 2 Mutation rates and frequencies. Cellular DNA replication is several orders of magnitude more accurate than RNA virus genome replication or retrotranscription. An average mutation rate of 10 24 means that one misincorporation error at a given nucleotide template will occur once every 10 000 times that the polymerase copies this specific nucleotide. For both cellular DNA and viral RNA, mutation rates and frequencies much higher than average have been described (hypermutability). For RNA genomes mutation rates and frequencies lower than average may also occur (hypomutability). For specific studies on mutation rates and frequencies see the References and Further Reading. Terms are defined in Table 3. Sequence distribution Average sequence Figure 3 Viral quasispecies. Lines represent genomes and symbols on lines represent mutations. A selection event (or a transmission bottleneck) leads to a change in the average sequence (right). Sequence distribution is also termed the mutant spectrum of the quasispecies. The virus population size that is transmitted (different sizes of the shaded large arrows) has an important influence in RNA virus evolution (see text). Virus Evolution 5ENCYCLOPEDIA OF LIFE SCIENCES / & 2001 Nature Publishing Group / www.els.net mutants in that environment: those that include tolerated mutations(Table3).However,amutationmaybeneutralin one environment and not in another environment. In the course of their normal life cycles viruses often encounter di?erentcelltypesevenwithinasinglehostindividual,and heterogeneous defensive responses in hosts (due to polymorphismsingenesoftheMHCcomplex,andothers). In addition, some viruses, alternate among di?erent host types (distinct vertebrate species, mammals and insects, plantsandinsects,etc.),thusenhancingtheenvironmental heterogeneity that must be faced by viruses. Quasispecies are huge reservoirs of genetic and phenotypic variants. Indeed in some infections, such as with Humanimmuno- de?ciencyvirus type 1 (HIV-1) or the HepatitisBandC viruses, viral loads reach 10 10 to 10 12 infectious units in infectedindividuals,withacontinuousproductionofnew variants. Implications of quasispecies extend to a number of predictions concerning viral disease prevention and therapy(Table 4). Virus Fitness Fitness is an important parameter in biology that, in the case of viruses, measures the relative ability of a virus to produce infectious progeny under a set of environmental conditions.Itisgenerallymeasuredingrowth-competition experiments between a reference virus and the virus to be tested, in cultured cells, animals or plants (Domingo and Holland, 1997). Fitness quantitates the adaptation of the virus to a given environment. Many studies on ?tness variation of RNA viruses have established the following conclusions: . Virus ?tness varies greatly from one environment to another.Vesicularstomatitisvirusadapted to persist in insectcellsdisplayed2C210 6 -foldhigher?tnessininsect cells thanin mammaliancells. . The population size of transmitted virus has an important in?uence in ?tness evolution. Repeated plaque-to-plaque passages (serial bottleneck events as represented by the small arrow in Figure 3) lead to progressiveaverage?tnesslossesduetoaccumulationof deleterious mutations. This phenomenon is known as Muller?s ratchet, as it was ?rst proposed by Muller in 1964 for asexual organisms, and it has found its best experimentalcon?rmationwith RNAviruses. . Largepopulationpassagesresultin?tnessgain.Thisisa consequence of the optimization of the quasispecies mutant spectra when large numbers of genomes are allowedto competein anunrestrictedmanner. . Inthecourseof?tnessgaintwoimportantprinciplesof population genetics were shown to operate: the Red Queen hypothesisandthe competitiveexclusion princi- ple. When twomutantsofequal ?tnesswere allowedto compete, they could coexist for many generations, but eventually one quasispecies displaced the other and became dominant (competitive exclusion). However, duringthecompetitionprocessboththewinnersandthe losersgained?tnesssothattheirrelativepositionsinthe ?tness landscape remained as they were at the onset of the competitions (Red Queen hypothesis) (reviewed in Domingo andHolland, 1997). Table 3 Quantification of mutations, and their effects on virus evolution Mutation rate Frequency of occurrence of a mutation event during template copying Mutation frequency Proportion of mutants (at a specific site or the average for an entire nucleotide sequence) in a genome population Rate of evolution or accumulation of mutations Number of mutations which become dominant per unit time (usually a year). It may refer to changes in an infected individual (during acute or chronic infections) or among many infected hosts (during an epidemic outbreak) Several types of mutations can be distinguished according to their chemical nature and their effects in coding and fitness Transition mutation Replacement of a purine by a purine or of a pyrimidine by a pyrimidine Transversion mutation Replacement of a purine by a pyrimidine or of a pyrimidine by a purine Deletion mutation Loss of one or several nucleotides Addition mutation Acquisition of one or several nucleotides Synonymous (or silent) mutation One that does not give rise to an amino acid substitution in the encoded protein Nonsynonymous mutation One that gives rise to an amino acid substitution in the encoded protein Frameshift mutation One that changes the reading frame in a coding genomic region Neutral mutation One that does not affect fitness in a given environment. A ?silent? mutation need not be ?neutral? as the structure of an RNA molecule (genomic RNA, replicative RNA intermediate or messenger RNA) may influence the viral phenotype Virus Evolution 6 ENCYCLOPEDIA OF LIFE SCIENCES / & 2001 Nature Publishing Group / www.els.net Fitness changes accompany the movements of RNA viruses in sequence space (Eigen and Biebricher, 1988). Sequencespacereferstoallpossiblesequencesavailableto areplicon.ForanRNAgenomeof10000nucleotides(four types of base) the total sequence space is 4 10000 ,an immensenumber,eventhoughonlyatinyportionofitcan be occupied by any functional RNA virus. RNA viruses canuseanadaptivestrategyverysimilartothatproposed for simple replicons during early stages of life because of their limited information content (Eigen and Biebricher, 1988).This mayhavecontributedtotheirgreatsuccess as parasitesofcells.Morethan70%ofthepathogenicviruses are RNAviruses. Evolution and Disease DiseasesassociatedwithDNAandRNAvirusesareoften linked to evolutionary processes, either microevolution within infected hosts, or macroevolution during the expansion of viral pathogens among susceptible hosts. Recombinationandmutationcana?ectviralpathogenesis in a direct way. Avirulent CoxsackievirusB3 evolved to become virulent (cardiopathic) when it replicated in selenium-de?cientmice.Inindependentinfectionsofmice the virulent mutant that was selected di?ered from the avirulentparentinsixpointmutationsatpreciselyde?ned genomic sites. This modi?ed virus was virulent even for nutritionally normal mice. It is believed that selenium de?ciency impaired the immune responses, allowing ampli?cation of quasispecies swarms and selection of the multiply mutated coxsackievirus B3 genome (reviewed in Domingo and Holland, 1997). Reversion of attenuated poliovirusvaccinestrainsresultsinrarepoliomyelitiscases amongvaccineesandtheircontacts.Bothpointmutations and recombination events have been identi?ed in the polioviruspopulationsexcretedbyvaccinees.Occasionally oneofthevariantsmayacquireneurotropicpotentialand cause paralyticdisease. Mutationsthatalterthehostcellspeci?cityofvirusesare potential contributors to pathology, and this may be isolate-dependent. Some lymphocytic choriomeningitis virus(LCMV)isolatescauseagrowthhormonede?ciency syndrome in certain strains of newborn mice, and other isolatesdonotproducethesyndrome.Also,someLCMV mutations are associated with a tropism for neurons, and others with a tropism for cells of the immune system (reviewedinDomingoandHolland,1997).Inthecourseof an HIV-1 infection, a continuous production of mutant viral genomes occurs. Some are selected for their altered cell tropism, and can adapt to the available repertoire of receptors and coreceptors as the infection progresses. For this and other chronic infections there is increasing evidence that antibody-escape and cytotoxic T lympho- cyte-escapemutantsmaybeanimportantelementofviral survival anddisease progression. Whendiseaseisanalysedatthelevelofthecommunityof susceptible hosts and potential new hosts, interesting di?erences between DNA and RNA viruses become apparent. Pandemics (worldwide epidemics) are often associated with RNA viruses (in?uenza virus, HIV-1) and rarely with DNA viruses. Variations in the surface antigens of viruses allows them to replicate in hosts that hadpreviouslybeenexposedtoadi?erentantigenictypeof the same virus. This has been extensively analysed in the case of in?uenza virus. Important human in?uenza pandemics have been associated with a drastic antigenic change(antigenicshift)promptedbytheincorporationof Table 4 Biological implications ofRNA virus quasispecies Derivedfrom theexistenceofamutantspectrum Preexistence (or frequentgeneration) ofviralmutants resistantto antiviral agents,or able to escape immuneresponses Mutantswith altered cellreceptor speci?city or host range Mutantswith di?erent ability toinduceinterferon Nutritional e?ects on virus populationsize and on theevolution ofvirulence Reversionto virulence of attenuatedvaccine strains ofviruses Rapid diversi?cation asa result of selection and randomsampling events. Antigenicshiftand drift Derivedfrom quasispeciesactingasaunitofselection Thresholdsfor phenotypicexpression:adeviantphenotypemaynotbemanifesteddependingonthesizeandcompositionof the mutant spectrum that surrounds the relevant mutant genome Association between pathogenic potentialand mutant spectrumcomplexity Predictionsfor viraldiseasecontrol Vaccines must be multivalent (including multiple B-cell and T-cell epitopes) Antiviraldrug treatmentsshouldbebasedoncombinationtherapy(drugsthataredirectedtoindependentviralgenes,and that are not antagonistic) Speci?c examples can be foundin thetext andin theReferences and FurtherReading. Virus Evolution 7ENCYCLOPEDIA OF LIFE SCIENCES / & 2001 Nature Publishing Group / www.els.net entirely di?erent haemagglutinin or neuraminidase genes into a human in?uenza virus genome constellation. Genomesegmentreassortmentbetweenahumanin?uenza virus and an animal in?uenza virus is often involved. In addition,thevirusmayundergogradualantigenicchange (antigenicdrift)duetopointmutationsinthesamegenes. Interestingly the majority of emergent and reemergent viral diseases recorded in recent years are linked to RNA viruses, and often to those with active recombination. Although evolutionary strategies di?er from the more complexDNAvirusestothesimplestRNAreplicons,there is actually a continuum of slightly di?erent mechanisms aimedatacommonend.Evolutionofvirusesisintimately linked to their survival, persistence and pathogenesis. Changes in the genetic composition of populations of di?erentiatedorganismsmaytakemillionsofyears,while genetic variations of viruses can be detected, quantitated andanalysedindaysormonths.Virusesprovideexcellent model systems for basic studies on molecular evolution. But it is now evident that by understanding more about evolutionary mechanisms of viruses we will also gain insightintostrategiestocombatthediseasestheyproduce. Acknowledgements This work was supported by grants from DGES PM97- 0060-C02-01, FIS 98/0054-01 and Fundacio´ n Ramo´ n Areces. References AckermannH-W(1998)Tailedbacteriophages:TheorderCaudovirales. AdvancesinVirusResearch51: 135?201. Alcam?´ A, Symons JA, Khanna A and Smith GL (1998) Poxviruses: Capturing cytokines and chemokines. SeminarsinVirology 5: 419? 427. Botstein D (1980) A theory of modular evolution for bacteriophages. AnnalsoftheNewYorkAcademyofSciences354: 484?491. DomingoEandHollandJJ(1997)RNAvirusmutationsand?tnessfor survival.AnnualReviewofMicrobiology51: 151?178. DomingoE,HollandJJandAhlquistP(eds)(1988)RNAGenetics.Boca Raton,FL: CRC Press. Eigen M and Biebricher C (1988) Sequence space and quasispecies distribution. In: Domingo E, Holland JJ and Ahlquist P (eds)RNA Genetics, vol. 3,pp.211?245.Boca Raton,FL: CRC Press. FieldsBN,KnipeDM,HowleyPMetal.(eds)(1996)FieldsVirology,3rd edn.Philadelphia: Lippincott-Raven. McGeochDJ,CookS,DolanA,JamiesonFEandTelfordEAR(1995) Molecular phylogeny and evolutionary timescale for the family of mammalianherpesviruses.JournalofMolecularBiology247:443?458. ShadanFFandVillarrealLP(1996)TheevolutionofsmallDNAviruses of eukaryotes: past and present considerations.VirusGenes11: 239? 257. Shub DA (1991) The antiquity of group I introns. CurrentOpinionin GeneticsandDevelopment1: 478?484. SousaR(1996)Structuralandmechanisticrelationshipsbetweennucleic acidpolymerases.TrendsinBiochemicalSciences21: 186?190. StraussEG,StraussJHandLevineAJ(1996)Virusevolution.In:Fields BN,KnipeDM,HowleyPMetal. (eds)FieldsVirology, 3rdedn,pp. 153?171.Philadelphia: Lippincott-Raven. Webster RG and Grano? A (eds) (1999)EncyclopediaofVirology,2nd edn.London: Academic Press. Further Reading Carlile MJ and Skehel JJ (1974)EvolutionintheMicrobialWorld. The Society for General Microbiology, symposium 24. Cambridge: Cam- bridge University Press. DomingoE,HollandJJandAhlquistP(eds)(1988)RNAGenetics,vols 1?3. BocaRaton,FL: CRC Press. Domingo E, Webster RG and Holland JJ (eds) (1999) Origin and EvolutionofViruses. London: Academic Press. Fields BN, Knipe DM, Howley PMetal. (eds) (1996) Fields Virology, 3rdedn.Philadelphia: Lippincott-Raven. GibbsA,CalisherCHandGarc?´a-ArenalF(eds)(1995)MolecularBasis ofVirusEvolution. Cambridge:Cambridge. University Press. KurstakE,MarusykRG,MurphyFAandVanRegenmortelMHV(eds) (1990) Applied Virology Research. Vol. 2: Virus Variability, Epide- miology,andControl. New York: Plenum. Morse SS (ed.) (1993) Emerging Viruses. Oxford: Oxford University Press. Morse SS (ed.) (1994)TheEvolutionaryBiologyofViruses. New York: RavenPress. Palese P and Roizman B (eds) (1980) Genetic variation of viruses. Introductory remarks.AnnalsoftheNewYorkAcademyofSciences 354:ix. Webster RG and Grano? A (eds) (1999)EncyclopediaofVirology,2nd edn.London: Academic Press. Virus Evolution 8 ENCYCLOPEDIA OF LIFE SCIENCES / & 2001 Nature Publishing Group / www.els.net A0436 1..8