characterization of a putative ancestor of coxsackievirus b5bayesian evolutionary analysis. we...

JOURNAL OF VIROLOGY, Oct. 2010, p. 9695–9708 Vol. 84, No. 190022-538X/10/$12.00 doi:10.1128/JVI.00071-10Copyright © 2010, American Society for Microbiology. All Rights Reserved.

Characterization of a Putative Ancestor of Coxsackievirus B5�

Maria Gullberg,1 Conny Tolf,1 Nina Jonsson,1 Mick N. Mulders,2 Carita Savolainen-Kopra,3Tapani Hovi,3 Marc Van Ranst,4 Philippe Lemey,4

Susan Hafenstein,5 and A. Michael Lindberg1*School of Natural Sciences, Linnaeus University, SE-391 82 Kalmar, Sweden1; Centre for Infectious Disease Control Netherlands,

National Institute for Public Health and the Environment (RIVM), Bilthoven, Netherlands2; Department ofInfectious Disease Surveillance and Control, National Institute for Health and Welfare (THL), Helsinki,

Finland3; Department of Microbiology and Immunology, Rega Institute for Medical Research,Katholieke Universiteit, Leuven, Belgium4; and Department of Microbiology and Immunology,

the Pennsylvania State University College of Medicine, 500 University Drive,Hershey, Pennsylvania 170335

Received 13 January 2010/Accepted 6 July 2010

Like other RNA viruses, coxsackievirus B5 (CVB5) exists as circulating heterogeneous populations of geneticvariants. In this study, we present the reconstruction and characterization of a probable ancestral virion ofCVB5. Phylogenetic analyses based on capsid protein-encoding regions (the VP1 gene of 41 clinical isolates andthe entire P1 region of eight clinical isolates) of CVB5 revealed two major cocirculating lineages. Ancestralcapsid sequences were inferred from sequences of these contemporary CVB5 isolates by using maximumlikelihood methods. By using Bayesian phylodynamic analysis, the inferred VP1 ancestral sequence dated backto 1854 (1807 to 1898). In order to study the properties of the putative ancestral capsid, the entire ancestralP1 sequence was synthesized de novo and inserted into the replicative backbone of an infectious CVB5 cDNAclone. Characterization of the recombinant virus in cell culture showed that fully functional infectious virusparticles were assembled and that these viruses displayed properties similar to those of modern isolates interms of receptor preferences, plaque phenotypes, growth characteristics, and cell tropism. This is the firstreport describing the resurrection and characterization of a picornavirus with a putative ancestral capsid. Ourapproach, including a phylogenetics-based reconstruction of viral predecessors, could serve as a starting pointfor experimental studies of viral evolution and might also provide an alternative strategy for the developmentof vaccines.

The group B coxsackieviruses (CVBs) (serotypes 1 to 6)were discovered in the 1950s in a search for new poliovirus-likeviruses (33, 61). Infections caused by CVBs are often asymp-tomatic but may occasionally result in severe diseases of theheart, pancreas, and central nervous system (99). CVBs aresmall icosahedral RNA viruses belonging to the Human en-terovirus B (HEV-B) species within the family Picornaviridae(89). In the positive single-stranded RNA genome, the capsidproteins VP1 to VP4 are encoded within the P1 region,whereas the nonstructural proteins required for virus replica-tion are encoded within the P2 and P3 regions (4). The 30-nmcapsid has an icosahedral symmetry and consists of 60 copies ofeach of the four structural proteins. The VP1, VP2, and VP3proteins are surface exposed, whereas the VP4 protein linesthe interior of the virus capsid (82). The coxsackievirus andadenovirus receptor (CAR), a cell adhesion molecule of theimmunoglobulin superfamily, serves as the major cell surfaceattachment molecule for all six serotypes of CVB (5, 6, 39, 60,98). Some strains of CVB1, CVB3 and CVB5 also interact withthe decay-accelerating factor (DAF) (CD55), a member of thefamily of proteins that regulate the complement cascade. How-

ever, the attachment of CVBs to DAF alone does not permitthe infection of cells (6, 7, 59, 85).

Picornaviruses exist as genetically highly diverse populationswithin their hosts, referred to as quasispecies (20, 57). Thisgenetic plasticity enables these viruses to adapt rapidly to newenvironments, but at the same time, it may compromise thestructural integrity and enzymatic functionality of the virus.The selective constraints imposed on the picornavirus genomeare reflected in the different regions used for different types ofevolutionary studies. The highly conserved RNA-dependentRNA polymerase (3Dpol) gene is used to establish phyloge-netic relationships between more-distantly related viruses (e.g.,viruses belonging to different genera) (38), whereas the vari-able genomic sequence encoding the VP1 protein is used forthe classification of serotypes (13, 14, 69, 71, 72).

In 1963, Pauling and Zuckerkandl proposed that compar-ative analyses of contemporary protein sequences can beused to predict the sequences of their ancient predecessors(73). Experimental reconstruction of ancestral characterstates has been applied to evolutionary studies of severaldifferent proteins, e.g., galectins (49), G protein-coupledreceptors (52), alcohol dehydrogenases (95), rhodopsins(15), ribonucleases (46, 88, 110), elongation factors (32),steroid receptors (10, 96, 97), and transposons (1, 45, 87). Inthe field of virology, reconstructed ancestral or consensusprotein sequences have been used in attempts to developvaccine candidates for human immunodeficiency virus type 1

* Corresponding author. Mailing address: School of Natural Sci-ences, Linnaeus University, SE-391 82 Kalmar, Sweden. Phone: 46 480446240. Fax: 46 480 446262. E-mail: [email protected].

� Published ahead of print on 14 July 2010.

9695

(21, 51, 66, 81) but rarely to examine general phenotypicproperties.

In this study, a CVB5 virus with a probable ancestral virion(CVB5-P1anc) was constructed and characterized. We firstanalyzed in detail the evolutionary relationships between struc-tural genes of modern CVB5 isolates and inferred a time scalefor their evolutionary history. An ancestral virion sequence wassubsequently inferred by using a maximum likelihood (ML)method. This sequence was then synthesized de novo, clonedinto a replicative backbone of an infectious CVB5 cDNAclone, and transfected into HeLa cells. The hypotheticalCVB5-P1anc assembled into functional virus particles that dis-played phenotypic properties similar to those of contemporaryclinical isolates. This is the first report describing the recon-struction and characterization of a fully functional picornaviruswith a putative ancestral capsid.

MATERIALS AND METHODS

Cell lines and viruses. African green monkey kidney (GMK), human colonadenocarcinoma (HT29), Chinese hamster ovary (CHO), human lung carcinoma(A549), and human rhabdomyosarcoma (RD) cell lines were purchased from theAmerican Type Culture Collection. HeLa Ohio cells were kindly provided by M.Roivainen (Helsinki, Finland). Recombinant CHO cells expressing CAR (CHO-CAR) or DAF (CHO-DAF) were constructed by H.-C. Selinka (77, 84). Cellswere propagated in Dulbecco’s modified Eagle’s medium (DMEM) supple-mented with 2 mM L-glutamine, 100 U/ml penicillin, 0.1 mg/ml streptomycin, and10% newborn calf serum (NCS). Recombinant CHO cells were grown in selec-tive medium supplemented with 1 mg/ml G418 (Sigma) for cells expressing CARand 0.75 mg/ml hygromycin B (Invitrogen) for cells expressing DAF.

The clinical CVB5 isolates used in this study were propagated in GMK cellsaccording to standard procedures (63). CVB5 strain 1954UK85 (CVB5UK)(109) was provided by J. W. McCauley, Newbury, United Kingdom, and strainDalldorf (CVB5D) was provided by R. L. Crowell, Philadelphia, PA (16, 79). Aspreviously described (47), the genome sequence of the CVB5D virus is identicalto that of prototype strain Faulkner (CVB5F) (55), except for one amino acidchange in the VP1 protein. Viral titers of propagated viruses were determined bythe 50% tissue culture infectious dose (TCID50) method according to standardprocedures (41).

Flow cytometry analysis. Flow cytometry analysis was performed as describedpreviously (77). Briefly, CHO, CHO-CAR, CHO-DAF, and HeLa cells werestained with an anti-CAR (RmcB) antibody (43) (the hybridoma was kindlyprovided by L. Philipson and R. Pettersson, Karolinska Institute, Sweden [alsoavailable from the American Type Culture Collection {ATCC CRL-2379}]), ananti-DAF (BRIC110) antibody (Cymbus Biotechnologies), or a mouse IgG1control antibody (X0931; Dako). After 1 h of incubation at 4°C, cells werewashed and stained with a secondary R-phycoerythrin-labeled rabbit anti-mouseantibody (R0439; Dako). Data were acquired by using a FACSCalibur instru-ment (Becton Dickinson) and analyzed with CellQuest, version 3.3, software(Becton Dickinson).

Extraction, amplification, and sequencing. Viral RNA was extracted frominfected cell cultures (QIAamp viral RNA minikit; Qiagen), reverse transcribed(Superscript III; Invitrogen), and PCR amplified (PicoMaxx; Stratagene) byusing virus-specific primers. PCR amplicons were visualized in agarose gels andpurified (QIAquick gel extraction kit; Qiagen). The nucleotide sequences weredetermined with an ABI Prism 3130 automated sequencer (Applied Biosystems)by a primer-walking strategy on both strands using BigDye chemistry (ABI PrismBigDye Terminator cycle sequencing ready-reaction kit, version 1.1; AppliedBiosystems). Sequences were analyzed by using the Sequencher, version 4.6,software package (Gene Codes Corporation).

Phylogenetic analysis. Viral nucleotide sequences were aligned by using Clust-alW (94), and phylogenetic signals were evaluated by using likelihood mapping(90). The presence of nucleotide substitution saturation was tested by using anapproach described previously by Xia et al. (104). Phylogenetic relationshipsbetween the different CVB5 strains, based on the genomic VP1 and P1 regions,were inferred by the ML method as implemented in PhyML (35). Branch supportvalues for inferred phylogeny were estimated by nonparametric bootstrappingconsisting of 1,000 pseudoreplicates (30). The general time-reversible (GTR)substitution model (53) with a gamma-distributed rate heterogeneity was used

for the VP1 analysis, while the same model including the proportion of invariablesites was used for the P1 sequence data. The phylogenetic relationship betweenviruses was also examined by the neighbor-joining method, as implemented inMEGA, version 4.0 (92). Previously determined CVB5 sequences (CVB5UK[GenBank accession number X67706] and CVB5D [47]) and swine vesiculardisease virus (SVDV) sequences (Svdh3jap76 [accession number D00435],Svdj1jap73 [accession number D16364], Svd27uk72 [accession number X54521],Svd1spa93 [accession number AF039166], and Svd1net92 [accession numberAF268065]) were included in the phylogenetic analysis of the VP1 gene. Thesequences of CVB4 Tuscany (CVB4T) (accession number DQ480420) andCVB6 Schmitt (CVB6S) (accession number AF114384) were used as an out-group. Phylogenetic trees were visualized with MEGA 4.0. Root-to-tip diver-gence as a function of sampling time was examined by using Path-O-Gen (avail-able at http://tree.bio.ed.ac.uk/software).

Bayesian evolutionary analysis. We inferred the time scale and tempo ofCVB5 evolution using a Bayesian statistical approach implemented in BEAST(25). This approach employs a full probabilistic model of sequence evolutionalong rooted, time-measured phylogenies with a coalescent prior, using either afixed or relaxed molecular clock model (23, 24). For rapidly evolving viruses, themolecular clock is calibrated based on the divergence accumulation betweensequences sampled at different points in time. We used the SRD06 model ofnucleotide substitution (86) with gamma-distributed rate variation among sites,an uncorrelated log-normal relaxed-clock model, and a Bayesian skyline treeprior (26). Markov chain Monte Carlo analyses were run for 10 million gener-ations and diagnosed by using Tracer (http://beast.bio.ed.ac.uk/Tracer). The evo-lutionary history was summarized in the form of a maximum clade credibility treeby using TreeAnnotator (http://beast.bio.ed.ac.uk/TreeAnnotator) and visual-ized with FigTree (http://tree.bio.ed.ac.uk/software/figtree). Bayesian credibleintervals for continuous parameters are reported as the highest posterior densityintervals, which are the smallest intervals that contain 95% of the posteriordistribution.

Ancestral sequence reconstruction. The ancestral sequences were recon-structed from the CVB5 ingroup taxa of the phylogenetic trees based on thegenetic VP1 and P1 regions but without the reference strains (CVB5D andCVB5UK). These reference strains were not included in the reconstructionbecause their passage history is unknown. ML ancestral sequences were inferredby using a standard codon substitution model (M0, which assumes a homogenousnonsynonymous/synonymous substitution rate among sites and among lineages),as implemented in codeml of the PAML package (34, 105). After ML optimi-zation under the codon model, this approach considers the assignment of a set ofcharacters to all interior nodes at a site as a reconstruction and selects thereconstruction that has the highest posterior probability, i.e., so-called jointreconstruction (106). This procedure was efficiently performed by using an al-gorithm described previously by Pupko and colleagues (78). A correspondingreconstruction was also performed by using the best-fitting empirical amino acidmodel. The inference of the ancestor sequence was facilitated by the absence ofgaps or evidence of recombination within the genomic P1 region, as confirmed byusing the phi test (12).

Construction of CVB5-P1anc. The complete CVB5D genome was amplifiedand cloned into the pCR-Script Direct SK(�) vector (Stratagene) by using theAscI and NotI restriction enzyme cleavage sites, as previously described (Fig. 1)(56). In this infectious full-length cDNA clone of CVB5D (pCVB5Dwt), a ClaIsite was introduced at nucleotide position 3340 to generate a cassette vector(pCVB5D-cas). This modification resulted in one amino acid change in the 2Aprotein (valine to leucine at amino acid position 17). This substitution wasaccepted, as leucine 17 is present in the 2A protein of other enteroviruses,including echovirus 30 and echovirus 21. In addition, a synonymous mutation wasintroduced into the P1 ancestral sequence to remove a SalI site. In order toconstruct a CVB5 clone with the inferred ancestral capsid sequence (pCVB5-P1anc), a pUC57 plasmid containing the ancestral P1 sequence flanked by SalIand ClaI sites was purchased (GenScript). Subsequently, the P1 genomic regionof this pUC57 plasmid was amplified and then cloned into pCVB5D-cas. Theconstructs were propagated in Escherichia coli DH5� cells and purified (Mi-diprep kit; Promega). The nucleotide sequences of all constructs were verified bysequencing as described above.

HeLa cell monolayers were transfected with 2.5 �g of the CVB5 cDNA clonesby using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s pro-tocol. Recombinant viruses were collected on day 5 posttransfection and subse-quently sequenced to confirm their identity. The titers of these viruses weredetermined by a TCID50 assay with HeLa cells. In order to ensure the safety oflaboratory workers, the environment, and the public, the generation of CVB5-P1anc was performed under conditions of biosafety level 2 containment.

9696 GULLBERG ET AL. J. VIROL.

Virus infection. Cell monolayers were infected with virus according to stan-dard procedures (63). Briefly, subconfluent monolayers of cells grown in 25-cm2

flasks were inoculated with viruses at a multiplicity of infection (MOI) of 10TCID50/cell. Following virus adsorption at room temperature for 1 h, the cellswere washed three times before the addition of DMEM supplemented with 2mM L-glutamine, 100 U/ml penicillin, and 0.1 mg/ml streptomycin. Incubation ofthe virus infection at 37°C in a 7.5% CO2 atmosphere was continued for 5 daysor until a cytopathic effect (CPE) was observed.

Viral replication was quantified by analyses of samples taken immediately afterinfection and 5 days postinfection (p.i.) or when a complete CPE was observed.After three cycles of freezing and thawing, the viral titers were determined by aTCID50 assay with HeLa cells as described above.

The molecular evolutions of CVB5-P1anc, 151rom70, 4378fin88, and wild-typeCVB5 (CVB5Dwt) were compared after 10 serial rounds of infection in GMK,HeLa, and RD cells. After the 10th passage, the P1 regions of these viruses weresequenced as described above.

Immunofluorescence. Infected HeLa cells cultured on Lab-Tek II chamberglass slides (Nalge Nunc International) were fixed in 4% formaldehyde for 30min at 4°C and stained for 1 h at room temperature with an enterovirus-specificpolyclonal rabbit antiserum (KTL-482) (42). The primary antibody was visualizedwith a secondary goat anti-rabbit antibody labeled with Alexa Fluor 488 (A11034;Molecular Probes Inc.). Finally, slides were mounted with Vectashield (Im-munkemi) containing 4�,6-diamidino-2-phenylindole (DAPI), and images werecaptured with an epifluorescence microscope.

Plaque formation assay. A semisolid gum tragacanth medium was used aspreviously described (18) to assess the plaque morphology of viruses. Briefly,confluent monolayers of HeLa cells in six-well plates were incubated with 1 ml ofvirus in 10-fold dilutions for 1 h at 37°C. Following adsorption, the virus inocu-lum was aspired, and cells were overlaid with DMEM supplemented with 1%NCS, 2 mM L-glutamine, 100 U/ml penicillin, 0.1 mg/ml streptomycin, and 0.8%(wt/vol) gum tragacanth (Sigma). The plaques were visualized by staining cellswith a crystal violet-ethanol solution after 48 h of incubation at 37°C.

Viral growth kinetics. To assess the growth kinetics of viruses, HeLa cells in24-well plates were infected with virus at an MOI of 10 TCID50/cell as describedabove. Cells and medium were harvested and frozen at various time pointspostinfection. Virus titers in collected samples were determined by a TCID50

assay.Neutralization assay. Serial 2-fold dilutions of antiserum against CVB1,

CVB2, CVB3, or CVB4 (kindly provided by H. Norder and L. Magnius, SwedishInstitute for Infectious Disease Control, Sweden); CVB5F (V032-501-560, 1965;NIH research reagent); or CVB6 (V033-501-560, 1965; NIH research reagent)was mixed with an equal volume of virus (100 TCID50) in DMEM with 2 mML-glutamine, 100 U/ml penicillin, and 0.1 mg/ml streptomycin and then incubatedat 37°C for 1 h. The virus-antibody mixtures were applied onto HeLa cells in96-well plates (quadruplicate). After 5 days of incubation at 37°C, cells wereexamined microscopically for evidence of virus-induced CPE. The highest serumdilution that completely inhibited CPE was determined to be the endpoint titer.

Virus binding assay. Viruses were metabolically labeled with [35S]methionine-cysteine (Perkin-Elmer) and purified by sucrose gradient centrifugation as de-scribed previously for echoviruses 1 and 8 (8). Virus attachment to cells was

measured according to a method described previously (2, 62). Briefly, adherentCHO, CHO-CAR, CHO-DAF, and HeLa cells were detached by EDTA treat-ment. Approximately 2.5 � 105 cells were incubated with radiolabeled virus(�30,000 dpm) in DMEM with 2% NCS, 2 mM L-glutamine, 100 U/ml penicillin,and 0.1 mg/ml streptomycin. Following virus adsorption to cells for 2 h at roomtemperature with gentle agitation, unbound virus was removed by three washes,and cell-bound radioactivity was quantified by liquid scintillation counting. Thevirus binding was analyzed in triplicate, and data are presented as means �standard errors of the means (SEM).

Mapping of structural differences between CVB5-P1anc and the clinical CVB5isolates. Based on a sequence alignment of CVB3 strain M, the clinical CVB5isolates, and CVB5-P1anc (ClustalW) (94), the sequence-equivalent residues inthe ancestor were mapped to the X-ray crystallographic structure of CVB3(Protein Data Bank [PDB] accession number 1COV) (64). The footprints ofCAR (39) and DAF (36) on the CVB3 virion surface were used to locate theequivalent CVB5-P1anc residues within the receptor binding footprints. TheX-ray crystal structure of the CVB3 capsid was modeled and visualized by usingChimera (75, 83).

Nucleotide sequence accession numbers. The VP1 and P1 nucleotide se-quences determined in this study were submitted to the GenBank sequencedatabase under accession numbers GU300033 to GU300065 and GU323569 toGU323576, respectively. Inferred ancestral P1 and VP1 nucleotide sequenceswere deposited under accession numbers GU323568 and GU323577, respec-tively.

RESULTS

Phylogenetic relationships of CVB5 viruses. The gene en-coding the VP1 capsid protein is considered to be the mostinformative genomic region for studying molecular epidemiol-ogy and evolutionary relationships among enteroviruses (13,14, 69, 71, 72). Hence, in order to evaluate the phylogeneticrelationships among 41 clinical CVB5 strains, the VP1-encod-ing gene was sequenced. These viruses were isolated in Eu-rope, Asia, North America, and South America between 1970and 1996. The ML analysis of aligned VP1 nucleotide se-quences of these clinical isolates together with two CVB5 ref-erence strains (CVB5D and CVB5UK) and five SVDV isolatesrevealed a dichotomous phylogenetic relationship, which dis-tinguishes the existence of two coevolving clusters of geneticlineages (indicated as clusters I and II) (Fig. 2A). Some CVB5strains isolated in Finland clustered together in cluster I,whereas other Finnish strains were more closely related toviruses isolated in other parts of Europe and North America,indicating a geographic mixture of cluster I and II viruses. The

FIG. 1. Genomic structures of CVB5D and CVB5-P1anc. An illustration of the genome organization of CVB5 is shown at the top, includingpositions of relevant restriction enzyme sites used to construct infectious viral cDNA clones. The ClaI site (*) was introduced to construct a cassettevector. Infectious CVB5D cDNA clone variants, as they were inserted into the pCR-Script Direct SK(�) vector, are depicted at the bottom. Theancestral P1 sequence is indicated in gray. Names of constructed cDNA clones and viruses generated from these clones (given within parentheses)are used throughout the text. UTR, untranslated region.

VOL. 84, 2010 CVB5 PHYLOGENETICS AND ANCESTRAL RECONSTRUCTION 9697

phylogenetic analysis also showed that SVDV was more closelyrelated to CVB5 viruses in cluster II. In addition to the VP1sequences, the complete sequence of the structural genes (theentire P1 region) of four isolates from each of the two clusters,separated both in time of isolation and by geographic location,was determined. The result from the subsequent phylogeneticanalysis based on the P1 sequences showed a relationship be-tween viruses corresponding to the phylogeny of the VP1 gene(Fig. 2B). Tree topologies corresponding to those resultingfrom the ML analyses were also observed for trees inferredwith the neighbor-joining method (data not shown).

Timed evolutionary history of CVB5 viruses. By plottingroot-to-tip divergence from the ML tree as a function of sam-pling time, we observed a clear accumulation of nucleotidesubstitutions over the sampling time interval (Fig. 3). To es-tablish a time scale for the CVB5 evolutionary history andestimate the viral rate of evolution, we performed Bayesianevolutionary analysis of the VP1 sequences sampled over timeusing BEAST (25). In this approach, we used Markov chainMonte Carlo analyses to average over tree space, with eachtree having branch lengths in units of time. Figure 4 shows themaximum clade credibility tree that summarizes the evolution-

FIG. 2. Phylogenetic relationships among CVB5 isolates. (A) Phylogram based on the nucleotide sequences of the VP1 gene. F, CVB5 isolatesselected for sequencing of the entire P1 region. (B) Phylogeny of selected CVB5 isolates based on the nucleotide sequences of the P1 region. Thescale bars represent the genetic distance (nucleotide substitutions per site). Both ML trees were evaluated by nonparametric bootstrap analysis and1,000 pseudoreplicates. Only bootstrap values �70% are denoted. Clusters are indicated by roman numerals (clusters I and II). The trees wererooted with CVB4T and CVB6S. Abbreviations in isolate names are as follows: den, Denmark; dor, Dominican Republic; ecu, Ecuador; est,Estonia; fin, Finland; fra, France; hon, Honduras; jap, Japan; kyr, Kyrgyzstan; net, Netherlands; pak, Pakistan; rom, Romania; rus, Russia; spa,Spain; uk, United Kingdom; usa, United States of America. The last two numbers of isolates depict the year of isolation, except for isolates fromFrance, where the year of isolation is shown by the two first numbers after fra.


ary history estimated by using a relaxed molecular clock (23).The most recent common ancestor (MRCA) of all CVB5 lin-eages (clusters I and II) dated back to 1854 (1807 to 1898).Cluster II had a somewhat older MRCA than did cluster I

(1913 versus 1933) albeit with overlapping credible intervals(1887 to 1935 versus 1916 to 1947). The evolutionary rateestimate resulted in 0.0042 (0.0033 to 0.0052) nucleotide sub-stitutions per site per year. Despite the rate of evolution andthe relatively old MRCA, no signal for substitution saturationwas detected by using the saturation index reported previouslyby Xia et al. (104). The coefficient of variation (0.24 [0.09 to0.42]) indicated that the rate of evolution varies amongbranches within about 25% of the mean rate. The highest rate(Fig. 3) was observed for the branch leading to the most recentSVDV lineages, which probably reflects the ongoing process ofadaptation to a new host species after interspecies transmissionto pigs. These SVDV lineages were also noted as outliers in theroot-to-tip divergence plot (Fig. 3).

Ancestral reconstruction. In order to reconstruct a putativeancestral character state for the CVB5 virion, we applied acodon-based ML approach to the inferred CVB5 phylogeny. Inthe sequence alignment of the reconstructed CVB5 capsidresidue state (P1anc) and eight clinical isolates, 51 variableamino acid residue positions were observed (Table 1). Therobustness of the P1anc reconstruction was assessed by com-paring the ancestral reconstruction of the P1 sequence with anancestor sequence based on the VP1 gene alone (VP1anc).

FIG. 3. Root-to-tip divergence plot. Shown is a linear regressionplot for root-to-tip divergence versus sampling year. The SVDV iso-lates included in the analysis are encircled.

FIG. 4. Maximum clade credibility tree representing the CVB5 evolutionary history inferred by using Bayesian evolutionary analysis. The treehas branch lengths in time units and is depicted on a time scale. The uncertainty (95% highest posterior density intervals) for the node times isindicated with blue bars. Branches with an asterisk are supported with posterior probabilities higher than 0.85. Rate variation among branches isindicated by using a blue-black-red (slow-average-high) color scheme. Clusters are indicated by roman numerals (clusters I and II).


TABLE 1. Primary structural differences between the capsid proteins of CVB5-P1anc and those of eight clinical CVB5 isolatesa

ResidueAmino acid

Locationb

CVB5-P1anc Cluster I Cluster II SVDV CVB3

VP17 V V V/I M/T I Disordered, N terminus19 G G/S G G G Interior, N terminus52 H H/R H H H Interior54 K K/R K K K Interior75 Y Y Y/F F Y Exterior85 D D D/N N - Exterior, BC loop, antigenic site87 D D/N D D K Exterior, BC loop, antigenic site95 S S/N S/N N T Buried, hydrophobic pocket, antigenic site99 V V V/A V A Buried125 T T T/S T T Exterior156 V V/I V V V Exterior180 M M/I M M M Buried, hydrophobic pocket246 T T/A T T A Buried272 E E E/K E T Exterior, antigenic site273 S S S/G G T Exterior, C terminus275 T T T/A T Q Exterior, C terminus, antigenic site276 D E D D S Exterior, C terminus279 T A T T T Exterior, C terminus281 Q Q Q/R K T Exterior, C terminus

VP237 V T/I V/I V V Interior, N terminus45 D E D/E E S Interior76 P P/S P P P Exterior, BC loop104 A A A/S A T Buried108 I V I I I Buried137 L L/I/V L L L Exterior, EF loop, putative DAF binding site153 S N S/N S/N K Exterior, EF loop154 M T M M/T E Exterior, EF loop, antigenic site156 E E/D E E A Exterior, EF loop160 T T S S S Exterior, EF loop, putative DAF binding site161 T T T/S T G Exterior, EF loop, putative DAF binding site185 F F F/Y F F Buried, EF loop205 I I I/V I T Interior223 I I I/V V V Buried233 T T T/A T/A P Exterior, HI loop, antigenic site

VP335 D A/E D/E/N D R Interior, N terminus39 Q Q Q/R Q E Interior58 T T/I T/A T V Exterior, �-B knob63 L L/S L/M M N Exterior, �-B knob, putative DAF binding site67 A A A/S A A Buried, �-B knob80 S S S/T S T Exterior, BC loop88 T T/I T T Q Exterior168 I I/V I I I Buried181 M M/V M/T M S Exterior, putative CAR binding site183 E E/D E E E Exterior, putative CAR binding site232 K K K/Q K S Exterior, putative DAF binding site235 N N/S N/S N N Exterior, C terminus

VP416 G G G/S S G Disordered18 S S S/N S/N N Disordered20 S S S/A A S Disordered45 D D/E D D D Interior47 T T/A T T T Interior

a Amino acid differences identified in aligned sequences of CVB5-P1anc, the eight clinical CVB5 isolates, SVDV (strain SPA/2�/93), SVDV (strain UKG/27/72), andCVB3 (strain M). The crystallographic information from CVB3 (PDB accession number 1COV) (64) and SVDV (PDB accession numbers 1OOP and 1MQT) (31, 101)as well as the footprints of CAR (39) and DAF (36) on the surface of CVB3 were used. The numbering of amino acids is according to the residue positions in the CVB5sequence. Identified antigenic sites of SVDV are indicated (9, 48, 68).

b Disordered, the corresponding residue in the X-ray crystallographic structure of CVB3 was not ordered; exterior, exposed on the outside of the capsid; interior,exposed on the inside of the capsid; buried, completely contained within the capsid protein shell.


The reconstruction of VP1 was based on the sequences of 41clinical isolates. When comparing the ancestral VP1 sequencederived from P1anc with the corresponding VP1 ancestorbased on the clinical strains, the two sequences matched com-

pletely except for one amino acid residue (aspartic acid inP1anc to asparagine in VP1anc at position 85). The high levelof similarity between the two ancestors indicates that the pre-diction of P1anc is consistent among genome regions of dif-

FIG. 5. Structural differences between CVB5-P1anc and clinical CVB5 isolates. (A, left) A CVB3 (PDB accession number 1COV) (64)protomer in a ribbon diagram with VP1, VP2, VP3, and VP4 (light blue, light green, pink, and light yellow, respectively) with a symmetry-relatedcopy of VP3 included to complete the canyon. Based on a ClustalW alignment of CVB5-P1anc and the CVB5 isolates with CVB3, thesequence-equivalent residues that differ between CVB5-P1anc and the CVB5 isolates are depicted as spheres, and VP1, VP2, VP3, and VP4 areshown in dark blue, green, red, and yellow, respectively. The asymmetric unit is indicated by a black triangle. (Right) A single pentamer of the viruscapsid is surface rendered to show the location of the amino acid differences exposed to the viral surface. (B) Surface-rendered close-up of thepocket with VP1 residues 93 and 178 (i.e., residues 95 and 180 of CVB5) that line the pocket. The pocket factor is shown in orange (64). On theright is the pentamer showing the amino acid differences exposed to the interior surface of the capsid. (C) Residues predicted to interact with CARand DAF. (Left) The protomer is shown in a ribbon diagram, with residues within the CAR and DAF footprints shown as magenta and cyanspheres (36, 39). A symmetry-related copy of VP3 is shown to provide the entire CAR footprint on CVB3 in one asymmetric unit. (Right) In thesurface-rendered pentamer, the CAR footprint on CVB3 is in magenta, and the DAF binding footprint is in cyan.


ferent sizes and robust to sampling variations. Reconstructionsusing an empirical amino acid model, however, resulted inthree different amino acid residues in the reconstructed P1ancestor, i.e., in the VP2 protein (threonine to serine at aminoacid position 160), the VP3 protein (aspartic acid to glutamicacid at position 35), and the VP1 protein (aspartic acid toasparagine at position 85).

To predict the localization of amino acid differences be-tween the phylogenetically reconstructed ancestral virionand the clinical isolates on which this ancestor sequence wasbased, the CVB5 P1 sequences were aligned with that ofCVB3, and the equivalent residues were mapped onto thecrystal structure of CVB3 (PDB accession number 1COV)(64). The structural analysis showed that amino acid differ-ences were clustered into five defined regions, i.e., around theVP3 �-B knob, in a linear cluster on the internal surface, in thehydrophobic pocket, as well as around regions correspondingto the CAR (39) and DAF (36) binding sites on CVB3 (Fig. 5and Table 1). Furthermore, in a CVB5-SVDV comparison, 7of the 51 variable residue positions mapped to regions corre-sponding to antigenic sites of SVDV (Table 1).

Construction and characterization of CVB5-P1anc. In thepicornavirus genome, the P2-P3 region is generally highly con-served among members of a given HEV species, as it codes forproteins essential for viral replication (44). The P1 region, onthe other hand, seems less crucial for RNA replication. Thiswas clearly illustrated when a recombinant poliovirus clonelacking large parts of the P1 region was able to replicate itsRNA in transfected cells (74). In this study, a replication-competent backbone (pCVB5D-cas) based on a full-lengthinfectious CVB5D cDNA clone (pCVB5Dwt) was constructed(Fig. 1). To generate a CVB5 construct with an ancestral cap-sid (pCVB5-P1anc), the P1 region of the pCVB5D-cas vectorwas replaced with a synthesized ancestral P1 sequence. Viruseswere derived from the generated constructs, pCVB5Dwt,pCVB5D-cas, and pCVB5-P1anc, by transfection into HeLacells. All the recombinant CVB5 viruses caused complete cy-tolysis at 96 h posttransfection and replicated to high titers (109

TCID50/ml) in HeLa cells (Fig. 6). The identity of progenyviruses was confirmed by sequencing. Continued functionalcharacterization of HeLa cells infected with CVB5-P1anc(MOI of 10) showed that the virus induced a completedestruction of the cell monolayer within 12 h p.i. (Fig. 7A).

Furthermore, infection with CVB5-P1anc was verified by thedetection of viral antigens using an enterovirus-specific an-tiserum (Fig. 7B).

Comparative analyses of CVB5-P1anc and modern clinicalisolates. To further characterize CVB5-P1anc, properties in-cluding molecular evolution, receptor preferences, plaquemorphology, and cell tropism were analyzed and comparedwith the features of CVB5Dwt and two clinical isolates, onefrom each phylogenetic cluster (4378fin88 from cluster I and151rom70 from cluster II).

Like other RNA viruses, CVB5 has a high mutation rate,which facilitates a rapid adaptation to new environmental con-ditions. In this study, the accumulation of mutations duringreplication in the P1 regions of CVB5-P1anc, 4378fin88,151rom70, and CVB5Dwt, after 10 consecutive passages inthree different cell lines (i.e., GMK, HeLa, and RD cells), wasanalyzed. Sequence analyses of the CVB5-P1anc progeny viruscollected after passages in GMK cells (two independent exper-iments) revealed that the original ancestral P1 sequence wascompletely retained (Table 2). After passages in HeLa cells, asingle-amino-acid substitution in the VP2 protein (in the firstexperiment) or the VP1 protein (in the second experiment)was fixed in the final progeny population. In a correspondingexperiment with CAR-deficient RD cells (77), CVB5-P1ancreplicated without signs of CPE, and serial infections resultedin an introduction of three nonsynonymous mutations, locatedin the VP1 and the VP3 genes. Interestingly, one substitution(lysine to methionine at position 259 in the VP1 protein) wasintroduced into both viral progeny populations after passagesin RD cells. Comparative studies of the accumulation of mu-tations of CVB5-P1anc and 4378fin88, 151rom70, andCVB5Dwt after 10 passages in GMK, HeLa, and RD cellsshowed that the number of mutations introduced into the

FIG. 6. Viral titers of cDNA clone-derived viruses. The titers weredetermined at 5 days after transfection of cDNA clones into HeLa cellsby endpoint titration. Values shown are means � SEM (n 3).

FIG. 7. CVB5-P1anc infection in HeLa cells. (A) Light micro-scopic image of CVB5-P1anc-infected (MOI of 10) HeLa cells (12h p.i.). Bar, 100 �m. (B) Production of viral antigen in HeLa cellsinfected with CVB5-P1anc (MOI of 10) and analyzed at 5 h afterinfection. Antigen was detected with an enterovirus-specific poly-clonal rabbit antibody (KTL-482) and a secondary antibody labeledwith Alexa Fluor 488 (green). The cellular nuclei were visualizedwith DAPI (blue). Bar, 50 �m.


ancestral sequence was not higher than that in sequences ofcontemporary viruses. In conclusion, these results showed thatthe ancestral P1 region of CVB5-P1anc was tolerated duringserial passages in cell culture.

Some viruses use multiple cell surface receptors for initialhost cell attachment. Previously, it was reported that CVB5binds to CAR as a primary receptor (6, 60) but that it also usesDAF as a coreceptor (6, 85). The capacity of radiolabeledCVB5 viruses to bind either CAR or DAF alone or the tworeceptors in combination was assessed by utilizing CHO celllines transfected with either CAR or DAF and HeLa cellsexpressing both CAR and DAF. The expression of cell surfacereceptors was verified by flow cytometry (Fig. 8). At the level ofvirus binding, no significant interaction between radiolabeledCVB5 variants and CHO cells was detected (Fig. 9A). Theviruses bound with equal efficiencies to HeLa and CHO-DAFcells, whereas a lower level of binding to CHO-CAR cells wasmeasured. Hence, the expression of CAR on CHO cells re-sulted in a 10-fold increase in the binding of the CVB5 strains,whereas a significantly higher level of binding was observed inthe presence of DAF (450-fold). Overall, these results indi-cated that CVB5-P1anc, CVB5Dwt, and two clinical CVB5

isolates use both CAR and DAF as cellular attachment mole-cules.

We further complemented our comparative analyses ofCVB5-P1anc by investigating the plaque morphology andgrowth properties of the virus in HeLa cells. The plaque phe-notype of CVB5-P1anc in HeLa cells was similar to those ofplaques formed by CVB5Dwt and the two clinical isolates at48 h p.i. (Fig. 9B). The one-step growth curve analysis of theCVB5-P1anc infection in HeLa cells showed that virus produc-tion started as early as 4 h p.i. (Fig. 9C). This initial sign of viralreplication was followed by a steep rise in virus titers and aplateau that was reached 8 h after infection. Signs of cyto-pathogenicity were already observed at 6 h after viral exposure.Furthermore, CVB5-P1anc replicated as efficiently asCVB5Dwt and the two clinical isolates.

Human enteroviruses infect cells mainly of human or otherprimate origin. Consequently, studies of infectivity in a varietyof cell lines were undertaken as an additional approach toanalyze the host cell specificity of the CVB5-P1anc phenotype.As shown in Fig. 9D, CVB5-P1anc bound to and replicatedefficiently in HeLa, A549, HT29, and GMK cells. This virusreplication caused a CPE typical of enteroviruses. However,

TABLE 2. Mutations and amino acid substitutions in the genomic P1 region of CVB5-P1anc, 151rom70, 4378fin88, or CVB5Dwt after 10passages in different cell lines

Virus and cell lineSample 1 Sample 2

Genomic region Mutation Amino acid substitution Genomic region Mutation Amino acid substitution

CVB5-P1ancGMKa

HeLa VP2 1661T3A 238Y3N VP1 3219A3G 258Q3RRDb VP3 1908A3G 59E3G VP1 3222A3T 259K3M

VP3 2164A3A/GVP1 2524G3G/T 26E3DVP1 3222A3T 259K3M

151rom70GMK VP1 2099A3A/G 132Q3Q/R VP1 2099A3A/G 132Q3Q/RHeLaa

RDb VP3 1163C3T 58T3I VP3 1163C3T 58T3IVP1 2099A3G 132Q3R VP3 1335G3G/A

VP1 2099A3G 132Q3RVP1 2480A3T 259K3M

4378fin88GMK VP1 2090C3T 129S3F VP1 2090C3T 129S3FHeLa VP1 1883C3C/T 60S3S/F VP1 1883C3C/T 60S3S/F

VP1 2090C3T 129S3F VP1 2090C3T 129S3FRDb VP1 1986C3T VP1 2098C3A 132Q3K

VP1 2098C3A 132Q3K VP1 2218A3A/G 172T3T/AVP1 2228G3A 175S3N VP1 2377G3A 225D3NVP1 2377G3A 225D3N

CVB5DwtGMK VP1 2148T3C VP1 2087A3A/G 128Q3Q/R

VP1 2148T3CHeLa VP1 1998A3A/G VP1 1998A3A/GRDb VP3 1057T3C 23S3P VP2 723C3T

VP1 2087A3G 128Q3R VP1 2087A3G 128Q3RVP1 2109C3C/T VP1 2148T3CVP1 2148T3C VP1 2299G3A 199A3TVP1 2248A3T 182I3L

a No mutations were detected.b Noncytolytic infection.


although CVB5-P1anc caused a productive infection in RDcells, no signs of CPE were detected. In contrast, no activereplication or induction of CPE was observed for CHO cells inspite of the detected virus binding to the cell surface. In addi-tion, CVB5Dwt and the two clinical isolates showed a celltropism corresponding to that of CVB5-P1anc (Table 3). Con-clusively, progeny CVB5 virus production was clearly detect-able in all cells assessed except CHO cells.

In order to analyze if specific antiserum generated againstCVB5F (which is the CVB5 prototype strain and identical toCVB5Dwt except for one amino acid substitution in the VP1protein) could neutralize CVB5-P1anc, a neutralization assaywas performed. Infections by CVB5Dwt and two field strainsincluded in this comparative study were neutralized by theantiserum. Interestingly, this anti-CVB5F antiserum wasequally efficient in protecting HeLa cells from infection byCVB5-P1anc (Table 3). These results suggest that the CVB5-P1anc virion shares neutralizing epitopes with the CVB5F lab-oratory strain as well as with recently isolated CVB5 viruses.Further studies of CVB5-P1anc serology showed that neitherantiserum against CVB1 to CVB4 nor antiserum against CVB6protected HeLa cells against infection.

Taken together, these results showed that the CVB5-P1ancvirus constructed using ML phylogenetic methods displayedcharacteristics corresponding to those of present-day circulat-ing CVB5 viruses, including properties such as receptor bind-

ing, plaque morphology, growth kinetics, cell tropism, andoverall structure.

DISCUSSION

Previous studies have provided some insight into the geneticdiversity within the CVB5 serotype (50, 80, 93). In the presentstudy, an ML approach was applied to extend previous studiesand to analyze the global genetic diversity among contempo-rary CVB5 isolates collected over a 25-year period. Consistentwith a study of local CVB5 isolates in Belgium (93), our phy-logenetic analysis of isolates from Europe, Asia, North Amer-ica, and South America showed that CVB5 viruses haveevolved from their MRCA into two major evolutionary lin-eages. The two major cocirculating clades were observed byanalysis of the VP1 gene but also in a corresponding analysis ofthe entire P1 region. There are several possible explanations tothis bimodal CVB5 evolution, including the geographic sepa-ration of ancestral viruses or adaptation processes during earlyhost switch events. Possibly, in the future, viruses in these twoclusters will evolve into two distinct serotypes. Interestingly,the evolution of other serotypes within the HEV-B species,based on a phylogenetic analysis of the VP1 gene, exhibits adifferent tree topology. For example, CVB4 shows a multifur-cating topology (65), whereas echovirus 30 displays a tree-trunk-like pattern (70). However, the evolutionary events lead-ing to these differences in tree topology are not known.

The SVDV isolates included in the phylogenetic analysesconstituted a monophyletic group that was most closely relatedto CVB5 viruses of cluster II, in agreement with data from aprevious report (108). The adaptation of CVB5 from human topig has been estimated to have occurred between 1945 and1965 (108). This highly contagious porcine CVB5 variantcauses symptoms similar to those of foot-and-mouth diseasevirus and is therefore economically important (27, 54, 67).

The high rate of RNA virus evolution allows a rapid adap-tation to new environments (20, 57, 76, 102). Among picorna-viruses, it has been shown for poliovirus that every progenyRNA molecule on average contains one mutation (22). Byusing a relaxed-clock model, the accumulation of mutations inthe heterochronous CVB5 sequence data was transformed intoan estimated evolutionary rate of approximately 0.004 nucleo-tide substitutions per site per year, which in turn correspondsto an MRCA dating back to the mid-19th century. This evo-lutionary rate is within the range estimated for other picorna-viruses (11, 29, 108). For example, an evolutionary rate of

TABLE 3. Comparisons of cell tropisms and CVB5 antiserumcross-reactivities between CVB5-P1anc and

modern CVB5 isolates

VirusCell tropisma

Neutralizationdilutionc

CHO GMK HeLa RDb A549 HT29

CVB5-P1anc � � � � � 1:4,096CVB5Dwt � � � � � 1:8,192151rom70 � � � � � 1:4,0964378fin88 � � � � � 1:4,096

a Plus sign, cytolytic infection; minus sign, no infection.b Noncytolytic infection.c Highest dilution of serum that neutralizes virus infection.

FIG. 8. Flow cytometric analysis of CAR and DAF expression onCHO, CHO-CAR, CHO-DAF, and HeLa cells. For the detection ofCAR and DAF, monoclonal anti-CAR (RmcB) and anti-DAF(BRIC110) antibodies were used (black histograms). A mouse IgG1antibody was used as a negative control (gray histograms).


0.0038 nucleotide substitutions per site per year was estimatedfor enterovirus 70 based on VP1 sequence analyses (91). Theuse of a relaxed-clock model allowed a more confident esti-mate of divergence times in the face of rate variation amonglineages, and it also enabled us to investigate sources of ratevariation. For example, a rapid rate of evolution was observedfor the branch leading to the most recent SVDV isolates,strongly suggesting adaptation after interspecies transmission.

In studies of picornaviruses, the evolution of virus sequenceshas been assayed by serial transfers of large viral populationspertubated by bottleneck events (3, 19, 28), while molecularstructure-function relationships of viral proteins have beenevaluated by site-directed mutagenesis (37, 40, 58, 100, 107).Recent advances in phylogenetics and DNA synthesis tech-niques enable ancestral sequence reconstruction, providing animportant new approach to investigate evolutionary events andthe conservation of structural features (96). Although it isimpossible to claim that reconstructed ancestral sequences arecorrect in all details, since computational models simplify bi-ological processes, the resurrection of molecular ancestorswith a detectable biological activity, such as enzymes and vi-ruses, offers a possibility to verify their structural integrity. The

proper folding of viral capsid proteins, plus intra- and inter-molecular interactions, plays a crucial role for the capsid tomaintain its multifaceted properties, including structural sta-bility to protect the genome and, at the same time, flexibility toenable uncoating after interactions with host cell receptors(103). Despite the advantage of using virus infectivity to verifythe functionality of hypothetical ancestral sequences, few stud-ies of resurrected viral sequences have been presented (21, 51,66, 81).

In the present study, an ancestral state, i.e., the most likelyancestral CVB5 capsid sequence, was inferred from sequencesof contemporary CVB5 isolates by ML ancestral reconstruc-tion (78) and de novo gene synthesis. Although CVB5 mostlikely exists as a swarm of closely related variants within hosts,similar to poliovirus and foot-and-mouth disease virus (20),this has little impact on our evolutionary reconstruction be-cause it essentially involves an interhost evolutionary process.All variants sampled from a patient at a particular point in timewill generally coalesce to a common ancestor prior to the timeof infection. Therefore, whatever variants are sampled fromthe patients, we expect the same common ancestor for virusesobtained from different patients. Inference techniques and

FIG. 9. Receptor preferences, plaque phenotypes, and virus growth kinetics of CVB5-P1anc, CVB5Dwt, 151rom70, and 4378fin88 as well asCVB5-P1anc infectivity in different cell lines. (A) Binding of radiolabeled CVB5 viruses to CHO, CHO-CAR, CHO-DAF, or HeLa cells. Cellswere incubated with [35S]methionine-cysteine-labeled CVB5-P1anc, CVB5Dwt, 151rom70, or 4378fin88 at room temperature for 2 h. Followingthe removal of unbound virions, the cell-associated radioactivity was determined by scintillation counting. Results are presented as means � SEM(n 3). (B) Plaques were visualized at 48 h p.i. by crystal violet staining of HeLa cells. Virus-infected cell lysates were diluted 107 times in orderto distinguish individual plaques. (C) HeLa cells were infected with the indicated viruses at an MOI of 10 TCID50/cell. At various timespostinfection, samples were frozen, and the total yield of infectious virus was quantified by the TCID50 method. Results shown are representativeof three independent experiments. (D) CVB5-P1anc titer determined at time point zero and after complete CPE or 5 days p.i. by endpoint titrationin HeLa cells. Results are presented as means � SEM (n 3).


evolutionary models may have a more important impact onancestral reconstructions. We have used likelihood-basedcodon models, which can accommodate a detailed evolutionaryprocess. For a few highly variable positions, however, this wasnot always consistent with amino acid reconstructions. Furtherresearch that can take into account the uncertainty of recon-structed ancestral sequences is therefore needed.

Characterization of the inferred ancestral P1 sequenceshowed that capsid proteins expressed from the CVB5-P1ancclone assembled into functional infectious virions. In addition,the recombinant CVB5-P1anc virus shared several featureswith present-day clinical isolates, including immunogenicepitopes, preferences for the CAR and DAF receptors, celltropism, plaque morphology, and growth characteristics. Sofar, no evidence of functional impairments has been observed,suggesting that the proposed ancestral capsid proteins fold intonative conformations, which facilitate the assembly of func-tional virions. Sequence analysis of the P1 region after 10passages of CVB5-P1anc and two clinical CVB5 isolates inthree different cell lines demonstrated that a number of sub-stitutions had been introduced into the VP3 �-B knob regionas well as the VP1 C terminus. This is consistent with previousreports describing these regions as neutralizing immunogenicepitopes (9, 48, 68). This analysis also showed that no substi-tutions were introduced at the “ancestral” amino acid positionsin the CVB5-P1anc sequence and, hence, no indication of anevolution toward the sequence of present-day isolates. How-ever, it is important that these 10 passages in cell culturerepresent a simple model system that is lacking a selectivepressure imposed by an immune response, including antibod-ies. Possibly, the evolution that is observed for CVB5-P1anchas more to do with adaptation to the different cell types usedin this experiment. Taken together, the data presented shedlight on the properties of current CVB5 isolates but alsoshowed that the likelihood-based phylogenetic method en-abled an ancestral reconstruction of the four different struc-tural proteins of CVB5. In future studies of CVB5-P1anc,additional phenotypic properties, including virion stability andpathogenicity in animal models, will be investigated.

The ancestral capsid sequence described is a reflection ofeight related CVB5 sequences on which the reconstruction wasbased. As more CVB5 sequences become available, it will be ofinterest to compare the ancestral states of these sequences withCVB5-P1anc. Encouraged by the initial characterizations ofthe CVB5-P1anc virus and with consideration of modern an-cestral reconstruction techniques, including methods for infer-ences of insertion-deletion scenarios (17), the resurrection ofthe ancestors of more distantly related viruses, such as theancestor of all CVBs, offers new possibilities in future evolu-tionary studies and possibly also for the development of newantiviral treatments.

Overall, the present study shows that phylogenetic toolscan be used to construct functional ancestral virions as wellas provide information on conserved structural features im-portant for the function of the virus. In the future, this typeof ancestral reconstruction may also contribute to an in-creased understanding of the molecular evolution of entero-viruses.

ACKNOWLEDGMENTS

We thank J. M. Bergelson for his contribution to the binding studies;K. Edman for valuable discussions; and M. Roivainen, H. Norder, L.Magnius, J. W. McCauley, R. L. Crowell, L. Philipson, and R. Petters-son for providing cells, viruses, and antibodies.

Funding was provided by grants from the Swedish Knowledge Foun-dation, the Graninge Foundation, the Helge Ax:son Johnsons Foun-dation, the Sparbanken Kronan Foundation, and the University ofKalmar. S.H. was supported by NIH grant K22 AI-07927. Figure 5 wasproduced by using the UCSF Chimera package from the ComputerGraphics Laboratory, University of California, San Francisco (sup-ported by NIH grant P41 RR-01081).

REFERENCES

1. Adey, N. B., T. O. Tollefsbol, A. B. Sparks, M. H. Edgell, and C. A.Hutchison III. 1994. Molecular resurrection of an extinct ancestral pro-moter for mouse L1. Proc. Natl. Acad. Sci. U. S. A. 91:1569–1573.

2. Arnberg, N., K. Edlund, A. H. Kidd, and G. Wadell. 2000. Adenovirus type37 uses sialic acid as a cellular receptor. J. Virol. 74:42–48.

3. Baranowski, E., N. Sevilla, N. Verdaguer, C. M. Ruiz-Jarabo, E. Beck, andE. Domingo. 1998. Multiple virulence determinants of foot-and-mouth dis-ease virus in cell culture. J. Virol. 72:6362–6372.

4. Bedard, K. M., and B. L. Semler. 2004. Regulation of picornavirus geneexpression. Microbes Infect. 6:702–713.

5. Bergelson, J. M., J. A. Cunningham, G. Droguett, E. A. Kurt-Jones, A.Krithivas, J. S. Hong, M. S. Horwitz, R. L. Crowell, and R. W. Finberg.1997. Isolation of a common receptor for coxsackie B viruses and adeno-viruses 2 and 5. Science 275:1320–1323.

6. Bergelson, J. M., J. F. Modlin, W. Wieland-Alter, J. A. Cunningham, R. L.Crowell, and R. W. Finberg. 1997. Clinical coxsackievirus B isolates differfrom laboratory strains in their interaction with two cell surface receptors.J. Infect. Dis. 175:697–700.

7. Bergelson, J. M., J. G. Mohanty, R. L. Crowell, N. F. St. John, D. M. Lublin,and R. W. Finberg. 1995. Coxsackievirus B3 adapted to growth in RD cellsbinds to decay-accelerating factor (CD55). J. Virol. 69:1903–1906.

8. Bergelson, J. M., N. St. John, S. Kawaguchi, M. Chan, H. Stubdal, J.Modlin, and R. W. Finberg. 1993. Infection by echoviruses 1 and 8 dependson the alpha 2 subunit of human VLA-2. J. Virol. 67:6847–6852.

9. Borrego, B., E. Carra, J. A. Garcia-Ranea, and E. Brocchi. 2002. Charac-terization of neutralization sites on the circulating variant of swine vesiculardisease virus (SVDV): a new site is shared by SVDV and the relatedcoxsackie B5 virus. J. Gen. Virol. 83:35–44.

10. Bridgham, J. T., E. A. Ortlund, and J. W. Thornton. 2009. An epistaticratchet constrains the direction of glucocorticoid receptor evolution. Na-ture 461:515–519.

11. Brown, B. A., M. S. Oberste, J. P. Alexander, Jr., M. L. Kennett, and M. A.Pallansch. 1999. Molecular epidemiology and evolution of enterovirus 71strains isolated from 1970 to 1998. J. Virol. 73:9969–9975.

12. Bruen, T. C., H. Philippe, and D. Bryant. 2006. A simple and robuststatistical test for detecting the presence of recombination. Genetics 172:2665–2681.

13. Caro, V., S. Guillot, F. Delpeyroux, and R. Crainic. 2001. Molecular strat-egy for ‘serotyping’ of human enteroviruses. J. Gen. Virol. 82:79–91.

14. Casas, I., G. F. Palacios, G. Trallero, D. Cisterna, M. C. Freire, and A.Tenorio. 2001. Molecular characterization of human enteroviruses in clin-ical samples: comparison between VP2, VP1, and RNA polymerase regionsusing RT nested PCR assays and direct sequencing of products. J. Med.Virol. 65:138–148.

15. Chang, B. S., K. Jonsson, M. A. Kazmi, M. J. Donoghue, and T. P. Sakmar.2002. Recreating a functional ancestral archosaur visual pigment. Mol. Biol.Evol. 19:1483–1489.

16. Crowell, R. L., and J. T. Syverton. 1961. The mammalian cell-virus rela-tionship. VI. Sustained infection of HeLa cells by coxsackie B3 virus andeffect on superinfection. J. Exp. Med. 113:419–435.

17. Diallo, A. B., V. Makarenkov, and M. Blanchette. 2010. Ancestors 1.0: aWeb server for ancestral sequence reconstruction. Bioinformatics 26:130–131.

18. Dobos, P. 1976. Use of gum tragacanth overlay, applied at room tempera-ture, in the plaque assay of fish and other animal viruses. J. Clin. Microbiol.3:373–375.

19. Domingo, E., C. Escarmis, E. Baranowski, C. M. Ruiz-Jarabo, E. Carrillo,J. I. Nunez, and F. Sobrino. 2003. Evolution of foot-and-mouth diseasevirus. Virus Res. 91:47–63.

20. Domingo, E., V. Martin, C. Perales, and C. Escarmis. 2008. Coxsackievi-ruses and quasispecies theory: evolution of enteroviruses. Curr. Top. Mi-crobiol. Immunol. 323:3–32.

21. Doria-Rose, N. A., G. H. Learn, A. G. Rodrigo, D. C. Nickle, F. Li, M.Mahalanabis, M. T. Hensel, S. McLaughlin, P. F. Edmonson, D. Monte-fiori, S. W. Barnett, N. L. Haigwood, and J. I. Mullins. 2005. Humanimmunodeficiency virus type 1 subtype B ancestral envelope protein is


functional and elicits neutralizing antibodies in rabbits similar to thoseelicited by a circulating subtype B envelope. J. Virol. 79:11214–11224.

22. Drake, J. W., and J. J. Holland. 1999. Mutation rates among RNA viruses.Proc. Natl. Acad. Sci. U. S. A. 96:13910–13913.

23. Drummond, A. J., S. Y. Ho, M. J. Phillips, and A. Rambaut. 2006. Relaxedphylogenetics and dating with confidence. PLoS Biol. 4:e88.

24. Drummond, A. J., G. K. Nicholls, A. G. Rodrigo, and W. Solomon. 2002.Estimating mutation parameters, population history and genealogy simul-taneously from temporally spaced sequence data. Genetics 161:1307–1320.

25. Drummond, A. J., and A. Rambaut. 2007. BEAST: Bayesian evolutionaryanalysis by sampling trees. BMC Evol. Biol. 7:214.

26. Drummond, A. J., A. Rambaut, B. Shapiro, and O. G. Pybus. 2005. Bayes-ian coalescent inference of past population dynamics from molecular se-quences. Mol. Biol. Evol. 22:1185–1192.

27. Escribano-Romero, E., M. A. Jimenez-Clavero, and V. Ley. 2000. Swinevesicular disease virus. Pathology of the disease and molecular character-istics of the virion. Anim. Health Res. Rev. 1:119–126.

28. Fares, M. A., E. Barrio, N. Becerra, C. Escarmis, E. Domingo, and A. Moya.1998. The foot-and-mouth disease RNA virus as a model in experimentalphylogenetics. Int. Microbiol. 1:311–318.

29. Faria, N. R., M. de Vries, F. J. van Hemert, K. Benschop, and L. van derHoek. 2009. Rooting human parechovirus evolution in time. BMC Evol.Biol. 9:164.

30. Felsenstein, J. 1985. Confidence limits on phylogenies: an approach usingthe bootstrap. Evolution 39:783–791.

31. Fry, E. E., N. J. Knowles, J. W. Newman, G. Wilsden, Z. Rao, A. M. King,and D. I. Stuart. 2003. Crystal structure of swine vesicular disease virus andimplications for host adaptation. J. Virol. 77:5475–5486.

32. Gaucher, E. A., J. M. Thomson, M. F. Burgan, and S. A. Benner. 2003.Inferring the palaeoenvironment of ancient bacteria on the basis of resur-rected proteins. Nature 425:285–288.

33. Gifford, R., and G. Dalldorf. 1951. The morbid anatomy of experimentalcoxsackie virus infection. Am. J. Pathol. 27:1047–1063.

34. Goldman, N., and Z. Yang. 1994. A codon-based model of nucleotidesubstitution for protein-coding DNA sequences. Mol. Biol. Evol. 11:725–736.

35. Guindon, S., and O. Gascuel. 2003. A simple, fast, and accurate algorithmto estimate large phylogenies by maximum likelihood. Syst. Biol. 52:696–704.

36. Hafenstein, S., V. D. Bowman, P. R. Chipman, C. M. B. Kelly, F. Lin, M. E.Medof, and M. G. Rossmann. 2007. Interaction of decay-accelerating factorwith coxsackievirus B3. J. Virol. 81:12927–12935.

37. Hammerle, T., C. U. Hellen, and E. Wimmer. 1991. Site-directed mutagen-esis of the putative catalytic triad of poliovirus 3C proteinase. J. Biol. Chem.266:5412–5416.

38. Harvala, H., and P. Simmonds. 2009. Human parechoviruses: biology,epidemiology and clinical significance. J. Clin. Virol. 45:1–9.

39. He, Y., P. R. Chipman, J. Howitt, C. M. Bator, M. A. Whitt, T. S. Baker,R. J. Kuhn, C. W. Anderson, P. Freimuth, and M. G. Rossmann. 2001.Interaction of coxsackievirus B3 with the full length coxsackievirus-adeno-virus receptor. Nat. Struct. Biol. 8:874–878.

40. Hellen, C. U., M. Facke, H. G. Krausslich, C. K. Lee, and E. Wimmer. 1991.Characterization of poliovirus 2A proteinase by mutational analysis: resi-dues required for autocatalytic activity are essential for induction of cleav-age of eukaryotic initiation factor 4F polypeptide p220. J. Virol. 65:4226–4231.

41. Hierholzer, J. C., and R. A. Killinton. 1996. Quantitation of virus, p. 35–37.In B. W. J. Mahy and R. A. Kangro (ed.), Virology methods manual.Academic Press, London, United Kingdom.

42. Hovi, T., and M. Roivainen. 1993. Peptide antisera targeted to a conservedsequence in poliovirus capsid VP1 cross-react widely with members of thegenus Enterovirus. J. Clin. Microbiol. 31:1083–1087.

43. Hsu, K. H., K. Lonberg-Holm, B. Alstein, and R. L. Crowell. 1988. Amonoclonal antibody specific for the cellular receptor for the group Bcoxsackieviruses. J. Virol. 62:1647–1652.

44. Hughes, A. L. 2004. Phylogeny of the Picornaviridae and differential evo-lutionary divergence of picornavirus proteins. Infect. Genet. Evol. 4:143–152.

45. Ivics, Z., P. B. Hackett, R. H. Plasterk, and Z. Izsvak. 1997. Molecularreconstruction of Sleeping Beauty, a Tc1-like transposon from fish, and itstransposition in human cells. Cell 91:501–510.

46. Jermann, T. M., J. G. Opitz, J. Stackhouse, and S. A. Benner. 1995.Reconstructing the evolutionary history of the artiodactyl ribonuclease su-perfamily. Nature 374:57–59.

47. Jonsson, N., M. Gullberg, and A. M. Lindberg. 2009. Real-time polymerasechain reaction as a rapid and efficient alternative to estimation of picorna-virus titers by tissue culture infectious dose 50% or plaque forming units.Microbiol. Immunol. 53:149–154.

48. Kanno, T., T. Inoue, Y. Wang, A. Sarai, and S. Yamaguchi. 1995. Identifi-cation of the location of antigenic sites of swine vesicular disease virus withneutralization-resistant mutants. J. Gen. Virol. 76(Pt. 12):3099–3106.

49. Konno, A., T. Ogawa, T. Shirai, and K. Muramoto. 2007. Reconstruction of

a probable ancestral form of conger eel galectins revealed their rapidadaptive evolution process for specific carbohydrate recognition. Mol. Biol.Evol. 24:2504–2514.

50. Kopecka, H., B. Brown, and M. Pallansch. 1995. Genotypic variation incoxsackievirus B5 isolates from three different outbreaks in the UnitedStates. Virus Res. 38:125–136.

51. Kothe, D. L., Y. Li, J. M. Decker, F. Bibollet-Ruche, K. P. Zammit, M. G.Salazar, Y. Chen, Z. Weng, E. A. Weaver, F. Gao, B. F. Haynes, G. M. Shaw,B. T. Korber, and B. H. Hahn. 2006. Ancestral and consensus envelopeimmunogens for HIV-1 subtype C. Virology 352:438–449.

52. Kuang, D., Y. Yao, D. Maclean, M. Wang, D. R. Hampson, and B. S. Chang.2006. Ancestral reconstruction of the ligand-binding pocket of family C Gprotein-coupled receptors. Proc. Natl. Acad. Sci. U. S. A. 103:14050–14055.

53. Lanave, C., G. Preparata, C. Saccone, and G. Serio. 1984. A new methodfor calculating evolutionary substitution rates. J. Mol. Evol. 20:86–93.

54. Lin, F., and R. P. Kitching. 2000. Swine vesicular disease: an overview. Vet.J. 160:192–201.

55. Lindberg, A. M., and C. Polacek. 2000. Molecular analysis of the prototypecoxsackievirus B5 genome. Arch. Virol. 145:205–221.

56. Lindberg, A. M., C. Polacek, and S. Johansson. 1997. Amplification andcloning of complete enterovirus genomes by long distance PCR. J. Virol.Methods 65:191–199.

57. Lukashev, A. N. 2005. Role of recombination in evolution of enteroviruses.Rev. Med. Virol. 15:157–167.

58. Marc, D., G. Drugeon, A. L. Haenni, M. Girard, and S. van der Werf. 1989.Role of myristoylation of poliovirus capsid protein VP4 as determined bysite-directed mutagenesis of its N-terminal sequence. EMBO J. 8:2661–2668.

59. Martino, T. A., M. Petric, M. Brown, K. Aitken, C. J. Gauntt, C. D.Richardson, L. H. Chow, and P. P. Liu. 1998. Cardiovirulent coxsackievi-ruses and the decay-accelerating factor (CD55) receptor. Virology 244:302–314.

60. Martino, T. A., M. Petric, H. Weingartl, J. M. Bergelson, M. A. Opavsky,C. D. Richardson, J. F. Modlin, R. W. Finberg, K. C. Kain, N. Willis, C. J.Gauntt, and P. P. Liu. 2000. The coxsackie-adenovirus receptor (CAR) isused by reference strains and clinical isolates representing all six serotypesof coxsackievirus group B and by swine vesicular disease virus. Virology271:99–108.

61. Melnick, J. L., E. W. Shaw, and E. C. Curnen. 1949. A virus isolated frompatients diagnosed as non-paralytic poliomyelitis or aseptic meningitis.Proc. Soc. Exp. Biol. Med. 71:344–349.

62. Milstone, A. M., J. Petrella, M. D. Sanchez, M. Mahmud, J. C. Whitbeck,and J. M. Bergelson. 2005. Interaction with coxsackievirus and adenovirusreceptor, but not with decay-accelerating factor (DAF), induces A-particleformation in a DAF-binding coxsackievirus B3 isolate. J. Virol. 79:655–660.

63. Minor, P. 1985. Growth, assay and purification of picornaviruses, p. 25–41.In B. Mahy (ed.), Virology, a practical approach. IRL Press, Oxford, UnitedKingdom.

64. Muckelbauer, J. K., M. Kremer, I. Minor, G. Diana, F. J. Dutko, J.Groarke, D. C. Pevear, and M. G. Rossmann. 1995. The structure ofcoxsackievirus B3 at 3.5 Å resolution. Structure 3:653–667.

65. Mulders, M. N., M. Salminen, N. Kalkkinen, and T. Hovi. 2000. Molecularepidemiology of coxsackievirus B4 and disclosure of the correct VP1/2A(pro) cleavage site: evidence for high genomic diversity and long-termendemicity of distinct genotypes. J. Gen. Virol. 81:803–812.

66. Mullins, J. I., D. C. Nickle, L. Heath, A. G. Rodrigo, and G. H. Learn. 2004.Immunogen sequence: the fourth tier of AIDS vaccine design. Expert Rev.Vaccines 3:S151–S159.

67. Nardelli, L., E. Lodetti, G. L. Gualandi, R. Burrows, D. Goodridge, F.Brown, and B. Cartwright. 1968. A foot and mouth disease syndrome inpigs caused by an enterovirus. Nature 219:1275–1276.

68. Nijhar, S. K., D. K. Mackay, E. Brocchi, N. P. Ferris, R. P. Kitching, andN. J. Knowles. 1999. Identification of neutralizing epitopes on a Europeanstrain of swine vesicular disease virus. J. Gen. Virol. 80(Pt. 2):277–282.

69. Norder, H., L. Bjerregaard, and L. O. Magnius. 2001. Homotypic echovi-ruses share aminoterminal VP1 sequence homology applicable for typing.J. Med. Virol. 63:35–44.

70. Oberste, M. S., K. Maher, M. L. Kennett, J. J. Campbell, M. S. Carpenter,D. Schnurr, and M. A. Pallansch. 1999. Molecular epidemiology and ge-netic diversity of echovirus type 30 (E30): genotypes correlate with tempo-ral dynamics of E30 isolation. J. Clin. Microbiol. 37:3928–3933.

71. Oberste, M. S., K. Maher, D. R. Kilpatrick, M. R. Flemister, B. A. Brown,and M. A. Pallansch. 1999. Typing of human enteroviruses by partialsequencing of VP1. J. Clin. Microbiol. 37:1288–1293.

72. Oberste, M. S., K. Maher, D. R. Kilpatrick, and M. A. Pallansch. 1999.Molecular evolution of the human enteroviruses: correlation of serotypewith VP1 sequence and application to picornavirus classification. J. Virol.73:1941–1948.

73. Pauling, L., and E. Zuckerkandl. 1963. Chemical paleogenetics: molecularrestoration studies of extinct forms of life. Acta Chem. Scand. 17:S9–S16.

74. Percy, N., W. S. Barclay, M. Sullivan, and J. W. Almond. 1992. A poliovirusreplicon containing the chloramphenicol acetyltransferase gene can be used


to study the replication and encapsidation of poliovirus RNA. J. Virol.66:5040–5046.

75. Pettersen, E. F., T. D. Goddard, C. C. Huang, G. S. Couch, D. M. Green-blatt, E. C. Meng, and T. E. Ferrin. 2004. UCSF Chimera—a visualizationsystem for exploratory research and analysis. J. Comput. Chem. 25:1605–1612.

76. Pfeiffer, J. K., and K. Kirkegaard. 2005. Increased fidelity reduces poliovi-rus fitness and virulence under selective pressure in mice. PLoS Pathog.1:e11.

77. Polacek, C., J. O. Ekstrom, A. Lundgren, and A. M. Lindberg. 2005. Cyto-lytic replication of coxsackievirus B2 in CAR-deficient rhabdomyosarcomacells. Virus Res. 113:107–115.

78. Pupko, T., I. Pe’er, R. Shamir, and D. Graur. 2000. A fast algorithm forjoint reconstruction of ancestral amino acid sequences. Mol. Biol. Evol.17:890–896.

79. Reagan, K. J., B. Goldberg, and R. L. Crowell. 1984. Altered receptorspecificity of coxsackievirus B3 after growth in rhabdomyosarcoma cells.J. Virol. 49:635–640.

80. Rezig, D., A. Ben Yahia, H. Ben Abdallah, O. Bahri, and H. Triki. 2004.Molecular characterization of coxsackievirus B5 isolates. J. Med. Virol.72:268–274.

81. Rolland, M., M. A. Jensen, D. C. Nickle, J. Yan, G. H. Learn, L. Heath, D.Weiner, and J. I. Mullins. 2007. Reconstruction and function of ancestralcenter-of-tree human immunodeficiency virus type 1 proteins. J. Virol.81:8507–8514.

82. Rossmann, M. G. 2002. Picornavirus structure overview, p. 27–38. In B. L.Semler and E. Wimmer (ed.), Molecular biology of picornaviruses. ASMPress, Washington, DC.

83. Sanner, M. F., A. J. Olson, and J. C. Spehner. 1996. Reduced surface: anefficient way to compute molecular surfaces. Biopolymers 38:305–320.

84. Selinka, H. C., A. Wolde, A. Pasch, K. Klingel, J. J. Schnorr, J. H. Kupper,A. M. Lindberg, and R. Kandolf. 2002. Comparative analysis of two cox-sackievirus B3 strains: putative influence of virus-receptor interactions onpathogenesis. J. Med. Virol. 67:224–233.

85. Shafren, D. R., R. C. Bates, M. V. Agrez, R. L. Herd, G. F. Burns, and R. D.Barry. 1995. Coxsackieviruses B1, B3, and B5 use decay accelerating factoras a receptor for cell attachment. J. Virol. 69:3873–3877.

86. Shapiro, B., A. Rambaut, and A. J. Drummond. 2006. Choosing appropriatesubstitution models for the phylogenetic analysis of protein-coding se-quences. Mol. Biol. Evol. 23:7–9.

87. Smit, A. F., and A. D. Riggs. 1996. Tiggers and DNA transposon fossils inthe human genome. Proc. Natl. Acad. Sci. U. S. A. 93:1443–1448.

88. Stackhouse, J., S. R. Presnell, G. M. McGeehan, K. P. Nambiar, and S. A.Benner. 1990. The ribonuclease from an extinct bovid ruminant. FEBS Lett.262:104–106.

89. Stanway, G., F. Brown, P. Christian, T. Hovi, T. Hyypia, A. M. Q. King,N. J. Knowles, S. M. Lemon, P. D. Minor, M. A. Pallansch, A. C. Palmen-berg, and T. Skern. 2005. Family Picornaviridae, p. 757–778. In C. M.Fauquet, M. A. Mayo, J. Maniloff, U. Desselberger, and L. A. Ball (ed.),Virus taxonomy: eighth report of the International Committee on Taxon-omy of Viruses. Elsevier Academic Press, London, United Kingdom.

90. Strimmer, K., and A. von Haeseler. 1997. Likelihood-mapping: a simplemethod to visualize phylogenetic content of a sequence alignment. Proc.Natl. Acad. Sci. U. S. A. 94:6815–6819.

91. Takeda, N., M. Tanimura, and K. Miyamura. 1994. Molecular evolution ofthe major capsid protein VP1 of enterovirus 70. J. Virol. 68:854–862.

92. Tamura, K., J. Dudley, M. Nei, and S. Kumar. 2007. MEGA4: MolecularEvolutionary Genetics Analysis (MEGA) software version 4.0. Mol. Biol.Evol. 24:1596–1599.

93. Thoelen, I., P. Lemey, I. Van Der Donck, K. Beuselinck, A. M. Lindberg,and M. Van Ranst. 2003. Molecular typing and epidemiology of enterovi-ruses identified from an outbreak of aseptic meningitis in Belgium duringthe summer of 2000. J. Med. Virol. 70:420–429.

94. Thompson, J. D., D. G. Higgins, and T. J. Gibson. 1994. CLUSTAL W:improving the sensitivity of progressive multiple sequence alignmentthrough sequence weighting, position-specific gap penalties and weightmatrix choice. Nucleic Acids Res. 22:4673–4680.

95. Thomson, J. M., E. A. Gaucher, M. F. Burgan, D. W. De Kee, T. Li, J. P.Aris, and S. A. Benner. 2005. Resurrecting ancestral alcohol dehydroge-nases from yeast. Nat. Genet. 37:630–635.

96. Thornton, J. W. 2004. Resurrecting ancient genes: experimental analysis ofextinct molecules. Nat. Rev. Genet. 5:366–375.

97. Thornton, J. W., E. Need, and D. Crews. 2003. Resurrecting the ancestralsteroid receptor: ancient origin of estrogen signaling. Science 301:1714–1717.

98. Tomko, R. P., R. Xu, and L. Philipson. 1997. HCAR and MCAR: thehuman and mouse cellular receptors for subgroup C adenoviruses andgroup B coxsackieviruses. Proc. Natl. Acad. Sci. U. S. A. 94:3352–3356.

99. Tracy, S., and C. Gauntt. 2008. Group B coxsackievirus virulence. Curr.Top. Microbiol. Immunol. 323:49–63.

100. Vance, L. M., N. Moscufo, M. Chow, and B. A. Heinz. 1997. Poliovirus 2Cregion functions during encapsidation of viral RNA. J. Virol. 71:8759–8765.

101. Verdaguer, N., M. A. Jimenez-Clavero, I. Fita, and V. Ley. 2003. Structureof swine vesicular disease virus: mapping of changes occurring during ad-aptation of human coxsackie B5 virus to infect swine. J. Virol. 77:9780–9789.

102. Vignuzzi, M., J. K. Stone, J. J. Arnold, C. E. Cameron, and R. Andino. 2006.Quasispecies diversity determines pathogenesis through cooperative inter-actions in a viral population. Nature 439:344–348.

103. Whitton, J. L., C. T. Cornell, and R. Feuer. 2005. Host and virus determi-nants of picornavirus pathogenesis and tropism. Nat. Rev. Microbiol.3:765–776.

104. Xia, X., Z. Xie, M. Salemi, L. Chen, and Y. Wang. 2003. An index ofsubstitution saturation and its application. Mol. Phylogenet. Evol. 26:1–7.

105. Yang, Z. 1997. PAML: a program package for phylogenetic analysis bymaximum likelihood. Comput. Appl. Biosci. 13:555–556.

106. Yang, Z., S. Kumar, and M. Nei. 1995. A new method of inference ofancestral nucleotide and amino acid sequences. Genetics 141:1641–1650.

107. Yu, S. F., and R. E. Lloyd. 1991. Identification of essential amino acidresidues in the functional activity of poliovirus 2A protease. Virology 182:615–625.

108. Zhang, G., D. T. Haydon, N. J. Knowles, and J. W. McCauley. 1999.Molecular evolution of swine vesicular disease virus. J. Gen. Virol. 80(Pt.3):639–651.

109. Zhang, G., G. Wilsden, N. J. Knowles, and J. W. McCauley. 1993. Completenucleotide sequence of a coxsackie B5 virus and its relationship to swinevesicular disease virus. J. Gen. Virol. 74(Pt. 5):845–853.

110. Zhang, J., and H. F. Rosenberg. 2002. Complementary advantageous sub-stitutions in the evolution of an antiviral RNase of higher primates. Proc.Natl. Acad. Sci. U. S. A. 99:5486–5491.


characterization of a putative ancestor of coxsackievirus b5bayesian evolutionary analysis. we...

Documents