dating the origin of the genus flavivirus in the light of

Dating the origin of the genus Flavivirus in the lightof Beringian biogeography

John H.-O. Pettersson and Omar Fiz-Palacios

Correspondence

John H.-O. Pettersson

[email protected] or

[email protected]

Received 28 February 2014

Accepted 4 June 2014

Department of Systematic Biology, Evolutionary Biology Centre, Uppsala University,Uppsala, Sweden

The genus Flavivirus includes some of the most important human viral pathogens, and its

members are found in all parts of the populated world. The temporal origin of diversification of the

genus has long been debated due to the inherent problems with dating deep RNA virus evolution.

A generally accepted hypothesis suggests that Flavivirus emerged within the last 10 000 years.

However, it has been argued that the tick-borne Powassan flavivirus was introduced into North

America some time between the opening and closing of the Beringian land bridge that connected

Asia and North America 15 000–11 000 years ago, indicating an even older origin for Flavivirus.

To determine the temporal origin of Flavivirus, we performed Bayesian relaxed molecular clock

dating on a dataset with high coverage of the presently available Flavivirus diversity by combining

tip date calibrations and internal node calibration, based on the Powassan virus and Beringian

land bridge biogeographical event. Our analysis suggested that Flavivirus originated ~85 000

(64 000–110 000) or 120 000 (87 000–159 000) years ago, depending on the circumscription

of the genus. This is significantly older than estimated previously. In light of our results, we

propose that it is likely that modern humans came in contact with several members of the genus

Flavivirus much earlier than suggested previously, and that it is possible that the spread of several

flaviviruses coincided with, and was facilitated by, the migration and population expansion of

modern humans out of Africa.

INTRODUCTION

The genus Flavivirus, within the family Flaviviridae, currentlyconsists of .70 virus species distributed all over the globe(Gould et al., 2003; Gubler et al., 2007; Lindenbach et al.,2007). It includes numerous viruses of major human healthconcern, such as Dengue virus (DENV), Japanese enceph-alitis virus, West Nile virus and the type species forflaviviruses, yellow fever virus (flavus: ‘yellow’), giving thefamily and the genus its name (Gould & Solomon, 2008;Mackenzie et al., 2004). The genome of flaviviruses consistsof a positive-sense ssRNA molecule of ~11 kbp. One singleORF encodes three structural proteins (capsid, pre-mem-brane and envelope) and seven non-structural proteins (NS1,NS2A, NS2B, NS3, NS4A, NS4B and NS5) flanked byuntranslated regions (Chambers et al., 1990; Lindenbach &Rice, 2003).

The genus Flavivirus has been divided into four maingroups based on ecological characteristics, molecular phylo-genetic analyses, vector specificity and virus performance inhost cells (Gaunt et al., 2001; Gould et al., 2001; Kuno,2007). Most of the recognized flaviviruses belong to themosquito-borne flaviviruses (MBFs) that are commonly, but

not exclusively, vectored by either Aedes or Culex mosqui-toes. The second group, tick-borne flaviviruses (TBFs), theonly monophyletic group of Flavivirus, are vectored byixodid ticks, and are further subdivided into a mammal anda seabird group based on their host specificity (Grard et al.,2007; Gritsun et al., 2003). The no-known vector flaviviruses(NKVFs) are associated with either bats or rodents andinfect vertebrates without an apparent arthropod vectortransmitting the viruses (Porterfield, 1980). The majority ofthe known flaviviruses are zoonotic, i.e. pathogenic virusesthat can be transmitted between humans and other animals.However, the fourth group, the insect-specific flaviviruses(ISFs; flaviviruses that are only capable of replicating ininsect cells) (Kuno, 2007), is most likely an undersampledand very diverse group (Cook et al., 2012).

The idea of a molecular clock has been used to addressmany hypotheses in the study of emerging viral diseases,especially for diseases caused by RNA viruses (Bromham &Penny, 2003), e.g. to reject the hypothesis of a potentialspread of human immunodeficiency virus (HIV) in the1950s through a contaminated polio vaccine (Korber et al.,2000). However, dating the origin of viruses is a complexand challenging task. For viruses with high rates of evolu-tion, the original phylogenetic signal is difficult to deduceeven with complex evolutionary models because the signal

Three supplementary figures and three supplementary tables areavailable with the online version of this paper.

Journal of General Virology (2014), 95, 1969–1982 DOI 10.1099/vir.0.065227-0

065227 G 2014 The Authors Printed in Great Britain 1969

diminishes with time due to repeated substitutions at thesame site (Holmes, 2003a). The extremely high substitutionrate observed for RNA viruses is mostly due to their error-prone replication and repair machinery (Bromham &Penny, 2003; Duffy et al., 2008). Unequal substitution ratesamong lineages of the same genus further adds to thecomplexity of accurately estimating the time to the mostrecent common ancestor (tMRCA) (Duffy et al., 2008;Holmes, 2003a; Sanjuan, 2012).

Reconstruction of divergence times requires a temporalreference to convert branch lengths of a phylogenetic treeinto time. This temporal reference is usually in the form ofa fossil or biogeographical event (i.e. internal node calibra-tion), as in most eukaryote studies, or isolation dates (i.e.tip date calibration), as with bacteria and virus studies.From the biogeographical events, fossils or isolation dates,the ages of internal nodes can be estimated. The matchbetween virus and host phylogenies has led to the sugges-tion that some virus lineages originated millions of yearsago. However, this has been strongly rejected by molecularclock studies indicating a virus origin of only a few thousandyears ago (Holmes, 2003a; Worobey et al., 2010). Thiscontroversy has sparked a debate regarding the validity andutility of molecular clock reconstructions using tip datesversus internal (deep) calibration to study and infer thetemporal scale of virus evolution (Sharp & Simmonds,2011).

Previous studies based on molecular clocks have suggestedthat the Flavivirus clade originated ~10 000 years ago inAfrica (Gould et al., 2001) from a non-vectored mam-malian virus ancestor (Gould et al., 2003), followed by theradiation of TBFs and MBFs during the last 5000 and 3000years, respectively (Zanotto et al., 1996), and ISFs ~3000years ago (Crochu et al., 2004). Other studies have alsodated small clades of individual flaviviruses using tip datesas calibration points. These estimates are broadly congruentwith the notion of a Flavivirus origin within the last 10 000years (e.g. Dunham & Holmes 2007; May et al., 2011; Panet al., 2011; Twiddy et al., 2003). However, recent analyseson TBFs, based on complete genomes analysed with arelaxed molecular clock and Bayesian methods, have esti-mated the age for the TBF clade to be at least 16 000 years(Heinze et al., 2012), indicating that the age for theFlavivirus genus as a whole should be significantly olderthan suggested previously.

Within the TBFs, a close relationship between the bio-geography of the Beringian land bridge, the historicallandmass that connected north-east Siberia with Alaska andnorth-west Canada (Hulten, 1937), and the evolution of thePowassan virus (POWV) clade, the only TBFs present inNorth America, has been pointed out recently (Ebel, 2010;Heinze et al., 2012). Geological studies estimate that the landbridge was open for mammal land migration before the endof the last glaciation between 15 000 and 11 000 years ago(Elias, 2001; Elias et al., 1996; Kelly, 2003; Mandryk et al.,2001). The absence of tick-borne encephalitis virus (TBEV)

and Russian POWV lineages in North America is consistentwith the idea that the colonization of POWV into NorthAmerica happened under a single event by a land route, alsosupported by the lack of evidence for continuous movementof POWV and TBEV between Russia and North America byeither seabirds or mosquitoes (Heinze et al., 2012).Therefore, it is unlikely that POWV emerged in NorthAmerica before the land bridge became accessible again afterthe last glacial maximum. Consequentially, it is improbablethat POWV emerged after the Bering Strait was formed. Themost probable explanation, given the current knowledge andpresent distribution of the POWV lineages, is that during thecourse of TBF evolution POWV diverged from other TBFlineages in a single introduction event by the land routeduring the presence of the Beringian land bridge 15 000–11 000 years ago (Heinze et al., 2012).

POWV could have been introduced into North Americaby either humans or other mammals that colonized theAmericas during the existence of the Beringian landbridge via one of two routes: the interior route or coastalroute. (i) The interior route became accessible when theprocess of deglaciation created a corridor between theLaurentide and Cordilleran glaciers. The corridor did notexist during the last glacial maximum. The corridorstarted to open ~15 000 years ago (Dixon, 2013; Dyke,2004), and became habitable for humans and othermammals ~13 500 years ago (Dixon, 2013). (ii) Towardsthe end of the last glacial maximum, the glaciers border-ing to the south-west coast of Alaska through westernCanada started to melt. Around 16 000 years ago, theglaciers had receded to the extent that the coastal habitatscould support human populations (Dixon, 2013 andreferences therein). The coastal route is also likely to bethe route by which humans first entered the Americas(Achilli et al., 2013; Bodner et al., 2012; Fagundes et al.,2008; Goebel et al., 2008; Schurr, 2004).

Presently, there are coding nucleotide genomes available from.70 unique strains of the genus Flavivirus, including virusesrecognized by the International Committee on Taxonomy ofViruses (http://www.ictvonline.org/virusTaxonomy.asp?version=2013). Here, we combine this information together withrelaxed molecular clock methods implementing a Bayesianapproach, again to estimate and shed some light on diver-gence times of the genus Flavivirus and groups within. Aidedby biogeographical calibration, we report the first study, tothe best of our knowledge, estimating the age and rates ofsubstitution of the genus Flavivirus as a whole, including itsmajor groups, using complete coding nucleotide genomes.We explore multiple internal calibration points usingBayesian methods in combination with a relaxed molecularclock to reconstruct divergence times. We compare ourresults with previous studies reconstructing divergencetimes for different clades within the genus Flavivirus. Wealso discuss the incongruence between virus-dating studiesusing molecular clocks and RNA virus evolution. Finally, wepropose a temporal and biogeographical scenario for theevolution of the genus Flavivirus.

J. H.-O Pettersson and O. Fiz-Palacios

1970 Journal of General Virology 95

RESULTS AND DISCUSSION

Phylogenetic reconstruction of the genusFlavivirus

The phylogeny of Flavivirus and its major groups wasinferred using a genus-wide sampling approach includ-ing 86 complete coding flavivirus genomes with knownisolation dates (Table 1) analysed by both Bayesian andmaximum-likelihood methodologies. The overall treetopology from all of the different analyses was in agree-ment with previous published phylogenies based on theNS3 gene (Billoir et al., 2000; Cook & Holmes, 2006;Grard et al., 2007), multiple genes (Medeiros et al., 2007)and complete genomes (Cook & Holmes, 2006; Cook et al.,2012; Grard et al., 2007, 2010; Kuno & Chang, 2006; Loboet al., 2009). Our results were consistent with what iscommonly referred to as an NS3-like topology (Fig. 1).

Two datasets were analysed, a nucleotide and an aminoacid sequence alignment, using Bayesian and maximum-likelihood approaches. In our Bayesian phylogenetic anal-ysis, the nucleotide and amino acid 50 % majority ruleconsensus trees from MrBayes (Figs S1 and S2, available inthe online Supplementary Material) and the maximumclade credibility tree from BEAST (Figs 1 and S3), all rootedwith bovine viral diarrhea virus 1 (BVDV-1), producedtopologies with strong support for the focal nodes (A–O).The Bayesian trees were also supported by the topology ofthe maximum-likelihood trees, although the sister rela-tionships between Tamana bat virus (TABV) and all othergroups were unresolved in the maximum-likelihood trees(data not shown, available upon request).

There was maximum posterior probability (1.0 pp) for thefour major clades roughly corresponding to the conven-tional grouping of Flavivirus (Gould et al., 2001, 2003);the TBF clade (node M), the NKVF clade (NKVFa, nodeN), a MBF-dominated group (MBFdom, node F) and theISF clade (ISFa, node E), sister to the other three groups(Fig. 1). The MBFdom group (node F) includes (i) theAedes MBF clade (MBFAedes, node H), (ii) a NKVF clade(NKVFb), sister to the MBFAedes clade, (iii) a grade oflikely insect-specific flaviviruses consisting of Chaoyangvirus, Donggang virus, Lammi virus and Nounane virus(nodes I and J) (Huhtamo et al., 2009; Kolodziejek et al.,2013; Lee et al., 2013) and (iv) the Culex MBF clade(MBFCulex, node K). The present study supports the factthat the majority of the members of the genus Flavivirusshow a very strong pattern of host–vector association(Gould et al., 2001, 2003), as all groups showed distinctclustering depending on vector association. It is possiblethat the positioning of the NKVFb clade is a consequenceof secondary loss of vector capability (Gould et al., 2001,2003). Likewise, it is possible that the grade of ISFs (nodesI and J) within the MBFdom group is also due to secondaryloss of vector capability.

The main disagreement between our findings and otherstudies relates to the position of TABV. TABV has

previously been suggested to be the sister group to theISFa clade (node E) based on complete coding amino acidsequence analysis (Cook et al., 2012). However, in thepresent study, TABV appears to have diverged before theISFa clade (node B, 1.0 pp, based on Bayesian nucleotideand amino acid phylogenetic analyses), which instead issister to the TBFs, NKVFs and MBFs (Figs 1, S1 and S2).Our results are consistent with the topology of previouscomplete genome and gene-based phylogenetic studies (deLamballerie et al., 2002; Hoshino et al., 2009; Lobo et al.,2009).

Divergence times of the genus Flavivirus

Viruses, and RNA viruses in particular, provide an excel-lent opportunity to study evolutionary change because oftheir relatively high rate of substitution that allows forevolution to be observed within human timescales (Bromham& Penny, 2003; Drummond et al., 2003; Duffy et al., 2008;Holmes, 2003a). Many studies have used molecular clocks todate recent divergences (hundred-year scale; see: Kumar et al.,2010; Mehla et al., 2009; Mohammed et al., 2011; Patil et al.,2011; Ramırez et al., 2010; Weidmann et al., 2013), but herewe show the utility of using internal calibration that allows forthe reconstruction of deeper divergence times.

The divergence times under different calibration schemesare summarized in Tables 2 and S1. All BEAST runs wereperformed with an uncorrelated log-normal relaxedmolecular clock, as the null hypothesis of a strict globalsubstitution rate was rejected by the maximum-likelihoodmolecular clock test (data not shown).

Divergence times of the genus Flavivirus:calibration schemes compared

To examine the temporal origin of Flavivirus, we performedseveral analyses in BEAST as outlined in Methods. Using onlytip date calibration resulted in 12–14 % older mean diver-gence times for nodes A–O than using the internal Beringiancalibration (node O calibrated to 15 000–11 000 years ago)together with tip dates. For example, the age of the root(node A) varied between 265 000 [95 % highest posteriordensity (HPD): 25 600–2 768 000] and 230 500 (156 100–322 700) years ago, respectively (Fig. 1, Table 2). When usingTBF internal calibration (node M, calibrated to 16 100–44 929 years ago) based on the results from Heinze et al.(2012) using only tip dates as calibration, 17–18 % youngermean divergence times were recovered for nodes A–O com-pared with using tip dates only (Table 2). Both Beringianand TBF calibration recovered similar mean divergencetimes, although TBF calibration recovered 4–5 % youngermean estimates for nodes A–O than Beringian calibration(Table 2). Also, allowing for a wider range of the Beringiancalibration, i.e. 16 000–10 000 years, following Dixon (2013),TBF recovered similar ages for the split of POWV and itssister (node O) at 12 500 (10 000–15 600) years ago and forthe root (node A) at 228 500 (142 900–332 500) years ago as

Origin of the genus Flavivirus

http://vir.sgmjournals.org 1971

Table 1. Flaviviruses used in this study

Vector groups include insect-specific (IS), mosquito-borne (MB), no-known vector (NKV) and tick-borne (TB) flaviviruses.

Virus Virus abbreviation Strain Source of isolate Vector

group

Year GenBank

accession no.

Aedes virus AEFV AEFV_SPFLD_MO_2011_MP6 Aedes albopictus IS 2011 KC181923.1

Alfuy virus ALFV MRM3929 Centropus phasianius MB 1966 AY898809.1

Alkhurma virus AHFV 1176 Homo sapiens TB 1995 AF331718.1

Apoi virus APOIV ApMAR-Kitaoka Apodemus spp. NKV 1954 AF160193.1

Bagaza virus BAGV Spain H/2010 Alectoris rufa MB 2010 HQ644143.1

Bagaza virus BAGV DakAr-B209 Mixed Culex MB 1966 AY632545.2

Baiyangdian virus BYDV BYD-1 Duck MB 2010 JF312912.1

Banzi virus BANV SAH-336 Homo sapiens, Culex, Mansonia

africana

MB 1956 DQ859056.1

Bouboui virus BOUV DAK-AR-B490 Anopheles paludis MB 1967 DQ859057.1

Bovine viral diarrhoea virus 1 BVDV-1 NADL 1963 NC_001461.1

Bussuquara virus BSQV BeAn-4073 Alouatta beelzebul MB 1956 NC_009026.2

Cell fusing agent virus CFAV Aedes spp., Aedes egypti IS 1975 NC_001564.1

Chaoyang virus CHAOV ROK144 Aedes vexans nipponii IS 2003 JQ068102.1

Culex flavivirus CxFV Tokyo Culex pipiens IS 2003 AB262759.2

Deer tick virus DTV CTB30-CT95 Ixodes dammini TB 1995 NC_003218.1

Dengue virus type 1 DENV-1 45AZ5 Homo sapiens MB 1974 U88536.1

Dengue virus type 2 DENV-2 New-Guinea-C Homo sapiens MB 1944 AF038403.1

Dengue virus type 3 DENV-3 H87 Homo sapiens MB 1956 M93130.1

Dengue virus type 4 DENV-4 H241 Homo sapiens MB 1956 M14931.2

Donggang virus DGV DG0909 Aedes spp. IS 2009 NC_016997.1

Edge Hill virus EHV YMP-48 Aedes vigilax MB 1969 DQ859060.1

Entebbe bat virus ENTV UgIL-30 Tadarida limbata NKV 1957 DQ837641.1

Gadgets Gully virus GGYV CSIRO122 Ixodes uriae TB 1975 DQ235145.1

Greek goat encephalitis virus GGEV Vergina Capra aegagrus hircus TB 1969 DQ235153.1

Hanko virus HANKV Mosquitoes IS 2005 JQ268258.1

Iguape virus IGUV SPAn-71686 Sentinel mouse MB 1979 AY632538.4

Ilheus virus ILHV Original Aedes spp. and Psorophora spp. MB 1944 AY632539.4

Japanese encephalitis virus JEV JaOArS982 Mosquito? MB 1982 M18370.1

Jugra virus JUGV P9-314 Aedes spp. MB 1968 DQ859066.1

Kadam virus KADV AMP6640 Rhipicephalus pravus TB 1967 DQ235146.1

Kamiti River virus KRV SR-82 Aedes macintoshi IS 1999 NC_005064.1

Karshi virus KSIV LEIV_2247 Ornithodoros papillipes TB 1972 NC_006947.1

Kedougou virus KEDV DakAar-D1470 Aedes minutus MB 1972 AY632540.2

Kokobera virus KOKV AusMRM-32 Culex annulirostris MB 1960 NC_009029.2

Koutango virus KOUV DakAr-D-5443 Tatera kempi MB 1968 EU082200.1

Kunjin virus KUNV MRM61C Culex annulirostris MB 1960 D00246.1

Kyasanur forest disease virus KFDV G11338 Haemaphysalis spinigera TB 1957 JF416959.1

Lammi virus LAMV Mosquitoes IS 2004 FJ606789.1

J.H.-O

Pettersson

andO

.Fiz-P

alacios

19

72

Journal

of

General

Viro

logy

95

Table 1. cont.

Virus Virus abbreviation Strain Source of isolate Vector

group

Year GenBank

accession no.

Langat virus LGTV TP21 Ixodes granulatus Supino TB 1956 AF253419.1

Louping ill virus LIV 369/T2 Ixodes ricinus TB 1963 NC_001809.1

Meaban virus MEAV T70 Ornithodoros maritimus TB 1981 DQ235144.1

Modoc virus MODV M544 Peromyscus maniculatus NKV 1958 AJ242984.1

Montana myotis leukoencephalitis virus MMLV ATCC-VR-537 Myotis lucifugus NKV 1958 AJ299445.1

Murray Valley encephalitis MVEV MVE-1-51 Homo sapiens MB 1956 AF161266.1

Nakiwogo virus NAKV Uganda08 Mansonia africana nigerrima H4A1 IS 2008 GQ165809.2

New Mapoon Virus NMV CY1014 Culex annulirostris MB 1998 KC788512.1

Nounane virus NOUV Uranotaenia mashonaensis IS 2004 FJ711167.1

Ntaya virus NTAV IPDIA Mosquitoes MB 1966 JX236040.1

Omsk hemorrhagic fever virus OHFV Kubrin Homo sapiens TB 1947 AY438626.1

Omsk hemorrhagic fever virus OHFV Guriev Homo sapiens TB 1947 AB507800.1

Palm Creek Virus PCV 56 Coquillettidia xanthogaster IS 2010 KC505248.1

Potiskum virus POTV IBAN-10069 Cricetomys gambianus MB 1989 DQ859067.1

Powassan virus POWV LB Homo sapiens TB 1958 L06436.1

Quang Binh virus QBV VN180 Culex tritaeniorhynchus IS 2002 NC_012671.1

Rio Bravo virus RBV RiMAR Tadarida brasiliensis mexicana NKV 1954 AF144692.1

Rocio virus ROCV SPH-34675 Homo sapiens MB 1975 AY632542.4

Royal Farm virus RFV Eg-Art-371 Argas hermanni TB 1968 DQ235149.1

Saboya virus SABV DakAR-D4600 Tatera kempi MB 1968 DQ859062.1

Saumarez Reef virus SREV CSIRO-4 Ornithodoros capensis TB 1974 DQ235150.1

Sepik virus SEPV MK7148 Mansonia septempunctata MB 1966 DQ859063.1

Sitiawan virus STWV Broiler chicken MB 2000 JX477686.1

Spanish sheep encephalitis virus SSEV 87/2617 TBF 1987 DQ235152.1

Spondweni virus SPOV SM-6V-1 Mansonia spp.? MB 1955 DQ859064.1

St. Louis encephalitis virus SLEV MSI-7 MB 1975 DQ359217.1

Tamana bat virus TABV Tr127154 Pteronotus parnellii NKV 1973 NC_3996.1

Tick-borne encephalitis virus TBEV-Sib Vasilchenko Homo sapiens TB 1969 AF069066.1

Tick-borne encephalitis virus TBEV-FE Sofjin-HO Homo sapiens TB 1937 AB062064.1

Tick-borne encephalitis virus TBEV-Eu Salem Macaca sylvanus TB 2006 FJ572210.1

Tick-borne encephalitis virus TBEV-FE Oshima-5-10 Dog TB 1995 AB062063.2

Tick-borne encephalitis virus TBEV-Eu Neudoerfl Ixodes ricinus TB 1971 TEU27495

Tick-borne encephalitis virus TBEV-Sib Est54 Ixodes persulcatus TB 2000 GU183384.1

Tick-borne encephalitis virus TBEV 886-84 Clethrionomys rufocanus TB 1984 EF469662.1

Tick-borne encephalitis virus TBEV 178-79 Ixodes persulcatus TB 1979 EF469661.1

Tembusu virus TMUV ZJ GH-2 Goose MB 2010 JQ314465.1

Turkish sheep encephalitis virus TSEV TTE80 TB 1969 DQ235151.1

Tyuleniy virus TYUV 6017 TB 1971 DQ235148.1

Uganda S virus UGSV ORIGINAL Aedes spp. MB 1947 DQ859065.1

Usutu virus USUV SAAR-1776 Culex neavei MB 1958 AY453412.1

Orig

inof

theg

enusF

lavivirus

http://vir.sgm

journals.org1

97

3

compared with the more narrow Beringian calibration(Table S1).

The use of the Yang96 model with tip dates as calibrationrecovered older ages and broader intervals (95 % HPD)compared with using the SRD06 model (Table S1).However, both SRD06 and Yang96 model analyses underBeringian calibration recovered similar ranges for the 95 %HPD intervals (Table S1). The difference in estimates seenbetween SRD06, allowing for the third codon position tovary independently of the linked first and second codonpositions, and Yang96, allowing for all three codonpositions to vary, is an indication that divergence timeestimates become biased towards the present, with highersubstitution rates per nucleotide (Worobey et al., 2010).

The differences seen between the calibration schemes areexplained by the fact that estimating the tMRCA for entirevirus families or genera requires a broad sampling schemein order to cover as much of the variation as possible fromthe group in question. However, the wider the sampling, thehigher the variation in substitution rates among lineages.This will consequently lead to an increasing departure froma constant molecular clock (Bromham & Penny, 2003; Duffyet al., 2008; Holmes, 2003a). A high variation in substitutionrates will lead to uncertainty in the reconstructed divergencetimes unless calibration points restrict the node ages.Therefore, using tip dates alone results in a high level ofuncertainty for the deeper nodes and thus internalcalibration becomes crucial (Ho & Phillips, 2009). As theBeringian calibration constraint (15 000–11 000 years) isnarrower than the TBF calibration constraint (16 100–42 300years; Heinze et al., 2012), the resulting 95 % HPD intervalsfrom the Beringian analysis are also narrower.

Our study shows how incorporation of internal calibrationin large-scale virus phylogenies can help reconstruct moreprecise divergences times, independent of the substitutionmodel, given a robust and reliable calibration by narrowingdown the 95 % HPD intervals. Furthermore, as the meantMRCA estimates are generally congruent between thecalibration schemes applied (Table S1), we will hereafteronly report and discuss the results from the combined tipdates and the internal node calibration based on the Beringian-POWV biogeographical event, i.e. the Beringian calibration,with the most narrow 95 % HPD intervals, unless otherwisestated. This is because the incorporation of biogeographicalinformation is essential to date virus origins, especially forancient events (Katzourakis et al., 2009; Wertheim &Kosakovsky Pond, 2011), as is the case with Flavivirus.

Divergence times of the genus Flavivirus:congruences and incongruences

The genus Flavivirus is defined as sensu lato (node B) orsensu stricto (node C). Our analysis, inferred from completecoding nucleotide genomes, indicated that Flavivirus sensulato originated ~119 800 (87 100–158 900) years ago ifTABV is to be considered a part of the genus, or ~84 700T

ab

le1.

cont

.

Vir

us

Vir

us

abb

revi

atio

nS

trai

nS

ou

rce

of

iso

late

Vec

tor

gro

up

Yea

rG

enB

ank

acce

ssio

nn

o.

Wes

sels

bro

nvi

rus

WE

SS

VS

AH

-17

7-9

98

71

-2M

erin

osh

eep

MB

19

55

DQ

85

90

58

.1

Wes

tN

ile

viru

sW

NV

NY

99

-fla

min

go3

82

-99

Ph

oen

icop

teru

sch

ilen

sis

MB

19

99

AF

19

68

35

.2

Wes

tN

ile

viru

sW

NV

B9

56

Hom

osa

pie

ns

MB

19

37

NC

_1

56

3.2

Wes

tN

ile

viru

sW

NV

Rab

ensb

urg

-97

-10

3C

ule

xp

ipie

ns

MB

19

97

AY

76

52

64

.1

Yao

un

de

viru

sY

AO

VD

akA

r-Y

27

6C

ule

xn

ebu

losi

sM

B1

96

8E

U0

82

19

9.1

Yel

low

feve

rvi

rus

YF

VT

rin

idad

-79

A-7

88

37

9H

aem

ago

gus

speg

azz

ini

MB

19

79

AF

09

46

12

.1

Yo

ko

sevi

rus

YO

KV

Oit

a-3

6M

inio

pte

rus

fuli

gin

osu

sN

KV

19

71

AB

11

48

58

.1

Zik

avi

rus

ZIK

VM

R-7

66

Rh

esu

sm

on

key

MB

19

47

DQ

85

90

59

.1



(63 700–109 600) years ago if TABV is excluded, i.e.Flavivirus sensu stricto. Either way, this first moleculardating for the whole genus indicates a significantly olderage than the 10 000 years that was previously suggested forFlavivirus sensu stricto (Gould et al., 2001).

Our results also contrast the 3 000 years age of the MBFdom

group (node F), including MBFs, ISFs and NKVFs,estimated in Zanotto et al. (1996). In the present analysis,the MBFdom group (node F) was instead suggested to haveemerged .41 500 (32 600–51 600) years ago, with asubsequent diversification of the MBFAedes clade (nodeH) and the MBFCulex clade (node K) clade ~25 900 (19 800–32 000) and ~26 900 (21 000–33 400) years ago, respect-ively. Previously, the ISFa clade (node E) was indicated tohave emerged between 3500 and 350 000 years ago (Crochuet al., 2004). The combination of tip date calibration andinternal node calibration allows us to more preciselypinpoint the origin of the ISFa clade to have occurred~40 700 (30 800–52 300) years ago, i.e. ~44 000 years afterthe split from the last common ancestor of Flavivirus sensustricto (node C) that emerged ~84 700 (63 700–110 000)years ago. We also showed the first proposed dating of theNKVFa clade (node N), here estimated to have emerged~24 100 (18 000–30 900) years ago, ~15 000 years after thesplit from the last common ancestor of TBFs and NKVFs(node L).

The sharp contrast between the estimates in the presentstudy and previous studies can possibly be explainedbecause evolutionary rates in our study are inferred from agenus-wide sampled dataset using a codon-based substi-tution model as compared with a nucleotide-basedsubstitution model used in many other studies. It is likelythat nucleotide-based substitution models will not be ableto account for variation in selective pressure throughoutevolutionary history, where the effect of purifying selectionis likely to cause underestimation of the actual ages(Wertheim & Kosakovsky Pond, 2011). Using the SLAC, FEL,IFEL, FUBAR and MEME methods available within the HyPhypackage (Pond et al., 2005) and at the Datamonkeywebserver (Delport et al., 2010), there were none-to-weaksigns of positive selection, but strong signs of purifyingselection in the alignment used in the present study (resultsavailable upon request). Signs of strong purifying selectionwithin Flavivirus are in concordance with what has beenfound previously for members of the genus Flavivirus andfor vector-borne RNA viruses in general (Holmes, 2003b;Pybus et al., 2007; Woelk & Holmes, 2002). A codon-basedsubstitution model can to some extent compensate forpurifying selection, producing older tMRCA estimatesthan nucleotide-based substitution models (Wertheim &Kosakovsky Pond, 2011). However, and more importantly,to accurately date deep RNA virus origins, the use of othersources of evidence (such as biogeography) will eventuallybecome necessary as even codon-based models cannotaccount for the complex interactions of events and factorsthat have occurred throughout evolution (Katzourakis &Gifford, 2010; Katzourakis et al., 2009).

Our study is supported by the fact that both Beringiancalibration and only tip date calibration give similar meantMRCA estimates, although only tip date calibration givesbroader 95 % HPD intervals. In contrast to several studies(Sharp & Simmonds, 2011), we have not found any conflictbetween internal node and tip date calibrations. Our resultsare also congruent with several other studies (Table S3).The estimated tMRCA for the TBFs clade [node M, at27 500 (21 700–34 200) years ago] is supported by previousestimates for the whole clade (Heinze et al., 2012) (28 600years ago) and for the lineages within (Bertrand et al., 2012;Uzcategui et al., 2012) (Table S3). For the split of thePOWV clade and its sister (node O), our tMRCA estimatesof 14 800 years ago from tip date analysis also matchbroadly with previous molecular clock studies (12 300 yearsago; Heinze et al., 2012) and support the use of theBeringian biogeographical calibration.

Evolutionary rates of the genus Flavivirus andmolecular clocks in RNA viruses

Rates of nucleotide substitution were estimated from theBeringian calibrated BEAST analyses for the entire ORF ofthe genome (Table 3). Rates for the Beringian, TBF and tipdate calibration are summarized in Table S2.

The genomic substitution rate of positive-sense ssRNAviruses varies between 1022 and 1025 substitutions site–1

year21 (Hanada et al., 2004; Jenkins et al., 2002; Sanjuan,2012). Even though the estimated mean rates in the presentstudy for Flavivirus sensu lato (including TABV; node B),461025 (261025 to 661025) substitutions site–1 year21,are within range of those estimated for positive-sensessRNA viruses, they are at least an order of magnitudeslower than previous estimates for different lineages withinFlavivirus, i.e. 161023 and 261024 substitutions site–1

year21 as found for DENV-4 and St. Louis encephalitisvirus, respectively (Sanjuan, 2012 and references therein).Our results are not very surprising given that ratedifferences of up to 125 times have been shown for simianimmunodeficiency virus as compared with its sister, HIV(Worobey et al., 2010).

Furthermore, we did not find that the MBFdom group(node F) evolved at a significantly faster rate than the TBFclade (node M), as reported previously (Zanotto et al.,1995, 1996). Instead, our study showed that the MBFdom

group (node F) evolved at a mean rate of 3.261025

(1.761025 to 5.061025) substitutions site–1 year–1 andthat the TBF clade evolved at a mean rate of 3.961025

(2.361025 to 5.761025) substitutions site–1 year–1 (Table3). Therefore, the implications for Flavivirus evolutionderived from Zanotto et al. (1995, 1996) need to be revisedand reconsidered. The fastest rate of nucleotide substi-tution was found for the ISFa clade (node E), with a meanrate of 6.261025 (4.061025 to 8.661025) substitutionssite–1 year–1. Their fast rate of evolution could perhaps beexplained by their vertical mode of cycling transmission incombination with their specificity to insects with relatively



TBF calibration

(16 100–44 929 years)

Beringian calibration

(11 000–15 000 years)

DENV-3_1956_M93130.1

JEV_1982_M18370.1

KUNV_1960_D00246.1

MMLV_1958_AJ299445.1

RBV_1954_AF144692.1

KOUV_1968_EU082200.1

UGSV_1947_DQ859065.1

YAOV_1968_EU082199.1

TBEV-Eu_1971_TEU27495

NTAV_1966_JX236040.1

HANKV_2005_JQ268258.1

GGEV_1969_DQ235153.1

DENV-1_1974_U88536.1

NOUV_2004_FJ711167.1

TBEV-FE_1995_AB062063.2

DGV_2009_NC_016997.1

DENV-4_1956_M14931.2

ZIKV_1947_DQ859059.1

BVDV-1_1963_NC_001461.1

MVEV_1956_AF161266.1

POTV_1989_DQ859067.1

SEPV_1966_DQ859063.1

NAKV_2008_GQ165809.2

CFAV_1975_NC_001564.1

SABV_1968_DQ859062.1

ENTV_1957_DQ837641.1

DTV_1995_NC_003218.1

EHV_1969_DQ859060.1

OHFV_1947_AY438626.1

QBV_2002_NC_012671.1

OHFV_1947_AB507800.1

LGTV_1956_AF253419.1

DENV-2_1944_AF038403.1

APOIV_1954_AF160193.1

WNV_1997_AY765264.1

KEDV_1972_AY632540.2

PCV_2010_KC505248.1

SSEV_1987_DQ235152.1

TSEV_1969_DQ235151.1

TBEV-Sib_1969_AF069066.1

MEAV_1981_DQ235144.1

KADV_1967_DQ235146.1

BAGV_2010_HQ644143.1

MODV_1958_AJ242984.1

ROCV_1975_AY632542.4

BAGV_1966_AY632545.2

KFDV_1957_JF416959.1

JUGV_1968_DQ859066.1

YOKV_1971_AB114858.1

NMV_1998_KC788512.1

LAMV_2004_FJ606789.1

AEFV_2011_KC181923.1

TBEV-Sib_2000_GU183384.1

RFV_1968_DQ235149.1

STWV_2000_JX477686.1

WESSV_1955_DQ859058.1

AHFV_1995_AF331718.1

SREV_1974_DQ235150.1

GGYV_1975_DQ235145.1

TBEV_1984_EF469662.1

IGUV_1979_AY632538.4

KRV_1999_NC_005064.1

SLEV_1975_DQ359217.1

SPOV_1955_DQ859064.1

YFV_1979_AF094612.1

BSQV_1956_NC_009026.2

TABV_1973_NC_003996.1

TMUV_2010_JQ314465.1

BYDV_2010_JF312912.1

KOKV_1960_NC_009029.2

ILHV_1944_AY632539.4

BOUV_1967_DQ859057.1

TBEV-Eu_2006_FJ572210.1

TBEV_1979_EF469661.1

TBEV-FE_1937_AB062064.1

USUV_1958_AY453412.1

WNV_1937_NC_001563.2

ALFV_1966_AY898809.1

CHAOV_2003_JQ068102.1

TYUV_1971_DQ235148.1

LIV_1963_NC_001809.1

BANV_1956_DQ859056.1

CxFV_2003_AB262759.2

POWV_1958_L06436.1

KSIV_1972_NC_006947.1

WNV_1999_AF196835.2

A

C

B

D

E

F

G

I

L

M

N

O

J

K

H

TBF

MBF

MBF

ISF

ISF

NKVF

NKVF

NKVF

ISF

<10 500 years

glacier

13 000–11 000 years

land bridgeglacier

16 000–14 000 years

glacier

land bridge

25 000–18 000 years

glacier

land bridge

050100150200322 Time (×103 years)

POWV migration pathPOWV migration path

Fig. 1. Chronogram based on tip date calibrated terminals and internal nodes using Beringian calibration (see text for details).All named nodes have maximum posterior probability support (1.0 pp). Terminal tips are named by virus name (abbreviations,



short generation times (Bolling et al., 2012; Lutomiah et al.,2007).

Estimates of substitution rates are often found to be muchhigher for tips than for deep nodes (Wertheim &Kosakovsky Pond, 2011). From our perspective, a highlyvariable non-constant molecular clock with pulses of veryhigh substitution rates for short periods of time may reflectthe difference of orders of magnitude seen when comparingevolutionary rates in short and recent times (e.g. 161023

for DENV-4 in Klungthong et al., 2004) with long and deeptimes (e.g. 461025 for Flaviviruses sensu lato in this study).

Beringia and the emergence of POWV in NorthAmerica

The Beringian land bridge was open for land migrationbetween 15 000 and 11 000 years ago until it was floodedand the Bering Strait was formed (Dixon, 2013; Elias, 2001;Elias et al., 1996; Kelly, 2003), effectively blockingmigration between Asia and North America for animalsother than humans (Fig. 1). Our molecular reconstructionsusing only tip dates are congruent with POWV beingintroduced into North America when the Beringian landbridge was accessible and supports the use of tip dates incombination with an internal Beringian calibration (Fig. 1,Table 2).

How then did POWV enter North America? Althoughhumans might have been responsible for carrying POWV,

as humans do get infected with POWV, perhaps it is morelikely that ticks and associated tick hosts jointly broughtPOWV into North America. Human infections withPOWV in present times are infrequent (Ebel, 2010),although numbers appear to be on the increase (Hintenet al., 2008). Between 15 000 and 11 000 years ago, themigrating human population who settled the Americas hadan effective population size of perhaps ,80 individuals(Hey, 2005). This, in combination with the fact thatPOWV is considered to be maintained solely in an enzooticcycle between ixodid ticks and their vertebrate hosts (Ebel,2010; Ebel et al., 2000), is a strong argument againsthumans being the likely vectors by which POWV firstentered North America. Nor is it likely that POWV enteredthe Americas by the coastal route. Even though the coastalareas did support marine mammals and had terrestrialmammals living in refugia (Heaton & Grady, 2003), thedeglaciated coastal areas had sparse vegetation that did notsupport large populations of terrestrial mammals (Dixon,2013). Also, animals were isolated and movement wasinhibited due to the glaciers, and the coastal route was, ingeneral, not considered to be a migratory route formammals other than humans (Dixon, 2013).

In view of the results from the present study, the mostlikely time period of introduction was after the deglaciationcorridor became habitable ~13 500 years ago (Catto, 1996;Dixon, 2013) and before sea levels rose by 40 m and theBering strait reached its present width before 10 500 yearsago (Elias et al., 1996). During this period, populations of

see Table 1), year of isolation and GenBank accession number. Development of the Beringia region between the last glacialmaximum until the Bering Strait was formed (25 000–10 500 years ago). Beringia development after Elias et al. (1996) andDixon (2013).

Table 2. Mean estimated tMRCA (�103 years) and 95 % HPD interval: results for the Beringian, TBF and tip date calibration analyses

Node Beringian (11 000–15 000 years) TBF (16 100–42 300 years) Tip dates only

tMRCA 95 % HPD tMRCA 95 % HPD tMRCA 95 % HPD

A: Root 230.5 156.1–322.7 219.4 110.6–387.6 265.0 25.6–2.768.1

B: TABV/ISF–MBF–TBF–NKVF 119.8 87.1–158.9 114.0 62.2–194.6 138.7 12.9–1.456.1

C: ISFa/MBFdom–TBF–NKVFa 84.7 63.7–109.6 80.6 45.0–135.1 98.4 9.8–1.024.7

D: MBFdom/TBF–NKVFa 47.2 37.3–58.7 44.9 26.4–73.9 54.0 5.5–560.2

E: ISFa 40.7 30.8–52.3 38.5 21.7–64.6 46.8 4.6–486.7

F: MBFdom 41.5 32.6–51.6 39.4 23.1–64.9 47.4 4.9–492.7

G: NKVFb/MBFAedes 33.2 25.7–42.0 31.5 18.2–52.3 37.9 4.0–390.7

H: MBFAedes 25.9 19.8–33.0 24.6 14.0–41.0 29.5 2.9–304.1

I: ISFb/NOUV–MBFCulex 33.8 26.6–42.1 32.1 18.7–52.9 38.7 3.8–401.9

J: NOUV/MBFCulex 31.0 24.3–38.6 29.5 17.2–48.7 35.5 3.8–369.4

K: MBFCulex 26.9 21.0–3343 25.5 14.8–42.1 30.7 3.3–320.0

L: TBF/NKVFa 39.3 30.7–49.2 37.5 21.9–61.4 44.9 4.2–468.5

M: TBF 27.5 21.7–34.2 26.2 16.1–42.3 31.5 3.1–329.4

N: NKVFa 24.1 18.0–30.9 22.9 12.8–38.4 27.5 2.9–286.6

O: POWV/sister clade of TBF 12.8 11.0–14.8 12.2 7.0–20.0 14.8 1.5–152.3



ixodid ticks and land mammals (including humans) couldhave migrated from eastern Beringia through the thenhabitable deglaciated corridor to the south of the glaciers(Dixon, 2013; Goebel & Buvit, 2011; Shapiro et al., 2004).Ticks and tick hosts maintaining POWV might initiallyhave been restricted to eastern Beringia, north-west of theglaciers. As the deglaciated corridor became habitable andanimals could migrate southwards, and after the BeringStrait was formed preventing land migration back to Asia,POWV became established in North America. Using aninternal calibration of 15 000–11 000 or 16 000–10 000 yearsfor the split of POWV from closely related TBFs givessimilar tMRCA dates for POWV crossing into the Americas(12 800 and 12 500 years ago, respectively), which is inagreement with dates of the opening and closing of theBeringian land bridge.

Did flaviviruses and humans spread throughoutthe globe together?

Viruses, and RNA viruses in particular, are among the mostsuccessful evolving entities on the planet, having colonizedand adapted to an array of different environments andmodes of transmission within all domains of life (Wasik &Turner, 2012). The most significant event that might havehelped shape the distribution of flaviviruses is perhaps thespread of modern humans out of Africa. Modern humansarose in Africa (Cavalli-Sforza & Feldman, 2003; McDougallet al., 2005; Stringer & Andrews, 1988), and began migratingto other continents between 80 000 and 40 000 years ago,having populated all continents except Antarctica by 10 000years ago (Blome et al., 2012; Henn et al., 2012; Macaulayet al., 2005; Rasmussen et al., 2011). The migration ofhumans out of Africa has been accompanied by several otherpathogens. The pathogenic bacterium Helicobacter pyloriappears to have accompanied human migration out of Africaand is estimated to have spread from east Africa ~58 000

years ago (Falush et al., 2003; Linz et al., 2007; Moodley et al.,2012). Likewise, the Mycobacterium tuberculosis complex isestimated to have started to diverge ~70 000 years ago,consistent with an African origin linked to human expansionand migration (Comas et al., 2013; Wirth et al., 2008).Similarly, several human pathogenic viruses also showpatterns of co-divergence associated with human diversifica-tion (Holmes, 2004). Therefore, it is perhaps not unlikelythat the spread and evolution of several Flavivirus strains,especially if much of the Flavivirus diversity originated inAfrica (Gould et al., 2001, 2003), was influenced by humanmigration and expansion across the globe.

METHODS

Virus genomes, sequence alignments and models of molecular

evolution. Complete coding genomes of the genus Flavivirus,excluding the untranslated 59 and 39 flanking regions, were retrievedfrom GenBank (Table 1). Genomes were selected based on availablebackground information on the year of virus strain isolation (isolatedbetween 1937 and 2011). BVDV-1 of the genus Pestivirus within thefamily Flaviviridae was included as an outgroup based on a previouslypublished Flavivirus phylogeny (Cook et al., 2012). To construct arobust alignment without frameshifts, all sequences were translated toamino acids in SeaView 4.4.2 (Gouy et al., 2010), aligned with MAFFT

7.037b using the default G-INS-i algorithm and parameters (Katoh &Standley, 2013), and then back-translated into nucleotides. Theresulting alignments were inspected visually and edited with AliView0.8 (http://ormbunkar.se/aliview). To estimate the best-fitting modelof evolution for the sequence alignments, model tests were performedfor the nucleotide alignment using jModelTest 2.1.3 (Darriba et al.,2012) and for the amino acid alignment using ProtTest 3.2.1 (Darribaet al., 2011). To test the null hypothesis of a strict molecular clock, amaximum-likelihood clock test was performed using MEGA 5.22(Tamura et al., 2011). The best-fitting models were then specified inthe respective analyses.

Phylogenetic analysis: maximum-likelihood and Bayesian

analysis. Maximum-likelihood trees were estimated using RAxML7.6.3 (Stamatakis, 2006) with 1000 rapid bootstraps under the GTR+I+G model for the nucleotide alignment and the WAG+I+G modelfor the amino acid alignment, respectively. Bayesian phylogenetic treeswere inferred using MrBayes 3.2.1 (Ronquist et al., 2012) by executingtwo parallel runs with four Metropolis-coupled chains for 20 million and9 million Markov chain Monte Carlo (MCMC) generations, usingGTR+I+G for nucleotides and WAG+I+G for amino acids as modelof evolution, respectively, sampling every 1000 generations and run withdefault priors, discarding the first 25 % as burn-in.

Estimating the tMRCA: Bayesian analysis with BEAST. To explorethe temporal scale of the entire genus Flavivirus, evolutionary ratesand the tMRCA were estimated using a Bayesian MCMC method asimplemented in BEAST 1.7.5 (Drummond et al., 2012). All BEAST runswere performed with calibrated tip dates where the sequence with themost recent sampling date (2011) was set to represent the present.The SRD06 codon-based partition model (Shapiro et al., 2006) andHKY85+G nucleotide substitution model were used together with anon-parametric Bayesian skyline population coalescent tree priorwith a piecewise-constant skyline demographic model (Drummondet al., 2005), along with a log-normal uncorrelated relaxed molecularclock. The effect of using the Yang96 codon model (Yang, 1996) wasalso explored in a separate analysis. All analyses were run for 100million generations in triplicate, sampling every 1000 generations toensure mixing of chains and that a sufficiently effective sample size

Table 3. Mean estimated evolutionary rates (�10”5 substitu-tions site–1 year–1) and 95 % HPD intervals for the Beringiancalibration (11 000–15 000 years)

Node Mean rate 95 % HPD

B: TABV/ISF–MBF–TBF–NKVF 3.96 2.04–6.14

C: ISFa/MBFdom–TBF–NKVF 3.60 1.81–5.57

D: MBFdom/TBF–NKVFa 2.56 1.57–3.70

E: ISFa 6.24 4.00–8.65

F: MBFdom 3.24 1.73–5.00

G: NKVFb/MBFAedes 2.92 1.63–4.43

H: MBFAedes 3.11 1.70–4.72

I: ISFb/NOUV–MBFCulex 4.02 2.28–6.03

J: NOUV/MBFCulex 3.95 2.12–6.06

K: MBFCulex 4.19 2.30–6.37

L: TBF/NKVFa 4.27 2.39–6.38

M: TBF 3.89 2.32–5.73

N: NKVFa 5.18 2.32–6.15

O: POWV/sister clade of TBF 4.30 3.31–7.23



(.200) was reached. Convergence of chains and effective sample size

statistics were analysed with Tracer 1.5 (http://beast.bio.ed.ac.uk/

Tracer). Log and tree files were combined in LogCombiner (BEAST

package) to produce consensus files of the different runs. Finally,

maximum clade credibility trees of the three different analyses were

produced using TreeAnnotator (BEAST package). Chronograms were

viewed and annotated in FigTree 1.3.1 (http://tree.bio.ed.ac.uk/software/

figtree). Computations were performed at the Bioportal webportal at the

University of Oslo (www.bioportal.uio.no), the CIPRES webportal

(Miller et al., 2010) and at the Uppsala Multidisciplinary Center for

Advanced Computational Science (www.uppmax.uu.se).

Calibration schemes. In BEAST, three calibration schemes were

applied. (i) The first analysis was run with default uniform priors

using tip dates only (i.e. uncalibrated internal nodes) allowing the

MCMC chains to freely explore the treespace and the node height. (ii)

The second analysis was run using tip dates together with secondary

calibration data for the TBF node (split of mammalian TBFs from

seabird TBFs), previously estimated to have originated between 44 929

and 16 100 years ago (Heinze et al., 2012), hereafter referred to as the

TBF calibration. These dates were incorporated as upper and lower

bounds using a uniform distribution. (iii) The third analysis was run

using tip dates together with an internal node calibration based on the

biogeographical event for the split of the POWV and closely related

TBFs during the existence of the Beringian land bridge, estimated to

have been open for mammal land migration between 15 000 and

11 000 years ago (see Introduction). Here, the ages of the opening and

closing times of the Beringian land bridge were used to specify a

maximum bound (15 000 years) and minimum bound (11 000 years)

for the age of this split, given that the POWV clade is the only North

American TBF and that all other TBFs are either African, Eurasian or

Oceanian (Heinze et al., 2012), hereafter referred to as the Beringian

calibration. In addition to the Beringian calibration, we also tested the

effect of allowing for a wider range of the opening and closing of the

Beringia land bridge (Dixon, 2013). Here, the age for the split

between POWV and closely related TBFs was set with an internal

calibration to an upper bound of 16 000 years ago and a lower bound

of 10 000 years ago.

ACKNOWLEDGEMENTS

We are grateful to Allison Perrigo, Martin Ryberg and Petra Korall for

their constructive comments and feedback on the manuscript. We

acknowledge Allison Perrigo and Stina Weststrand for their help with

the design of the figures. We are also grateful to James E. Dixon for

valuable information and discussions on Beringian biogeography.

Finally, we wish to acknowledge all the hard work by people all over

the world making the sequences used in the present study available to

everyone. J. H.-O. P. was supported by Stiftelsen Olle Engkvist

Byggmastare.

REFERENCES

Achilli, A., Perego, U. A., Lancioni, H., Olivieri, A., Gandini, F.,Hooshiar Kashani, B., Battaglia, V., Grugni, V., Angerhofer, N. &other authors (2013). Reconciling migration models to the Americas

with the variation of North American native mitogenomes. Proc Natl

Acad Sci U S A 110, 14308–14313.

Bertrand, Y., Topel, M., Elvang, A., Melik, W. & Johansson, M. (2012).First dating of a recombination event in mammalian tick-borne

flaviviruses. PLoS ONE 7, e31981.

Billoir, F., de Chesse, R., Tolou, H., de Micco, P., Gould, E. A. & deLamballerie, X. (2000). Phylogeny of the genus flavivirus using

complete coding sequences of arthropod-borne viruses and viruses

with no known vector. J Gen Virol 81, 781–790.

Blome, M. W., Cohen, A. S., Tryon, C. A., Brooks, A. S. & Russell, J.(2012). The environmental context for the origins of modern humandiversity: a synthesis of regional variability in African climate

150,000–30,000 years ago. J Hum Evol 62, 563–592.

Bodner, M., Perego, U. A., Huber, G., Fendt, L., Rock, A. W.,Zimmermann, B., Olivieri, A., Gomez-Carballa, A., Lancioni, H. &other authors (2012). Rapid coastal spread of First Americans: novel

insights from South America’s Southern Cone mitochondrialgenomes. Genome Res 22, 811–820.

Bolling, B. G., Olea-Popelka, F. J., Eisen, L., Moore, C. G. & Blair,C. D. (2012). Transmission dynamics of an insect-specific flavivirus in a

naturally infected Culex pipiens laboratory colony and effects of co-

infection on vector competence for West Nile virus. Virology 427, 90–97.

Bromham, L. & Penny, D. (2003). The modern molecular clock. Nat

Rev Genet 4, 216–224.

Catto, N. R. (1996). Richardson Mountains, Yukon-Northwest

territories: the northern portal of the postulated ‘Ice-Free Corridor’.

Quartern Int 32, 3–19.

Cavalli-Sforza, L. L. & Feldman, M. W. (2003). The application of

molecular genetic approaches to the study of human evolution. Nat

Genet 33 (Suppl), 266–275.

Chambers, T. J., Hahn, C. S., Galler, R. & Rice, C. M. (1990). Flavivirus

genome organization, expression, and replication. Annu Rev Microbiol

44, 649–688.

Comas, I., Coscolla, M., Luo, T., Borrell, S., Holt, K. E., Kato-Maeda,M., Parkhill, J., Malla, B., Berg, S. & other authors (2013). Out-of-Africa migration and Neolithic coexpansion of Mycobacterium

tuberculosis with modern humans. Nat Genet 45, 1176–1182.

Cook, S. & Holmes, E. C. (2006). A multigene analysis of thephylogenetic relationships among the flaviviruses (Family: Flavi-

viridae) and the evolution of vector transmission. Arch Virol 151,

309–325.

Cook, S., Moureau, G., Kitchen, A., Gould, E. A., de Lamballerie, X.,Holmes, E. C. & Harbach, R. E. (2012). Molecular evolution of theinsect-specific flaviviruses. J Gen Virol 93, 223–234.

Crochu, S., Cook, S., Attoui, H., Charrel, R. N., De Chesse, R.,Belhouchet, M., Lemasson, J. J., de Micco, P. & de Lamballerie, X.(2004). Sequences of flavivirus-related RNA viruses persist in DNA

form integrated in the genome of Aedes spp. mosquitoes. J Gen Virol

85, 1971–1980.

Darriba, D., Taboada, G. L., Doallo, R. & Posada, D. (2011). ProtTest

3: fast selection of best-fit models of protein evolution. Bioinformatics27, 1164–1165.

Darriba, D., Taboada, G. L., Doallo, R. & Posada, D. (2012).jModelTest 2: more models, new heuristics and parallel computing.Nat Methods 9, 772.

de Lamballerie, X., Crochu, S., Billoir, F., Neyts, J., de Micco, P.,Holmes, E. C. & Gould, E. A. (2002). Genome sequence analysis of

Tamana bat virus and its relationship with the genus Flavivirus. J Gen

Virol 83, 2443–2454.

Delport, W., Poon, A. F. Y., Frost, S. D. W. & Kosakovsky Pond, S. L.(2010). Datamonkey 2010: a suite of phylogenetic analysis tools for

evolutionary biology. Bioinformatics 26, 2455–2457.

Dixon, E. J. (2013). Late Pleistocene colonization of North America

from Northeast Asia: new insights from large-scale paleogeographicreconstructions. Quartern Int 285, 57–67.

Drummond, A. J., Pybus, O. G., Rambaut, A., Forsberg, R. & Rodrigo,A. G. (2003). Measurably evolving populations. Trends Ecol Evol 18,481–488.



Drummond, A. J., Rambaut, A., Shapiro, B. & Pybus, O. G. (2005).Bayesian coalescent inference of past population dynamics frommolecular sequences. Mol Biol Evol 22, 1185–1192.

Drummond, A. J., Suchard, M. A., Xie, D. & Rambaut, A. (2012).Bayesian phylogenetics with BEAUti and the BEAST 1.7. Mol Biol Evol29, 1969–1973.

Duffy, S., Shackelton, L. A. & Holmes, E. C. (2008). Rates ofevolutionary change in viruses: patterns and determinants. Nat RevGenet 9, 267–276.

Dunham, E. J. & Holmes, E. C. (2007). Inferring the timescale ofdengue virus evolution under realistic models of DNA substitution.J Mol Evol 64, 656–661.

Dyke, A. S. (2004). An outline of the deglaciation of North Americawith emphasis on central and northern Canada. In QuaternaryGlaciations, Extent and Chronology Part II: North America, pp. 373–424. Edited by J. Ehlers & P. L. Gibbard. Amsterdam: Elsevier.

Ebel, G. D. (2010). Update on Powassan virus: emergence of aNorth American tick-borne flavivirus. Annu Rev Entomol 55, 95–110.

Ebel, G. D., Campbell, E. N., Goethert, H. K., Spielman, A. & Telford,S. R., III (2000). Enzootic transmission of deer tick virus in NewEngland and Wisconsin sites. Am J Trop Med Hyg 63, 36–42.

Elias, S. A. (2001). Beringian paleoecology: results from the 1997workshop. Quatern Sci Rev 20, 7–13.

Elias, S. A., Short, S. K., Nelson, C. H. & Birks, H. H. (1996). Life andtimes of the Bering land bridge. Nature 382, 60–63.

Fagundes, N. J. R., Kanitz, R., Eckert, R., Valls, A. C. S., Bogo, M. R.,Salzano, F. M., Smith, D. G., Silva, W. A., Jr, Zago, M. A. & otherauthors (2008). Mitochondrial population genomics supports asingle pre-Clovis origin with a coastal route for the peopling of theAmericas. Am J Hum Genet 82, 583–592.

Falush, D., Wirth, T., Linz, B., Pritchard, J. K., Stephens, M., Kidd, M.,Blaser, M. J., Graham, D. Y., Vacher, S. & other authors (2003).Traces of human migrations in Helicobacter pylori populations.Science 299, 1582–1585.

Gaunt, M. W., Sall, A. A., de Lamballerie, X., Falconar, A. K.,Dzhivanian, T. I. & Gould, E. A. (2001). Phylogenetic relationships offlaviviruses correlate with their epidemiology, disease association andbiogeography. J Gen Virol 82, 1867–1876.

Goebel, T. & Buvit, I. (editors) (2011). From the Yenisei to the Yukon:Interpreting Lithic Assemblage Variability in Late Pleistocene/EarlyHolocene Beringia. College Station, TX: Texas A&M University Press.

Goebel, T., Waters, M. R. & O’Rourke, D. H. (2008). The latePleistocene dispersal of modern humans in the Americas. Science 319,1497–1502.

Gould, E. A. & Solomon, T. (2008). Pathogenic flaviviruses. Lancet371, 500–509.

Gould, E. A., de Lamballerie, X., Zanotto, P. M. & Holmes, E. C.(2001). Evolution, epidemiology, and dispersal of flaviviruses revealedby molecular phylogenies. Adv Virus Res 57, 71–103.

Gould, E. A., de Lamballerie, X., Zanotto, P. M. & Holmes, E. C.(2003). Origins, evolution, and vector/host coadaptations within thegenus Flavivirus. Adv Virus Res 59, 277–314.

Gouy, M., Guindon, S. & Gascuel, O. (2010). SeaView version 4: Amultiplatform graphical user interface for sequence alignment andphylogenetic tree building. Mol Biol Evol 27, 221–224.

Grard, G., Moureau, G., Charrel, R. N., Lemasson, J.-J., Gonzalez,J.-P., Gallian, P., Gritsun, T. S., Holmes, E. C., Gould, E. A. & deLamballerie, X. (2007). Genetic characterization of tick-borneflaviviruses: new insights into evolution, pathogenetic determinantsand taxonomy. Virology 361, 80–92.

Grard, G., Moureau, G., Charrel, R. N., Holmes, E. C., Gould, E. A. &de Lamballerie, X. (2010). Genomics and evolution of Aedes-borneflaviviruses. J Gen Virol 91, 87–94.

Gritsun, T. S., Nuttall, P. A. & Gould, E. A. (2003). Tick-borneflaviviruses. Adv Virus Res 61, 317–371.

Gubler, D., Kuno, G. & Markoff, L. (2007). Flaviviruses. In FieldsVirology, 5th edn, vol. 1, pp. 1153–1253. Edited by D. Knipe,P. Howley, D. Griffin, R. Lamb, M. Martin, B. Roizman & S. Straus.Philadelphia, PA: Lippincott-Raven.

Hanada, K., Suzuki, Y. & Gojobori, T. (2004). A large variation in therates of synonymous substitution for RNA viruses and its relationshipto a diversity of viral infection and transmission modes. Mol Biol Evol21, 1074–1080.

Heaton, T. H. & Grady, F. (2003). The late Wisconsin vertebrate historyof Prince of Wales Island, Southeast Alaska. In Ice Age Cave Faunas ofNorth America, pp. 17–53. Edited by B. W. Schubert, J. I. Mead &R. W. Graham. Bloomington, IN: Indiana University Press.

Heinze, D. M., Gould, E. A. & Forrester, N. L. (2012). Revisiting theclinal concept of evolution and dispersal for the tick-borne flavivirusesby using phylogenetic and biogeographic analyses. J Virol 86, 8663–8671.

Henn, B. M., Cavalli-Sforza, L. L. & Feldman, M. W. (2012). The greathuman expansion. Proc Natl Acad Sci U S A 109, 17758–17764.

Hey, J. (2005). On the number of New World founders: a populationgenetic portrait of the peopling of the Americas. PLoS Biol 3, e193.

Hinten, S. R., Beckett, G. A., Gensheimer, K. F., Pritchard, E.,Courtney, T. M., Sears, S. D., Woytowicz, J. M., Preston, D. G., Smith,R. P., Jr & other authors (2008). Increased recognition of Powassanencephalitis in the United States, 1999–2005. Vector Borne ZoonoticDis 8, 733–740.

Ho, S. Y. W. & Phillips, M. J. (2009). Accounting for calibrationuncertainty in phylogenetic estimation of evolutionary divergencetimes. Syst Biol 58, 367–380.

Holmes, E. C. (2003a). Molecular clocks and the puzzle of RNA virusorigins. J Virol 77, 3893–3897.

Holmes, E. C. (2003b). Patterns of intra- and interhost nonsynon-ymous variation reveal strong purifying selection in dengue virus.J Virol 77, 11296–11298.

Holmes, E. C. (2004). The phylogeography of human viruses. MolEcol 13, 745–756.

Hoshino, K., Isawa, H., Tsuda, Y., Sawabe, K. & Kobayashi, M.(2009). Isolation and characterization of a new insect flavivirus fromAedes albopictus and Aedes flavopictus mosquitoes in Japan. Virology391, 119–129.

Huhtamo, E., Putkuri, N., Kurkela, S., Manni, T., Vaheri, A., Vapalahti,O. & Uzcategui, N. Y. (2009). Characterization of a novel flavivirusfrom mosquitoes in northern europe that is related to mosquito-borne flaviviruses of the tropics. J Virol 83, 9532–9540.

Hulten, E. (1937). Outline of the History of Arctic and Boreal Biota Duringthe Quaternary Period. Stockholm: Bokforlagsaktiebolaget Thule.

Jenkins, G. M., Rambaut, A., Pybus, O. G. & Holmes, E. C. (2002).Rates of molecular evolution in RNA viruses: a quantitativephylogenetic analysis. J Mol Evol 54, 156–165.

Katoh, K. & Standley, D. M. (2013). MAFFT multiple sequencealignment software version 7: improvements in performance andusability. Mol Biol Evol 30, 772–780.

Katzourakis, A. & Gifford, R. J. (2010). Endogenous viral elements inanimal genomes. PLoS Genet 6, e1001191.

Katzourakis, A., Gifford, R. J., Tristem, M., Gilbert, M. T. P. & Pybus,O. G. (2009). Macroevolution of complex retroviruses. Science 325,1512.



Kelly, R. L. (2003). Maybe we do know when people first came to

North America; and what does it mean if we do? Quartern Int 109–110, 133–145.

Klungthong, C., Zhang, C., Mammen, M. P., Jr, Ubol, S. & Holmes,E. C. (2004). The molecular epidemiology of dengue virus serotype 4

in Bangkok, Thailand. Virology 329, 168–179.

Kolodziejek, J., Pachler, K., Bin, H., Mendelson, E., Shulman, L.,Orshan, L. & Nowotny, N. (2013). Barkedji virus, a novel mosquito-

borne flavivirus identified in Culex perexiguus mosquitoes, Israel,

2011. J Gen Virol 94, 2449–2457.

Korber, B., Muldoon, M., Theiler, J., Gao, F., Gupta, R., Lapedes, A.,Hahn, B. H., Wolinsky, S. & Bhattacharya, T. (2000). Timing theancestor of the HIV-1 pandemic strains. Science 288, 1789–1796.

Kumar, S. R. P., Patil, J. A., Cecilia, D., Cherian, S. S., Barde, P. V.,Walimbe, A. M., Yadav, P. D., Yergolkar, P. N., Shah, P. S. & otherauthors (2010). Evolution, dispersal and replacement of American

genotype dengue type 2 viruses in India (1956–2005): selection

pressure and molecular clock analyses. J Gen Virol 91, 707–720.

Kuno, G. (2007). Host range specificity of flaviviruses: correlation

with in vitro replication. J Med Entomol 44, 93–101.

Kuno, G. & Chang, G.-J. J. (2006). Characterization of Sepik and

Entebbe bat viruses closely related to yellow fever virus. Am J Trop

Med Hyg 75, 1165–1170.

Lee, J. S., Grubaugh, N. D., Kondig, J. P., Turell, M. J., Kim, H.-C.,Klein, T. A. & O’Guinn, M. L. (2013). Isolation and genomic

characterization of Chaoyang virus strain ROK144 from Aedes vexansnipponii from the Republic of Korea. Virology 435, 220–224.

Lindenbach, B. D. & Rice, C. M. (2003). Molecular biology offlaviviruses. Adv Virus Res 59, 23–61.

Lindenbach, B. D., Thiel, H.-J. & Rice, C. M. (2007). Flaviviridae: the

viruses and their replication. In Fields Virology, 5th edn, vol. 1, pp.1101–1152. Edited by D. Knipe, P. Howley, D. Griffin, R. Lamb,

M. Martin, B. Roizman & S. Straus. Philadelphia, PA: Lippincott-Raven.

Linz, B., Balloux, F., Moodley, Y., Manica, A., Liu, H., Roumagnac, P.,Falush, D., Stamer, C., Prugnolle, F. & other authors (2007). An

African origin for the intimate association between humans andHelicobacter pylori. Nature 445, 915–918.

Lobo, F. P., Mota, B. E. F., Pena, S. D. J., Azevedo, V., Macedo, A. M.,Tauch, A., Machado, C. R. & Franco, G. R. (2009). Virus–hostcoevolution: common patterns of nucleotide motif usage in

Flaviviridae and their hosts. PLoS ONE 4, e6282.

Lutomiah, J. J. L., Mwandawiro, C., Magambo, J. & Sang, R. C. (2007).Infection and vertical transmission of Kamiti river virus in laboratory

bred Aedes aegypti mosquitoes. J Insect Sci 7, 1–7.

Macaulay, V., Hill, C., Achilli, A., Rengo, C., Clarke, D., Meehan, W.,Blackburn, J., Semino, O., Scozzari, R. & other authors (2005).Single, rapid coastal settlement of Asia revealed by analysis ofcomplete mitochondrial genomes. Science 308, 1034–1036.

Mackenzie, J. S., Gubler, D. J. & Petersen, L. R. (2004). Emergingflaviviruses: the spread and resurgence of Japanese encephalitis, West

Nile and dengue viruses. Nat Med 10 (Suppl), S98–S109.

Mandryk, C. A., Josenhans, H., Fedje, D. W. & Mathewes, R. W.(2001). Late Quaternary paleoenvironments of Northwestern North

America: implications for inland versus coastal migration routes.

Quat Sci Rev 20, 301–314.

May, F. J., Davis, C. T., Tesh, R. B. & Barrett, A. D. T. (2011).Phylogeography of West Nile virus: from the cradle of evolution inAfrica to Eurasia, Australia, and the Americas. J Virol 85, 2964–2974.

McDougall, I., Brown, F. H. & Fleagle, J. G. (2005). Stratigraphic

placement and age of modern humans from Kibish, Ethiopia. Nature433, 733–736.

Medeiros, D. B. A., Nunes, M. R. T., Vasconcelos, P. F. C., Chang,G.-J. J. & Kuno, G. (2007). Complete genome characterization ofRocio virus (Flavivirus: Flaviviridae), a Brazilian flavivirus isolatedfrom a fatal case of encephalitis during an epidemic in Sao Paulostate. J Gen Virol 88, 2237–2246.

Mehla, R., Kumar, S. R. P., Yadav, P., Barde, P. V., Yergolkar, P. N.,Erickson, B. R., Carroll, S. A., Mishra, A. C., Nichol, S. T. & Mourya,D. T. (2009). Recent ancestry of Kyasanur Forest disease virus. EmergInfect Dis 15, 1431–1437.

Miller, M. A., Pfeiffer, W. & Schwartz, T. (2010). Creating the CIPRESScience Gateway for inference of large phylogenetic trees. Presented atthe Gateway Computing Environments Workshop, 2010.

Mohammed, M. A. F., Galbraith, S. E., Radford, A. D., Dove, W.,Takasaki, T., Kurane, I. & Solomon, T. (2011). Molecular phylogeneticand evolutionary analyses of Muar strain of Japanese encephalitisvirus reveal it is the missing fifth genotype. Infect Genet Evol 11, 855–862.

Moodley, Y., Linz, B., Bond, R. P., Nieuwoudt, M., Soodyall, H.,Schlebusch, C. M., Bernhoft, S., Hale, J., Suerbaum, S. & otherauthors (2012). Age of the association between Helicobacter pyloriand man. PLoS Pathog 8, e1002693.

Pan, X.-L., Liu, H., Wang, H.-Y., Fu, S.-H., Liu, H.-Z., Zhang, H.-L., Li,M.-H., Gao, X.-Y., Wang, J.-L. & other authors (2011). Emergence ofgenotype I of Japanese encephalitis virus as the dominant genotype inAsia. J Virol 85, 9847–9853.

Patil, J. A., Cherian, S., Walimbe, A. M., Patil, B. R., Sathe, P. S., Shah,P. S. & Cecilia, D. (2011). Evolutionary dynamics of the AmericanAfrican genotype of dengue type 1 virus in India (1962–2005). InfectGenet Evol 11, 1443–1448.

Pond, S. L. K., Frost, S. D. W. & Muse, S. V. (2005). HyPhy: hypothesistesting using phylogenies. Bioinformatics 21, 676–679.

Porterfield, J. S. (1980). Antigenic characteristics and classification ofTogaviridae. In The Togaviruses: Biology, Structure, Replication, pp.13–46. Edited by R. W. Schlesinger. New York: Academic Press.

Pybus, O. G., Rambaut, A., Belshaw, R., Freckleton, R. P.,Drummond, A. J. & Holmes, E. C. (2007). Phylogenetic evidence fordeleterious mutation load in RNA viruses and its contribution to viralevolution. Mol Biol Evol 24, 845–852.

Ramırez, A., Fajardo, A., Moros, Z., Gerder, M., Caraballo, G.,Camacho, D., Comach, G., Alarcon, V., Zambrano, J. & other authors(2010). Evolution of dengue virus type 3 genotype III in Venezuela:diversification, rates and population dynamics. Virol J 7, 329.

Rasmussen, M., Guo, X., Wang, Y., Lohmueller, K. E., Rasmussen, S.,Albrechtsen, A., Skotte, L., Lindgreen, S., Metspalu, M. & otherauthors (2011). An Aboriginal Australian genome reveals separatehuman dispersals into Asia. Science 334, 94–98.

Ronquist, F., Teslenko, M., van der Mark, P., Ayres, D. L., Darling, A.,Hohna, S., Larget, B., Liu, L., Suchard, M. A. & Huelsenbeck, J. P.(2012). MrBayes 3.2: efficient Bayesian phylogenetic inference andmodel choice across a large model space. Syst Biol 61, 539–542.

Sanjuan, R. (2012). From molecular genetics to phylodynamics:evolutionary relevance of mutation rates across viruses. PLoS Pathog8, e1002685.

Schurr, T. G. (2004). The peopling of the new world: perspectivesfrom molecular anthropology. Annu Rev Anthropol 33, 551–583.

Shapiro, B., Drummond, A. J., Rambaut, A., Wilson, M. C., Matheus,P. E., Sher, A. V., Pybus, O. G., Gilbert, M. T. P., Barnes, I. & otherauthors (2004). Rise and fall of the Beringian steppe bison. Science306, 1561–1565.

Shapiro, B., Rambaut, A. & Drummond, A. J. (2006). Choosingappropriate substitution models for the phylogenetic analysis ofprotein-coding sequences. Mol Biol Evol 23, 7–9.



Sharp, P. M. & Simmonds, P. (2011). Evaluating the evidence forvirus/host co-evolution. Curr Opin Virol 1, 436–441.

Stamatakis, A. (2006). RAxML-VI-HPC: maximum likelihood-basedphylogenetic analyses with thousands of taxa and mixed models.Bioinformatics 22, 2688–2690.

Stringer, C. B. & Andrews, P. (1988). Genetic and fossil evidence forthe origin of modern humans. Science 239, 1263–1268.

Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M. & Kumar,S. (2011). MEGA5: molecular evolutionary genetics analysis usingmaximum likelihood, evolutionary distance, and maximum par-simony methods. Mol Biol Evol 28, 2731–2739.

Twiddy, S. S., Holmes, E. C. & Rambaut, A. (2003). Inferring the rateand time-scale of dengue virus evolution. Mol Biol Evol 20, 122–129.

Uzcategui, N. Y., Sironen, T., Golovljova, I., Jaaskelainen, A. E.,Valimaa, H., Lundkvist, A., Plyusnin, A., Vaheri, A. & Vapalahti, O.(2012). Rate of evolution and molecular epidemiology of tick-borneencephalitis virus in Europe, including two isolations from the samefocus 44 years apart. J Gen Virol 93, 786–796.

Wasik, B. R. & Turner, P. E. (2013). On the biological success ofviruses. Annu Rev Microbiol 67, 519–541.

Weidmann, M., Frey, S., Freire, C. C. M., Essbauer, S., Ruzek, D.,Klempa, B., Zubrikova, D., Vogerl, M., Pfeffer, M. & other authors

(2013). Molecular phylogeography of tick-borne encephalitis virus incentral Europe. J Gen Virol 94, 2129–2139.

Wertheim, J. O. & Kosakovsky Pond, S. L. (2011). Purifying selectioncan obscure the ancient age of viral lineages. Mol Biol Evol 28, 3355–3365.

Wirth, T., Hildebrand, F., Allix-Beguec, C., Wolbeling, F., Kubica, T.,Kremer, K., van Soolingen, D., Rusch-Gerdes, S., Locht, C. & otherauthors (2008). Origin, spread and demography of the Mycobac-terium tuberculosis complex. PLoS Pathog 4, e1000160.

Woelk, C. H. & Holmes, E. C. (2002). Reduced positive selection invector-borne RNA viruses. Mol Biol Evol 19, 2333–2336.

Worobey, M., Telfer, P., Souquiere, S., Hunter, M., Coleman, C. A.,Metzger, M. J., Reed, P., Makuwa, M., Hearn, G. & other authors(2010). Island biogeography reveals the deep history of SIV. Science329, 1487.

Yang, Z. (1996). Maximum-likelihood models for combined analysesof multiple sequence data. J Mol Evol 42, 587–596.

Zanotto, P. M., Gao, G. F., Gritsun, T., Marin, M. S., Jiang, W. R.,Venugopal, K., Reid, H. W. & Gould, E. A. (1995). An arbovirus clineacross the northern hemisphere. Virology 210, 152–159.

Zanotto, P. M., Gould, E. A., Gao, G. F., Harvey, P. H. & Holmes, E. C.(1996). Population dynamics of flaviviruses revealed by molecularphylogenies. Proc Natl Acad Sci U S A 93, 548–553.



dating the origin of the genus flavivirus in the light of

Documents