s Lu, L., Lycett, S. J., and Leigh Brown, A. J. (2014) Determining the phylogenetic and phylogeographic origin of highly pathogenic avian influenza (H7N3) in Mexico. PLoS ONE, 9 (9). e107330. ISSN 1932-6203 Copyright © 2014 The Authors http://eprints.gla.ac.uk/100137 Deposited on: 11 December 2014 Enlighten – Research publications by members of the University of Glasgow http://eprints.gla.ac.uk Determining the Phylogenetic and Phylogeographic Origin of Highly Pathogenic Avian Influenza (H7N3) in Mexico Lu Lu1, Samantha J. Lycett2, Andrew J. Leigh Brown1* 1InstituteofEvolutionaryBiology,UniversityofEdinburgh,AshworthLaboratories,Edinburgh,UnitedKingdom,2UniversityofGlasgow,InstituteofBiodiversity,Animal HealthandComparativeMedicine,Glasgow,UnitedKingdom Abstract Highlypathogenic(HP)avianinfluenzavirus(AIV)H7N3outbreaksoccurred3timesintheAmericasinthepast10yearsand caused severe economic loss in the affected regions. In June/July2012, new HP H7N3 outbreaks occurred at commercial farmsinJalisco,Mexico.OutbreakscontinuedtobeidentifiedinneighbouringstatesinMexicotillAugust2013.Toexplore the origin of this outbreak, time resolved phylogenetic trees were generated from the eight segments of full-length AIV sequences in North America using BEAST. Location, subtype, avian host species and pathogenicity were modelled as discretetraitsuponthetreesusingcontinuoustimeMarkovchains.Afurtherjointanalysisamongsegmentswasperformed using a hierarchical phylogenetic model (HPM) which allowed trait rates (location, subtype, host species) to be jointly inferred across different segments. The complete spatial diffusion process was visualised through virtual globe software. Our result indicated the Mexico HP H7N3 originated from the large North America low pathogenicity AIV pool through complicated reassortment events. Different segments were contributed by wild waterfowl from different N. American flyways.Fiveoftheeightsegments(HA,NA,NP,M,NS)wereintroducedfromwildbirdsmigratingalongthecentralNorth Americanflyway, andPB2,PB1and PAwereintroducedviathewesternNorth Americanflyway.Theseresultshighlighta potential role for Mexico as a hotspot of virus reassortment as it is where wild birds from different migration routes mix duringthe winter. Citation:LuL,LycettSJ,LeighBrownAJ(2014)DeterminingthePhylogeneticandPhylogeographicOriginofHighlyPathogenicAvianInfluenza(H7N3)in Mexico.PLoSONE9(9):e107330.doi:10.1371/journal.pone.0107330 Editor:FlorianKrammer,IcahnSchoolofMedicineatMountSinai,UnitedStatesofAmerica ReceivedMay26,2014;AcceptedAugust16,2014;PublishedSeptember16,2014 Copyright:(cid:2)2014Luetal.Thisisanopen-accessarticledistributedunderthetermsoftheCreativeCommonsAttributionLicense,whichpermitsunrestricted use,distribution,andreproductioninanymedium,providedtheoriginalauthorandsourcearecredited. Data Availability: The authors confirm that all data underlying the findings are fully available without restriction. All data analyzed in this paper were downloadedfromGenBankandAccessionNumbersaregivenintheSupplementaryInformation. Funding:ThisworkwasfundedbyaChinaScholarshipsCouncil(www.csc.edu.cn)andtheUniversityofEdinburgh(www.ed.ac.uk)ScholarshiptoLL.SJLwas supportedbytheWellcomeTrust(www.wellcome.ac.uk;grantnumber092807).Thefundershadnoroleinstudydesign,datacollectionandanalysis,decisionto publish,orpreparationofthemanuscript. CompetingInterests:Theauthorshavedeclaredthatnocompetinginterestsexist. *Email:[email protected] Background Although predictors of such outbreaks have long been sought, surveillanceinwildbirdsinNorthAmericahasfailedtoprovidea Migratorybirdsaremajorcandidatesforlong-distancedispersal clear early warning signal. Three H7N3 HPAI events in poultry of zoonotic pathogens and low pathogenicity (LP), avian-origin haveoccurredinNorthAmericassince2000,and,inonecase,it influenza A viruses (AIVs) are widely distributed in free-ranging wasreportedthattheoutbreakH7N3AIVweretransmittedfrom waterbirds[1].Wildbirdsspreadtheirvirusestootherwildaswell poultry to humans [8]. Phylogenetic analyses indicated that each as domestic birds as they migrate through an area, allowing of these H7N3 HPAI strains had a close relationship with LPAI extensive reassortment [2]. Once introduced into poultry (espe- isolated from wild birds sampled in neighbouring provinces ciallychickensandturkeys),LPAImayswitchtohighpathogenic [9,10,11]. viruses(HPAI)withtheintroductionofbasicaminoacidresidues In June 2012, H7N3 HPAI outbreaks were found in poultry into the haemagglutinin cleavage site, which is associated with a farms in Jalisco state in Mexico, a region of high poultry density highmortalityrateinpoultry[3,4].Wehaverecentlyshownthata [12] and concurrent infections of humans with this HPAI A higher inter-subtype reassortment rate can be found in wild (H7N3) virus (2 cases) have been confirmed [13]. The outbreak Anseriformes than domestic Galliformes in the internal segments has beenaffecting broilers,breeders,layers andbackyard poultry of Eurasian AIV, indicating the wild bird population was the in the Mexican States of Jalisco,Aguascalientes, Guanajuato and source ofthenewreassortants, ratherthan domestic poultry [5]. Puebla: the latest outbreak reported by the World Organization Migrating wild birds have been implicated in the spread and for Animal Health (OIE) was on 19th May 2014. Ongoing emergence of HPAI such as HP H5N1 and H7N3. Viral epidemiological investigations have implicated contact with wild transmission between wild birds and domestic poultry, and birdsasafactorintheoutbreaks[12].However,thespecificorigin consequent genetic exchange, has contributed to genomic of the novel outbreak strain and its relationship to the previous reassortment which confounded disease control efforts [6,7]. outbreak strains isnot known. PLOSONE | www.plosone.org 1 September2014 | Volume 9 | Issue 9 | e107330 OriginofHighlyPathogenicAvianInfluenza(H7N3)inMexico The aim of our study was to investigate the origin of the 427 AIV strains collected over a 12 year period (2001 to 2012). precursor strain of the Mexico H7N3, using a Bayesian Thetimetothemostrecentcommonancestor(TMRCA)foreach phylogeographic inference framework by reconstructing the segmentofthenovelHPAIH7N3inMexicowasestimatedfrom spatiotemporal spread of AIVfromwildbirdsinNorthAmerica. the time-scaled phylogenetic trees. The HPAI H7N3 strains sampled in Mexico shared similar common ancestors among Results different genes between October 2011 and March 2012, i.e. duringthewinterof2011–2012(Table1).Thecommonancestor Phylogenetics of the HPAI H7N3 Mexico with north oftheHPAIH7N3Mexicooutbreakandtheclosestrelatedavian America AIV influenzastrainsexistedbetween1.1to3.9yearsago,whichvaried To investigate the origin of the AIV causing the HPAI H7N3 amongtheirdifferentgenomicsegments(Table1).Thedifference outbreakinMexicoin2012,aninitialphylogeneticanalysisusing in unsampled diversity among gene segments suggested that the Maximumlikelihoodwasperformedforeachsegmentofboththe reassortment of North American AIV lineages which led to the outbreaksequencesandabackgrounddatasetwhichcomprisedall H7N3Mexicooutbreakmayhaveinvolvedseveraleventsspread available AIV of North American AIV lineages (Figure 1). The over this time period. This can be seen by comparing the closest phylogenetictreesofallavailableH7segmentsinnorthAmerican related strains in the phylogenetic tree for any segment: they can showedthatAIVisolatedinrecentyearshavedivergedfromthose be quite distant from the H7N3 Mexico strain in the other before 1990 (Figure 1A). In addition, in the HA segment a sub- segments.Thisresultsupportsourhypothesisoftheoccurrenceof lineage mainly composed of H7N2 AIV from domestic birds in multiple reassortment events. New York state is clearly separate from the recent lineage Co-circulationofmultipleH7cladeswasobservedinHAacross composed of AIV from wild birds, which indicates extensive NorthAmerica.Interestingly,theHPAIH7N3Mexicostrainsare diversityofLPAIVinwildanddomesticbirds.Since2000,theN3 notrelatedinHAtotheHPH7N3outbreakinBritishColumbia NA segment of North American AIV has split into two separate in2004and2007,butinsteadarecloselyrelatedtoasubgroupof lineages (Figure 1B). The mechanism for maintenance of this H7 AIV (H7N3, H7N8 and H7N9) from wild waterfowl isolated divergence remains unknown as viruses from both lineages co- from Nebraska, Illinois, Missouri and Mississippi in 2010 and circulate in geographically overlapping host populations, mainly 2011. The mean estimate of the date of the common ancestor is wild waterfowl. February2010(Figure2).Ontheotherhand,thepictureinNAis Diversereassortmenteventsinvolvingthesixinternalsegments different:theclosestrelatedstraintothatoftheMexicooutbreakis canbeinferredfromthemaximumlikelihoodphylogeniesof2343 asubtypeH2N3AIVisolatedfromagreenwingedtealinIllinois NorthAmerican AIV.Clades identified inthephylogenyfor one in2010(Figure S2). segment(e.g.,PB2)arenotmaintainedinthephylogeniesofother Incontrast,thethreepolymeraseencodinggenesegmentsPB2, internal segments (Figure S1). In addition, internal segments of PB1 and PA of the Mexico outbreak strain belong to lineages AIV viruses isolated in distant locations can be closely related to composed mainly of AIV found in wild waterfowl in California each other within the same time period, which suggests not just fromthebeginningof2012(FigureS3,S4,S5),withsegmentsPB1 frequent reassortment but also rapid movement of influenza andPAhavingthesamemostcloselyrelatedstrain:A/American viruses acrossNorthAmerica. green-winged teal/California/123/2012 (H1N1). The other Sequences for time-scaled phylogenetic analysis were selected internal segments of the Mexico H7N3 strains have a different from the closest clades to the novel H7N3 HPAI viruses on the origin. From the Bayesian phylogenetic tree of the NP segment, maximumlikelihoodtreeofeachsegment.Thisdatasetcomprised theclosestAIVstraintoH7N3MexicoisanH11N9strainisolated Figure1.PhylogeniesoftheH7andN3segmentsofallavailableNorthAmericanAIV.A:HA.Sequencesingreyarefrombefore1990;the cladecoloredblueiscomposedofH7N2AIVisolatedfromasinglesurveillanceinpoultryinNewYork;thecladecoloredpinkwasselectedfortime- scaled phylogenetic analysis. B: NA tree. The uncoloured sequences are from before 2000; the AIV clade in pink was selected for time-scaled phylogeneticanalysis. doi:10.1371/journal.pone.0107330.g001 PLOSONE | www.plosone.org 2 September2014 | Volume 9 | Issue 9 | e107330 OriginofHighlyPathogenicAvianInfluenza(H7N3)inMexico Table1. Timeof themostrecent commonancestors forthe MexicoH7N3 virus. Gene TMRCAa Mostcloselyrelatedstrainb HA 20/3/2012(29/11/2011,23/5/2012) A/northshoveler/Missouri/2010(H7N3) NA 4/1/2012(10/8/2011,23/4/2012) A/Americangreen-wingedteal/Illinois/2010(H2N3) PB2 30/10/2011(8/6/2011,25/2/2012) A/mallard/California/198/2012(H11N9) PB1 10/10/2011(10/10/2010,28/1/2012) A/Americangreen-wingedteal/California/123/2012(H1N1) PA 18/11/2011(3/7/2011,26/3/2012) A/Americangreen-wingedteal/California/123/2012(H1N1) NP 21/1/2012(16/8/2011,5/5/2012) A/Americangreen-wingedteal/Mississippi/2012(H11N9) M 2/11/2011(7/0/2011,2/5/2012) A/Americangreen-wingedteal/Illinois/2008(H10N7) NS 25/10/2011(25/3/2011,17/1/2012) A/Americangreen-wingedteal/Illinois/2010(H10N7) aTimeofthemostrecentcommonancestors(TMRCA)ofeachsegmentofthenovelMexicoH7N3virusarerepresentedintheorderofdate/month/year.Thevaluesin parenthesesrepresentthe95%HPDintervals. bThestrainsareidentifiedarethosemostcloselyrelatedtotheoutbreakstrainsineachtreephylogenyinthisstudy. doi:10.1371/journal.pone.0107330.t001 fromMississippiin2012(Table1).TheNPsegmentoftheH7N3 domestic Galliformes (gal) such chickens and wild Anseriformes Mexico outbreak strain is also in the same lineage as a small (ans)suchasmallards,withAnseriformescomprisingthemajority numberofH7N7AIVstrainscarriedbynortherntealinIllinois/ of our AIV data set (n=366/427), see table S1. Among all four Missouri in the fall of 2010; these strains belong to the same strongly supported transitions with Bayes Factor (BF) .100 with lineage in the HA segment as well (see above and Figures 2 and mean diffusion rate (R) between 0.01 and 0.07, the highest S6).TheNSsegmentwasderivedfromanH10N7AIVwhichwas diffusion rate was found between Charadriiformes and Passer- alsocirculatinginthesameregionatthattime(FigureS8).Inthe iformes. The other three were found between Anseriformes and M segment, however, Mexico H7N3 strains are distinct from all other bird orders, and the HPAI H7N3 outbreak in poultry currently available AIV in North America, suggesting a surveil- (labelled as Galliformes Mexico) is linked to Anseriformes with lance gap(Figure S7). strongsupport(R=0.02,BF.100)(Figure 3AandTableS5).The These results indicate that the HPAI H7N3 virus that caused resultsconfirmthattherehasbeenextensivemixingofinfluenzaA theoutbreaksinMexicoisnotrelatedtoanyofthepreviousH7N3 virus betweendifferent ordersof birds, bothwildand domestic. HPAI outbreaks in North America, nor related to other AI Toexplorewhichhostspeciesmighthavebeenthedirectdonor outbreaks(HPH5N2outbreaksinMexicoin1994–1995,LPH7 of the Mexico outbreak strains, AIV belonging to Anseriformes outbreaks in Canada in 2009) in domestic birds in recent years were further divided into the five predominant species: mallard [14,15]. In addition, no clear pattern of association among the (Anas platyrhynchos), northern pintail (Anasacuta), northern segments of the Mexico H7N3 strains was observed, indicating shoveller (Anas clypeata), blue-winged teal (Anas discors) and green-wingedteal(Anascarolinensis)(TableS2).Speciesthatwere multiple segment exchange events occurred among North sampled at relatively low levels were combined as ‘‘other American influenza strains togiverisetoit. Anseriformes’’ (Table S2). Multiple statistically supported transi- tions (with R from 0.15 to 2.13, BF from 6 to over 100) were GeneflowoftheprecursoroftheHPAIH7N3outbreakin identifiedamongdifferenthostspecieswithinthisOrder,andboth Mexico mallard and green winged teal are linked to 3 other host species To further explore the origin of the Mexico outbreak strain, a (Figure 3B and Table S6). This analysis indicates the Mexican jointanalysisofdiscretetraitmodelswasperformedtoestimatethe outbreak strains were most likely to have been transmitted from overall genetic transmission process. In this the phylogenetic tree green winged teals (R=0.15and BF=6). space was sampled independently for each segment, while the The phylogeographic analysis for each segment of the Mexico transition pattern was jointly estimated in a single analysis as the HPAI H7N3 strain was summarised by a MCC tree in a diffusion parameters being applied in the discrete trait models geographiccontext.However,tovisualizetheevolutionprocessin were assumed to be the same (see methods). Four major factors a spatiotemporal mode we converted the spatial annotated time- including seven specific traits were tested by implementing scaled phylogeny to an annotated map (Figure 4 and Figure S9). Bayesian stochastic search variable selection (BSSVS): i) host Five segments (HA, NA, NP, M, and NS) of the Mexico HPAI population of AIV (order/species); ii) geographic location of H7N3 strain were introduced directly from different states in sampled AIV (bird migration flyways/provinces and states of centralUS,whilePB2,PB1andPAwereintroducedfromstatesin North America); iii) subtype of AIV and iv) virulence (pathoge- the western region. The introductions of segments from several nicity/cleavage sites). For each trait, the evolving process of the different geographic locations indicate multiple reassortment HPAIH7N3inMexicoandcloselyrelatedAIVcanbeseenfrom events were likely to have been involved in the generation of the the reconstructed time-scaled phylogeny of each segment inde- novelH7N3 MexicoAIV. pendently (with exception of pathogenicity and cleavage sites Joint discrete trait analysis of all eight segments indicated whichonlyappliedtotheHAsegment),withthebranchescolored frequent gene transfer among locations (states and provinces) by thespecific traitaccordingtotheancestor traitintheinternal wherethebackgroundAIVsequenceswereisolated.However,in nodes(Figure 2 andFigure S2,S3,S4,S5,S6,S7,S8). this initial analysis, no significant support was found between Host populations of AIV in this study were first analysed by Jalisco(theoutbreakstate)andanyotherlocation,probablydueto Order: wild Charadriiformes (cha) such as gulls, wild Gruiformes thelargenumberofpossibletransitions(26states,325irreversible (gru) such as cranes, wild Passeriformes (pas) such as sparrows, transition pairs) and the limited number of outbreak strains (3; PLOSONE | www.plosone.org 3 September2014 | Volume 9 | Issue 9 | e107330 OriginofHighlyPathogenicAvianInfluenza(H7N3)inMexico Figure2.Maximumcladecredibility(MCC)phylogeniesfortheHAsegment.Branchesarecolouredaccordingtothe4discretetraits(host order,hostspecies,flywayandlocation)oninternalnodes.Mexicanoutbreakstrainsarehighlightedwithpink.A:Hostorder.Fivehostordersare labelled on HA tree: wild birds of the order Anseriformes (ans-wild); wild birds of the order Charadriiformes (cha-wild); wild birds of the order Passeriformes(pas-wild);domesticbirdsoftheorderGalliformesandMexicoH7N3outbreakintheorderGalliformes(gal-domestic-Mexico).B:Host species.WildAnseriformesareclassifiedintothefivemainspeciesandagroupcomprisingtheotherrarerspeciesofAnseriformesinthisstudy: mallard(Anasplatyrhynchos),northernpintail(Anasacuta),northernshoveller,blue-wingedteal,green-wingedtealandotherAnseriformes(other ans);TheorderGalliformesareshownas‘‘outbreak’’(theH7N3Mexicooutbreak)and‘‘other_gal’’;Theotherordersareshownas:Charadriiformes (cha)andGalliformes(gal),Gruiformes(gru)andPasseriformes(pas).C:Flyway.FourspecificNorthAmericanflywaysarelabelledontheHAtree:the Atlantic,Mississippi,Central,andPacific.D:State.22statesandprovincesoftheviralsamplelocationsarelabelledontheHAtree.TheoriginalMCC treefileswithalltaxanamesaredepositedinDryad(doi:10.5061/dryad.j5bf8),andtreesfortheother7segmentswithouttaxanamescanbefoundin FigureS2,S3,S4,S5,S6,S7,S8. doi:10.1371/journal.pone.0107330.g002 TableS3).Previousstudieshaveshownthatincorporatinglocation summarized in Table S4, showing a wide range in rate and greatly improves phylogeographic descriptions of the pattern of statistical support (R=0.02 to 0.25; BF=3 to .100). Highly virus gene flow [16,17]. Therefore, we enhanced the statistical significant links [BF.100, Indicator (I) =1] were found between poweroftheanalysisintwoways:firstbydecreasingthenumber majorNorthAmericanflyways,particularlybetweenAtlanticand oflocationsbycombiningstatesandprovincesintomajorregions Mississippi (R=0.17 exchange/year); Mississippi and Pacific (flyways) and secondly by reducing the number of pairwise (R=0.22 exchange/year), Central and Atlantic (R=0.25 ex- transitions possible. change/year) and Central and Pacific (R=0.18 exchange/year) Toaggregatelocationsweusedtheknownmigrationroutes,or (Figure 5AandTable S7). Linkagesbetween Atlantic and Pacific ‘‘flyways’’: Atlantic, Mississippi, Central, or Pacific (Figure S10). flyways, Mississippi and Central flyways were also identified The distributions of avian influenza virus for each flyway are althoughwithweakersupport(3,BF,6).Thetransitionsbetween PLOSONE | www.plosone.org 4 September2014 | Volume 9 | Issue 9 | e107330 OriginofHighlyPathogenicAvianInfluenza(H7N3)inMexico Figure3.InferredhosttransmissionnetworksofMexicanoutbreakAIV.A:Hostorder.Nodelabelsinthenodesarehostordersidentified following the abbreviations used in the colored phylogenetic trees (Figure2 and Figures S2, S3, S4, S5, S6, S7, S8): wild birds of the order Anseriformes(ans-wild);wildbirdsoftheorderCharadriiformes(cha-wild);wildbirdsoftheorderPasseriformes(pas-wild);domesticbirdsofthe orderGalliformesandMexicoH7N3outbreakintheorderGalliformes(gal-domestic-Mexico).Arrowsshowthedirectionoftransmissionbetweentwo hostorders;thearrowweightandthenumberaboveeacharrowindicatesthepercapitatransmissionrate.NodesizereflectsthenumberofAIVfor eachhostorder(TableS1).LinecoloursindicatetheoverallBayesFactortestsupportforepidemiologicallinkagebetweenhostorders,Redlines indicatestatisticalsupportwithBF.100(verystrongsupport),darkpinklinesindicatesupportwith30,BF,100(strongsupport),pinklinesindicate support with 3,BF,30. B: Host species (Anseriformes only). Wild Anseriformes are further classified into the five main species and a group comprisingtheotherrarerspeciesofAnseriformesinthisstudy:mallard(Anasplatyrhynchos),northernpintail(Anasacuta),northernshoveller,blue- wingedteal,green-wingedtealandotherAnseriformes(otherans);Asabove,arrowsshowthedirectionoftransmissionbetweentwohostspecies; thearrowweightandthenumberaboveeacharrowindicatesthepercapitatransmissionrate.NodesizereflectsthenumberofAIVforeachhost species(TableS3).LinecoloursindicatetheoverallBayesFactortestsupportforepidemiologicallinkagebetweenhostspecies,Redlinesindicate statisticalsupportwithBF.100(verystrongsupport),darkpinklinesindicatesupportwith30,BF,100(strongsupport),pinklinesindicatesupport with3,BF,30. doi:10.1371/journal.pone.0107330.g003 flyways and the H7N3 HPAI in Mexico are not that strongly explain the gene flow of North America AIV than a model supported compared to the between flyway transitions, probably incorporating thebetweenflyway transitions. due to the limited number of sequences available from the Considering individual locations, we found 11 locations were outbreakAIV,butthreedirectdonorsamongtheseflywaystothe linked to the HPAI H7N3 in Mexico, among which 7 showed predecessor of the Mexico outbreak were identified: the Pacific, significantlystronglinks(BF.100).Thesewere:Alaska(R=0.59), CentralandMississippiflyways.Amongtheseflows,thetransition Alberta (R=1.07), California (R=0.59), Illinois (R=1.61), ratefromMississippi(R=0.05exchange/year)isthemoststrongly Missouri (R=2.26), Ohio (R=0.66) and Wisconsin (R=1.39), supported with BF=86. In comparison, the link with the Pacific belonging to the Pacific, Central and Mississippi flyways. AIV (R=0.02 exchange/year) and Central (R=0.04 exchange/year) from other four states/provinces (Minnesota, New Brunswick, were weaker (BF=4; Figure 5B and Table S7). There was no Quebec and Washington) are also linked to Mexico H7N3 AIV supported link between the Atlantic flyway and Mexico. The but with weaker support (BF=21 to 30) and lower rate (0.7 to results indicated that the HPAI H7N3 in Mexico probably 0.84) (Figure 5A and Table S8). This result indicates the donor originatedfromAIVtransmittedbywildbirdsfromthreedifferent locations of the Mexico outbreak are spread widely across North flyways. America. Giventhesefindingsitappearedlikelythattheprecursorstrains In addition, an extremely complex pattern of linkage between were generated somewhere near Mexico as it is the place where the 52 ancestral subtypes (Table S9) was identified, which birds from the different migration routes meet during winter. To confirmed the extent of the reassortment events which had test thishypothesis,wefurther reduced thenumber oftransitions occurredbetween,especiallyintheinternalsegments(fulltreewith in the locations transition matrix: transitions between two flyway annotation can be found on the Dryad Digital Repository: regionswereswitchedoff(byforcingtheinitialindicatorofagiven doi:10.5061/dryad.j5bf8).However,asforlocations,nosignificant transition pair from1 to0,so that this transitionpair willnot be linkage between H7N3 in Mexico and other subtypes was counted), and those for within flyway transitions and transitions identified due tothelarge number ofcandidate donor subtypes. linked to Mexico were maintained. There are 98 non-reversible MexicoH7N3isconfirmedinthephylogenetictreesoftheHA transitionpairsinthenewreducedmatrix(TableS10).AICMtests segment (Figure S11) to have mutated from a LPAI (low pathogenic avian influenza) virus to HPAI after the ancestral (seeMethods)revealedthatthenon-reversibleBSSVSmodelwith virus was introduced into poultry from wild birds (Figure S11A), reducednumberoftransitionswassignificantlyfavouredoverthe rather than being associated with previous HP outbreaks. The other models with the original matrix (Table2), indicating the virulence of HP avian influenza viruses is associated with the numberoftransitionshasaneffectontheperformanceofdiscrete appearance of an insertion of multiple basic amino acids at the trait models. This reduced model has better support than a cleavage site of the HA protein [18]. Categorizing the cleavage randomlyreducedmodelwiththesamenumberoftransitionpairs sites in this dataset into three types: 1) Insertion, 2) Partial andthesamenon-reversibleBSSVSsetting(Table 3);weconclude insertion,3) No insertion (Figure S11 C), we find that the H7N3 that gene transitions within flyways and Mexico alone better Mexicostrainhasauniqueinsertion-DRKSRHRR-compared PLOSONE | www.plosone.org 5 September2014 | Volume 9 | Issue 9 | e107330 OriginofHighlyPathogenicAvianInfluenza(H7N3)inMexico Figure4.SpatialdiffusionofAIVsegmentsoftheMexicooutbreakAIV.Thefirstthreepanelsrepresentthreesegmentsseparately(A:HA,B: NA,C:PB2)andDrepresentsthespatialtransmissionofall8segmentsjointly.TheplottedlinesrepresentthebranchesoftheMCCtreesfordifferent segments,distinguishedbycolor;thesizeofeachcirclerepresentsthenumberoflineageswiththatlocationstate.Themapsourceforthisfigurewas OpenStreetMap(http://www.openstreetmap.org/).Thespatialdiffusionsofotherfivesegments(PB1,PANP,MandNS)onthemapareshownin FigureS9. doi:10.1371/journal.pone.0107330.g004 to other HPAI strains (Figure S11 C). We conclude the Mexico Ourstudyconfirmedtheassumptionsofearlierstudiesbasedon H7N3 strains originated from a lineage composed of LP strains the HA segment that theoutbreak strain derived from wild birds with partial insertions in the cleavage sites. Similarly, the H7N3 [12,19,20]. We have now shown using powerful Bayesian strainswhichcausedanoutbreakofHPAIinCanadain2004were phylogenetic methods that the origin of the HA segments of the also derived from LP strains with the same partial insertions, but MexicoH7N3strainscanbedatedtoMarch2012,andthatthey fromacompletelyseparateHAclade,indicatingparallelevolution fall into a subgroup of H7 AIV (H7N3, H7N8 and H7N9) from with respect to the acquisition of the multibasic cleavage site, wild waterfowl isolated from Nebraska, Illinois, Missouri and starting fromdifferent lineages(Figure S11 B). Mississippi with a common ancestor around February 2010. We have extended this analysis to all eight segments, thus obtaining Discussion thecompleteevolutionaryhistoryoftheoutbreakavianinfluenza viruses. Wehaveinvestigatedtheoriginoftherecenthighlypathogenic The predecessors of HPAI H7N3 in Mexico were transmitted H7N3 outbreaks in Mexico. Our analysis found that the from migrating waterfowl in North America. Previous cases of progenitoroftheHPAIH7N3wasareassortantviruswithseveral periodic transmission of H7N3 viruses from wild birds to different origins among the eight segments. We also found that gallinaceous poultry in the Americas suggests that these viruses gene segments of AIV in North American wild birds are continuously circulate in wild birds, and their propensity to exchanged at a very high frequency, with no evidence of any become highly pathogenic after transmission suggests that they restriction whichmightimply linkage of segments. have a gene constellation conducive to generating pathogenic PLOSONE | www.plosone.org 6 September2014 | Volume 9 | Issue 9 | e107330 OriginofHighlyPathogenicAvianInfluenza(H7N3)inMexico Figure5.InferredphylogeographictransmissionnetworksofMexicanoutbreakAIV.A:Flyway.AIVtransmissionamong4N.American flywayswithlinkstotheMexicanoutbreakstrains.Arrowsshowthedirectionoftransmissionbetweentwoflyways;arrowweightandthenumber above each arrow indicates the per capita transmission rate. Node size reflects the number of AIV for each flyway (Table S6). B: Location. AIV transmissionamongstates/provincesinNorthAmericaandJalisco(theMexicanstatewheretheoutbreakstrainswereisolated).Arrowsshowthe directionoftransmissionbetweenthetwostates;thearrowweightandthenumberaboveeacharrowindicatesthepercapitatransmissionrate. NodesizereflectsthenumberofAIVforeachflyway(TableS5). doi:10.1371/journal.pone.0107330.g005 variants [21]. H7N3 has been responsible for all lethal influenza major ecological barriers, including spatial distance and avian outbreaks in poultry in the Americas over the past decade [21]. flyway[17].EarlierstudiessuggestedthatAIVexhibitsastrongly ExperimentalstudieshavealsoindicatedthatH7influenzaviruses spatially structured population in North America, with relatively from the North American lineage have acquired sialic acid- infrequentgeneflowamonglocalitiesandespeciallybetweenthose binding properties that more closely resemble those of human that are spatially distant or belong to different flyways using influenzavirusesandhavethepotentialtospreadtonaiveanimals phylogeographic analysis [27]. This hypothesis was supported by [22]. In parallel, H7 influenza viruses from East Asian migratory studies showed that AIV isolates from mallard were linked by waterfowl were introduced into domestic ducks in China on migration between sites in central Canada and Maryland but several occasions during the past decade and subsequently limited reassortment occurred along the inter-migratory flyway reassorted with enzootic H9N2 viruses to generate a novel routes [28]. However, more recently, the opposite was seen in a H7N9 influenza A virus, resulting in 44 human deaths in China anotherstudy,whichemphasizedthatthelong-termpersistenceof (WHOreportedinDec.3rd2013)sinceitsfirstdetectioninMarch theinfluenzaAvirusgenepoolinNorthAmericanwildbirdsmay 2013[23].TheseresultsindicatedthatAIVofH7subtypescarried be independent of migratory flyways, and the short-term by wildbirds arepotential threat tomammalian hosts. evolutionary consequences of these ecological barriers may be EarlierstudiesshowedthatshorebirdsandgullsintheAmericas rapidly erased by East-Westvirus migration [29]. are more frequently the source of the potential precursors to HP We also found there are genetic interactions between flyways, H5 and H7 avian influenza viruses, while in Eurasia, the usingasimilardiscretetraitmodel.However,tofindthestrongest precursors of HP influenza viruses are usually from duck species link between the Mexican outbreak and potential precursors we [24,25]. However, we found that wild Anseriformes (ducks and found the model had more power when we switched off the geese)weretheoriginoftheprecursorofHPAIH7N3inMexico. between flyway transitions, keeping only the links within flyways Anseriformes showed substantial diversity of AIV in North and with Mexico. We found that gene flow from three flyways America, and were divided into five avian species groups in our (Pacific,CentralandMississippi)generatedthereassortantswhich dataset – mallards, northern pintails, northern shovellers, blue- acted as the predecessor of HPAI H7N3 in Mexico, and it is winged teals, green-winged teals, which have specific ranges for possible that the reassortment events occurred in Mexico or breeding, migration and wintering [17]. We found the green fartherafield.Flywayboundariesarenotsharplydefinedandboth winged teal was the species most strongly supported as thedirect in the northern breeding grounds and the southern wintering donor of the predecessor HPAI H7N3 in Mexico. This species groundsthereisoverlaptosomedegree.Forexample,inPanama nests as far north as Alaska, and migrates along all four flyways. parts of all four flyways merge into one (http://www.birdnature. However, the genetic transitions of different segments showed a com/flyways.html).Birdsthatarelong-distancemigrantstypically complicated interactioninvolving different bird species. haverangesthatextendfromtheUnitedStatesandCanadainthe Migratingbirdsmayexchangeviruseswithotherpopulationsat summertoMexicoandfurthersouthinthewinterandnearlyall staging, stopover or wintering sites [26]. Many studies have been ofthemigratorybirdsoftheeasternUnitedStates,aswellasmany performedonAIVgeneflowsinNorthAmericaduringwildbird western species, use the western Mexican Gulf during migration migration:Onesuchrevealedthatavianinfluenzavirusexhibitsa [30]. stronglyspatiallystructuredpopulationinNorthAmerica,andthe Whiletheresolutionanddetectionofmigrationeventshasbeen intra-continental spread of AIV by migratory birds is subject to enhanced through increased surveillance in recent years, critical PLOSONE | www.plosone.org 7 September2014 | Volume 9 | Issue 9 | e107330 OriginofHighlyPathogenicAvianInfluenza(H7N3)inMexico n m information for wild bird surveillance remains sparse. Only one u ol AIV in Mexico has been published (A/cinnamon teal/Mexico/ c Mod8 6582 6475 6091 5870 5719 5480 2498 2 edinthe 2oM8f1tsh7ee/g2mei0ge0hn6tt)soaefngmdtheietnitHss7n(dNoat3tareonluaotttebdsrhetaookwtngh)ee.nWneeewwfaoosuundtbdisrtteihnaecktovfrirrioguimsnionoftahtnheyer at c North American AIV in the dataset, as seen in the phylogenetic di Mod7 4084 3977 3593 3372 3221 2982 2 22498 eother(in tparnreedecse.Sd2iIn,ngS3tah,dedSi4oti,uotSnb5,r,eatShk6e,grrSoe7ula,ptSivi8ne)lyosuthggegrreeasgttesenrtehlesereneggmtmheanyotsfb(eFbirmgaunirsceshisne2gs th intermediates, possibly through insufficient AIV surveillance in ver CentralandSouthAmerica.Togetherwithpreviousphylogenetic o Mod6 1102 995 611 390 239 2 22982 25480 umn1) sstaumdpielsinwgh(iwchithalsroemcoerndtioonfesdamthpeliinmgpotirmtaencaenodffliollcinatgiogna)psinintvhiersael ol regions[31,32],thishighlightstheneedforincreasedsurveillance c n inthose regions. d. (i Mod5 863 756 372 151 2 2239 23221 25719 edinbol nemodel hHoPOstAvdIeisrHtarl7ilb,Nub3tyiocAnoImaVnbditnheiacntogloctaghuyesipendhoyaulorgsaeenrniaeeltsyicsoishf,iswptooeursylhtrooyfwAotIuhVteborweriiatghkisnthionef at o Mod4 712 605 221 2 2151 2390 23372 25870 erformanceisindic ntsthesupportfor MwtCmAhaIeereVtnonxeuttirrcgfiaeoonhlvwoeAwilunssmtitlsdfae.rtnrohOibmceinaurodrtdmivsmirefieflageesnhurrdteleptnasbetisrenasioomdraerdiictgapaurnsotsatetetutefidunoccilnoateirmadfrolf,iplreyhrdaweoidnhatesdbeypn,ynsotsitmiifmvnyfeiooingrNrgersuaoAttrrhiiIvtmoVheenpiltAlrohaoerrmnafteacsaesnewrtotiirlciolytnda-f, p e Mod3 490 384 2 2221 2372 2611 23593 26091 withthebest nacellrepres SMoeutthhaonddsCentral Americaishighly desirable. Mod2 107 2 2384 2605 2756 2995 23977 26475 data.Themodel hepositivevaluei DM1p2uaeT2btx8lahii3sceh/opec2rdo0em1pi(nA2paf;/llrueacAtehteni/ioczgMkaneeneAnox/imcvJoaier/luiosIscnfosDte/hqRCrueEPee7Ano2c1u1e/ts8b2/or02ef1a0N2k1;o2sr)ttrhAaain/nAdcsmhoieacfrlkliHecnap7n/rNeJl3vainiloifesrucaosogml/ye AICMestimatesforthefitofdifferentdiscretetraitmodels.Table2. bcdJointDiscretetraitmodelsAICMS.E.Mod1 a22Mod1Reduce_BSSVS_sym111031/7.5+ 22Mod2Reduce_nonBSSVS_sym111137/9.449107+ 22Mod3Reduce_BSSVS_asym111521/11.798490+ 22Mod4Reduce_nonBSSVS_asym111742/7.822712+ 22Mod5Original_BSSVS_sym111893/6.668863+ 22Mod6Original_nonBSSVS_sym112132/13.4551102+ 22Mod7Original_BSSVS_asym115115/14.6624084+ 22Mod8Original_nonBSSVS_asym117612/26.5846582+ aThenamesofthe8models(Mod1-8)inthecomparisontest.bTheestimatedAICMscoreoftheposterior:lowervaluesofmarginallikelihoodindicateabetterfittothecThestandarderroroftheAICMestimatedusing1000bootstrapreplicates.dTheAICMcomparisonsareshowninthematrixcomposedofcolumns4to11.Ineachrowofthematrix,ttitles).AdifferenceofAICM=10isconsideredtoindicateastrongpreferenceforonemodeloveranother.doi:10.1371/journal.pone.0107330.t002 (1un[5csdasictstooti(ToHods((obcn29nnooheeee3oovs0fvnursiiaeoh28gvgqsqtd7it4idamn=etc0testntmeeemm72iuuib]hetirrksMcricaglrrh,p1id2tseerrtagbaepabgeeeracea;eennnot8eecrbaoutnnlehuMrqnbanioneesatt9ccarl1roencttti)heutkeltodeeucctr).;oie2e)ttescU,dioeor;esshhves.nrfesneHetnsaynsdeewyPgFSdTtrft1ltfg(nolhsce2oaeramoa7ACohp3beithemadfaaHe0Atrpslrthrsni1ies,eiLeeihmr1losapgneIptacfte7oents.neNE3huVnnnwehHpitseloaNmtngio.WeHcrdsm=f.egeiSvcenfohitnAcyi3vrseomohaSeehe,gTAlmeriei8iesN3uulnnlsnpoeeaHnyqleo(9;nsh.oodtqtarHdog35me;entubttetnwPjeu=thsadgM(holotesr7h2W(tadeoltselA[nmetis2eenayrfeo)1f4eu3:ncirLtadi0fb)nIcnGrhsht422e3asedicNk0yeesnlgeganris8tves3]eiytwTsll1iqsieete.lrAieNets(o4oxmhhr8nengMeuntsRacwf3dnMutwheteA3rmoetterttoo(ra;sesedesN)heoeer8fniear;teaGrtNndeaxerfs2reficase3exfNanautnatoejsdri0lenAosoeeca)ielyoe2tfibrsteo,smss1PtrnroosifgcghwMn2oet)tnuwt.n2tnt,hr.seimyaahluh7hnatr,lHa)eontnppnrMlFn=7emeTyehacceloegfieleraol7Ags;l=tnoneaner1ohsuAioenrr4Nenanaa,tdntsmdl0vlaeddnm7asnse2yeidtultieii3wk0ttyeeslde9vegso7detaehcyaeoufthdpsdremrmherrdtls(ssielxnnb(eisrlee1toAsaietid1acaerishiussoerer;sea4gimmnat4eveygIoattefnofstlrim1irVneeos9eiaminiftdatsotudotnnse0ohsirdurrs4uthieaendmgmeeg,erehneTtsbnsgteeon(fesmainietnreast1oont)mxtrRaf,(rtsa.r;tltystaM6ar)nestaibshe)Gescc;pkwHrmnAeo;T9inwielPlhtleneBenPneaLeiroaem8gMetxrasAmaBelst)twsatlBns)rmieMmbeiytssSheo1cnesifiBPet2hdscflpnosoo9kuaoerd,ieteoliLs,,toaina)yhsfnacr.gnnosecner;nletnnneiSllrcyhltdlldbtnouivooe,esodktlg9l(e===haa7whdorcchtutanw.Nahdsdatt.titrgeeowinno0g368hheeeceaIeteraesdeeAshhhnndddddd4976leeeeeerrss-t.l PLOSONE | www.plosone.org 8 September2014 | Volume 9 | Issue 9 | e107330 OriginofHighlyPathogenicAvianInfluenza(H7N3)inMexico A of e). Ran10 407822 407804 407752 407714 407693 407664 407664 407659 407651 407301 2 menumber columntitl Ran9 521 502 451 413 391 363 363 357 350 2 2407301 whichthesa theother(the elsin over 8 0 7651 mod mn1) Ran 171 153 101 63 42 13 13 8 2 235 240 0are colu AICMestimatesforthefitofdifferentmodelswithreducednumberoftransitions.Table3. JointreducedabcdAICMS.E.FlywayRan1Ran2Ran3Ran4Ran5Ran6Ran7models 22/5.21870108129157158163Flyway111031+ 222Ran1111049/4.4185290111139139145+ 2222Ran2111101/5.970523859878893+ 22222Ran3111138/14.9108903821505055+ 222222Ran4111160/6.51291115921282834+ 2222222Ran5111188/5.615713987502806+ 2222222Ran6111188/4.315813988502806+ 222222222Ran7111194/12.816314593553466+ 222222222Ran8111201/8.6171153101634213138+ 222222222Ran9111551/9.9521502451413391363363357+ 222222222Ran10518853/3495.6407822407804407752407714407693407664407664407659+ aThenamesofthe11models(FlywaymodelandRan1-10)inthecomparisontest:Model1isthenon-reversibleBSSVSmodelwithoutbetweenflywaytransitions(Flyway);Ran1-1transitionshasbeenswitchedoff,chosenatrandom(Ran1-10).bTheestimatedAICMscoreoftheposterior:lowervaluesofmarginallikelihoodindicateabetterfittothedata.Themodelwiththebestperformanceisindicatedinbold.cThestandarderrorofAICMestimatedusing1000bootstrapreplicates.dDifferencesofmodelsareshowninthematrixbeingcomposedbycolumn4to13.Ineachrowofthematrix,thepositivevalueinacellrepresentedthesupportforonemodel(indifferenceofAICM=10isconsideredastrongpreferenceforonemodeloveranother.doi:10.1371/journal.pone.0107330.t003 PLOSONE | www.plosone.org 9 September2014 | Volume 9 | Issue 9 | e107330