The Interrelationships of Land Plants and the Nature of the Ancestral Embryophyte PDF
Preview The Interrelationships of Land Plants and the Nature of the Ancestral Embryophyte
This is a repository copy of The Interrelationships of Land Plants and the Nature of the Ancestral Embryophyte. White Rose Research Online URL for this paper: http://eprints.whiterose.ac.uk/128197/ Version: Published Version Article: Puttick, M.N., Morris, J.L., Williams, T.A. et al. (8 more authors) (2018) The Interrelationships of Land Plants and the Nature of the Ancestral Embryophyte. Current Biology, 28 (5). pp. 733-745. ISSN 0960-9822 https://doi.org/10.1016/j.cub.2018.01.063 Reuse This article is distributed under the terms of the Creative Commons Attribution (CC BY) licence. This licence allows you to distribute, remix, tweak, and build upon the work, even commercially, as long as you credit the authors for the original work. More information and the full terms of the licence here: https://creativecommons.org/licenses/ Takedown If you consider content in White Rose Research Online to be in breach of UK law, please notify us by emailing [email protected] including the URL of the record and the reason for the withdrawal request. [email protected] https://eprints.whiterose.ac.uk/ Article The Interrelationships of Land Plants and the Nature of the Ancestral Embryophyte Graphical Abstract Authors MarkN.Puttick,JenniferL.Morris, TomA.Williams,...,HaraldSchneider, green algae ggrreeeeeennn aallggaaee DavidePisani,PhilipC.J.Donoghue lhmivooersnrwswoorrtt “bryophytes” mhmlmlhmiivvooooeerrssnnnrrwwwsswwwooooorrrrtttt bryophytes“bryophytes” Correspondence Tracheophyta TTrraaaccchhheeoopphhyyta [email protected](H.S.), [email protected](D.P.), [email protected](P.C.J.D.) ggrreeeeeenn aallggaaee green algae ggrreeeeeennn aallggaaee In Brief lmlhlhmiiivvvooooeerrsssnnrrwwssswwwoooorrrrtttt ypy“bryophytes” llhmiivvooeersnrrrwwswooorrrttt Bryophyta lhmhhmlhhiivvooooooeerrssnnrrwwwssswwooooorrrrtttt bryophytes“bryophytes” Pusunupitttpiincogkrtleivftoearrlwb.rroeyrsotsoplhavynetdeamm‘‘Sooesntsoaeppshhayyntlyad.’’Tchlaedire TTrrraaaccchhheeeoopphhyyta Tracheoppphhyyttaa TTrraaccchhheeeoopphhyytta resultsindicatethattheancestralland plantwasmorecomplexthanhasbeen envisagedbasedonphylogenies green algae ggrreeeeennn aallggaaee recognizingliverwortsasthesister lhmivooersnrwswoorrtt “bryophytes” hmhmhhllliiivvvoooooeeerrrssnnrrrwwwssswwooooorrrrrttttt “bryophytes”“bryophytes” lineagetoallotherembryophytes. Tracheophyta TTrraaaccchhheeeooopphhyyttaa Highlights d Earlylandplantrelationshipsareextremelyuncertain d Weresolvethe‘‘Setaphyta’’cladeofliverwortsplusmosses d Thesimplebodyplanofliverwortsresultsfromlossof ancestralcharacters d Theancestrallandplantwasmorecomplex Putticketal.,2018,CurrentBiology28,733–745 March5,2018ª2018TheAuthor(s).PublishedbyElsevierLtd. https://doi.org/10.1016/j.cub.2018.01.063 Current Biology Article The Interrelationships of Land Plants and the Nature of the Ancestral Embryophyte MarkN.Puttick,1,2,3,9JenniferL.Morris,1,4,9TomA.Williams,2CymonJ.Cox,5DianneEdwards,4PaulKenrick,6 SilviaPressel,3CharlesH.Wellman,7HaraldSchneider,3,8,*DavidePisani,1,3,*andPhilipC.J.Donoghue1,10,* 1SchoolofEarthSciences,UniversityofBristol,LifeSciencesBuilding,TyndallAvenue,BristolBS81TQ,UK 2SchoolofBiologicalSciences,UniversityofBristol,LifeSciencesBuilding,TyndallAvenue,BristolBS81TQ,UK 3DepartmentofLifeSciences,TheNaturalHistoryMuseum,CromwellRoad,LondonSW75BD,UK 4SchoolofEarthandOceanSciences,CardiffUniversity,MainBuilding,ParkPlace,CardiffCF103AT,UK 5CentrodeCieˆnciasdoMar,UniversidadedoAlgarve,Gambelas,8005-319Faro,Portugal 6DepartmentofEarthSciences,TheNaturalHistoryMuseum,CromwellRoad,LondonSW75BD,UK 7DepartmentofAnimalandPlantSciences,UniversityofSheffield,AlfredDennyBuilding,WesternBank,SheffieldS102TN,UK 8CenterofIntegrativeConservation,XishuangbannaTropicalBotanicalGarden,ChineseAcademyofSciences,Menglun,Yunnan,China 9Theseauthorscontributedequally 10LeadContact *Correspondence:[email protected](H.S.),[email protected](D.P.),[email protected](P.C.J.D.) https://doi.org/10.1016/j.cub.2018.01.063 SUMMARY and the fossil record, and prompts re-evaluation of the phylogenetic affinity of early land plant The evolutionary emergence of land plant body fossils, the majority of which are considered stem plans transformed the planet. However, our under- tracheophytes. standing of this formative episode is mired in the uncertainty associated with the phylogenetic rela- tionships among bryophytes (hornworts, liverworts, INTRODUCTION and mosses) and tracheophytes (vascular plants). Theevolutionaryemergenceoflandplantbodyplansisoneof Hereweattempttoclarifythisproblembyanalyzing the most formative episodes in the evolution of our planet alargetranscriptomicdatasetwithmodelsthatallow [1, 2]. Land plant innovations, including stomata, vascular and for compositional heterogeneity between sites. rootingsystems,symbioseswithfungi,and,eventually,leaves, Zygnematophyceae is resolved as sister to land expanded the sequestration of CO through photosynthesis 2 plants, but we obtain several distinct relationships and silicate weathering [1]. These weathering effects resulted between bryophytes and tracheophytes. Concate- in changes in atmospheric CO over long timescales [1, 3, 4], 2 natedsequenceanalysesthatcanexplicitlyaccom- aswellaskeychangestotheenvironmentsuchasthedevelop- modate site-specific compositional heterogeneity ment of soils [5]. Early land plants have even been invoked in give more support for a mosses-liverworts clade, shapingterrestriallandscapesbyconstrainingsedimentological ‘‘Setaphyta,’’ as the sister to all other land plants, processes [6]. Unfortunately, a detailed understanding of this episodeisobscuredbyuncertaintyassociatedwiththephyloge- and weak support for hornworts as the sister to all netic relationships among bryophytes (hornworts, liverworts, other land plants. Bryophyte monophyly is sup- and mosses) and tracheophytes (vascular plants), for which ported by gene concatenation analyses using almosteverypossiblesolutionhasbeenproposed(e.g.,[7];Fig- models explicitly accommodating lineage-specific ures1A–1G).Intheabsenceofphylogeneticresolution,itisnot compositional heterogeneity and analyses of gene possibletoestablishthesequenceinwhichembryophyte,bryo- trees. Both maximum-likelihood analyses that phyte,andtracheophytebodyplancharacterswereassembled. compare the fit of each gene tree to proposed spe- This is a prerequisite for determining their intrinsic molecular ciestreesandBayesiansupertreeestimationbased developmental causes and extrinsic environmental conse- ongenetreessupportbryophytemonophyly.Ofthe quences,thephylogeneticinterpretationoffossilembryophytes, 15 distinct rooted relationships for embryophytes, and,consequently,establishingthetimescaleoverwhichthese we reject all but three hypotheses, which differ only characters evolved—facilitating tests of hypotheses on their roleintransformingtheplanet. in the position of hornworts. Our results imply that Although the monophyly of Embryophyta (land plants) and the ancestral embryophyte was more complex than Tracheophyta(vascularplants)isuniversallyaccepted,various has been envisaged based on topologies recog- hypothesesontheinterrelationshipsofbryophytes(hornworts, nizing liverworts as the sister lineage to all other liverworts,andmosses)andtracheophyteshavegainedsupport embryophytes.Thisrequiresmanyphenotypicchar- inthelast30years.Thefundamentaldistinctionisbetweenbryo- acterlossesandtransformationsintheliverwortline- phyte monophyly (e.g., [7–11]; Figure 1A) and paraphyly (e.g., age, diminishes inconsistency between phylogeny [12–17]; Figures 1B–1G). Indeed, at least seven alternative CurrentBiology28,733–745,March5,2018ª2018TheAuthor(s).PublishedbyElsevierLtd. 733 ThisisanopenaccessarticleundertheCCBYlicense(http://creativecommons.org/licenses/by/4.0/). green algae hornwort Bryoliverwortphyta moss Tracheophyta green algae hornwort“bryoliverwortphytes”moss Tracheophyta green algae hornwort“bryoliverwortphytes”moss Tracheophyta green algae liverwort“bryomossphytes”hornwort Tracheophyta A B C D Bryophyta Hornwort-sister Hornworts-sister Liverworts-Mosses sister (i) Bayesian & mul(cid:2)-species coalescent supertrees (i) Bayesian supertree* (iv) CAT-GTR (3/30 jack-knife) (iv) CAT-GTR (10/30 jack-knife) (iv) CAT-GTR (10/30 jack-knife) (v) NDCH2 across-branch composi(cid:2)onal heterogeneity green algae liverwort“bryomossphytes”hornwort Tracheophyta green algae liverwort“bryohornwortphytes”moss Tracheophyta green algae liverwort“bryohornwortphytes”moss Tracheophyta E F G Liverworts-Mosses-Hornworts Liverworts-Hornworts-Mosses Liverworts plus Hornworts-Mosses Figure1. TheSevenCompetingTopologiesIdentifiedbetweenthePrincipalEmbryophyteLineagesandTracheophytes,IncludingaSum- maryoftheSupportfortheThreeMainTopologiesIdentifiedintheAnalyses Thesevenprincipalcompetinghypothesesofbryophyteandtracheophyterelationships:(A)bryophytemonophyly;(B)hornwortssistertoacladeofmossesplus liverworts,itselfsistertotracheophytes;(C)mosses,liverworts,andhornwortsaresuccessivesisterlineagesoftracheophytes;(D)acladeofliverwortsand mossesasasisterlineageofhornwortsplustracheophytes;(E)hornworts,mosses,andliverwortsaresuccessivesisterlineagesoftracheophytes;(F)mosses, hornworts,andliverwortsaresuccessivesisterlineagesoftracheophytes;and(G)hornwortsplusmossescompriseaclade,sistertotracheophytes,and liverwortsareanoutgrouptoallthree.ThereissignificantsupportformonophyleticbryophytesfromanalysesemployingBayesiansupertreeestimationfrom genetrees(i),significanttestsofthemaximum-likelihoodfitofgenetrees(ii)andsequencedata(iii)tothetopologies,andCAT-GTRanalysesinPhylobayes(iv). However,bothBayesiansupertreeestimation(i)andmaximum-likelihoodfitofgenetrees(ii)significantlyrejecthornwortssisterandmosses-liverwortssister,but bothofthesetopologiesarefoundmoreconsistentlyinjack-knifeCAT-GTRanalysesthanmonophyly(iv). topologieshavebeenproposedfrommorphologicalandmolec- [11] and the use of overly simplistic models with assump- ular analyses (Figures 1A–1G; Table 1), leading to the current tions—stationarity, reversibility, and homogeneity—that are consensus of a polytomy between mosses, liverworts, horn- violated in real sequence data [22]. Therefore, any attempt to worts,andtracheophytes[18].Theidentityoftheembryophyte understand these early relationships requires substitution outgroupremainsequallyuncertain,witholderstudiessupport- models and analytic methods that can incorporate these ing the morphologically complex stoneworts (Charales) as the complexities. immediatesistercladeandmorerecentstudiessupportingthe Hereweanalyzealargetranscriptomicaminoacidalignment ‘‘pondscum’’Zygnematophyceaeasthelandplantsistergroup from103speciesofalgae(ChlorophytaandStreptophyta)and (e.g.,[7,19,20]). Embryophyta (mosses, hornworts, liverworts, and tracheo- The failure of different studies to reach congruence on the phytes)fromWickettetal.[7]usingamodel(CAT-GTR+G[23]) fundamental relationships among embryophytes is most likely thataccountsforcompositionalheterogeneityacrosssites[23] duetoinadequatephylogeneticmodelsratherthaninsufficient andadata-recodingstrategy(Dayhoff-6)thatreduceslineage- data [11, 21, 22]. For example, the lack of consensus at the specific compositional heterogeneity; the same recoded data- rootofembryophyteshasbeenattributedtodirectionalevolu- sets were also analyzed using the node-discrete composition tion in nucleotide sequences leading to compositional biases heterogeneity (NCDH) model [24] that explicitly accounts for 734 CurrentBiology28,733–745,March5,2018 Table1. SummaryoftheTopologiesIdentifiedbetweenthePrincipalEmbryophyteLineages,aswellastheAlgalOutgroupto EmbryophytaandtheDataTypeAnalyzed EmbryophyteSisterGroupa Data Reference BryophyteMonophyly Charales nuclearrDNAb Horietal.[56] Charales nuclearrDNAb Steeleetal.[57] Coleochaetales/Charales spermultrastructure Garbaryetal.[8] Coleochaetales spermultrastructure Mishleretal.[14] Charales spermultrastructure Renzagliaetal.[58] charophytes/chlorophytes chloroplastaminoacids Nishiyamaetal.[9] Chlorophyta/Coleochaetales chloroplastnucleotidesb Quandtetal.[59] Coleochaetales chloroplastnucleotidesc GoremykinandHellwig[10] Coleochaetales chloroplastaminoacidsb Rodrı´guez-Ezpeletaetal.[60] Zygnematales chloroplastaminoacidsandnucleotidesb Turmeletal.[61] Zygnematales chloroplastaminoacids Lemieuxetal.[62] Zygnematales chloroplastaminoacids Coxetal.[11] Zygnematales chloroplastaminoacids Civanetal.[63] Zygnematales nuclearaminoacids Wickettetal.[7] (Hornwort,((Liverwort,Moss),Tracheophyta)) Coleochaetales nuclearrDNA Heddersonetal.[38] Coleochatales/Klebsormidiales nuclearrDNA Heddersonetal.[64] Coleochaetales spermultrastructure GarbaryandRenzaglia[40] Chlorophytes mitochondrialrDNA DuffandNickrent[65] Chlorophyta/Coleochaetales nuclearandchloroplastnucleotidesc NishiyamaandKato[66] Charales/Zygnematales nuclear,mitochondrial,chloroplastnucleotidesand Nickrentetal.[67] nuclearrDNAc Coleochaetales morphology,nuclearandmitochondrialnucleotides Renzagliaetal.[58] Zygnematales nuclearnucleotidesc Wickettetal.[7] (Hornwort,(Liverwort,(Moss,Tracheophyta))) Charales morphology Bremeretal.[68] Coleochaetales morphology,nuclearrDNA Mishleretal.[14] ((Liverwort,Moss),(Hornwort,Tracheophyta)) Coleochaetales chloroplastnucleotidesd Nishiyamaetal.[9] Zygnematales chloroplastnucleotides Turmeletal.[61] Zygnematales chloroplastnucleotides Lemieuxetal.[62] Zygnematales chloroplastnucleotidesc Qiuetal.[17] Zygnematales chloroplastnucleotidesc Gaoetal.[20] Zygnematales chloroplastnucleotides Karoletal.[69] Zygnematales chloroplastaminoacids Ruhfeletal.[19] Zygnematales chloroplastaminoacids Lemieuxetal.[70] (Liverwort,(Moss,(Hornwort,Tracheophyta))) ‘‘Charophyceae’’ morphology Parenti[71] Coleochaetes/Charales chloroplastnucleotides Lewisetal.[72] Charales chloroplastnucleotides Delwicheetal.[73] Coleochaetales chloroplastgenomestructure Kelchetal.[74] Coleochaetales chloroplastnucleotidesc Nishiyamaetal.[9] Coleochaetales chloroplastnucleotides Wolfetal.[75] Charales/Coleochaetales mitochondrialnucleotides Groth-Maloneketal.[76] CharalesorZygnematales chloroplastnuclecotideandrDNA,andmitochondrial Qiuetal.[17] andnuclearrDNA Charales chloroplastandmitochondrialnucleotides,andnuclear, Qiuetal.[77] chloroplast,andmitochondrialrDNA (Continuedonnextpage) CurrentBiology28,733–745,March5,2018 735 Table1. Continued EmbryophyteSisterGroupa Data Reference Charales chloroplastnucleotides Smithetal.[27] Zygnematales chloroplastnucleotides Gaoetal.[20] Zygnematales chloroplastnucleotides ChangandGraham[28] Chlorophytes/charophytes chloroplastnucleotidesandnuclearrDNA Fiz-Palaciosetal.[78] Coleochaetales chloroplastnucleotides Magallo´netal.[79] CharalesorZygnematales mitochondrialnucleotides Turmeletal.[80] Charales/Coleochaetales chloroplastnucleotides Kimetal.[81] Charales mitochondrialnucleotidesandaminoacids Liuetal.[21] Zygnematales chloroplastnucleotides Ruhfeletal.[19] Zygnematales chloroplastnucleotides Zhongetal.[29] (Liverwort(Hornwort(Moss,Tracheophyta))) ‘‘charophytes’’ morphology MishlerandChurchill[13] Charales/Coleochaetales morphology Bremer[31] Coleochaetales morphology Bremeretal.[68] Coleochaetales morphologyandchloroplastandnuclearrRNA Mishleretal.[14] Charales/Coleochaetales morphology KenrickandCrane[15] Charales chloroplastandmitochondrialnucleotides,nuclearrDNA Karoletal.[45] (Liverwort((Moss,Hornwort),Tracheophyta)) Klebsormidiales/Coeleochaetales nuclearrRNA Watersetal.[12] Klebsormidiales chloroplastandnuclearrDNA Mishleretal.[14] Chlorophytes mitochondrialrDNA DuffandNickrent[65] (Moss,(Liverwort,(Hornwort,Tracheophyta))) Charales mitochondrialnucleotides Liuetal.[21] Coleochaetales chloroplastaminoacids Lemieuxetal.[82] (Moss,(Hornwort,(Liverwort,Tracheophyta))) Charales nuclearaminoacids Floydetal.[83] aEitherthedesignatedoutgroupsor,ifidentified,themostcloselyrelatedtaxontolandplants. bNotethatalthoughbryophytesweremonophyletic,theywereembeddedinaparaphyletictracheophyte. cProteincoding-genes1+2codonpositionsonly[7,9,10,20,67]. dExcludingthenucleotidesitescodingleucineandthirdcodonpositionsandfourfolddegeneratesites. lineage-specific compositional heterogeneity. Finally, we CompositionalHeterogeneityamongSites exploretherelativefitamongsequencedataandgenetreesto The position of hornworts is unresolved in a focal analysis of seven proposed hypotheses for the relationships of early land genespresentinatleast95%oftaxa(20,512aminoacids).Anal- plants (Figures 1A–1G). Analyses under the CAT-GTR model ysisofthesedatawiththeCAT-GTRmodelyieldsaphylogenyin are not conclusive, supporting several topologies. However, whichtherelationshipsbetweenthebryophytephylaandthetra- bryophytemonophylyissignificantlyfavoredoverallalternative cheophytes are unresolved (Figures 3 and 4). Using posterior topologies (1) in superalignment analyses using the NDCH2 predictive tests to estimate the number of Dayhoff-recoded model,(2)whencomparingthelikelihoodfitofgenetreesandse- aminoacidsateachsiteinthealignment,thereisasignificant quences to different hypotheses, and (3) through Bayesian difference between the empirical data and those estimated supertreeanalysesofgenetrees. from the CAT-GTR model (Z score: 3.85), but this model fit is betterthanthepredictionfromtheGTRmodelwithnocomposi- RESULTS tionalheterogeneityandtheempiricaldata(Zscore:27.24).An- alyses without the hornworts did not change the relationships Theresultsofourphylogeneticanalysesoverwhelminglysupport amongtheremainingbryophyteandtracheophyteclades. the clade uniting liverworts and mosses that we name ‘‘Seta- Fivefullyresolvedalternativetopologiesaresupportedbythe phyta.’’Toalesserdegree,wefindsupportforasisterrelation- jack-knifeanalysesofthefulldataset(Figure4).Thejointmost ship between Setaphyta and hornworts and, hence, the commonlysampledconsensustree(10/30replicates)fromthe monophylyofthebryophytes(Figures2,3,and4).Supportfor jack-knifeanalysesshowshornwortsassistertotracheophytes, the monophyly of the three bryophyte groups is mainly from withliverwortsandmossescomprisingasistercladeatthebase the use of hypothesis tests using prior topologies, but also ofembryophytes(Figure4B).Thepositionofhornwortschanges supertreeestimationfromgenetrees. in the remaining phylogenies to be either at the base of 736 CurrentBiology28,733–745,March5,2018 Pyramimonas parkeae C NUMreoopnnhoermmosaaes lstmipx iso ppiysrtihfoorsmtigisma hlorop Mesostigma viride Ch CSphilroortoakeynbiuas m aitnmuotaphyticus hkby. Entransia fimbriata K Klebsormidium subtile lb Chara vulgaris C Chaetosphaeridium globosum C Coleochaete irregularis o SCpoilreoogcyhraae stpe scutata Zle. Mesotaenium endlicherianum y Cylindrocystis brebissonii gn Cylindrocystis cushleckae e Mougeotia sp m Netrium digitus a Roya obtusa to Penium margaritaceum ph Cosmarium ochthodes . Nothoceros aenigmaticus H Nothoceros vincentianus o Bazzania trilobata L Metzgeria crassipilis iv Sphaerocarpos texanus e Ricciocarpos natans rw SMMpaahrrccahhgaannnuttmiiaa lpeeomslcyaumrrgioiinrpahtaa ort Sebryo BRCPPorhoeylysryuasutmtrcloiaoc dbahmorruyigntmure epm nclulta oercp mufpu mcamraeteupunnislselare Mos taphytaphytes Hedwigia ciliata se Leucodon brachypus s Thuidium delicatulum Rhynchostegium serrulatum Anomodon attenuatus Selaginella moellendorffii 1kp L Selaginella moellendorffii genome y Huperzia squarrosa co Pseudolycopodiella caroliniana p Dendrolycopodium obscurum h. Equisetum diffusum Pteridium aquilinum Alsophila spinulosa P Angiopteris evecta te S GPOsipnihlkoigotuogm lbo isnlosubudamum petiolatum rid. trep Zamia vazquezii to Cycas micholitzii G p Cycas rumphii y h Cedrus libani m y EPipnhuesd traae sdianica no ta s Gnetum montanum p PWreulmwnitsocphitiyas m ainradbinilais erm Sciadopitys verticillata a Taxus baccata e Juniperus scopulorum Cunninghamia lanceolata Amborella trichopoda T Nuphar advena r AKcaodrsuusr aa mheetreicroacnluitas achEm Dioscorea villosa eb OBZSSYSCueoramoracyarblcicgczl aahmhhaalxi yu cabfs pmiubyelaoamosrt dmmibnv iiaauaecuunmondtoltau ooxdnmrsiasantaaclehyon SpermatopEupophyrtaryophyta HSPLSieaaorirrruoscutdeatmueanyan dan drhmiaraeo egnncrlro aiytcrubiadlrinaapatiafera Angioshytahyllop PAEoqscduhoilespcghhiyaoll ulfzomiram pcoeaslltiaafotrunmica perm hyta a Vitis vinifera e Larrea tridentata Medicago truncatula Boehmeria nivea Populus trichocarpa Hibiscus cannabinus Arabidopsis thaliana Carica papaya Kochia scoparia Diospyros malabarica Tanacetum parthenium Inula helenium Rosmarinus officinalis Ipomoea purpurea Catharanthus roseus Allamanda cathartica Figure2. TheMainTopologySupportedbyAnalysesRecoversaMonophyleticBryophyta Abbreviations: Chloro, Chlorophyta; Chkb, Chlorokybophyceae; Klb, Klebsormidiophyceae; C, Charophyceae; Cole, Coleochaetophyceae; Zygnemato, Zygnematophyceae;Ho,hornworts;Lycophy,Lycophyta;Pterid.,Pteridophyta.ImagesofKlebsormidium,Chara,andSpirogyraarefromPhylopic. CurrentBiology28,733–745,March5,2018 737 A strict consensus B 50% majority−rule consensus Monomastix opisthostigma Monomastix opisthostigma Nephroselmis pyriformis C Nephroselmis pyriformis Uronema sp h Uronema sp Pyramimonas parkeae l Pyramimonas parkeae MSCpheilsrooortosatkeigynbmiuaas m vaiitrnmiudoteaphyticus Ch ChlorokSyMbpeuirssoo taastetmignomiap ahm yvitniircuiduteas Entransia fimbriata K Entransia fimbriata Klebsormidium subtile l Klebsormidium subtile Chara vulgaris C Chara vulgaris Chaetosphaeridium globosum C Chaetosphaeridium globosum Coleochaete irregularis o Coleochaete irregularis Coleochaete scutata l Coleochaete scutata SRNpoeiytrraoiu gomybr tadu issgapitus Zyg MesCotyaleinndiuromc yesntidSsl pibcirhreoebgriyisarsnaou snmpii MPCeoenssimoutmaare imunmiaur mgoac erhintthadocliecdhueemsrianum nem CylindrocysNtiMest orciuuugmseh odletigciakit auseps Cylindrocystis brebissonii Roya obtusa Cylindrocystis cushleckae a Cosmarium ochthodes Mougeotia sp t Penium margaritaceum Nothoceros aenigmaticus H Nothoceros aenigmaticus Nothoceros vincentianus Nothoceros vincentianus Bazzania trilobata Bazzania trilobata Metzgeria crassipilis L Metzgeria crassipilis Sphaerocarpos texanus iv Sphaerocarpos texanus RMMiaacrrccciohhcaaannrttpiiaao pes omnlyaamtragoninrspahtaa er Se bry MMaarRrccihchacaninottciaiaa rpepomolysam rngoairntpaahntaas Sphagnum lescurii t o Sphagnum lescurii Polytrichum commune a p Polytrichum commune Ceratodon purpureus p h Physcomitrella patens Physcomitrella patens m h y Ceratodon purpureus HRBreoydsuuwmliag baiarryg cueimlniat etcauf mcapillare oss yta tes RosulaBbrryyHuuemmdw acifrg gciaea npctiiellliauarmtea Anomodon attenuatus e Anomodon attenuatus Leucodon brachypus s Leucodon brachypus Rhynchostegium serrulatum Rhynchostegium serrulatum Thuidium delicatulum Thuidium delicatulum Selaginella moellendorffii 1kp Selaginella moellendorffii 1kp Selaginella moellendorffii genome L Selaginella moellendorffii genome PDseenuddroollyyccooppooddiiuemlla ocbasrcoulinruiamna yco DendrolycHouppoedriuzmia soqbusacurrrousma Huperzia squarrosa S Pseudolycopodiella caroliniana OAEqnpughiiisooepgttlueomrsiss ud eimfvfeu pcsetuatmiolatum Pte tre EAAqlsunoigspieohtpiultame rs idpsi ifnefuuvesloucstmaa APPlstseiolroipdtuhiumilma n saupqdinuuuimlilnousma rid pto OphioPgtleorsPisdsuiiulmomt up ameqt iunoiullaidntuuummm GZainmkigao v baizloqbuaezii G p ZaGmiinak vgaoz bqiuloebzaii Cycas micholitzii y Eh Cycas micholitzii Cycas rumphii m my Cycas rumphii PCiendursu sta leibdaani no bta CPeidnruuss tlaibeadnai GEpnheetudmra msionnictaanum sp ry GnetEupmh emdoran tsainnuicma PSWrceuialmwdniotsopcpithiytiysas vm aeinrrtadicbiniilllaiasta erm op ScWPiaerdulwompitnsitocyphsi itavye smr taiircnaidlblaiinltiaas NATJCuamuunxnpbuipnhoseia nrbrergu alalhsacda c svtmarceitocinaahpa oulaplonorcdueamolata ae ETrachyta CunJAnuimnnigpbheoarrNeumTulsalipa axsh utclarasoircn p bhacuaoedlcopovcreolaaundttmaaaa SKaardcsaunrad rhae gtelarborcalita Suphe KaSdasrucraan hderate grolacblirtaa Houttuynia cordata pho Houttuynia cordata PLSAieacriorrosurdeumeasan aa dhmmreoeennrrr iiytccuiaalinnpuaifsera ermyllophy LirAioPcdeoerrnsSuedsaar raou amnmm etauer lihricpiecaifannenruryaasi Sabal bermudana apt Yucca filamentosa EPADSCYOZSBueosqormioroacycadurlscicgczhl oiachlmhhaasepoxiyu ccgahfs rphmiubiyyelaaoaomoslat l dmiulbnf vzo ivmiaauaiecrailu−mmon l optncltu oooseadomsraslxlitisaaafnottaarucnlmehicyaon Angiosperm tophytahytaa BrEasPcCchohoydslpcoAScoShDphqSadimhcioubiooyuuliiaszrlllmmelgalOciu aghxbo m r da iuercaybZie umarsz poeafmttlaoeuni aa fb ruvolma cstmimiadrc−halnnltaoooyntauaiinolcssvoyomlaaeaaansxr Kochia scoparia Kochia scoparia DViitoiss pvyinroifes rmaalabarica ae DiospyrIonsu lma ahlealbeanriuicma Inula helenium Tanacetum parthenium Tanacetum parthenium Ipomoea purpurea Rosmarinus officinalis Rosmarinus officinalis Ipomoea purpurea Allamanda cathartica Allamanda cathartica Catharanthus roseus Catharanthus roseus Vitis vinifera Boehmeria nivea Boehmeria nivea Populus trichocarpa Larrea tridentata Hibiscus cannabinus Medicago truncatula Larrea tridentata Populus trichocarpa Medicago truncatula Hibiscus cannabinus Arabidopsis thaliana Arabidopsis thaliana Carica papaya Carica papaya Figure3. ResultofAnalysisof30Sampled20,000AminoAcid,Dayhoff-RecodedDatasetsAnalyzedundertheCAT-GTRModel Strict(A)and50%majority-rule(B)consensus.Abbreviations:Chl,Chlorophyta;Ch,Chlorokybophyceae;Kl,Klebsormidiophyceae;C,Charophyceae;Col, Coleochaetophyceae;Zygnemat,Zygnematophyceae;H,hornworts;Liv,liverworts;Lyco,Lycophyta;Pterid.,Pteridophyta. embryophytes withthe moss-liverwort clade sisterto tracheo- consensusoftheresultsfromeachjack-knifeanalysisindicates phytes (10/30 replicates; Figure 4) or within monophyletic a polytomy between the bryophyte phyla and tracheophytes bryophytes at the base of Embryophyta (4/30 replicates; Fig- (Figure3A)andthe50%majority-ruleconsensusshowsasimilar ure4C).Theremainingtopologiesshowvariationsofapolytomy polytomy, but with mosses and liverworts forming a clade containing bryophytes at the base of embryophytes. A strict (Figure3B). 738 CurrentBiology28,733–745,March5,2018 M L H T H M L T M L H T M L H T M L H T L M H T M H L T Figure4. AssessingSupportforCompeting Topologies A B C D E F G Thenumberoftimesseventopologieswerepro- ducedfromthe30jack-knifeanalysesofDayhoff- 10 recodeddataundertheBayesianCAT-GTRmodel (A–G)andthetopologyfromBayesiansupertree c(10, 10, 3, 3, 2, 1, 1)fe replicatesjackkni468 lfif(e[eA2rgsoa5)tumsi]martbecaecasotalsiaumoecdmndhpemouooossanfifirtniit8lzoghiv5enete2sahrljewtglayhecomehnkreet-cestktoehntnarroieosfndeegedsneornsmuf(HuenSosos)tseuatsiersnneeldgdseaeesnoanpdpnseersnRtoahd(odIees)du.inrscT1ittge4hleydo8er: lineage of hornworts plus tracheophytes; (B) 2 hornwortssistertoacladeofmossesplusliver- worts,itselfsistertotracheophytes;(C)bryophyte 0 Index monophyly;(D)mosses,liverworts,andhornworts are successive sister lineages of tracheophytes; (E)mossesplusliverwortscompriseacladesister 852 ML gene trees 148 least heterogeneous gene trees totracheophytes,thepositionofhornwortsunre- Chlorophyta solved;(F)hornworts,mosses,andliverwortsare Chlorokybophyceae Charales successivesisterlineagesoftracheophytes;and Coleochaetales (G)thepositionofhornwortsandliverwortsunre- Zygnematophyceae solved, sister to tracheophytes, with mosses hornworts liverworts branchingearlier.TheBayesiansupertreeanalysis 0.96 of852genesrecoversabryophyteclade,including mosses thesetaphytecladeofliverwortsandmosses(H); hornworts however,bryophytesareparaphyleticinthetree Lycophyta Monilophyta 0.98 of 148 genes, with hornworts sister to tracheo- phytes (I).Abbreviations: H, hornwort;M,moss; Gymnospermae L,liverwort;T,tracheophyta. Angiospermae 0.98 mainingembryophytesisrecovered(pos- teriorprobability=0.98).Further,the148 least heterogeneous genes indicate that H I Charaissistertoembryophytes,whereas all other estimated supertrees support a Across-BranchCompositionalHeterogeneity Zygnematophyceae-Embryophyta relationship (Figure 4I). The TheNDCH2analysisoftheDayhoff-recoded148leastheteroge- effective sample size (ESS) of all parameters was >200, and neousgenesrecoveredmonophyleticbryophyteswithmaximal themedianestimateofthefreebetaparameterwas2.06. (posteriorprobability[PP]=1.0)posteriorprobability.CAT-GTR We also estimated multi-species coalescent supertrees in analysisofthesedatasupportsaliverwortplusmosscladesister Astral[26].Thesetreessupportbryophytemonophylywithlow toahornwortsplustracheophytesclade,withmaximalposterior support(52.5)andalsosupportasister-grouprelationshipbe- probability(PP=1.0). tweenCharaandtheembryophytes. FitofSequenceAlignmentstoProposedTopologies FitofGeneTreestoProposedTopologies Analyses of the sequence data rejected all topologies except Comparisonsofthelikelihoodofgenetreesandsevenproposed monophyleticbryophytes(Table2).Noneofthefiveotheralter- embryophyte relationships support bryophyte monophyly native topologies was sampled during the bootstrap analyses (Table 3). All alternative hypotheses can be rejected using the (monophyletic bryophytes proportion = 1). Bryophyte mono- AUandKHtests,althoughthemoreconservativeSHtestcannot phylyreceived thehighestsupportofsummedlikelihoodsand rejectthetopologyinwhichliverwortsandmossesarerecovered issignificantlysupportedcomparedtoallalternativetopologies asthesistercladetotheremainingembryophytes(Table3,top). using approximately unbiased (AU), Kishino-Hasegawa (KH), However,whenusingtheleastheterogeneous148genedataset, andShimodaira-Hasegawa(SH)tests(Table2). itisnotpossibletorejectliverworts-mossessister,monophyletic bryophytes,orliverwortsplushornworts-mossestopologiesus- SupertreeEstimationusingGeneTrees ingtheAUtest(Table3,bottom). Results from analyses of gene trees using Bayesian supertree inference [25] strongly support monophyletic bryophytes. The DISCUSSION posteriorprobabilityofthesplitsupportingmonophyleticbryo- phytes was 1 across two independent chains (Figure 4H). Thephylogeneticrelationshipamongearlylandplantshasbeen When the 148 least compositionally heterogeneous genes are oneofthemostrecalcitrantproblemsinphylogenetics.Wefind analyzedalone(Figure4I),aliverwort-mosssistercladetothere- support for three topologies of bryophytes and tracheophytes CurrentBiology28,733–745,March5,2018 739 Table2. TheFitoftheWeightedAlignmenttotheSevenHypothesesofEarlyLandPlantTopologyUsingRELLBootstrapping RELLBootstrap Topology Likelihood DLikelihood Proportion AU SH KH Monophyleticbryophytes (cid:2)19,802,012.115 0.000 1 1 1 1 Hornwortssister (cid:2)19,802,243.57 231.459 0 0 0.0013 0 Liverworts-mossessister (cid:2)19,802,248.226 236.111 0 0 0.0004 0 Liverworts-mosses-hornworts (cid:2)19,803,217.59 1,205.477 0 0.0002 0 0 Liverwortsplushornworts-mosses (cid:2)19,803,340.12 1,328.01 0 0.0019 0 0 Hornworts-liverworts-mosses (cid:2)19,803,399.22 1,387.107 0 0.0057 0 0 Liverworts-hornworts-mosses (cid:2)19,803,472.14 1,460.029 0 0 0 0 Thesummedlikelihoodandbootstrapproportionofalignmentssupportsthemonophyleticbryophytestopologyandsignificantlyrejectsallalternative topologies.RELL,resamplingofestimatedloglikelihoods;AU,approximatelyunbiased;KH,Kishino-Hasegawa;SH,Shimodaira-Hasegawa. fromanalysesusingmodelsthataccommodatesequencehet- uponmorphology(e.g.,[15])andtranscriptomes(e.g.,[7]).Our erogeneity:(1)monophylyofthebryophytes,(2)hornwortsalone analyses suggest that this continuing controversy is due to a earliest branching, and (3) liverworts plus mosses earliest combinationoffactors:embryophytesequencedataexhibitsig- branching. Bryophyte monophyly is consistently supported by nificant compositional heterogeneity that is difficult to model analysesthatcomparethefitofsequencedataandgenetrees even with the best current approaches [22]; there is a paucity tohypothesizedrelationships,aswellasbyanalysesperformed ofsequencedataforkeyhornwortlineages;andbiological ef- usingASTRAL.Supportforbryophytemonophylyisalsofoundin fects, such as incomplete lineage sorting, may be masking theresultsofanalysesaccommodatingacross-branchcomposi- early-branchingrelationships. tional heterogeneity. Overall, these results suggest bryophyte Inouranalyses,thechoice ofmodel didimpacttheinferred monophyly, and support for alternative topologies might be a tree topology, and models accounting for across-site or consequenceofincompletelineagesortingandlineagespecific across-tree heterogeneity did not agree with each other. Site compositionalheterogeneityinthedata.Bryophytemonophyly specific heterogeneity is very high in this dataset and cannot is congruent with recent studies that accommodate across- be accounted for in full also when applying Dayhoff recoding branchcompositionalheterogeneity[11]butconflictwithresults (Z = j6j). In addition, it is evident from the results of the CAT- fromalarge-scaleconcatenationanalysisusingsimplermodels GTR+G analyses, which resolve different tree topologies, that [7](Figure2). theproblem mayalso befurthercomplicatedbythe presence of incomplete lineage sorting in the data. However, both the RejectionofProposedRelationships use of ASTRAL coalescent methods and the use of models In the last 30 years, at least seven topologies have been pro- thataccommodatelineage-specificcompositionalheterogene- posedforearlylandplants(Figures1A–1G;Table1).Ourana- ity support bryophyte monophyly. Thus, from a modeling lysesallowustonarrowthistopologicaluncertaintydownsolely perspective, resolving the deepest relationships among land tothepositionofhornworts:asthesistertoacladeofmosses plantsisachallengingproblem.Theapplicationofsite-hetero- and liverworts in monophyletic bryophytes (Figure 1A), as the geneous models has resolved phylogenetic controversy else- sisterlineagetootherembryophytes(Figure1B),orasthesister whereinthetreeoflife(e.g.,[32,33]),but,atpresent,itisunable lineageoftracheophytes(Figure1D).Withallmethods,wecan to definitively resolve early land plant relationships (Figure 4). reject previous hypotheses that do not find a moss-liverwort Furthermore,modelingofcompositionalheterogeneitybetween clade(i.e.,Table1;Figures1Dand1E–1G),suchasthesucces- sitesmaynotbesufficientlycomplextoresolvelandplantrela- sivebranchingofhornworts,liverworts,andmossessistertothe tionships, as it will not capture compositional heterogeneity tracheophytes [13, 19–21, 27–31]. The three remaining topol- among lineages [24] or non-reversibility in sequence evolution ogies represent a fundamental split in the topology witheither [34].Anadditionalfactoristhatthegenome-scaleaminoacidda- monophyly or paraphyly of bryophytes (hornworts or liver- tasetsthatarenowavailableformanyembryophytelineagesare worts-mosses as the sister lineage to all remaining embryo- too large to be easily modeled using the best-fitting models phytes). Wickett et al. [7] found support for both bryophyte available. In particular, achieving convergence with the CAT- paraphyly(hornwortssister—basedonconcatenationanalyses GTRmodelinPhylobayesortheNDCH2modelofP4becomes of nucleotide data) and bryophyte monophyly (based upon a challengingbeyonddatasetsofaround20,000alignedpositions, coalescentanalysisofareduceddatasetofaminoacids).Here evenwhenusingparallelization. wefindequivocalsupportforthesehypothesesthroughestima- Asecondimportantfactorrelatestothesamplingofkeyline- tionofphylogeniesbasedonCAT-GTR-model-basedanalyses ages.Despitethemodelingdifficultiesoutlinedabove,ourana- ofaminoaciddata(Figures3and4). lysesenabledustoresolverobustlyallofthemainbranchesof the land plant tree with the exception of the hornworts—the DifficultiesintheResolutionofEarlyLandPlant most poorly sampled major lineage in our dataset. Hornworts Phylogeny are currently represented by just two transcriptomes from Despitethreedecadesofresearch,thephylogenyofearlyland congenericspecies(Nothoceros),leavingfourofthefiveorders plants remains unresolved with large-scale analyses based unrepresented. Thus, improved genomic or transcriptomic 740 CurrentBiology28,733–745,March5,2018
Description:The list of books you might like
