RESEARCHARTICLE Feeding on Host Plants with Different Concentrations and Structures of Pyrrolizidine Alkaloids Impacts the Chemical-Defense Effectiveness of a Specialist Herbivore CarlosH.Z.Martins1,2☯,BeatrizP.Cunha3‡,VeraN.Solferini3‡,JoséR.Trigo1☯* 1 LaboratóriodeEcologiaQuímica,DepartamentodeBiologiaAnimal,InstitutodeBiologia,UNICAMP, CaixaPostal6109,13083–970,Campinas,SãoPaulo,Brazil,2 ProgramadePós-GraduaçãoemBiologia FuncionaleMolecular,InstitutodeBiologia,UNICAMP,CaixaPostal6109,13084–970,Campinas,São Paulo,Brazil,3 DepartamentodeGenéticaeEvolução,InstitutodeBiologia,Unicamp,CaixaPostal6109, 13083–970,Campinas,SãoPaulo,Brazil ☯Theseauthorscontributedequallytothiswork. ‡Theseauthorsalsocontributedequallytothiswork. *[email protected] Abstract OPENACCESS Citation:MartinsCHZ,CunhaBP,SolferiniVN,Trigo Sequestrationofchemicaldefensesfromhostplantsisastrategywidelyusedbyherbivo- JR(2015)FeedingonHostPlantswithDifferent rousinsectstoavoidpredation.LarvaeofthearctiinemothUtetheisaornatrixfeedingon ConcentrationsandStructuresofPyrrolizidine unripeseedsandleavesofmanyspeciesofCrotalaria(Leguminosae)sequesterN-oxides AlkaloidsImpactstheChemical-Defense EffectivenessofaSpecialistHerbivore.PLoSONE ofpyrrolizidinealkaloids(PAs)fromthesehostplants,andtransferthemtoadultsthrough 10(10):e0141480.doi:10.1371/journal.pone.0141480 thepupalstage.PAsconferprotectionagainstpredationonalllifestagesofU.ornatrix. Editor:PeterSchausberger,UniversityofNatural AsU.ornatrixalsousesotherCrotalariaspeciesashostplants,weevaluatedwhetherthe ResourcesandLifeSciences,AUSTRIA PAchemicaldefenseagainstpredationisindependentofhostplantuse.Wefedlarvae Received:March26,2015 fromhatchingtopupationwitheitherleavesorseedsofoneofeightCrotalariaspecies (C.incana,C.juncea,C.micans,C.ochroleuca,C.pallida,C.paulina,C.spectabilis,and Accepted:October8,2015 C.vitellina),andtestedifadultswerepreyeduponorreleasedbytheorb-weavingspider Published:October30,2015 Nephilaclavipes.Wefoundthattheprotectionagainstthespiderwasmoreeffectivein Copyright:©2015Martinsetal.Thisisanopen adultswhoselarvaefedonseeds,whichhadahigherPAconcentrationthanleaves.The accessarticledistributedunderthetermsofthe exceptionswereadultsfromlarvaefedonC.paulina,C.spectabilisandC.vitellinaleaves, CreativeCommonsAttributionLicense,whichpermits unrestricteduse,distribution,andreproductioninany whichshowedhighPAconcentrations.WithrespecttothePAprofile,wedescribeforthe medium,providedtheoriginalauthorandsourceare firsttimeinsect-PAsinU.ornatrix.ThesePAs,biosynthesizedfromthenecinebaseretro- credited. necineofplantorigin,ormonocrotaline-andsenecionine-typePAssequesteredfromhost DataAvailabilityStatement:Allrelevantdataare plants,wereequallyactiveinmothchemicaldefense,inadose-dependentmanner.These withinthepaper. resultsarealsopartiallyexplainedbyhostplantphylogeny,sincePAsofthehostplantsdo Funding:Financialsupportwasprovidedby haveaphylogeneticsignal(cladeswithhighandlowPAconcentrationsinleaves)whichis FAPESPtoJRT(11/17708-0)andVNS(12/02526-7), reflectedintheadultdefense. andCNPqtoJRT(306103/2013-3).Thefundershad noroleinstudydesign,datacollectionandanalysis, decisiontopublish,orpreparationofthemanuscript. CompetingInterests:Theauthorshavedeclared thatnocompetinginterestsexist. PLOSONE|DOI:10.1371/journal.pone.0141480 October30,2015 1/27 ChemicalDefenseinaSpecialistHerbivore Introduction Defensesevolvedbyanimalstoavoidpredationareubiquitousinnature,anddifferentdefen- sivestrategieshaveevolvedinresponsetodifferentlifestyles.Themyriaddefensivestrategies includeavoidingdetection,preventingattack,anddeceivingpredators[1].Someherbivorous insectsfeedingonchemicallyprotectedhostplantsareabletoovercometheseplantdefenses bysequesteringplantsecondarycompounds,andusingthemfortheirowndefenseagainst predatorattack[2–4].Sequestrationmayhaveevolvedindependentlyindifferenttaxaofher- bivorousinsects.ItpredominatesinColeopteraandLepidoptera,butalsooccursfrequentlyin theHeteroptera,Hymenoptera,OrthopteraandSternorrhyncha[4].Thesequestereddefensive compoundscompriseavastarrayofnaturalproductsofdifferentbiosyntheticpathways,such ascardenolides,cyanogenicandiridoidglycosides,aristolochicacids,glucosinolatesandpyrro- lizidinealkaloids.Thesesubstancesareeffectiveagainstavarietyofpredators,rangingfrom invertebratesincludingspidersandants,tovertebratesincludingbirdsandmammals[2–5]. Generally,theeffectivenessofchemicaldefensesagainstpredationisdose-and/orstruc- ture-dependent[3].Highconcentrationsofsequesteredcompoundsintheherbivorousprey aremoreeffectiveagainstpredators,andtheireffectivenessisusuallyrelatedtotheconcentra- tionofthesecompoundsinthehostplants.Differentstructuresofthesameclassofcom- poundsmayalsoshowdifferentactivitiesagainstpredation.Forinstance,apalatability spectrumofthemonarchandqueenbutterflies,DanausplexippusandD.gilippus,wasfound tobeassociatedwiththeamountofcardenolidessequesteredfromdifferenthostplants[6,7]. ThispatternhasbeenalsoobservedintheadultsofthenymphalidbutterflyEuphydryaspha- eton,whichacquireddefensiveiridoidglycosidesaslarvaefromtwodifferenthostplants[8,9]; andintheadultsoftheheliconiinebutterflyHeliconiuserato,whoselarvaefeedonfourspecies ofPassiflora[10].Additionally,SilvaandTrigo[11]demonstratedthatpyrrolizidinealkaloids haveapositivedose-dependentactivityintheprotectionofinsectsagainsttheorb-weavingspi- derNephilaclavipes(Nephilidae).Inthesamestudy,theyshowedthatdifferentPAstructures haddifferentantipredationactivities. Amongthedefensivecompoundssequesteredbyherbivorousinsects,theroleofpyrrolizi- dinealkaloids(PAs)hasbeenwelldocumented[12].Thesecompoundsareproducedbyplants inseveralfamilies(e.g.Asteraceae,BoraginaceaeandLeguminosae),andaresequesteredby specialistgrasshoppers,hemipterans,beetles,mothsandbutterflies[5,13,14],wheretheyshow defensiveandsexualcommunicationfunctions[14].PAsinspecializedinsectsarealwayspres- entinN-oxideform[15].ManyarctiinemothsconvertPAssequesteredfromtheirlarvalhost plantsinto“insect-PAs”inwhichtheacidcomponentsofthealkaloidsarereplacedbysmall, branchedaliphatic2-hydroxyacidsofinsectorigin[16].ThesePAsofinsectoriginareprecur- sorsofthemalesexpheromoneinthesemoths[16]. ThearctiinerattleboxmothUtetheisaornatrixisoneofthemost-studiedspecieswith respecttoitsecologicaldependenceonPAs[5,14,17–19].U.ornatrixoccursintheNeotropics andwarmNearcticregions[20],whereitfeedsonmanyspeciesofCrotalarialegumes[21]. ThegenusCrotalariahasaPantropicaldistribution,andsomemembershavecolonizedthe warmerpartsoftheNearcticregion[22].ThelarvaeofU.ornatrixsequesterthesealkaloids, andpassthemtopupaeandadults.Duringmating,malestransferPAstofemales,whichtrans- ferthemtotheeggs.Therefore,alllifestagesofU.ornatrixareprotectedbyPAsagainstpreda- tors[14].Eisner[23]firstdemonstratedthattheunpalatabilityofU.ornatrixadultsisdueto PAsintheirtissues,andthatthealkaloidoriginatesfromthelarvalhostplant,Crotalariapal- lida(formerlyC.mucronata).EisnerfoundthatadultsofU.ornatrixwereprotectedagainst theorb-weavingspiderNephilaclavipes,aswellasfromotherspiders,andbirds.Additional evidenceofthedefensiveroleofPAsinU.ornatrixcamefrombioassays,withPAstopically PLOSONE|DOI:10.1371/journal.pone.0141480 October30,2015 2/27 ChemicalDefenseinaSpecialistHerbivore appliedonpalatableorganismsorbyofferingdietswithorwithoutPAstoPA-specialist insects,andtestingtheseorganismsagainstvariouspredators[5]. AlthoughitissuggestedthattheunpalatabilityofU.ornatrixmaybecloselyrelatedtothe amountofPAsintheirhost-planttissues,fewstudieshavedemonstratedthisrelationship.Eis- neretal.[14,23–25]testedtheadults,withandwithoutPAs,againstspidersandbirds,butno dose-activitybioassaywascarriedout.Ferroetal.[26]foundthatthedifferencesinadultpalat- abilityincreasedwhentheirlarvaefedonunripeseedsandleavesofC.pallida.Theleaveshad alowerPAcontentthanunripeseeds,andconsequentlythelarvaefedonleavesweremore consumedbyN.clavipesthanthosefedonunripeseeds.Likewise,HristovandConner[27] alsoshowedthatU.ornatrixfedonleavesofC.spectabilisweremorepredatedbythebatEpte- sicusfuscuscomparedtomothsfedonseeds;mothsraisedinadietfreeofPAswerepalatable tothebats. AnimportantissuethatremainsunclearistheroleofPAstructureinthechemicaldefense ofU.ornatrix.Dodifferentstructuresconferdifferentlevelsofdefense?Eisneretal.[23–25] raisedU.ornatrixonitsusaraminePA-containinghostplant(C.pallida)andobservedthatthe mothwasrejectedbythespidersN.clavipesandLycosaceratiola,whilemothsraisedonadiet freeofPAswerepreyedupon.SimilarresultswerefoundforL.ceratiola,whichreleasedU. ornatrixraisedonadietsuppliedwithPAmonocrotaline,themainPAofC.spectabilis[24,25]. IntwootherstudiesonchemicaldefenseofU.ornatrixadults,larvaewerealsoraisedonC.pal- lidaorC.spectabilis,andthereforethePAsinvolvedinthedefensewererespectively,usara- mineandmonocrotaline[26,27].However,larvaeofU.ornatrixcanfeedonmanyCrotalaria species[21,28],withdifferentPAconcentrationsandprofiles[29].ThefindingthatC.pallida andC.spectabilisconferredsimilarlevelsofdefenseinspiteofcontainingdifferentPAsledus tohypothesizethatthestructureisunimportantindeterminingpalatability.Ourpredictions were:(1)adultsfromlarvaethatfedonplantsorplantpartswithhighPAcontentswouldbe betterdefendedthanthosethatfedonplantsorplantpartswithlowPAcontents,and(2)PA structurewouldplaynoroleinthemothchemicaldefenses. Anotherpointthathasnotbeenexaminedinstudiesofchemicaldefensesofherbivores sequesteredfromtheirhostplantsistheroleofhostplantphylogeny.Itwouldbeexpectedthat phylogeneticallyrelatedhostplantsproducesimilarcompoundsavailableforsequestrationby insects.Thereforewecannotconsidereachhostplantspeciesasindependentforcomparative analysis(e.g.[30]).Ifphylogeneticinertiaisstrong,thepotentialadaptationsthatrelatedspe- ciesmayevolvewillbesimilarlyconstrained,withtheeffectthatspeciescannotberegardedas independentofeachother[31].Consequently,itwasnecessarytotakeintoaccountthehost plantphylogeny,inordertocomparethesequesteredchemicaldefensesinanherbivorous insect.Toourknowledge,thisapproachhasneverbeentakeninpublishedstudies.Forexam- ple,evenforthemonarchbutterfly,awell-studiedmodelofsequesteredchemicalcompounds, theconnectionbetweeninsectdefensesandhostplantphylogeneticrelationshipshasnever beenexplored,althoughthephylogenetictrendsrelatedtochemicaldefensesarewellknownin Asclepias[32,33].Therefore,weexploredthequestionofwhetherCrotalariaphylogenydirects anytrendforthechemicaldefensesofU.ornatrix.Wehypothesizedthatmothsthatfedon phylogeneticallyrelatedhostplantswouldshowsimilardefensivepatterns. Toaddressourhypothesisandpredictions,wefedlarvaeonleavesorseedsofeightdifferent CrotalariaspeciesfoundintheNeotropics,whichwerenative,non-nativeorofuncertainori- gin,withdifferentPAconcentrationsandprofiles[29].Theadultsthatemergedfromlarvae fedthedifferentdietswereofferedtothespiderN.clavipesinapredationbioassay.Weana- lyzedthePAconcentrationsandprofilesforU.ornatrixadultsandCrotalariaspecies,andcor- relatedthePAconcentrationsofadultswiththePAcontentsoftheplantpartstheyfedonas larvae.WealsocorrelatedthePAconcentrationandprofilebetweenthemothsandthespider PLOSONE|DOI:10.1371/journal.pone.0141480 October30,2015 3/27 ChemicalDefenseinaSpecialistHerbivore responseinthepredatorbioassay.Thiscorrelationwillorwillnotsupportourfirstprediction. SincewefounddifferentclassesofPAsinU.ornatrix,wealsobioassayedthethreemostcom- monclassesagainstthespidertotestoursecondprediction.Finally,wemappedthePAprofiles andconcentrationsinCrotalariaspecies,thePAconcentrationsinthemoth,andthedefensive responseofN.clavipesagainstanindependentphylogenetichypothesisfortheeightCrotalaria species,inordertotestourlasthypothesis. MaterialsandMethods StudySystem TherattleboxmothUtetheisaornatrix(Erebidae:Arctiinae)isprimarilyNeotropicaland extendstowarmerareasoftheNearcticregion[20].U.ornatrix,togetherwithfivespeciesthat occuronlyintheGalapagosIslands,aretheextantUtetheisaspeciesintheNeotropics[34,35]. TheadultsaregenerallyfoundflyingnearpatchesofCrotalariainpastures,theedgeofwoods androadsides,wherethelarvaecanbefoundfeedingonbothleavesorseedsinsidetheunripe podsofCrotalaria[21,26,28]. Crotalaria(Leguminosae:Papilionoideae:Crotalarieae)isaPantropicalgenusofweeds, comprisingapproximately702species[22,36].IntheNeotropics,particularlyinBrazil,31 nativeand11non-nativespecieshavebeenrecorded[37].CrotalariaspeciesarerichinPAs [34],whicharefoundinhigherconcentrationsinseedsthaninleaves[26].InadditiontoPAs, Crotalariaspecieshaveotherdefensivetraitsagainstherbivoresincludingantifeedingdeter- rentssuchaslectins[38],non-proteinaminoacids[39]andproteaseinhibitors[40].Theyalso haveextrafloralnectaries(EFNs)thatattractpredatoryantsandwasps[26,41–43].Weused eightspeciesofCrotalaria,includingthreenatives,threenon-natives,andtwowithuncertain origins(Table1),thathavebeenplantedinanopenareaneartheAnimalBiologyDepartment, InstituteofBiologyattheStateUniversityofCampinas,Campinas,SãoPaulo,Brazil(22° 49'15.38"S,047°04'8.87"W).Innaturalenvironments,wehaveobservedU.ornatrixusingC. incana,C.micans,C.spectabilis,C.pallidaandC.vitellineashostplants.Fortheotherthree species,wehavenoinformationabouttheirnaturalusebyU.ornatrix. Theneotropicalorb-weavingspiderNephilaclavipes(Nephilidae)isapredatorthatbuilds itswebinforestclearingsandcorridors,whichareflightpathsforinsects[44].Thisspider preysongrasshoppers,bees,wasps,moths,andbutterflies,butisabletodiscriminatePA- containinginsects,releasingthemunharmed[5](Fig1).Intheedgesofwoods,thisspiderco- occurswithU.ornatrix. ThepermitforresearchwithwildanimalswasprovidedbyIBAMA-ICMBio(Ministériodo MeioAmbiente,Brazil). RearingUtetheisaornatrixLarvae WeobtainedlarvaeofU.ornatrixfromadultscollectedontheFazendaSantaMariana,Campi- nas,SãoPaulo,Brazil(22°47’02.91”S,47°00’36.03”W),wherethehostplantC.pallidaisabun- dant.Webroughttheadultstothelaboratory,sexedthemfollowingTravassos[45],and allowedthemtomate(20malesand10femalespercage)inapapercage(ca.3.2L)following Cogni[28].Wesuppliedfoodinavialcontaining5.0%aqueoushoney.After3–4days,the adultsstartedtoovipositonthepaper-cagesurfaces.Theeggsweremixedtorandomizethe parentalorigin.Aftereclosion,werearedthefirstinstarsonleavesorunripeseedsofoneofthe eightspeciesofCrotalaria.WealsorearedagroupoffirstinstarsonaPA-freediet,following Cognietal.[46].Foreachtreatment(dietsofleavesorunripeseedsofeachspeciesofCrota- laria),werearedca.20individualsinaplasticcontainer(6cmhigh,20cmdiameter)until pupation.Forsometreatments,dependingontheavailabilityofseedsandleaves,wehad2or PLOSONE|DOI:10.1371/journal.pone.0141480 October30,2015 4/27 ChemicalDefenseinaSpecialistHerbivore Table1. SpeciesofCrotalariausedtofeedlarvaeofUtetheisaornatrixforNephilaclavipesbioassayandPAanalysis. Species Section Nativerange C.incanaL. Chrysocalycinae Pantropical(uncertainorigin)2 C.junceaL. Calycinae India,Asia2 C.micansLink Chrysocalycinae Neotropics1,2 C.ochroleucaG.Don Hedriocarpae TropicalAfrica2 C.pallidaAiton Hedriocarpae Pantropical(uncertainorigin)2 C.paulinaSchrank Calycinae Neotropics1,2 C.spectabilisRoth Crotalaria Asia2 C.vitellinaKerGawler Chrysocalycinae Neotropics1 1Flores[38] 2Polhill[22] doi:10.1371/journal.pone.0141480.t001 3containers,giving40–60individualspertreatment.Thelarvaewerefedadlibitum.We inspectedthecontainersdaily,removingfecesanddeadindividuals,andreplacingtheleavesor seedswithnewones.Wemovedthepupaetoanothercontainerwiththesamedimensions untiladultemergence,sincethelarvaecancannibalizepupae[47].Wecarriedoutthesame procedureforpupaefromthePA-freediet.Wemovednewlyemergedadultstopapercages, separatingthembytreatmentandsex,untilthespider-predationbioassays.Wekeptboththe papercageswithadultsandtheplasticcontainerswithlarvaeandhostplants,orpupae,at roomtemperature. PredationBioassaywithLivingMoths WecarriedoutthebioassayswithNephilaclavipesinasmallpatchofwoodsinCampinas(22° 48'21.26"S,47°4'43.12"W)fromMarchtoMayin2012and2013,whenthespiderswereabun- dant.Weconducedallbioassaysbetween09:00and16:00hs.Weusedonlyadultfemalespi- dersthatrespondedimmediatelywhenanypreywastossedintotheirweb.Wedidnotuse spidersthatwereinthecourseoffeedingonapreyinsect,butwedidnotcontrolforsatiation beforeeachbioassay.Weusedaround80spidersforthebioassays.Sincethenumberofspiders wasalimitingfactor,sometimeswerepeatedthebioassayswiththesamespider,buteachindi- vidualwasusedonlyonceperweek.Webioassayed481adultmoths,testingaround20individ- ualsforeachsex,hostplantandplantpart.Wemadeasmallcutinthemoth'shindwings beforeplacingitontheweb,topreventthemothfromflyingawayifreleasedbythespider.We placedthemothontheweb,andrecordedifitwaseatenorreleased.Weconsideredthespider responsetobepredationwhenitbit,wrappedandkilledthemoth.Whenthespidercutthe webaroundthemothaftertouchingitandfreeditunharmed,werecordedtheresponseasa release.WedidnotrecordtherejectionbehaviordescribedbyVasconcellos-NetoandLewin- sohn[48](thepreywasinitiallysuckedandthenfreed),sincethepreywasremovedfromthe webbeforethisstage.Ifthespiderreleasedthemoth,weofferedapalatablefreeze-killedmeal- wormTenebriomolitor(Coleoptera:Tenebrionidae)asacontrol.Werecordedthetrialasa rejectiononlyifthespiderfedonthemealworm.Ifthespiderkilledtheexperimentalmoth,we didnotusethecontrol.Whenpredationwasrecorded,weimmediatelyremovedthemoth fromthewebandplaceditina1.5-mLEppendorftubefilledwithMeOHforfurtherPAanaly- sis.Allmothswererecoveredintact;eventhepartsbittenbythespiderhadnovisiblesignsof damage.WealsopreservedthereleasedmothsinMeOHasabove.Weanalyzed310individu- als;171werenotusedforPAanalysis,sincetheywerelostafterthespiderbioassay. PLOSONE|DOI:10.1371/journal.pone.0141480 October30,2015 5/27 ChemicalDefenseinaSpecialistHerbivore Fig1.FemaleofNephilaclavipeshandlinganadultmaleofUtetheisaornatrix.Notethatthespidertouchesthemothwithherpedipalps,probably evaluatingthecontentofdefensivepyrrolizidinealkaloids. doi:10.1371/journal.pone.0141480.g001 PLOSONE|DOI:10.1371/journal.pone.0141480 October30,2015 6/27 ChemicalDefenseinaSpecialistHerbivore PyrrolizidineAlkaloids:Extraction,Quantification,Characterizationand Isolation ToquantifythePAs,weextractedthefreeze-driedsamplesofunripeseeds,leavesofhostplants oradultmothsthreetimeswithEtOH(10xvolume:weight).WecentrifugedtheEtOHextract at10,416rcffor10min,andrecoveredtheEtOHlayer.WecompletedtheEtOHlayerto20 mLandtookanaliquot(0.1or1.0mL)tocarryoutthecolorimetricanalysisaccordingto Trigoetal.[49,50].Weusedmonocrotalineforthereferencecurve.ThetotalPAconcentration wasgiveninμgofPAs/mgofdryweight. ForPAcharacterizationbygaschromatography-massspectrometry(GC-MS),weextracted theplantorinsectsamplesusingtheacid-baseproceduresdescribedbyTrigoetal.[49,50]. TheGC-MSanalysiswascarriedoutinelectronimpactmodeaccordingtoFloresetal.[29]. Theretentionindicesandmassfragmentationpatternswerecomparedwithpublisheddescrip- tions(seeTable2,andreferences[29,51,52]). ForPAcharacterizationbyliquidchromatography-massspectrometrythesampleswere extractedinEtOHasabove.TheEtOHlayerwasevaporatedinarotaevaporatorat40°C,recov- eredwith1.5mL2%aqueousaceticacidandcleaned3timeswiththesamevolumeofhexane. Theacidsolutionwasaddedtoaclean2-mLvial,capped,andstoredat-20°CuntilLC-MS analysis.WeusedanAgilent1260InfinityQuaternaryLC,equippedwithanEclipsePlusC-18 column,4.6x250mm,5μm,andaguardcolumnwiththesamephase,coupledwithanAgi- lent6130singlequadrupoleinelectrosprayionizationmode.Thecolumnwaskeptat40°C, andtheinjectionvolumewas5to50μl.ThePAwasseparatedusingalineargradientcontain- ingaqueous20mMammoniumacetateandMeOHatflowrate0.5mL/min.Thegradient startedat95%ammoniumacetate:5%MeOH(3min),MeOHwasraisedto100%in25min, andkeptat100%for3min.Themassspectrometerwasruninpositivemode,scanningthe productionsfrom100–500amu(seeTable2). ForthepredationbioassayswithpurePAsinN-oxideform,weisolatedmonocrotaline fromC.spectabilis,integerriminefromSeneciobrasiliensis(Asteraceae:Senecioneae)andthe putativemixtureinsectPAscreatonotineB:iso-creatonotineB(Fig2)fromadultsofU.orna- trixthatfedaslarvaeonleavesandunripeseedsofC.vitellina.WeextractedPAsusingthe acid-baseprocedureasdescribedabove,andisolatedthemusinganadsorptioncolumnchro- matograph(40cmlength,2.5cmdiameter),usingSilicaGel70–325meshasthestationary phase,andagradientfromCHCl toCHCl :MeOH:Et N85:14:1asthemobilephase.Wefol- 3 3 3 lowedthePAsin10-mLfractionsbysilica-gelthin-layerchromatographywithCHCl :MeOH: 3 NH OH85:14:1aseluentandDragendorff'sreagentfordetection.WeN-oxidizedthePAsand 4 purifiedthemusingtheproceduredescribedbyCraigandPuroshothaman[53]. PredationBioassaywithPurePyrrolizidineAlkaloids WefoundthatadultsofU.ornatrixgenerallycontainedPAN-oxidesofmonocrotalineand integerriminetypes,andthespecificmixtureofinsectPAscreatonotineB:iso-creatonotineB (seeResults).WecarriedoutpredationbioassayswithN.clavipestodetermineifPAN-oxides withdifferentstructureshadsimilaractivityagainstthespider.Thebioassayswerecarriedout inthesameareawherewecarriedoutthepredationbioassaywithlivingmoths. Wetreatedpalatablefreeze-killedlarvaeofthemealwormT.molitor(hereafterbaits)with 3.0μg/mgdryweightofeachalkaloid,whichwereobtainedasdescribedabove.Thisconcen- trationwasdeterminedbycalculatingthelogisticregressionoftheN.clavipesresponseinrela- tiontoPAconcentrationinadultsofU.ornatrixfedaslarvaeononeoftheeightdifferenthost plants(seeStatisticalanalyses),whereat3.0μg/mgtheprobabilityofreleasebythespiderwas 87%.ToevaluatewhetherthePAconcentrationalsoplayedanyroleinthechemicalprotection PLOSONE|DOI:10.1371/journal.pone.0141480 October30,2015 7/27 ChemicalDefenseinaSpecialistHerbivore Table2. MassfragmentationpatternofpyrrolizidinealkaloidsfoundinCrotalariaspeciesandinadultsofUtetheisaornatrixfedonthesehost plants. Theanalyseswerecarriedoutbygaschromatography-massspectrometry(GC-MS)inelectronimpactmodeandliquidchromatography-massspec- trometry(LC-MS)inelectrosprayionizationmode. Pyrrolizidinealkaloids RIa Rtb DiagnosticionsforGC-MSc,m/z(%) Diagnosticionsfor Referencefor LC-MSd,m/z(%) GC-MSe Retronecine 1487 6.085 [M+]155(23),111(60),94(17),80(100) [2M+H]+343(9),[M [29]1 +H]+172(100) 9-(2’-Hydroxy)-ethanoylretronecine- 1795 nd [M+]227(2),183(6),138(62),120(6),93 nd [29]1 like (100),80(40) 9-(2-Hydroxy-3-methylpentanoyl)- 1831 nd [M]+255(2),142(6),125(10),124(100),96 nd [52]1 trachelanthamidine (4),83(20),82(12) CreatonotineB-like 1840 nd [M]+255(3),211(7),138(99),124(13),120 nd [51]2 (11),94(40),93(100),80(25) 7-Senecioylretronecine-like 1861 nd [M]+237(5),137(29),136(18),94(25),93 nd [29]1 (10),80(100) 9-Senecioylretronecine-like 1890 nd [M]+237(3),193(9),154(15),138(27),137 nd [29]1 (28),136(20),94(24),93(100),80(20) Iso-creatonotineB 2024 22.075 [M]+269(3),251(10),138(40),124(17), [M+Na]+308(11),[M [51]1 120(26),111(61),106(51),94(26),80 +H]+286(100) (100) 7-Octanoylretronecine-like 2031 nd [M]+281(50),220(19),154(8),136(22), nd [29]1 124(22),111(69),106(47),94(24),80 (100) CreatonotineB 2048 22.969 [M]+269(<1),251(1),225(5),138(100),93 [M+Na+]308(6),[M [51]1 (88),80(18) +H]+286(100) 9-Octanoylretronecine-like 2052 nd [M]+281(5),236(6),138(62),120(26),106 nd [29]2 (13),94(47),93(100),80(44) 1,2-DihydrocreatonotineB 2064 nd [M]+271(4),210(12),171(20),140(25), nd [52]1 139(11),114(10),96(34),95(70),82(100) 9-(5’-Hydroxy)- 2082 24.162 [M]+283(1),224(4),155(24),138(65),93 [M+Na]+322(14),[M [29]1 heptanoylretronecine-like (100),80(19) +H]+300(100) Crispatine-like 2175 22.598 [M]+309(2),222(7),136(88),120(83),119 [M+Na]+348(11),[M [29]3 (100),93(62),80(29) +H]+326(100) Unknownmonocrotaline-type 2243 24.936 [M]+323(3),236(4),208(10),136(97),120 [M+Na]+362(9),[M [29]3 (87),119(100),93(54),80(27) +H]+340(100) Incanine-like 2315 27.280 [M]+337(6),264(11),250(5),222(8),136 [M+Na+]376(15),[M [29]1 (100),120(76),119(79),93(49),80(27) +H+]354(100) Monocrotaline 2336 18.610 [M+]325(1),254(3),236(46),136(52),120 [M+Na]+364(6),[M [29]1 (100),93(38),80(19) +H]+342(100) Senecionine/Integerrimine 2339/ 26.789/ [M]+335(6),291(12),248(12),220(21), [M+Na)+]374(13),[M [29]1 2410 25.111 136(97),120(100),93(83),80(34) +H]+352(100) Trichodesmine-like 2341/ 25.326 [M]+353(6),264(100),136(31),120(35), [M+Na]+392(17),[M [29]1 2348 93(44),80(14) +H]+370(100) Unknownmonocrotaline-type 2346 26.050 [M]+337(2),222(5),136(58),120(89),119 [M+Na]+376(20),[M [29]3 (100),93(40),80(18) +H]+354(100) Unknownmonocrotaline-type 2346 nd [M]+353(2),264(45),136(75),120(100), nd [29]2 93(70),80(25) Senecionine-like 2376 nd [M]+337(6),222(18),136(75),120(100), nd [29]2 93(67),80(32),55(47) Methylmonocrotaline-like 2384 nd [M]+339(1),250(58),136(53),120(100), nd [29]1 93(39),80(20) Senecionine-like 2386 nd [M]+321(13),247(7),136(98),120(100), nd [29]1 93(76),80(37) Trichodesmine-like 2418 23.415 [M]+353(2),264(60),136(47),120(100), [M+Na]+392(13),370 [29]1 93(35),80(18) [M+H]+370(100) (Continued) PLOSONE|DOI:10.1371/journal.pone.0141480 October30,2015 8/27 ChemicalDefenseinaSpecialistHerbivore Table2. (Continued) Pyrrolizidinealkaloids RIa Rtb DiagnosticionsforGC-MSc,m/z(%) Diagnosticionsfor Referencefor LC-MSd,m/z(%) GC-MSe Incanine-like 2430 nd [M]+337(5),264(27),222(20),136(62), nd [29]1 120(100),93(65),80(27) Trichodesmine-like 2437 nd [M]+353(2),264(83),222(8),136(58),120 nd [29]1 (100),93(34),80(18) Senecionine-like 2540 nd [M]+337(1),155(12),138(71),136(28),93 nd [29]2 (100),80(19),55(27) Senecionine-like 2551 nd [M]+351(31),220(10),136(43),120(100), nd [29]2 119(90),93(75),80(33) PlatyphorineC-like 2556 25.260 [M]+383(2),281(13),267(11),252(100), [M+Na]+422(19),[M [52]1 138(24),136(27),120(51),93(51),80(18) +H]+400(100) Unknownseco-PA 2615 nd [M]+365(20),321(15),276(22),238(100), nd [29]1 168(70),122(38),110(32),94(31),83(46) Retrorsine/Usaramine 2621/ 22.989/ [M]+351(7),246(5),136(100),120(99), [M+Na]+390(7),[M [29]1 2647 22.731 119(84),93(80),80(35) +H]+368(100) Senecionine-like 2675 nd [M]+351(3),224(8),143(100),136(65), nd [29]2 120(86),119(67),93(51),80(20) Senecionine-like 2684 nd [M]+351(2),143(24),136(51),120(100), nd [29]2 119(91),93(79),80(23) Unknownseco-PA 2728 nd [M]+381(25),338(55),320(63),250(58), nd [29]1 238(77),168(100),150(32),122(52),110 (44) Unknownseco-PA 2815 nd [M]+379(43),334(26),238(22),168(74), nd [29]1 151(29),139(57),122(100),110(62),94 (57) Unknownseco-PA 2866 nd [M]+421(31),376(22),168(77),150(79), nd [29]1 139(67),122(100),110(64),94(45),43 (96) Unknownseco-PA 2907 nd [M]+437(34),250(79),226(41),197(78), nd [29]1 183(78),168(44),122(100),110(52),94 (52) a.RetentionindexinGC-MSanalyses. b.Retentiontime(min)inLC-MSanalyses.SomecompoundsdetectedbyGC-MSwerenotdetected(nd)byLC-MS. c.infreebaseform. d.inN-oxideform. e.CharacterizationofPAsbymassspectrausing:1Massspectramatchingwithliterature,2.Interpretationofthemassspectrumfromliterature, 3Erroneouslydescribedin29asunknownsenecionine-type. doi:10.1371/journal.pone.0141480.t002 againstthespider,wealsobioassayedthesePAN-oxidesinaone-tenthconcentration.Addi- tionally,weconductedbioassayswith5.0μg/mgN-oxidefortheinsectPA.Allbioassayswere conductedusing10baitsforeachPAconcentration.ThePAsweredilutedinMeOHand appliedtopicallywitha10-μLsyringeonthesurfaceofthebait,whichwaskilledbyfreezing. WeusedaMeOH-treatedbaitasacontroltodeterminewhethertoacceptthetrial,as describedabovefortheN.clavipespredationbioassaywithlivingmoths. HostPlantPhylogeneticHypothesisandCharacterMapping WemappedthecharactersofPAsinvolvedinthechemicaldefensesofU.ornatrixinaninde- pendentphylogenetichypothesisfortheeightCrotalariaspecies.Thisproceduremayreveal theevolutionarytrendsofmappedcharacters[54].Wedidthisonlyforthosethatwerefedon leaves,asallindividualsfedonseedswerewellprotectedagainstthespider,regardlessofthe PLOSONE|DOI:10.1371/journal.pone.0141480 October30,2015 9/27 ChemicalDefenseinaSpecialistHerbivore Fig2.Somepyrrolizidinealkaloidsfoundinthisstudy.ThealkaloidsaredrawnintheirN-oxideform.Formorestructures,see,e.g.,Floresetal.[29]. doi:10.1371/journal.pone.0141480.g002 hostplant.AsPAcharactersweusedPAconcentration(high>3.0μg/mgandlow<1.0μg/ mg)andPAtype(monocrotaline,senecionine,senkerkineorinsectPA)inalleightCrotalaria speciesandU.ornatrix,andtheresponseofthepredatorN.clavipesinrelationtoU.ornatrix (percentageofrelease). WeconstructedaphylogenetichypothesisbasedonITSsequencesretrievedfromGenBank forsevenCrotalariaspeciesandfortheout-groupBolusiaamboensis[36].TotalgenomicDNA ofC.vitellina(unavailableinGenBank)wasextractedfromfreshleaves,collectedinthegarden ofUNICAMP,usingthetwo-foldhexadecyltrimethylammoniumbromide(CTAB)method [55].ThenuclearribosomalITSregion(ca.700bp)wasamplifiedasdescribedbyleRouxetal. [36],usingtheprimersITS17SEandITS26SE[56,57].PCRproductswerepurifiedwitha GFXPCRDNAandGelBandPurificationKit(GEHealthcare)andsequencedattheCenterof PLOSONE|DOI:10.1371/journal.pone.0141480 October30,2015 10/27
Description: