Epichloë endophytes alter inducible indirect defences in host grasses

Epichloe ¨ Endophy es Al e Inducible Indi ec De ences in Hos G asses Tao Li 1 *, James D. Blande 1 , Ped o E. Gundel 2¤ , Ma jo Helande 3 , Ka i Saikkonen 2 1Depa men o En i onmen al Science, Uni e si y o Eas e n Finland, Kuopio, Finland, 2Plan P oduc ion Resea ch, MTT Ag i ood Resea ch Finland, Jokioinen, Finland, 3Depa men o Biology, Uni e si y o Tu ku, Tu ku, Finland Abs ac Epichloe ¨endophy es a e common symbion s li ing asymp oma ically in pooid g asses and may p o ide chemical de ences agains he bi o ous insec s. While he mechanisms unde lying hese ungal de ences ha e been well s udied, i emains unknown whe he endophy e p esence a ec s he hos ’s own de ences. We add essed his issue by examining a ia ion in he impac o Epichloe ¨on cons i u i e and he bi o e-induced emissions o ola ile o ganic compounds (VOC), a well-known indi ec plan de ence, be ween wo g ass species, Schedono us phoenix (ex. Fes uca a undinacea; all escue) and Fes uca p a ensis (meadow escue). We ound ha eeding by a gene alis aphid species, Rhopalosiphum padi, induced VOC emissions by unin ec ed plan s o bo h g ass species bu o a ying ex en s, while mechanical wounding ailed o do so in bo h species a e one day o damage. In e es ingly, ega dless o damage ea men , Epichloe ¨uncina a-in ec ed F. p a ensis emi ed signi ican ly lowe quan i ies o VOCs han hei unin ec ed coun e pa s. In con as , Epichloe ¨coenophiala-in ec ed S. phoenix did no di e om hei unin ec ed coun e pa s in cons i u i e VOC emissions bu ended o inc ease VOC emissions unde in ense aphid eeding. A mul i a ia e analysis showed ha endophy e s a us imposed s onge di e ences in VOC p o iles o F. p a ensis han damage ea men , while he e e se was ue o S. phoenix. Addi ionally, bo h endophy es inhibi ed R. padi popula ion g ow h as measu ed by aphid d y biomass, wi h he inhibi ion appea ing g ea e in E. uncina a-in ec ed F. p a ensis. Ou esul s sugges , no only ha Epichloe ¨endophy es may play impo an oles in media ing hos VOC esponses o he bi o y, bu also ha he magni ude and di ec ion o such esponses may a y wi h he iden i y o he Epichloe ¨–g ass symbiosis. Whe he Epichloe ¨-media ed hos VOC esponses will e en ually ansla e in o e ec s on highe ophic le els me i s u u e in es iga ion. Ci a ion: Li T, Blande JD, Gundel PE, Helande M, Saikkonen K (2014) Epichloe ¨Endophy es Al e Inducible Indi ec De ences in Hos G asses. PLoS ONE 9(6): e101331. doi:10.1371/jou nal.pone.0101331 Edi o : Benedic e Ribe Alb ec sen, Umea ˚Plan Science Cen e, Umea ˚Uni e si y, Sweden Recei ed Ma ch 10, 2014; Accep ed June 4, 2014; Published June 30, 2014 Copy igh : ß2014 Li e al. This is an open-access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License, which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal au ho and sou ce a e c edi ed. Da a A ailabili y: The au ho s con i m ha all da a unde lying he indings a e ully a ailable wi hou es ic ion. All da a a e included wi hin he pape and i s Suppo ing In o ma ion iles. Funding: This esea ch was unded by he Academy o Finland (P ojec nos. 256050, 251898 and 137909). The unde s had no ole in s udy design, da a collec ion and analysis, decision o publish, o p epa a ion o he manusc ip . Compe ing In e es s: The au ho s ha e decla ed ha no compe ing in e es s exis . * Email: [email p o ec ed] ¤ Cu en add ess: Ca ´ ed a de Ecologı ´a, Facul ad de Ag onomı ´a (UBA), IFEVA-CONICET, Ciudad Au o ´noma de Buenos Ai es, Buenos Ai es, A gen ina In oduc ion Plan s o m in ima e associa ions wi h a my iad o mic oo gan- isms, ei he de imen al o bene icial [1–5]. One o he mos widely s udied associa ions is he symbio ic de ensi e mu ualism be ween Po¨o¨ideae g asses and endophy ic ungi o he genus um Epichloe¨, due o hei signi ican impac s on insec and mammalian he bi o es, pa icula ly in ag icul u al pas o al sys ems [2–4,6–8]. In exchange o shel e , nu i ion and ansmission ia hos seeds, he endophy es may inc ease hos g ow h and ep oduc ion, s ess ole ance and he bi o e and pa hogen esis ance (e.g. [1,9]). Mo eo e , hey may also ha e communi y-wide impac s by a ec ing seconda y consume s and al e ing in e plan compe i ion [4,10–12]. The endophy e-con e ed he bi o e esis ance is o en a ibu - ed o he di ec induc ion o biologically ac i e alkaloids by he endophy e, which may ad e sely a ec he bi o e pe o mance [9,13–16]. Howe e , he alkaloid p o iles and concen a ions may a y conside ably among g ass-endophy e sys ems and en i on- men al condi ions, leading o no e ec s, o e en posi i e e ec s, on he bi o es [7,15–19]. These a iable e ec s indica e ha apa om endophy e-con e ed alkaloid de ence, addi ional as-ye - undisco e ed mechanisms such as endophy e-media ed changes in hos de ence chemis y a e likely o be implica ed in complex endophy e-hos -insec in e ac ions. In plan -mic obe in e ac ions some bene icial mic obes (e.g. myco hizal ungi and plan g ow h- p omo ing hizobac e ia) ha e been shown o a ec he bi o e pe o mance by inducing hei hos plan ’s own de ences [20,21]. Ye , his has been la gely dis ega ded in p e ious s udies conce ning g ass-endophy e symbioses [16,22]. Among he mul i ude o plan esponses o bio ic and abio ic s imuli is he elease o complex blends o ola ile o ganic compounds (VOCs) ha consis mainly o e penoids, a y acid de i a i es and phenylp opanoids. Plan VOCs play an impo an ecological ole in shaping he assemblage o , and in e ac ions be ween, o ganisms wi hin a plan ’s communi y [23–27]. Fo example, he bi o ous insec s ely, o a ying ex en s, on plan VOCs as ol ac o y cues o ecognize hos plan s and a oid non- hos plan s [23,24]. On he o he hand, he bi o e a ack elici s PLOS ONE | www.plosone.o g 1 June 2014 | Volume 9 | Issue 6 | e101331 bo h local and sys emic emission o plan VOCs, which in u n may p o ec plan s om u he damage ei he di ec ly h ough de e ing o epelling he bi o es o indi ec ly h ough a ac ing na u al enemies o he bi o es [24,25]. In e es ingly, accumula ing e idence has e ealed ha coloniza ion by plan g ow h-p omo - ing hizobac e ia o a buscula myco hizal ungi can also modi y cons i u i e o he bi o e-induced VOC emissions o colonized plan s and he eby al e VOC-media ed mul i ophic in e ac ions [28–36], al hough o en in unp edic able ways. Li le is known abou he e ec s o ungal endophy es on VOC emissions o hos plan s and he cascading e ec s on he beha iou o insec s a di e en ophic le els. The ew s udies on endophy e- induced al e a ion o VOC blends ha e ound ha VOC emissions may be enhanced [37], educed [38], o unchanged by endophy e p esence [20] depending on plan and endophy e species. Howe e , hese s udies mainly in ol e ho izon ally ansmi ed ungal endophy es and cons i u i e VOC emissions. In con as o ho izon ally ansmi ed endophy es, e ically ansmi ed ungal Epichloe¨ endophy es o m a li e-long and he i able symbiosis wi h hei hos g asses (e.g. [1,9]), and hence may al e he hos ’s chemis y di e en ly. Fu he mo e, sys emic endophy es ha e long domina ed he li e a u e p esumably because o hei ag onomic impac (e.g. [2,9]), bu hei e ec s on cons i u i e and he bi o e- induced VOC emissions ha e been poo ly explo ed. The only in o ma ion on his comes om an ea ly s udy o Epichloe¨ coenophiala—in ec ed all escue (Schedono us phoenix) [39] and a e y ecen s udy o Epichloe¨ es ucae a . lolii-in ec ed pe ennial yeg ass (Lolium pe enne) [40], bo h demons a ing he po en ial e ec o Epichloe¨ endophy es on hos ’s VOC elease bu ailing o de e mine he impac o he bi o y. Rega dless o which g oup hey belong o, elucida ing he endophy e e ec s on bo h cons i u i e and he bi o e-induced VOC emissions in endophy e-g ass in e ac ions would no only help o unde s and he a iable endophy e e ec s on he bi o y, bu also o e alua e he po en ial o endophy es as biological con ol ools. We in es iga ed whe he and how Epichloe¨ endophy es al e ed cons i u i e and he bi o e-induced VOC emissions in wo escue species, all escue (symbio ic wi h Epichloe¨ coenophiala) and meadow escue (Fes uca p a ensis, symbio ic wi h Epichloe¨ uncina a). To examine he bi o e-induced VOC emissions, we used a gene alis aphid species, Rhopalosiphum padi (bi d che y-oa aphid), a common pes o g asses ha has been ex ensi ely used as a model phloem eede in g ass endophy e esea ch (e.g. [9]). Speci ically, we add essed he ollowing ques ions: (1) Does aphid eeding induce VOC emissions in g asses? (2) Does Epichloe¨ in ec ion modi y VOC emissions by i s hos ? (3) Is he e any a ia ion in hese esponses o he wo g ass species? We discuss how VOC emissions by he hos could in e ac wi h endophy e-con e ed de ence. Ma e ials and Me hods E hics s a emen This s udy was conduc ed in he labo a o y and did no in ol e any endange ed o p o ec ed species; he insec species used in his s udy is a se ious ce eal pes wo ldwide, including Finland. Hence, no speci ic e hical app o al was equi ed o his s udy. Plan s, endophy es and Insec s Seeds o na u ally endophy e in ec ed (E+) and endophy e- ee (E-) all escue (Schedono us phoenix) cul i a ‘Ken ucky 319and meadow escue (Fes uca p a ensis) cul i a ‘Kaspe ’ we e collec ed om expe imen al ields in he Uni e si y o Tu ku Ruissalo Bo anical Ga den. Fo all escue, in addi ion o E+and E- plan s, manipula i ely endophy e- ee (ME-) plan s, which we e ob ained om E+plan using hea ea men , we e included o sepa a e he e ec s o endophy e in ec ion om plan esponses. Because hos plan s o di e en endophy e s a us a e om a single cul i a , gene ic a ia ion and i s impac on plan de ence esponse migh be limi ed compa ed o wild plan s. The in ec ion s a us o he plan s was e i ied by g owing ou he ungus om su ace s e ilized g ass lea cu ings pla ed on po a o dex ose aga (PDA) in Pe i dishes [41,42]. Plan s we e g own indi idually in 12-cm- diame e plas ic po s illed wi h a s anda d po ing soil in a g eenhouse [pho osyn he ically ac i e adia ion (PAR) a canopy le el ca. 300 mmol m 22 s 21 ]. App oxima ely 60 days a e seedling ansplan a ion, plan s we e ans e ed o a clima e-con olled oom (18–24uC, 70% RH, L16:D8 pho ope iod, and ca. 250 mmol m 22 s 21 PAR) in he labo a o y, whe e all expe imen s we e conduc ed. Rhopalosiphum padi we e ob ained om a colony a he Depa men o Ecology a he Swedish Uni e si y o Ag icul u al Sciences in Uppsala, Sweden, and we e ea ed on ba ley in a clima e-con olled oom (16L:8D, 18–24uC, and 70% RH). Plan ea men Expe imen s o bo h escue species ollowed a ull- ac o ial expe imen al design wi h wo ac o s, endophy e in ec ion and damage in lic ed ei he by aphid in es a ion o mechanical wounding. Be o e he s a o he expe imen s, a o al o 18 plan s wi hou isible damage we e selec ed om each endophy e s a us (E-, ME- and E+ o all escue; E- and E+ o meadow escue) and di ided in o six g oups acco ding o plan size, wi h plan s in each g oup cha ac e ized by simila size. The h ee plan s wi hin each g oup we e hen andomly subjec ed o he ollowing h ee ea men s: 1) aphid in es a ion (A): plan s we e in es ed by placing i e ba ley lea segmen s con aining a o al o 50 mixed-ins a nymphs and ap e ae be ween ille s and emo ing hese segmen s a e all aphids had mo ed on o he plan s; 2) wounding (W): plan s we e mechanically damaged by i s cu ing ca. 4 cm o e e y lea ip wi h scisso s and hen squeezing he emaining lea blade eigh imes wi h o ceps; and 3) con ol (C): con ol plan s ecei ed no damage. Mechanical wounding was mean o mimic he damage caused by animal g azing and ampling. In o al, he e we e nine and six endophy e 6damage ea men s o all and meadow escue, espec i ely, each con aining six plan s. To a oid aphids mo ing o neighbou ing plan s, all plan s including hose wi hou aphids we e placed in sc eened cages wi h cage posi ions o a ed daily o con ol o any di e ences in ligh o empe a u e condi ions. VOCs om wounded and aphid-in es ed plan s we e collec ed a di e en imes ollowing ea men . Mechanical damage ypically elici s apid elease o VOCs, pa icula ly g een lea ola iles (GLVs; e.g. [43]). Consis en wi h his, we ound in a p elimina y s udy wi h ed escue (Fes uca ub a) ha 40 min a e mechanical wounding simila o ha in his s udy, GLV emissions we e se e al hund ed- old highe compa ed o p e-damage emissions (Figu e S1). A e one day, GLV emissions emained subs an ially highe han p e-damage emissions; howe e , an induc ion o ew e penoids (e.g. b-ocimene) mani es ed i sel . The e o e, we collec ed VOCs om mechanically wounded plan s one day a e wounding o cap u e esponses o bo h GLV and e penoids compounds. Fo meadow escue, VOC collec ion was also pe o med a six days pos -wounding. Howe e , since VOC esponses induced by phloem eede s – which in lic minimal issue damage – p oceeds slowly wi h a delay o se e al days [34,44], we collec ed VOCs om aphid-in es ed plan s a 6 and 12 days a e in es a ion. This iming was chosen acco ding o he ime cou ses Endophy es Modi y Hos Vola ile Release PLOS ONE | www.plosone.o g 2 June 2014 | Volume 9 | Issue 6 | e101331 o aphid-induced VOC induc ion epo ed in he li e a u e [34,44]. A e he las VOC sampling, he abo eg ound plan pa s, which had been enclosed du ing VOC collec ion, we e ha es ed, o en-d ied and weighed. Owing o high numbe s o aphid o sp ing on in es ed plan s, in pa icula ly on in es ed E- plan s, o al d y weigh o aphids pe plan (including dead ones p esen on he plan s) a he han aphid numbe was de e mined and used as a p oxy o aphid p opaga ion and g ow h o assess he endophy e e ec s on aphid pe o mance. VOC collec ion and analysis VOC collec ion was conduc ed as in Li e al. [27]. In b ie , he po s plus soil we e ca e ully w apped wi h aluminium oil o p e en con amina ion wi h soil-de i ed ola iles. The ae ial pa o each po ed plan was hen enclosed in a polye hylene e eph hala e (PET) bag and sealed wi h a plas ic-coa ed wi e. Cha coal il e pu i ied ai was pushed h ough Te lon ubing in o each bag (230 ml min 21 ) and pulled ou by a acuum pump (200 ml min 21 ) h ough a s ainless s eel ap packed wi h 150 mg o Tenax TA and 150 mg o Ca bopack B (Ma kes In e na ional, Llan isan , RCT, UK). VOCs we e collec ed o 1 h and simul aneously om plan s o di e en ea men s; pe iodic collec ions o VOCs om emp y PET bags we e also made. Aphids we e kep on in es ed plan s du ing collec ion since he p esence o aphids has been shown o con ibu e li le i any hing o he VOC blends emi ed by he plan -aphid complex [34,44]. Fu he mo e, i is likely ha emo ing aphids om plan s will esul in a deg ee o mechanical damage o he plan and may esul in aphids emi ing hei ala m phe omones. VOC samples we e analyzed by GC-MS (Hewle -Packa d GC 6890; MSD 5973; Wilming on, DE, USA). T apped ola iles we e deso bed wi h a he mal deso p ion uni (ATD400; Pe kin Elme , Wal ham, MA, USA) a 250uC o 10 min, ocused a 210uCona cold ap and ans e ed on o an HP-5 capilla y column (50 m60.2 mm; ilm hickness 0.5 mm) wi h helium as he ca ie gas (1.2 ml min 21 ). The column empe a u e was ini ially held a 40uC o 1 min, hen amped o 210uCa 5uC min 21 , and inally o 250uCa 20uC min 21 . Indi idual VOCs we e en a i ely iden i ied by compa ing mass spec a wi h hose in NIST and Wiley spec al lib a ies and e i ied by ch oma og aphy wi h au hen ic s anda ds when a ailable. Al hough o e 50 p ominen peaks could be de ec ed, only hose ha we e ound consis en ly highe in he samples han in he blanks we e conside ed in u he analyses, allowing iden i ica ion o 17 compounds in bo h g ass species. Fo quan i ica ion, peak a eas o cha ac e is ic quan i ie ions we e in eg a ed and he amoun o each compound was calcula ed based on ex e nal calib a ion cu es gene a ed wi h au hen ic s anda ds. Fo compounds whose e e ence s anda ds we e no a ailable, quan i ica ion was assessed ela i e o he ex e nal s anda d 1-chlo ooc ane. Emission a es we e p esen ed in nanog ams pe g am d y weigh pe hou (ng g 21 DW h 21 ). S a is ical analysis To analyse he main e ec s o endophy e (E- e sus E+), aphid (C e sus A), sampling da e (6 days e sus 12 days pos -in es a ion) and hei in e ac ion on emissions o o al and single VOCs, we used a linea mixed model (LMM) o each g ass species, wi h endophy e and aphid as he be ween-subjec s ac o s and ime as he wi hin-subjec s ac o . Since R. padi aphids pe o med badly on E+plan s ela i e o E- plan s (see ‘Resul s’), E+plan s may expe ience less lea damage o e he in es a ion pe iod. To accoun o any po en ial e ec s o di e en ial aphid damage on VOC emissions, we used he o al aphid d y weigh pe plan as an Figu e 1. To al VOC emissions (mean ±1SE) om na u ally endophy e ee (E-; ci cles) s in ec ed (E + ; iangles) all escue (a) and meadow escue (b) in esponse o aphid o mechanical damage. C: un ea ed con ol; W: mechanical wounding; A: aphid in es a ion. Fo compa ison, manipula i ely endophy e ee all escue (ME-; squa es) was included. All damage ea men s we e ini ia ed a he same ime, hen VOC collec ions conduc ed a 1 day a e mechanical wounding, o a 6 and 12 days a e aphid addi ion. S a is ical de ails a e shown in ables 1 and 2. doi:10.1371/jou nal.pone.0101331.g001 Endophy es Modi y Hos Vola ile Release PLOS ONE | www.plosone.o g 3 June 2014 | Volume 9 | Issue 6 | e101331 Table 1. Summa y o LMM esul s showing he le el o signi icance o he e ec o endophy e (E), aphid (A) and sampling ime (T) on VOC emissions o all escue. E A T In e ac ion Compound nd F P F P F P P Te penoids a-pinene 56.89 10.92 , 0.001 1.46 0.231 100.09 , 0.001 6-me hyl-5-hep en-2-one { 57.39 0.38 0.683 1.02 0.317 0.14 0.708 b-my cene 53.21 0.84 0.437 1.21 0.277 2.32 0.133 E x A: 0.010 b-pinene 36.72 0.31 0.738 0.00 0.970 3.25 0.080 d-ca ene 52.75 0.86 0.427 2.72 0.105 1.63 0.208 (Z)-b-ocimene { 58.65 2.97 0.059 4.47 0.039 0.04 0.833 d-limonene 49.70 0.69 0.506 0.00 0.997 1.39 0.244 E x A: 0.069 b-phelland ene { 47.08 0.22 0.806 0.09 0.764 5.71 0.021 (E)-b-ocimene 59.54 1.61 0.208 2.97 0.090 0.10 0.759 a- e pinolene 55.75 0.23 0.796 0.04 0.838 1.96 0.167 linalool* 57.57 1.15 0.325 10.23 0.002 1.46 0.232 Unkown mono e pene { 59.93 5.03 0.010 1.12 0.295 6.60 0.013 (E)-b-ca yophylene* 56.06 5.75 0.005 5.57 0.022 0.60 0.442 To al Te penoids 59.16 2.75 0.072 2.19 0.144 1.58 0.213 G een lea ola iles (GLV) (Z)-3-hexen-1-ol* 60.00 0.72 0.491 2.76 0.102 0.02 0.899 E x A: 0.061 (Z)-3-hexen-1-ol ace a e 59.96 0.76 0.472 0.92 0.341 2.35 0.130 To al GLV 60.00 0.74 0.482 1.13 0.292 2.01 0.161 O he compounds 1-oc en-3-ol* 42.60 1.43 0.250 282.62 , 0.001 44.21 , 0.001 AxT: , 0.001 me hyl salicyla e 57.75 3.32 0.043 5.67 0.021 0.06 0.811 A x T: 0.043 To al VOCs 59.99 0.68 0.512 2.07 0.155 2.23 0.142 Bold numbe s indica ed e ec s wi h P alues less han 0.1 as de e mined by LMM. nd ep esen s he nume a o deg ees o eedom; he denomina o deg ees o eedom (dd ) we e 2 o E and 1 o A and T. Fo de ails on emission a es see Table S1 and S2. *Compounds ha we e iden i ied by PLS-DA analysis as mos in luen ial o sepa a ion o indi idual ea men s a 12 days pos aphid eeding. { Compounds en a i ely iden i ied. doi:10.1371/jou nal.pone.0101331. 001 Endophy es Modi y Hos Vola ile Release PLOS ONE | www.plosone.o g 4 June 2014 | Volume 9 | Issue 6 | e101331 indica o o he ex en o plan damage, and e-analyzed aphid- induced VOCs by including aphid d y weigh as a co a ia e in he LMM. In his model, endophy e, sampling da e and hei in e ac ion we e ixed ac o s. To examine endophy e by wounding e ec s, we pe o med wo-way ANOVA. E ec s o endophy e on aphid d y weigh and e ec s o endophy e by aphid on plan d y mass we e analysed by one-way and wo-way ANOVA, espec i ely. Da a we e log ans o med [log(X+1)] o mee no mali y and homoscedas ici y. All analyses we e pe o med using he s a is ical package SPSS 19.0 o windows. To isualize di e ences in VOC blends o di e en ly ea ed plan s, da a we e also analysed wi h Pa ial Leas P ojec ion o La en S uc u es-Disc iminan Analysis (PLS-DA) (SIMCA- P11.0; Ume ics, Umea˚, Sweden). To p ep ocess da a, emission a es o indi idual VOCs we e no malized [log(X+1)], mean- cen ed and scaled o uni a iance. The numbe o signi ican PLS componen s was de e mined by c oss alida ion [45]. This me hod allows no only o isualiza ion o high dimensional da a in sco e plo s, bu also o iden i ica ion o a iables (i.e. ola ile compounds) ha a e impo an o he di e ences in complex VOC blends among ea men s. In gene al, a iables wi h he Va iable Impo ance in he P ojec ion (VIP) sco es la ge han 1 a e conside ed mos in luen ial o he model. Resul s VOC emissions by all escue The e was no signi ican e ec o endophy es (LMM; F (2, 59.99) = 0.68, P= 0.512) and aphids (F (1, 59.99) = 2.23, P= 0.155) on o al VOC emissions, hough aphid-in es ed E+plan s appea ed o inc ease o al emissions a e 12 days o eeding (Figu e 1a; Tables 1, S1 and S2). Rega ding indi idual compounds, endo- phy e p esence dec eased emissions o (Z)-b-ocimene (P= 0.059) and (E)-b-ca yophyllene (P= 0.005) i espec i e o eeding ea - men , whe eas endophy e emo al inc eased emissions o a-pinene (P,0.001), an unknown mono e pene (P= 0.010) and me hyl salicyla e (P= 0.043). A e con olling o possible e ec s o di e en ial le els o lea damage, simila esul s we e obse ed o he e ec s o endophy e p esence on VOC emissions om aphid- in es ed plan s (Table S3). Aphid eeding induced di e en ial esponses o se e al compounds depending on endophy e s a us and eeding du a ion. A 12 days pos - eeding, o example, in es ed E+plan s emi ed highe amoun s o (Z)-3-hexen-1-ol (P= 0.073) and (Z)-3-hexen-1-ol ace a e (P= 0.093) han con ol E+plan s, while in es ed E- plan s emi ed subs an ially mo e linalool (P= 0.016), (E)-b-ca yophyllene (P= 0.088) and (Z)-3- hexen-1-ol (P= 0.038) han con ol E- plan s (Figu e 2; Table S1). The mos no able aphid e ec was he de no o induc ion o 1- oc en-3-ol (P,0.001), which was eleased exclusi ely om in es ed plan s independen ly o endophy e s a us and in signi ican ly highe amoun s a 12 days a e aphid addi ion han a 6 days (P, 0.001). The e was no signi ican e ec o mechanical damage on ei he o al o indi idual VOC compounds (Table S4). VOC emissions by meadow escue Unlike all escue, meadow escue exhibi ed signi ican ly educed emissions o o al VOCs in he p esence o endophy e (LMM; F (1, 36.99) = 17.47, P,0.001; Figu e 1b; Tables 2, S5 and S6) as well as emission o nine componen s, including 6-me hyl-5- hep en-2-one (P= 0.012), b-my cene (P,0.001), d-limonene (P, 0.001), b-phelland ene (P,0.001), a- e pinolene (P= 0.001), (Z)-3- hexen-1-ol (P= 0.006), (Z)-3-hexen-1-ol ace a e (P= 0.002), 1- oc en-3-ol (P= 0.002), and he same unknown mono e pene (P, 0.001) de ec ed in all escue. These emission educ ions we e consis en ly obse ed a all h ee sampling imes ega dless o whe he plan s we e damaged o no . Mo eo e , educed emissions o VOCs by aphid-in es ed E+plan s we e s ill e iden a e adjus ing o po en ial e ec s o di e en ial lea damage (Table S7). Aphid in es a ion did no a ec o al emissions (P= 0.755) ei he a 6 o 12 days o in es a ion, bu induced emissions o b-my cene (P= 0.036), (Z)-b-ocimene (P= 0.002), d- limonene (P= 0.005), (E)-b-ocimene (P= 0.025), (E)-b-ca yophyl- lene (P= 0.063) and 1-oc en-3-ol (P,0.001) a 12 days whils dep essing (Z)-3-hexen-1-ol ace a e emission (P= 0.008) (Figu e 3; Table S5). Again, he mos p onounced e ec o aphid eeding was he de no o induc ion o 1-oc en-3-ol as obse ed in all escue. Addi ionally, he e was a signi ican o ma ginally signi ican in e ac ion be ween endophy es and aphids on emissions o 6- me hyl-5-hep en-2-one (P= 0.051), (Z)-3-hexen-1-ol (P= 0.070), (Z)-3-hexen-1-ol ace a e (P= 0.079) and 1-oc en-3-ol (P= 0.002), wi h in es ed E+plan s emi ing less o hese compounds han plan s in any o he ea men . Nei he mechanical damage no i s in e ac ion wi h endophy es a ec ed VOC emissions a 1 day ollowing damage (Table S8). Visualiza ion o di e ences in VOC p o iles among ea men s Fo all escue VOC p o iles o aphid-in es ed plan s we e dis inc ly di e en om con ol plan s and he di e ence became mo e appa en a 12 days pos -in es a ion (Figu e 4) han a 6 days (Figu e S2), whe eas he p o iles o ME-, E- and E+plan s we e ela i ely simila o each o he . In con as , in meadow escue a clea dis inc ion be ween VOC blends o E- and E+plan s could be depic ed a all h ee sampling days al hough a small o e lap was seen (Figu es 4, S2 and S3). VOC blends om con ol and aphid- in es ed plan s g ouped sepa a ely om each o he only a 12 days a e in es a ion (Figu e 4), wi h he mos p onounced sepa a ion occu ing be ween in es ed E- plan s and plan s in all o he Figu e 2. Indi idual VOCs o all escue whose elease changed signi ican ly in esponse o endophy e in ec ion and 12 days o aphid eeding. E-: na u ally endophy e ee; E+: na u ally endophy e in ec ed; ME-: manipula i ely endophy e ee; C: un ea ed con ol; A: aphid damage. S a is ical de ails a e shown able S1. doi:10.1371/jou nal.pone.0101331.g002 Endophy es Modi y Hos Vola ile Release PLOS ONE | www.plosone.o g 5 June 2014 | Volume 9 | Issue 6 | e101331 Table 2. Summa y o LMM esul s showing he le el o signi icance o he e ec o endophy e (E), aphid (A) and sampling ime (T) on VOC emissions o meadow escue. E A T In e ac ion Compound nd FP FP FP P Te penoids a-pinene 29.60 0.02 0.879 0.50 0.484 25.59 , 0.001 6-me hyl-5-hep en-2-one { * 38.88 6.98 0.012 0.13 0.720 5.40 0.025 ExA:0.051 b-my cene* 30.14 19.77 , 0.001 4.79 0.036 1.33 0.259 b-pinene 33.42 0.20 0.659 0.87 0.356 7.70 0.009 d-ca ene 39.56 0.07 0.796 0.83 0.369 11.66 0.001 (Z)-b.ocimene { * 35.66 0.00 0.961 11.42 0.002 1.36 0.250 d-limonene* 39.24 17.98 , 0.001 8.97 0.005 9.38 0.004 b-phelland ene { 39.86 15.75 , 0.001 0.25 0.621 0.10 0.752 (E)-b-ocimene* 35.52 0.37 0.548 5.46 0.025 0.96 0.334 a- e pinolene 39.62 12.47 0.001 1.43 0.239 29.77 , 0.001 linalool 37.38 1.77 0.192 0.90 0.349 11.45 0.002 Unkown mono e pene { * 37.64 14.57 , 0.001 0.99 0.326 1.17 0.286 (E)-b-ca yophylene* 35.23 0.89 0.353 3.69 0.063 0.03 0.856 E x A x T: 0.073 To al Te penoids 35.78 16.14 , 0.001 12.22 0.001 0.01 0.913 G een lea ola iles (GLV) (Z)-3-hexen-1-ol* 39.46 8.35 0.006 1.08 0.305 0.17 0.685 E x A: 0.070 (Z)-3-hexen-1-ol ace a e* 38.73 10.95 0.002 7.77 0.008 0.01 0.908 E x A: 0.079 To al GLV 38.72 11.42 0.002 7.56 0.009 0.01 0.903 E x A: 0.069 O he compounds 1-oc en-3-ol* 29.31 11.83 0.002 51.63 , 0.001 39.63 , 0.001 All: , 0.002 me hyl salicyla e 38.16 3.41 0.072 0.23 0.632 0.25 0.618 To al VOC 36.99 17.47 0.000 0.10 0.755 0.73 0.397 *and{as in Table 1. See Table S5 and S6 o de ails on emission a es. doi:10.1371/jou nal.pone.0101331. 002 Endophy es Modi y Hos Vola ile Release PLOS ONE | www.plosone.o g 6 June 2014 | Volume 9 | Issue 6 | e101331 ea men s. In bo h species, wound-induced VOC blends we e no sepa a ed om cons i u i e VOC blends (Figu e S3). In bo h species, he compounds ha con ibu ed mos s ongly o he di e ences among VOC blends o di e en ly ea ed plan s we e ound o be he same compounds ha had been demon- s a ed abo e o be signi ican ly induced by endophy es, aphids and/o hei in e ac ion (Tables 1 and 2). Among hem, 1-oc en-3- ol had he s onges disc imina o y powe in di e en ia ing in es ed and con ol plan s. Aphid and plan g ow h Endophy e p esence educed aphid pe o mance as aphid colonies g own on E+plan s had much lowe d y weigh s han on E- plan s o bo h all (F (2, 18) = 12.19, P,0.001) and meadow escue (F (1, 12) = 52.37, P,0.001; Figu e S4). In e es ingly, in all escue aphids pe o med be e on ME- plan s han on E+plan s bu no as well as on E- plan s, and ME- plan s had lowe d y mass han bo h E- and E+plan s (Figu e S5). This sugges s ha ei he endophy e emo al o hea ea men o emo e endophy es o bo h may hinde plan g ow h and aphid pe o mance. Addi ion- ally, sho - e m aphid in es a ion did no in luence plan g ow h o ei he all (F (2, 54) = 0.92, P= 0.406) o meadow escue (F (2, 36) = 0.05, P= 0.955; Figu e S5). Discussion Ou esul s e eal ha eeding by he gene alis aphid R. padi induces VOC emissions in all and meadow escue ha a e known o se e in plan di ec and indi ec de ences agains he bi o e a ack. Mos impo an ly, we demons a e o he i s ime ha Epichloe¨ endophy es a e capable o modi ying bo h cons i u i e and aphid-induced VOC emissions o hos g asses. Mo eo e , he wo s udied escues di e subs an ially in VOC emission pa e ns in esponse o aphid in es a ion and endophy e in ec ion. Al hough in ensi ely cha ac e ized in a ange o plan species (e.g. [24]), VOC p o iles ha e a ely been s udied in Fes uca, pa icula ly in esponse o he bi o y. The only ele an s udy in ol es induced VOC emissions by all escue upon exogenous applica ion o jasmonic acid (JA) [39], a phy oho mone widely used o mimic induced de ence esponses o lea -chewing he bi o es. Ou esul s he e o e expand upon p e ious indings, showing clea induc ion o VOC emissions om all escue, and also om meadow escue, in esponse o he phloem-sucking he bi o e R. padi. Compa ed o JA ea men , which inc eased amoun s o se e al cons i u i e VOCs [39], we ound ha aphid eeding on all escue enhanced emissions o ewe componen s. This comes as no su p ise gi en ha lea -chewe s (o JA applica ion) and phloem-sucke s ac i a e JA and salicylic acid (SA) signaling pa hways, espec i ely and ha hese wo pa hways o en an agonize each o he and elici emissions o a di e en se o ola ile compounds [16,26]. Addi ionally, phloem-sucke s, which usually cause limi ed cell damage, do no elici ola ile esponses as s ongly as lea -chewe s [34,44]. Se e al compounds, including wo b-ocimene isome s, we e induced by aphid eeding in meadow escue bu no in all escue, sugges ing ha meadow escue is mo e esponsi e han all escue when challenged by aphids. Howe e , he e was some commonali y in he induced VOCs, mos no ably 1-oc en-3-ol, which was he only de no o syn hesized compound induced by aphid eeding and posi i ely ela ed o aphid densi y. 1-oc en-3-ol is emi ed by squash plan s in ec ed wi h powde y mildew (Podosphae a sp.), and is a pa icula componen o ‘mouldy odou ’ and a ac i e o mycophagous wen y-spo ed ladybi d bee les (Psyllobo a igin imacula a) [46]. This compound is also an induced ola ile o T i olium p a ense a e damage by Spodop e a Figu e 3. Indi idual VOCs o meadow escue which changed signi ican ly in esponse o endophy e in ec ion and 12 days o aphid eeding. E-: na u ally endophy e ee; E+: na u ally endophy e in ec ed; C: un ea ed con ol; A: aphid damage. S a is ical de ails a e shown able S5. doi:10.1371/jou nal.pone.0101331.g003 Endophy es Modi y Hos Vola ile Release PLOS ONE | www.plosone.o g 7 June 2014 | Volume 9 | Issue 6 | e101331 li o alis ca e pilla s [47]. The e o e, 1-oc en-3-ol may se e as a componen o mul iple indi ec de ence esponses. Unlike he bi o y, mechanical damage in lic ed by single wounding e en s o en igge s apid VOC elease, which hen d ops o he p e ea men le el wi hin a ew hou s (e.g. [43]). Howe e , in some cases VOC esponses induced by mechanical damage may equi e days o appea [48]. In line wi h hese indings, ou p elimina y s udy wi h ed escue e ealed a apid ise and p olonged emission. Unexpec edly, in his s udy we did no de ec induced VOC emissions in ei he all o meadow escue one day a e mechanical wounding. Howe e , we may ha e missed a apid VOC esponse by p o iding a one-day eco e y pe iod ollowing wounding be o e collec ing VOCs. Ne e heless, hese s udies sugges ha he speed and du a ion o VOC esponses induced by mechanical damage may di e among plan species. In e es ingly, ou s udy e eals ha Epichloe¨ endophy es a ec bo h cons i u i e and aphid-induced VOC emissions o hos g asses. In he absence o aphids, E. uncina a-in ec ed meadow escue had signi ican lowe emission a es han unin ec ed coun e pa s, whe eas E. coenophiala-in ec ed all escue did no di e om unin ec ed plan s. When subjec ed o aphid a ack, endophy e-in ec ed meadow escue s ill eleased signi ican ly lowe amoun s o VOCs, bu in ec ed all escue ended o inc ease VOC emissions a e 12 days o con inuous eeding. In pa icula , endophy es and aphids appea ed o ac syne gis ically o supp ess p oduc ion o he wo dominan compounds (Z)-hexen-1-ol and (Z)-3-hexen-1-ol ace a e in meadow escue bu p omo e hei elease in all escue. The di e ences in aphid-induced VOC elease be ween endophy e- ee and in ec ed plan s emained p onounced e en a e accoun ing o po en ial e ec s o di e en ex en s o damage, sugges ing ha endophy e-media ed changes in hos VOC elease migh occu ega dless o in ensi y o he bi o y. One ca ea o ou s udy is ha e en hough single cul i a s we e used (Ken ucky 31 o all escue; Kaspe o meadow escue), hos geno ype was no s ic ly con olled and may ha e in luenced VOC emissions alone o in e ac i ely wi h endophy es. In line wi h ou indings, an ea ly s udy wi h all escue has shown ha E. coenophiala- in ec ed plan s did no di e in cons i u i e VOC emissions om unin ec ed coun e pa s, bu doubled (Z)-3-hexen-1-ol ace a e emission in esponse o JA ea men while dec easing emissions o a ew e penoids such as (E)-b-ocimene [39]. Howe e , a ecen s udy o pe ennial yeg ass has ound g ea e quan i ies o bo h cons i u i e and pa hogen- elici ed VOC emissions emi ed by E. es ucae a . lolii-in ec ed plan s han unin ec ed coun e pa s [40]. Toge he , all hese s udies sugges ha endophy e-media ed adjus men o hos VOC p oduc ion may a y wi h he iden i y o Epichloe¨-g ass symbiosis and he ype o bio ic s ess. Vola ile compounds may ac as plan de ensi e semiochemicals ha dis u b he bi o e se lemen and p oli e a ion and/o ec ui he bi o es’ na u al enemies [23–25]. Fo example, s udies on plan -aphid-pa asi oid in e ac ions ha e e ealed ha 6-me hyl-5- hep en-2-one, (Z)-3-hexen-1-ol, (Z)-3-hexen-1-ol ace a e, (E)-b- ca yophyllene and (E)-b-ocimene [49–52] can a ac aphid pa asi oids and linalool can di ec ly epel aphids [50,53], while me hyl salicyla e seemingly ac s in bo h ways [52,54,55]. In p e ious and cu en s udies, g ass hos s ha e been ound o change emission pa e ns o some o hese compounds in esponse o Epichloe¨ in ec ion ei he alone o in conjunc ion wi h bio ic s ess. The e o e, i is likely ha he al e ed ola ile p o iles may modi y VOC-media ed mul i ophic in e ac ions. This hypo hesis needs o be es ed in u u e s udies o disclose he ecological consequences o Epichloe¨-media ed change in hos VOC emissions o he bi o es, na u al enemies, and hus plan i ness. While ou esul s show ha Epichloe¨-media ed hos VOC esponses depend on endophy e and hos species, he ecological and e olu iona y p ocesses ha lead o such a ia ion emain unclea and me i u u e s udy. As wi h he exp ession o cons i u i e and inducible plan de ences, ha bou ing endophy es is cos ly because endophy es mus p ocu e all o hei nu ien s om he hos , including p ecu so s in he syn hesis o seconda y me aboli es such as alkaloids (e.g. [1,2]). Thus, e olu iona y ade- o s may occu in Epichloe¨–g ass symbioses. In o he wo ds, Epichloe¨-g ass symbioses which ha e de eloped high le els o endophy e-de i ed esis ance unde na u al o a i icial selec ion may ha e e ol ed low le els o hos de ence, and ice e sa. Ou Figu e 4. PLS-DA plo s based on compa isons among VOC blends emi ed by di e en ly ea ed plan s a 12 days a e aphid addi ion. E-: na u ally endophy e ee; E+: na u ally endophy e in ec ed; ME-: manipula i ely endophy e ee; C: con ol; A: aphid eeding. Fo all escue (uppe panel), a clea sepa a ion was seen be ween con ol and in es ed plan s, which clus e ed mainly on he igh and le side o he plo , espec i ely, whe eas plan s o di e en endophy e s a us wi hin each damage ea men la gely o e lapped. Fo meadow escue (lowe panel), E- and E+plan s g ouped sepa a ely while o e lapping somewha , wi h he s onges sepa a ion occu ing be ween in es ed E- plan s and any o he ea men . S a is ical de ails conce ning compounds esponsible o he clus e ing a e gi en in ables S1 and S5. doi:10.1371/jou nal.pone.0101331.g004 Endophy es Modi y Hos Vola ile Release PLOS ONE | www.plosone.o g 8 June 2014 | Volume 9 | Issue 6 | e101331 obse a ion o Epichloe¨ species ela ed di e ences in hos VOC esponses pa ially suppo his idea. Speci ically, he meadow escue–E. uncina a symbiosis which has high endophy e-con e ed cons i u i e de ence [13,14,17,56] eleased low amoun s o VOCs by he hos g ass ega dless o he bi o e p esence. By compa ison, in he all escue–E. coenophiala symbiosis whe e E. coenophiala p o ides ela i ely low cons i u i e de ence [13,14,17,56], he hos g ass exhibi ed induced VOC esponses in he p esence o high aphid densi y. In conclusion, ou s udy has shown ha Epichloe¨ endophy es may modula e VOC esponses o hos g asses, wi h he sign and s eng h o such esponses depending on he iden i y o he Epichloe¨–g ass symbiosis. Ou esul s illus a es he impo ance o assessing hos plan ola iles and hei impac s on he bi o e hos - sea ching beha iou o in es iga e al e na i e mechanis ic links be ween Epichloe¨ endophy es and he bi o e esponses. Gi en ha bo h Epichloe¨ endophy es and he bi o es can manipula e hei sha ed hos s in di e se ways and ha endophy e-p o ided esis ance o he bi o es a ies conside ably among Epichloe¨–g ass associa ions, Epichloe¨–media ed hos VOC esponses and hei impac s on mul i ophic in e ac ions should be a iable. Fu he s udies wi h di e en Epichloe¨–g ass associa ions would shed mo e ligh on endophy e-p o ided de ence and i s in e ac ion wi h he hos ’ own de ence. Suppo ing In o ma ion Figu e S1 Kine ics o VOC emissions om ed escue (Fes uca ub a) ollowing mechanical wounding. (a) To al ion cu en (TIC) ch oma og ams o VOCs om a ep esen a i e plan sample o ed escue be o e and a e mechanical wounding. (b) Emissions (peak a ea 6SE; n = 4) o he dominan VOCs om ed escue. 1 = (Z)-3-hexenal, 2 = (E)-2-hexenal, 3 = (Z)-3-hexen- 1-ol, 4 = (Z)-3-hexen-1-ol ace a e, 5 = (Z)-b-ocimene, 6 = (E)-b- ocimene. (TIF) Figu e S2 PLS-DA plo s o VOC blends emi ed by di e en ly ea ed plan s a 6 days a e aphid addi ion. E-: na u ally endophy e ee; E+: na u ally endophy e in ec ed; ME-: manipula i ely endophy e ee; C: con ol; A: aphid eeding; W: mechanical wounding. Fo all escue (uppe panel), a clea sepa a ion was seen be ween con ol and in es ed plan s, whe eas o meadow escue (lowe panel) he sepa a ion was mainly ound be ween E- and E+plan s. S a is ical de ails conce ning com- pounds esponsible o he clus e ing a e gi en in able S2 and S6. (TIF) Figu e S3 PLS-DA plo s o VOC blends emi ed by di e en ly ea ed plan s a 1 day a e mechanical wounding. In all escue (uppe panel) he s onges sepa a ion was obse ed be ween ME- plan s and ei he o he E- and E+ plan s, wi h he la e wo la gely o e lapping. In meadow escue (lowe panel) E- and E+plan s, while o e lapping somewha , emained la gely sepa a ed om each o he (E-: na u ally endophy e ee; E+: na u ally endophy e in ec ed; ME-: manip- ula i ely endophy e ee; C: un ea ed con ol; W: mechanical wounding). S a is ical de ails conce ning compounds esponsible o he clus e ing a e gi en in ables S4 and S8. (TIF) Figu e S4 E ec s o endophy e on aphid popula ion g ow h, which was es ima ed by o al aphid d y mass pe plan . Di e en le e s o e he ba s indica e signi ican di e ence acco ding o one-way ANOVA. ME-: manipula i ely endophy e ee; E-: na u ally endophy e ee; E+: na u ally endophy e in ec ed. (TIF) Figu e S5 E ec s o endophy e by aphid on plan g ow h as es ima ed by abo eg ound d y weigh . Di e en le e s o e he ba s indica e signi ican di e ence acco ding o wo-way ANOVA. NS: no signi ican . ME-: manipula i ely endophy e ee; E-: na u ally endophy e ee; E+: na u ally endophy e in ec ed. C: con ol; W: mechanical wounding; A: aphid in es a ion. (TIF) Table S1 VOC emissions (ng gDW 21 h 21 ) om all escue a 12 days pos eeding. E-: na u ally endophy e ee; ME-: manipula i ely endophy e ee; E+: na u ally endophy e in ec ed. (DOCX) Table S2 VOC emissions (ng gDW-1 h-1) om all escue a 6 days pos eeding. E-: na u ally endophy e ee; ME-: manipula i ely endophy e ee; E+: na u ally endophy e in ec ed. (DOCX) Table S3 LMM esul s showing he endophy e e ec s in all escue a e con olling o aphid damage. In he model, endophy e, sampling ime and hei in e ac ion we e ixed ac o s, and aphid d y weigh (an indica o o he bi o y) was included as a co a ia e. (DOCX) Table S4 VOC emissions (ng gDW-1 h-1) om all escue a 1 day a e mechanical wounding. E-: na u ally endophy e ee; ME-: manipula i ely endophy e ee; E+: na u ally endophy e in ec ed. (DOCX) Table S5 VOC emissions (ng gDW-1 h-1) om meadow escue a 12 days pos eeding. E-: na u ally endophy e ee; E+: na u ally endophy e in ec ed. (DOCX) Table S6 VOC emissions (ng gDW-1 h-1) om meadow escue a 6 days pos damage. E-: na u ally endophy e ee; E+: na u ally endophy e in ec ed. (DOCX) Table S7 LMM esul s showing he endophy e e ec s in meadow escue a e con olling o aphid damage. In he model, endophy e, sampling ime and hei in e ac ion we e ixed ac o s, and aphid d y weigh (an indica o o he bi o y) was included as a co a ia e. (DOCX) Table S8 VOC emissions (ng gDW-1 h-1) om meadow escue a 1 day a e mechanical wounding. E-: na u ally endophy e ee; E+: na u ally endophy e in ec ed. (DOCX) Acknowledgemen s Many hanks o S an Fae h and an anonymous e iewe o hei cons uc i e commen s on he manusc ip , s a a he Uni e si y o Eas e n Finland Kuopio Campus esea ch ga den (Elina Ha¨ikio¨, Ma ja a Puu unen and Toi o Ku onen) o g owing plan s, Timo Oksanen o echnical suppo , Robe Glinwood o p o iding Rhopalosiphum padi insec s, and Aleksand a S ach (Adam Mickiewicz Uni e si y, Poland) and Sand o A. Rhoden (Uni e sidade Es adual de Ma inga´, B azil) o checking endophy e s a us. Endophy es Modi y Hos Vola ile Release PLOS ONE | www.plosone.o g 9 June 2014 | Volume 9 | Issue 6 | e101331