Lolitrem B and Indole Diterpene Alkaloids Produced by Endophytic Fungi of the Genus Epichloë PDF
Preview Lolitrem B and Indole Diterpene Alkaloids Produced by Endophytic Fungi of the Genus Epichloë
toxins Review Lolitrem B and Indole Diterpene Alkaloids Produced by Endophytic Fungi of the Genus Epichloë and Their Toxic Effects in Livestock GuerrePhilippe UniversitédeToulouse,INP,ENVT,URMycotoxicologie,F-31076Toulouse,France;[email protected]; Tel.:+33-056-119-3840 AcademicEditor:KevinWelch Received:15January2016;Accepted:3February2016;Published:15February2016 Abstract: Differentgroupofalkaloidsareproducedduringthesymbioticdevelopmentoffungal endophytes of the genus Epichloë in grass. The structure and toxicity of the compounds vary considerablyinmammalianherbivoresandincroppests. Alkaloidsoftheindole-diterpenegroup,of whichlolitremBisthemosttoxic,werefirstcharacterizedinendophyte-infectedperennialryegrass, andareresponsiblefor“ryegrassstaggers.”Ergotalkaloids,ofwhichergovalineisthemostabundant ergopeptidealkaloidproduced,arealsofoundinryegrass,butgenerallyatalowerratethanlolitrem B.Otheralkaloidssuchaslolinesandperaminearetoxicforcroppestsbuthaveweaktoxicological propertiesinmammals. Thepurposeofthisreviewistopresentindole-diterpenealkaloidsproduced inendophyteinfectedryegrassfromthefirstcharacterizationofryegrassstaggerstothedetermination ofthetoxicokineticsoflolitremBandoftheirmechanismofactioninmammals, focusingonthe differentfactorsthatcouldexplaintheworldwidedistributionofthedisease. Otherindolediterpene alkaloids than lolitrem B that can be found in Epichloë infected ryegrass, and their tremorgenic properties,arepresentedinthelastsectionofthisreview. Keywords:alkaloids;endophyticfungi;Epichloë;lolitremB;mycotoxins;livestock;toxicology;staggers 1. Introduction ThesymbioticdevelopmentoffungalendophytesofthegenusEpichloë ingrassresultsinthe productionofdifferentgroupsofalkaloids(Figure1)whoseprofileofproductioninplantsexplains signs of toxicity. Lolitrem B was recognized as the main indole-diterpene alkaloid produced in Loliumperenne(perennialryegrass)infectedbyE.festucaevar. lolii(Neotyphodiumlolii),responsible for“ryegrassstaggers”inlivestock[1,2]. Otherindole-diterpenealkaloidshavebeencharacterized inendophyteinfectedryegrass,butdifferfromlolitremBintheirtremorgenicproperties[3]. Ergot alkaloids have been found in L. perenne, in L. arundinaceum (tall fescue), and in other grasses [4,5]. Ergovaline has been recognized as the most abundant and the most toxic ergopeptide alkaloid producedinE.coenophiala-infectedtallfescue,andisresponsibleforfescuetoxicosisinlivestock[5]. Bycontrast,inE.festucaevar. lolii-infectedperennialryegrass,thesignsofergotalkaloidtoxicityare oftenmaskedbytetanicspamsandstaggers,whicharelinkedtotheingestionoflolitremB[6]. Other alkaloidsproducedinendophyte-infectedperennialryegrassincludelolines,ofwhichN-formylloline isthemostabundant,andperamine,whichiswelltoleratedbylivestockbuttoxicforcroppests[6,7]. Becausethealkaloidsproducedbythefungalendophyteareresponsibleforseriousdiseasesand economiclossesinlivestock,asimplesolutiontoavoidtoxicitycouldbetoeliminatetheendophyte fromthegrass. However, alkaloidsarealsotoxicforinsectsandnematodesandthecultivarsthat arefreeofendophytesaremoresensitivetocroppeststhanthecorrespondingendophyte-infected cultivars. So,mostrecentresearchanddevelopmentonendophyte-infectedgrasseshasfocusedonthe Toxins2016,8,47;doi:10.3390/toxins8020047 www.mdpi.com/journal/toxins Toxins2016,8,47 2of16 Toxins 2016, 8, 47 2 of 15 sgerlaescsteios nhaosf Efopciuchsleodë sotnra tihnes tshelaetcatiroenu onfa bElpeictholopër ostdruaicnest htheaat lakrael ouidnasbthlea ttoa rperohdiguhcley tthoex iaclkinalloividess ttohcakt, ia.ree., heirggholvya tloinxeica innd livloelsittorecmk, iB.e,.,b eurtgostvialllianbel eantod plorloitdreumce Bt,h beuatl sktailllo aidbslet thoa tpraordeutocex itchefo arlkinasloeicdtss athnadt nareem taotxoicd efosr[ 6in].seOctns atnhde ontehmerathoadneds ,[6n]o. nO-nse tlheec toetdheern hdaonpdh,y nteosn‐(saelsleoctkendo ewnndoapsh“ywteisld (”alseon dkonpohwynt eass) a“rweilsdti”ll epnrdeosepnhtyitnesg) raarses setsililn psreevseenrat linc oguranstrsieess i,na nsedveqruael sctoiounnstrpieesr,s aisntdc oqnuceesrtnioinnsg ptherespisrto cdounccteironninogf athlkea ploroiddsuicntipolna notfs aalknadlothidesir into pxilcaintyts iannldiv tehsetoirc kto.xTihcietyp iunr pliovseestoofctkh. iTshreev pieuwrpiossteo opfr tehsiesn rtetvhieewin diso tloe dpirteesrepnetn tehael kinadlooildes dthitaetrpareenpe raoldkualcoeiddsb ytheantd aorpe hpyrtoicdfuucnegdi boyf tehnedgoepnhuystiEcp ficuhnlogëi aonf dthteh egierntouxsi cEpeficfehclotsë ianndliv thesetior ctko,xfioc ceuffseinctgs oinn lliovleistrteomck,B foacnudsionngt ohne lfoalcittorersmr eBs panodns oibnl ethfeo rfavcatorirast rioenspsoinnsiitbsllee vfoerl vinargiaratisosness ainn ditsi nleivtselt oinx igcirtayssinesl iavnedst ionc kit.s toxicity in livestock. FFiigguurree 11.. MMaaiinn aallkkaallooiiddss pprroodduucceedd bbyy EEppiicchhllooëë iinn ppeerreennnniiaall rryyeeggrraassss.. 22.. RRyyeeggrraassss SSttaaggggeerrss, ,ffrroomm IIttss FFiirrsstt CChhaarraacctteerriizzaattiioonn ttoo tthhee DDiissccoovveerryy ooff LLoolliittrreemm BB TThhee fifirrsstt ddeessccrriippttiioonn ooff mmuussccuullaarr iinnccoooorrddiinnaattiioonn iinn ccaattttllee aanndd hhoorrsseess ggrraazziinngg oonn rryyeeggrraassss wwaass mmaaddee iinn NNeeww ZZeeaallaanndd iinn 11990066.. BBeeccaauussee tthhee ddiisseeaassee wwaass oobbsseerrvveedd wwhheenn sseeeeddss wweerree ffoorrmmeedd,, tthhee sscclleerroottiiaa ooff CClalavviciceeppss ppuurprpuurerae awwereer ecocnosnidsiedreerde tdo tboe bceaucsaautisvaeti avgeeangtse nutnstiul nthtiel 1th95e01s9. D50usr.inDgu trhien gsatmhee spaemrioedp,e ar ifoudn,gaufsu wngaus sfowuansdf toou inndfetcot iLn. fpeecrteLn.npee sreeendnse isne eddifsfeinrednitf fceoruennttrcioesu n[8t]r.i eTsh[e8 ]f.uTnhguesf uwnagsu csawlleads c“aelnleddop“heyntdeo”p bheyctaeu”sbe etchaeu smeytcheelimumyc edleiuvemlodpeevde ilnopsiedde itnhsei dpelatnhte cpellalsn tacnedl lsinavnaddeindv tahdee wdhthoelew phlaonlet pexlacneptte txhcee protothtse. Irto wotass. sIetewd atrsasnesemditttreadn,s bmuitt taepdp,abruenttalyp pnaorte tnratlnysmnoitttetrda nbsym thitet epdolbleynt h[9e].p Inol 1le9n35[,9 a]. review on “staggers” in livestock pointed to the different etiology of the syndrome: a metabolic origin of the signs in cows after parturition, a sclerotia origin in some cases of Paspalum staggers, whereas there was some doubt about the sclerotia origin in at least some cases of ryegrass staggers, especially Toxins2016,8,47 3of16 In 1935, a review on “staggers” in livestock pointed to the different etiology of the syndrome: a metabolicoriginofthesignsincowsafterparturition,asclerotiaorigininsomecasesofPaspalum staggers,whereastherewassomedoubtaboutthesclerotiaorigininatleastsomecasesofryegrass staggers,especiallybecausethediseaseoccurredwhenthegrasswasparticularlyshort,afteraperiod ofdryness,intheabsenceofseeding[10]. However,feedingendophyte-infectedseedstomice,rat, chickens, and sheep failed to reveal deleterious effects, and the toxicity of the fungal endophyte remained unheeded for several years [11]. Although the etiologic agent of ryegrass staggers was unknown, it was clear that the disease occurred in sheep grazing extremely short grass [12], but experimentalreproductionofthediseasebyfeedingsheepwithergotsofC.purpureafailedtoproduce symptomsofryegrassstaggers[13]. Bycontrast,staggerswereexperimentallyproducedinsheep grazingthebaseoftheryegrassplant,whereasnosignswereobservedinsheepthatwereprevented fromgrazingthispartoftheplant[14]. PenitremA,verruculogen,fumitremorginB,andpaxilline, which are tremorgenic mycotoxins of different fungal origin, were suspected to be the causative agent of ryegrass staggers [15]. It was hypothesized that mycotoxins are produced in the soil by saprophyticfungi,thentranslocatedtotheplantviatheryegrassroots[16]. In1981,afeedingtrial conductedtocomparetherateandseverityofstaggersinsheepwiththerateofendophyteinfectionof ryegrassdefinitivelyconfirmedtheimplicationoftheendophyte[1]. LolitremAandBwerethemain mycotoxinsisolatedinendophyte-infectedperennialryegrass[2],whereasfeedingsheepwithseeds thatcontainedlolitremsmadeitpossibletoreproduceryegrassstaggers[17]. Consequently,lolitrem BproducedbyE.festucaevar. loliiduringitssymbioticdevelopmentinL.perennewasconsideredto beresponsibleforryegrassstaggers. Thetoxicthresholdwasestablishedasbeingbetween1800and 2000µglolitremB/kgfeedincattleandsheep[18]. However,althoughotheralkaloidsthanlolitremB canbeproducedbyEpichloëinendophyte-infectedryegrass,littleisknownabouttheireffectsduring concomitantexposuretololitremB.Thiswasthecaseforergotalkaloidsandergovaline,forwhich signsoftoxicityarerarelyreportedinlivestockfedendophyte-infectedryegrass,whereastheyare themainalkaloidsinvolvedinthetoxicityofendophyte-infectedtallfescue[6]. Thisdifferencewas linkedtotheprofileofalkaloidproductionbyEpichloëintheinfectedplant[5]. Theconcentrationof lolitremBisusually5to10foldhigherthanergovalineinendophyte-infectedryegrass[19–23],so moststudiesfocusedonlolitremB,andlittleinformationwasavailableonergovalineintheseplants. InNewZealandinparticular,mostofthetoxicityofergotalkaloidsinendophyte-infectedperennial ryegrassremainedunknownuntilthedevelopmentofanovelendophyte(knownas“endosafe”)that wasunabletoproducelolitremB[6]. Interestingly,variousstudiesconductedinlambsandinlactating ewesalsosuggestedthatthetoxicthresholdofergovalinewaslowerinendophyte-infectedryegrass thaninendophyte-infectedtallfescue[24–26]. Bycontrast,asynergisticeffectbetweenergotamine andlolitremBwasobservedinsmoothmusclecontractiletensioninlongitudinalpreparationofthe distalcoloninsheep[27]. Itwassuggestedthatthiseffectcouldcontributetothemoreprevalentrate ofnoninfectiousdiarrhea,observedinsheepgrazingendophyte-infectedpastures[27]. Interaction betweenlolitremBandergovalinecouldalsocontributetodecreasesinmilkproductionobservedin dairycowsgrazingonendophyte-infectedryegrass[28–31]. Inthesameway,littleisknownaboutthe productionandtoxicityofindole-diterpenealkaloidsotherthanlolitremB,includingtheintermediate metabolitesofitssynthesisbyEpichloë. 3. LolitremB,aTremorgenicMycotoxin Signsofstaggersarecommonlyobservedinanimalsfedwithmycotoxinssuchaspaspalitrem, paxilline, penitrem, aflatrem, verruculogen, fumitremorgin, janthitrem, and lolitrem (Figure 2). Paspalitrems are produced by different species of the genus Claviceps that parasitize the seeds of Paspalumgrasses,Bermudagrass,andothergrasses[32]. Paxilline,penitrem,aflatrem,verruculogen, fumitremorgin,andjanthithremareproducedbyfungiofthegeneraAspergillusorPenicillium,mostof whicharesaprophyticincerealsandplantsduringstorage,andsomearealsophytopathogenic[33]. Paxilline, epoxy-janthitrem, and lolitrem are produced by endophytic fungi of the genus Epichloë Toxins2016,8,47 4of16 duringtheirsymbioticdevelopmentinsomegrasses[34]. Althoughthesecompoundsareofdifferent fungalorigin,severalstudieswereconductedtocomparetheirmechanismofactioninthecourseof elucidationoflolitremBmechanismofaction. Boththetoxicokineticandpharmacologicalproperties oftremorgenicmycotoxinscanvaryintermsofthelocation(peripheralvs. central),nature(excitatory vTso.xiinnsh 2ib01it6o, r8y, )4,7a ndseverity(amplitude,duration)oftheeffects. 4 of 15 Figure2.Structuresofsometremorgenicmycotoxins. Figure 2. Structures of some tremorgenic mycotoxins. 33..11.. FFrroomm SSttaaggggeerrss ttoo BBKK-‐CChhaannnneellss TThhee tteerrmm ““ssttaaggggeerrss”” iiss uusseedd ttoo ddeessccrriibbee aa vvaarriieettyy ooff ddiisseeaasseess iinn lliivveessttoocckk wwhhoossee eettiioollooggyy ccaann ddiiffffeerr ccoonnssiiddeerraabbllyy,, rryyeeggrraassss iinnggeessttiioonn bbeeiinngg oonnllyy oonnee ppoossssiibbllee ddiisseeaassee eettiioollooggyy [[1100]].. SSttaaggggeerrss tthheemmsseellvveess vvaarryy iinn bbootthh mmoorrbbiiddiittyy aanndd sseevveerriittyy dduurriinngg tthhee ccoouurrssee ooff rryyeeggrraassss ttooxxiicciittyy.. FFrroomm 55%% ttoo 7755%% ooff tthhee aanniimmaallss iinn tthhee ssaammee hheerrdd aarree aaffffeecctteedd,, aanndd ssiiggnnss vvaarryy ffrroomm sslliigghhtt ttrreemmbblliinngg ooff tthhee nneecckk aafftteerr hhaarrdd eexxeerrcciissee ttoo sseevveerree tteettaanniicc ssppaassmm aanndd ccoolllalappssee, ,aa sysystsetmem ofo sfcsocroirnign gbebienign gusuesde dto taosasesssse stsheth seevseervietyri toyf othfet hdeisdeaisseea [s1e2[,1142,,3154],3. 5U]n.tUil nthtiel tdhisecdovisecroyv oefr ylooliftrleomlitsr einm 1s9i8n1,1 s9e8v1e,rsaelv terreamlotrregmenoircg menyiccomtoyxcinost owxienrse wsuesrpeescutesdp etcot ebdet othbee cthauescaatuivsea taivgeenatg eonf troyfergyreagssr asstsasgtgaegrgse r[2s,1[25,]1 5a]nadn dvavraioriuosu sini nvvivivoo aanndd eexx vviivvoo eexxppeerriimmeennttss wweerree ccoonndduucctteedd ttoo uunnddeerrssttaanndd tthheeiirr mmeecchhaanniissmm ooff aaccttiioonn.. SSttuuddiieess uussiinngg sshheeeepp aanndd rarat tsysynnapaptotsoosmomese srerveevaelaelde dthtahta vtevrreurrcuuclougloegne annadn pdenpietnreitmre mA aAct abcyt binyteirnfteerrinfegr iwngithw thiteh rtehleeasree loefa asemoinfoa amciidn oneaucriodtrnaenusmroittrtaenrss.m Tihttee trrse.mTohrse ctoruelmd obres dcuoeu tlod abneomdualeoutos arenloemasael oouf sborethle aesxeciotaftbooryth aenxdc iitnahtoibryitoarnyd tirnahnisbmitiotrteyrstr aatn scmenittrtaelr saantdc penertripalhaenradl psyenriapphseersa, llseyadnainpgs etso, lleoasds ionfg cotoorldoisnsaotifocno oofr dthine anteiounraol fmthecehnaenuisrmalsm theacth caonnistrmosl mthuastccleo natcrtoiolnm [u36s]c.l Eenahctainocne[d3 6u]n.sEtinmhualnacteedd release of the excitatory amino acid neurotransmitters aspartic acid and glutamic acid was measured in cerebrocortical synaptosomes prepared from sheep showing severe symptoms of ryegrass staggers [37]. Studies in rat brain membranes revealed that the four tremorgenic mycotoxins tested (aflatrem, paspalinine, paxilline, and verruculogen) inhibited GABA‐induced 38CI− influx into rat brain microsacs, while a non‐tremorgenic mycotoxin, verruculotoxin, had no effect [38]. It was suggested that the relatively nonpolar properties of these mycotoxins enable them to pass the blood‐brain barrier and gain rapid access to many of the synapses present in the brain, where they have a central effect [36]. Toxins2016,8,47 5of16 unstimulatedreleaseoftheexcitatoryaminoacidneurotransmittersasparticacidandglutamicacid wasmeasuredincerebrocorticalsynaptosomespreparedfromsheepshowingseveresymptomsof ryegrassstaggers[37]. Studiesinratbrainmembranesrevealedthatthefourtremorgenicmycotoxins tested (aflatrem, paspalinine, paxilline, and verruculogen) inhibited GABA-induced 38CI´ influx into rat brain microsacs, while a non-tremorgenic mycotoxin, verruculotoxin, had no effect [38]. Itwassuggestedthattherelativelynonpolarpropertiesofthesemycotoxinsenablethemtopassthe blood-brainbarrierandgainrapidaccesstomanyofthesynapsespresentinthebrain,wherethey haveacentraleffect[36]. Abioassaywasdevelopedinmicetoassessthetremorgenicpropertiesofmycotoxins[39]. After intraperitonealadministration,lolitremBwasresponsiblefortremorswhoseonsetwasslowerbut lastedlongerthanthosecausedbyaflatrem[40]. LolitremBalsohadamuchlongerandmorepotent tremorgeniceffectthanpaxilline[41]. Electromyographicactivityofskeletalmusclerecordedinsheep receivingdifferentlevelsofpenitrem,paxilline,andlolitremBrevealedthattheexcitatoryproperties ofallthecompoundsagreewiththepreviouslyobservedeffect[42]. Theeffectsoftremorgenicmycotoxinswerealsoinvestigatedonsmoothmuscles,butweremore difficulttointerpret. Verruculogen,penitremB,andpaxillineenhancedtheelectrically-stimulated contractions of guinea pig ileum but did not influence the contractions caused by exogenous acetylcholine,suggestingthatthesecompoundsenhancethereleaseofacetylcholine[43]. Penitrem, paxilline, and lolitrem B induced variable responses of the smooth muscle of the reticulorumen, abomasum, and duodenum in sheep [44]. An inhibitory effect on the abomasum was observed whereasbothexcitatoryandinhibitoryeffectswereobservedontheduodenum[42]. Bycontrast,an excitatoryeffect,whichwaspartiallyblockedbyatropine,wasobservedonthereticulorumen[45]. Thedifferentresponsesofthegastro-intestinaltracttololitremBinsheepcouldbepartlyexplainedby interferenceinthereleaseofacetylcholinebytheparasympatheticnerve,whichisrequiredforthe cyclicalcontractionsofthereticulorumen,butdoesnotoccurintheabomasumandduodenum[46]. Sarcolemmal membrane vesicles of bovine aortic smooth muscle were used to compare the properties of lolitrem B and paspalicine, a non-tremorgenic analog of paspalinine that is dehydroxylated[47].Becausebothcompoundsblockedlargeconductancecalcium-activatedpotassium channels(BKchannels),itwasconcludedthatalthoughsomeofthepharmacologicalpropertiesof lolitremBcanbeexplainedbyinhibitionofBKchannels,tremorgenicitymaynotberelatedtothis mechanismofaction[47]. TwentyyearsafterthefirstdiscoveryoflolitremB,toxicthresholdshave been established for the occurrence of staggers [18], but its pharmacological mechanism of action remainsunclearasbothperipheralandcentraleffectsaresuspected. Knowledge of the mechanism of action of lolitrem B at the pharmacological level advanced considerablywhenpaxillineandlolitremBtoxicitywerecomparedinwildmiceandinknock-out micedeficientinBKchannels(Kcnma1). BKchannelsareexpressedincellmembranesofallthetissues wheretheyenableoutflowofK+,whichisresponsibleforhyperpolarizationofthecellsandareduction incellularactivity[48]. ThefactthattremorsoccurredinwildmicebutnotinthemicelackingKcnma1 demonstratedtheimportantroleofBKchannelsinryegrassstaggers. Theobservationthatknown lethaldosesoflolitremBinthewild-typemicehadnoeffectinknock-outmicealsosuggestedthat inhibition of BK channels by lolitrem B is probably the only mechanism of action involved in the acute toxicity of lolitrem B [49]. Relationship structure activities were performed on BK channels, makingitpossibletocomparedifferentstructurally-relatedlolitrems[50]. Thisconfirmedthatonlya smallchangeinstructuremodifiesthebindingpropertiesofthetoxins,aspreviouslyobservedfor tremorgenicityinthemousebioassay[51]. ComparisonofpaxillineandlolitremBalsorevealedthat lolitremBismorepotentthanpaxillineatinhibitingBKchannelsinvitroandthatinhibitioncannotbe reversedbywashing[52]. Thisresultisinagreementwithpreviousobservationsininvivostudiesin mouse,inwhichlolitremBhadamuchlongerandmorepotenteffectonmotorfunctionthanpaxilline, whereaspaxillinecausedamorerapidonsetoftremor,buttremorslastedforashorterperiodthan thosecausedbylolitremB[41]. Toxins2016,8,47 6of16 3.2. Toxicokinetics BothcentralandperipheraleffectsoflolitremBcouldcontributetotoxicity;however,because lolitremBisinsolubleinaqueousmedia,itisdifficulttoobtain,andrelativelylittleisknownaboutits toxicokinetics. AllthestudiesperformedafteroraldosingoflolitremBrevealedverylowabsorption ofthetoxin[31,53]. NostudyhasreportedlolitremBconcentrationsinthebrain, but14C-labeled paxillineinjectedintraperitoneallyinthemouseshowedthatthetoxinwaspresentinthebrainbut atextremelylowconcentrations[41]. AfterhighdosesoflolitremBintravenouslyadministeredin sheep(75µg/kgBW),theconcentrationoflolitremBinserumdecreasedrapidly,whereastremors continuedfor16hourspostdosing[41]. RapideliminationoflolitremBfromserumwasconfirmedin intravenouslydosed(23µg/kgBW)lactatinggoats,andahalf-lifeof14minwascalculated[53]. Both therapideliminationofthetoxinfromserumandthetimeatwhichtremorsoccurredafterIVdosing suggestthatlolitremBmaybestoredinaspecificcompartmentofthebody,thenprogressivelyreleased into the blood at very low levels for transport to the brain. This hypothesis was strengthened by analysisoflolitremBlevelsinbodyfluidsandtissues. Ingoats,lolitremBwaspresentinmilkfor32h aftertheintravenousinjection(onedoseof23µg/kgBW)withanexcretionrateof3%. Thisresultwas confirmedafteroraldosing(onedoseof100µg/kgBW)inlactatinggoats: thedurationofelimination in milk was 75 h and the excretion rate of the administered dose was 0.19% [53]. A similar result wasobservedindairycowsafterprolongedexposuretothetoxin,whenonly0.23%ofthelolitremB consumedwasexcretedintothemilk[31]. BecauselolitremBisverysolubleinorganicsolventsandinlipidmedia,itwashypothesized thatitcouldbestoredinfatstorage[41]. HighlevelsoflolitremBinfatwereconfirmedinallthe experimentsinwhichlolitremBwasmeasured,inbothcattleandsheep,andinbothgrowingand lactatinganimals[25,31,54–56]. Whensheepgrazedforaprolongedperiodinpasturesthatcontained lolitremB,thetoxinconcentrationinthefatrapidlyincreasedwithanincreaseintheconcentrationin thepasture,anddecreasedwithadecreaseintheconcentrationsinthepasture[55]. Thislastresult suggestedthat,ratherthanaccumulating,thequantitiesinthefatrespondrapidlytotheamounts beingconsumedbytheanimals[55]. BoththetimeatwhichtremorsareobservedafterIVdosingandthelipophilicnatureoflolitremB suggestthatthetoxinmaybemetabolized. However,verylittleisknownaboutthebiotransformation oflolitremBortheinteractionsbetweenthetoxinanddrug-metabolizingenzymes. Astudyconducted in lactating ewes fed with endophyte-infected perennial ryegrass hay containing lolitrem B and ergovalinerevealedslighteffectsontheactivitiesofsomedrug-metabolizingenzymes[25]. These resultsweredifficulttointerpretbecausefeedingendophyte-infectedtallfescuehaythatcontained ergovalinealonechangedsomedrugmetabolizingenzymesactivitiesinthisspeciesatlowerlevelsof exposure[24]. However,comparisonoftheresultsobtainedinthetwostudiessuggeststhatinteraction between lolitrem B and ergovaline may affect the activities of drug-metabolizing enzymes [24,25]. Alternatively,theimportanceofcytochromeP450inthebiosynthesisoflolitremB,andtheroleofthe transformationscatalyzedbytheseenzymesonthebiologicalactivityofthetoxinsarenotable[57,58]. 4. LolitremBinPlantsandRyegrassStaggers 4.1. WorldwideDistributionoftheDisease WhereasinfectionofL.perennebyE.festucaevar. loliiisreportedworldwide,mostoutbreaksof ryegrassstaggershavebeenreportedinAustraliaandNewZealand[59]. Bycontrast,onlysporadic atypical cases are observed in Europe and America. Most cases reported in Europe involved one animaloralimitednumberofanimalsafterconsumptionofhay. Thediseasehasbeendiagnosed in horses, bulls cattle, dairy cows, and sheep following observation of the typical clinical signs of staggers[23,60–63]. ThediagnosiswasconfirmedbydemonstratinghighlolitremBconcentrationsin thesuspecthay. AlongthenortherncoastofCalifornia,somecasesofstaggershavebeendiagnosed insheepandcattlewithahistoryofingestionofperennialryegrass[64]. Staggerswasalsoobserved Toxins2016,8,47 7of16 inJapanincattleandhorsesfedwithryegrassstrawimportedfromOregon[65]. InSouthAmerica, atypicaloutbreakofryegrassstaggerswasobservedinArgentinainabout50%oftheheifersofa herdof560animalsaftergrazing26daysinapaddockofpureL.perenne. Microscopicexaminationof plantsconfirmedthepresenceofanendophyticfungus[66]. Several studies conducted in Australia and New Zealand have demonstrated that the rate of infestation of ryegrass generally correlates with the frequency and severity of staggers [59,67–69]. A high rate of endophyte infection in perennial ryegrass was selected by plant breeders because it enhances growth and persistence of ryegrass in the pasture. For example, in New Zealand, improvedpersistenceofendophyte-infectedryegrasshasbeenassociatedwithimprovedresistance to the Argentine stem weevil (Listronotis bonariensis) [70]. Resistance to other pests has been demonstrated [71–73]. In Europe, the rate of Epichloë infestation of perennial ryegrass varies considerably,dependingonthecountryandthegeographicallocationoftheplantsconcerned[74–79]. ArelativelylowerrateofinfestationinEuropecomparedwithAustralasiacouldpartlyexplainthe lowerprevalenceofstaggers,buttherateofinfestationalonewasprobablynotsufficienttoexplainall thedifferences. ThegeneticsoftheendophytesandotherfactorsthatchangetheleveloflolitremBin plantmaycontributetodifferencesintheprevalenceofryegrassstaggersworldwide. 4.2. LolitremBBiosynthesis SincelolitremBwasrecognizedasthemaintoxinresponsibleforstaggersinendophyte-infected ryegrass, several studies have been conducted to determine the genetic factors responsible for its synthesis. Althoughalargenumberofendophytestrainshavebeenidentifiedinplants,perennial ryegrasshasbeenfoundtobethehostofalimitednumberofdistinctendophytetaxa[80,81]. Genetic analysisofE.festucaevar. loliistrainsdemonstratedthatsomestrainssynthesizeergovalinealone,and otherssynthesizelolitremBalone,butmostproducevariableconcentrationsofbothtoxins[82,83]. Analysis of the genetic factors required for the production of lolitrem B revealed that 10 different genesarepresentinacomplexlocus(ltm)organizedinthreeclustersinterspersedwithtransposon relics[83,84]. Paspalinewasproposedastheintermediatemetaboliteformingthestructuralbackbone requiredfortheproductionofmorecomplexcompoundssuchaslolitremsandterpendoles[58,80]. ThedifferentmetabolitesproducedduringlolitremBsynthesisandtheirpropertieswithrespectto staggersarereviewedinthefollowingparagraph. Althoughitisclearthatthelackofageneinvolved inthesynthesisoflolitremBmakesitpossibletopredictthelackofitssynthesisanditsabsencein plants,thepresenceofltmgenesdoesnotmakeitpossibletopredicttheirlevelofexpressionorthe leveloflolitremBinpastures. Indeed,comparisonsoflolitremBconcentrationsinthewholeplant revealedmajorvariationsdependingonwhichpartoftheplantwasanalyzedandonthelocationofthe study. FieldstudiesinAustraliarevealedconcentrationsoflolitremBrangingfrom 0to4.44mg/kg in perennial ryegrass, and of more than 1.8 mg/kg in 37% of pastures [85]. Other studies of the strawofendophyte-infectedperennialryegrassinOregonalsorevealedawiderangeoflolitremB concentrations,from0tomorethan5mg/kg[22]. Selection of E. festucae var. lolii strains in Australasia also revealed marked variations in the productionofalkaloids[6,86]. WhereaswildstrainsproducedlolitremB,ergovaline,andperamine, the “endosafe,” “AR1,” and AR37” strains did not produce lolitrem B. Unfortunately, because the “endosafe”strainstillproducedergovaline,unexpectedcasesoftoxicitycharacterizedbyintolerance toheatstresswerereported[6,87]. Bycontrast,the“AR1”strainonlyproducedperamine,whichis nottoxicformammalianherbivores,butthisstrainprovidesonlymoderateprotectionagainstcrop pests[55,86,87]. Alkaloidsproducedby“AR37”arediscussedinthefollowingsection. InEurope,the beneficialroleofendophyteinfectioninplantpersistenceorresistancetostressandinvertebrateisless wellunderstood[88–91].However,studiesondifferentecotypeshaverevealedthattheorigin/genetics oftheecotypeinfluencetheconcentrationoflolitremBinplants[92,93]. Interestingly,comparison of lolitrem B concentrations in ecotypes in Germany and in New Zealand also revealed that the environmental conditions of growth (mainly nutrients and climatic factors, also known as abiotic Toxins2016,8,47 8of16 factors)thatoccurredduringplantgrowthhaveastrongerinfluenceontheconcentrationofalkaloids inplantsthantheorigin/geneticsoftheecotype[92]. However,littleisknownabouttheinfluenceof abioticfactorsontheleveloflolitremBinplants. 4.3. VariationsinConcentrationsofLolitremBinPlants Environmental conditions can affect the level of alkaloids in plants; differences depend on plant-endophyte genotype interactions and on the compounds produced [91]. Lolitrem B concentrations in endophyte-infected perennial ryegrass vary considerably depending on the ecotype studied but also on the time of year and the location of the study. In Australia and NewZealand, lolitremBmaximaarereachedinMarch, whichcorrespondstofallinthenorthern hemisphere[85,94,95]. Bycontrast,inFranceandGermany,thehighestconcentrationsoflolitremB areobservedinJunetoAugust,withsomedifferencesdependingontheecotypesanalyzed[92,93,96]. The maxima were seen to range from 2 to 10 mg lolitrem B/kg in the whole plant [22,85,92–96]. Peak concentrationsoftheotheralkaloidsproducedweregenerallyreachedatthesametimeasthe peakinlolitremB,althoughwithsomedifferences. Thepeakergovalineconcentrationhasgenerally beenobservedinspringduringflowering[20,93,94,96–98]. Anotherergovalinepeak,corresponding toplantregrowth,hasalsobeenreportedinthefall,andwassometimeshigherthanthepeakinthe spring,especiallyinAustralasia[20,85,93–98]. ThedistributionoflolitremBintheplantalsovariesdependingonthepartoftheplantanalyzed andthestageofmaturationoftheplant. Astudyconductedovertwoyearsinendophyte-infected perennialryegrassinFrancerevealedaconcentrationoflolitremBinthebase,leaves,andinflorescence ranging from 0.01 to 3 mg/kg dry matter [96]. Over the study period, the highest concentrations oflolitremBwerealwaysobservedatthefullyripestageintheinflorescence,followedbythebase ofplants,whiletheleavescontainedthelowestlevels[96]. Similarresultshavebeenobservedfor ergovaline. Likewise,lolitremBbutnotergovalinewasreportedtoaccumulateinthebaseoftheplant andinsenescenttissues;thedistributionoftheendophytedidnotplayamajorroleinthedistribution ofalkaloidsintheplant[99,100]. Nitrogenfertilization,temperatureanddroughtareknowntoinfluencethelevelofergovalinein endophyte-infectedtallfescue[95,101–106].Howeverlessisknownabouttheinfluenceofthesefactors ontheconcentrationoflolitremBinendophyte-infectedperennialryegrass. Dataobtainedunder controlledconditionsrevealedthathighnitrogeninputreducedthelevelsofendophyteintheplant, which,inturn,reducedalkaloidcontent[107,108]. Astudyconductedinfieldconditionswithnitrogen fertilizationaccordingtotheusualrecommendationsforpasturesuggestedthatnitrogeninputhadno effectontheleveloflolitremBinthewholeplant,whereasitincreasedthelevelofergovaline[96]. In thesameway,littleisknownabouttheinfluenceofdroughtandrainfallonlolitremB.Ananalysis oftheinfluenceofclimaticfactorsonendophyte-infectedperennialryegrassunderfieldconditions suggestedapositivecorrelationbetweenthecumulativerainfallandthelolitremBlevelsinthewhole plant[96]. OutbreaksofperennialryegrasstoxicosisinAustraliawerealsomostsevereduringperiods ofhighrainfallinspringandsummer,buttheconcentrationsoflolitremBinthewholeplantwereno greaterthantheconcentrationsrecordedearlierintheseason[109]. Finally,thestageofmaturationof theplantappearstobethemostimportantnon-geneticfactorresponsibleforvariationsinthelevelof lolitremBintheplant. 5. Indole-DiterpeneAlkaloids,NotOnlyLolitremB Indole-diterpenealkaloidsareformedbyacyclicditerpene-derivedskeletonandanindolemoiety derivedfromtryptophan(Figure3). Differentstrategiesarefoundinfungalsecondarymetabolismto incorporatetheindoleinthefinalalkaloidmetabolites[110]. Byanalogywiththeknownpathways forpaspalinebiosynthesisinP.janthinellumandthepaxillinebiosynthesisinP.paxilli,itissuggested thatthecyclicditerpeneskeletonderivedfromfourisopreneunitsofgeranylgeranyldiphosphate (GGDP),whereastheindolemoietyderivedfromindole-3-gycerophosphate(I3GP)[111]. Paspaline, Toxins2016,8,47 9of16 the simplest indole-diterpene alkaloid, is considered as a key intermediate in the biosynthesis of morecomplexcompounds[58,80]. Paspalineisnottremorgenic. Otherindole-diterpenes, suchas paspalinineandpaspalitrem,alsoproducedbyClavicepspaspaluminPaspalumdilatatuminfectedseeds, areresponsibleforpaspalumstaggers[112,113]. Paspalinecanbeoxidizedinto13-desoxypaxilline then to paxilline by monooxygenases of the cytochrome P450 system [80]. Although paxilline has low tremorgenic properties, few data are available concerning its level in endophyte-infected ryegrass[114–116]. Oxidationofpaspalinecanalsoproducemorethanadozenterpendoles,whichare labeledbyadifferentletter(AtoM)dependingonthenumberandpositionofhydroxylsubstituents on the diterpene moiety of the molecule. Not all the terpendoles characterized were found in endophyte-infectedryegrassandnotalltheterpendolesfoundweretremorgenic. TerpendoleIwas prenylatedandcyclizedtoforman“Iring”andterpendoleC,whichwasfoundinendophyte-infected rTyoexginrsa 2s0s1a6n, 8d, h47a stremorgenicpropertiessimilartothoseofpaxillineinthemousebioassay[1169] .of 15 FFiigguurree 33.. SSttrruuccttuurreess ooff ssoommee iinntteerrmmeeddiiaarriieess ooff ssyynntthheessiiss ooff lloolliittrreemm BB.. Lolitrems can be formed from paspaline and terpendoles and other intermediate metabolites by Lolitrems can be formed from paspaline and terpendoles and other intermediate metabolites adding an “A” and “B” ring to the indole moiety of the molecule at the C20–C21 position [80]. More by adding an “A” and “B” ring to the indole moiety of the molecule at the C20–C21 position [80]. than a dozen lolitrems have been characterized and labeled by a letter (A to N). They differ by the Morethanadozenlolitremshavebeencharacterizedandlabeledbyaletter(AtoN).Theydifferby presence or absence of an I ring and the number and position of hydroxyl and aryl substituents. The thepresenceorabsenceofanIringandthenumberandpositionofhydroxylandarylsubstituents. tremorgenic properties of these compounds vary considerably. Although the relationship structure Thetremorgenicpropertiesofthesecompoundsvaryconsiderably. Althoughtherelationshipstructure activity of the different lolitrems identified is complex, the presence of the I ring appeared to be activity of the different lolitrems identified is complex, the presence of the I ring appeared to be necessary to cause prolonged tremors in the mouse bioassay [116]. For example, lolitrem E has no I ring and low tremorgenic properties [117]. Comparison of four stereoisomers of lolitrem B (including lolitrem F, a natural stereoisomer of lolitrem B) also revealed that the stereochemistry of the A/B ring junction does not influence the tremorgenic properties unless the A ring is oriented toward the alpha‐face, which stops activity [44]. Lolilline, lolitriol, and lolicines are intermediate metabolites of lolitrems that have an A and B ring on the indole moiety of the molecule but no I ring. These compounds can be found in endophyte‐infected ryegrass, but lolilline and lolitriol did not present tremorgenic properties [19,116]. Janthitrems also have an “A” and “B” ring linked to the indole moiety of the molecule, but at the C21–C22 position (Figure 2). These compounds were first isolated from Penicillium janthinellum strains obtained from pastures in which ryegrass staggers have been described [118]. Seven janthitrems were characterized and labelled by a letter (A to G). They differ in the number and position of hydroxyl and acetate substitution on the H ring [119]. In contrast to lolitrems, most janthitrems are not epoxidized at the C11–C12 position, but the epoxidized form of janthitrem G was characterized in perennial ryegrass infected with E. festucae var. lolii strain AR37 [3]. Epoxy‐janthitrems Toxins2016,8,47 10of16 necessarytocauseprolongedtremorsinthemousebioassay[116]. Forexample,lolitremEhasnoI ringandlowtremorgenicproperties[117]. ComparisonoffourstereoisomersoflolitremB(including lolitrem F, a natural stereoisomer of lolitrem B) also revealed that the stereochemistry of the A/B ringjunctiondoesnotinfluencethetremorgenicpropertiesunlesstheAringisorientedtowardthe alpha-face, which stops activity [44]. Lolilline, lolitriol, and lolicines are intermediate metabolites of lolitrems that have an A and B ring on the indole moiety of the molecule but no I ring. These compoundscanbefoundinendophyte-infectedryegrass,butlolillineandlolitrioldidnotpresent tremorgenicproperties[19,116]. Janthitremsalsohavean“A”and“B”ringlinkedtotheindolemoietyofthemolecule,butatthe C21–C22position(Figure2). ThesecompoundswerefirstisolatedfromPenicilliumjanthinellumstrains obtainedfrompasturesinwhichryegrassstaggershavebeendescribed[118]. Sevenjanthitremswere characterizedandlabelledbyaletter(AtoG).Theydifferinthenumberandpositionofhydroxyland acetatesubstitutionontheHring[119]. Incontrasttololitrems,mostjanthitremsarenotepoxidized at the C11–C12 position, but the epoxidized form of janthitrem G was characterized in perennial ryegrass infected with E. festucae var. lolii strain AR37 [3]. Epoxy-janthitrems can pass through the digestive tract of animals and are found as residues in milk and fat [31,55]. Janthitrems and epoxy-janthitremshavebeenshowntohavetremorgenicpropertiesinthemousebioassay,butthey arelesspotentthanlolitremB[55,118]. Interestingly, theprotectionagainstpestcropsobservedin perennial ryegrass infected with AR37 is close to that observed with “wild” endophytes [73]. So, despitehighconcentrationsofepoxy-janthitremsfoundinthefatandmilkofanimalsgrazingthese pastures,becausethesignsofstaggerswereweakandbecausetheprotectiveeffectagainstcroppests werehigh,itwasestimatedthattheratioofrisktobenefitofAR37waspositiveinNewZealand[120]. Inconclusion,thesymbioticdevelopmentofEpichloëinperennialryegrassleadstotheproduction ofdifferentgroupsofalkaloids. Amongthem,lolitremBhasbeenshowntobethemosttoxicalkaloid oftheindolediterpenegroup,responsibleforryegrassstaggers,probablybecauseofitsbindingto the BK channels. A correlation between the rate of infestation and occurrence of the disease has beenobservedinAustralasia,buthighratesofendophyteinfectionofperennialryegrasswithhigh concentrationsoflolitremBinplantshavealsobeenobservedworldwide,withonlyafewatypical casesofstaggersoutsideAustralasia. Severalfactorscouldexplainthesedifferences,suchastheuse ofperennialryegrassasamonocropinAustralasiaandthelivestockraisingpractices. Indeed,the leveloflolitremBingrassvariesconsiderablydependingonthestageofmaturityoftheplant,hence theriskoftoxicityvarieswiththeperiodoftheyearandthepartoftheplantconsumed. Duetothe severityofryegrassstaggersandthehighprevalenceofcroppestsinAustralasia,mostresearchhas focusedontheselectionofEpichloëunabletoproducetoxicalkaloidsformammalianherbivoresbut stillabletoproducealkaloidsthataretoxicforinsectsandnematodes. Thistargetwasdifficultto achieve. AmongtheEpichloëstrainsstudied,thestrainsunabletoproducelolitremBandergovaline butabletoproducehighlevelsofepoxy-janthitremsappeartobeagoodcompromisebetweentherisk ofstaggersinlivestockandtheneedforeffectivecroppestscontrolinAustralasia. Althoughother indolediterpenealkaloidsbesideslolitremsandjanthitremswereproducedbyEpichloëandhavetoxic properties,littleisknownabouttheirlevelinplants. Acknowledgments:Allsourcesoffundingofthestudyshouldbedisclosed.Pleaseclearlyindicategrantsthat youhavereceivedinsupportofyourresearchwork.Clearlystateifyoureceivedfundsforcoveringthecoststo publishinopenaccess. ConflictsofInterest:Theauthorsdeclarenoconflictofinterest. References 1. Fletcher,L.R.;Harvey,I.C.AnAssociationofaLoliumEndophytewithRyegrassStaggers.N.Z.Vet.J.1981, 29,185–186.[CrossRef][PubMed] 2. Gallagher,R.T.;White,E.P.;Mortimer,P.H.RyegrassStaggers:IsolationofPotentNeurotoxinsLolitremA andLolitremBfromStaggers-ProducingPastures.N.Z.Vet.J.1981,29,189–190.[CrossRef][PubMed]
Description:The list of books you might like
