A Blind Circadian Clock in Cavefish Reveals that Opsins Mediate Peripheral Clock Photoreception


previous item | next item

Citation
A Blind Circadian Clock in Cavefish Reveals that Opsins Mediate Peripheral Clock Photoreception

Material Information

Title:
A Blind Circadian Clock in Cavefish Reveals that Opsins Mediate Peripheral Clock Photoreception
Series Title:
PLOS Biology
Creator:
Cavallari, Nicola
Frigato, Elena
Vallone, Daniela
Fröhlich, Nadine
Lopez-Olmeda, Jose Fernando
Foà, Augusto
Berti, Roberto
Sánchez-Vázquez, Francisco Javier
Bertolucci, Cristiano
Foulkes, Nicholas S.
Publication Date:
Language:
English

Subjects

Subjects / Keywords:
Circadian Clock, Photoreception, Cavefish ( local )
Genre:
serial ( sobekcm )

Notes

Abstract:
The circadian clock is synchronized with the day-night cycle primarily by light. Fish represent fascinating models for deciphering the light input pathway to the vertebrate clock since fish cell clocks are regulated by direct light exposure. Here we have performed a comparative, functional analysis of the circadian clock involving the zebrafish that is normally exposed to the day-night cycle and a cavefish species that has evolved in perpetual darkness. Our results reveal that the cavefish retains a food-entrainable clock that oscillates with an infradian period. Importantly, however, this clock is not regulated by light. This comparative study pinpoints the two extra-retinal photoreceptors Melanopsin (Opn4m2) and TMT-opsin as essential upstream elements of the peripheral clock light input pathway.
Original Version:
PLOS Biology, Vol. 9, no. 9 (2011-09-06).

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
This item is licensed with the Creative Commons Attribution License. This license lets others distribute, remix, tweak, and build upon this work, even commercially, as long as they credit the author for the original creation.
Resource Identifier:
K26-05077 ( USFLDC: LOCAL DOI )
k26.5077 ( USFLDC: LOCAL Handle )

USFLDC Membership

Aggregations:
University of South Florida
Karst Information Portal

Postcard Information

Format:
serial

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

ABlindCircadianClockinCavefishRevealsthatOpsins MediatePeripheralClockPhotoreceptionNicolaCavallari1,2,ElenaFrigato1,DanielaVallone2,NadineFro ¨ hlich2,JoseFernandoLopez-Olmeda2,3, AugustoFoa `1,RobertoBerti4,FranciscoJavierSa ´ nchez-Va ´ zquez3,CristianoBertolucci1* ,NicholasS. Foulkes21 DepartmentofBiologyandEvolution,UniversityofFerrara,Ferrara,Italy, 2 InstituteofToxicologyandGenetics,KarlsruheInstituteofTechnology,Eggenstein,Germany, 3 DepartmentofPhysiology,FacultyofBiology,UniversityofMurcia,Murcia,Spain, 4 DepartmentofEvolutionaryBiology‘‘LeoPardi,’’UniversityofFirenze,Firenze,ItalyAbstractThecircadianclockissynchronizedwiththeday-nightcycleprimarilybylight.Fishrepresentfascinatingmodelsfor decipheringthelightinputpathwaytothevertebrateclocksincefishcellclocksareregulatedbydirectlightexposure.Here wehaveperformedacomparative,functionalanalysisofthecircadianclockinvolvingthezebrafishthatisnormallyexposed totheday-nightcycleandacavefishspeciesthathasevolvedinperpetualdarkness.Ourresultsrevealthatthecavefish retainsafood-entrainableclockthatoscillateswithaninfradianperiod.Importantly,however,thisclockisnotregulatedby light.Thiscomparativestudypinpointsthetwoextra-retinalphotoreceptorsMelanopsin(Opn4m2)andTMT-opsinas essentialupstreamelementsoftheperipheralclocklightinputpathway.Citation: CavallariN,FrigatoE,ValloneD,Fro ¨ hlichN,Lopez-OlmedaJF,etal.(2011)ABlindCircadianClockinCavefishRevealsthatOpsinsMediatePeripheral ClockPhotoreception.PLoSBiol9(9):e1001142.doi:10.1371/journal.pbio.1001142 AcademicEditor: UeliSchibler,UniversityofGeneva,Switzerland Received May15,2011; Accepted July29,2011; Published September6,2011 Copyright: 2011Cavallarietal.Thisisanopen-accessarticledistributedunderthetermsoftheCreativeCommonsAttributionLicense,whichpermits unrestricteduse,distribution,andreproductioninanymedium,providedtheoriginalauthorandsourcearecredited. Funding: ThisworkwassupportedbyfundingfromtheMax-Planck-InstituteforDevelopmentalBiology,Tu ¨ bingen,KarlsruheInstituteofTechnology(KIT, Germany),theCNRS(France),theUniversityofFerrara(Italy),MIUR(Italy)projectsPRIN2008andAzioneIntegrataItalia-Spagna,theVIGONIprog ramoftheDAAD andtheAIT-MIUR,andtheMICINN(Spain)projectsCRONOSOLEAandAQUAGENOMICS.JFLOhasapostdoctoralfellowshipfromFundacionSeneca(Murcia, Spain).Thefundershadnoroleinstudydesign,datacollectionandanalysis,decisiontopublish,orpreparationofthemanuscript. CompetingInterests: Theauthorshavedeclaredthatnocompetinginterestsexist. Abbreviations: CF,cavefish;Clk,Clock;Cry,Cryptochrome;CT,circadiantime;DD,constantdarkness;d-old,daysold;LD,light/dark;LRM,lightresponsive module;per1b,period1b;per2,period2;TMT,teleostmultipletissue;UTR,untranslatedregion;wk-old,weeksold;zf,zebrafish;ZT,Zeitgeberti me *E-mail:bru@unife.itIntroductionThecircadianclockisahighlyconserved,physiological timingmechanismthatallowsorganismstoanticipateand adapttodailyenvironmentalchangesanditissynchronized primarilybylight.Inmammals,intrinsicallyphotosensitive retinalganglioncellsserveastheprincipalcircadianphotoreceptors[1].Innon-mammalianvertebrates,photoreceptors locatedoutsideoftheretina(inthepinealcomplexandinthe deepbrain)havealsobeenimplicatedintheregulationofthe circadiantimingsystem[2].Atthecoreofthevertebrate circadianclockisatranscriptiontranslationfeedbackloop mechanismcomposedofactivatorandrepressorclockproteins [3].Light-inducedexpressionofcertainclockgenesrepresents akeystepintherelayoflightinginformationtothecoreclock machinery[4,5]. Thezebrafish( Daniorerio )representsafascinatingmodelto studythemechanismswherebylightregulatestheclock.The ‘‘peripheral’’clocksinmostzebrafishtissuesandevencelllinesare entrainedbydirectexposuretolight[6].However,fundamental questionsconcerningtheidentityofthewidelyexpressed photoreceptormoleculesandhowtheysignaltoperipheralclocks remainunanswered.Todate,asetofwidelyexpressedopsins,one cryptochromehomolog,andflavin-containingoxidaseshaveall beenimplicatedascandidateperipheralphotoreceptors[7–9]. Incertainextremeenvironmentssuchascavessomefish specieshaveremainedcompletelyisolatedfromtheday-night cycleformillionsofyears[10].Theyshowconvergentevolution, sharingarangeofstrikingphysical‘‘troglomorphic’’properties includingnotablydegenerationoftheeyesduringearly development.However,manyaspectsofcavefishbiologystill remainincompletelyunderstood.Doesevolutioninconstant darknessleadtolossofotheraspectsofphotoreceptorfunction includingregulationofperipheralcircadianclocksbylight? Furthermore,dotheseremarkableanimalsevenretainnormal circadianclocks? Inthisreportweexplorethecircadianclockanditsregulation bylightin Phreatichthysandruzzii ,aSomaliancavefishthatshowsan extremetroglomorphicphenotype.Wecomparethecircadian clockmechanismofthiscavefishwiththatofthezebrafishwiththe goalofidentifyingkeycomponentsofthelightinputpathway.We revealthat P.andruzzii possessesaclockthatisentrainedby periodicfoodavailability,displaysalonginfradianperiod,and lackstemperaturecompensation.However,importantly,this cavefishclockisnolongerentrainablebylight.Strikingly,inthe cavefishweencountermutationsinthecandidatenon-visual photoreceptorsMelanopsin(Opn4m2)andTMT(teleostmultiple tissue)-opsinandweprovidedirectevidenceforalight-sensing functionofthesenon-visualopsinsintheregulationofvertebrate peripheralclocks. PLoSBiology|www.plosbiology.org1September2011|Volume9|Issue9|e1001142

PAGE 2

Results ExploringtheCircadianClockintheCavefish PhreatichthysandruzziiWechosetostudyaspeciesofcavefishexpressinganextreme ‘‘troglomorphic’’phenotype, P.andruzzii (Figure1B).This Somaliancavefishevolvedfromsurfacedwellingancestors, isolatedinatotallydarkenvironmentbeneaththedesertata nearlyconstanttemperaturefor1.4–2.6millionyears[10], approximately1millionyearslongerthanthewell-studiedcavefish Astyanaxmexicanus [11]. P.andruzzii showstotaleyedegeneration, noscales,andcompletedepigmentation[12]. Asafirststepwewishedtoexplorethecircadianclockandits regulationbylightinthiscavefishspecies.Weinitiallymeasured locomotoractivityincavefishexposedtoa12:12light-dark(LD) cycle(Figures1D,FandS1B).Wedocumentedstrikingarrhythmic locomotoractivitythatcontrastswiththeclearrhythmic,diurnal patternobservedforzebrafishunderthesameconditions (Figures1A,C,EandS1A).Inordertoexploreatthemolecular levelthislackofbehaviouralrhythmicityincavefish,wenext subclonedasetof P.andruzzii clockgenehomologswiththeaimof examiningtheirexpressionpatternunderLDcycles(TableS1). Thesegeneswereselectedsincetheirzebrafishcounterpartsare eitherclock-orlight-regulated.Sequenceanalysisconfirmeda closesimilaritybetweenzebrafishand P.andruzzii genes(TableS2 andFigureS2),consistentwithbothspeciesbelongingtothe Cyprinidaefamily.Wemeasuredtheexpressionofasubsetof clock-regulated( Clk1a , Clk2 , Per1 , Per1b )andlight-regulatedgenes ( Per2 , Cry1a ,and Cry5 )[4,5,8,13,14] invivo inadulttissuesandin wholelarvaefrombothspecies(Figure1G–JandFigureS3).Inthe zebrafish,inagreementwithpreviousreports,exposuretoaLD cycleresultsinrobustrhythmicexpressionofthesegenes (Figure1G,IandFigureS3A,B,E)[6,14].Remarkably,arrhythmic geneexpressionwasencounteredincavefishtissuesandlarvae (Figure1H,JandFigureS3C,D,F,G),evenduringthefirstdayof lifewheneyerudimentsstillpersist[12].Importantly,theseclock geneexpressionpatternsareconsistentwiththebehavioural activityprofilesobservedunderLDcyclesforbothspecies. Circadianclocksareencounteredinmostvertebratecelltypesand evenincellcultures invitro .Furthermore,thepersistenceof circadianclocks invitro hasenabledmanymoredetailed mechanisticstudies.Therefore,weestablishedcelllinesfrom adultcavefishcaudalfins(CFcells)totestwhether P.andruzzii cells invitro alsolackrhythmicgeneexpressionunderLDconditions. Againweobservedarrhythmicclockgeneexpressionin P.andruzzii cells(Figure2C,D)thatcontrastswiththerobustrhythms documentedinzebrafishcells(Figure2A,B)[6,15,16].Thus,we failedtodetectrhythmicclockgeneexpression invivo or invitro in P.andruzzii underLDcycles.Soourdatarevealthateither P. andruzzii lacksthecircadianclockitselforithasaclocklackinga functionallightinputpathway. Todistinguishbetweenthesetwohypotheses,wefirstassessed whetherthecavefishcircadianclockcouldbeentrainedbyan alternativeenvironmentaltimesignal( zeitgeber ),periodicfood availability.Adultcavefishandzebrafishwerefedonceatthesame timeeachdayforonemonthunderconstantdarkconditions,and duringthisperiod,locomotoractivitywasmeasured(Figure3).For bothspeciesweobservedfoodanticipatoryactivity(FAA,[17]),a characteristicincreaseinlocomotoractivityencounteredafew hourspriortomealtime(Figure3A,B,D,E)andastrong entrainmentofrhythmiclocomotoractivity(Figure3C,F).FAA isindicativeofregulationbyafoodentrainableoscillator(FEO, [18]);thuswetestedcircadianclockgeneexpressioninvarious tissuesofbothspeciesduringthefinaldayoffoodentrainmentand thentwodaysoffastingunderconstantconditions(Figure4A). Consistentwithpreviousresults[19],rhythmicclockgene expression( Clk1a and Per1b )wasobservedinzebrafishbrain, heart,fin,andliver(Figures4B,D,FandS4A),withtheonly exceptionof Clk1a inthezebrafishbrain(Figure4B,redtrace). However,inallcavefishtissuesincludingthebrain,robust circadianrhythmsofboth Clk1a and Per1 expressionwere observed(Figures4C,E,GandS4B).Inbothspecies,differences inthephaseandamplitudeofrhythmicexpressionforeachgene wereobservedbetweendifferenttissues.Importantly,theseresults pointto P.andruzziihavingafunctionalclockthatisentrainableby feedingbutnotbyLDcycles.Thiscontrastswiththesituationin zebrafishwherebothlight-andfood-entrainableoscillatorsare present. Wealsotestedtheeffectofalternativezeitgebersonclockgene expressionincavefishCFcells.Transienttreatmentswith glucocorticoidsarewidelyusedtoinducerhythmicgene expressioninculturedcells[20].CFcellstransfectedwitha clock-regulatedzebrafishreporterconstruct( zfPer1b-Luc )[13]were treatedtransientlywith100nMdexamethasone,anagonistofthe glucocorticoidreceptor(FigureS5;Figure5B,greentrace)[20]. Surprisingly,thisinducedabioluminescencerhythmincavefish cellsthatpersistedforalmostthreecycleswithanextremelylong period( t =43hat25 u C).Thiscontrastswiththecircadian bioluminescencerhythmobservedupondexamethasonetreatment ofzebrafishcells( t =24.2hat25 u C)(Figure5A,greentrace). Theseresultsrevealtheexistenceofanabnormalcircadianclock in P.andruzzii thatdisplaysaninfradianperiodunderconstant conditions.Wenextwishedtodeterminewhetherothercircadian clockpropertiesareabnormalinthecavefish.Onehighly conservedfeatureofthecircadianclockisthatitsperiodremains relativelyconstantoveraphysiologicalrangeoftemperatures(socalledtemperaturecompensation)[21].Thus,wemeasuredthe periodofrhythmsinducedbydexamethasonepulsesinCFand zebrafishcellsheldattwoadditionalconstanttemperatures22 u C and29 u C(Figure5).Withanincreaseintemperaturetheperiod lengthofthecavefishclockdecreasedsignificantly( t =47(22 u C), 43(25 u C),and38h(29 u C))revealingreducedtemperature compensationwithQ10< 1.35,whileinzebrafishcells,asexpected, AuthorSummaryThecircadianclockisaphysiologicaltimingmechanism thatallowsorganismstoanticipateandadapttothedaynightcycle.Sinceittickswithaperiodthatisnotprecisely 24h,itisvitalthatitisresetonadailybasisbysignals suchaslighttoensurethatitremainssynchronizedwith theday-nightcycle.Themolecularmechanismswhereby lightregulatestheclockremainincompletelyunderstood. Herewehavestudiedacavefishthathasevolvedfor millionsofyearsintheperpetualdarknessofsubterranean cavesinSomalia.Likemanyothercaveanimals,thesefish displaystrikingadaptationstotheirextremeenvironment, includingcompleteeyedegeneration.Weshowthat despiteevolvinginaconstantenvironment,thisblind cavefishstillretainsacircadianclock.However,thisclock tickswithanextremelylongperiod(nearly47h),and importantlyitdoesnotrespondtolight.Werevealthat eyelossdoesnotaccountforthis‘‘blind’’clock. Specifically,mutationsoftwowidelyexpressednon-visual opsinphotoreceptors(MelanopsinandTMTopsin)are responsiblefortheblindclockphenotypeinthecavefish. Ourworkillustratesthegreatutilityofcavefishfor studyingtheevolutionandregulationofthecircadian clock.CavefishBlindClockRevealsPhotoreceptors PLoSBiology|www.plosbiology.org2September2011|Volume9|Issue9|e1001142

PAGE 3

arelativelyconstantperiodwasobserved( t =23.6(22 u C),24.2 (25 u C),and24.6(29 u C))h,respectively(Q10< 1[22]).Thus, togetherourresultsindicatethat P.andruzzii possessesacircadian timingsystemwithanaberrantcoreclockmechanism. Couldthediscrepancybetweenthestrikinginfradianperiodof the P.andruzzii clock( t . 30h)andtheperiodoftheLDcycle (T=24h)havebeentheoriginoftheobservedarrhythmicity underLDcycles?Insuchascenario,thecavefishclockmightstill beentrainablebylight.Inzebrafishcelllines,exposuretobrief lightpulsesresultsinadvancesordelaysinpre-existingclockgene expressionrhythmsdependingontheprecisetimewhenthecells areilluminated[13].Thus,wewishedtotestwhetherlightisable toregulatethecavefishclockinasimilarmanner.We Figure1.LackofrhythmicityinP.andruzziiunderlightdark cycles. Adultzebrafish(A)and P.andruzzii (B).(C,D)Representative actogramsofzebrafish(C)andcavefish(D)maintainedunderLD conditions(eachcycle:12hlight,12hdark)andfedrandomly.Records aredoubleplottedona48htimescaletoaidinterpretation;the y -axis progressesinsingledayswitheachdaybeingplottedtwice(day1on therightsideisrepeatedonday2ontheleftside).Theactivitywas binnedevery10min,theheightofeachpointrepresentingthenumber ofinterruptionsoftheinfraredlightbeam.(E,F)Meanwaveformsof zebrafish(E)andcavefish(F)arerepresented.Eachpointinthemean waveformhasbeencalculatedasthemean 6 SEMfrom10minbinned dataacrossalltheexperimentaldaysshown( n =16)oneachactogram andforallexperimentalaquaria( n =6forzebrafishand n =6for cavefish).For x2periodogramanalysis(confidencelevel,95%),see FigureS1.(G–J)QuantitativeRT-PCRanalysisofendogenousclockgene expressionfor2consecutivedaysinadultzebrafishand P.andruzzii fins (G,IandH,J,respectively; n =6pertimepoint)exposedto12:12LD cycles.Forallpanels,eachpointrepresentsthemean 6 SEM.In P. andruzzii ,neitherclock-regulated( Per1 (blue), Clk1a (red),and Clk2 (black)(H)),norlight-regulatedgenes( Cry1a (purple), Cry5 (orange),and Per2 (green)(J)),showsignificantcycling( p . 0.1)comparedwithhigh amplituderhythmsforthesamegenesobservedinthezebrafish ( p , 0.0001)(panelsGandI).Whiteandblackbarsrepresentlightand darkperiods,respectively.Onthe y -axesareplottedrelativeexpression levels,whileonthe x -axestimeisexpressedaszeitgebertime(ZT, whereZT0representslightson). doi:10.1371/journal.pbio.1001142.g001 Figure2.LackofrhythmicclockgeneexpressioninP.andruzziicelllinesexposedtoLDcycles. QuantitativeRT-PCRanalysisof clockandlight-regulatedclockgeneexpressioninzebrafish(A,B)and cavefish(CF;C,D)celllines.Thepresentation,colorcodefortheplotted data,andtheexperimentalconditionsareidenticaltoFigure1G–J.Each pointrepresentsthemean 6 SEM. doi:10.1371/journal.pbio.1001142.g002 CavefishBlindClockRevealsPhotoreceptors PLoSBiology|www.plosbiology.org3September2011|Volume9|Issue9|e1001142

PAGE 4

Figure3.Behavioralentrainmentbyperiodicfoodavailability. Representativeactogramsofzebrafish(A)andcavefish(D)maintainedunder constantdarknessandfedonceadayatafixedtime(ZT=0).Feedingtimeisindicatedbytheblackarrowatthetopofeachactogram(formore detailsonactogramrepresentation,seelegend;Figure1).(B,E)Meanwaveformsofzebrafish(B)andcavefish(E)arerepresented.Eachpointinthe meanwaveformhasbeencalculatedasthemean 6 SEMfrom10minbinneddataacrossalltheexperimentaldays( n =30)shownoneachactogram andallexperimentalaquaria( n =5forzebrafishand n =3forcavefish).(C,F) x2periodogramanalysis(confidencelevel,95%)forzebrafish(C)and cavefish(F)actograms.Theperiodogramindicatesthepercentageofvariance(%V)oftherhythmexplainedbyeachanalyzedperiodwithinarange of20–28h.Theslopeddottedlinesrepresentthethresholdofsignificance,setat p =0.05.Periodogramanalysisshowedthepresenceofbehavioral activityrhythmssynchronizedtothe24hfeedingcycles. doi:10.1371/journal.pbio.1001142.g003 CavefishBlindClockRevealsPhotoreceptors PLoSBiology|www.plosbiology.org4September2011|Volume9|Issue9|e1001142

PAGE 5

synchronizedrhythmicclockgeneexpressioninCFcellsby dexamethasonetreatmentandthenexposedthecellstoa15min lightpulseatfivedifferenttime-pointsdistributedthroughoutthe 43hcycle(FigureS6)[13,23].Noneofthelightpulseschanged eitherthephaseorlevelsofrhythmicclockgeneexpression. Hence,weconcludethat P.andruzzii indeedpossessesatrulyblind clock.Thus,thiscavefishrepresentsapowerfulcomplementary modelforexploringthefunctionofthelightinputpathwayin vertebrates.DefiningMolecularDefectsintheLightInputPathwayof P.andruzziiOursystematiccloningandsequencingstrategyfailedtodetect anymutationssignificantlyaffectingclockgenecodingsequences (unpublisheddata).Light-inducedtranscriptionofclockgenes representsakeystepinphoticentrainmentofthezebrafishclock [4,5].Therefore,wespeculatedthatmutationsinpromoter sequencesoflight-inducibleclockgenescouldaccountforthe cavefishblindclockphenotype.Wehavepreviouslydefinedan essentiallightresponsivemodule(LRM)containingE-andD-box enhancerelementsinthezebrafish per2 promoter[5].IntheLRM promoterregionfromthecavefish per2 gene,weidentifiedseveral basesubstitutionscomparedwiththezebrafishsequence,including oneintheE-boxelement(Figure6A).Wewishedtotestwhether thecavefishLRMwasstillabletodirectlight-inducedgene expression.Withthisaimwetransfectedzebrafishcellswitha cavefish per2 promoterdrivingluciferaseexpression( cfPer2-Luc ) (Figure6B,greentrace).Thesecellsexhibitedrobustlight-driven Figure4.Anon-photiczeitgeberentrainsthecavefishclock. (A)Adultzebrafishandcavefishwerefedonceperday(atZT0)for30d(dark bluearrowheads).Fishwerethenstarvedandafter24h,brains,hearts,andfinswereharvestedeach4h( n =4)duringthelastdayoffood entrainmentandthefirst2doffasting(blackarrowheads).(B–G)RealtimePCRanalysisofrhythmicendogenous Clk1a (red,orange)and Per1 (green, darkblue)expressioninthebrain(B,C),heart(D,E),andfin(F,G)ofzebrafish(B,D,andF)andcavefish(C,E,andG).TimeisexpressedasZTor CircadianTime(CT)duringstarvation.Ineachpanel,asolid,verticalline(atZT0)indicatesthelastfeedingtime.Subsequentlyduringstarvatio n,the twoverticaldottedlines(atCT0andCT24)denotewhenthefeedingwouldnormallyhaveoccurredaccordingtothepreviousregularfeeding regime.Eachpointrepresentsthemean 6 SEM.Inallcavefishtissuestested(C,E,G)robustcircadianrhythmsof Clk1a and Per1 expressionwere observed( p , 0.01).Rhythmicclockgeneexpressionwasalsoobservedinzebrafishheart(D)andfin(F)( p , 0.01).However,inthezebrafishbrain(B) only Per1b (greentrace)wasrhythmicallyexpressed( p , 0.01).SeealsoFigureS4. doi:10.1371/journal.pbio.1001142.g004 CavefishBlindClockRevealsPhotoreceptors PLoSBiology|www.plosbiology.org5September2011|Volume9|Issue9|e1001142

PAGE 6

reporterexpressioncomparabletothatofthezebrafish per2 promoter( zfPer2-Luc )(Figure6B,bluetrace).Conversely,zebrafish zfPer2-Luc transfectedintocavefishcellsfailedtoshowlightinducedexpression(Figure6C,greentrace).Together,these resultsindicatethatmutationsdisruptingthecavefishlightinput pathwayshouldlieupstreamofdirectlylight-regulatedclockgene promoters.TMT-OpsinandMelanopsinAreEssentialPhotoreceptors forPeripheralClocksMutationsaffectingperipheralphotoreceptorscouldalso accountfortheblindcavefishclock.Althoughtheidentityofthe teleostperipheralphotoreceptorremainsunclear,candidates includetheopsinsMelanopsin(Opn4m2)andTMT-opsinthat arewidelyexpressedinmosttissues[9,24].Melanopsinwas originallyisolatedfromthephotosensitivemelanophoresof Xenopus [25].Subsequently,orthologsofMelanopsinwereisolatedfrom othernon-mammalianvertebratesincludingzebrafish[26].TMTopsinwasoriginallyidentifiedbyvirtueofitsopsinsequence homology,andtodatehasonlybeenisolatedfromteleosts[9].We chosetocloneandcharacterizethesetwoopsinsinthecavefish. Figure5.ReducedtemperaturecompensationoftheP. andruzziiclock. Followingtransfectionwith zfper1b-Luc ,zebrafish(A) andCF(B)cellsweretransientlytreatedfor30minwithdexamethasoneandthenincubatedeitherat22 u C(blue),25 u C(green),or29 u C (red).Dexamethasoneinducesrhythmicclockgeneexpressionincells ateachofthethreetemperatures( p , 0.0001).Whileinzebrafishcells (A)theperiodremainsrelativelyconstantwithchangesintemperature (Q10< 1),CFcells(B)showedamarkedshorteningofperiodlengthwith increasingtemperatureindicatingapoorlytemperaturecompensated clock(Q10< 1.35)(forcalculations,seeMaterialsandMethods).Each pointrepresentsthemean 6 SEM. doi:10.1371/journal.pbio.1001142.g005 Figure6.Thedefectinthecavefishlightinputpathwaylies upstreamoftheperiod2genepromoter. (A)Alignmentofthe cavefish period2 promoterLRMwiththecorrespondingsequencefrom thezebrafishandothervertebrates[5].Boxinghighlightsthehighly conservedE-andD-boxenhancers.Withineachenhancer,yellow highlightedsequencesareconservedinallspecies.Bluehighlighting indicatesthemutationinthecavefishE-box.Below,asterisks(*)indicate perfectlyconservedsequences.(B)Zebrafishcellstransfectedwith cfPer2-Luc (greentrace)and zfPer2-Luc (bluetrace)andexposedtoLD cycles.Bothconstructswererhythmicallyexpressed( p , 0.0001).The strongestinductioncoincidedwiththeonsetofthelightperiod.(C)In vivoluciferaseassaysofcavefish(CF,greentrace)andzebrafishcells (ZF,bluetrace)transfectedwith zfPer2-Luc .Theabsenceofrhythmic reporterexpressionincavefishcells( p . 0.1)contrastswithhigh amplituderhythmsobservedinzebrafishcells( p , 0.00001).Theplotted valueswerecalculatedbysubtractingthebasalexpressionlevelsofthe luciferasereporter.Foreachtimepoint,mean 6 SEMisplotted.White andblackbarsrepresentthelightanddarkperiods,respectively. doi:10.1371/journal.pbio.1001142.g006 CavefishBlindClockRevealsPhotoreceptors PLoSBiology|www.plosbiology.org6September2011|Volume9|Issue9|e1001142

PAGE 7

WedocumentedmRNAexpressionofboth Melanopsin and TMTopsin invariouscavefishtissuesincludingtheCFcellline(Figure S7),consistentwithpreviousresultsthatrevealedwidespread expressionforthezebrafishhomologs[9,24].Furthermore, sequencealignmentwithotherteleosthomologsrevealedstrong conservation(unpublisheddata).Interestingly,however,prematurestop-codonswereencounteredinthecodingsequencesof bothTMT-opsinandMelanopsin(attheC-terminusofthe5thtransmembranedomainandtheN-terminusofthe6thtransmembranedomain,respectively)(Figure7A).TheseC-terminal truncationsweretheresultofframeshiftmutationsinthetwo codingsequences(aninsertionofoneTin TMT-opsin atposition + 654anddeletionofoneGin Melanopsin atposition + 842relative totheATGinitiationcodonsofthezebrafishhomologsequences). Bothopsinmutationswouldbepredictedtoeliminatebindingof theessentialchromophore,retinaldehyde,thatnormallyoccursat the7thtransmembranedomain.Therefore,wedecidedtotest whetherectopicexpressionofthezebrafishhomologsofthesetwo opsinswouldrescuelight-inducibilityofa zfPer2-Luc reporterin cavefishcells.Simplysupplementingtheculturemediumwith retinaldehydefailedtoinducerhythmicexpressionof zfPer2-Luc (Figure7B).Strikingly,uponcotransfectionwithsingleopsin expressionvectors, zfPer2-Luc wasrobustlyinducedduringthelight phaseandsubsequentlydecreasedduringthedarkphase (Figure7C,bluetraceand7D,redtrace).Incontrast,expression incavefishcellsofzebrafishMelanopsinandTMT-opsincarrying mutationsintroducingprematurestopcodonsequivalenttothe twocavefishopsins(zfOpn4m2K286XandzfTMTY224X)failedto Figure7.Rescueoflight-inducedgeneexpressionincavefish. (A)Schematicrepresentationoftheseventransmembranedomainsofa genericopsinprotein.RedandbluecrossesdenotethepositionofstopcodonsincavefishTMT-opsinandMelanopsin(Opn4m2),respectively.Below, theC-terminiofthetruncatedcavefishopsinproteinsarealignedwiththecorrespondingzebrafishsequences(asterisks(*)indicateidentical aminoacids;periods(.),non-conservativeaminoacids;andcolons(:),conservativeaminoacids.(B)Arrhythmicbioluminescenceexpression( p . 0.5)of zfper2-Luc inCFcellsunderLDcyclesinthepresence(black)orabsence(grey)of100nM9-cis-/all-trans-retinal.(C–D)Rescuedlight-inducible expressioninCFcellsof zfper2-Luc co-transfectedwithzebrafishOpn4m2(bluetrace,C)andTMT-opsin(redtrace,D)( p , 0.0001).Incontrast, expressionofthezebrafishOpn4m2K286X(C)andTMTY224X(D)carryingequivalenttruncationmutationstothetwocavefishopsins(greytraces)or TMTK295A(withlysine295mutatedtoanalanine)(greentrace)(D)failedtorescuerhythmiclightinduced zfper2-Luc expression( p . 0.1).Foreachtime point,mean 6 SEMisplotted.Whiteandblackbarsrepresentthelightanddarkperiods,respectively. doi:10.1371/journal.pbio.1001142.g007 CavefishBlindClockRevealsPhotoreceptors PLoSBiology|www.plosbiology.org7September2011|Volume9|Issue9|e1001142

PAGE 8

rescuelightinducible zfper2-Luc expression(Figure7CandD,grey traces).Thesedataareconsistentwiththepredictedlossofthe chromophorebindingdomainsinthecavefishopsins.Furthermore,ourresultsconstitutestrongevidencethatMelanopsinand TMT-opsinindeedfunctionasperipheralclockphotoreceptors. WhileMelanopsinrepresentsawell-characterizedphotoreceptor[25,26],untilnowTMT-opsinhasbeenimplicatedasa photopigmentbasedonlyuponsequencehomology[9].The cavefishrepresentsapowerfulmodeltoconfirmthatTMT-opsin indeedfunctionsasaphotopigment.Inallopsinsahighly conservedlysineresiduewithinthe7thtransmembranedomain providesaSchiffbaselinkagewiththechromophoreretinalthatis criticalforphototransduction[27].InzebrafishTMT-opsin,a lysineresidueatposition295wouldbepredictedtofulfilthisrole [9].Totestthisprediction,wemutatedlysine295toanalanine (TMTK295A).ExpressionofTMTK295Afailedtorescuelightinduced zfper2-Luc expressionincavefishcells(Figure7D,green trace),thusstronglyindicatingthatTMT-opsinindeedfunctions asanopsinphotopigment. Itiswellestablishedthattheresponseofopsinstolightis wavelengthdependent.Melanopsinhasbeenshowntorespond preferentiallytotheblueregionofthelightspectrum[25],while thewavelengthsensitivityofTMT-opsinisunknown.Are MelanopsinandTMT-opsinalonesufficienttoaccountforthe wavelengthsensingpropertiesofzebrafishperipheralclocks?Asa firststeptowardsaddressingthiskeyquestion,werepeatedour cavefish zfper2 expressionrescueassayunderthreedifferent monochromaticlightsources(blue,green,andred,FigureS8). Wethencomparedtheseresultswiththeresponseofthe zfper2 promotertothesamelightsourcesinzebrafishcells.Exposureof MelanopsinorTMT-opsintransfectedcavefishcellstoblue (468nm)orgreen(530nm)lightisabletoactivatethe zfper2 promoter(Figure8A–D).Incontrast,norescuewasobserved underred(657nm)light(Figure8E).However,exposureof zebrafishcellstothesesamemonochromaticlightsources revealedactivationbyallthreelightsources,withthestrongest inductionbyblue(Figure8F).Theseresultsconfirmthe wavelengthsensitivityofMelanopsin[28]andprovidethefirst evidence,toourknowledge,thatTMT-opsincanrespondtoblue andgreenbutnotredwavelengthsoflight.Thedifferencesinthe redlightresponseofthe zfper2 promoterbetweenthezebrafish cellsandtherescuedcavefishcellsystempointtotheexistenceof additionalperipheralphotoreceptors.Importantly,comparable resultswereobtainedinallourrescueexperimentswhenwe substituted zfper2-Luc fortheminimallight-responsivepromoter fragmentofthezebrafish per2 gene 2 0.43per2:Luc (unpublished data)[5].Thispredictsthatlight-drivengeneexpression mediatedbyMelanopsinandTMT-opsinisdependentupon theLRMofthe per2 promoter.DiscussionInsummary,during1.4–2.6millionyearsofisolationfromthe day-nightcycle,theevolutionofthecavefish P.andruzzii haslead toanaberrantcircadianclock.Contrarytothesituationinmost organisms,thisclockisnolongerentrainedbylight.Furthermore, uponexposuretoalternativezeitgebersitcycleswitharemarkably longinfradianperiodandshowsreducedtemperaturecompensation.Itispossiblethatthisreflectsprogressivelossofamechanism thatprovidesnoselectiveadvantageforanimalsthatliveunder constantdarknessandtemperature.Insupportofthishypothesis, onerecentreporthasdocumentedasevereattenuationof circadianclockrhythmicityinanarcticmammalnaturally exposedtotheextremepolarphoticenvironment,thereindeer, Rangifertarandus [29].Tofurthertestthishypothesis,itwillbe fascinatingtocomparethecircadianclockphenotypeof P. andruzzii withthatofothercavefishspeciesrepresentingthefull rangeoftroglomorphicphenotypes. Ourstudyrevealsthataregulardailyfeedingtimedoesentrain thecavefishclock.Interestingly,acomparisonoftheclockgene expressionrhythmsencounteredinvarioustissuesoftheregularly fedcavefishrevealsmanydifferencesinphaseandamplitude betweentissues.Thissituationisstronglyreminiscentofthe differentpatternsofcyclinggeneexpressionobservedinindividual lightentrainedperipheralclocksofzebrafish[30]andmayreflect differencesinthemolecularregulatorynetworksassociatedwith eachtissue.Severalreportspointtothegeneralimportanceof feedingentrainmentforthecircadiantimingsysteminfish[17,19]. Itistemptingtospeculatethatfoodavailabilityinthe subterraneanenvironmentofthiscavefishmightindeedbe periodic,andthereforeaclockrespondingtoandanticipating feedingtimemayconferasurvivaladvantage.Interestingly, severallinesofevidencepointtotheexistenceofafoodentrainableoscillatorinvertebratesdistinctfromthelightentrainableoscillator,howevertheanatomicallocationand molecularmechanismremainsunclear[18,19,31,32].Inthis regard,fishcouldemergeaspowerfulmodelsfortheinvestigation offoodentrainment.Acomparativestudyinvolvingzebrafishthat possessbothlightandfoodentrainableoscillatorsand P.andruzzii thatretainsonlythefoodentrainableoscillatorcouldprovide importantinsightintothebasisofthefeedingentrainment mechanisminvertebrates. Byemployingacomparativefunctionalanalysisinvolving zebrafishand P.andruzzii ,wehavebeenabletoprovidedirect evidencethatTMT-opsinandMelanopsinserveasperipheral tissuephotoreceptorsinteleosts.Furthermore,ourresultsstrongly indicatethatthetwoopsinsactivategeneexpressionviatheLRM promoterelementthatmediateslight-driven per2 expression[5]. Previously,theidentityofthewidelyexpressedphotoreceptorsthat mediatethedirectentrainmentofperipheralclockswasunclear. Inadditiontoopsins,othercandidatesincludeCry4,amemberof thecryptochromefamilyinzebrafish,andflavin-containing oxidases[7,8].Itwillnowbeimportanttore-evaluatetherelative contributionoftheseothercandidatephotoreceptorstoperipheral clockentrainment.Basedonourstudyoftheeffectsofdifferent wavelengthsoflightontheinductionofclockgeneexpressionwe predictthatMelanopsinandTMT-opsinarenottheonly photoreceptorsforperipheralclocks.Thisfindingbegsthe questionastowhyseveraldifferentphotoreceptorsmight contributetothephoticentrainmentpropertiesofperipheral tissues.Changesinthephotoperiod,intensity,aswellasthe spectralcompositionofsunlightrepresentreliableindicatorsof day-nightaswellasseasonalchangesintheenvironment[33].The presenceofmultiplephotoreceptors,eachonedifferentially extractingtiminginformationfromsunlight,couldenablethe circadiansystemtomorereliablyindicatethetimingofdawnand dusk. Finallyfromabroaderperspective,inadditiontodisplayinga uniqueandfascinatingcollectionofadaptationstoitsextreme environment,wehavedemonstratedthat P.andruzzii servesasa powerfulcomplementarymodeltodissectthemolecularpathways thatrespondtolight.MaterialsandMethods EthicsStatementTheanimalhandlingproceduresandresearchprotocolswere approvedbytheUniversityofFerrara(Italy),UniversityofFirenzeCavefishBlindClockRevealsPhotoreceptors PLoSBiology|www.plosbiology.org8September2011|Volume9|Issue9|e1001142

PAGE 9

(Italy),UniversityofMurcia(Spain),andKarlsruheInstituteof Technology(Germany)InstitutionalAnimalCareandUse Committees.CavefishandCavefishCellLinesP.andruzzii originallycollectedfromSomalia(FigureS9)were maintainedandbredattheUniversityofFirenze,Italy.The Figure8.Rescueoflight-inducedgeneexpressionincavefishcellsunderdifferentmonochromaticlightsources. Rescuedrhythmic light-inducibleexpressionof zfper2-Luc inCFcellstransientlyco-transfectedwithzebrafishOpn4m2(bluetrace,(AandC))andTMT-opsin(redtrace, (BandD))exposedtoblue(468nm,AandB)orgreen(530nm,CandD)light( p , 0.0001).Incontrast,norescuewasobservedunderred(657nm) light( p . 0.1)(E).ThetruncatedzebrafishOpn4m2K286XandTMTY224X(greytraces)failedtorescuerhythmiclight-induced zfper2-Luc expressioninall lightingconditions( p . 0.2)(A–D).Interestingly,inzebrafishcells(PAC-2 2 1.7per2:Luc )exposedtoeachofthesamethreemonochromaticlight sources, zfper2-Luc reporterexpressionislight-driven( p , 0.0001)(blacktrace,(F)),withthestrongestinduction( + 35%ofpeakvalues)observedunder bluelight.Aboveeachpanel,colouredandblackbarsindicatethetypesoflightsourcesutilizedandthedurationofthelightanddarkperiods.For eachtimepoint,mean 6 SEMisplotted. doi:10.1371/journal.pbio.1001142.g008 CavefishBlindClockRevealsPhotoreceptors PLoSBiology|www.plosbiology.org9September2011|Volume9|Issue9|e1001142

PAGE 10

cavefishwerekeptindarknessataconstant27 u Cexceptduring foodadministrationandaquariamaintenance.Threetimesper weekthefishwerefedwithfrozenchironomidlarvae.Fertilized eggswerecollectedevery30minandaliquotsof10–20eggswere transferredinto75cm2tissuecultureflasks(BDGmbH).Flasks weresealedandsubmergedinalargevolume,thermostatically controlledwaterbath(Tetra).Fromthethird/fourthdayafter hatching,larvaewerefedonceaday. Cavefish(CF)celllineswerederivedfromfinclipsofadultfish andmaintainedusingstandardmethodsdescribedelsewhere[34]. CellsweretransientlytransfectedusingFuGeneHDreagent accordingtothemanufacturer’srecommendations(Roche)inthe absenceofserum.Rhythmicclockgeneexpressionoriginally establishedbyserumtreatmentduringtheseedingoftheCFcells wasresynchronizedbya30mintreatmentwitharangeof dexamethasone(Sigma)concentrationsfrom50nMto1mM(see FigureS5)[20].Subsequentlythedexamethasone-containing mediumwasreplacedbyfreshmediumagainlackingserumand containingluciferin.LightingConditionsTotestforphoticentrainmentofrhythmicclockgene expression,cavefishandzebrafishadultsandlarvaeaswellas celllinesweremaintainedat27 u Cundera12:12LDcyclewitha lightintensityof350mW/cm2(full-spectrumcoolfluorescent tubes,OsramGmbH).Forbehaviouralanalysis,zebrafishand cavefishweremaintainedunderfull-spectrumcoolfluorescent tubelightsourceswithalightintensityof20mW/cm2.For monochromaticlightsources,light-emittingdiodes(LED,Kopa) sourceswereused(blue: lpeak=468nm,green: lpeak=530nm, red: lpeak=657nm;white:450nm , l , 700nm)(FigureS8).The lightintensityofeachLEDlightsourcewasadjustedtoensurean equivalentnumberofphotonswereemittedfromeachsource (1.42 6 10186 0.04 6 1018photons/s/m2).ZebrafishandZebrafishCellLinesAllexperimentsusingadultandlarvalzebrafishaswellasthe zebrafishcelllinesAB9[35],PAC-2,andastablePAC-2cellline expressing 2 1.7per2:Luc wereperformedusingstandardmethods describedelsewhere[5,13].Inductionofrhythmicclockgene expressioninzebrafishcellsusingtransientdexamethasone treatmentwasperformedasdescribedforthecavefishCFcell line(seealsoFigureS5).RecordingofAdultLocomotorActivityCavefishandzebrafishlocomotoractivitywasregistered continuouslybymeansofaninfraredphotocell(E3S-AD62, Omron)placedattheaquariumwall,inthecornerwherefoodwas provided.Thephotocellwasplaced5cmfromthewatersurface and20cmfromthebottom.Thenumberoflight-beam interruptionswascountedandstoredevery10minbyacomputer connectedtothephotocell.Theanalysisoflocomotoractivity records,representationofactograms,andcalculationsofmean waveformsand x2periodograms(Sokolove-Bushelltest)were performedusingthechronobiologysoftware ElTemps (version 1.228).CloningCavefishcDNASequencesToobtainpartialcDNAsequences,single-strandedcDNAwas synthesizedusingSuperScriptIIIReverseTranscriptase(Invitrogen).CavefishgeneswereamplifiedbyPCRusingTaqDNA Polymerase(Invitrogen)withprimersdesignedbyPrimer3 softwareonthebasisofsequenceofthezebrafishhomologs (TableS1).Bandsofthepredictedsizeswereclonedintothe pGEM-TEasyVector(Promega).ThecavefishgenecDNA fragmentsweresequenced(QIAGENGmbH)andcomparedwith theGenBankdatabasebyusingtheBLASTalgorithm.Additional cDNAsequencesweresubsequentlyclonedusinga5 9 -3 9 SMART RACEcDNAamplificationkit(BDBioscience),andthencoding sequencesweredepositedinGenBank(TableS2).Bythis approach,wecloned13clockgenes( Per1 , Per2 , Per3 , Cry1a , Cry1b , Cry2a , Cry2b , Cry3 , Cry4 , Cry5Clk1a , Clk1b , Clk2 )and2opsins ( Opn4m2 , TMT-opsin )from P.andruzzii .PhylogeneticAnalysisSequencesfromthecavefishPERandCLKproteinfamilies werealignedwithhomologsfromotherteleostspecies( Takifugu rubripes , Tetraodonnigroviridis , Daniorerio , Gasterosteusaculeatus ,and Oryziaslatipes )[36,37]usingClustalW.Alignmentsweremanually verifiedandphylogenetictreesweregeneratedusingNeighbourjoiningmethods[38]withacompletedeletionmode.Bootstrap testswereperformedwith1,000replications.Poissoncorrection distancewasadoptedandratesamongsitesweresetasuniform. Drosophilamelanogaster PERandCLKsequenceswereusedasan out-grouptorootthetrees.GeneExpressionAnalysisSingle-strandedcDNAwassynthesizedusingSuperScriptIII ReverseTranscriptase(Invitrogen).QuantitativePCRwas performedfor P.andruzzii andzebrafishclockgenesusingthe pairsofprimersshowninTableS1.TheStepOnePlusReal-Time PCRSystem(AppliedBiosystems)wasemployedusingSYBRgreen-primer-mastermixaccordingtothemanufacturer’srecommendationswiththefollowingcycleconditions:15minat95 u C, then40cyclesof15sat95 u C,and30sat60 u C.Therelative levelsforeachRNAwerecalculatedbythe22 DD CTmethod. Relativeexpressionlevelswerenormalizedto b -actin .EachCT valueisthemeanofthreebiologicalreplicatesandeachassaywas performedaminimumofthreetimes.LuciferaseReporterConstructsandBioluminescence AssayszfPer1b-Luc containsapromoterregionextending3.3kb upstreamofthe5 9 endofthe Period1b cDNA(equivalenttothe zfperiod4 promoterconstructin[13]).The zfPer2-Luc reporterhas beendescribedpreviouslyas 2 1.7per2:Luc thatcontainsafragment of1,571bpupstreamofthetranscriptionstartsiteand129bpof the5 9 UTRofthezebrafish Period2 gene[5].Inaddition,the minimallightresponsive per2 promoterreporterconstruct 2 0.43per2:Luc wasalsotestedinthecavefishcellrescue experiments.Thisconstructcontains431bpupstreamofthe transcriptionstartsiteand164bpofthe5 9 UTRofthezebrafish Period2 gene[5].Both zfPer2-Luc reportersrespondedinan equivalentmanner. cfPer2-Luc contains876bpupstreamofthe transcriptionstartsiteand112bpofthe5 9 UTRofthecavefish Period2 gene.All invivo bioluminescenceassayswereperformedas describedpreviously[13,34].OpsinExpressionVectorsThefull-lengthzebrafishTMT-opsin(ENSDART000 00081729)wasamplifiedwiththeprimersFwd5 9 -AATGGATTGCGGATTGGATCCATTGTGTCCAACTTG-3 9 and Rev5 9 -CTGCAGAATTCACTAGTGATTTCGCCTGTA-3 9 resultinginthemutationoftheATGtranslationinitiationcodon sequenceintoTTCandthecreationofBamHIandEcoRI restrictionsites,respectively,forcloningintoamodifiedCavefishBlindClockRevealsPhotoreceptors PLoSBiology|www.plosbiology.org10September2011|Volume9|Issue9|e1001142

PAGE 11

pcDNA3.1( + )expressionvector(Invitrogen)thatincorporatesan HA-TagintotheN-terminusoftheexpressedprotein[39]. Similarly,thefulllengthzebrafish Opn4m2 (ENSDART 00000018501)wasamplifiedwiththeFwd5 9 -GCTCGGATCCGCCTTGAGCCATCACTCTTCA-3 9 andRev5 9 -GCCCTCTAGACTCTTAGTTCCCTCCAAGCAA-3 9 primers,thusmutatingtheATGinitiationcodonintoTTGaswellascreatingBamHI andXbaIsites,respectively,forcloningintothepcDNA3.1( + )HAtagmodifiedexpressionvector.PCRreactionswereperformed usingthePerkinElmerGeneAmpXLPCRkitaccordingtothe manufacturer’sinstructions.Constructionoftheexpressionvectors forzfTMTK295AandthetruncatedformszfOpn4m2K286Xand zfTMTY224Xinvolvedsite-directedmutagenesisusingtheQuikChangeMultiSite-DirectedMutagenesisKit(Stratagene)accordingtothemanufacturer’sinstructions.Forthetruncatedforms,in thecaseofzfTMT-opsinthecodonatposition224(TAT,tyrosine) wasmutatedtoastopcodon(TAA)whileforMelanopsin (zfOpn4m2),thecodonatposition286(AAA,lysine)wasmutated toaStopcodon(TAA).Forthemutationofakeylysineinthe7thtransmembranedomainofzfTMT-opsintoanalanineresidue,the codonsequenceAAGwasmutatedtoGCG.Followingsequence analysis,theintegrityandfunctionofallopsinexpressionvectors wastestedbytransienttransfectionintoamammaliancelllineand thenWesternblottinganalysisusingananti-HA-tagspecific antibody.TestingofOpsinExpressionConstructsHepa1-6cells(1 6 105)(ATCC)weretransfectedwith2mgof eachHA-Tagopsinexpressionconstructusingthestandard Promo-Fectintransfectionprotocol(PromoKine).Twenty-four hoursaftertransfection,proteinextractswerepreparedaccording toastandardmethod[40],withsomemodifications.Cellswere solubilizedandincubatedat4 u Cinamixtureofequalvolumesof TSAbuffer(2mMTris-HCl,pH8.0,140mMNaCl,0.025% NaN3)andLysisbuffer(TSAbufferplus2%TritonX-100,5mM iodoacetamide,0.2U/mlaprotinin,and1mMphenylmethylsulfonylfluoride).After1h,0.2volumesof5%sodiumdeoxycholate wereadded,andthemixturewasincubatedonicefor10min. Thelysatewascentrifugedat2,800gfor10minat4 u C,andthe supernatantwascollectedandstoredat 2 80 u Cuntiluse.Before electrophoresisin10%or15%SDS-polyacrylamidegels,the proteinsweredilutedinLaemmlibuffer(finalconcentration 0.05MTris,pH6.8,2%SDS,100mMDTT,10%glycerol, 0.05%bromophenolblue)withoutheating.Followingtransferto nitrocellulose,themembraneswereincubatedwithahighaffinity anti-HAratmonoclonalantibody(clone3F10,Roche)according tothemanufacturer’sinstructionsandtheboundantibodywas visualizedusingtheECLdetectionsystem(AmershamBiosciences) (FigureS10).StatisticalAnalysisAlltheresultswereexpressedasmeans 6 SEM.Datawere analyzedbyone-ortwo-wayanalysisofvariance(ANOVA)to determinesignificantdifferencesusingthesoftwareGraphPad Prism4.0(GraphPadSoftwareInc.). p values , 0.05were consideredstatisticallysignificant.Toevaluatetheperiodlength ofgeneexpression,wemeasuredthetimespanbetweentwo consecutivepeaks.Atrigonometricstatisticalmodelwasappliedto evaluateperiodicphenomena.Thesinglecosinorprocedure[41] wasusedtodefinethemainrhythmicparameters(circadianperiod andpeaktime). BioluminescencedatawereanalyzedusingMicrosoftExcelor CHRONOsoftware[13,42].Periodestimatesmeasuredafter2d inDDweremadebylinearregressionfollowingpeakfinder analysiswithCHRONO.ForQ10temperaturecoefficient calculations,periodlengthestimatesforcellsheldat22 u C, 25 u C,and29 u Cwerecalculatedascyclesperhourandthen plottedagainsttemperature.Linearregressionanalysisrevealeda goodfittoastraightline(cavefishR2=0.99;zebrafishR2=0.98). Meanperiodlengthsat22 u Cand29 u Cwerethensubstitutedinto theequationQ10=(R2/R1)10/(T2 2 T1),whereRisrateandTis temperature.GenbankAccessionNumbersThesequencesreportedinthisarticlearedepositedinGenBank underaccessionnumbersGQ404475–GQ404490(seeTableS2).SupportingInformationFigureS1PeriodogramanalysisofbehavioralactivityinLD cycles. x2periodogramanalysis(confidencelevel,95%)forthe zebrafish(A)andcavefish(B)actogramsshowninFigure1.The periodogramindicatesthepercentageofvariance(%V)ofthe rhythmexplainedbyeachanalyzedperiodwithinarangeof20– 28h.Theslopeddottedlinesrepresentthethresholdof significance,setat p =0.05.Periodogramanalysisconfirmsthata behavioralactivityrhythmissynchronizedwiththe24hLDcycle inzebrafishbutnotincavefish. (TIF)FigureS2Phylogeneticanalysisofcavefishclockgenes. ComparisonofPERIOD(toppanel)andCLOCK(bottompanel) cavefishproteinswithotherpublishedteleosthomologs[36,37] usingphylogenetictreeanalysis.Thisconfirmstheclosesimilarity betweenzebrafishand P.andruzzii sequences. (TIF)FigureS3Absenceofrhythmicclockgeneexpressionincavefish brainandlarvae.QuantitativeRT-PCRanalysisofendogenous clockgeneexpressionunderLDcyclesinthewholebrainof zebrafish(A,B)andcavefish(C,D)( n =6pertimepoint)aswellas in5-d-oldzebrafishlarvae(E)and1-d-or4-wk-oldcavefishlarvae (FandG,respectively).ResultsareplottedasdescribedinFigure1. Inthecaseofcavefish,norhythmicexpressionwasdetectedeither inthebrain(C,D; p . 0.1),1-d-oldlarvaewhichstillretaineye rudiments(F, p . 0.07),oreveninlarvaeexposedfor4wktoLD cycles(G, p . 0.1). (TIF)FigureS4Rhythmicclockgeneexpressionintheliverfollowing feedingentrainment.Real-timePCRanalysisofrhythmic endogenous Clk1a (redtrace)and Per1b (greentrace)expression intheliverofzebrafish(A)and Clk1a (orangetrace)and Per1 (dark bluetrace)expressioninthecavefishliver(B).Timeisexpressedas ZTtimeorCircadianTime(CT)duringstarvation.Ineachpanel, asolid,verticalline(atZT0)indicatesthelastfeedingtime. Subsequentlyduringstarvation,theverticaldottedlines(atCT0 andCT24)denotewhenthefeedingwouldnormallyhave occurredaccordingtothepreviousregularfeedingregime.Each pointrepresentsthemean 6 SEM.Inbothspeciesrobust circadianrhythmsofclockgeneexpressionwereobserved ( p , 0.01)(seealsoFigure4). (TIF)FigureS5Effectsofdifferentdexamethasonedosestosynchronizerhythmicclockgeneexpression.Bioluminescence(cps)of cavefishcellstransfectedwiththezebrafishreporterconstruct, zfPer1b-Luc andtransientlytreatedwithdifferentdexamethasone concentrations(50nM,100nM,500nM,and1mM).Acontrol lackingdexamethasone(‘‘Control,’’blacktrace)wasalsoincluded.CavefishBlindClockRevealsPhotoreceptors PLoSBiology|www.plosbiology.org11September2011|Volume9|Issue9|e1001142

PAGE 12

All4dexamethasonetreatmentswereabletoadvanceby5 6 0.5h thephaseofapre-existingoscillationevidentinthecontrolcells. Serumtreatmentduringtheseedingofthecellsisresponsiblefor theoriginalestablishmentofthisoscillation(unpublisheddata). Theverticaldottedandsolidlinesindicatethepeaksofthecontrol anddexamethasonepulsedcells,respectively. (TIF)FigureS6Lightpulsesfailtoshiftthephaseofadexamethasone-entrainedcavefishclock.(A)FivesetsofCFcellswere transientlytransfectedwiththe zfPer1b-luc reporterandthenat8h intervals(attimepointsI–V)weretransientlytreatedwith100nM dexamethasonebeforebeingsimultaneouslyexposedtoa15min lightpulse.(B–F)Resultingbioluminescenceprofilesofthefivesets ofcells,comparedwithnon-light-pulsedcontrols.Ateachtime point,themean 6 SEMisplotted.Duringtheassay,cellswere maintainedataconstanttemperatureandinconstantdarkness.In eachpanel,coloursofthebioluminescencetracesmatchthosein theexperimentaldesign(A).Palecolouredtracesrepresent constantdarkcontrols,whiledarkcolouredtracesrepresentthe lightpulsedsetsofcells(exceptD,wherethelightpulsedtraceis shownasyellowandthedarkcontroltraceisshownasolive). Noneofthelightpulsedsetsofcellsshowasignificantdifferencein phaserelativetotheirconstantdarkcontrols( p . 0.4). (TIF)FigureS7Tissuelocalizationof Melanopsin(Opn4m2) and TMTopsin incavefish.RT-PCRresultsusingbrain,heart,muscle,fin, andCFcellRNAfortheamplificationof Melanopsin , TMT-opsin , and b -actin (ascontrol).EqualvolumesofthethreePCRreactions performedforeachtissueweresubsequentlymixedandthen productswerevisualizedbyagarosegelelectrophoresis. Melanopsin wasexpressedinallsamples,whereas TMT-opsin wasnotfoundin theheartandtheskeletalmuscle(Sk.Muscle).Anegativecontrol lackingreversetranscriptaseatthecDNAsynthesisstepwasalso included(RT( 2 )). (TIF)FigureS8Irradiancecurvesfordifferentmonochromaticlight sources.Light-emittingdiodesourceswereusedtoproducewhite light(blacktrace)(450nm , l , 700nm)andmonochromaticlight intheblue( lpeak=468nm),green( lpeak=530nm),andred ( lpeak=657nm)regionofthespectrum.Thelightintensityof eachsourcewasadjustedtoensurethataconstantnumberof photonswasemittedbyeachsource(1.42 6 10186 0.04 6 1018photons/s/m2). (TIF)FigureS9Originof Phreatichthysandruzzii .Adultcavefishwere originallycollectedinthewildintheoasisofBud-Bud (04 u 11 9 19 0 N–46 u 28 9 27 0 E)inthecentreoftheSomaliandesert duringseveralexpeditionstoAfrica(1968–1982).Ancestorsof P. andruzzii enteredthelargephreaticlayersoftheSomaliandesert thatdevelopedinEocenehorizontallimestoneformationsatthe endofthePliocene(1.4–2.6millionyearsago)andbecame isolatedwiththeextinctionofepigeansisterspeciesastheresultof extremeclimaticchanges. (TIF)FigureS10ExpressionofHA-taggedopsinsinHepa1-6cells. WesternblotanalysisofcellstransfectedwithHA-tagged expressionvectorswhichencode(A)zfOpn4m2(ca.57kDa), zfTMT-opsin(ca.45kDa),zfTMTK295A(ca.45kDa),and(B) zfTMT-opsinY224X(ca.27kDa)andzfOpn4m2K286X(ca. 34kDa).ThepositionofHA-immunoreactiveopsinbandsis indicatedbyarrowheads.Thepresenceofdoubletbandsforsome opsinsmayresultfromthedenaturingconditionsusedtoprepare the7-transmembranedomainproteinextractsforSDSPAGE analysis.Proteinextractspreparedfromcellstransfectedwiththe emptyexpressionvectorwereloadedasnegativecontrols. (TIF)TableS1PCRprimers.Summaryofthesequencesofforward (F)andreverse(R)PCRprimersusedtoamplifytheinitial P. andruzzii cDNAfragments(basedonevolutionaryconserved D. rerio nucleotidesequences)aswellasthecavefishandzebrafish sequenceprimersusedforquantitativePCRanalysis. (DOC)TableS2P.andruzzii circadianclockcDNAs.Summaryofthe Genbankaccessionnumbersforeachofthe P.andruzzii clockrelatedcDNAsclonedandsequenced.ForeachcDNA,the percentageofaminoacidsimilaritywiththezebrafish( D.rerio ) homologsisalsoindicated. (DOC)AcknowledgmentsWethankElenaKage,AndreaMargutti,andPietroFarinelliforexcellent technicalassistance;SauloBambi(MuseumofNaturalHistoryofthe UniversityofFlorence)forthephotographs;FerencMu ¨ller,RalfDahm, RobertJ.Lucas,ThomasDickmeis,OlivierKassel,andYoavGothilffor criticallyreadingthemanuscriptandmanyveryconstructivediscussions;as wellasPhilippMracek,MariaLauraIdda,andLuisaM.Veraforhelpful comments.WethankOlivierKasselforthegiftofthemodified pcDNA3.1( + )HAtagexpressionvector.WealsothankPaulSpiringand summerpracticalstudentsfromtheEuropeanSchoolKarlsruhefor technicalhelp.AuthorContributionsTheauthor(s)havemadethefollowingdeclarationsabouttheir contributions:Conceivedanddesignedtheexperiments:CBNSFFJSV AFNC.Performedtheexperiments:NCEFDVNFJFLORB.Analyzed thedata:CBDVNCEFJFLO.Contributedreagents/materials/analysis tools:CBNSFRBAFFJSV.Wrotethepaper:NSFCBNCDVFJSV.References1.BailesHJ,LucasRJ(2010)Melanopsinandinnerretinalphotoreception.Cell MolLifeSci67:99–111. 2.BertolucciC,FoaA(2004)Extraocularphotoreceptionandcircadian entrainmentinnonmammalianvertebrates.ChronobiolInt21:501–519. 3.Wager-SmithK,KaySA(2000)Circadianrhythmgenetics:fromfliestomiceto humans.NatGenet26:23–27. 4.TamaiTK,YoungLC,WhitmoreD(2007)Lightsignalingtothezebrafish circadianclockbyCryptochrome1a.ProcNatlAcadSciUSA104: 14712–14717. 5.VatineG,ValloneD,AppelbaumL,MracekP,Ben-MosheZ,etal.(2009)Light directszebrafishperiod2expressionviaconservedDandEboxes.PLoSBiol7: e1000223.doi:10.1371/journal.pbio.1000223. 6.WhitmoreD,FoulkesNS,Sassone-CorsiP(2000)Lightactsdirectlyonorgans andcellsinculturetosetthevertebratecircadianclock.Nature404:87–91. 7.HirayamaJ,ChoS,Sassone-CorsiP(2007)Circadiancontrolbythereduction/ oxidationpathway:catalaserepresseslight-dependentclockgeneexpressionin thezebrafish.ProcNatlAcadSciUSA104:15747–15752. 8.KobayashiY,IshikawaT,HirayamaJ,DaiyasuH,KanaiS,etal.(2000) Molecularanalysisofzebrafishphotolyase/cryptochromefamily:twotypesof cryptochromespresentinzebrafish.GenesCells5:725–738. 9.MoutsakiP,WhitmoreD,BellinghamJ,SakamotoK,David-GrayZK,etal. (2003)Teleostmultipletissue(tmt)opsin:acandidatephotopigmentregulating theperipheralclocksofzebrafish?BrainResMolBrainRes112:135–145. 10.ColliL,PagliantiA,BertiR,GandolfiG,TagliaviniJ(2009)Molecular phylogenyoftheblindcavefish Phreatichthysandruzzii and Garrabarreimiae within thefamilyCyprinidae.EnvironBiolFish84:95–107. 11.JefferyWR(2009)Regressiveevolutionin Astyanax cavefish.AnnuRevGenet43: 25–47.CavefishBlindClockRevealsPhotoreceptors PLoSBiology|www.plosbiology.org12September2011|Volume9|Issue9|e1001142

PAGE 13

12.BertiR,DurandJP,BecchiS,BrizziR,KellerN,etal.(2001)Eyedegeneration intheblindcave-dwellingfish Phreatichthysandruzzii .CanJZool79:1278–1285. 13.ValloneD,GondiSB,WhitmoreD,FoulkesNS(2004)E-boxfunctionina periodgenerepressedbylight.ProcNatlAcadSciUSA101:4106–4111. 14.WhitmoreD,FoulkesNS,StrahleU,Sassone-CorsiP(1998)Zebrafishclock rhythmicexpressionrevealsindependentperipheralcircadianoscillators.Nat Neurosci1:701–707. 15.DickmeisT,LahiriK,NicaG,ValloneD,SantorielloC,etal.(2007) Glucocorticoidsplayakeyroleincircadiancellcyclerhythms.PLoSBiol5:e78. doi:10.1371/journal.pbio.0050078. 16.PandoMP,PinchakAB,CermakianN,Sassone-CorsiP(2001)Acell-based systemthatrecapitulatesthedynamiclight-dependentregulationofthe vertebrateclock.ProcNatlAcadSciUSA98:10178–10183. 17.Lopez-OlmedaJF,Sanchez-VazquezFJ(2010)Feedingrhythmsinfish:from behavioraltomolecularapproach.In:KulczykowskaE,PopekW,KapoorBG, eds.Biologicalclocksinfish.EnfieldNH:SciencePublishers.pp155–184. 18.StephanFK(2002)The‘‘other’’circadiansystem:foodasaZeitgeber.JBiol Rhythms17:284–292. 19.Lopez-OlmedaJF,TartaglioneEV,delaIglesiaHO,Sanchez-VazquezFJ (2010)Feedingentrainmentoffood-anticipatoryactivityandper1expressionin thebrainandliverofzebrafishunderdifferentlightingandfeedingconditions. ChronobiolInt27:1380–1400. 20.BalsalobreA,MarcacciL,SchiblerU(2000)Multiplesignalingpathwayselicit circadiangeneexpressioninculturedRat-1fibroblasts.CurrBiol10: 1291–1294. 21.PittendrighCS(1954)Ontemperatureindependenceintheclocksystem controllingemergencetimeinDrosophila.ProcNatlAcadSciUSA40: 1018–1029. 22.LahiriK,ValloneD,GondiSB,SantorielloC,DickmeisT,etal.(2005) Temperatureregulatestranscriptioninthezebrafishcircadianclock.PLoSBiol 3:e351.doi:10.1371/journal.pbio.0030351. 23.CarrAJ,WhitmoreD(2005)Imagingofsinglelight-responsiveclockcellsreveals fluctuatingfree-runningperiods.NatCellBiol7:319–321. 24.BellinghamJ,WhitmoreD,PhilpAR,WellsDJ,FosterRG(2002)Zebrafish melanopsin:isolation,tissuelocalisationandphylogeneticposition.BrainRes MolBrainRes107:128–136. 25.ProvencioI,JiangG,DeGripWJ,HayesWP,RollagMD(1998)Melanopsin: anopsininmelanophores,brain,andeye.ProcNatlAcadSciUSA95: 340–345. 26.BellinghamJ,ChaurasiaSS,MelyanZ,LiuC,CameronMA,etal.(2006) Evolutionofmelanopsinphotoreceptors:discoveryandcharacterizationofa newmelanopsininnonmammalianvertebrates.PLoSBiol4:e254.doi:10.1371/ journal.pbio.0040254. 27.MenonST,HanM,SakmarTP(2001)Rhodopsin:structuralbasisofmolecular physiology.PhysiolRev81:1659–1688. 28.NewmanLA,WalkerMT,BrownRL,CroninTW,RobinsonPR(2003) Melanopsinformsafunctionalshort-wavelengthphotopigment.Biochemistry 42:12734–12738. 29.LuW,MengQJ,TylerNJ,StokkanKA,LoudonAS(2010)Acircadianclockis notrequiredinanarcticmammal.CurrBiol20:533–537. 30.KanekoM,Hernandez-BorsettiN,CahillGM(2006)Diversityofzebrafish peripheraloscillatorsrevealedbyluciferasereporting.ProcNatlAcadSciUSA 103:14614–14619. 31.EscobarC,CailottoC,Angeles-CastellanosM,DelgadoRS,BuijsRM(2009) Peripheraloscillators:thedrivingforceforfood-anticipatoryactivity. EurJNeurosci30:1665–1675. 32.StorchKF,WeitzCJ(2009)Dailyrhythmsoffood-anticipatorybehavioral activitydonotrequiretheknowncircadianclock.ProcNatlAcadSciUSA106: 6808–6813. 33.RoennebergT,FosterRG(1997)Twilighttimes:lightandthecircadiansystem. PhotochemPhotobiol66:549–561. 34.ValloneD,SantorielloC,GondiSB,FoulkesNS(2007)Basicprotocolsfor zebrafishcelllines:maintenanceandtransfection.MethodsMolBiol362: 429–441. 35.KwokC,KornRM,DavisME,BurtDW,CritcherR,etal.(1998) Characterizationofwholegenomeradiationhybridmappingresourcesfor non-mammalianvertebrates.NucleicAcidsRes26:3562–3566. 36.WangH(2008)Comparativeanalysisofperiodgenesinteleostfishgenomes. JMolEvol67:29–40. 37.WangH(2008)Comparativeanalysisofteleostfishgenomesreveals preservationofdifferentancientclockduplicatesindifferentfishes.Mar Genomics1:69–78. 38.TamuraK,DudleyJ,NeiM,KumarS(2007)MEGA4:MolecularEvolutionary GeneticsAnalysis(MEGA)softwareversion4.0.MolBiolEvol24:1596–1599. 39.KasselO,SchneiderS,HeilbockC,LitfinM,GottlicherM,etal.(2004)A nuclearisoformofthefocaladhesionLIM-domainproteinTrip6integrates activatingandrepressingsignalsatAP-1-andNF-kappaB-regulatedpromoters. GenesDev18:2518–2528. 40.WaksmanY,OlsonJM,CarlisleSJ,CabralGA(1999)Thecentralcannabinoid receptor(CB1)mediatesinhibitionofnitricoxideproductionbyratmicroglial cells.JPharmacolExpTher288:1357–1366. 41.NelsonW,TongYL,LeeJK,HalbergF(1979)Methodsforcosinorrhythmometry.Chronobiologia6:305–323. 42.RoennebergT,TaylorW(2000)Automatedrecordingsofbioluminescencewith specialreferencetotheanalysisofcircadianrhythms.MethodsEnzymol305: 104–119.CavefishBlindClockRevealsPhotoreceptors PLoSBiology|www.plosbiology.org13September2011|Volume9|Issue9|e1001142


printinsert_linkshareget_appmore_horiz

Download Options

close


  • info Info

    There are only PDFs associated with this resource.

  • link PDF(s)



Cite this item close

APA

Cras ut cursus ante, a fringilla nunc. Mauris lorem nunc, cursus sit amet enim ac, vehicula vestibulum mi. Mauris viverra nisl vel enim faucibus porta. Praesent sit amet ornare diam, non finibus nulla.

MLA

Cras efficitur magna et sapien varius, luctus ullamcorper dolor convallis. Orci varius natoque penatibus et magnis dis parturient montes, nascetur ridiculus mus. Fusce sit amet justo ut erat laoreet congue sed a ante.

CHICAGO

Phasellus ornare in augue eu imperdiet. Donec malesuada sapien ante, at vehicula orci tempor molestie. Proin vitae urna elit. Pellentesque vitae nisi et diam euismod malesuada aliquet non erat.

WIKIPEDIA

Nunc fringilla dolor ut dictum placerat. Proin ac neque rutrum, consectetur ligula id, laoreet ligula. Nulla lorem massa, consectetur vitae consequat in, lobortis at dolor. Nunc sed leo odio.