<%BANNER%>

UFIR



An unbiased approach to identify genes involved in development in a turtle with temperature-dependent sex determination
http://www.biomedcentral.com ( Publisher's URL )
CITATION DOWNLOADS PDF VIEWER
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/IR00001162/00001
 Material Information
Title: An unbiased approach to identify genes involved in development in a turtle with temperature-dependent sex determination
Series Title: BMC Genomics 2012, 13:308
Physical Description: Journal Article
Creator: Chojnowski, Jena L.
Braun, Edward L.
Publisher: BioMed Central
Place of Publication: BMC Genomics
Publication Date: 15 July 2012
 Notes
Abstract: Background Many reptiles exhibit temperature-dependent sex determination (TSD). The initial cue in TSD is incubation temperature, unlike genotypic sex determination (GSD) where it is determined by the presence of specific alleles (or genetic loci). We used patterns of gene expression to identify candidates for genes with a role in TSD and other developmental processes without making a priori assumptions about the identity of these genes (ortholog-based approach). We identified genes with sexually dimorphic mRNA accumulation during the temperature sensitive period of development in the Red-eared slider turtle (Trachemys scripta), a turtle with TSD. Genes with differential mRNA accumulation in response to estrogen (estradiol-17β; E2) exposure and developmental stages were also identified. Results Sequencing 767 clones from three suppression-subtractive hybridization libraries yielded a total of 581 unique sequences. Screening a macroarray with a subset of those sequences revealed a total of 26 genes that exhibited differential mRNA accumulation: 16 female biased and 10 male biased. Additional analyses revealed that C16ORF62 (an unknown gene) and MALAT1 (a long noncoding RNA) exhibited increased mRNA accumulation at the male producing temperature relative to the female producing temperature during embryonic sexual development. Finally, we identified four genes (C16ORF62, CCT3, MMP2, and NFIB) that exhibited a stage effect and five genes (C16ORF62, CCT3, MMP2, NFIB and NOTCH2) showed a response to E2 exposure. Conclusions Here we report a survey of genes identified using patterns of mRNA accumulation during embryonic development in a turtle with TSD. Many previous studies have focused on examining the turtle orthologs of genes involved in mammalian development. Although valuable, the limitations of this approach are exemplified by our identification of two genes (MALAT1 and C16ORF62) that are sexually dimorphic during embryonic development. MALAT1 is a noncoding RNA that has not been implicated in sexual differentiation in other vertebrates and C16ORF62 has an unknown function. Our results revealed genes that are candidates for having roles in turtle embryonic development, including TSD, and highlight the need to expand our search parameters beyond protein-coding genes.
Funding: This work was facilitated by National Science Foundation grants to ELB and collaborators (DEB-0228682 and DUE-0920151) as well as an American Society of Ichthyology and Herpetology grant and a Sigma Xi grant to JLC. Publication of this article was funded in part by the University of Florida Open-Access Publishing Fund.
Acquisition: Collected for University of Florida's Institutional Repository by the UFIR Self-Submittal tool. Submitted by Edward Braun.
 Record Information
Source Institution: University of Florida Institutional Repository
Holding Location: University of Florida
Rights Management: All rights reserved by the submitter.
System ID: IR00001162:00001

Downloads
Full Text

PAGE 1

RESEARCHARTICLEOpenAccessAnunbiasedapproachtoidentifygenesinvolved indevelopmentinaturtlewith temperature-dependentsexdeterminationJenaLChojnowski1,2*andEdwardLBraun2AbstractBackground: Manyreptilesexhibittemperature-dependentsexdetermination(TSD).TheinitialcueinTSDis incubationtemperature,unlikegenotypicsexdetermination(GSD)whereitisdeterminedbythepresenceof specificalleles(orgeneticloci).Weusedpatternsofgeneexpressiontoidentifycandidatesforgeneswitharolein TSDandotherdevelopmentalprocesseswithoutmaking apriori assumptionsabouttheidentityofthesegenes (ortholog-basedapproach).WeidentifiedgeneswithsexuallydimorphicmRNAaccumulationduringthe temperaturesensitiveperiodofdevelopmentintheRed-earedsliderturtle( Trachemysscripta ),aturtlewithTSD. GeneswithdifferentialmRNAaccumulationinresponsetoestrogen(estradiol-17 ;E2)exposureanddevelopmental stageswerealsoidentified. Results: Sequencing767clonesfromthreesuppression-subtractivehybridizationlibrariesyieldedatotalof581 uniquesequences.Screeningamacroarraywithasubsetofthosesequencesrevealedatotalof26genesthat exhibiteddifferentialmRNAaccumulation:16femalebiasedand10malebiased.Additionalanalysesrevealedthat C16ORF62 (anunknowngene)and MALAT1 (alongnoncodingRNA)exhibitedincreasedmRNAaccumulationatthe maleproducingtemperaturerelativetothefemaleproducingtemperatureduringembryonicsexualdevelopment. Finally,weidentifiedfourgenes( C16ORF62,CCT3,MMP2 ,and NFIB )thatexhibitedastageeffectandfivegenes ( C16ORF62,CCT3,MMP2,NFIB and NOTCH2 )showedaresponsetoE2exposure. Conclusions: HerewereportasurveyofgenesidentifiedusingpatternsofmRNAaccumulationduringembryonic developmentinaturtlewithTSD.Manypreviousstudieshavefocusedonexaminingtheturtleorthologsofgenes involvedinmammaliandevelopment.Althoughvaluable,thelimitationsofthisapproachareexemplifiedbyour identificationoftwogenes( MALAT1 and C16ORF62 )thataresexuallydimorphicduringembryonicdevelopment. MALAT1 isanoncodingRNAthathasnotbeenimplicatedinsexualdifferentiationinothervertebratesand C16ORF62 hasanunknownfunction.Ourresultsrevealedgenesthatarecandidatesforhavingrolesinturtle embryonicdevelopment,includingTSD,andhighlighttheneedtoexpandoursearchparametersbeyond protein-codinggenes.BackgroundTurtleshavebeencharacterizedforasmallnumberof fascinatingdifferencesfromthebetter-studiedgroupsof amniotes,liketheregulationofsexualdevelopmentby temperatureandthepresenceofthecarapace.However, littleisknownaboutgenesinvolvedinthemanyaspects ofturtledevelopment.Thedevelopmentoftheeye[1,2], brain[3,4],carapace[reviewedin[5]],andgonads (Table1);andtheroleofspecifichormones[reviewedin [6]]haveallbeenbrieflystudiedbutthesestudieshave focusedontheorthologsofgenesalreadyknowntohave aroleinhumanandmousedevelopment.Althoughthe useofstudiesfocusedontheturtleorthologsofgenes identifiedinotherorganismsareclearlyimportant, screensforcandidategenesthatavoidmaking apriori assumptionsrepresentacomplementaryapproachwith excellentpotentialtorevealnoveldevelopmentalgenes. *Correspondence: kixs4@uga.edu1GeneticsDepartment,UniversityofGeorgia,500DWBrooksDr.,Coverdell CenterRm270,Athens,GA30602,USA2DepartmentofBiology,UniversityofFlorida,POBox118525,Gainesville, FL32607,USA 2012ChojnowskiandBraun;licenseeBioMedCentralLtd.ThisisanOpenAccessarticledistributedunderthetermsofthe CreativeCommonsAttributionLicense(http://creativecommons.org/licenses/by/2.0),whichpermitsunrestricteduse, distribution,andreproductioninanymedium,providedtheoriginalworkisproperlycited.ChojnowskiandBraun BMCGenomics 2012, 13 :308 http://www.biomedcentral.com/1471-2164/13/308

PAGE 2

Thisstudyusessuchanapproachtoidentifygeneswith sexuallydimorphicexpression;thissetofgenesis enrichedforgenesinvolvedintemperature-dependent sexdetermination(TSD),butwillalsoincludegenesin otheraspectsofsexualdifferentiationandgeneraldevelopmentalprocessesrelatedtotemperatureandhormone regulationthatmightnotbedirectlyrelatedtosexual differentiation. Manyreptiliantaxa,includingthemajorityofturtle species,exhibitTSD[20,21].Incubationtemperatureis theinitialcueforsexualdevelopmentinTSD,incontrasttogeneticsexdetermination(GSD)thatisevident inanumberofvertebrategroupssuchasamphibians, snakes,birds,andmammals[22].GSDisbestcharacterizedintherianmammalsandisinitiatedbythe SRY gene,locatedontheYchromosome,whichcauses organismsexpressingthegenetodevelopasmales[23]. SRY orthologshavenotbeenidentifiedinothergroups ofvertebratesregardlessofwhethertheyexhibitTSDor GSD,suggestingthat SRY isaninnovationuniqueto therianmammals.Infact,onlyoneotherunique “ trigger ” geneforsexualdevelopmenthasbeenidentifiedinavertebratetaxon,afishwithGSD(medaka;see[24,25]).Itis unclearwhetheratriggergeneexistsinorganismsthat exhibitTSD,sincethereareseveralmodelsthatcanexplainTSD.Forexample,TSDmayreflectregulationofa triggergene(orsetoftriggergenes)byincubation temperature,itmayreflecttheimpactoftemperature upontheactivityofspecificenzymesthathavearolein signaling,oritmayreflectacombinationofbothphenomena[9,12,13].Regardless,itisclearbaseduponthe studiesinorganismswithknowntriggersthatagene homologoustoaknowntriggerinotherorganismsdoes notregulateTSDinturtles[26]. Althoughtriggergene(s)arenotconserved,ifoneor moreareevenpresent,anumberofgenesinvolvedin gonadaldifferentiationandotheraspectsofsexualdevelopmentareconservedamongvertebrates,including organismswithdifferentsexdeterminingsystems[27,28]. Anumberoforthologsofgenesfirstidentifiedinmammalshavebeenidentifiedandcharacterizedindifferent vertebrategroups[29-33],includingturtles(Table1). Studiesfocusedonorthologsofgenesknowntoplaya roleinmammaliansexualdevelopmenthaveprovided valuableinformation,althoughthisapproachhaslimits forbroaderinvestigations.OnesetbacktoanorthologapproachisthatthecompletesetofgenesinvolvedinmammalianGSDremainsunknownandtherefore,itis restrictinginfindingnovelgenesorpathwaysspecifictoa newsystem.Acomplementaryapproachistoidentifycandidategenesusingpatternsofexpressionratherthan orthology.Thisraisesthequestionofthemostappropriate tissuetoassaygeneexpression,sincesexuallydimorphic geneexpressionhasbeennotedinmultipletissues[13,3436].Thus,focusingonspecifictissuesmayresultinthe omissionofcriticalcandidategenes.Theuseofwhole embryosisanunbiasedstrategyforfindingcandidate genesforsexdetermination,butitalsoallowsforthe identificationofgenesinotherpathwaysthatoccurwithin thesametimepoints.Also,temperatureandhormonesdo notexclusivelyaffectsexdetermination,andtherefore novelgenescanalsobeidentifiedthatarerelatedto temperatureandtheeffectsofhormonalexposure. Thegoalofthisstudyistousethered-earedslider turtle( Trachemysscripta ),aturtlewithTSD,toidentify candidategeneswithintheembryonicdevelopment pathway.Genesthatexhibitedasexuallydimorphicexpressionpatternduringthetemperature-sensitiveperiod (TSP),whichcontainsthecriticalstageforcommitment toaspecificsex[37],weretargeted.Theapproachwe usedwillalsorevealgenesthatexhibitadifferential expressionpatternduetotemperatureorhormonaldifferencesthatarenotdirectlyrelatedtosexualdevelopment.Toaccomplishthissearchweidentifiedgenesthat showincreasedmRNAaccumulationundereitherthe maleorfemaleproducingtemperaturesaswellasgenes thatexhibitresponsetoestrogenexposure.Sincewealso examinedembryosatvarioustimesduringtheTSPit wasalsopossibletoidentifystageeffectsuponmRNA accumulationduringtheTSP.Thesedifferentsearches allowforabroaderinvestigationfornotonlysexually specificfactorsbutalsoforgeneraldevelopmentalgenes previouslyunknownwithinthetimepointsoftheTSP. Toaccomplishthis,weproducedthreesubtractionlibraries.Twooftheselibrarieswereenrichedforgenesthat showhighermRNAaccumulationduringtheTSPinone specifictemperatureregime(i. e.,genesthatexhibitgreater mRNAaccumulationatthefemal e-producingtemperature [31C]thanatthemale-producingtemperature[26C]and viceversa).Thethirdlibrarywasenrichedforgenesthat Table1Ageneraloverviewofsexuallydimorphicgene expressioninturtleswithTSDstage17lateinTSP GeneTestisOvaryTestisOvaryTestisOvaryReference SF1 +++[ 7 8 ] WT1 samesamesamesamesamesame[ 9 11 ] DAX1 samesamesamesamesamesame[ 12 ] SOX9 samesame++[ 11 13 ] DMRT1 +++[ 14 16 ] CYP19 samesame++[ 8 15 17 ] SOX8 samesamesamesamesamesame[ 18 ] FOXL2 samesame++[ 12 ] MIS samesame++[ 8 18 ] R-SPONDIN samesame++[ 19 ] WNT4 samesamesamesame+[ 12 ] ChojnowskiandBraun BMCGenomics 2012, 13 :308 Page2of12 http://www.biomedcentral.com/1471-2164/13/308

PAGE 3

showincreasedmRNAaccumulationduringtheTSPin embryosproducedatthemale-producingtemperatureafter treatmentwithexogenousestr ogen,ultimatelyproducing thefemalephenotype.SubsetsofthecDNAsfromthese librarieswereexaminedmorethoroughlybymacroarray hybridizationandsemi-quantitativePCRorquantitative real-timePCR.Thisapproachhasthepotentialtoidentify developmentalgenesthatexhibitdifferentialexpression whenred-earedsliderturtlesareexposedtodifferenttemperaturesandhormonalconditionsduringtheTSPwithoutmaking apriori assumptionsabouttheidentityofthe genes.ResultsanddiscussionSuppressionsubtractionhybridization(SSH)librariesSSHwasusedtoconstructlibrariesenrichedforcDNAs thatcorrespondtomRNAsthatexhibitdifferentlevels ofaccumulationduringtheTSP.ThreesubtractedcDNA librarieswereconstructed:oneenrichedformRNAsthat accumulateathigherlevelsatthefemale-producing temperaturethanthemale-producingtemperature(hereaftercalledthe “ femalelibrary ” );anotherenrichedfor mRNAsthataccumulateathigherlevelsatthemaleproducingtemperaturethanthefemale-producing temperature(hereaftercalledthe “ malelibrary ” );anda thirdenrichedformRNAsthataccumulateathigher levelsatthemale-producingtemperaturewithexogenous estradiol-17 (sufficientforsexreversal)thaninsimilar embryostreatedwiththevehiclealone(hereaftercalled the “ E2library ” ).Atotalof767sequenceswereobtained andwerepreviouslydepositedindbEST(FG341000: FG341832).TheSSHexpressedsequencetags(ESTs)were processedasdescribed[38, 39],yieldingatotalof581 contigsandsingletons(unigenes)afterassemblyusing CAP3[40].Theresultsforthehomologysearches canbefoundinAdditionalfile1:Homologysearchfor subtractionlibraries. SSHlibrariestypicallycontainsomehousekeeping genes[41-43]sinceitisdifficulttocompletelyeliminate genesthatdonotexhibitdifferentialexpressionbetween thetwoexperimentalconditions.ThequalityofSSHlibrariescanbeassessedbyexaminingtheproportionof broadlyexpressedgenes,althoughthemostappropriate setsofgenestoviewas “ housekeeping ” canbeproblematic todefineprecisely[44].However,analysesofGC-content [39]providealineofevidencethattheproportionof cDNAsthatcorrespondtohousekeepinggenetranscriptsis greatlyreducedinourSSHlibraries;housekeepinggenes tendtohaveahigherGC-contentthangenesthatexhibit lowerlevelsofexpression[45,46]anditprovidesan efficientandpracticalmethodtodefinehousekeeping genesthatavoidsconflictsamongtheavailablelistsof housekeepinggenes.WefoundthattheGC-content oftheturtletranscriptswaslowerthanexpectedfor otherreptilianESTefforts[39].Thus,theSSHmethoddid appeartoenrichforgeneswi thlowerlevelsofmRNAaccumulationdespitebeingunable,asexpected,toeliminateall housekeepinggenecDNAs.GenesfoundintheSSHlibrariesGeneMergewasusedtotestforover-representedGO (GeneOntology)termssignifyingbiologicalprocesses fromgeneswithhumanhomologsfoundinallthreeSSH libraries[47](Additionalfile2:GeneMergeforbiological processes).Atotalof34over-representedGOtermswere significant(p < 0.05)andtheyrepresentabroadrangeof biologicalprocesses.Afewumbrellacategoriesthat includeanumberofover-representedGOtermsare anatomicalstructuremorphogenesis(GO:0009653; includesfacemorphogenesis[GO:0060325]andskeletal systemmorphogenesis[GO:0048705]),cellularprocessing (GO:0009987;includesribosomalsmallsubunitbiogenesis[GO:0042274],Tcelldifferentiationinthethymus [GO:0033077],cellularmembraneorganization[GO: 0016044],DNApackaging[GO:0006323],regulationof cellcycle[GO:0051726],negativeregulationofapoptosis [GO:0043066]),andmetabolicprocessing(GO:0044267; includestranslation[GO:0006412],transcription[GO: 0006350],proteinfolding[GO:0006457],translationalinitiation[GO:0006413],andtranslationalelongation[GO: 0006414]).Thesecategoriesshowthatthegenesfoundin theSSHlibrariesinvolveactivecelldifferentiation andprocessing,asexpectedformRNAsexpressedin developingembryos. Inaddition,genesthathavehumanhomologsfrom theSSHlibrarieswereclusteredintofunctionallyrelated groupswithinthesubsetofbiologicalprocessesbythe DAVIDtoolforfunctionalannotationclusteringwith highstringency[48,49](Additionalfile3:DAVIDfunctionalannotationclustering).Thedifferentgroupsrepresentthediversityofthelibraries ’ genes.Apertinent clustertoourstudythatemergedisoneenrichedfor genesinvolvedindevelopmentalprocesses(Figure1). Thoughthesegenesareidentifiedasbeingassociated withhumandevelopmentalprocesses,thisstudyoffersa chancetodetermineiftheyhavebeenco-optedforsimilarfunctionsintheturtle.Oneofthegenesfromthis cluster MMP2 ,starredinFigure1,isofparticularinterestbecauseitisoneofthefirst MIS (Mllerianinhibitingsubstance)-targetgenesinvolvedinMllerianduct regressionandisinvolvedinthebreakdownofextracellularmatrixinnormalphysiologicalprocesses,suchas embryonicdevelopment,reproduction,andtissueremodelling[50]. MMP2 ’ sinvolvementinmammaliandevelopmentleadsustobelieveithaspotentialtobea candidategeneforrelevanttotheaspectsofturtledevelopmentthatrepresentedthefocusofthisstudy,potentiallyincludingTSD.ChojnowskiandBraun BMCGenomics 2012, 13 :308 Page3of12 http://www.biomedcentral.com/1471-2164/13/308

PAGE 4

Anumberofdistinctgenes(7)intheSSHlibrariesencodetemperatureresponsiveproteinsorregulatory genesinvolvedintheheatshockresponse.Ten temperatureresponsivecDNAswerefoundinthefemale library(twoofwhichexhibitedwithin-libraryredundancy)whereasonlyoneofeachwasfoundinthemale andE2libraries(Table2).Sincethefemalelibrarywas enrichedforgenesexpressedatatemperature5C higherthaneitherthemaleortheE2librariesthelarger numberofheatshockcDNAscouldsimplyreflecta temperatureeffect.However,specifictemperatureresponsivemRNAsaccumulatedifferentiallyduringgonadaldifferentiationinanotherreptilewithTSD ( Alligatormississippiensis ;[51]).Furthermore,specific heatshockproteinsplayacriticalroleinthetranscriptionalcomplexofsteroidhormonereceptorsandtheir correspondingchaperonesandcofactors[52].Giventhat temperatureistheinitialsignalinTSD,temperatureresponsivegenesrepresentgoodcandidatesforinvolvementintheTSDcascade.DifferentialexpressionrevealedbymacroarrayanalysesAmacroarrayassaywasusedtorefinethesetofgenes identifiedbysequencingtheSSHlibrariesforsexual dimorphismandplaceouranalysisoftranscriptaccumulationunderdifferentexperimentalconditionsina quantitativeframework(Figure2).Atotalof26signals weredetectedashavingdifferentialexpressionpatterns: 16femalebiasedsignalsand10malebiasedsignals.However,thedegreeofdifferentialexpressionrevealedbythe macroarrayanalyseswastypically < 2-fold.Thus,our macroarrayanalyseswereabletoshowthatanumberof cDNAspresentintheSSHlibrariesdoexhibitsexualdimorphicpatternsofexpressionundertheconditionswe tested,althoughthedifferencesintheamountofmRNA presentwastypicallylimited.Sinceourexperimentswere conductedonwholeembryostoensureunbiasedcandidategeneidentification,itmightbethecasethatsome genesidentifiedexhibitstrongersexualdimorphismina specifictissueorasubsetoftissues(e.g.brain[34,35],liver [36],andgonad[13]).However,ourresultsindicatethat sexuallydimorphicgeneexpressionisdetectableatthe wholeembryolevel,indicatingthattheunbiasedapproach isfeasible. Genesthatarefoundtobesexuallydimorphicarenot automaticallyconsideredtobepartofTSDsincedifferingtemperaturesandhormonescanaffectmorethan justsexualdevelopment.Thegenesthatemergedfrom themacroarrayasbeingdifferentiallyexpressedhavea mixtureofbiologicalrolesinhumansbasedonDAVID. Somegenesoverlapintheirbiologicalroleswhileothers havemoredistinctroles.Forexample,10genes ( GTPBP4,HSP90AA1,ARID4A,RAN,HBZ,SERPINA3, BRIP1,NFE2L1,CDK6 ,and NFIB )areinvolvedinthe regulationofmetabolicprocessingand6genes ( GTPBP4,BRIP1,RAN,KATNA1,CDK6 ,and NFIB )are involvedincelldivisionandproliferation(5ofthe6in thelatercategoriesarealsofoundintheearliercategory ofregulationofmetabolicprocessing).Moreover, AFP and LAPTM4A areinvolvedinreproductionandtransport,respectively.Thoughthesegenesarenotdirectly relatedtohumansexualdevelopment,theyprovideus withaclueastowhichtypesofprocessesoccurduring turtledevelopmentunderdifferenttemperaturesand hormoneexposures. Table2TemperatureresponsivegenesfoundinSSHNameLibrarywherefoundRedundancywithinlibrary HSPA8 Male1 HSP90B1 E21 CIRBP Female2 HSBP1 Female3 HSP90AA1 Female1 HSPD1 Female1 SERPINH1 Female1 Figure1 DevelopmentClusterfromDAVIDfunctionalannotationclusteringwithhighstringency. Functionalclusterofdevelopmental genesandGOtermsfromtheresultingknowngenesfromtheSSHlibrarieswithhighstringency. ChojnowskiandBraun BMCGenomics 2012, 13 :308 Page4of12 http://www.biomedcentral.com/1471-2164/13/308

PAGE 5

Semiquantitative(semiQ)PCRvalidationofTSD candidategenesFurtherexperimentsforvalidationwereconductedona subsetofcandidategenesselectedfromboththesubtractionlibraryandmacroarrayanalyses.Thesegenes werechosenbecausetheywereimplicatedinmammaliansexualdevelopment( MMP2 :[53]),mammaliandevelopment( CCT3,NFIB, and NOTCH2 [54-56]),orhave anunknownfunction( C16ORF62 ).Thefivecandidate geneswereusedinthreedifferentcategoriesforsemiQPCR:sexualdimorphicexpressionatstage17,differencesbetweenstages14and17,andanE2timetrial conductedduringstage14includingfast(6hours)and slow(24hours)responses(Figure3). Onlytheturtleorthologof C16ORF62 ,ageneofunknownfunction,showedevidenceofsexuallydimorphicexpressionatstage17.Thetranscriptofthisgeneshowed greateraccumulationinmalesthaninfemales(~2.5foldincrease),thesametrendthatwasevidentinthemacroarray results.Itisconservedinmammals,birds,fish,insects, nematodes,andplants,itsGOtermisIntegraltomembrane,anditisfoundinawidevarietyofadultandembryonictissuesinmammals[57,58].Since C16ORF62 isagene ofunknownfunctionanditwasfoundtobesexuallydimorphicinturtleembryosduringtheTSP,itrepresents trulyanovelcandidateforageneinvolvedinturtledevelopment,potentiallyincludingTSD;andpotentiallyinsexualdevelopmentinothervertebratesaswell. Fourgenesexhibitedincreasedaccumulationduring stage17relativetostage14inembryosincubatedatthe maleproducingtemperature(26C).Thisstageeffectwas evidentfor C16ORF62,CCT3,MMP2 ,and NFIB (Figure3B);themoststrikingisa~8-foldincreasein mRNAaccumulationbetweenstage14and17for C16ORF62 .Theothersshowedarangeofrelativeincrease inmRNAaccumulationof1.2-fo ldto4.8-foldwithstage progression. CCT3,MMP2 ,and NFIB havebeenimplicated ingonaddevelopment( MMP2 :[59])orotheraspectsofdevelopment( NFIB and CCT3 :[54,55])inmammals.When thisinformationiscombinedw ithourobservationthatturtleorthologsexhibitedinc reasedmRNAaccumulationas developmentproceededfromstage14tostage17(earlyin TSP)itisreasonabletospeculatethatthesegenesplaya roleinturtledevelopment,potentiallysexualdevelopment inthecaseof MMP2 .Though NOTCH2 doesnotshowa stageeffectforturtleithasbeenpreviouslyseenasadevelopmentalgeneinneuronaldevelopmentofamammaland mighthaveamoresignificantaffectonturtledevelopment duringdifferentstagesorthestagedifferencesaretoolow toidentifywiththisstudy[56]. Allfivegenesshowarapid(6hours)responsetoE2exposure. C16ORF62,CCT3, and MMP2 allshowa downregulationofmRNAexpressionand NFIB and NOTCH2 showanupregulation.Fourofthegenes ( CCT3,MMP2 NFIB ,and NOTCH2 )exhibitedsimilar mRNAaccumulationboth6hoursand24hoursafterE2 Figure2 Macroarrayresultsshowingsexuallydimorphicexpressionpatterns. Logviewofthefoldchangebetweenfemaleandmale expressionpatternsdeterminedfromamacroarray. ChojnowskiandBraun BMCGenomics 2012, 13 :308 Page5of12 http://www.biomedcentral.com/1471-2164/13/308

PAGE 6

exposure;accumulationofthe C16ORF62 mRNAalmost returnedtopre-exposurelevelsafter24hours. MMP2 is affectedbyE2baseduponpreviousstudieswhereasthe impactofE2exposureuponmRNAaccumulationfor othergeneshasnotbeenexamined.Mahmoodzadeh etal.[59]showedthatE2inhibits MMP2 geneexpressioninratfibroblastsandthoseresultscorroborateour findingsfor MMP2 ’ sinvolvementwithE2.ExpressionofalongnoncodingRNA(ncRNA)is sexually-dimorphicAnumberofcDNAsonthemacroarray(61)couldnot beidentifiedusingBLASTX,suggestingthattheycorrespondeithertocDNAsforwhichonlyuntranslated regionwasincludedintheESTreadornoncoding RNAs(ncRNA).ToidentifysomeofthesecDNAswe conductedBLASTNsearchesandrevealedthatoneof thecDNAsthatexhibitssexualdimorphismisancRNA, MALAT1 MALAT1 isalong(~7kb)ncRNAthatundergoesa cleavagethatproducestwoRNAs,asmallertRNA-like cytoplasmicRNA(~61nt)anda6.7kbRNAthatlocalize totwodifferentsubcellularcompartments,cytoplasm andnuclearspecklesrespectively[60].Characteristically, ithasshortblocksofhighconservationacrosstheentire transcript,especiallyin3 ’ halfofRNA,andlacksrepetitiveelementsexceptforaSINEandLINEelementnear its5 ’ end[60].Thesmaller(~61nt)transcriptgenerated Figure3 Semi-quantitativePCRResults. A.Anexampleofasemi-quantitativegelimage.B.Semi-quantitativeresultsforsexualdimorphism (M=male,F=female),stageeffectbetweenmales(greenbars),andanE2timetrial(orangebars). ChojnowskiandBraun BMCGenomics 2012, 13 :308 Page6of12 http://www.biomedcentral.com/1471-2164/13/308

PAGE 7

bycleavageishighlyconservedacrossmanyspecies,includingmouse,human,dog,lizard,frog,andstickleback. Ithasnotyetbeenfoundinanyofthefouravailable sequencedbirdgenomes(chicken,turkey,zebrafinch, andduck).Therearetwopossibleexplanations:(1)birds lostthelocusor(2)itislocatedinaregionthatwas consistentlyunderrepresentedinalloftheavailable aviangenomeassemblies[61-64].Bothexplanationsare plausiblesincegenelossisknowntobeanimportant processduringevolution[65]buttheaviangenomeassemblies(likeothervertebrategenomeassemblies)are incomplete. MALAT1 showsabroaddistributionofexpressionin normalhumanandmousetissuesbutitsmisregulation iscorrelatedwiththeprogressionofcancersanditis upregulatedinmanyhumancarcinomas[66-69].More importantlyforthisstudy, MALAT1 accumulationis higherinadultmammalianovariesthanadulttestes [60,70].However,thepatternofdifferentialexpression for MALAT1 inadultmammaliangonadsisdistinct fromthepatternweobservedusingthemacroarray assay,inwhichthemRNAaccumulationappeared1.6foldhigherinwholemaleturtleembryos.Since MALAT1 isancRNAthatshowsdimorphicexpression intheTSPwefeltitwasanexcellentcandidatefora geneinvolvedinTSDsoweusedquantitativereal-time PCR(qRT-PCR)toverifythepatternofexpressionsuggestedbythemacroarray. WeusedqRT-PCRtoexamine MALAT1 RNAaccumulationbecauseitrepresentsarigoroustestofdifferentialexpression. MALAT1 RNAaccumulationwas examinedindependentlyformultipleindividuals(n=5) andthetwostagesduringtheTSP(stages17and19)ratherthanusingpooledsamples.Thisanalysisrevealeda slightbutsignificantsexualdimorphism(about1.4-fold higherinmales)intheamountof MALAT1 RNAduring bothstagesweexamined(Figure4). MALAT1 RNAexpressionalsoshowsamodestincreaseasdevelopment progressesfromstage17tostage19inbothmalesand females.Theseobservationsareconsistentwiththehypothesisthat MALAT1 playsaroleinturtleTSD.ConclusionsHerewereportedasurveyofgenesidentifiedbased upontheirpatternsofmRNAaccumulationduringembryonicdevelopmentintheRed-earedsliderturtle.We usedanon-orthologbasedstrategyandidentifiedfour genesthatexhibitedincreasedmRNAexpressionasdevelopmentproceededfromstage14tostage17,asetof genesthatrespondedtoE2exposure,andtwogenes ( MALAT1 and C16ORF62 )thatshowgreateraccumulationatthemaleproducingtemperaturethanatthefemaleproducingtemperature.Thissurveyfocusedon changesinmRNAaccumulationinwholeembryos. Thus,itremainspossiblethatsomeorallofthesegenes exhibitevenmorestronglydimorphicexpressioninspecifictissues(e.g.,thedevelopinggonadorbrain).Moreover,thegenesweidentifiedarelikelytobesignificant sincescreeningfordifferentialexpressionatthewhole embryolevelisexpectedtobeaconservativewayto examinegeneexpressionduringdevelopment. MMP s(matrixmetalloproteinases)areinvolvedinthe breakdownofextracellularmatricesinphysiologicalprocesses,includingcancer[71]. MMP2 issexuallydimorphicindevelopingmalemicebecauseitfunctionsas aparacrinedeathfactorinMllerianductregression downstreamofthe MIS cascade[50].Inaddition,Kim etal.2008[72]found MMP2 tobesexuallydimorphic andregulatedbytestosteroneinsongbirdsinrelationto thevocalcontrolcenterduringadultneurogenesis.Furthermore,estrogenaffectstheMMPpathwayinhumans byincreasingMMP2enzymaticactivity[73].ItisuncleariftheincreaseinMMP2isthroughanincreasein mRNAaccumulationorthroughothermechanisms, suchasbindingaffinitychanges.Though MMP2 was notfoundtobesexuallydimorphicinturtlesitwas foundtobeinhibitedbyE2,theoppositeoftheregulationinmammalsbutpotentiallysimilartobirds[72]. Togetherwiththepriorknowledgeofitsinvolvementin mammalianandaviandevelopment, MMP2 isanovel candidategenefordevelopmentintheturtle. ncRNAsarebelievedtoplayalargenumberofbiologicalroles(reviewedin[74]),buttheirroleindevelopmentremainspoorlycharacterized[75].Althoughthere issomeevidencethatncRNAshaverolesinsexualdevelopmentinbothmammals[75]andbirds[76],thisis thefirstevidencethatancRNAmayhavearoleinsexualdevelopmentforanorganismwithTSD.Thishypothesisiscorroboratedbythefactthat MALAT1 exhibitsdifferentialexpressioninmammaliangonads (expressionishigherinadultovariesthanintestes). However,thepatternofsexualdimorphismreportedfor mammalsisdistinctfromthatevidentinturtles(where theRNAaccumulationishigherinmaleembryosthan infemaleembryos).OurfindingshighlighttheimportanceofexaminingncRNAswheninvestigatingvertebratedevelopmentingeneralandsexualdevelopment specifically. Littleisknownaboutthegenesinvolvedinturtledevelopment,includingtheprocessesrelatedtoTSD,regardlessofwhethertheyareprotein-codinggenesor ncRNAs.Muchoftheinformationavailableisfocused ontheexaminationoftheturtleorthologsofgenes involvedinmammaliandevelopment.Thelimitationson thistypeofanexaminationareexemplifiedbyournovel resultsoftwogenes( MALAT1 and C16ORF62 )that showsexualdimorphismduringTSP.Thesetwotypes ofgenesarenoveltoturtlesexualdevelopmentbecauseChojnowskiandBraun BMCGenomics 2012, 13 :308 Page7of12 http://www.biomedcentral.com/1471-2164/13/308

PAGE 8

oneisancRNA( MALAT1 ),nottypicallyfoundinsexual differentiationinothervertebrates,andtheother ( C16ORF62 )isofunknownfunction.Ourresultshighlighttheneedtodivergefromfocusingonlyonproteincodinggeneswhenlookingfordevelopmentalgenesand toexpandintothemorediverseworldofRNAingeneral,specificallyincludingncRNAs.MethodsIncubationandexperimentalmanipulationsFreshlylaid Trachemysscripta eggs(500)werepurchasedfromKliebertTurtleFarmsinHammond,Louisianain2004and2006.Theywerekeptatroom temperatureforlessthan48hoursuntiltheywereestablishedasviablebycandling.Thoseviablewererandomly separatedequallyintofourexperimentalgroupsincontainerswithmoistenedvermiculite(1:1vermiculiteto water).Theexperimentalgroupswerethefemaleproducingtemperatureof31C(hereaftercalledfemale),the maleproducingtemperatureof26C(hereaftercalled male),26Cpaintedwithexogenousestradiol-17 (E2)in 1 g/ Lin95%ethanol(non-denatured)(phenotypically female;hereaftercalledE2),and26Cpaintedwithexogenous95%ethanol(non-denatured)asthevehicle control(phenotypicallymale;hereaftercalledvehicle). ApplicationofE2andvehicleoccurredatstage14.The eggboxeswererotateddailywithintheincubatorsand wecheckedarandomselectionofeggsperiodicallyto determinetheirdevelopmentalstageusingthestaging guidelinessuggestedbyYntema[77].Thetemperature wasmonitoreddailywithHOBOdataloggersandinincubatorthermometers.Sexwasdeterminedforeachexperimentalgroupthroughavisualinspectionbytwoindependentresearchersathatching(gonadsarevisually distinctathatchingbutnotbefore)of10embryosperexperimentalgroupforrelevantgrossanatomy.Allexperimentscompliedwiththeappropriateethicalguidelines.IsolationofRNAWholeembryosweretakenbetweenstages17and20 fromeachexperimentalgroupandquicklyfrozeninliquidnitrogenandstoredat 80C.Asubsetofwhole embryoswasalsocollectedat0,6,and24hoursafterE2andvehicleapplicationatstage14.Anaverageof5 embryoswascollectedperstageandexperimental group.ConductingtheE2timetrialatstage14before theTSPremovesE2effectswithingonadaldifferentiationandleavesjustthosefromthetrial.TotalRNA wasextractedfromeachembryobyhomogenizationin Tri-Reagent(TRIzol,SigmaUSA),followedbyextraction inchloroform,andprecipitationinisopropanolaccordingtoSambrookandRussell[78].TotalRNAyieldand qualitywereassessedwiththeND-1000Nanodropspectrophotometer(NanoDropTechnologies,ThermoFisher Scientific,Wilmington,DE19810,USA),andtheintegritywasverifiedbyelectrophoresisona1%agarosegel.Suppressionsubtractivehybridization(SSH)Threelibrarieswereselectivelyinducedforfemale againstmale,maleagainstfemale,andE2againstvehicle. TestersanddriversweremadefrompooledRNAfrom stages17 – 20(twoindividualsperstage)fromeachexperimentalgroupfrom2004(statedabove).cDNAsynthesiswasperformedwiththeBDSMART ™ PCRcDNA SynthesisKit(Clontech,MountainView,CA)according tothemanufacturer ’ sprotocol.ThreesubtractionlibrarieswereconstructedwiththeClontechPCR – SelectTMcDNASubtractionKit(Clontech,MountainView,CA) Figure4 QuantitativeRT-PCRshowingsexualdimorphicexpressionatstage17and19. Theexpressionof MALAT1 (n=5)consistsofsex andstage(e.g.M17=male,stage17)andisrelativetothecontrolgene( PP1 ).TherelativeexpressionforM17wassetas1toalloweasier comparisonbetweengroups.Geneexpressionwasanalysedusingtwo-tailed t -testtoexaminedifferencesbetweensexesatthesamestage (stage17, P =0.04andstage19, P =0.01)andbetweenstagesofthesamesex(male, P =0.18andfemale, P =0.24). ChojnowskiandBraun BMCGenomics 2012, 13 :308 Page8of12 http://www.biomedcentral.com/1471-2164/13/308

PAGE 9

accordingtothemanufacturer ’ sprotocolexceptaPEG (PolyEthyleneGlycol)precipitationfollowedbyanethanol washwasusedtopurifythePCRproductsaftercDNA synthesisinsteadofthecolumnchromatography.The resultingcDNAwasligatedintoapGEM-TEasyvector andtransformedinto E.cloniW10Gelectrocompetent cells(Lucigen)bythemanufacturer ’ sprotocol.Individual colonieswerepickedandstoredin96-wellplateswith 50%glycerolat 80C.Plasmidinsertswerepurified usingamodified96-wellPerfectprepWPlasmidprotocol (5Prime,Gaithersburg,MD),accordingtoSambrookand Russell[78]oraTempliPhiAmplificationkit(asrecommendedbymanufacturer;GEHealthcare).Single-passsequencingwasconductedonanABIPrism ™ 3100-Avant geneticanalyzer(PEAppliedBiosystems)usingtheABI BigDyeWTerminatorv.3.1chemistry.AnalysisofSSHresultsSequencesfromthelibrarieswithredundancywere alignedandallsequences(bothindividualandaligned) wereeditedinSequencher ™ 4.1(GeneCodesCorp.). SequenceswereputintoFASTAformatandrunin GOannafromtheAgBasev.2.0databasetodetermine thetopBlasthitforeachsequenceandtosimultaneously determineGOtermsforeachhit[79].GOannauses BLASTX(determinesgeneproductsfromsequences) thereforeanysequencesthatdidnothaveahitwererun throughBLASTN(determinesallaspectsofRNAtranscriptsincludinguntranslatedregionsandnon-protein codingRNAs)ontheNCBIserver[57]. GeneMergecategorizedthehumanhomologsofgenes thatresultedfromallthreeSHHlibrarieswithoverrepresentedGO(GeneOntology)termsfromthebiologicalprocessescategorygivenahumanbackground setofgenes[47].Thesignificancecut-offwassetat P < 0.05. ThegenesthathavehumanhomologsfromtheSSH librarieswereclusteredintofunctionallyrelatedgroups withinthesubsetofbiologicalprocessesbyDAVIDv.6.7 [48,49].TheDAVIDtoolforfunctionalannotationclusteringusesGOtermsandthetermenrichmentscore wasusedathighstringency(basedon P < 0.05).MacroarraypreparationandanalysesThreehundredandseventyfourclones(including322 knownclones[thosewithBLASThits]and61unknown clones)thatwereobtainedfromallthreeSSHlibraries (discussedabove)werespottedontomembranes(Pall BiodyneBNylon,Nunc)using100nanoliterpinsona Biomek2000(BeckmanCoulter,USA).Positivecontrols ( ArabidopsisthalianaRCA [X14212], CAB [X56062], and RBCL [U91966])andnegativecontrolswerealso spottedontothemembranes.Allsampleswerespotted induplicatewithfourreplicatesperexperimentalgroup (femaleandmale).TotalRNAwascollected(asstated above)from2embryosfromstage17and2embryos fromstage19.Allcollectedembryosfromthemaleexperimentalgroupin2004werepooledaswellasforthe femaleexperimentalgroup.PooledtotalRNAwasmixed withcontrolcDNAsandthenreversetranscribedbefore labellingwith 33P-dATPasdescribedinBlumetal. [80].Afterhybridization[80],themembraneswere rinsedandexposedtoaphosphorimagerandscanned usingaMolecularDevicesTyphoonScanner. Table3Semi-quantitativeandquantitativereal-timePCRprimers,andoptimalconditionsforsemi-quantitativePCRGenesymbolPrimernamePrimerSequenceAnnealing Temperature(C) MgCl2Concentration(mM)* Cycle Number Cct3 Cct3FGGATGCCTAAAATTAGCCTCCTA62.51.530 Cct3RGAAGCTACGGCAAATGATGG Malat1 } Malat1FGTACGCGGGCAGACTAACAC57.11.536 Malat1RTGCGTCTAGACACCACAACC C16orf62 C16orf62FCGGCCGAGGTACAAATTAAG58.32.536 C16orf62RTGCAAGTGCATTATGGAAGC Mmp2 Mmp2FATGAAGAAGCCCCGCTGTGGTAATCC62.54.527 Mmp2RAAAGGCATCGTCTACTGTTTCGGAGTCC Nf1a NfibFAAACACACTGCGTCAAGTGC61.41.524 NfibRCTTGCCCTGGATAGCGATTA Notch2 Notch2FTATTTCTGTGGCTGCCTGGA62.51.536 Notch2RGGGACAGGGACCTTTGTTGT Pp1 } Pp1FACCTCTTCCTGGGCGACTAT62.51.527 Pp1RTGATGTTGTAGCGCCTCTTG*totalconcentration,includinganyMgCl2inTaqbuffer. } SameprimersusedforqRT-PCRbutat60Cannealingtemperatureand40cycles.ChojnowskiandBraun BMCGenomics 2012, 13 :308 Page9of12 http://www.biomedcentral.com/1471-2164/13/308

PAGE 10

SignalintensitieswerequantifiedusingImageQuant 5.1(AmershamBiosciencesUKLimited,Amersham PlaceLittleChalfontBuckinghamshireEngland)andintensitydifferenceswerecalculatedasdescribedbyHelbing etal.[81].Briefly,imagedatawereconvertedtoastandard 8-bitTIFFfilebeforeaccountingforsignalsaturation. Positivecontrolsusedtostandardizeacrossarraysfor hybridizationefficiencywerechosenbasedontheircoefficientofvariation(cv)acrossall8arrays(thecvdidnot exceed0.2).Eacharraywasnormalizedtothegeometric meanforallpositivecontrolschosenforthatarray[82]. Thenon-signalbackgroundwasdeterminedusingthemedianintensityvalueplusonestandarddeviationforthe blankpositionsandnegativecontrols,andthemaximum valueforallarrayswassetasthe “ nosignal ” value.All valuesacrossallarraysequaltoorbelowthatnumberis consideredzero.Afternormalizingacrossarraysanddiscardingvaluesequaltoorbelowthefloorvalue,theaverageoftheduplicateswithinarrays,theestimatedstandard deviationacrossreplicatearrays,themedianacross replicatearrays,andthefoldchangebetweentreatmentgroupswerecalculated.Ifthestandarddeviationacrossreplicateswasgreaterthanorequalto2 thenthosespotswerenotreliableforfurtherexamination.BaseduponHelbingetal.[81],weconsidered a1.5-folddifferenceinsignalintensityrelativetocontroltreatmentsassufficientinthearrayexperiments basedonthedetectionlimitationsofthecDNAarray analyses.Thosegenesthatexhibiteddifferentialexpressionwereusedforfurtherexamination.Semi-quantitativePCRpreparationandanalysisEmbryoswerecollectedandtotalRNAwasextractedas statedabovefromthe31Cexperimentalgroupatstage 17,the26Cexperimentalgroupatstage17,andfrom theE2timetrialexperimentalgroupatstage14from 2006.Fourembryosfromeachgroupwerepooledand cDNAwasmadeusingInvitrogen ’ sSuperscriptIIIReverse Transcriptionkit(Invitrogen,USA)followingthemanufacturer ’ sinstructions. Pilotexperimentswereconductedtodetermine optimumPCRconditionsforthecandidategenesanda controlgene( PP1 ,[12]).PCRprimersweredesigned usingPrimer3[83]andalistofsequencesandPCRconditionsafterpilotexperimentscanbefoundinTable3. Controlsweresystematicallyrunineachsetofsemiquantitativeassays:1)acDNApositivecontrolfora knownsample;2)aPCRpositivecontrol;and3)anegative control.Aninternalexogenousstandard(cDNAsynthesis withnoreversetranscriptase)wasalsorunseparatelyfor eachcDNAmixture.PCRproductswereloadedontoa 1.5%TBEgelwithethidiumbromideanda1kbpDNAladdermolecularweightmarker(MinnesotaMolecular)and electrophoresedat90Vfor45minutes. AnalysisofgelimageswasconductedusingImageJ [84].Eachexperimentalgenewasstandardizedtothe controlgene, PP1 [12],andthennormalizedtothefemale group.QuantitativeReal-timePCR(qRT-PCR)preparationand analysisFivewholeembryoswerecollectedfromstage17and5 fromstage19fromtwoexperimentalgroups,maleandfemale,andtotalRNAwasextractedasstatedabove.cDNA wasgeneratedusingImProm-II ™ ReverseTranscriptase andrandomprimersfollowingthemanufacturer ’ sinstructions.Relativegeneexpressionlevelswerequantifiedusing anABIStepOnePlus ™ Real-timePCRcycler(STepOne ™ Softwarev2.1)withthefollowingcyclingparameters:initialdenaturingfor10minat95C,followedby40cycles of35sat95C,30sat60C,and30sat72C.Thefinal cyclewasfollowedbyameltingcurveanalysistoverify theamplificationofasingleproductineachwell.Specificitiesofallprimerpairswerealsoverifiedbysequencing PCRproducts.RepeatingtheaboveproceduresonRNA samples(priortoreversetranscription)verifiedthatno productswereamplifiedfromcontaminatinggenomic DNA.Allsampleswereruninduplicateandincluded3.75 Lofa1:100dilutedsample,1 Mofeachprimer,and2x SYBRGreenMasterMix(AppliedBiosystems)inatotal of15 L.PCRefficiencieswerecalculatedfromagenespecificstandardcurvefroma10-folddilutionseries. Relativetranscriptabundancewasnormalizedtotheexpressionof PP1 byusingtherelativestandardcurve method[85].Todetermineifexpressiondifferedbetween experimentalgroupsatwo-tailedStudent ’ s t -testanda standarderroranalysiswereperformed.Primersusedto assaygeneexpressionweredesignedusingPrimer3[85] andAmplify[86](Table3).AdditionalfilesAdditionalfile1: Homologysearchesforsubtractionlibraries. Homologysearchesforsequencesfromall3libraries.Thetableincludes cloneidentifier,match,accessionnumber,e-value,library,and redundancy. Additionalfile2: GeneMergeforbiologicalprocesses. GeneMerge determinedover-representedGOtermsformygenesetgivenahuman genesetbackground.Thecut-offwassettoP < 0.05. Additionalfile3: DAVIDfunctionalannotationclustering. The DAVIDclusteringtoolwasusedtodeterminefunctionallyrelated groupingsbasedonmygenesetfromtheSSHlibraries.Theterm enrichmentwassetathighstringency. Competinginterests Theauthorsdeclaretheyhavenocompetinginterests. Acknowledgements WearegratefultoRebeccaKimball,GordonBurleigh,TeresaBryan,Marta Wayne,KarenKochandthemembersoftheBraun-Kimballlabgroupfor helpfuldiscussions.ThisworkwasfacilitatedbyNationalScienceFoundationChojnowskiandBraun BMCGenomics 2012, 13 :308 Page10of12 http://www.biomedcentral.com/1471-2164/13/308

PAGE 11

grantstoELBandcollaborators(DEB-0228682andDUE-0920151)aswellas anAmericanSocietyofIchthyologyandHerpetologygrantandaSigmaXi granttoJLC.PublicationofthisarticlewasfundedinpartbytheUniversity ofFloridaOpen-AccessPublishingFund. Authors ’ contributions JLCcarriedoutalldatacollection,participatedintheexperimentaldesign anddraftedthemanuscript.ELBparticipatedintheexperimentaldesign, bioinformaticanalyses,andhelpedtodraftthemanuscript.Allauthorsread andapprovedthefinalmanuscript. Received:12September2011Accepted:15July2012 Published:15July2012 References1.Francisco-MorcilloJ,Hidalgo-SnchezM,Martn-PartidoG: Spatialand temporalpatternsofproliferationanddifferentiationinthedeveloping turtleeye. BrainRes 2006, 1103: 32 – 48. 2.MehtaV,SernagorE: Earlyneuralactivityanddendriticgrowthinturtle retinalganglioncells. EurJNeurosci 2006, 24: 773 – 786. 3.BernierB,BarI,PieauC,LambertDeRouvroitC,GoffinetA: ReelinmRNA expressionduringembryonicbraindevelopmentintheturtle Emys orbicularis JCompNeurol 1999, 413: 463 – 479. 4.MetinC,AlvarezC,MoudouxD,VitalisT,PieauC,MolnarZ: Conserved patternoftangentiealneuronalmigrationduringforebrain development. Development 2007, 134: 2815 – 2827. 5.KurataniS,KurakuS,NagashimaH: Evolutionarydevelopmental perspectivefortheoriginofturtles:thefoldingtheoryfortheshell basedonthedevelopmentalnatureofthecarapacialridge. EvolDev 2011, 13: 1 – 14. 6.RamseyM,CrewsD: Steroidsignalingandtemperature-dependentsex determination — Reviewingtheevidenceforearlyactionofestrogen duringovariandeterminationinturtles. SeminCellDevBiol 2009, 20: 283 – 292. 7.FlemingA,WibbelsT,SkipperJK,CrewsD: Developmentalexpressionof steroidogenicfactor1inaturtlewithtemperature-dependentsex determination. GenCompEndocrinol 1999, 116: 336 – 346. 8.RamseyM,ShoemakerC,CrewsD: Gonadalexpressionof Sf1 and aromataseduringsexdeterminationinthered-earedsliderturtle ( Trachemysscripta ),areptilewithtemperature-dependentsex determination. Differentiation 2007, 75: 978 – 991. 9.ValenzuelaN: Relicthermosensitivegeneexpressioninaturtlewith genotypicsexdetermination. Evolution 2008, 62: 234 – 240. 10.SchmahlJ,YaoHH,Pierucci-AlvesF,CapelB: ColocalizationofWT1and cellproliferationrevealsconservedmechanismsintemperaturedependentsexdetermination. Genesis 2003,35: 193 – 201. 11.SpotilaLD,SpotilaJR,HallSE: SequenceandexpressionanalysisofWT1 andSox9inthered-earedsliderturtle, Trachemysscripta JExpZool 1998, 281: 417 – 427. 12.ShoemakerCM,QueenJ,CrewsD: Responseofcandidatesexdetermininggenestochangesintemperaturerevealstheirinvolvement inthemolecularnetworkunderlyingtemperature-dependentsex determination. MolEndocrinol 2007, 21: 2750 – 2763. 13.RamseyM,CrewsD: Adrenal-kidney-gonadcomplexmeasurementsmay notpredictgonad-specificchangesingeneexpressionpatternsduring temperature-dependentsexdeterminationinthered-earedsliderturtle ( Trachemysscriptaelegans ). JExpZoolAEcolGenetPhysiol 2007, 307: 463 – 470. 14.ShoemakerC,RamseyM,QueenJ,CrewsD: Expressionof Sox9 Mis ,and Dmrt1 inthegonadofaspecieswithtemperature-dependentsex determination. DevDyn 2007, 236: 1055 – 1063. 15.MurdockC,WibbelsT: ExpressionofDmrt1inaturtlewithtemperaturedependentsexdetermination. CytogenetGenomeRes 2003, 101: 302 – 308. 16.TorresMaldonadoLC,LandaPiedraA,MorenoMendozaN,Marmolejo ValenciaA,MezaMartinezA,MerchantLariosH: Expressionprofilesof Dax1 Dmrt1 ,and Sox9 duringtemperaturesexdeterminationingonadsofthe seaturtle Lepidochelysolivacea GenCompEndocrinol 2002, 129: 20 – 26. 17.MurdockC,WibbelsT: Cloningandexpressionofaromataseinaturtle withtemperature-dependentsexdetermination. GenCompEndocrinol 2003, 130: 109 – 119. 18.TakadaS,DiNapoliL,CapelB,KoopmanP: Sox8 isexpressedatsimilar levelsingonadsofbothsexesduringthesexdeterminingperiodin turtles. DevDyn 2004, 231: 387 – 395. 19.SmithCA,ShoemakerCM,RoeszlerKN,QueenJ,CrewsD,SinclairAH: Cloningandexpressionof R-Spondin1 indifferentvertebratessuggestsa conservedroleinovariandevelopment.BMCDevBiol 2008, 8: 72. 20.CrewsD,BergeronJM,BullJJ,FloresD,TousignantA,SkipperJK,WibbelsT: Temperature-dependentsexdeterminationinreptiles:proximate mechanisms,ultimateoutcomes,andpracticalapplications. DevGenet 1994, 15: 297 – 312. 21.JanzenFJ,PhillipsPC: Exploringtheevolutionofenvironmentalsex determination,especiallyinreptiles. JEvolBiol 2006, 19: 1775 – 1784. 22.EzazT,StiglecR,VeyrunesF,MarshallGravesJA: Relationshipsbetween vertebrateZWandXYsexchromosomesystems. CurrBiol 2006, 16: 736 – 743. 23.SinclairAH,BertaP,PalmerMS,HawkinsJR,GriffithsBL,SmithMJ,FosterJW, FrischaufAM,Lovell-BadgeR,GoodfellowPN: Agenefromthehuman sex-determiningregionencodesaproteinwithhomologytoa conservedDNA-bindingmotif. Nature 1990, 346: 240 – 244. 24.VolffJN,KondoM,SchartlM: Medaka dmY/dmrt1Y isnottheuniversal primarysex-determininggeneinfish. TrendsGenet 2003, 19: 196 – 199. 25.MatsudaM,ShinomiyaA,KinoshitaM,SuzukiA,KobayashiT,Paul-Prasanth B,LauEL,HamaguchiS,SakaizumiM,NagahamaY: DMY geneinduces maledevelopmentingeneticallyfemale(XX)medakafish. ProcNatlAcad SciUSA 2007, 104: 3865 – 3870. 26.HerpinA,BraaschI,KraeusslingM,SchmidtC,ThomaEC,NakamuraS, TanakaM,SchartlM: Transcriptionalrewiringofthesexdetermining dmrt1 geneduplicatebytransposableelements. PLoSGenet 2010, 6: e1000844. 27.JohnstonCM,BarnettM,SharpePT: Themolecularbiologyof temperature-dependentsexdetermination. PhilosTransRSocLondBBiol Sci 1995, 350: 297 – 303.discussion303 – 294. 28.WesternPS,SinclairAH: Sex,genes,andheat:triggersofdiversity. JExp Zool 2001, 290 (6):624 – 631. 29.ClintonM: Sexdeterminationandgonadaldevelopment:abird'seye view. JExpZool 1998, 281: 457– 465. 30.GuanG,KobayashiT,NagahamaY: Sexuallydimorphicexpressionoftwo typesofDM( Doublesex/Mab-3 )-domaingenesinateleostfish,the Tilapia( Oreochromisniloticus ). BiochemBiophysResCommun 2000, 272: 662 – 666. 31.WesternPS,HarryJL,MarshallGravesJA,SinclairAH: TemperaturedependentsexdeterminationintheAmericanalligator:expressionof SF1 WT1 and DAX1 duringgonadogenesis. Gene 2000, 241: 223 – 232. 32.SmithCA,SinclairAH: Sexdeterminationinthechickenembryo. JExp Zool 2001, 290: 691 – 699. 33.ShibataK,TakaseM,NakamuraM: The Dmrt1 expressioninsex-reversed gonadsofamphibians. GenCompEndocrinol 2002, 127: 232 – 241. 34.DewingP,ShiT,HorvathS,EricVilainE: Sexuallydimorphicgene expressioninmousebrainprecedesgonadaldifferentiation. BrainRes MolBrainRes 2003, 118: 82 – 90. 35.LeeSI,LeeWK,ShinJH,HanBK,MoonS,ChoS,ParkT,KimH,HanJY: Sexuallydimorphicgeneexpressioninthechickbrainbeforegonadal differentiation. PoultSci 2009, 88: 1003 – 1015. 36.JnneM,HogeveenKN,DeolHK,HammondGL: Expressionandregulation ofhumansexhormone-bindingglobulintransgenesinmiceduring development. Endocrinology 1998, 140: 4166 – 4174. 37.WibbelsT,BullJJ,CrewsD: Chronologyandmorphologyoftemperaturedependentsexdetermination. JExpZool 1991, 260: 371 – 381. 38.ChojnowskiJL,FranklinJ,KatsuY,IguchiT,GuilletteLJJr,KimballRT,Braun EL: Patternsofvertebrateisochoreevolutionrevealedbycomparisonof expressedmammalian,avian,andcrocodiliangenes. JMolEvol 2007, 65: 259 – 266. 39.ChojnowskiJL,BraunEL: Turtleisochorestructureisintermediatebetween amphibiansandotheramniotes. IntegrCompBiol2008, 48: 454 – 462. 40.HuangX,MadanA: CAP3:ADNAsequenceassemblyprogram. Genome Res 1999, 9: 868 – 877. 41.DiatchenkoL,LauYF,CampbellAP,ChenchikA,MoqadamF,HuangB, LukyanovS,LukyanovK,GurskayaN,SverdlovED, etal : Suppression subtractivehybridization:amethodforgeneratingdifferentially regulatedortissue-specificcDNAprobesandlibraries. ProcNatlAcadSci USA 1996, 93: 6025 – 6030. 42.LuoMJ,LaiMD: Identificationofdifferentiallyexpressedgenesinnormal mucosa,adenomaandadenocarcinomaofcolonbySSH. WorldJ Gastroenterol 2001, 7: 726 – 731.ChojnowskiandBraun BMCGenomics 2012, 13 :308 Page11of12 http://www.biomedcentral.com/1471-2164/13/308

PAGE 12

43.ChenAQ,WangZG,XuZR,YuSD,YangZG: Analysisofgeneexpression ingranulosacellsofovineantralgrowingfolliclesusingsuppressive subtractivehybridization. AnimReprodSci 2009, 115: 39 – 48. 44.DeFerrariL,AitkenS: MininghousekeepinggeneswithaNaiveBayes classifier. BMCGenomics 2006, 7: 277 – 291. 45.KudlaG,LipinskiL,CaffinF,HelwakA,ZyliczM: Highguanineandcytosine contentincreasesmRNAlevelsinmammaliancells. PLoSBiol 2006, 4: e180. 46.ArhondakisS,ClayO,BernardiG: GClevelandexpressionofhuman codingsequences. BiochemBiophysResCommun 2008, 367: 542 – 545. 47.Castillo-DavisCI,HartlDL: GeneMerge – post-genomicanalysis,data mining,andhypothesistesting. Bioinformatics 2003, 19: 891 – 892. 48.DennisGJr,ShermanBT,HosackDA,YangJ,GaoW,LaneHC,LempickiRA: DAVID:DatabaseforAnnotation,Visualization,andIntegratedDiscovery. GenomeBiol 2003, 4: P3. 49.daHuangW,ShermanBT,LempickiRA: Systematicandintegrative analysisoflargegenelistsusingDAVIDbioinformaticsresources. Nat Protoc 2009, 4: 44 – 57. 50.RobertsLM,VisserJA,IngrahamHA: Involvementofamatrix metalloproteinaseinMIS-inducedcelldeathduringurogenital development. Development 2002, 129: 1487 – 1496. 51.KohnoS,KatsuY,UrushitaniH,OhtaY,IguchiT,GuilletteLJJr: Potential contributionsofheatshockproteinstotemperature-dependentsex determinationintheAmericanalligator. SexDev 2010, 4: 73 – 87. 52.PicardD: Chaperoningsteroidhormoneaction. TrendsEndocrinolMetab 2006, 17: 229 – 235. 53.RobinsonLL,SznajderNA,RileySC,AndersonRA: Matrix metalloproteinasesandtissueinhibitorsofmetalloproteinasesinhuman fetaltestisandovary. MolHumReprod 2001, 7: 641 – 648. 54.ChaudhryAZ,LyonsGE,GronostajskiRM: Expressionpatternsofthefour nuclearfactorIgenesduringmouseembryogenesisindicateapotentialroleindevelopment. DevDyn 1997, 208: 313 – 325. 55.WalkleyNA,DemaineAG,MalikAN: Cloning,structureandmRNA expressionofhumanCctg,whichencodesthechaperoninsubunitCCT gamma. BiochemJ 1996, 313 (Pt2):381 – 389. 56.WeinmasterG,RobertsV,LemkeG: Notch2 :asecondmammalian Notch gene. Development 1992, 116: 931 – 941. 57.NCBI:http://www.ncbi.nlm.nih.gov. 58. Bgee:adataBaseforGeneExpressionEvolution .http://bgee.unil.ch/bgee/bgee. 59.MahmoodzadehS,DworatzekE,FritschkaS,PhamTH,Regitz-ZagrosekV: 17beta-Estradiolinhibitsmatrixmetalloproteinase-2transcriptionvia MAPkinaseinfibroblasts. CardiovascRes 2010, 85: 719 – 728. 60.WiluszJE,FreierSM,SpectorDL: 3'endprocessingofalongnuclearretainednoncodingRNAyieldsatRNA-likecytoplasmicRNA. Cell 2008, 135: 919 – 932. 61.MorissonM,DenisM,MilanD,KloppC,LerouxS,BardesS,PitelF,Vignoles F,GrusM,FillonV,DouaudM,VignalA: ThechickenRHmap:current stateofprogressandmicrochromosomemapping. CytogenetGenomeRes 2007, 117: 14 – 21. 62.HillierLW,MillerW,BirneyE,WarrenW,HardisonRC,PontingCP,BorkP, BurtDW,GroenenMAM,DelanyME,DodgsonJB,ChinwallaAT,CliftenPF, CliftonSW,DelehauntyKD,FronickC,FultonRS,GravesTA,KremitzkiC, LaymanD,MagriniV,McPhersonJD,MinerTL,MinxP,NashWE,NhanMN, NelsonJO,OddyLG,PohlCS,Randall-MaherJ: InternationalChicken GenomeSequencingConsortium, etal :Sequenceandcomparative analysisofthechickengenomeprovideuniqueperspectiveson vertebrateevolution. Nature 2004, 432: 695 – 716. 63.DalloulRA,LongJA,ZiminAV,AslamL,BealK,AnnBlombergL,BouffardP, BurtDW,CrastaO,CrooijmansRPMA,CooperK,CoulombeRA,DeSupriyo, DelanyME,DodgsonJB,DongJJ,EvansC,FredericksonKM,FlicekP,Florea L,FolkertsO,GroenenMAM,HarkinsTT,HerreroJ,HoffmannS,Megens H-Jan,JiangA,deJongP,KaiserP,KimH, etal : Multi-platformnextgenerationsequencingofthedomesticturkey( Meleagrisgallopavo ): Genomeassemblyandanalysis. PLoSBiol 2010, 8: e1000475. 64.WarrenWC,ClaytonDF,EllegrenH,ArnoldAP,HillierLW,KnstnerA,Searle S,WhiteS,VilellaAJ,FairleyS,HegerA,KongL,PontingCP,JarvisED,Mello CV,MinxP,LovellP,VelhoTAF,FerrisM,BalakrishnanCN,SinhaS,BlattiC, LondonSE,LiY,LinY,GeorgeJ,SweedlerJ,SoutheyB,GunaratneP, WatsonM, etal : Thegenomeofasongbird. Nature 2010, 464: 757– 762. 65.BraunEL: Innovationfromreduction:geneloss,domainlossand sequencedivergenceingenomeevolution. ApplBioinformatics 2003, 2: 13 – 34. 66.JiP,DiederichsS,WangW,BoingS,MetzgerR,SchneiderPM,TidowN, BrandtB,BuergerH,BulkE, etal : MALAT-1,anovelnoncodingRNA,and thymosinbeta4predictmetastasisandsurvivalinearly-stagenon-small celllungcancer. Oncogene 2003, 22: 8031 – 8041. 67.GuffantiA,IaconoM,PelucchiP,KimN,SoldaG,CroftLJ,TaftRJ,RizziE, Askarian-AmiriM,BonnalRJ, etal : Atranscriptionalsketchofaprimary humanbreastcancerby454deepsequencing. BMCGenomics 2009, 10: 163. 68.KoshimizuTA,FujiwaraY,SakaiN,ShibataK,TsuchiyaH: Oxytocin stimulatesexpressionofanoncodingRNAtumormarkerinahuman neuroblastomacellline. LifeSci 2010, 86: 455 – 460. 69.GuoF,LiY,LiuY,WangJ,LiY,LiG: Inhibitionofmetastasis-associated lungadenocarcinomatranscript1inCaSkihumancervicalcancercells suppressescellproliferationandinvasion. ActaBiochimBiophysSin (Shanghai) 2010, 42: 224 – 229. 70.HutchinsonJN,EnsmingerAW,ClemsonCM,LynchCR,LawrenceJB,Chess A: AscreenfornucleartranscriptsidentifiestwolinkednoncodingRNAs associatedwithSC35splicingdomains. BMCGenomics 2007, 8: 39. 71.BourbouliaD,Stetler-StevensonWG: Matrixmetalloproteinases(MMPs) andtissueinhibitorsofmetalloproteinases(TIMPs):Positiveand negativeregulatorsintumorcelladhesion. SeminCancerBiol 2010, 20: 161 – 168. 72.KimDH,LilliehookC,RoidesB,ChenZ,ChangM,MobasheryS,Goldman SA: Testosterone-inducedmatrixmetalloproteinaseactivationisa checkpointforneuronaladditiontotheadultsongbirdbrain. JNeurosci 2008, 28: 208 – 216. 73.GrandasOH,MountainDH,KirkpatrickSS,CassadaDC,StevensSL,Freeman MB,GoldmanMH: Regulationofvascularsmoothmusclecellexpression andfunctionofmatrixmetalloproteinasesismediatedbyestrogenand progesteroneexposure. JVascSurg 2009, 49: 185 – 191. 74.PontingCP,OliverPL,ReikW: Evolutionandfunctionsoflongnoncoding RNAs. Cell 2009, 136: 629 – 641.75.McFarlaneL,WilhelmD: Non-codingRNAsinmammaliansexual development. SexDev 2010, 3: 302 – 316. 76.ZhaoD,McBrideD,NandiS,McQueenHA,McGrewMJ,HockingPM,Lewis PD,SangHM,ClintonM: Somaticsexidentityiscellautonomousinthe chicken. Nature 2010, 464: 237 – 242. 77.YntemaC: Aseriesofstagesintheembryonicdevelopmentof Chelydra serpentina JMorph 1968, 125: 219 – 252. 78.SambrookJ,RussellD: MolecularCloning:ALaboratoryManual .3rdedition. ColdSpringHarbor,NewYork:ColdSpringHarborLaboratoryPress;2001. 79.McCarthyFM,WangN,MageeGB,NanduriB,LawrenceML,CamonEB, BarrellDG,HillDP,DolanME,WilliamsWP, etal : AgBase:afunctional genomicsresourceforagriculture. BMCGenomics 2006, 7: 229. 80.BlumJL,PruchaMS,PatelVJ,DenslowND: UseofcDNAmacroarraysand geneprofilingfordetectionofeffectsofenvironmentaltoxicants. MethodsMolBiol 2008, 410: 43 – 53. 81.HelbingCC,BaileyCM,JiL,GundersonMP,ZhangF,VeldhoenN,Skirrow RC,MuR,LesperanceM,HolcombeGW,KosianPA,TietgeJ,KorteJJ,Degitz SJ: Identificationofgeneexpressionindicatorsforthyroidaxisdisruption ina Xenopuslaevis metamorphosisscreeningassay:Part1.Effectson thebrain. AquaticToxicology 2007, 82: 227 – 241. 82.PfafflMW,TichopadA,PrgometC,NeuviansTP: Determinationofstable housekeepinggenes,differentiallyregulatedtargetgenesandsample integrity:BestKeeper – Excel-basedtoolusingpair-wisecorrelations. BiotechnolLett 2004, 26: 509 – 515. 83. Primer3 .http://frodo.wi.mit.edu. 84. ImageJ .Bethesda,MD:WayneRasband,NationalInstitutesofHealth;http:// rsb.info.nih.gov/ij. 85.LarionovA,KrauseA,MillerW: Astandardcurvebasedmethodfor relativerealtimePCRdataprocessing. BMCBioinformatics 2005, 6: 62. 86. Amplify .http://engels.genetics.wisc.edu/amplify.doi:10.1186/1471-2164-13-308 Citethisarticleas: ChojnowskiandBraun: Anunbiasedapproachto identifygenesinvolvedindevelopmentinaturtlewith temperature-dependentsexdetermination. BMCGenomics 2012 13 :308.ChojnowskiandBraun BMCGenomics 2012, 13 :308 Page12of12 http://www.biomedcentral.com/1471-2164/13/308