The evolution of a series of behavioral traits is associated with autism-risk genes in cavefish


previous item | next item

Citation
The evolution of a series of behavioral traits is associated with autism-risk genes in cavefish

Material Information

Title:
The evolution of a series of behavioral traits is associated with autism-risk genes in cavefish
Series Title:
BMC Evolutionary Biology
Creator:
Yoshizawa, Masato
Settle, Alexander
Hermosura, Meredith C.
Tuttle, Lilian J. Cetraro, Nicholas et al
Publisher:
BMC
Publication Date:
Language:
English

Subjects

Subjects / Keywords:
Vertebrate Model ( local )
Hapflk ( local )
Systemic Regulation ( local )
Psychiatric Disease ( local )
Adaptation ( local )
Astyanax Mexicanus ( local )
Genre:
serial ( sobekcm )

Notes

Abstract:
Background An essential question in evolutionary biology is whether shifts in a set of polygenic behaviors share a genetic basis across species. Such a behavioral shift is seen in the cave-dwelling Mexican tetra, Astyanax mexicanus. Relative to surface-dwelling conspecifics, cavefish do not school (asocial), are hyperactive and sleepless, adhere to a particular vibration stimulus (imbalanced attention), behave repetitively, and show elevated stress hormone levels. Interestingly, these traits largely overlap with the core symptoms of human autism spectrum disorder (ASD), raising the possibility that these behavioral traits are underpinned by a similar set of genes (i.e. a repeatedly used suite of genes). Result Here, we explored whether modification of ASD-risk genes underlies cavefish evolution. Transcriptomic analyses revealed that > 58.5% of 3152 cavefish orthologs to ASD-risk genes are significantly up- or down-regulated in the same direction as genes in postmortem brains from ASD patients. Enrichment tests suggest that ASD-risk gene orthologs in A. mexicanus have experienced more positive selection than other genes across the genome. Notably, these positively selected cavefish ASD-risk genes are enriched for pathways involved in gut function, inflammatory diseases, and lipid/energy metabolism, similar to symptoms that frequently coexist in ASD patients. Lastly, ASD drugs mitigated cavefish’s ASD-like behaviors, implying shared aspects of neural processing. Conclusion Overall, our study indicates that ASD-risk genes and associated pathways (especially digestive, immune and metabolic pathways) may be repeatedly used for shifts in polygenic behaviors across evolutionary time.
Original Version:
BMC Evolutionary Biology, Vol. 18, no. 89 (2018-06-18).

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:
K26-05201 ( USFLDC: LOCAL DOI )
k26.5201 ( 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

RESEARCHARTICLEOpenAccess Theevolutionofaseriesofbehavioraltraitsisassociatedwithautism-riskgenesincavefishMasatoYoshizawa1* ,AlexanderSettle1,MeredithC.Hermosura2,LillianJ.Tuttle1,NicolasCetraro1,CourtneyN.Passow3andSuzanneE.McGaugh3AbstractBackground:Anessentialquestioninevolutionarybiologyiswhethershiftsinasetofpolygenicbehaviorsshareageneticbasisacrossspecies.Suchabehavioralshiftisseeninthecave-dwellingMexicantetra,Astyanaxmexicanus.Relativetosurface-dwellingconspecifics,cavefishdonotschool(asocial),arehyperactiveandsleepless,adheretoaparticularvibrationstimulus(imbalancedattention),behaverepetitively,andshowelevatedstresshormonelevels.Interestingly,thesetraitslargelyoverlapwiththecoresymptomsofhumanautismspectrumdisorder(ASD),raisingthepossibilitythatthesebehavioraltraitsareunderpinnedbyasimilarsetofgenes(i.e.arepeatedlyusedsuiteofgenes).Result:Here,weexploredwhethermodificationofASD-riskgenesunderliescavefishevolution.Transcriptomicanalysesrevealedthat>58.5%of3152cavefishorthologstoASD-riskgenesaresignificantlyup-ordown-regulatedinthesamedirectionasgenesinpostmortembrainsfromASDpatients.EnrichmenttestssuggestthatASD-riskgeneorthologsinA.mexicanushaveexperiencedmorepositiveselectionthanothergenesacrossthegenome.Notably,thesepositivelyselectedcavefishASD-riskgenesareenrichedforpathwaysinvolvedingutfunction,inflammatorydiseases,andlipid/energymetabolism,similartosymptomsthatfrequentlycoexistinASDpatients.Lastly,ASDdrugsmitigatedcavefish’sASD-likebehaviors,implyingsharedaspectsofneuralprocessing.Conclusion:Overall,ourstudyindicatesthatASD-riskgenesandassociatedpathways(especiallydigestive,immuneandmetabolicpathways)mayberepeatedlyusedforshiftsinpolygenicbehaviorsacrossevolutionarytime.Keywords:Vertebratemodel,hapFLK,Systemicregulation,Psychiatricdisease,Adaptation,AstyanaxmexicanusBackgroundAnimalspecieshaveevolvedtochangingenvironmentsbymodifyingmorphological,physiologicalandbehavioraloutputs[1].Onechallengingquestioninevolutionarybiologyishowanimalsevolvedmultiplebehaviorswithindependentgeneticbases.Indeed,somebehavioralsyndromesdemonstratethatcorrelatedbehaviorscanbeunderpinnedbydifferentgeneticfactors[2,3].Currently,itislargelyunknownifanyparticularsuiteofgenes,orso-called‘genetictoolkit’[4–6],aremodifiedacrossevolutionarytimetoaffectasetofgeneticallyindependentmultiplebehaviors.Toinvestigatehowmultiplebehaviorsevolve,wefocusedonbehaviorswhosephysiologicalandmolecularpathwaysmaybecomparableacrossspecies.Forex-ample,almostallanimalspeciesexhibitasleep-likestate,characterizedbyextendedperiodsofbehavioralquiescencethatcorrelatewithelevatedarousalthresh-oldstosensorystimuli[7,8].Moreover,themolecular(e.g.melatonin)andcellularmechanismsofsleep-likestatesaresharedamongsomeanimalspecies[9,10].Similarly,despitelargedifferencesinthecomplexitiesoftheirbehavioraltraits,vertebratessharemanycorecharacteristicsofneuralconnectivityandmolecularpathwaysintheirinnatesocialbehaviors(e.g.mesolimbic *Correspondence:yoshizaw@hawaii.edu1DepartmentofBiology,UniversityofHawai‘iatMnoa,2538McCarthyMall,EDM216,Honolulu,HI96822,USAFulllistofauthorinformationisavailableattheendofthearticle ©TheAuthor(s).2018OpenAccessThisarticleisdistributedunderthetermsoftheCreativeCommonsAttribution4.0InternationalLicense(http://creativecommons.org/licenses/by/4.0/),whichpermitsunrestricteduse,distribution,andreproductioninanymedium,providedyougiveappropriatecredittotheoriginalauthor(s)andthesource,providealinktotheCreativeCommonslicense,andindicateifchangesweremade.TheCreativeCommonsPublicDomainDedicationwaiver(http://creativecommons.org/publicdomain/zero/1.0/)appliestothedatamadeavailableinthisarticle,unlessotherwisestated.Yoshizawaetal.BMCEvolutionaryBiology (2018) 18:89 https://doi.org/10.1186/s12862-018-1199-9

PAGE 2

systemandoxytocin/vasopressin)[6,11–15],stress-relatedbehaviors[16],attention/cognition-relatedmolecularpathways[17],andstarvation/satiationpathways[18].Someofthesebehaviors,includingsocialresponses,showevendeeperconservationwithinMetazoa,especiallyinmolecularpathways[5,6].Thisconservedmolecularpath-waywassuggestedasagenetic‘toolkit’forrepeatedevolu-tionofsocialbehavior[6].Sincemanyofthesebehavioralpathwaysaresharedacrossvertebrates,weexaminedtheMexicanteleostAs-tyanaxmexicanus,whichconsistsofbothcave-dwellingandsurface-dwellingpopulations.Notably,thecavemorphshavesignificantlydivergedfromthesurfacemorphsinmultiplebehaviorsduringevolution.Suchtraitsincludelossofschooling(reducedsocialinteraction),per-formingrepetitivebehaviors,sleepdeficits,hyperactivity,behavioraladherencetoaparticularvibrationstimulusat40Hz(behaviorallyadheretoaparticularstimulus,calledvibrationattractionbehavior,orVAB[19]),andhighercortisollevels(relatedtohigheranxietylevels)[20].Unlikecavefish,theconspecificsurface-dwellingpopulationsreadilyschool,donotexhibitrepetitivebehaviororhyper-activity,havenormativesleep,donotshowstrongadher-encetoavibrationstimulus,andhavelowercortisollevelsthancavemorphs.Thispolymorphicsuiteoftraitsthatarepresentinoneecotypeandabsentinanotherisrarewithinthatnaturalworld.Surprisingly,manyofthesecavefishbehaviorsoverlapwiththecoresymptomsofahumanpsychiatricdisease,autismspectrumdisorder(ASD)(e.g.,reducedsocialinteraction,performingrepeti-tivebehavior,sleepdeficits,hyperactivity,adherencetoaparticularstimulusorobject,andhigheranxietylevel[21–23]).Inaddition,manyoftheseASD-liketraitsincavefishshowlargevariations,rangingfromthelevelsofsurface-tocave-type[19,24,25].Thisisreminis-centofthelargevariationwithinASD:fromseveretohighfunctioning[22].Accordingly,thebehavioralsimilaritiesandthehomologiesofthevertebratener-voussystemmotivatedustoinvestigatewhethershiftsinorthologsofASD-riskgenesmayunderpintheevolutionofmultiplebehaviorsofbothhumansandateleostspecies.Thisstudy,therefore,seekstoanswerthefollowingquestions:(1)AreASDriskgenesinhumansandA.mexicanusexpressedinsimilardirectionswhencompar-ingASDpatientswithhumancontrols,andcavefishwithsurfacefish?(2)DohumanASD-riskgenesexhibitsignaturesofmolecularevolutionincavefishthataredi-vergentfromtherestofthegenomeandmayindicateselection?(3)Docavefishrespondtopharmacologicaltreatmentsforautisminasimilarwayaspatients,suggestingasharedneuralbasisintheregulationofASD-likebehaviors(e.g.dopaminergic,serotonergic,adrenergiccircuits)[26–29]?ResultsWefirstqueriedtheAstyanaxgenometoidentifyortho-logsoftheASD-riskgenes,whicharelistedinthedata-baseofSimonsFoundationAutismResearchInitiative(SFARI)(sfari.org)[30].Wefoundthat92.5%of493humanASD-riskgenes(SFARIGenedatabaseCategory1to4andCategoryS—highevidencedASD-riskgenes—accessedinMarch2017[20,30–32].SeeMethods)haveorthologsintheAstyanaxmexicanusgenomev1.2(Table1,Additionalfiles1and2).Inhumanstudies,someASD-riskgenesexhibitdiffer-entialexpressionbetweenpeoplewithandwithoutASD[33–35];thus,weanalyzedgeneexpressiondifferencesbetweencavefishandsurfacefishusingapreviouslypublishedRNAseqdatasetforA.mexicanus[36,37].ThisdatasetincludesgeneexpressiondatacollectedfromwholeindividualsofbothsurfacefishandPachóncavefishatkeydevelopmentaltimepoints:10hpost-fertilization(hpf;endofthegastrulation),24hpf(endofsomitogenesis;hatching),36hpf(livewithyolk)and72hpf(mostoftheorgans,includinggutandjaw,havedeveloped)(GenBankSRA;accessioncode:PRJNA258661[31,36,38]).SinceASDsymptomsinhumansemergeatanearlydevelopmentalstage(evenbefore1–2yearsold[32,39]),weinvestigatedboththeinteractionofage×morph(surfacefishandcavefish)foralltimepointsandtheexpressiondifferencebetweenmorphsat72hpfinASD-riskgenes.Wehypothesizethatthis72hpfisacomparabletimepointto‘justbeforebirth’inhumanswhenthebasicneuralcircuithasbeenformedandisreadytoprunesynapsesandrewiretoformtheproperneuralcircuitsinresponsetoenviron-mentalstimuli[40,41].Remarkably,genesinthecategorieswithstrongerevidenceofassociationwithASDinhumans(Cat-egories1and2inSFARIGene),weremoreoftensig-nificantlydifferentlyexpressedbetweensurfacefishandcavefishthanincategorieswithweakerevidenceofassociationwithASDinhumans(Categories3,4inSFARIGene)(Table1,Additionalfiles1and2).Thistrendwasobservedinboththeinteractionofage×morph(Table1:Cat.1–2range65–72%vs.Cat.3-Srange57–63%)andtheexpressiondifferencebetweenmorphsat72hpf(Table1:Cat.1–2range68–94%vs.Cat.3-Srange63–71%)oforthogroups.Note,anorthogroupconsistsofmultipleparalogsthatsharethesameancestorwitheachhumangene[42].ThistrendofhigherratesofdifferentialexpressionforgenesincategorieswithstrongerevidencewasalsoseenatthelevelofindividualA.mexicanusgenes(i.e.,paralogs;Table1,Additionalfile1).Thissuggeststhat,althoughparalogsevolvedfromgeneduplicationeventsandmaybeunderdifferentexpres-sionregulation,differentiallyexpressedorthologsofYoshizawaetal.BMCEvolutionaryBiology (2018) 18:89 Page2of16

PAGE 3

ASD-riskgenesaremorecommoninthehigherconfidenceSFARIgenes.Totestwhethertheobservedlevelsofdifferentialex-pressionfororthologsofCategory1and2aresignifi-cantlyhigherthanforarandomsubsetofgenesacrosstheAstyanaxgenome,weperformedbootstrappingusing9999randomsamplingsof500genesfromthe22,767geneswithexpressiondataoutof23,042totalgenesinthegenome(Ensembl.orgAssembly;Ast-Mex102;GenebuildlastupdatedJuly2016[31,43–45]).Thenumber500waschosenbecauseweused493SFARIgenesinouranalysis.Ofthisrandomsamplingofasubsetofgenes,ourresultsindicatedthat48.0±4.2%(mean±95%confidenceinterval)hadasignificantage×morphinteraction,and49.0±4.4%weresignificantlydif-ferentiallyexpressedbetweencaveandsurfacefishat72hpf(Table1).Incontrast,wefoundthat61.1and60.7%ofCategory1and2genes,respectively,exhibitedasig-nificantage×morphinteractionand75and64.3%ofCategory1and2genes,respectively,weredifferentiallyexpressedat72hpf.ThisresultindicatesthatcavefishorthologsofhumangenesinCategory1and2ASD-riskgenesareenrichedfordifferentialexpressionbetweencaveandsurfacefish(>99.9percentileofbootstrappingprobability).ToevaluatewhetherthisobservedgeneenrichmentisspecifictoASDorapplicabletootherpsychiatricdis-eases,wealsoexaminedgenesinvolvedinschizophrenia(SCZ),whichsharesmanysymptomswithcavefishandASD[46].OnedatabaseforSCZ-riskgeneslists44genesasbeingtightlyassociatedwiththedisease,andanothermorerecentdatabasecontains304genes(www.szgene.organdwww.szdb.org,respectively).Un-likeASD-riskgenes,theseSCZ-riskgenesdonotshowenrichmentfordifferentialexpressionbetweencavefishandsurfacefishcomparedwiththerandomsamplingofgenesubsets(oftheSCZ-riskgenesinszdb.org,39.8to55.2%showsignificantlydifferentexpressionbetweencavefishandsurfacefish,Additionalfile3).Thus,differ-entialgeneexpressionbetweencavefishandsurfacefishappearstohavemoresimilaritiestoASDthanSCZ.TheASD-riskgenesareincludedintheSFARIdatabasemainlybasedongeneticassociationstudies(evaluationofgeneticvariationinhumancohorts),inwhichtheexpressiondirection—downorupregula-tion—isnottakenintoaccount[30,45,47].Thespecific Table1TheenrichmentoftheexpressionshiftsbetweensurfacefishandcavefishinASD-riskgenesHumanASD-riskgenesCavefishgenesRiskcategory#ofListedGenesinSfari.org%(#)humanASD-riskgeneswithcavefishorthologs%(#)oforthogroupsthatshowsignificantage×morphinteraction†%(#)oforthogroupsthatshowsignificantexpressiondifferencebetweenmorphsat72hpf†%(#)ofallparalogsthatshowsignificantage×morphinteraction†%(#)ofallparalogsthatshowsignificantexpressiondifferencebetweenmorphsat72hpf†Category11994.7%(18of19)72.2%(13of18)94.4%(17of18)61.1%(>99.9percentile)(22of36)75.0%(>99.9percentile)(27of36)Category24393.0%(40of43)65.0%(26of40)67.5%(27of40)60.7%(>99.9percentile)(34of56)64.3%(>99.9percentile)(36of56)Category313995.0%(132of139)60.6%(80of132)70.5%(93of132)50.3%(84.8percentile)(95of189)58.2%(>99.9percentile)(110of189)Category424489.8%(219of244)62.6%(137of219)63.0%(138of219)51.9%(96.1percentile)(167of322)52.5%(94.4percentile)(169of322)CategoryS(notalreadyincludedamongCat1–4genes)4897.9%(47of48)57.4%(27of47)63.8%(30of47)44.2%(4.7percentile)(34of77)49.4%(59.0percentile)(38of77)Total49392.5%(456of493humangenes)62.1%(283of456)66.9%(305of456)53.8%(352of654)54.6%(357of654)Bootstrappingscore:mean±95%confidenceinterval48.0±4.2%49.0±4.4%The72hpost-fertilization(hpf)representsthestageinwhichfishhavehatched,buthavenotyetdevelopedaswimbladderandthejawisunderdeveloped,comparabletothelateembryonicstageofmammals[111].†P<0.05afterBenjamini-Hochbergadjustment.Percentilesinthetablesarefrom9999-bootstrappedvalues.SF:surfacefish.CF:Pachóncavefish.Hpf:hourspostfertilization.Weavoidedtestingthedifferencesbetweenmorphsineachdevelopmentaltimepoint(i.e.10,24and36hpf)duetosavestatisticalpower.Thenumberofgenesareindicatedinparentheses.SeealsoAdditionalfile1foreachstatisticaltestofagexmorphandexpressiondifferenceat72hpfYoshizawaetal.BMCEvolutionaryBiology (2018) 18:89 Page3of16

PAGE 4

directionofexpressionofgenesinapathwayornetworkcanprovideevidenceastowhetherthesemolecularpathwaysarestrengthenedorattenuated(co-expressionnetwork)[33,35,48].Toaddressthis,wecomparedthedirectionofgeneexpression(upordown)betweencave-fishversussurfacefishwiththatobservedinASDpatientsversuscontrols,byutilizingthehumanbraintranscriptomeofpost-mortemASDpatients[33–35].Ofthe58and3442humanorthologsthatexhibitedsignifi-cantlydifferentgeneexpressionbetweenASDpatientsandcontrolsintwodifferentstudiesthatbothusepost-mortemcortices(Voineaguetal.2011andParikshaketal.,2016respectively)[33–35],74.1to74.6%werealsodifferentiallyexpressedbetweencavefishandsurfacefish(Table2,Additionalfile4).Amongthedifferentiallyexpressedgenes,58.5to60.7%ofcavefishgenesshowedthesamedirectionofex-pression(i.e.cavefishrelativetosurface)asthehumangenesfromtheASDtranscriptomestudies(i.e.ASDpa-tientsrelativetocontrols)(Table2,Additionalfile4)[33–35].Thisincludesthedown-regulationofdistal-lesshomeobox1(DLX1)andtheup-regulationofbagfamilymolecularchaperoneregulator3(BAG3)andchlorideintracellularchannelprotein1(CLIC1)cavefishortho-logsrelativetosurfacefish,whichconsistentlyshowedsimilarpatternsofexpressioninhumanswithASDrela-tivetocontrols[33,35].IncontrasttothehighpercentageofASDorthologsthatshowdifferentialgeneexpressionbetweencavefishandsurfacefish(Table2),otherASDmodels—includingaclassicASDmousemodel(BTBRmouse)[49],thebloodcellsofASDpatients[50],andtheneuronsde-rivedfrominducedpluripotentstemcell(iPScell)ofASDpatients[51]—exhibitedmuchlowerconcordancewithhumanbrainASDtranscriptomicstudies(Table2).ForseveralorthologsofCategory1SFARIgenes,wealsoobservedexpressiondifferencesbetweenthebrainsofsurfacefishandcavefishbyquantitativeRT-PCR(Additionalfile5)inlaterdevelopmentalstages(from1monthto1yearold).Overall,thesetranscriptomicanalysesindicatethatA.mexicanuscavefishandhumanswithASDsharesimilarpatternsofASD-riskgeneex-pressionsthatcouldunderliesharedASD-likebehaviors.Tosurveyadditionalparallels,weexaminedpatternsofmolecularevolutioninDNAsequencesofASD-riskgenes.WetestedwhetherASD-riskgenesinA.mexicanusarehighlydivergent(i.e.,potentiallyunderpositiveselection)betweencavefishandsurfacefish[52].WeidentifiedthenumberofgeneswithintheASD-riskgenesthataredivergenceoutliersbe-tweencavefishandsurfacefishbasedonthreemet-ricsfrompopulationgenetics:(i)top5%ofFSTforallgenesacrossthegenome(indicatesthedifferenceinallelefrequenciesbetweenpopulations),(ii)top20%ofDXY(pairwisenucleotidedifferencesbetweentwopopulations)and/or(iii)P-value<0.05forhapFLK[53](detectsselectionsignaturesbasedonpopulationhaplotypefrequenciesandismoreanexplicittestforpositiveselection,whereasFSTandDXYaremeasuresofdivergence)(Table3,Additionalfiles6and7).SincehapFLKisanexplicittestofselection,weonlyusedhapFLKtotestforenrichmentofASD-riskgenesforpositiveselectionrelativetothegenomeasawhole.Wetestedwhetherthecavefishorthologsof493ASD-riskgenes(SFARIgeneCategory1–4andCategoryS)wereenrichedforgenesidentifiedtobeunderpositiveselectionviahapFLK.WeincludedanyAstyanaxparalogsforASD-riskgenes(seeAdditionalfile7),andusedaFisher’sexacttestwithYate’scor-rection[54]totestwhetherthenumberofASD-riskgeneswithP-values<0.05viahapFLKwasoverrepre-sentedrelativetototalnumberofgenesinthegen-omewithP-values<0.05viahapFLK.WefoundthattheASD-riskgenesareenriched(13.5%)relativetoallthegenesinthegenome(7.3%)forsigna-turesofselectionusinghaplotypefrequency(hapFLK,P<0.05;Table3,Additionalfile7).Thus,ASD-riskgeneswere~2×morelikelytoexhibitasignatureofpositivese-lectionthanwerethegenesinthegenomegenerally(odds-ratio1.94;95%confidenceinterval:1.54–2.45).ASD-riskgenesinhumansarealsohypothesizedtobeenrichedforsignaturesofpositiveselection[55].Sinceweneededtosetthecut-offsforFSTandDXYmeasurestopercentagesofthegenome,enrichmenttestswerenotlogical,andweperformedKruskal-WallisteststoassessdifferencesbetweentheASDgenesetcomparedtoallgenesintheAstyanaxmexicanusgen-omeforFSTandDXY.Forallcomparisons,geneswithnodatawereremoved.NotethathapFLKisbasedoncaveandsurfacecomparisonsacross45Astyanaxsamplesfromfivepopulations(seeMethods),whichincludedadditionalpopulationsthatarebeyondthescopeoftheworkhere.FSTandDXY,however,wereonlyfocusedoncomparisonsbetweenPachóncavepopula-tionandChoysurfacepopulation.AlthoughwedocumentedthattheASD-riskgeneswereenrichedforsignificanthapFLKtestsrelativetothegenome,wefoundthatofthe661CategorySandCategories1–4SFARIgeneswithsequencedata,19weredivergenceoutliersforFST(cut-offtop5%ofallgenesacrossthegenome;Additionalfile7;sevenoftheseweredivergenceoutlierswithmultiplemetrics)and51weredivergenceoutliersforDXY(cut-offtop20%;13oftheseweredivergenceoutliersdefinedbymultiplemetrics;Additionalfile7),whicharebothalowernumberofgenesthanwhatwouldbeexpectedsolelyfromourpercentcut-offmetrics(Note:FSTandDXYplotsforYoshizawaetal.BMCEvolutionaryBiology (2018) 18:89 Page4of16

PAGE 5

divergenceoutlierswhichmeetatleasttwometricsareshowninAdditionalfile6).ASD-riskgenesthatpassouroutlierthresholdforFST(top5%ofallgenesgenome-wide)areonaveragenomoredivergentbetweencavefishandsurfacefishthangenome-widenon-ASD-riskgenesthatalsopassedtheoutlierthresholdFST(top5%)(FST:Kruskal-Wallistest;X2=1.727,df=1,P-value=0.189,Additionalfiles6and7).WhencomparinggenesthatdidnotpassourFSTthreshold(i.e.,thebottom95%ofFSTvaluesacrossthegenome),ASD-riskgenesweresignificantlymoredivergentthannon-outlier,non-ASD-riskgenes,buttheeffectsizewassmall(meanASD-riskgenes=0.20,meannon-ASD-riskgenes=0.23,FST:Kruskal-Wallistest;X2=4.781,df=1,P-value=0.029,Additionalfiles6and7).Thus,wehavelittleevidencethatASD-riskgenesareonaveragemoredivergentforFSTthannon-ASD-riskgenes.Interestingly,wefoundevidencethatdivergenceoutlierASD-riskgenesforDXYweremoreconservedthandiver-genceoutliernon-ASDgenes.DXYoutlierASD-riskgeneswere~0.5×lessdivergentbetweencaveandsurfacefishthanoutliernon-ASD-riskgenes(DXY:Kruskal-Wallistest; Table3GenesetenrichmentanalysisbasedonFisher’sexacttestwithYate’scontinuitycorrectionDivergencemetricsComparisonsSignificantASDgenesTotalASDgenesSignificantgenesingenomeTotalgenesingenomeYate’sChiSquareX2DegreesoffreedomYate’sP-valuehapFLKWholegenome86635a164822,71034.5571<0.0001GeneenrichmentforASDgeneswasbasedoncomparisonsbetweenChoysurface(Choy)andCave(Pachón).Calculationswereperformedusingthechisq.testfunctioninRaNotethattotalASDgenesforhapFLKislowerthanthetotalfororthologsandparalogsduetomissingdata Table2Directionofgeneexpression(up-ordown-regulated)inthisandpreviouslypublishedstudies(cavefishcomparedwithsurfacefish,andcasescomparedwithcontrols)#orthologsofSFARIGenethatexpressdifferently(upordown)TranscriptomefromthecorticesofASDpatients(Voineaguetal.,2011)TranscriptomesfrommultipletissuesofASDpatients(review:Anseletal.,2017)TranscriptomefromthecorticesofASDpatients(Parikshaketal.,2016)A.mexicanus(WholeEmbryo10–72hpf)Reanalyzedinthisstudy335of409†(81.9%)43‡of58orthologsexpresseddifferentlybetweenCFandSF(74.1%)51‡§of77orthologsexpresseddifferentlybetweenCFandSF(66.2%)2567‡of3442orthologsexpresseddifferentlybetweenCFandSF(74.6%)31of51‡genesexpressedinthesamedirection:(60.7%)27of45‡§directionallyexpressedgenesareinthesamedirection:(60.0%)1843of3152‡directionallyexpressedgenesareinthesamedirection:(58.5%)BTBRMouse(Hippocampus)Provenzanoetal.,201630of493(6.1%)2of65orthologsexpresseddifferentlyinBTBRmouse(3.0%)a9§of105orthologsexpresseddifferentlyinBTBRmouse(8.6%)b216of4042orthologsexpresseddifferentlyinBTBRmouse(5.3%)c0of2genesexpressedinthesamedirection(0.0%)4of7§directionallyexpressedgenesareinthesamedirection(57.1%)109of216directionallyexpressedgenesareinthesamedirection(50.5%)ASDpatients(Blood)Pramparoetal.,201551of493(10.3%)11of65humangenesexpresseddifferentlyinASDpatient’sblood(16.7%)d16§of107humangenesexpresseddifferentlyinASDpatient’sblood(14.8%)e485of4425humangenesexpresseddifferentlyinASDpatient’sblood(11.0%)f7of11genesexpressedinthesamedirection(63.6%)4of7§directionallyexpressedgenesareinthesamedirection(57.1%)179of485directionallyexpressedgenesareinthesamedirection(36.9%)NeuralcellsderivedfromiPScellofASDpatientMarianietal.,201595of493(19.3%)18of65ofhumangenesexpresseddifferentlyinneuronsderivedfromiPScellsofASDpatients(27.7%)g21§of107humangenesexpresseddifferentlyinneuronsderivedfromiPScellsofASDpatients(19.6%)h527of4425humangenesexpresseddifferentlyinneuronsderivedfromiPScellsofASDpatients(11.9%)i0of18humangenesexpressedinthesamedirection(0%)1of16§directionallyexpressedgenesareinthesamedirection(6.3%)163of527directionallyexpressedgenesareinthesamedirection(30.9%)†84of493orthologswerenotfoundintheAstyanaxgenebuildat2016(Ensembl.orgAssembly:AstMex102;GenebuildatJul2016)‡SomeorthologsofhumangeneshavemultipleparalogsinA.mexianus(i.e.shank3aandshank3b)§Weexcludedgenesthatshowedinconsistentexpressiondirectionsbetweenmultiplereports(i.e.up-regulatedinonepaperbutdown-regulatedinanother[34])2testsfordifferentiallyexpressedgenesagainsttotalgenesbetweenA.mexicanusandBTBRMouse(a:2=30.2,P=1.17×107;b:2=31.3,P=6.57×108;andc:2=1785.6,P<1.0×1010),betweenA.mexicanusandpatients’blood(d:2=14.9,P=3.48×104;e:2=21.7,P=9.40×106;andf:2=1445.5,P<1.0×1010),andbetweenA.mexicanusandpatients’iPScells(g:2=8.1,P=0.0136;h:2=16.3,P=1.66×104;andi:2=1377.2,P<1.0×1010).Allofthesetestshavedf=1.P-valuesweremultipliedbythenumberofthetests(Bonferronicorrection)Yoshizawaetal.BMCEvolutionaryBiology (2018) 18:89 Page5of16

PAGE 6

X2=29.285,df=1,P-value<0.001,Additionalfile6).TheseresultssuggestthatASD-riskgenes,whichfallwithinouroutliercut-offsforDXY,maybeexperiencingpurifyingselectionrelativetooutliergenesintherestofthegenome.Incontrast,non-outlierASD-riskgenesandnon-outlier,non-ASD-riskgenesintheremainderofthegenomedonotdifferintheirlevelofdivergenceforDXY(DXY:Kruskal-Wallistest;X2=0.046,df=1,P-value=0.830,Additionalfiles6and7).ConsideringthatmanyASD-riskgenesinhumanwerefoundasconstrained[56,57],thisfindingindicatesanothersimilaritybetweencavefishandhumanintheevolutionofASD-genes.AnalysiswithIngenuityPathwayAnalysisComparisonAnalysis(IPA)[48]highlightsafurtherpotentialrela-tionshipbetweentheevolutionofASD-riskgenesandcavefishtraits(Additionalfile8).Whilemanyfunctionalcategoriesareenrichedfornon-outlierandoutlierASD-riskgenes,somefunctionalcategoriesareonlyenrichedinoutlierASD-riskgenes(Additionalfile8).Incomparisontonon-outlierASD-riskgenes,outlierASD-riskgenes(definedbytop5%ofFST,top20%ofDXY,and/orsignificanthapFLKscores)areenrichedforfunctionsthatincludeauditorydisease,digestivesystemdevelopmentandfunction,inflammatorydiseases,lipidmetabolism,ophthalmicdisease,andothers.Thesere-sultswereconsistentevenwhenweimposedamorestringentcut-offforDXY(top5%).Manyofthesefunc-tionalcategorieshavebeenobservedasco-morbidsymptomswithASD[58,59](Additionalfile8).Thesefunctionalcategoriesmapwelltophenotypeslikelyunderselectionincavefish[20],aswellasknownsymp-tomsinASDpatients.MultiplecavefishorthologsofASD-riskgenesoverlapwithknownquantitativetraitlocus(QTL)intervalsforbehavioralandsensorytraits(Fig.1,Additionalfile9).TheASD-riskgenesthataredivergenceoutliersbyoneormoredivergencemetricsandareunderpreviouslymappedQTLareabca5,cacna1fb,chd7,dock8,erbin(i.e.errbb2ip),grip1,hdac4,pah,pax6,plxna4,scn1a,slc1a2b(Fig.1).ManyofthedivergentoutliergenesarenotunderQTL,whichmaybebecauseofthefragmen-tednatureofthecurrentgenomesequenceofA.mexi-canus.However,thisinitialanalysisrevealedthatmanyoftheoutliergenesunderQTLforeyesize,aminoacidresponse,andtastebudnumberaremembersoftwomajorgenenetworksthathavebeensuggestedtobein-volvedinASD:synapticfunction(cacna1fb,dock8,erbin,grip1,plxna4,scn1a,slc1a2b)andepigeneticregulation(chd7,hdac4)[23,56,60].Consideringthateye-sizeisassociatedwithadherencebehavior[19]andchemosen-soryorganscanmodifywakefulness[61],thisresultim-pliesthatsomeoftheputativelyselectedgenesincavefishmayalsobeassociatedwithASD-likebehavioralphenotypes.TheresultspresentedaboveindicatethattheASD-riskgenesincavefishandhumansshareevolutionaryandgeneexpressionsignatures.Additionally,wesoughttounderstandifcavefishandASDpatientsrespondsimilarlytodrugsusedtotreatASD.Accordingly,wetreatedA.mexicanuswiththeU.S.FoodandDrugAdministration-approvedASDdrugs,aripiprazoleandrisperidone,andclassicantipsychoticdrug,clozapine[62,63].Thesedrugsactasagonistsand/orantago-nistsofmultiplereceptorsfortheneurotransmittersdopamine,serotonin,histamine,adrenalin/noradrena-lin,and/oracetylcholine[62,63].Wealsotreatedse-lectiveserotoninreuptakeinhibitor,fluoxetine,andanopioidblocker,naltrexone,whichareusedforASDpatientsunderthephysicians’direction[62–64].Wefoundthattreatingcavefishwitharipiprazole,risperidone,clozapineandfluoxetinemitigatedASD-likebehaviorsincavefish.Thedrugssignificantlyreducedadherencetoaparticularvibrationstimulus(Fig.2a,b,d),significantlyre-ducedhyperactivity(swimmingdistanceinFig.2f,h,i)andincreasedsleepduration(Fig.2k,m,n)(seeAdditionalfile10),whicharesimilartotheresponsesob-servedinASD-patients[62–64].Incontrast,thesedrugsshowedlittleeffectonsurfacefishbehaviors(Fig.2).Thedrugnaltrexone,anopioidblockerthatcanmitigatehyper-activityandrestlessnessbutnotthecoresymptomsofASDanddidnotchangecavefishbehaviors(Fig.2e,j,o)[64].ThesepharmacologicalstudiesindicatethatcavefishmaysharesimilarneuralpathwayswithASDpatientssincechemicalinterventionalterssimilarbehaviors.DiscussionFordecades,evolutionarybiologistshavebeeninterestedinunderstandinghowanimalsevolvemultiplebehaviorswhosegeneticbasesarefrequentlyindependentandcom-plex.HereweshowthatcavefishandASDpatientsexhibitsimilaritiesinexpressiondirectionamongASD-riskgenes,evolutionarysignaturesforASD-riskgenes,andresponsestoASDsymptom-treatingdrugs.Theseoverlapsmayindi-catepotentialutilizationofasuiteofgenes(ASD-riskgenes)fortheevolutionofASD-likebehaviorsinbothhu-manandcavefish.TheseASD-riskgenesarealsoknowntobeinvolvedinsocialbehaviorofhoneybeesandmam-mals[5],suggestingthatanimalsmayfrequentlymodifyASD-riskgenesintheevolutionofbehaviors.ModificationstomostASD-riskgenesmaynotresultinacutedeleteriouseffects(exceptforsomeASD-riskgenescontributingtocoreneuralactivitiesthatareoftenseenindenovomutation)[23,65,66].Indeed,inhumans,manyofthecommonvariantsforASD-riskgeneshavesmalleffectsthatmightmodifybrainsystemsmoresubtly;insomecases,thesemodificationsmayevenproviderathersmallbeneficialeffects[55,67].Accordingly,ithasbeenreportedthatASDpositivelyYoshizawaetal.BMCEvolutionaryBiology (2018) 18:89 Page6of16

PAGE 7

correlateswithchildhoodintelligence,collegecomple-tion,andyearsofschooling;inaddition,humanASD-riskgenesshowenrichmentforsignaturesofpositiveselection[55].Manyofthesesmall-effectal-lelesarethoughttoprovidecumulativeeffectsthatcanleadtoASD[68].Incontrasttothesecommonsmall-effectvariants,ASD-riskgenescontributingtocoreneuralactivitiesarefrequentlyfoundasrareal-lelesand/ordenovovariants,andonlyafewofthesevariantsarethoughttobeenoughtopromoteASD[21,23].InA.mexicanus,thequantitativetraitloci(QTL)mappingofcavefishbehaviors—adherencetovibrationstimulus,lossofschoolingandlossofsleep—showedsmall-effectsizedQTLornodetect-ableQTL[20].Wethereforeconsiderthatcavefishismoresimilartothecommonvariant-inducedASDthanASDinducedbyrareordenovovariants.WeexpectthatsomeofthecavefishASD-riskgenesunderQTLmaypotentiallyyieldasmallbeneficialeffectinthecaveenvironment(Fig.1). CYP11B1 MBOAT7SCOSPONDIN STYK1 RAD_17501 Am213d Am30bmc1r Am101a Am15c Am26cAm101bAm106a tph2Am123bAm216aAm229e Am120a RAD_47726RAD_89057 RAD_28591RAD_71383 5 RAD_20136RAD_71744 RAD_57607 RAD_28708 RAD_47914 RAD_98033 Am205e RAD_78967Am210a RAD_43116RAD_83369 RAD_14016 RAD_44245 oca2 Am132b Am218eAm43c shroom2 mchr1 RAD_82059 RAD_11443RAD_15266 RAD_38026RAD_85454Am30c Am105b 5ht2ARAD_95549 RAD_43090RAD_75134RAD_12122RAD_66364Am213f Am26a RAD_29965 RAD_91463 RAD_88030RAD_60780 RAD_52152 RAD_20037Am237a Am213e RAD_48253 RAD_56285 RAD_83582 Am16aAm33b RAD_37846 RAD_5195 RAD_53058 RAD_45260 1 RAD_87009RAD_9075 RAD_9536 RAD_85655 5ht1A Am43a RAD_25005 RAD_43143 RAD_98090 RAD_45734 RAD_16332RAD_73745 RAD_24659 RAD_29437RAD_95564 RAD_38160RAD_94613 RAD_22736 RAD_879215ht2C_like Am131a Am55aAm203eAm104c RAD_75522RAD_40841 RAD_46941 Am19b RAD_61843 RAD_19197 RAD_97170 Am144a Am207c RAD_78693Am218a RAD_31061 7Am214fRAD_38936RAD_75960Am27cRAD_47991 tfe3 RAD_90523 RAD_22966 RAD_66439 RAD_19643 RAD_75942RAD_43329 Am23b Am2b RAD_6124 RAD_3350 RAD_29967 Am209d Am145a Am239d RAD_12396RAD_52647 Am217d RAD_49153 14 RAD_49779 Am206a Am240b Am209e Am207e RAD_16578Am220a 20 Am3c Am3b RAD_91569 RAD_76294 RAD_85775 RAD_53453Am8c RAD_68183 RAD_72140 RAD_53343 RAD_79020Am227e 23 Am219b RAD_17299RAD_45739 pmchth2 Am111b Am219eAm226bRAD_33183 RAD_101042 RAD_18104 RAD_86740 RAD_98087RAD_98146 RAD_47054 Am224c RAD_44155 Am202bAm210d RAD_97807 RAD_68023 Am36d RAD_8919RAD_46535RAD_62524 RAD_11089 RAD_80900 RAD_89017Am221b Am112a RAD_67695 th RAD_3816Am113b Am44c mchr2 RAD_96195 RAD_53852 3Am222cRAD_79893RAD_98064RAD_81830RAD_45928 Am1b RAD_17772 Am130a Am142b RAD_44407 RAD_72042 Am122bAm207b Am215f Am231c RAD_26544 RAD_80615 Am215b Am221a RAD_80833 RAD_97104 RAD_16726 RAD_90557 RAD_49302 9Am09c Am3a RAD_48554 RAD_63219 RAD_8678 Am113c RAD_75211 RAD_35696Am119a RAD_4708Am222eAm227a RAD_31443 Am207d 16 Am39c RAD_82521 RAD_62716Am225b RAD_85599RAD_62994 RAD_36097 RAD_26997 RAD_21287RAD_63960RAD_22157RAD_54508 RAD_38799RAD_52882RAD_77528RAD_87988RAD_10852RAD_38716RAD_51590RAD_69886RAD_94022Am201eAm206b RAD_55567RAD_6621RAD_69332 RAD_74420 RAD_47069 RAD_84766 RAD_82106 RAD_16972 RAD_100562 RAD_64892 RAD_50884RAD_66630 25 RAD_92930RAD_80950 RAD_23993RAD_16601 RAD_87397 RAD_67181 PAX6 Am211a Am111c NYU30 RAD_45540 Am228cAm237c Am241dRAD_24181 RAD_31901 RAD_26916RAD_57718Am208d RAD_80019 Am107c RAD_59620RAD_72551 RAD_98329 RAD_47674 RAD_22219RAD_93450 Am231aRAD_66363Am235a RAD_28027RAD_36772 RAD_80188RAD_43142 6 Am216e RAD_72696shh Am210bAm230c RAD_16106RAD_20796 RAD_15432RAD_4491 3 RAD_75186RAD_7348 4 RAD_47274 RAD_41169RAD_41346RAD_81214RAD_99487 RAD_52346RAD_68409RAD_89917 RAD_33084 RAD_96608 RAD_53947NYU25 Am204d2 5ht2C 239e NYU01 13 RAD_78478 Am55bAm135c Am235b Am131c RAD_68111RAD_24567RAD_59957RAD_83812RAD_92195 RAD_36827 RAD_72171RAD_34877 RAD_24898RAD_77393sert Am208c Am202a Am217eAm240e Am224f RAD_83767 RAD_19702RAD_9988 3 12 RAD_92953 RAD_39547 RAD_52196 RAD_39360 Am11bAm226fAm229c RAD_87388 RAD_33773RAD_36939 Am203d Am136a Am42a RAD_31776RAD_19127RAD_19535 19 Am115c RAD_41764RAD_55145RAD_60158RAD_82798 RAD_13133RAD_8025 7 RAD_95416 RAD_37425RAD_57878RAD_62629RAD_73613RAD_97116Am228d RAD_94251 RAD_85634Am220e RAD_79171 RAD_73582 24 RAD_61685 RAD_27866Am33c RAD_33752Am216bAm222d Am120b Am204dAm204e Am135a Am142a RAD_27287 Am125c Am17bAm216f Am120c Am143c Am108c RAD_13262RAD_2536 RAD_55528Am203f RAD_92414RAD_80647 RAD_69113 RAD_68586 RAD_42043 Am110b RAD_19081Am103b Am119c RAD_79105 RAD_22542 eya1 2 RAD_67705RAD_21879 Am214bAm204fAm236b RAD_68234 RAD_52214RAD_86594 RAD_13173RAD_41073RAD_5656RAD_63406 RAD_34985RAD_36775 RAD_43022 Am218bAm37c RAD_44888 RAD_74554 RAD_87138RAD_94131 RAD_75955 RAD_31719 Am24bmc2r Am205f Am202dAm213b RAD_54293 Am23cRAD_53778 RAD_28529 8RAD_91791Am114aRAD_70946RAD_32846Am242aAm24cAm205a RAD_81138 Am232d Am226a RAD_64032RAD_6112 Am104b RAD_53800 RAD_26570RAD_80962 RAD_93616 RAD_15815 RAD_83571 RAD_99634 RAD_37299 RAD_50212 Am214e RAD_55497 15 NYU03 Am229bcryaa Am235dAm119b RAD_83939 RAD_89962 RAD_55110 RAD_90815 RAD_40480 RAD_27434 RAD_29233RAD_57134RAD_67007 21 RAD_33090 RAD_15502 RAD_39047 RAD_10746 RAD_38622 pfn2 5ht2B RAD_46723RAD_49749Am230e Am5c RAD_22130 RAD_66607Am237b 22 Am241cAm131b Am133bAm209b Am44bRAD_84956 RAD_9420 RAD_31844RAD_50614RAD_75597 RAD_70163RAD_66015 RAD_45265RAD_82002Am226d Am39aAm128aRAD_16205RAD_48750RAD_74656RAD_79497RAD_88212 RAD_95530RAD_77720 RAD_89262 Am209c Am16c RAD_79590RAD_9218 RAD_12916RAD_80517Am211c Am136b RAD_40023RAD_49979RAD_67706RAD_79353 RAD_90665 RAD_89065 RAD_70198 Am112b Am217b Am235eRAD_89910 Am6aAm117c RAD_11793 RAD_4424RAD_45328RAD_53036 4 RAD_70732 Am17c Am214a Am143a Am211e RAD_31051 Am103c Am227dAm228e RAD_70750 RAD_51215RAD_65638Am207aAm139b RAD_11021RAD_14743 10 RAD_27613 RAD_93780RAD_6611RAD_8992RAD_9320 RAD_32290 Am7b RAD_55775 RAD_25464 RAD_100015 Am236a RAD_85408 Am142c Am205dAm208e Am17a RAD_31011RAD_59634 RAD_15817Am220bRAD_70147 RAD_54124RAD_35861 RAD_65881 RAD_17072 RAD_11922RAD_86461 RAD_61584 RAD_33205RAD_35532 RAD_60079RAD_65504 RAD_20479 RAD_32109 RAD_33977 RAD_14070Am103a RAD_92588RAD_65631 RAD_80695 Am207f Am219a RAD_59676 RAD_83065 RAD_97151 RAD_67741RAD_97716 RAD_57763 RAD_75031 RAD_63022 17 (Yoshizawa et al., 2012bKowalko et al., 2013b) Eye sizeAmino acid responseTaste bud number(Protas et al., 2008) Eye sizeVABEO SN numberFeeding angle NSD1OLFM1ADAMTS2OLFML2SCN1aBRCC3HTR2cRIPPLY1MPPED6ADAM9OLFML2SPASTLMX1BSMARCA2BZRAP1FBN1NRXN3bOPHN1STXBP5GTF2lPPP2R1NF1bNIPBLBAsxl2Chd8Dyrk1aaPOGZaPTENbSCN1LabSETD5SHANK3SHANK3bSYNGAP1aANKRD11BCL11aaCACNA1HaCACNA2D3Ube3aCNTN4CNTNAP2aCTNND2aCTNND2bERBB2IPGABRb3GRIP1ILF2IRF2BP2KDM5bbKMT2cKMT2cbMETMYT1LaNLGN3aDRD3SLC6A1lWACbARXBRAFCACNA1CCHKbDHCR7EHMT1bTSC1bHCN1HEPACAMBCKDKMBD5GRIP2aNFIXbNR2F1aOCRLSLC9A6UPF3bPAX6PCDH19PRODH2PIK3R1DSTREREbARXaSLC9A6GRIN2B RAD_68774RAD_69693 RAD_88035 RAD_79830 RAD_5037 RAD_66300 RAD_40327 RAD_48716 Am214c RAD_81018RAD_33902RAD_53327 RAD_9864 RAD_81618Am229d Am226e Am214dRAD_39920RAD_85641RAD_91430RAD_13265 RAD_26335 RAD_29543Am6c RAD_61901Am216d RAD_58136 Am123a RAD_81324 RAD_3536 RAD_51379 RAD_28781RAD_69609RAD_9115 RAD_49522 18 ADNPbCDKL5PTPN11bREREa TbudLensMelADNPb ONLEyeTbud LensMelTeeth PTCHD1SEMA5APLXNA4SLC1A2bDEAF1ADK RIMS1aKHDRBS2GABRA4ATP1A3bSCN1A RAD_37221Am118c RAD_48937 RAD_70351 RAD_11927RAD_61566RAD_9562 Am114bmao RAD_25262RAD_87332 RAD_3486 RAD_33301RAD_61695 RAD_58175 RAD_86010 RAD_11257RAD_34713RAD_4979 RAD_51120 RAD_59730RAD_87881RAD_95271Am127a RAD_70721 RAD_88707 RAD_28710 RAD_21502Am106b RAD_69027 Am206e RAD_65270 RAD_70453 Am41cigfbp5RAD_19903RAD_93042 RAD_30959RAD_37938 RAD_23092RAD_78809 Am215dRAD_39121 RAD_73250 Am19c RAD_57717 RAD_44307 RAD_18293RAD_19878RAD_4051RAD_80482 11 HDAC4MYO16CHD7CACNA1IGSTM1MYT1LPLCB1 SNTG2SNX14CCT4 NFIXaSPASTCACNA1FbDOCK8 IQSEC2HYDINSUV420H1(KMT5B)SLC38A10 CACNA1G EP300aABCA5NRXN1TNRC6B MKL2a AHDC1CHD9MARK1 Fig.1ThecongruencybetweenquantitativetraitlociandASD-riskgeneshighlightspotentialgenetichubsforgeneregulation.Linkagemapconstructedfrom115F2hybridprogenyofacrossbetweenasinglesurfacefishfemaleandasinglemalePachóncavefish.Themapincludes699markersassembledinto25linkagegroupsthatcollectivelyspan1835.5cM.ColoredbarsrepresentapproximatepositionofQTLforeyesize,chemical(aminoacid)sensingability,tastebudnumber,VABlevel,andthenumberofmechanosensorysuperficialneuromastattheeyeorbit(EOSN)asindicated[74,109].Lens:lenssize,Mel:melanophorenumber,Teeth:teethnumber,Eye:eyesize,Tbud:tastebudnumber,ONL:thicknessintheouternuclearlayerofretina[110].Eachlinkagegroupisannotatedwithgenomicmarker(rightside)andanchoredASD-riskgenes(leftside).BluecharactersingenomicmarkersaretheonesthatsharethesamegenomicscaffoldastheASD-riskgenesontheleftside.RedcharactersinASD-genesaretheonesthatshowthesignaturesofdivergenceshowninAdditionalfile7.Othergenes(attheleft)aresuccessfullyanchoredCategory1and2SFARIGenes(Additionalfile9,alsoAdditionalfile1)Yoshizawaetal.BMCEvolutionaryBiology (2018) 18:89 Page7of16

PAGE 8

TheIPAanalysisrevealedthat,comparedtonon-selectedASDgenes,positivelyselectedand/orhighlydivergentASD-riskgenesincavefishareenrichedinthepathwaysofdigestivesystemdevelopment/function,in-flammatorydiseases,lipidmetabolismandenergymetab-olism.Indeed,somephenotypesareobservedincavefishrelativetosurfacefish:cavefishexhibitfattyliversandin-satiableappetite[69],lowermetabolicratethansurfacefish[70],higherinfectionsusceptibility,andmorpho-logicalchangeingut(personalobservations).Theseco-occurringsymptomsraiseapossibleavenueforfutureworktoexplorehowchangesinimmunefunctionandmetabolism(andperhapsgutfunction)influenceasetofASD-likebehaviorsincavefish.Indeed,metabolism,immuneandgutdefectshavebeensuggestedinASDetiologyinhumans[59,71–73].Notably,besidesASD-likebehaviors,cavefishevolvedeyedegeneration,pigment-loss,widenedjaws,anin-creaseinfattissue,anincreaseinnumberofteeth,andenhancementofnon-visualsensorysystems(mechano-sensorylateralline,tastebuds,andolfactoryepithelium)[20,69,74–82].Someofthesecave-associatedtraitsmaybegeneticallycorrelatedtoASD-likebehaviors.Forexample,adherencetoaparticularvibrationandlossofschoolingshowedsignificantcorrelationswitheyesize(vibrationattractionbehaviorandtheeyesize,andschoolingandeyesizeinF2intercross:r=0.26and=0.27,respectively[77,83]).WealsofoundthatmanyofthepositivelyselectedorthologsofASD-riskgenesareunderQTLintervalsforeyesize(Fig.1).How-ever,visualimpairmentitselfmaynotinduceasetofASD-likebehaviors.Rearingsurfacefishinthedarkdid 121086420 500040003000200010000 BeforeAfter 121086420 121086420Sleep Duration(hr/24hr)500040003000200010000500040003000200010000Swimming distance(m/24hr)Fluoxetine(5 µM)Clozapine(2 µM)Naltrexone(10 µg/g)Vibration Attraction Behavior 500040003000200010000 Aripiprazole(1 µM) 121086420 70006000500040003000200010000 121086420 Risperidone(1 µM) 543210543210543210BeforeAfterBeforeAfterBeforeAfterBeforeAfterBeforeAfterBeforeAfterBeforeAfterBeforeAfterBeforeAfterBeforeAfterBeforeAfterBeforeAfterSurface fish Pachón cavefish F1 *********************************mnoklhjigcde******* 543210543210BeforeAfterBeforeAfter*****abf Fig.2HumandrugsforASDmitigatedcavefish-typesymptomsinF1hybridandcavefish.(a-e)Adherenceto40-Hzvibrationstimulus.Vibrationattractionbehaviorisrepresentedbythesquare-rootednumberofapproachesduringa3-minassay.(f-j)Swimmingdistance(mper24-hassay).(k-o)Thechangesinsleepduration(hper24-hassay).BeforeandaftertreatmentofdrugsusedforASDpatients—aripiprazole(a,f,k),risperidone(b,g,l),fluoxetine(c,h,m),clozapine(d,i,n),naltrexone(e,j,o)—wereobservedfor24heachandplottedwithmeans±s.e.m.Inthesecases,therearesignificantshiftsofcavefishbehaviorsaftertreatmenttowardsthesurfacefishbehaviorsbeforetreatment,exceptforthenaltrexonetreatments.Starsindicatethesignificantbehavioralchangesbetweenbeforeandafterdrugtreatments(pairedt-testadjustedbyBonferronicorrection,***:P<0.001,**:P<0.01,*:P<0.05).Blackline:surfacefish,andorangeline:cavefish.AllstatisticsareavailableinAdditionalfile11.Blackdashedlinesina-eindicatethethresholdlevelofvibrationattractionbehavior(square-rootednumberofapproachesequals2)[20]Yoshizawaetal.BMCEvolutionaryBiology (2018) 18:89 Page8of16

PAGE 9

notincreasetheirvibrationattractionbehavior[19,84].Also,thereisnocorrelationbetweenlossofeyesandmanyofthesebehavioraltraits[85].Therefore,thesedatasuggestthateyeregressionisnotthemajordrivertoshiftasetofASD-likebehaviors.Wethenproposethat,inaregardoftheevolutionofmultiplebehaviors,theeyeregressionhaslittlecontributiontoASD-likebehaviorsexhibitedbythecavefish.Thepotentialforsharedgeneticunderpinningsbe-tweencavefishandASDcanofferfurtherinsightsintotheetiologyofASD.RecentstudieshavehelpedclarifythegeneticsofdenovovariantsofASD,whichlikelyac-countfor3–15%ofASDcases[23,56,65,86].However,giventhatASDishighlyheritableandthatallcommonandraregeneticvariantsareestimatedtoexplainasignifi-cantproportionofASDcases(17~50%)[23,65,66],ananimalmodelformultipleheritablevariantsisstillun-available.Here,cavefishmayservetouncovergene-geneandgene-environmentinteractions,andtoshedlightontheeffectofthegut-brainaxisonASD[20,71].ConclusionOverall,cavefishappearstobeanadvantageousplatformuponwhichtountanglethepolygenicevolutionarypro-cessesthatgenerateadiversebehavioralspectruminvertebrates.Arecentstudyinhoneybees—inwhichgeneexpressionmodificationsbetweenthebrainsofso-cialandless-socialhoneybeeswerefoundtobeenrichedinASD-riskgenes—highlightsfurtherthatASD-riskgenesforsocialbehaviorsaredeeplyconserved[5].Aboveall,inmanyofanimalspecies,includinghumanandcavefish,asetofASD-riskgenesmayimpacttheevolutionofmultiplebehaviors[5].MethodsFishmaintenanceandrearinginthelabAstyanaxmexicanussurfacefishusedinthisstudywerelaboratory-raiseddescendantsoforiginalcollectionsmadeinBalmorheaSpringsStatePark,Texas.Cavefishwerelaboratory-raiseddescendantsoriginallycollectedfromCuevadeElPachón(Pachóncavefish)inTamau-lipas,Mexico.Fish(surfacefishandPachóncavepopulations)werehousedintheUniversityofHawai‘iatMnoaAstyanaxfacilitywithtemperaturessetat21°C±0.5°Cforrearing,24°C±0.5°Cforbehaviorexperiments,and25°C±0.5°Cforbreeding[24,87].Lightsweremain-tainedona12:12light:darkcycles[24,87].Forrearingandbehaviorexperiments,lightintensitywasmaintainedbetween30and100Lux.Fishhusbandrywasperformedaspreviouslydescribed[20,24,87].Fishwereraisedtoadultsandmaintainedinstandard42Ltanksinacustom-madewater-flowtanksystem.Adultfishwerefedamixeddiettosatiationthreetimesdailystarting3hafterthelightscameon(Zeitgebertime3orZT3),ZT6andZT9(TetraColorTropicalFishFoodGranulesandTetra-MinTropicalFishFoodCrisps,Tetra,Blacksburg,VA;JumboMysisShrimp,HikariSalesU.S.A.,Inc.,Hayward,CA).Allfishinthebehavioralexperimentswerebetween2.5–5cminstandardlengthandbetween6and18monthsold.AllfishcareandexperimentalprotocolsareapprovedunderIACUC(17–2560)atUniversityofHawai‘iatMnoa.GenomesurveyandgeneexpressionofASD-andSCZ-riskgenesWeidentifiedalistofASD-riskgenesfromtheSimonsFoundationAutismResearchInitiative(https://sfari.org/resources/sfari-gene),whichhousesanextensivecollec-tionofdataongenespotentiallyimplicatedinASDinhumans.WequeriedtheSimonsFoundationAutismResearchInitiative(https://www.sfari.org/resource/sfari-gene/;up-datedMarch2017)databasesandselected493ASDgenesinCategories1–4andS(genesineachcategoryareclassifiedbasedonformofASD,amountofriskcon-ferred,andtypeofevidenceforassociationwithASD,withhighercategoriesindicatingmoreevidence.‘S’cat-egorygenesareassociatednotonlywithASDbutalsowithadditionalsymptoms).ForSCZ-riskgenes,wequeriedtheSchizophreniaResearchForum(http://www.szgene.org/;updated2012)and(http://szdb.org/;updatedMay2017)andextracted44genesinthe‘TopResults’and304genesbasedon‘Score2–4,’respectively.Thesewerebasedontheevidencefromhumangenome-wideassociationstudy,geneexpressionofpost-mortembrainsand/orexpressionQTL)[45,47].Forex-ample,inSCZgenes,score4groupincludesthegeneswhichmeets4categories:(1)significantlydifferentlyexpressedinpatients,(2)significantingenome-wideas-sociationstudy,(3)significantinlinkageand/orassoci-ationstudyand(4)significantinpathwayanalysis(Additionalfile3).Bothofthesegenesetsweresurveyedagainsttherecentcavefishgenome(Ensembl.orgAssem-bly;AstMex102;GenebuildlastupdatedJuly2016[31,43–45]).First,humangeneswerequeriedwiththeHomologue(OrthologousCavefishGenes)attributestoAstyanaxgenesinBioMart.Includingtheparalogs,wehavealistof677AstyanaxhomologsofhumanASDgenes(Additionalfile1).Similarly,wehavealistof766homologsofhumanSCZgenes(Additionalfile3).FortheRNAseqtranscriptomeanalysis,variationingeneexpressionwasanalyzedusingpreviouslypublishedRNAseqdata(Genbanksequencereadarchive(SRA),accessioncode:PRJNA258661)[31,38].Thisdatasetincludes50-pooledwholelarvaefromsurfacefishandPachóncavefish(caveandsurfacefishpooledseparately)atdifferentdevelopmentalYoshizawaetal.BMCEvolutionaryBiology (2018) 18:89 Page9of16

PAGE 10

stages(10hpostfertilization(hpf),24hpf,36hpfand72hpf).Librariesforeachpoolof50larvaewerepreparedonceandthensequencedontheIlluminaHiSeqinthreetechnicalreplicates[36,88].Datawereanalyzedbyfollowingpreviouslypublishedprotocols[89].Briefly,theexoninformationforA.mexicanuswasacquiredviatheGTFfile(Astyanax_mexicanus.AstMex102.89.gtf.gz)atensembl.org(http://www.ensembl.org/info/data/ftp/index.html)andRNAsequencesinfastqformatwerealignedtoA.mexicanusgenomesequence(“Astyanax_-mexicanus.AstMex102.dna_sm.toplevel.fa”downloadablefromftp://ftp.ensembl.org/pub/release-91/fasta/astyanax_mexicanus/dna/)usingSTARalignerversion2.5.1b[90].First,weindexedthegenomesequenceforSTARbyusing“–runModegenomeGenerate”“–sjdbGTFfile./Astyanax_mexicanus.AstMex102.84.gtf”“–genomeFas-taFiles./Astyanax_mexicanus.AstMex102.dna_sm.toplevel.fa”.WethenmappedtherawfastqreadstoAstya-naxgenomeusing“–outSAMtypeBAMUnsorted.”Afterthealignment,agenemodeldatabasewasbuiltbythefunctionmakeTxDbFromGFFintheGenomicFeaturespackage(ver.1.23.31)inR[91].Oncethedatabasewasbuilt,weusedthefunctionsummarizeOverlapsintheGenomicAlignmentspackage(ver.1.8.0[91])inRtoad-justthereadcountsbasedontheexoninformationofeachgene,whichconvertedthereadcountsintoFPKM(FragmentsPerKilobaseMillion).Expressionlevelswerecomparedusingtheadjustedreadcounts.ToquantifyexpressiondifferencesbetweensurfacefishandPachóncavefishat72hpf,weusedthe‘re-sults’functionintheDESeqpackageafterestimatingthedatavariance(ver.1.12.0[92]).Wealsotestedtheage×populationinteractionsbysettingtheparameter‘reduced=~population+age’inthefunctionofDESeq,followedbythe‘results’functiontoextractthestatistics[89].Allscriptshavebeenmadeavail-ableonGitHub(https://github.com/masa-yoshizawa/Asty-RNAseq).Theanalysisat72hpfwasselectedbe-cause(i)differencesinexpressionpatternsthroughdevelopmentalstagescouldaffectthenervoussystemde-velopment(age×populationinteractions),and(ii)fishstartmoving/swimmingaccordingtothesensoryinputssothattheneuralwiringslikelyarebeingelaborated(comparabletoinfants:72hpf).Benjamini-Hochbergad-justedP-valuesandlog2expressiondifferencesbetweencavefishandsurfacefishwereusedtodeterminesignifi-cance(asdescribedin[89]).Weperformedbootstrappingusing9999randomsamplingsof500genesfromthe22,767geneswithexpressiondataoutof23,042totalgenesinthegenomeofA.mexicanus(Ensembl.orgAssembly;AstMex102;GenebuildlastupdatedJuly2016[31,43–45]).Thenumberof500ischosenbecauseweused493SFARIgenesinouranalysis(Table1andAdditionalfile3:S2).PopulationgenomicsandselectionpressureanalysesSamplecollectionandpreparationAllfin-clipsoffishinthewildwerecollectedunderMexico’sNationalAquacultureandFishingCommission(CONAPESCA)permitPPF/DGOPA-106/2013toDr.ClaudiaPatriciaOrnelasGarcíaandMexico’sSecretariatofEnvironmentandNaturalResources(SEMARNAT)permit02241toDr.ErnestoMaldonado.Briefly,wein-cludedacoresetofsampleswhichcontainedthefollow-ing:Pachóncave,N=10(9newlyre-sequenced+thereferencereadsmappedbacktothereferencegenome)andsurface(RíoChoy),N=9.AfterDNAextractionwithGenomic-TipTissueMidikitsandDNeasyBloodandTissuekit(Qiagen),individualsampleswerebar-codedandnext-generationsequencinglibrarieswerepreppedwithIlluminaTruSeqNanoDNASamplePrepKit(v3reagents).EveryfivebarcodedsampleswerepooledandsequencedintwolanesoftheIlluminaHiSeq2000using100bppaired-endreads.AlignmentsofIlluminadatatotheA.mexicanusgenomever.1.0.2[31]werecreatedwiththeBWA-memalgorithminbwa-0.7.1[93].TheGenomeAnalysisToolkitv.3.3.0(GATK)andPicardv1.83(http://broadinstitute.github.io/picard/)wereusedtofilteralignmentsinaccordwithGATKBestPractices[94].Hardfilterswereappliedsep-aratelytoSNPsandindels/mixedsitesusingtheVariant-FiltrationandSelectVariantstoolstoremovelowconfidencecallsfromthedataset.Extensivedetailsofsamplecollectionandpopulationgenomicanalysesareprovidedin(Hermanetal.submitted)whichincludesadditionalsamples.Samplesusedintheanalysespre-sentedhereweresubmittedtotheProjectAccessionNumber:SRP046999[38].Weemployedmultiplemeasurestoidentifyregionsofexceptionalgenomicdivergencebetweencavefishandsurfacefishpopulationsandtoidentifyregionspoten-tiallyunderpositiveselectioninthecavefish.Forallpopulationgenomicmeasures,weexcludedmaskedre-petitiveelements,indels(ifpresentinanyofthecoresetofsamples),and10bpsurroundingthebasesaffectedbyeachindel.Weusedthemasking_coordinates.gzfileavailablefortheA.mexicanusgenomev1.0.2thoughNCBIgenomesFTPandperformedthefollowingmea-sureswithGATK-processeddataonaper-genebasis,unlessotherwisenoted.Wefocusedonmultiplepopula-tiongenomicstatisticalmetricsDXY,FST,andhapFLK.Specifically,weconductedanenrichmenttestsonthepositivelyselectedASD-genesrelativetotherestofthegenomeusingonlytheresultsfromhapFLK.Wedefined‘divergenceoutliers’asgenesthatwereamongthetop5%acrossthegenomeforFST,thetop20%forDXY(asthisisalesssensitivemeasurethanFST)and/orexhibitedsignificantp-valuesusingtheprogramhapFLK.Thetop5%iscommonlyusedasacut-offinYoshizawaetal.BMCEvolutionaryBiology (2018) 18:89 Page10of16

PAGE 11

outlieranalyses[95].Thecut-offforDXYasthetop20%waschosenasalessstringentcriteriathanforFSTbe-causethismetricislesssensitivetoallelefrequencyshifts[52].AcrossSFARIgenesCategory1–4,only35genesoutof113totaloutliergenesweredefinedasout-lierssolelybythecriteriaofbeinginthetop20%forDXY,andtheseweremainlyusedintheIPAanalysis.WealsoredidIPAanalysiswith5%asthecut-offforDXYandobtainedsimilarresults.Thesetwometricsexhibitdifferentsensitivitiesandassumptions(reviewedin[95]).Forexample,relativemeasuresofdivergence(e.g.FST)[96]detectdivergentregionsbetweentwopopulations,yetmayalsodetectoutliersinlowdiversityregionsthatarefalsepositives.Thus,weinterpretedanyoutliersde-finedbyrelativemeasuresofdivergenceinthecontextofpairwisenucleotidediversity(i.e.Pi),whichisameasureofdiversitywithinthepopulations.Absolutemeasuresofdivergence(DXY)[52],whicharenotcon-foundedbylowdiversity,arenotassensitivetobiologic-allymeaningfuldivergenceasrelativemeasures[52]andmayleadtofalsenegatives.Thus,weusedacombin-ationofevidencetounderstandthemolecularevolutionofASD-genes.Toidentifygenesinthetop5%forFSTandtop20%forDXY,weperformeddenserankingswhereeachmeas-ure(e.g.Pisurface,Picave,FSTandDXY)wasrankedforeachgeneinthegenome.Thehighertheranking,thehigherthevaluewasforthatmeasurewithdenseranks(e.g.0.02,0.03,0.04,0.04wasranked1,2,3,3).Toavoidissuesinregionsoflowdiversity,weexcludedgenesthatwerethelowestranked500genesforPiinthesurfacepopulationasthesegenesmayrepresentregionsoflowrecombinationinthegenome.Inaddition,tomeasuresofabsoluteandrelativedivergence,wealsoimplementedtheprogramhapFLK,whichfocusesondifferencesofhaplotypefrequenciesbetweenpopulations(seebelow).ForhapFLK,wefocusedongeneswithatleastonehapFLKP-valuethatwaslessthan0.05.Duetotheuseofmultiplemetrics(e.g.DXY,FST,hapFLK),weclassifiedafocalpsychiatricdisease-relatedgeneasadivergenceoutlierifthegenemetanyofthecriteriaoftop5%ofgenesforFST,thetop20%forDXYand/orhadoneP-value<0.05forhapFLKwithinthegene(Table3,Additionalfiles6and7).BasicpopulationgenomicmetricsWeusedVCFtools[97]tocalculatePiandFSTandcustompythonscriptstocalculatethesemetricsonaper-genebasis.Weidentifiedtheallelecountsperpopu-lationwithVCFtoolsandusedtheseforsubsequentDXYcalculations.Forallmetrics,weonlyusedsitesthatcontainedsixormoreindividualsperpopulation(seeAdditionalfile7).ForFSTandDXY,wefocusedoncomparisonsbetweenPachóncavefishandsurfacefish(RíoChoypopulation).TheRíoChoysurfacefishrepre-sentthepopulationclosestinourpopulationgenomicsamplingtothesurfacefishpopulationfromTexasusedinthecurrentstudy[98].hapFLKhapFLKisanexplicittestforpositiveselectionandde-tectschangesinhaplotypefrequenciesthatexceedwhatisexpectedforgeneticdriftgivenahierarchicalpopula-tionstructure[53].hapFLKmayberobustagainstbot-tlenecksandmigration,andinanalysesofvariousselectivesweepmeasuresacrossregionsofknownsweepsindogs,hapFLKdetectedeveryfocalsweep[95].WeusedhapFLKver.1.3https://forge-dga.jouy.inra.fr/projects/hapflk[53]with45Astyanaxsamplesfromfivetotalpopulations(6–10individualsperpopulation)andtwoadditionaloutgroupsandincludedthefollowingparameters:30haplotypeclusters(-K30),20EMrunstofittheLDmodel(nfit=10),andunphaseddata.P-valueswereestimatedbyfittingastandardnormaldistributiongenomewideinR(Table3,Additionalfile7)[53].QuantitativePCRFromthelab-rearedindividuals,weanesthetized(with0.5mg/mlofbufferedMS222inice-coldwater)fourin-dividualseachat1month,2months,and4monthsoldandtwoindividualsat12monthsoldfromboththesurfaceandPachóncavefishpopulations.Wholebrainsfromeachindividualwerethencarefullydissectedoutinice-coldPBS[99]andcollectedintoapre-chilled1.5ml-tubeondryice.Thebrainsofeachwerehomoge-nizedin1mlQIAzolLysisReagent(Qiagen,Valencia,CA)byusingaMicroTubeHomogenizerSystem(Wilmad-LabGlass,Vineland,NJ).ThetotalRNAextrac-tionwasperformedbyusingtheRNeasyPlusUniversalkit(Qiagen)withanelutionvolumeof20l.RNAqual-ityandquantityweredeterminedbasedonelectrophor-esisandQubit3.0Fluorometersystem(ThermoFisherScientific,Waltham,MA),respectively.iScriptgDNAClearcDNASynthesiskitwasusedtoeliminatethecar-ryoverofgenomicDNA,followedbysynthesisofcDNAbyusingiScriptReverseTranscriptionSupermixforRT-qPCR(Bio-RadLaboratories,Hercules,CA).ThequantitativeRT-PCRforthegenesassociatedwithASDwasperformed.Threehousekeepinggeneseef2.1a,rsp18andb2mwereusedtonormalizethequantifica-tioncycle(Cq).Quantitativereal-timePCRwasper-formedonaCFX96TouchReal-TimePCRDetectionSystem(Bio-RadLaboratories)usingtheSsoAdvancedUniversalSYBRGreenSupermix(Bio-RadLaboratories).Cyclingparameterswere:1cycleof95°Cfor15s,and40cyclesof98°Cfor5sand60°Cfor30s.Afterquantitativereal-timePCR,themeltcurveanalysiswasYoshizawaetal.BMCEvolutionaryBiology (2018) 18:89 Page11of16

PAGE 12

performedbetween65°C–95°Cwith0.5°Cstep.Dur-ationofeachstepwas2–5s.Thisidentifiedtheanneal-ingtemperatureforeachPCRproduct,whichinformsthetargetspecificityofthePCRreactionbymonitoringwhetherthesinglelengthofPCRproducts(i.e.asinglesharppeakofmeltingcurve)wasamplified[100].Mea-surementsofgeneexpressionateachdevelopmentalstage(1month,2months,4monthsand12monthsold)weretechnicallyrepeatedthreetimesbyusingthreewellsofaPCRplate.GeometricaverageofCqtargetwassubtractedbythegeometricaverageofthreerepeatsofthreehousekeepinggenesatthesamedevelopmentalstage(Cqreference:eef2.1a,rsp18andb2m),yieldingCq.Relativeexpression(Cq)ofeachgeneateachtargettissue(i.e.brainof1,2,4or12monthsfromsurfacefishorcavefish)wasthencalculatedbysubtractingtheCqofcavefishbrainat1montholdfromCqoftargettis-sue(Additionalfile5).ThesequencesforPCRprimersarereportedinTable4.DrugtreatmentASDdrugswereselectedaccordingtoclinicaltrialsandpractices[62,63,101,102].Drugconcentrationswerede-terminedbasedonpreviousexperimentsinmodelspecies[103–105].Fluoxetine(1.0–28.5M;Sigma-Aldrich,St.Louis,MO),clozapine(0.1–12.5M;Selleck,Houston,TX),naltrexone(5–10g/bodyg;Selleck),aripiprazole(1–5M;Selleck),orrisperidone(1–5M;Selleck)weredeliveredviabathapplicationwithPBSsolutioninconditionedwater(forfluoxetine),via0.1%dimethylfor-mamideinconditionedwater(forclozapine,aripiprazoleandrisperidone),orviainjectionthroughthebodycavity(fornaltrexone;lessthan20lwitha27Ginsulinsyringe)(Table5).Injectionswereperformedunderanesthesiausing66.7g/mlofMS-222(Tricaine,Sigma-Aldrich,St.Louis,MO).Forsleepandhyperactivityassays,fishwerebath-treatedorgivenanintraperitonealinjectionwitheachdrugatZeitgeberTime1(ZT1).InformationforeachdrugisreportedinTable5.Westartedvideorecord-ingsrightafterthetimeofinjection(ZT2)thenrecordedfor24htomeasuresleepandhyperactivitylevels(seebelow).Foradherenceassays(vibrationattractionbehav-ior;VABassay,seebelow),fishweretreatedwiththefocaldrugforatleast16h(overnight),andthensubjectedtoa3-minbehavioralassay(seeTable5).F1hybridsofsurfacefishandcavefishwerealsoassayedunderthetreatmentoffluoxetine,clozapineandnaltrexone,however,wehadtechnicaldifficultiesinraisingsufficientnumbersofF1hy-bridstoalsotesthybridsforthearipiprazoleandrisperi-donetreatments.SleepandhyperactivityFishwererecordedundernon-drugtreated-conditionsinacustom-designed10.0-Lacrylicrecordingchamber(457.2×177.8×177.8mminlength×width×height,respectively,with6.4mmthickness)withopaqueparti-tionsthatallowfor5individuallyhousedfishpertank(eachindividualchamberis88.9×177.8×177.8mm). Table4PCRprimersusedinquantitativeRT-PCRstudyPrimerNameEnsembleGeneIDForwardSequenceReverseSequenceshank3bENSAMXG00000009680AGTATGACCCACGGCTAGAGCGATCACATAATCACTGTAGGAGGshank3aENSAMXG00000004290CGAATTACACCAGCAGAAATCAGCCTCAGTAGCTCCGAAAGACadnp2aENSAMXG00000006355AGAGTCACTGGATGTGATTCACTTCTTGGTTCAAGTCGATGATCTCsuv420h1ENSAMXG00000020652AAATGAACACCAGGTTTCGACAGAAAGTGATGTCTGCTCCAkmt5bENSAMXG00000017258GGCTGATGATTGAAACAGAGACCATGTCGTCTTCCATTTACTACCpogzaENSAMXG00000020522TATTACACGCGTATTTCAGGGTAGACACAGATATCCACGAAGAGptenaENSAMXG00000010994CGGGACTACTTGATTCTAACTCTGTACAACTTCACCTTAAAGTTCGGGscn2aENSAMXG00000004964TCTTCACCTACATCTTCATCCTGCCAAAGACACATCTACAATGAGGb2mENSAMXG00000011344TTCACACCTCAGAAGAACGAACTGCATTCTCCATCTGGTeef2a.2ENSAMXG00000018020TATCATTGAGGAGTCTGGAGAGTGGGTCGGATTTCTTAATTGGrps18ENSAMXG00000007922CCATCAAGGGTGTTGGTAGGTGCATAATGGTCACCACCC Table5DruginformationusedinthisstudyDrugNameCommercialNameTargetApplicationMethodReferencesAripiprazoleAbilifyPartialagonistforthereceptorsofdopamine,serotoninandothersBath(1–5M)[62]RisperidoneRisperdalAntagonistforthereceptorsofdopamine,serotoninandothersBath(1–5M)[62,63]FluoxetineProzacSelectiveserotoninreuptakeinhibitor(SSRI)Bath(1.0–28.5M)[62]ClozapineClozarilAntagonistforthereceptorsofdopamine,serotoninandothersBath(0.1–12.5M)[62,63]NaltrexoneReviaOpiateantagonistInjection(5–10g/g)[63]Yoshizawaetal.BMCEvolutionaryBiology (2018) 18:89 Page12of16

PAGE 13

Thissetupisapproximatelythesameasusedin[24].Briefly,therecordingchamberwasilluminatedwithacustom-designedIRLEDsourcefor24handwithawhiteLEDlightthatwassetona12hon,12hoffcycle(Lampux12VFlexibleWaterproof5050LEDStripLights-DaylightWhite,LightingEVER,LasVegas,NV).Whitelightwasturnedonat7:00amandturnedoffat7:00pmeveryday.Behaviorwasrecordedafter4–5daysofacclimationfor24hbeginningtwohoursafterthelightwasturnedon(Zeitgebertime2).Videoswerere-cordedat15frames/secusingaUSBwebcamthatwasfittedwithazoomlens(Macro1.8/12.5-75mmC-mountzoomlens,KowaAmericanCorp.,Torrance,CA).AnIRhigh-passfilterwasplacedbetweenthecam-eraandthelenstoblockvisiblelight.Videoswerecap-turedbythesoftware,Virtualdub(Version1.10.4,http://www.virtualdub.org/)withx264vfwcodecandweresub-sequentlyprocessedusingSwisTrack(Version4,https://en.wikibooks.org/wiki/SwisTrack).Watertemperaturewasmonitoredthroughouttherecordings,andnode-tectabledifferenceswereobservedduringthelightanddarkperiods(24.0°C±0.5°C)[24].Thevisiblelightdur-ingbehaviorrecordingswasapproximately30–100Lux.Trackingparametersfordetectionweresetasfollows:thedetectionwassetto‘subjectbrighterthanbackground’andbrightnesscontrastfrom20to255;currentframeweightsetto15;videosampleratesetto15frames/sec,andpixelsmoothingwasturnedoff.Wemonitoredsleep,activity,andarousalthresholdviapreviouslyestablishedprotocolsinA.mexicanus[24].Sleepstatewasanalyzedaccordingtoapreviousstudy[24,106].Datawassubse-quentlyprocessedusingcustom-writtenPerlscripts(v5.10.0,www.perl.org)andExcelmacro(Microsoft,Redmond,WA).AdherencetoaparticularvibrationstimulusWeassayedVABasdescribedpreviously[19,77,107].Briefly,Wehaveacclimatedfishindividualsfor4–5dayspriortotheassayinthecylindricalassaychamber(325mlglassdish,10cmdiameter5cmhigh,VWR,Radnor,PA,USA)filledwithconditionedwater(pH6.8–7.0;conductivity600–800S).Duringtheas-says,vibrationstimuliwerecreatedbyusingaglassrodthatvibratedat35or40Hz.Thenumberofapproaches(NOA)tothevibratingrodwasvideo-recordedduringa3-minperiodunderinfraredilluminationandcountedusingImageJ1.50osoftware(NIH,Bethesda,MD,USA).StatisticsForstatisticalcomparisons,weperformedparametrictestsincludingstudent’st-testsandone-wayortwo-wayANOVAstocomparebetweensurfaceandcavefish.WeperformedLevene’sequalityofvariancetestandvisuallyinspectedthedistributionofthedata,tolookforviolationsoftheassumptionsofequalvarianceandnor-mality.Ifviolationsweredetected,wetransformedthedatabyapplyingsquare-rootorlogtransformation,ac-cordingly[108].Post-hocDunnettt-testsandBonferronicorrectionwereusedforunderstandingwhichcontrastsweresignificant.AbovecalculationswereconductedusingIBMSPSS24.0.0.1software(IBM,Somers,NY,USA)andallstatisticalscoresareavailableinAdditionalfiles10and11orfigurelegends.AdditionalfilesAdditionalfile1:GeneexpressionchangesinA.mexicanusorthologsofSFARIgenes(Category1to4andCategorySininMarch2017).(XLSX217kb)Additionalfile2:GeneexpressionplotsofA.mexicanusASD-riskgeneslistedinSFARIGene.ManyofA.mexicanusorthologsofASD-riskgenesinCategory1–4andCategoryShadsignificantlydifferentexpressionbetweensurfacefishandcavefishatthestagesof10,24,36and72hpostfertilization(hpf).(PDF1187kb)Additionalfile3:LittleenrichmentoftheexpressionshiftsbetweensurfacefishandcavefishinSCZ-riskgeneslistedinSZgene.orgorszdb.org.(PDF235kb)Additionalfile4:Thedirectionofgeneregulation(upordown)seeninA.mexicanuswiththatseeninASDpatientsversuscontrols.(XLSX559kb)Additionalfile5:ExampleexpressionpatternsofASD-riskgenesinSFARIGeneCategory1.QuantitativeRT-PCRshowedsimilaritiestoASDpatients.(PDF74kb)Additionalfile6:DXYandFSTstatisticsandplotsforthetopoutlierswhosedivergencemetricspassedourthreshold(i.e.top5%forFST,top20%forDXY,andP<0.05forhapFLK).(PDF139kb)Additionalfile7:DivergenceoutlierswithintheSFARIASD-riskgenesbetweencavefishandsurfacefishbasedonthreemetrics(i.e.top5%forFST,top20%forDXY,andP<0.05forhapFLK).(XLSX466kb)Additionalfile8:IngenuityPathwayAnalysisComparisonAnalysisrevealedenrichedpathwaycategoriesindiversifiedSFARIgenesincavefish.(PDF8084kb)Additionalfile9:ASDgenesunderdiversification/positiveselectionandunderQTLintervals.(XLSX71kb)Additionalfile10:Humandrugsforpsychiatricdiseasemitigatedcavefish-typesymptomsinadose-dependentmanner.(PDF93kb)Additionalfile11:StatisticalscoresforFig.2.(PDF45kb)AbbreviationsASD:Autismspectrumdisorder;CF:Cavefish;IPA:IngenuityPathwayAnalysis;iPScells:Inducedpluripotentstemcell;QTL:Quantitativetraitlocus;RNAseq:Ribonucleicacidsequencing;RT-PCR:Reversetranscriptionpolymerasechainreaction;SCZ:Schizophrenia;SF:Surfacefish;SFARI:SimonsFoundationAutismResearchInitiative;VAB:VibrationattractionbehaviorAcknowledgmentsWethankA.MaunakeaandK.Bennettforhelpfulcommentsandadvice,M.YoshidaandV.Fernandesfortechnicalassistance,C.Macaspac,C.Valdez,S.Fujita,J.Choi,L.Lu,N.SimonandN.Cetraroforfishmaintenance,M.Iwashitaforcriticalreading,andC.BrenneckaforscientificwritingservicesonbehalfofPerfectingPapersLtd.WealsothanktheaidofC.P.Ornelas-García,E.MaldonadoandW.R.JefferyforthesamplecollectionandfieldpermitsinMexico.MYconceivedtheprojectanddesignedtheexperiments.Yoshizawaetal.BMCEvolutionaryBiology (2018) 18:89 Page13of16

PAGE 14

FundingThisworkwassupportedbyTheHawaiianCommunityFoundation(16CON-78919)toMY,theUMN-MinnesotaSupercomputingInstitute,andtheEppleyFoundationforResearchtoSEM.CNPwassupportedbytheUniversityofMinnesotaGrandChallengesPostdoctoralProgram.Thesefundingdidnothavearoleinthedesignofthestudyandcollection,analysis,andinterpretationofdataandinwritingthemanuscript.AvailabilityofdataandmaterialsOriginalbehaviordataarekeptonlocalharddriveswithbackups,andsequencedatawereuploadedatNCBIGenbankSRAProjectAccessionNumber:SRP046999.AllscriptsusedinRNAseqprocessingareavailableonGitHub(https://github.com/masa-yoshizawa/Asty-RNAseq).Allotherdatageneratedoranalyzedduringthisstudyareincludedinthispublishedarticleanditsadditionalfiles.Authors’contributionsMY,AS,andLTperformedpharmacologicalexperimentsanddataanalysis.NCperformedthequantitativePCRexperimentanddataanalysis.MY,MH,andLTperformedtranscriptomicanalysesandSFARIGenesurvey.CNPandSEMperformedthegenomicDNAsequencinganddivergenceoutlieranalyses.Allofauthorswrote,contributedtoeditsandapprovedthemanuscript.EthicsapprovalandconsenttoparticipateAllfishcareandexperimentalproceduresinthelabwereperformedundertheapprovedprotocolsofTheInstitutionalAnimalCare&UseCommittee(IACUC)(17–2560)atUniversityofHawai‘iatMnoa.Allfin-clipsoffishinthewildwerecollectedunderMexico’sNationalAquacultureandFishingCommission(CONAPESCA)permitPPF/DGOPA-106/2013toDr.ClaudiaPatriciaOrnelasGarcíaandMexico’sSecretariatofEnvironmentandNaturalResources(SEMARNAT)permit02241toDr.ErnestoMaldonado.CompetinginterestsTheauthorsdeclarethattheyhavenocompetinginterests.Publisher’sNoteSpringerNatureremainsneutralwithregardtojurisdictionalclaimsinpublishedmapsandinstitutionalaffiliations.Authordetails1DepartmentofBiology,UniversityofHawai‘iatMnoa,2538McCarthyMall,EDM216,Honolulu,HI96822,USA.2DepartmentofCellandMolecularBiology,SchoolofMedicine,UniversityofHawai’iatMnoa,Honolulu,HI96813,USA.3DepartmentofEcology,Evolution,andBehavior,UniversityofMinnesota,St.Paul,MN55108,USA.Received:16February2018Accepted:18May2018 References1.MayrE.Animalspeciesandevolution.1sted.Cambridge:BelknapPress;1963.2.SihA,BellA,JohnsonJC.BehavioralSyndromes:AnEcologicalandEvolutionaryOverview.2004;19:372–8.3.BellAM.Behaviouraldifferencesbetweenindividualsandtwopopulationsofstickleback(Gasterosteusaculeatus).JEvolBiol.2004;18:464–73.4.CarrollSB,GrenierJ,WeatherbeeS.FromDNAtoDiversity:MolecularGeneticsandtheEvolutionofAnimalDesign.2nded.Massachusetts:BlackwellScience,Inc.;2004.5.ShpiglerHY,SaulMC,CoronaF,BlockL,CashAhmedA,ZhaoSD,RobinsonGE.Deepevolutionaryconservationofautism-relatedgenes.ProcNatlAcadSci.2017;114:201708127.6.RittschofCC,BukhariSA,SloofmanLG,TroyJM,Caetano-AnollésD,Cash-AhmedA,KentM,LuX,SanogoYO,WeisnerPA,ZhangH,BellAM,MaJ,SinhaS,RobinsonGE,StubbsL.Neuromolecularresponsestosocialchallenge:commonmechanismsacrossmouse,sticklebackfish,andhoneybee.ProcNatlAcadSci.2014;111:17929–34.7.CampbellSS,ToblerI.Animalsleep:areviewofsleepdurationacrossphylogeny.NeurosciBiobehavRev.1984;8:269–300.8.SiegelJM.Doallanimalssleep?TrendsNeurosci.2008;31:208–13.9.CirelliC.Thegeneticandmolecularregulationofsleep:fromfruitfliestohumans.NatRevNeurosci.2009;10:549–60.10.MiyazakiS,LiuC-Y,HayashiY.Sleepinvertebrateandinvertebrateanimals,andinsightsintothefunctionandevolutionofsleep.NeurosciRes.2017;118:3–12.11.InselTR.Thechallengeoftranslationinsocialneuroscience:areviewofoxytocin,vasopressin,andaffiliativebehavior.Neuron.2010;65:768–79.12.GoodsonJL.Thevertebratesocialbehaviornetwork:evolutionarythemesandvariations.HormBehav.2005;48(1SPEC.ISS):11–22.13.BukhariSA,SaulMC,SewardCH,ZhangH,BenskyM,JamesN,ZhaoSD,ChandrasekaranS,StubbsL,BellAM.Temporaldynamicsofneurogenomicplasticityinresponsetosocialinteractionsinmalethreespinedsticklebacks.PLoSGenet.2017;13:e1006840.14.SaulMC,SewardCH,TroyJM,ZhangH,SloofmanLG,LuX,WeisnerPA,Caetano-AnollesD,SunH,ZhaoSD,ChandrasekaranS,SinhaS,StubbsL.Transcriptionalregulatorydynamicsdrivecoordinatedmetabolicandneuralresponsetosocialchallengeinmice.GenomeRes.2017;27:959–72.15.BendeskyA,KwonYM,LassanceJM,LewarchCL,YaoS,PetersonBK,HeMX,DulacC,HoekstraHE.Thegeneticbasisofparentalcareevolutioninmonogamousmice.Nature.2017;544:434–9.16.DenverRJ.Structuralandfunctionalevolutionofvertebrateneuroendocrinestresssystems.AnnNYAcadSci.2009;1163:1–16.17.NithianantharajahJ,KomiyamaNH,McKechanieA,JohnstoneM,BlackwoodDH,ClairDS,EmesRD,vandeLagemaatLN,SaksidaLM,BusseyTJ,GrantSGN,StClairD,EmesRD,vandeLagemaatLN,SaksidaLM,BusseyTJ,GrantSGN.Synapticscaffoldevolutiongeneratedcomponentsofvertebratecognitivecomplexity.NatNeurosci.2012;16:16–24.18.vandePolI,FlikG,GorissenM.Comparativephysiologyofenergymetabolism:fishingforendocrinesignalsintheearlyvertebratepool.FrontEndocrinol(Lausanne).2017;8:36.19.YoshizawaM,GorikiS,SoaresD,JefferyWR.Evolutionofabehavioralshiftmediatedbysuperficialneuromastshelpscavefishfindfoodindarkness.CurrBiol.2010;20:1631–6.20.KeeneAC,YoshizawaM,McGaughSE.BiologyandevolutionoftheMexicancavefish.Amsterdam:ElsevierInc.;2015.21.BanerjeeS,RiordanM,BhatMA.Geneticaspectsofautismspectrumdisorders:insightsfromanimalmodels.FrontCellNeurosci.2014;8:58.22.AmericanPsychiatricAssociation.Diagnosticandstatisticalmanualofmentaldisorders(DSM-5).5thed.Arlington:AmericanPsychiatricAssociationPublishing;2013.23.BourgeronT.Fromthegeneticarchitecturetosynapticplasticityinautismspectrumdisorder.NatRevNeurosci.2015;16:551–63.24.YoshizawaM,RobinsonBG,DubouéER,MasekP,JaggardJB,O’QuinKE,BorowskyRL,JefferyWR,KeeneAC.Distinctgeneticarchitectureunderliestheemergenceofsleeplossandprey-seekingbehaviorintheMexicancavefish.BMCBiol.2015;1325.GalloND,JefferyWR.EvolutionofspacedependentgrowthintheteleostAstyanaxmexicanus.PLoSOne.2012;7:e41443.26.O’ConnellLA,HofmannHA.Thevertebratemesolimbicrewardsystemandsocialbehaviornetwork:acomparativesynthesis.JCompNeurol.2011;519:3599–639.27.O’ConnellLA,HofmannHA.Evolutionofavertebratesocialdecision-makingnetwork.Science(80-).2012;336:1154–7.28.BaranNM,McGrathPT,StreelmanJT.Applyinggeneregulatorynetworklogictotheevolutionofsocialbehavior.ProcNatlAcadSci.2017;114:5886–93.29.GrillnerS,RobertsonB.Thebasalgangliaover500millionyears.CurrBiol.2016;26:R1088–100.30.AbrahamsBS,ArkingDE,CampbellDB,MeffordHC,MorrowEM,WeissLA,MenasheI,WadkinsT,Banerjee-BasuS,PackerA.SFARIgene2.0:acommunity-drivenknowledgebasefortheautismspectrumdisorders(ASDs).MolAutism.2013;4:36.31.McGaughSE,GrossJB,AkenB,BlinM,BorowskyR,ChalopinD,HinauxH,JefferyWR,KeeneA,MaL,MinxP,MurphyD,O’QuinKE,RétauxS,RohnerN,SearleSMJJ,StahlBA,TabinC,VolffJ-NN,YoshizawaM,WarrenWC.Thecavefishgenomerevealscandidategenesforeyeloss.NatCommun.2014;5:5307.32.AmaralDG,SchumannCM,NordahlCW.Neuroanatomyofautism.TrendsNeurosci.2008;31:137–45.33.VoineaguI,WangX,JohnstonP,LoweJK,TianY,HorvathS,MillJ,CantorRM,BlencoweBJ,GeschwindDH.Transcriptomicanalysisofautisticbrainrevealsconvergentmolecularpathology.Nature.2011;474:380–4.Yoshizawaetal.BMCEvolutionaryBiology (2018) 18:89 Page14of16

PAGE 15

34.AnselA,RosenzweigJP,ZismanPD,MelamedM,GesundheitB:Variationingeneexpressioninautismspectrumdisorders:anextensivereviewoftranscriptomicstudies.FrontNeurosci2017,10(JAN):601.35.ParikshakNN,SwarupV,BelgardTG,IrimiaM,RamaswamiG,GandalMJ,HartlC,LeppaV,UbietaLdelaT,HuangJ,LoweJK,BlencoweBJ,HorvathS,GeschwindDH.Genome-widechangesinlncRNA,splicing,andregionalgeneexpressionpatternsinautism.Nature.2016;540:423–7.36.StahlBA,GrossJB.AlterationsinMc1rgeneexpressionareassociatedwithregressivepigmentationinAstyanaxcavefish.DevGenesEvol.2015;225:367–75.37.StahlBA,GrossJB.AcomparativetranscriptomicanalysisofdevelopmentintwoAstyanaxcavefishpopulations.JExpZoolBMolDevEvol.2017;328:515–32.38.LeinonenR,SugawaraH,ShumwayM,Internationalnucleotidesequencedatabasecollaboration:Thesequencereadarchive.NucleicAcidsRes2011,39(Databaseissue):D19–D21.39.WernerE,DawsonG,OsterlingJ,DinnoN.Recognitionofautismspectrumdisorderbeforeoneyearofage:aretrospectivestudybasedonhomevideotapes.JAutismDevDisord.2000;30:157–62.40.HuaJY,SmithSJ.Neuralactivityandthedynamicsofcentralnervoussystemdevelopment.NatNeurosci.2004;7:327–32.41.KatzLC,ShatzCJ.Synapticactivityandtheconstructionofcorticalcircuits.Science(80-).1996;274:1133–8.42.WapinskiI,PfefferA,FriedmanN,RegevA.Naturalhistoryandevolutionaryprinciplesofgeneduplicationinfungi.Nature.2007;449:54–61.43.YatesA,AkanniW,AmodeMR,BarrellD,BillisK,Carvalho-SilvaD,CumminsC,ClaphamP,FitzgeraldS,GilL,GirónCG,GordonL,HourlierT,HuntSE,JanacekSH,JohnsonN,JuettemannT,KeenanS,LavidasI,MartinFJ,MaurelT,McLarenW,MurphyDN,NagR,NuhnM,ParkerA,PatricioM,PignatelliM,RahtzM,RiatHS,etal.Ensembl2016.NucleicAcidsRes.2016;44:D710–6.44.BasuSN,KolluR,Banerjee-BasuS:AutDB:agenereferenceresourceforautismresearch.NucleicAcidsRes2009,37(Database):D832–D836.45.AllenNC,BagadeS,McQueenMB,IoannidisJPA,KavvouraFK,KhouryMJ,TanziRE,BertramL.Systematicmeta-analysesandfieldsynopsisofgeneticassociationstudiesinschizophrenia:theSzGenedatabase.NatGenet.2008;40:827–34.46.HorváthS,MirnicsK.Schizophreniaasadisorderofmolecularpathways.BiolPsychiatry.2015;77:22–8.47.WuY,YaoY-G,LuoX-J.SZDB:adatabaseforschizophreniageneticresearch.SchizophrBull.2016;43:459–71.48.KrämerA,GreenJ,PollardJ,TugendreichS.Causalanalysisapproachesiningenuitypathwayanalysis.Bioinformatics.2014;30:523–30.49.ProvenzanoG,CorradiZ,MonsornoK,FedrizziT,RicceriL,ScattoniML,BozziY.Comparativegeneexpressionanalysisoftwomousemodelsofautism:transcriptomeprofilingoftheBTBRandEn2(/)hippocampus.FrontNeurosci.2016;10:396.50.PramparoT,PierceK,LombardoMV,CarterBarnesC,MarineroS,Ahrens-BarbeauC,MurraySS,LopezL,XuR,CourchesneE.Predictionofautismbytranslationandimmune/inflammationCoexpressedgenesintoddlersfrompediatriccommunitypractices.JAMAPsychiatry.2015;72:386.51.MarianiJ,CoppolaG,ZhangP,AbyzovA,ProviniL,TomasiniL,AmenduniM,SzekelyA,PalejevD,WilsonM,GersteinM,GrigorenkoEL,ChawarskaK,PelphreyKA,HoweJR,VaccarinoFM.FOXG1-dependentdysregulationofGABA/glutamateneurondifferentiationinautismSpectrumdisorders.Cell.2015;162:375–90.52.CruickshankTE,HahnMW.Reanalysissuggeststhatgenomicislandsofspeciationareduetoreduceddiversity,notreducedgeneflow.MolEcol.2014;23:3133–57.53.FarielloMI,BoitardS,NayaH,SanCristobalM,ServinB.Detectingsignaturesofselectionthroughhaplotypedifferentiationamonghierarchicallystructuredpopulations.Genetics.2013;19354.CampbellI.Chi-squaredandFisher–Irwintestsoftwo-by-twotableswithsmallsamplerecommendations.StatMed.2007;26:3661–75.55.PolimantiR,GelernterJ.Widespreadsignaturesofpositiveselectionincommonriskallelesassociatedtoautismspectrumdisorder.PLoSGenet.2017;13:1–14.56.DeRubeisS,HeX,GoldbergAP,PoultneyCS,SamochaK,ErcumentCicekA,KouY,LiuL,FromerM,WalkerS,SinghTT,KleiL,KosmickiJ,FuS-C,AleksicB,BiscaldiM,BoltonPF,BrownfeldJM,CaiJ,CampbellNG,CarracedoA,ChahrourMH,ChiocchettiAG,CoonH,CrawfordEL,CrooksL,CurranSR,DawsonG,DuketisE,FernandezBA,etal.Synaptic,transcriptional,andchromatingenesdisruptedinautism.Nature.2014;515:209–15.57.SamochaKE,RobinsonEB,SandersSJ,StevensC,SaboA,McGrathLM,KosmickiJA,RehnströmK,MallickS,KirbyA,WallDP,MacArthurDG,GabrielSB,DePristoM,PurcellSM,PalotieA,BoerwinkleE,BuxbaumJD,CookEH,GibbsRA,SchellenbergGD,SutcliffeJS,DevlinB,RoederK,NealeBM,DalyMJ.Aframeworkfortheinterpretationofdenovomutationinhumandisease.NatGenet.2014;46:944–50.58.CouryDL,AshwoodP,FasanoA,FuchsG,GeraghtyM,KaulA,MaweG,PattersonP,JonesNE.Gastrointestinalconditionsinchildrenwithautismspectrumdisorder:developingaresearchagenda.Pediatrics.2012;130(Supplement2):S160–8.59.MacfabeD.Autism:metabolism,mitochondria,andthemicrobiome.GlobAdvHealMed.2013;2:52–66.60.KrummN,O’RoakBJ,ShendureJ,EichlerEE.Adenovoconvergenceofautismgeneticsandmolecularneuroscience.TrendsNeurosci.2014;37:95–105.61.McIntyreJC,ThiebaudN,McGannJP,KomiyamaT,RothermelM.Neuromodulationinchemosensorypathways.ChemSenses.2017;42:375–9.62.LeClercS,EasleyD.Pharmacologicaltherapiesforautismspectrumdisorder:areview.PT.2015;40:389–97.63.TsaiLY.Psychopharmacologyinautism.PsychosomMed.1999;61:651–65.64.RoyA,RoyM,DebS,UnwinG,RoyA.Areopioidantagonistseffectiveinattenuatingthecoresymptomsofautismspectrumconditionsinchildren:asystematicreview.JIntellectDisabilRes.2015;59:293–306.65.GauglerT,KleiL,SandersSJ,BodeaCA,GoldbergAP,LeeAB,MahajanM,ManaaD,PawitanY,ReichertJ,RipkeS,SandinS,SklarP,SvantessonO,ReichenbergA,HultmanCM,DevlinB,RoederK,BuxbaumJD.Mostgeneticriskforautismresideswithcommonvariation.NatGenet.2014;46:881–5.66.KleiL,SandersSJ,MurthaMT,HusV,LoweJK,WillseyAJ,Moreno-De-LucaD,YuTW,FombonneE,GeschwindD,GriceDE,LedbetterDH,LordC,ManeSM,MartinC,MartinDM,MorrowEM,WalshCA,MelhemNM,ChasteP,SutcliffeJS,StateMW,CookEH,RoederK,DevlinB,LeseMartinC,MartinDM,MorrowEM,WalshCA,MelhemNM,etal.Commongeneticvariants,actingadditively,areamajorsourceofriskforautism.MolAutism.2012;3:9.67.CrespiBJ.Autismasadisorderofhighintelligence.FrontNeurosci.2016;10:300.68.WeinerDJ,WigdorEM,RipkeS,WaltersRK,KosmickiJA,GroveJ,SamochaKE,GoldsteinJI,OkbayA,Bybjerg-GrauholmJ,WergeT,HougaardDM,TaylorJ,Bækvad-HansenM,DumontA,HansenC,HansenTF,HowriganD,MattheisenM,MoranJ,MorsO,NordentoftM,Nørgaard-PedersenB,PoterbaT,PoulsenJ,StevensC,AnttilaV,HolmansP,HuangH,KleiL,etal.Polygenictransmissiondisequilibriumconfirmsthatcommonandrarevariationactadditivelytocreateriskforautismspectrumdisorders.NatGenet.2017;49:978–85.69.AspirasAC,RohnerN,MartineauB,BorowskyRL,TabinCJ.Melanocortin4receptormutationscontributetotheadaptationofcavefishtonutrient-poorconditions.ProcNatlAcadSciUSA.2015;112:9668–73.70.MoranD,SoftleyR,WarrantEJ.EyelessMexicancavefishsaveenergybyeliminatingthecircadianrhythminmetabolism.PLoSOne.2014;9:e107877.71.CryanJF,DinanTG.Mind-alteringmicroorganisms:theimpactofthegutmicrobiotaonbrainandbehaviour.NatRevNeurosci.2012;13:701–12.72.MangiolaF,IaniroG,FranceschiF,FagiuoliS,GasbarriniG,GasbarriniA.Gutmicrobiotainautismandmooddisorders.WorldJGastroenterol.2016;22:361–8.73.GuptaS,EllisSE,AsharFN,MoesA,BaderJS,ZhanJ,WestAB,ArkingDE.Transcriptomeanalysisrevealsdysregulationofinnateimmuneresponsegenesandneuronalactivity-dependentgenesinautism.NatCommun.2014;5:5748.74.ProtasM,JefferyWR.Evolutionanddevelopmentincaveanimals:fromfishtocrustaceans.WileyInterdiscipRevDevBiol.2012;1:823–45.75.ProtasM,TabanskyI,ConradM,GrossJB,VidalO,TabinCJ,BorowskyR.Multi-traitevolutioninacavefish,Astyanaxmexicanus.EvolDev.2008;10:196–209.76.ProtasM,ConradM,GrossJB,TabinC,BorowskyR.RegressiveevolutionintheMexicancavetetra,Astyanaxmexicanus.CurrBiol.2007;17:452–4.77.YoshizawaM,YamamotoY,O’QuinKE,JefferyWR.Evolutionofanadaptivebehavioranditssensoryreceptorspromoteseyeregressioninblindcavefish.BMCBiol.2012;10Yoshizawaetal.BMCEvolutionaryBiology (2018) 18:89 Page15of16

PAGE 16

78.WilkensH.EvolutionandgeneticsofepigeanandcaveAstyanax-fasciatus(Characidae,Pisces)-Supportfortheneutralmutationtheory.In:HechtMK,WallaceB,editors.EvolBiol,vol.23.NewYork:PlenumPublishingCorporation;1988.p.271–367.79.WilkensH.Regressiveevolution:ontogenyandgeneticsofcavefisheyerudimentation.BiolJLinnSoc.2007;92:287–96.80.YamamotoY,ByerlyMS,JackmanWR,JefferyWR.Pleiotropicfunctionsofembryonicsonichedgehogexpressionlinkjawandtastebudamplificationwitheyelossduringcavefishevolution.DevBiol.2009;330:200–11.81.BibliowiczJ,AliéA,EspinasaL,YoshizawaM,BlinM,HinauxH,LegendreL,PèreS,RétauxS.DifferencesinchemosensoryresponsebetweeneyedandeyelessAstyanaxmexicanusoftheRioSubterráneocave.Evodevo.2013;4:25.82.HinauxH,DevosL,BlinM,ElipotY,BibliowiczJ,AliéA,RétauxS.Sensoryevolutioninblindcavefishisdrivenbyearlyembryoniceventsduringgastrulationandneurulation.Development.2016;143:4521–32.83.KowalkoJE,RohnerN,RompaniSB,PetersonBK,LindenTA,YoshizawaM,KayEH,WeberJ,HoekstraHE,JefferyWR,BorowskyR,TabinCJ.Lossofschoolingbehaviorincavefishthroughsightdependentandsight-independentmechanisms.CurrBiol.2013;23:1874–83.84.EspinasaL,BibliowiczJ,JefferyWR,RétauxS.EnhancedpreycaptureskillsinAstyanaxcavefishlarvaeareindependentfromeyeloss.Evodevo.2014;5:35.85.EspinasaL,BonarotiN,WongJ,PottinK,QueinnecE,RétauxS.Contrastingfeedinghabitsofpost-larvalandadultAstyanaxcavefish.SubterrBiol.2017;21:1–17.86.SandersSJ,MurthaMT,GuptaAR,MurdochJD,RaubesonMJ,WillseyAJ,Ercan-SencicekAG,DiLulloNM,ParikshakNN,SteinJL,WalkerMF,OberGT,TeranNA,SongY,El-FishawyP,MurthaRC,ChoiM,OvertonJD,BjornsonRD,CarrieroNJ,MeyerKA,BilguvarK,ManeSM,—estanN,LiftonRP,GünelM,RoederK,GeschwindDH,DevlinB,StateMW.Denovomutationsrevealedbywhole-exomesequencingarestronglyassociatedwithautism.Nature.2012;485:237–41.87.ElipotY,LegendreL,PèreS,SohmF,RétauxS.Astyanaxtransgenesisandhusbandry:howcavefishentersthelaboratory.Zebrafish.2014;11:291–9.88.GrossJB,StahlBA,PowersAK,CarlsonBM.Naturalbonefragmentationintheblindcave-dwellingfish,Astyanaxmexicanus:candidategeneidentificationthroughintegrativecomparativegenomics.EvolDev.2016;18:7–18.89.LoveMI,AndersS,KimV,HuberW,LoveMI,AndersS,KimV,HuberW.RNA-Seqworkflow:gene-levelexploratoryanalysisanddifferentialexpression.F1000Research.2016;4:1070.90.DobinA,DavisCA,SchlesingerF,DrenkowJ,ZaleskiC,JhaS,BatutP,ChaissonM,GingerasTR.STAR:ultrafastuniversalRNA-seqaligner.Bioinformatics.2013;29:15–21.91.LawrenceM,HuberW,PagèsH,AboyounP,CarlsonM,GentlemanR,MorganMT,CareyVJ.Softwareforcomputingandannotatinggenomicranges.PLoSComputBiol.2013;9:e1003118.92.LoveMI,HuberW,AndersS.ModeratedestimationoffoldchangeanddispersionforRNA-seqdatawithDESeq2.GenomeBiol.2014;15:550.93.LiH:Aligningsequencereads,clonesequencesandassemblycontigswithBWA-MEM.2013.94.vanderAuweraGA,CarneiroMO,HartlC,PoplinR,DelAngelG,Levy-MoonshineA,JordanT,ShakirK,RoazenD,ThibaultJ,BanksE,GarimellaKV,AltshulerD,GabrielS,DePristoMA.FromFastQdatatohighconfidencevariantcalls:thegenomeanalysistoolkitbestpracticespipeline.CurrProtocBioinforma.2013;43:1–33.95.SchlampF,vanderMadeJ,StamblerR,ChesebroughL,BoykoAR,MesserPW.Evaluatingtheperformanceofselectionscanstodetectselectivesweepsindomesticdogs.MolEcol.2016;25:342–56.96.CharlesworthB.Measuresofdivergencebetweenpopulationsandtheeffectofforcesthatreducevariability.MolBiolEvol.1998;15:538–43.97.DanecekP,AutonA,AbecasisG,AlbersCA,BanksE,DePristoMA,HandsakerRE,LunterG,MarthGT,SherryST,McVeanG,DurbinR.ThevariantcallformatandVCFtools.Bioinformatics.2011;27:2156–8.98.DowlingTE,MartasianDP,JefferyWR.Evidenceformultiplegeneticformswithsimilareyelessphenotypesintheblindcavefish,Astyanaxmexicanus.MolBiolEvol.2002;19:446–55.99.Nusslein-VolhardC,DahmR.Zebrafish:apracticalapproach.NewYork:OxfordUniversityPress;2002.100.TaylorS,WakemM,DijkmanG,AlsarrajM,NguyenM.ApracticalapproachtoRT-qPCR—publishingdatathatconformtotheMIQEguidelines.Methods.2010;50:S1–5.101.ChabrolH,BonnetD,RogeB.Psychopharmacologyofautism.Encephale.1996;22:197–203.102.KaplanG,McCrackenJT.PsychopharmacologyofautismSpectrumdisorders.PediatrClinNAm.2012;59:175–87.103.RihelJ,ProberDA,ArvanitesA,LamK,ZimmermanS,JangS,HaggartySJ,KokelD,RubinLL,PetersonRT,SchierAF.Zebrafishbehavioralprofilinglinksdrugstobiologicaltargetsandrest/wakeregulation.Science.2010;327:348–51.104.StewartA,WuN,CachatJ,HartP,GaikwadS,WongK,UtterbackE,GilderT,KyzarE,NewmanA,CarlosD,ChangK,HookM,RhymesC,CafferyM,GreenbergM,ZadinaJ,KalueffAV.Pharmacologicalmodulationofanxiety-likephenotypesinadultzebrafishbehavioralmodels.ProgNeuro-PsychopharmacologyBiolPsychiatry.2011;35:1421–31.105.StewartAM,NguyenM,WongK,PoudelMK,KalueffAV.Developingzebrafishmodelsofautismspectrumdisorder(ASD).ProgNeuro-PsychopharmacologyBiolPsychiatry.2014:27–36.106.DubouéER,KeeneAC,BorowskyRL.Evolutionaryconvergenceonsleeplossincavefishpopulations.CurrBiol.2011;21:671–6.107.YoshizawaM,AshidaG,JefferyWR.ParentalgeneticeffectsinacavefishadaptivebehaviorexplaindisparitybetweennuclearandmitochondrialDNA.Evolution(NY).2012;66:2975–82.108.LynchM,WalshB.GeneticsandAnalysisofQuantitativeTraits.Sunderland:SinauerAssociates,Inc;1998.109.YoshizawaM.TheevolutionofsensoryadaptationinAstyanaxmexicanus.In:KeeneAC,YoshizawaM,McGaughSE,editors.BiolEvolMexCavefish.Amsterdam:ElsevierInc;2015.p.247–63.110.O’QuinKE,YoshizawaM,DoshiP,JefferyWR.QuantitativegeneticanalysisofretinaldegenerationintheblindcavefishAstyanaxmexicanus.PLoSOne.2013;8:e57281.111.HinauxH,PottinK,ChalhoubH,PereS,ElipotY,LegendreL,RetauxS.AdevelopmentalstagingtableforAstyanaxmexicanussurfacefishandPachoncavefish.Zebrafish.2011;8:155–65. Yoshizawaetal.BMCEvolutionaryBiology (2018) 18:89 Page16of16


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.