Zr65ddd.com

RESEARCHOpenAccessCryptontransposons:identificationofnewdiversefamiliesandancientdomesticationeventsKenjiKKojimaandJerzyJurka*AbstractBackground:"Domestication"oftransposableelements(TEs)ledtoevolutionarybreakthroughssuchastheoriginoftelomeraseandthevertebrateadaptiveimmunesystem.
ThesebreakthroughswereaccomplishedbytheadaptationofmolecularfunctionsessentialforTEs,suchasreversetranscription,DNAcuttingandligationorDNAbinding.
CryptonsrepresentauniqueclassofDNAtransposonsusingtyrosinerecombinase(YR)tocutandrejointherecombiningDNAmolecules.
Cryptonswereoriginallyidentifiedinfungiandlaterintheseaanemone,seaurchinandinsects.
Results:HereinwereportnewCryptonsfromanimals,fungi,oomycetesanddiatom,aswellaswidelyconservedgenesderivedfromancientCryptondomesticationevents.
PhylogeneticanalysisbasedontheYRsequencessupportsfourdeepdivisionsofCryptonelements.
Wefoundthatthedomainofunknownfunction3504(DUF3504)ineukaryotesisderivedfromCryptonYR.
DUF3504issimilartoYRbutlacksmostoftheresiduesofthecatalytictetrad(R-H-R-Y).
GenescontainingtheDUF3504domainarepotassiumchanneltetramerizationdomaincontaining1(KCTD1),KIAA1958,zincfingerMYMtype2(ZMYM2),ZMYM3,ZMYM4,glutamine-richprotein1(QRICH1)and"withoutchildren"(WOC).
TheDUF3504genesarehighlyconservedandarefoundinalmostalljawedvertebrates.
Thesequence,domainstructure,intronpositionsandsyntenyblockssupporttheviewthatZMYM2,ZMYM3,ZMYM4,andpossiblyQRICH1,werederivedfromWOCthroughtworoundsofgenomeduplicationinearlyvertebrateevolution.
WOCisobservedwidelyamongbilaterians.
TherecouldbefourindependenteventsofCryptondomestication,andoneofthem,generatingWOC/ZMYM,predatedthebirthofbilateriananimals.
Thisisthethird-oldestdomesticationeventknowntodate,followingthedomesticationgeneratingtelomerasereversetranscriptase(TERT)andPrp8.
ManyCrypton-derivedgenesaretranscriptionalregulatorswithadditionalDNA-bindingdomains,andtheacquisitionoftheDUF3504domaincouldhaveaddednewregulatorypathwaysviaprotein-DNAorprotein-proteininteractions.
Conclusions:CryptonshavecontributedtoanimalevolutionthroughdomesticationoftheirYRsequences.
TheDUF3504domainsaredomesticatedYRsofanimalCryptonelements.
Keywords:tyrosinerecombinase,Crypton,domestication,transposon,DUF3504BackgroundThestructuralandmechanisticvarietyoftransposableele-ments(TEs)iswell-documented[1].
Theyencodeproteinsthatincludediversefunctionaldomainsinvolvedincataly-sisorinteractionwithDNA,RNAandotherproteins.
Becauseofthisdiverserepertoire,TEscansupplyfunc-tionalmodulestogeneratenewgenes.
"Moleculardomes-tication"oftransposableelements[2]ledtoevolutionarymilestonessuchastheoriginoftelomeraseandtheverte-brateadaptiveimmunesystem.
Telomerasereversetran-scriptase(TERT)providesasolutionforendreplicationproblemsaccompanyinglinearchromosomereplicationandwasderivedfromareversetranscriptase(RT)relatedtoPenelope-likeelementsintheveryearlystageofeukar-yoteevolution[3,4].
V(D)Jrecombinationisamechanismusedinjawedvertebratestogenerateavarietyofimmuno-globulinsandT-cellreceptors.
Itiscatalyzedbytherecombinationactivatinggene1(RAG1)derivedfromatransposaseencodedbytheTransibfamilyofDNAtrans-posons[5].
Differentkindsoftransposonproteinswere*Correspondence:jurka@girinst.
orgGeneticInformationResearchInstitute,1925LandingsDrive,MountainView,CA94043,USAKojimaandJurkaMobileDNA2011,2:12http://www.
mobilednajournal.
com/content/2/1/122011KojimaandJurka;licenseeBioMedCentralLtd.
ThisisanOpenAccessarticledistributedunderthetermsoftheCreativeCommonsAttributionLicense(http://creativecommons.
org/licenses/by/2.
0),whichpermitsunrestricteduse,distribution,andreproductioninanymedium,providedtheoriginalworkisproperlycited.
domesticated,includingtransposase,integrase,RT,envel-opeandgagproteins[6].
Hereinwereportin-depthstu-diesofanothertypeoftransposonenzyme,tyrosinerecombinase(YR),whichwasrepeatedlydomesticatedinthehistoryofanimals.
TodatefourtypesofenzymesareknowntocatalyzeDNAintegrationofeukaryotictransposons:DDE-trans-posase,YR,rolling-circlereplicationinitiatorandthecombinationofRTandendonuclease(EN)[7].
DDE-transposaseisthemostabundantgeneinnature[8]andiscarriedbymanyDNAtransposonsuperfamilies,self-synthesizingtransposons(Polinton),aswellaslongterm-inalrepeat(LTR)retrotransposons(Gypsy,Copia,BELandendogenousretroviruses)[1,9-11].
Theysharethreeconservedaminoacids(DDDorDDE)attheircatalyticsites,whichareseparatedbyaminoacidsequencesofvaryinglength.
SomedomesticatedDDE-transposasesbecameDNA-bindingproteins,suchasCENP-Binmam-malsandDaysleeperinArabidopsisthaliana[12,13].
Non-LTRretrotransposonsandPenelope-likeelementsuseacombinationofRTandENintheirtransposition[14-17].
Helitronistheonlygroupofeukaryotictranspo-sonsencodingrolling-circlereplicationinitiator[9].
YRgenesareubiquitousinprokaryotesbutrareineukaryotes[18,19].
AllYRsfoundineukaryotesareencodedbymobileelements:yeast2-microncircleplas-mids[20],ciliateEuplotescrassustransposons(Tec1,Tec2andTec3)[21,22],threegroupsofretrotransposons(DIRS/Pat,NgaroandVIPER)[23-25],andCryptons[19].
TheYRencodedbytheyeast2-micronplasmid,knownas"flippase"(FLP),iswidelyusedforsite-specificrecom-binationintheFLP-FRTsystem[26].
Tec1andTec2transposonsencodeaDDE-transposaseinadditiontoYR,andthereforetheYRdomainsinthesetransposonsareprobablyinvolvedinresolvingtranspositioninter-mediates.
TodatetheonlyYR-encodingtransposonsfoundinthevertebrategenomesareDIRSandNgaroret-rotransposons.
Cryptonswereoriginallyfoundinabasi-diomyceteCryptococcusneoformansandseveralpathogenicfungi.
Theirboundariesaredifficulttochar-acterizebecausetheyhaveneitherterminalinvertedrepeats(TIRs)norlongdirectrepeats.
Insteadtheyhaveshortdirectrepeatsatbothtermini.
These4-or6-bpdirectrepeatsareconsideredsubstratesforrecombina-tion.
ByanalogytoprokaryoticYR-encodingtransposons,Goodwinetal.
[19]proposedthatCryptonsareexcisedfromthehostgenomeasanextrachromosomalcircularDNAandintegratedatadifferentlocusinthegenome.
YRtypicallyrecognizesrecombinationsitesconsistingoftwoinvertedrepeatsthatare11to13bplongandsepa-ratedbyasegment6to8bplong[27].
Recently,transpo-sonsencodingonlyaYRhavebeenfoundinseaurchin,insectsandcnidariansandclassifiedasCryptons[28,29].
YRcontainsfourcatalyticallyimportantresidues(R-H-R-Y),buttheiroverallsequenceidentityisverylowamongdifferentgenesandtransposons[18,19].
TheconservedtyrosineresiduedirectlybindstoDNAintherecombina-tionreaction.
Inthispaper,wereportCryptonsfromvar-iousspecies,includingmedakafish,andsixhumangenesoriginatedfromancientdomesticationeventsofCryptonYRs.
ResultsThediversityofCryptonelementsintermsoftheirsequenceanddomainstructureWeidentified94Cryptonelementsfrom24speciesrepresentinganimals,fungiandstramenopilesthatincludeoomycetesanddiatom(Figure1,Table1andAdditionalfile1).
PhylogeneticclusteringofCryptonsonthebasisoftheirYRdomainsequencesrevealedfourgroupsreflectingthesystematicsoftheirhosts(Figure2,opencircles),buttwoofthemwerenotstronglysup-portedphylogeneticallybecauseofthelowbootstrapvalues.
HereinwedesignatethemasCryptonF,CryptonS,CryptonAandCryptonItoindicatetheircorrespondinghosts:fungi,stramenopiles,animalsandinsects.
Cryp-tonAandCryptonIarestructurallysimilar;however,CryptonF,CryptonSandCryptonA/CryptonIhavedistinctproteindomainstructures(seeFigure1anddetaileddescriptioninthenextthreesections).
Becauseofthelowresolutionofthephylogenetictree,wecouldnotdeterminewhetherthereisanyrelationshipbetweenthesefourCryptongroupsandtootherYR-encodingele-ments,andwecannotruleoutthepossibilitythattheyhaveoriginatedindependently.
CryptonFelementsfromfungiandoomycetes,andCryptonF-derivedgenesWeidentifiedCryptonFelementsinninespeciesoffungiandfourspeciesofoomycetes(Table1andAdditionalfile1).
TheseelementsencodeaproteinthatincludesYRandGCR1_CDNA-bindingdomains(Figure1).
MostofthefungalCryptonsandthefiveoomyceteCryptonsareasso-ciatedwith6-bpterminaldirectrepeats,whicharelikelysubstratesforCryptonintegration(Additionalfile1).
InFusariumoxysporum,CryptonisfusedwithaMariner-typeDNAtransposonandthiscompositetransposonishearafternamedMarCry-1_FO(Figure1).
TheanalysisoffourMarCry-1_FOcopieswithmorethan97%identitytoeachotherrevealedthepresenceof16-bpTIRsandtargetsiteduplications(TSDs)oftheTAdinucleotide,indicatingthattheirMariner-typeDDE-transposaseisresponsiblefortransposition.
CryptonF-2_PSfromPhytophthorasojaeandrelatedelementsencodeaC48peptidase(Ulp1pro-tease)inadditiontoaYR(Figure1).
TheoomyceteCryp-tonFelementsarenestedinfungalCryptonFelementsinthephylogenetictree(Figure2),indicatingahorizontaltransferbetweenfungiandoomycetes.
KojimaandJurkaMobileDNA2011,2:12http://www.
mobilednajournal.
com/content/2/1/12Page2of17FourgenesfromSaccharomycescerevisiaewerederivedfromCryptonFelements(Figure3andAdditionalfiles2and3).
ItwaspreviouslyreportedthattheGCR1_Cpro-teindomainencodedbyGcr1,Msn1andHot1genesissimilartotheC-terminalpartoffungalCryptons[19].
Inadditiontothesethreegenes,wefoundthatCbf2/Ndc10containsaC-terminaldomainsimilartoCryptonFpro-teins.
ThecentralportionsofCbf2andGcr1aresimilartoCryptonFYRdomains,butthecatalyticsiteisnotpre-served(datanotshown).
VanderwaltozymapolysporacarriestwoparalogousgenesofGcr1andMsn1.
Candidatropicalisandrelatedspecies(Candidaalbicans,PichiastipitisandPichiaguilliermondii)harboranothergenederivedfromaCryptonFelement,representedbyXP_002548716inC.
tropicalis.
ItisdesignatedhereinasCrypton-derivedgene1(Cdg1)(Figure3).
TheonlydomainsharedbyCryptonFelementsandallCrypton-derivedgenesistheGCR1_Cdomain.
ThephylogeneticanalysisofGCR1_Cdomains(Figure3C)indicatesthatHot1andMsn1areparalogousandthatthegenerelatedtoHot1/Msn1inC.
tropicalisrepresentsanoutgroupofbothgenes.
Therefore,itislikelythatfourdomesticationevents(forHot1/Msn1,Gcr1,Cbf2andCdg1)occurredinthisgroup.
WecouldnotfindanyCryptoninsertionsinthesub-phylumSaccharomycotina(includingS.
cerevisiae,C.
tropicalisandrelatedspecies).
ThedistributionofCrypton-derivedgenesindicatesthatCryptonwasactiveGCR1_CCrypton-Cn1YR600bp/200aaMarCry-1_FOGCR1_CDDEYRCryptonF-2_PSGCR1_CC48YRCryptonS-1_PIYRCryptonS-5_PRHTHC48YRCryptonA-1_OLYRCryptonI-1_RProYRintronFigure1SchematicstructuresofCryptons.
Crypton-Cn1andMarCry-1_FObelongtotheCryptonFgroup.
YR=tyrosinerecombinase;GCR1_C=DNA-bindingdomain;DDE=DDE-transposase;C48=C48peptidase;HTH=helix-turn-helixmotif.
Table1DistributionofCryptonelementsClassificationPhylumorClassSpeciesaFungiBasidiomycotaCryptococcusneoformans[19]AscomycotaCoccidioidesposadasii[19],Histoplasmacapsulatum[19],Chaetomiumglobosum,Fusariumoxysporum,Ajellomycescapsulatus,Coccidioidesimmitis,Microsporancanis,Talaromycesstipitatus,NeosartoryafischeriZygomycotaRhizopusoryzaeAnimalsChordataOryziaslatipesEchinodermataStrongylocentrotuspurpuratus[28]HemichordataSaccoglossuskowalevskiiMolluscaLottiagiganteaArthropodaNasoniavitripennis[29],Triboliumcastaneum[29],Rhodniusprolixus,Aedesaegypti,CulexquinquefasciatusCnidariaNematostellavectensis[28]StramenopilesOomycetesPhytophthorainfestans,Phytophthorasojae,Phytophthoraramorum,Pythiumultimum,Saprolegniaparasitica,Hyaloperonosporaarabidopsidis,AlbugolaibachiiDiatomsPhaeodactylumtricornutumaCryptonelementsfromspecieswithoutreferencesarefoundinthisstudy.
KojimaandJurkaMobileDNA2011,2:12http://www.
mobilednajournal.
com/content/2/1/12Page3of17$WB7L&U\)B36:2&B$FDO&U\6B38=0<0B+V&U\,B19L',56B6.
=0<0B*J:2&B6P&U\6B36.
&7'B+V633DW&U\)B76EB;W',56',56B19$$9;DB&P&U\B&Q&U\$B63&U\6B38&U\6B3,=0<0B+V/DPEGD&U\)B3,',56B'5;HU'&U\,B&4.
&7'B'U',56B'3X&U\)B76',56B19&U\$B63&U\,B7&',56B;7&U\$B/*',56B;7&U\6B$/DL&U\6B36&U\)B36&U\)B&*OR&U\)B523LRQHHU',56B;7=0<0B+V45,&+B+V',56B$&&5(&U\,B53UREB+VEB'U7Q)/3B.
ODF',56B19&U\$B2/&U\,B$$.
&7'B;W&U\6B3,45,&+B;W&U\)B&,:2&B'P',56B;745,&+B'U',56B63',56B19&U\6B37UL',56B&5DB+V&U\6B3,=0<0DB'U=0<0B'U/B*J&U\,B7&D=0<0EB'U.
DQJDURR&U\6B36',56%B'5',56B;7',56B$&D:2&B7F&U\$B6.
DB;W0DU&U\B)2',56B19&U\6B+$UD=0<0B;W',56B'5:2&B6S9OIB$F139&U\$B19&U\$B6.
$7),3B7J&U\$B6.
)/3=0<0B;W/B2DCryptonICryptonSCryptonFCryptonA&DUF3504genesFigure2PhylogenyofCryptons,DUF3504genesandothereukaryotictyrosinerecombinases.
Thenumbersatnodesarebootstrapvaluesover40.
OpencirclesindicatetheclustersofCryptons,andfilledcirclesshowtheclustersofDUF3504genes.
YR=tyrosinerecombinase.
Prefixesofnamesareasfollows.
Cry=Crypton;1958=KIAA1958.
AccessionnumbersofDUF3504genesareshowninAdditionalfile5.
SequencesofthetransposableelementsaredepositedinRepbasehttp://www.
girinst.
org/repbase/.
Otherabbreviationsandaccessionnumbersareasfollows.
FLP=FLPrecombinaseofthe2-micronplasmidinSaccharomycescerevisiae(NP_040488);FLP_Klac=FLPrecombinaseoftheplasmidpKD1inKluyveromyceslactis(YP_355327);CRE=CrerecombinaseoftheenterobacteriaphageP1(YP_006472);Vlf1_AcNPV=verylateexpressionfactor1fromtheAutographacalifornicanucleopolyhedrovirus(NP_054107);Tn916=Tn916transposasefromEnterococcusfaecalis(NP_0687929);XerD=XerDfromEscherichiacoli(NP_417370);Lambda=lambdaphagerecombinase(NP_040609);At_Ti=recombinasefromtheAgrobacteriumtumefaciensTiplasmid(NP_059767);SpPat1fromStrongylocentrotuspurpuratus(obtainedathttp://biocadmin.
otago.
ac.
nz/fmi/xsl/retrobase/home.
xsl).
Suffixesforspeciesnamesareasfollows.
Animals:Hs=human,Homosapiens;Oa=platypus,Ornithorhynchusanatinus;Gg=chicken,Gallusgallus;Tg=zebrafinch,Taeniopygiaguttata;Ac/ACa=lizard,Anoliscarolinensis;Xt/XT=frog,Xenopustropicalis;Dr/DR=zebrafish,Daniorerio;OL=medaka,Oryziaslatipes;Cm=chimaera,Callorhinchusmilii;SP=seaurchin,Strongylocentrotuspurpuratus;SK=acornworm,Saccoglossuskowalevskii;Dm=fruitfly,Drosophilamelanogaster;Tc/TC/TCa=beetle,Triboliumcastaneum;NVi=parasiticwasp,Nasoniavitripennis;CQ=southernhousemosquito,Culexquinquefasciatus;AA=yellowfevermosquito,Aedesaegypti;DPu=waterflea,Daphniapulex;Acal=seahare,Aplysiacalifornica;Sm=bloodfluke,Schistosomamansoni;NV=seaanemone,Nematostellavectensis.
Fungi:RO=Rhizopusoryzae;CGlo=Chaetomiumglobosum;TS=Talaromycesstipitatus;CI=Coccidioidesimmitis;FO=Fusariumoxysporum.
Stramenopiles:PI=Phytophthorainfestans;PS=Phytophthorasojae;PU=Pythiumultimum;HAra=Hyaloperonosporaarabidopsidis;ALai=Albugolaibachii;PTri=Phaeodactylumtricornutum.
Plants:CR=Chlamydomonasreinhardtii.
KojimaandJurkaMobileDNA2011,2:12http://www.
mobilednajournal.
com/content/2/1/12Page4of17GCR1_CCrypton-Cn1YRHOT1MSN1GCR1CBF2/NDC10CDG1500aaSaccharomycescerevisiaeCandidaglabrataVanderwaltozymapolysporaZygosaccharomycesrouxiiLachanceathermotoleransKluyveromyceslactisAshbyagossypiiCandidatropicalisGcr1Msn1Hot1Cbf2Cdg1++++++++++++++++++++++++++++++Msn1Hot1Cbf2Cdg1Gcr1+RWB$J+RWB&J+RWB.
O&U\)B&*OR&GJB&D&EIB6F0VQB/W*FUB=U&U\)B36&EIB/W*FUB&J&U\)B0&&EIB&J&U\)B&,*FUB.
O0VQB.
O&U\)B3,&EIB9S&U\)B1)&U\SWRQB&Q0VQEB9S+RWB=U+RWB/W*FUB$J&GJB3V&EIB.
O0VQB&J*FUB6F&EIB$J&GJB3J&U\)B6D3D&EIB=U+RWB&W0DU&U\B)20VQB6F0VQDB9S0VQB$J*FUB/W&U\)B76*FUDB9S+RWB6F&GJB&W&U\)B52&U\)B76&U\)B&*ORFigure3DistributionandschematicstructuresofCrypton-derivedgenesinSaccharomycetaceaefungi.
(A)SchematicproteinstructuresencodedbyCrypton-derivedgenesandCryptons.
(B)DistributionofCrypton-derivedgenes.
Eachgeneidentifiedinthehaploidgenomeisrepresentedbyaplussymbol.
(C)ThephylogenyofCrypton-derivedgenesandCryptonsusingtheGCR1_Cdomainsequences.
Thenumbersatnodesarebootstrapvaluesover50.
AccessionnumbersofgenesareshowninAdditionalfile2.
"Cry"standsforCrypton.
Suffixesforspeciesnamesareasfollows.
Sc=Saccharomycescerevisiae;Cg=Candidaglabrata;Vp=Vanderwaltozymapolyspora;Zr=Zygosaccharomycesrouxii;Lt=Lachanceathermotolerans;Kl=Kluyveromyceslactis;Ag=Ashbyagossypii;Ct=Candidatropicalis;Ca=Candidaalbicans;Ps=Pichiastipitis;Pg=Pichiaguilliermondii.
KojimaandJurkaMobileDNA2011,2:12http://www.
mobilednajournal.
com/content/2/1/12Page5of17inthepastandthattheDNA-bindingdomainGCR1_CwasmostlikelyderivedfromCryptons.
CryptonS,anewgroupofCryptonsfromoomycetesanddiatomWefoundCryptonSelementsinsevenoomyceteandonediatomspecies(Figure1,Table1andAdditionalfile1).
CryptonSelementsdonotencodeanyGCR1_Cdomain,buttheC-terminalregionisconservedamongCryptonSelements.
CryptonSelementsareassociatedwith5-or6-bpterminaldirectrepeats.
ThemajorityofCryptonSele-mentsshareTATGGtermini.
SomeCryptonSelementsencodeanadditionalproteincontainingaC48peptidasedomain.
ThepeptidasesencodedbyCryptonSandCryp-tonFelementsinoomycetesbelongtothesamefamilyandarerelatedtotheUlp1proteasefamily.
DomainshufflingbetweentwogroupsofCryptonelementscouldexplainthesimilarity,butmoredataareneededtodeterminetherelationshipbetweenthesepeptidasesandothercellularpeptidases.
Cryptonsinanimals(CryptonAandCryptonIgroups)WeidentifiedCryptonsinsevenmetazoananimalsbelong-ingtofivephyla(Table1andAdditionalfile1).
CryptonIelementswerefoundonlyininsects,whereasCryptonAelementswerefoundinvariousanimals,includingcnidar-ians.
AnimalCryptons(bothCryptonAandCryptonI)havenoC-terminaldomain(Figure1).
WedidnotfindanyterminalrepeatsinanimalCryptons.
CryptonI-1_RProfromRhodniusprolixushostsanon-autonomousderiva-tivefamily,CryptonI-1N1_RPro,inwhich5'438bpand3'260bpare98%identicaltothoseofCryptonI-1_RPro.
Thisisthefirstreportofnon-autonomousCryptonele-ments.
Comparisonof50copiesofCryptonI-1_RProandCryptonI-1N1_RProrevealednoterminalrepeats(neitherdirectnorinverted).
Inmedaka,wealsofoundtwofamiliesofnon-autonomousderivatives(CryptonA-1N1_OLandCryptonA-1N2_OL)ofCryptonA-1_OL.
AsinthecaseofotherDNAtransposons,Cryptonnon-autonomousele-mentsaremuchmoreabundantthantheirautonomouscounterparts.
Wecansafelyruleoutthetheoreticallypossiblecontam-inationofthegenomicsequencesfrommedakausedinthisstudy.
First,weidentifiedmorethan2,700copiesofautonomousandnon-autonomousCryptonelementswithDNAsequenceidentitiestoconsensusrangingfrom59%to98%.
ThenucleotidediversityofCryptonsfrommedakaisconsistentwiththeirlong-termpresenceinthemedakalineage.
Second,wefoundmanyCryptonsequencesinthedatabaseofexpressedsequencetags(ESTs)fromthreedif-ferentmedakastrains:Hd-rR,CABandHNI(datanotshown).
WealsofoundseveralCryptonswithinsertedmedaka-specifictransposonssuchaspiggyBac-N1_OLandRTE-1_OL(Table2).
Crypton-derivedsequencesintheATF7IPgeneIdentificationofCryptonsinthreedeuterostomespecies(medaka,seaurchinandacornworm)promptedustoextendanalysisofCryptonsinchordates,includingfoursequencedactinopterygianspecies(Fugurubripes,Tetrao-donnigroviridis,GasterosteusaculeatusandDaniorerio).
AlthoughmultiplecopiesofCryptonelementswerefoundonlyinmedaka,sequencessimilartoCryptonswerefoundinvariouschordatespecies(Table3).
Mostofthemdonotencodeanyfunctionalrecombinases,owingtoframeshifts,deletionsandsubstitutionsatcatalyticallyessentialresidues.
However,twosimilarsequences(ABQF01015803fromthezebrafinchTaeniopygiaguttataandAAVX01068049fromthechimaera(elephantshark)Callorhinchusmilii)includeanintactopenreadingframeofYR(Figure4A).
Wedidnotfurtheranalyzethesequencefromchimaera,becausethesequencedregionwasonly2,661bpinlength.
TheCrypton-likesequenceinzebrafinchisinsideanintronofagenecodingforactivatingtranscriptionfactor7interactingprotein(ATF7IP)(Figure4B).
ThereisaYRsequenceattheorthologouslocusofchickenGal-lusgallus,whichencodesaprotein97%identicaltothatofzebrafinch,butitcontainsaframeshiftinsidetheYRregion.
TheorthologousYRsequencefromtheturkeyMeleagrisgallopavocontainsaframeshiftatthesameposition(datanotshown).
Becausethedivergencebetweenchickenandzebrafinchoccurredsome107mil-lionyearsago(MYA)[30],thisunusuallyhighsimilarityindicatesastrongselectionoperatingontheseYRsequences.
Anexon-intronpredictionprogramwouldpredictalternativesplicingintheATF7IPgenefromzebrafinch,althoughatpresenttherearenomRNAorESTscorrespondingtothefusiontranscript.
ItispossiblethattheYRistranslatedaspartoftheATF7IPproteinandretainscatalyticactivityinsomebirds.
UsingtheUniversityofCaliforniaSantaCruz(UCSC)GenomeBrowserhttp://genome.
ucsc.
edu/,wefoundthattherearepartialCryptonsequencesattheortholo-gouspositionsoftheATF7IPgenefromthehuman,horse,kangarooandplatypusgenomes(Figures4Band4C).
Therearealsocloselyrelatedsequencespresentinthegenomesofrhesusmacaqueandtarsier.
Therefore,theinsertionofCryptonintheATF7IPgenemusthaveoccurredinthecommonancestorofamniotesmorethan325MYA[30].
Noneofthemammalianortholo-goussequencesencodeintactYRproteins,andmanymammalianspeciesaremissingtheYRsequence.
Thisindicatesonlyaslight,ifany,selectivepressureonthissequenceinmammals.
AncientdomesticationofCryptonsinanimalsMostvertebrategenessimilartoCryptoncodeforpro-teins(Additionalfile4).
Inthehumangenome,thereareKojimaandJurkaMobileDNA2011,2:12http://www.
mobilednajournal.
com/content/2/1/12Page6of17sevenproteinssimilartoCryptonYRs,whichareanno-tatedaspartsofsixgenes(Figure5andAdditionalfile5).
TheKIAA1958genecontainstwoisoforms,bothofwhichincludeYR-derivedsequences.
Theothergenesarepotassiumchanneltetramerizationdomaincontain-ing1(KCTD1),zincfinger,myeloproliferativeandmentalretardationtype2(ZMYM2)/zincfingerprotein198(ZNF198),ZMYM3/ZNF261,ZMYM4/ZNF262andglu-tamine-richprotein1(QRICH1)(Figure5).
APSI-BLASTsearchoftheseproteinsagainsttheNationalCenterforBiotechnologyInformation(NCBI)conserveddomaindatabase(CDD)revealedthattheyshareadomainofunknownfunction(DUF3504superfamily;E-value≤1e-29).
Thesixgenesarewidespreadamongvertebrates(Figure6)andarehighlyconservedamongphylogeneticallydistantspecies(Table4).
Thephyloge-neticrelationshipofeachgeneagreedwiththatofspecies(datanotshown).
ThenucleotidesequencescorrespondingtoallsevenDUF3504domainswerepresentintheNCBIESTdatabase,indicatingtheirexpression.
ThedataclearlyshowthattheyareneitherpseudogenesnordefectiveCryp-tons(seetheaccessionnumbersofDUF3504genesinAdditionalfile5).
However,noneofthempreservetheYRcatalyticsite.
Allofthemlostthecatalytictyrosineandthesecondconservedarginine,andallbutKCTD1alsolosttheconservedhistidine.
Althoughtheresolutionislowbecauseofhighdiver-genceandtheshortlengthoftheYRsequence,animalDUF3504genestendtococlusterwithanimalCryptons(CryptonA)intheYRphylogenetictree(Figure2).
TherearefourindependentclustersofDUF3504genes:KCTD1,KIAA1958a,KIAA1958b/KIAA1958LandWOC/ZMYM/QRICH1(Figure2,filledcircles).
KCTD1coclus-terswithseveralanimalCryptons,andtheclusteringissupportedby100%bootstrapvalue.
Cryptonsformaparaphyleticcluster,whichindicatesthattheDUF3504Table2CryptoncopiescontaininginsertionsofothertransposableelementsChromosomeStartaEndaElementStartbEndbDirectioncIdentitydchr13510034435100443Crypton-1N1_OL470573c0.
86143510044435100648piggyBAC-N1_OL1204d0.
95123510064935101118Crypton-1N1_OL1474c0.
8728chr2348442264844361Crypton-1N1_OL1147d0.
948948443624844566piggyBAC-N1_OL1205d0.
985448445674844980Crypton-1N1_OL144570d0.
9294chr2376400797640510Crypton-1N1_OL1441d0.
891876405117640716piggyBAC-N1_OL1205d0.
946676407177640832Crypton-1N1_OL438573d0.
8814chr520330282033143Crypton-1N1_OL3116d0.
896620331442033348piggyBAC-N1_OL1205c0.
965920333492033803Crypton-1N1_OL113573d0.
9132chr52219385222194261Crypton-1N1_OL144573c0.
90302219426222194466piggyBAC-N1_OL1205d0.
97072219446722194605Crypton-1N1_OL1147c0.
9078chr876668667667209Crypton-1N1_OL21376d0.
907576672107667896RTE-1_OL2,6663,352c0.
979676678977668065Crypton-1N1_OL372550d0.
8667aSequencecoordinatesinthecorrespondingchromosome.
bCorrespondingcoordinatesoftheconsensussequencesofrepeatelementsfromRepbase.
cSequenceorientationrelativetoconsensus:(d)irectand(c)omplementaryorientation.
dIdentitytotheconsensussequences.
Table3MolecularfossilsofCryptonsinchordatesSpeciesAccessionnumbersDaniorerioBX530066*XenopustropicalisNP_001120376,AAMC01135377*,AAMC01082917*CallorhinchusmiliiAAVX01521991*,AAVX01068049*,AAVX01132927*CionaintestinalisXP_002124034,XP_002125964CionasavignyiAACT01002283*,AACT01041791*HalocynthiaroretziBAB40645OikopleuradioicaCBY34656BranchiostomafloridaeXP_00260067,XP_002595788,XP_002613958,XP_002613959,XP_002587732,XP_002607491*NucleotidesequencesincludingCryptonfragments.
KojimaandJurkaMobileDNA2011,2:12http://www.
mobilednajournal.
com/content/2/1/12Page7of17RHRYABCATF7IP_ChickenATF7IP_ZebrafinchAAVX01068049Crypton-1_SPCryptonA-2_SPCryptonA-1_SKCryptonA-2_SKCryptonA-3_SKCryptonA-1_OLATF7IP_ChickenATF7IP_ZebrafinchAAVX01068049Crypton-1_SPCryptonA-2_SPCryptonA-1_SKCryptonA-2_SKCryptonA-3_SKCryptonA-1_OLATF7IP_ChickenATF7IP_ZebrafinchAAVX01068049Crypton-1_SPCryptonA-2_SPCryptonA-1_SKCryptonA-2_SKCryptonA-3_SKCryptonA-1_OLATF7IP_ChickenATF7IP_ZebrafinchAAVX01068049Crypton-1_SPCryptonA-2_SPCryptonA-1_SKCryptonA-2_SKCryptonA-3_SKCryptonA-1_OL111111111QLSEQKVVRCVPYTCGIFDGTTLINCFEKQLKQEETSSENKEAVEAAQMNFNFPEDVLFGLEEEEEDAKSIKNTHKQTGWAANLLKQWLAKN-GKDP-SFELVPVSELNDIQLSDHKVVRCVPYTCGIFDGTTLINCFEKQLKQEETPSENKEAVEAAQTNFNFPEDVLFGLEEDEEDAKSIKNTHKQTGWAANLLKQWLAKN-GKDP-SFELVPVSELNDIMTEDQLAGDELLDAARRHVSVTEEDLIVLEEERNE----KNTRKQTDWAVNILKQWLIEK-GQDE-HFEVMSVEELNGVMAELETLPRFQLLDDDQMRDLIDSADS----KNTKNSVKFSMKIFEDFLQVI-NTDLDSVNKLSNSELDNVMAAAAFKVPTFEILSTAQMQELLNGTDS----ASTKKSIKFGFAKFEAFLGLK-QIELGEI-TTDRLRLDDVNDCHPTAHNKLSFADFVFVRRRCYFFVVMASKRHAELSPSELDNILLENDS----KRTRMATSNAISTFKQYLGSKFGYGEGALETFTATTLDTAMIPQHEFIFIFIITNVTFXFVTENASLGEATAQTTDSDGAEAGPSRPCRFVPVTDDDVRRTIETQQN----KNTAKKTLYDMRLLGQFLKTPEINEAREIHEIPPNELETIMKTKREVAVFNDWKKSM-TQEKRDLKDIPPSQLDALMSDFVVSFNLFGEFDNFDDWDAATEKERQEHQKRERRHKEVSPEELNFLEDDEDE----VNTKKITKWAVNIFRDFLAQK-NMDI-NFENYTATTLNES110110746767921083694LREFYCKIRN-HDGNTYSVASYKSMRAGLNRHLKMPPYNRQICLMKDKEFASANMVFVSMLRMLRIQGKDETHHHP-PIAADDLRKIKLSGVLGLHSPVALVNKVWFDLQLLREFYYTIRN-HDGNTYSVASYKSMRAGLNRHLKMPPYNRQICLMKDKEFASANMVFMSVLKMLRMQGKDETHHHP-PIAAEDLRKIKQSGVLGLHSPLALVNKVWFDLQLLREFYGTVRN-HDGNTYSISSYKSIRAGLNRHLKMPPHLRQICLMQDKEFTSANNVFLGVLKILRKQGKDETNHHP-PINAADLRKIRTSGVLGLHTPLALVNKVWFDLQLLQKFYAGARR-KNGSYYMKKSMLAIRFGLQRHFLN---CNKIDIIKHEDFSNSTRVFKCFSAMLKQEGKGVVTHKP-AISAEDMDKIQ--GSLDLDDPVGLQDKVFMDVMLLCTFYSSLRR-EDGSWYSKKTMQVIRYGVQRHFLD---LHNFDIRDKATFSQSVRMFKAVLVKLKKEGKGSTKHKP-AISASDMEIIQNSSVLDCDTPAGLQNKVFVDFMTLRTFYAEVRT-QKGEVYTTKSLQSLRYGIQRYFNQPPMRRNFDLVNGDEFADSRKMFAAVSKQLKSQGKGNVCHKP-AILPGDLSKLTTYFSRYSTSAPVLLHKVWFSIMLLCKFFISVRK-PSGDDYEPTSLRSMLSSFERHLKR--HKYMYSLISSIEFNEVRETIKAKQKKLKKDGKGNLPRTAEAATDEEIEILWESGQFGSHSPEAIINTMWFYNTVLSQFYLGIRK-KNGDEYEPDSISLYQKSIDRYLRD--NDYGVSIIRGDAFASSRRTLAAKRKHSKASGFGNLPNKSQPITKDQEEALLKAGQLGVDSATAIINTVWYNNTKLRLFYASVQSTKEGGEYSVASLRSLRAGINRHLRDVNIISDTVFKSSNAVFKAIMKRYRKSGKDTSSHHP-RIPESDLEKIRCSSALSPDTPLGLVRKVWFDIQL219219183171173201216144198HFTKRGREILRDLAPDAFVVEKDKNGRRYAVF-RYPGKGKNGE---DPHKM---GKMYDMPGDPN-CPVFSLELYLSKLPPEPPAFYLHPLKLTAEQMKEQHFAKRGREILRDLAPDAFVVEKDKNGRRYAMF-RYPGKGKNGE---DPQKM---GKMYDMPGDPN-CPVFSLELYLSKLPPEPPAFYLHPLKLTSEQMQEQHFAKRGREILRDLPPDAFVIKRDPNGRRYAML-KYTGKGRNRE---DPLKL---GRMYDMPGDIN-CPVTSLDVYLSKLPPDPPAFYLHPLKLTPEQIQEQYFCNRGRENLREMTLDSFDI-CDEGGKCSITL--KDTLTKNNRADKLEKSQ--GGVMIPTNG-PR-CPVASFLRYKDKLNPQCKSFWQRPANAQAKRELKSNPSSSYFCNRGRENLRELKPDDFRLETDEDGLRYIT--KRDQLTKNNREDDDEVSN--NGVMYEIPGSSK-CPVESFMQFVSKLNKDCPFLWQKPKAKKPEDGYFCRRGREGQRDLRGSHFMVKSDDNGSKYVIQ-VGSEVSKNHQCDDDGVS---GGIMYANNSNPRHCPVRAFELYLSKANRRCDALFQRPRDNYSADDHFGLRGNTEHRNMCWGDVSLCTDSSGREYLEFSERQTKTRTGENPRDTRKV--KPKMWDVPANPRRCPVAIFKKYLSLRPAGYTNSDDPYYIATHSRGLPRPGLFGLRGNNEHRQMEWGDVVLQSDENG-EFLVYNERLTKTRHGE-PGNTRSF--APKAYATPKTPDICPVRAYKAYAAHRPDRMNCSSAPFYLGIEYVPCTIAKCLARRGREGCRELTMASFSIHRDEAGAEYLSL-SHNPDTKNHKTPNDPHKQNLRGFMFARPGDPL-CPIQSFKKYISKCPPDAKSFYLHPKRSVTAAA312312276270266295317243294PVWYKREPMGVNYLGAMMPRISVAARL--SQRYTNHSLRTTTIQLLCEAGLGPGEPVWYKREPMGVNYLGTMMPRISVAARL--SQRYTNHSLRTTTIQLLCEAGLGPREIMAVTGHRSESAIRHYWSVWYKREPMGVNYLGSMMPRISIAARL--SQRYTNHSLRTTTVRLLCDAGLGAREIMALTGHRSESSIRNYWDQWYCNAPLGKNSIGDKMKTISSRAG---TKAYTNHCLRATSISTLQNAGFRDREIMSVSGHKAETSLKHYADNWYCNAPVGKNTMGNKMKQISQKAGC--SKLYTNHCLRATCITTLDRAGFESRDIQSVSGHHSEQSLRNYCDVWFENKPIGKNTLSEMMALISKCASL--SQRYTNHCIRATSITILSEAGFNNRHTMSVSRHRXELRAGYXNEQWFRHQPIGINKIGSLMKNMATAANLPPNKRLTNHSARKHLIQKLSDQNIPPTQIMQISGHRNIQSVNAYSGIWFRNQPMGINKLTTIMKSMSKAADL--TGKLSNHSARKTCVQRLLDAGVPPNTAAQLSGHKNVSSLNRYSEVWYSREPMGVNYLGAMLKKISEEVGL--SQIYTNHSLRSTAVGRLSDAGLESRQIMSVTGHRCESSLQAYWATF7IPchickenzebrafinchplatypushorsehumanYRExon4567891010kbFigure4Crypton-derivedsequenceinanintronofATF7IPgene.
(A)AlignmentofproteinscodedbydeuterostomeCryptonsandCrypton-derivedsequences.
Catalyticallyessentialresiduesareshownbelowthealignment.
(B)IllustrationoftheconservationofATF7IPloci.
ThepositionoftheYRsequenceisindicatedbytheopenbox.
BlackboxesrepresentexonsofthechickenATF7IPgene.
GrayboxesindicateconservedblocksbetweenchickenandrespectivespeciesbasedontheNetTracksoftheUCSCGenomeBrowserhttp://genome.
ucsc.
edu/.
Linesbetweengrayboxesindicatethatboxesareconnectedbyunalignablesequences.
(C)AlignmentofnucleotidesequencesofCrypton-derivedsequences.
KojimaandJurkaMobileDNA2011,2:12http://www.
mobilednajournal.
com/content/2/1/12Page8of17domainofKCTD1wasderivedfromaCryptonYR.
ThepositionofKIAA1958aisdistinctfromeitherCryptonAorCryptonI,andWOC/ZMYM/QRICH1isclusteredasasistergroupofallanimalCryptonAelements.
Therefore,phylogenyalonedoesnotsupportthedomesticationofanimalCryptonsleadingtoWOC/ZMYM/QRICH1andKIAA1958a.
TheDUF3504domainwasderivedfromYR,notviceversa,becauseDUF3504lacksthecompletecatalytictetradessentialforYRactivity.
YRisessentialfortransposition,andrepeatedgenerationofactiveYRsfromdefectiveYRsishighlyimprobable.
ThedistributionsofWOC/ZMYM/QRICH1andKIAA1958aarerestrictedtobilateriansandjawedvertebrates,respectively.
ApartfromCryptons,theonlyotherpossiblesourcesofYRsinanimalgenomesaretheretrotransposonfamiliesDIRSandNgaro[23,24].
However,allsearchesoftheYRsfromCryptonA-1_OL,Crypton-1_SPandCryptonA-1_SKmatchedtheDUF3504sequencewithE-values≤8e-12,whereasYRsofDIRSandNgarodidnotmatchtheDUF3504sequenceatall(evenwhenthethresholdE-valuewassetat100).
Severalrepre-sentativesofDUF3504areactuallyCryptonsequences;forexample,XP_001639277istheproteincodedbyCrypton-1_NV.
ThepatchydistributionofCryptonsandtheincon-sistencybetweenCryptonandhostphylogeniesindicateancientamplificationandextinctioneventsinCryptonevolution.
Theancientamplificationswouldhavegener-atedmanylineagesofCryptons,anditislikelythatWOC/ZMYM/QRICH1andKIAA1958aderivedfromlineagesofCryptonsthatarenowextinct.
WecannotcompletelyruleoutthepossibilitythatthetwogenesandCryptonAele-mentswereindependentlyderivedfromDIRS-likeretro-transposonsorsomeasyetuncharacterizedtypesofmobileelements,butthisimpliesindependentoriginsofCryptonAandotherCryptongroups(CryptonF,CryptonSandCryptonI).
Therefore,fourindependentdomesticationeventsofanimalCryptonsarethemostparsimoniousexplanationfortheoriginsofanimalDUF3504genes.
ArepresentativeofDUF3504fromHalocynthiaroretzi(BAB40645)hasorthologsinothertunicates:Cionaintes-tinalis(XP_002125964),Cionasavignyi(AACT01002283andAACT10141791)andOikopleuradioica(CBY34656).
TheycouldalsorepresentadomesticationeventofCryp-ton.
Anotherrepresentative(YP_025778)iscodedinthemitochondrionofthegreenalgaPseudendocloniumakine-tum.
ItcouldalsobeacandidateCrypton-derivedgene;owingtothelackofrelatedsequences,however,wedidnotanalyzeitfurther.
AllDUF3504genesencodemuchlongerproteinsthantheirDUF3504domains,anditispossiblethatthepre-existinggenescapturedentireCryptonprotein-codingsequences.
However,theonlyrecognizabledomainencodedbyanimalCryptonsisYR(DUF3504),andthereislittlesequencesimilaritybeyondYRsamongCryptonsthemselves.
Therefore,itisunlikelythatanysignificantsequencesimilaritywaspreservedbeyondtheDUF3504domainsbetweentheDUF3504andCryptonproteins.
DUF3504BTB/POZIsoformbKCTD1WOCP-richZincfinger/MYM-typeZMYM4ZMYM3ZMYM2QRICH1Crypton-1_SPQ-richIsoformaIsoformbKIAA1958200aaCryptonA-1_OLFigure5SchematicstructuresofDUF3504proteins.
KIAA1958genehastwoisoforms,eachofwhichencodesaDUF3504domain.
ThestructuresofKCTD1,KIAA1958,QRICH1,ZMYM2,ZMYM3andZMYM4arefromhumans.
ThestructureofWOCisfromDrosophilamelanogaster.
KojimaandJurkaMobileDNA2011,2:12http://www.
mobilednajournal.
com/content/2/1/12Page9of17KCTD1geneTheKCTD1genecontainstheDUF3504domaincon-finedwithinasingleexon.
Amongthevertebrategenes,theKCTD1DUF3504domainistheclosesttotheCryp-tonYRsintermsofproteinsequencesimilarity.
ThesequenceidentitybetweentheKCTD1DUF3504domain200MyWOCKCTD1KIAA1958aKIAA1958bATF7IPSpeciesCryptonKCTD1bKIAA1958aKIAA1958bKIAA1958LQRICH1ZMYM2ZMYM3ZMYM4ATF7IPHumanMouseDogOpossumPlatypusChickenZebrafinchLizardFrogZebrafishSticklebackPufferfishMedakaChimaeraLamprey+LanceletSeasquirtSeaurchinAcornworm++Seahare*+Bloodfluke-+FruitflySeaanemoneWOCFigure6DistributionofCryptonsandCrypton-derivedgenes.
Eachgeneidentifiedinthehaploidgenomeisrepresentedbyaplussymbol.
MinussymbolsindicatetheabsenceofCryptonsorCrypton-derivedgenes.
Asterisksindicatethepresenceoftheirdisruptedfragments.
ThebranchagesarebasedonTimeTree[30].
Theunitoftimeisindicated.
Crypton-derivedgeneslistedatnodesofthetreeindicatethetimesoftheirdomesticationbasedontheirdistributionindifferentspecies.
KIAA1958L,QRICH1,ZMYM2,ZMYM3andZMYM4arenotshown,becausetheywerelikelyderivedbygeneduplications.
Table4ProteinidentitiesbetweenDUF3504domainsintwospeciesComparisonKCTD1KIAA1958bQRICH1ZMYM2ZMYM3ZMYM4Human-chicken98%(268of274)99%(282of287)97%(299of309)92%(287of313)85%(265of315)88%(302of347)Human-zebrafish90%(273of304)86%(245of287)75%(229of306)47%(145of310)-52%(ZMYM4a)(176of345)52%(ZMYM4b)(169of331)Numbersinparenthesesrepresentthenumberofidenticalresidues(left)andthealigneddomainlength(right).
Gapsareexcludedinthecomparison.
KojimaandJurkaMobileDNA2011,2:12http://www.
mobilednajournal.
com/content/2/1/12Page10of17andtheYRofCrypton-1_SPis32%,whichexceedstheanalogousidentityamongdifferentlineagesofCryptons.
Forexample,Crypton-1_SPintheCryptonAlineageandCrypton-1_TCintheCryptonIlineageshowlessthan30%sequenceidentitytoeachother.
KCTD1encodestwoproteinisoformsofdifferentlengths.
Thelongerisoform(isoformb)containsbothanN-terminalDUF3504domainandaC-terminalBTB/POZ(Broad-complex,TramtrackandBric-a-brac/poxvirusandzincfinger)domain(Figure5),whereastheshorterone(iso-forma)containsonlytheBTB/POZdomain.
Theshorterisoformisapproximately80%identicaltotheKCTD15geneattheproteinlevel,andrelatedgenesarefoundinvariousorganisms,includinglancelet,seaurchinandinsects(Additionalfile6).
TheKCTD15genedoesnothaveanyDUF3504domainandisfoundingnathostomesfrommammalstochimaera.
WeinferthatKCTD1andKCTD15wereduplicatedfromasinglegeneintheearlyevolutionofvertebratesandafterthataCryptoncopywasinsertedupstreamoftheKCTD1gene,whichgeneratedanewtranscriptionalvariantencodingtheisoformb.
ThisinsertionshouldhaveoccurredbeforethebranchingoftheChondrichthyes(sharks,rays,skatesandchimaeras)about527MYA[30].
KIAA1958geneTheKIAA1958gene,ofunknownfunction,ispresentintwoisoforms(aandb)whichcontaindifferentDUF3504domainsencodedbydifferentexons(Figure5).
DUF3504domainsinisoformsaandbareonly26%identicaltoeachotherinthehumangenome.
Althoughalternativesplicinghasnotbeenconfirmedexperimentally,thehighconservationofbothDUF3504-codingsequencesindi-catesthatbothencodefunctionalproteins(Table4andAdditionalfile4).
NeitherofthetwoDUF3504-codingsequencesisinterruptedbyintrons.
Wefoundbothiso-formsinawiderangeoftetrapods(Figure6).
Zebrafishlacksisoforma,whereasChondrichthyes(chimaera)lackisoformb.
SomeActinopterygii(medaka,stickleback,fuguandpufferfish)lackbothisoforms.
Chicken,zebrafinchandplatypushaveanadditionalgenesimilartoKIAA1958isoformb,designatedKIAA1958L.
Theexonsencodingisoform-specificregionsofKIAA1958barepositionedupstreamfromthoseofKIAA1958a.
TheKIAA1958LgenelikelyoriginatedfromaduplicationofthesegmentincludingKIAA1958b-specificexonsbutnotincludingKIAA1958a-specificexons,andthisduplicationeventpredatedthebranchingbetweenmammalsandbirds.
KIAA1958LislessconservedthanKIAA1958iso-formsaandb.
TheDUF3504domainsofKIAA1958Lproteinsfromplatypusandchickenareonly44%identi-cal,whereasthoseoftheKIAA1958bare97%identicalbetweenthespecies.
KIAA1958aisnonfunctionalinchickenandzebrafinch,butitisintactinlizard.
IsoformbwasnotfoundinChondrichthyes,anditispossiblethatitoriginatedlaterinthelineageswhichbranchedoffChondrichthyes.
KIAA1958amighthaveoriginatedinthecommonancestorofgnathostomes.
Anotherpossibi-lityisthatbothisoformswereacquiredinthecommonancestorofgnathostomesandisoformsaandbhadbeenlostinthelineagesofActinopterygiiandChondrichthyes,respectively.
ZMYM2,ZMYM3,ZMYM4andQRICH1genesTheZMYM2,ZMYM3andZMYM4genesarepresentindiversegnathostomesfromhumantochimaera(Figure6).
ZMYM2,ZMYM3andZMYM4aresimilartoarthropodWOCintermsoftheirsequenceandstructure[31,32].
TheDUF3504domainsfromtheDrosophilamelanogasterWOCgeneandthehumanZMYM2geneare41%identi-calattheproteinlevel.
SomeintronsarealsopositionedatthecorrespondingsitesofZMYM2andWOC(Figure7A).
Inadditiontochordatesandarthropods,wefoundsequencefragmentsrelatedtoWOCinechinoderms(Strongylocentrotuspurpuratus),hemichordates(Saccoglos-suskowalevskii),mollusks(PinctadamaximaandAplysiacalifornica)andplatyhelminthes(Schistosomamansoni,S.
japonicumandSchmidteamediterranea)(Additionalfile5).
ThereisnoevidencethatWOCformsmultiplegenefamiliesininvertebrates.
TheZMYM2,ZMYM3andZMYM4genesarelistedinthedatasetofohnologsreportedrecentlybyMakinoandMcLysaght[33],whichmeansthattheywereduplicatedfromasinglegeneduringtworoundsofwhole-genomeduplicationintheearlyevo-lutionofvertebratesbeforethesplitbetweenjawedverte-bratesandagnathans[34,35].
ThesyntenyblocksofZMYM2,ZMYM3andZMYM4shareseveralgenesinadditiontoZMYMgenes(Figure7B).
Themostparsimo-niousscenarioisthattheWOC/ZMYMgenefamilyorigi-natedfromthedomesticationofCryptoninthecommonancestorofbilaterians.
TherearethreeotherZMYMgenes(ZMYM1,ZMYM5andZMYM6)inthesyntenyblocks(Figure7B),buttheyhavenoDUF3504domain.
TheN-term-inalpartofZMYM5issimilartothatofZMYM2,whereasthoseofZMYM1andZMYM6aresimilartothatofZMYM4.
Thesethreegenesarepresentonlyamongeutherianmammals.
Thesedatasupportinde-pendentgeneduplicationeventsinsideeachsyntenyblock.
ItisnoteworthythattheC-terminalpartsofZMYM1,ZMYM5andZMYM6derivedfromtranspo-sasesofhAT-typeDNAtransposons,butthesehAT-derivedsequencesarenotclosetoeachother.
TheC-terminalpartofZMYM6isclosetoCharlieelementsinthehumangenome,whereasthatofZMYM1isclosertoplanthATelementssuchasHAT-1_Madfromapple(datanotshown).
KojimaandJurkaMobileDNA2011,2:12http://www.
mobilednajournal.
com/content/2/1/12Page11of17TheQRICH1genewasfoundindiversevertebrates,includinglamprey(Figure6).
TheDUF3504domaininQRICH1isquitesimilartothoseofZMYM2,ZMYM3andZMYM4.
Besides,fiveofeightintronsofQRICH1wereatthesitescorrespondingtothoseofZMYM2,ZMYM3andZMYM4(Figure7Aanddatanotshown).
ThehighstructuralandsequencesimilaritybetweenWOC,ZMYM2/3/4andQRICH1indicatesthatQRICH1originatedfromeitherWOCorZMYMgenes.
IntheneighborhoodofQRICH1,wecouldnotfindanygenesparalogoustogenesinthesyntenyblocksofZMYM2,ZMYM3andZMYM4.
However,becauseQRICH1ispresentinthelampreygenome,itmusthaveoriginatedatthetimeclosetothewhole-genomeduplicationevents.
DiscussionEvolutionofWOC:thethird-oldesteventoftransposondomesticationThemostancienttransposon-derivedgeneknowntodateisTERT,whichwasgeneratedbythedomesticationofaPenelope-likeretroelement[4],andPrp8,aspliceosomalZMYM2_humanAAGCATACTTCCAGATGgt.
.
agGGTCAATATTCTCTCGASILPDGSIFSRZMYM3_humanCACACTCCTCCCCAACAgt.
.
agATACGGTGTTCTCTCGATLLPNNTVFSRZMYM4_humanTACAATACTTCCTAATGgt.
.
agGTTACATGTTCTCTCGCTILPNGYMFSRQRICH1_humanCCGGGTCACTCCACTTGgt.
.
agGCTATGTCTTGCCCAGCRVTPLGYVLPSWOC_sea_urchinGACCGTGCACCCCACCA------CCGGTCTCCTCCTGAACTVHPTTGLLLNWOC_sea_hareTAAAGTCAACAGCCAAGgt.
.
agGTCAGACACTTTGCCGAKVNSQGQTLCRWOC_bloodflukeTATCAGTCCTGAAG-GTCTTCTTATCTGTCGTISPEGLLICRWOC_fluitflyGCTTTACAACGATTCGCgt.
.
agAATACATTGTCACTCGCLYNDSQYIVTRZMYM2_humanTGCTACTTGTCTAAAAGgt.
.
agTCCACAGAATCTTAATCYLSKSPQNLNZMYM3_humanTTCTATCTCTCAAAATGgt.
.
agTCCTGAAAGCCTCCGGFYLSKCPESLRZMYM4_humanTTTTACCTGTCAAAATGgt.
.
agTTCTGAAAGTGTGAAGFYLSKCSESVKQRICH1_humanTTCTACCTCTTCAAATGgt.
.
agCCCCCAGAGTGTGAAAFYLFKCPQSVKWOC_sea_urchinTTTTACCTCTCCAAATGgt.
.
agCCCTGAGAGCATCAAGFYLSKCPESIKWOC_sea_hareTTTTTCCTTTCCAAATGgt.
.
agTCCGGAGAGCATCAAAFFLSKCPESIKWOC_bloodflukeGCATACTTCGCCCGTTGgt.
.
agTCCCCCAGAACTACTAAYFARCPPELLWOC_fluitflyTTTTATCTCTCCAAATGgt.
.
agCCCGGAGAGCGTGAAGFYLSKCPESVKGgt.
.
agGGAgt.
.
agANGgt.
.
agGGGgt.
.
agGGACGgt.
.
agGAgt.
.
agANGgt.
.
agGGgt.
.
agGGgt.
.
agTGgt.
.
agTGgt.
.
agTGgt.
.
agCGgt.
.
agCGgt.
.
agTGgt.
.
agTGgt.
.
agCgt.
.
agTgt.
.
agTgt.
.
agTgt.
.
agCgt.
.
agCgt.
.
agTgt.
.
agTgt.
.
agCGgt.
.
agGG-------GGgt.
.
agGCgt.
.
agAQgt.
.
agA.
.
ABGJB6GJB2GJA3ZMYM2PSPC1ZMYM5GJB1ZMYM3NONOZMYM6SFPQZMYM4ZMYM1DLGAP3C1orf212GJA4GJB3GJB4GJB5P3ZMYM6SFPQChr13Chr1ChrXFigure7ParalogousrelationshipsofWOC/ZMYM/QRICH1genes.
(A)TwoconservedintronpositionsamongWOC,ZMYM2,ZMYM3,ZMYM4andQRICH1.
Intronsareprintedinlowercaselettersandshaded.
Proteinsequencesareshownbelownucleotidesequences.
Theupperandlowerintronpositionscorrespondtothe20thand22ndintronsofhumanZMYM2,respectively.
(B)ThesyntenyblocksofZMYM2,ZMYM3andZMYM4.
OhnologousrelationshipsreportedbyMakinoandMcLysaght[33]areindicatedbydottedlines.
GJB=gapjunctionproteinb;GJA=gapjunctionproteina;DLGAP3=discslargehomolog-associatedprotein3;C1orf212=chromosome1openreadingframe212.
Othergenenamesaredescribedinthetext.
KojimaandJurkaMobileDNA2011,2:12http://www.
mobilednajournal.
com/content/2/1/12Page12of17componentderivedfromaretrointron(groupIIself-spli-cingintron)[36].
TERTretainsthecatalyticactivityofRT,butPrp8doesnot.
Thesetwogenesaresharedbyalmostalleukaryotes.
Anotherexampleofanancientdomestica-tioneventistheRAG1gene[5].
Itisdistributedwidelyamonggnathostomes,butnoRAG1orthologwasfoundinagnathans,includinglampreyandhagfish.
Giventhatagnathanshaveadifferenttypeofadaptiveimmunesys-temcalled"variablelymphocytereceptors"[37],thedomesticationofRAG1likelyoccurredinthelastcommonancestorofgnathostomesaftertheirbranchingfromagnathans.
OthertransposonsdomesticatedinthedistantpastareinHARBI1andPBDG5genes,bothofwhicharepresentinvertebratesfromhumanstoactinopterygianfish[38,39].
TheKCTD1b,KIAA1958aandKIAA1958bgenesareasoldasorolderthantheHARBI1andPBDG5genes(Figure6).
Atransposon-derivedCENP-B,ahighlycon-servedmammaliancentromere,andthreeCENP-B-likeproteins(Abp1,Cbh1andCbh2)infissionyeastresembleeachotherintermsoftheirsequencesandfunctions,buttheyderivedindependentlyfromdifferentpogo-liketrans-posases[40].
Thehumangenomeharborsasignificantnumberofgenesderivedfromtransposons[6].
Someofthemweredomesticatedinthedistantpast,andtherearenotracesofrelatedrepetitivesequencesorTEsfromwhichtheywerederived.
Forexample,theHARBI1genewasderivedfromPIF/HarbingerandPHSA(THAPdomain-containingprotein9,orTHAP9)fromaP-likeelement[38,41].
BothHARBI1andPHSAwerefoundbyscreeningmammaliangenesagainstDNAtransposonsfromzebrafish.
Similarly,thekeytoourfindingsofCryp-ton-derivedgeneswasscreeningofgenesagainstCryptonspreservedinmedaka,becausethereareonlyafewrem-nantsofCryptonsleftinvertebrategenomessequencedtodate,exceptinmedaka.
TheancestralgeneforWOC/ZMYMprobablyorigi-natedinthecommonancestorofallbilateriansmorethan910MYA[30].
Thisisthethird-oldesttransposondomes-ticationeventknowntodate,followingthetwodomestica-tioneventsofRT[4,36].
OurstudyindicatesthatdomesticationofCrypton-likeelementsineukaryoteswasrelativelycommoninthedistantpast.
ThisimpliesthatCryptonsareveryancientand,giventheirrareoccurrenceinthegenomicfossilrecordandtheirgreatdiversity,theywereprobablymuchmoreactiveinthedistantpastthaninmorerecentevolutionaryhistory.
FunctionalimplicationsfordomesticatedCryptonYRsNofunctionofDUF3504domainshasbeenreportedtodate.
Evenso,theYRoriginofDUF3504domainsimpliestheirfunctionstosomeextent.
YRformsamultimerwhenitbindssubstrateDNAduringrecombination[42].
Onthatbasis,wecanenvisiontwopossiblefunctionsderivedfromYRs:DNAbindingandprotein-proteininteraction.
ThereareseveralindicationsforfunctionsofdomesticatedYRs.
First,manygenesderivedfromYRsaretranscrip-tionalregulators.
Gcr1,KCTD1,WOC,ZMYM2,ZMYM3ZMYM4,andATF7IPareeithertranscriptionalactivatorsorrepressors[43-48].
Cbf2actsasacentromericproteindirectlybindingtocentromere-specificsequencesandisessentialforspindlepolebodyformation[49,50].
AlthoughtheseproteinsusuallycontainaDNA-bindingdomainotherthanDUF3504,exemplifiedbytheGCR1_CdomainofGcr1andCbf2,theDUF3504domaincouldalsoworkasaDNA-bindingdomain.
Second,thereisaninterestingresemblancebetweenfunctionsofdomesticatedDDE-transposasesandYRs.
DaysleeperisatranscriptionfactorderivedfromahATDDE-transposaseandbindsaspecificmotiffortranscriptionregulation[12].
CENP-Bisacen-tromereproteinderivedfromtheDDE-transposaseofapogo-liketransposon[13].
Inthesegenes,transposase-deriveddomainsactasDNA-bindingdomains.
Third,alargefamilyofprokaryotictranscriptionalactivators,AraC/XylS,showsstructuralsimilaritytoYRs.
Theoverallfoldofthe129-aminoacidproteinMarA,amemberoftheAraC/XylSfamily,almostentirelyrecapitulatestheYRdomainofCrerecombinase[51].
MarAcansimulta-neouslybindRNApolymeraseIIandDNAtoformatern-arycomplex[52].
ThesedatasupporttheputativefunctionofDUF3504tobeDNAorproteinbinding.
TodaterelativelylittleisknownaboutCryptons.
Therehavebeennostudiesoftheirtransposition,transcription,translationorregulation.
ThesequencesimilaritybetweenCryptonsisverylow,especiallyintheirnon-protein-cod-ingregions.
WecomparedDNAsequencesofCryptonsfromdifferentspecies,butwecouldnotfindanycon-servednucleotidesequencesamongthem.
Furthermore,allCryptondomesticationeventsareveryold.
Therefore,itisverydifficulttoproposeanyspecificfunctionsofDUF3504domains.
Instead,hereinweproposepotentialpathwaysinwhichDUF3504domainscouldbeinvolved.
KCTD1andKCTD15areparalogsthathavedivergedduringtheearlyevolutionofvertebrates(Additionalfile6).
KCTD1isoformb,generatedbyaninsertionofCryptonupstreamoftheoriginalKCTD1gene,iswidelyconservedamongjawedvertebrates(Figure6),althoughitisunclearwhethertheagnathanscarrytheKCTD1bgene.
ThehighconservationofKCTD1b(Table4)indicatesitsessentialfunctionsharedamongjawedvertebrates.
KCTD1repressestheactivityoftheAP-2atranscriptionfactor,andtheBTB/POZdomainisresponsiblefortheinteraction[46].
AP-2aplaysanessentialroleinneuralborder(NB)andneuralcrest(NC)formationsduringembryonicdevelopment[53].
NBistheprecursorofNC.
KCTD15isexpressedinNBandinhibitsNCinduction[54].
TheNCcellsareatransient,multipotent,migratorycellpopulationuniquetovertebrates.
Theygiverisetodiversecelllineages.
WecanspeculatethatbyaddingaKojimaandJurkaMobileDNA2011,2:12http://www.
mobilednajournal.
com/content/2/1/12Page13of17newprotein-proteinorprotein-DNAinteractionKCTD1bcancontributetothenetworkofNCformationthroughtheregulationofAP-2a.
AmongDUF3504genes,thefunctionofWOC/ZMYMisofspecialinterestbecausetwoofthegenesinthisgroup,ZMYM2andZMYM3,arelinkedtohumandiseases.
AchromosomaltranslocationbetweenZMYM2andfibro-blastgrowthfactorreceptor1(FGFR1)causeslymphoblas-ticlymphomaandamyeloproliferativedisorder[55].
AtranslocationinvolvingZMYM3isassociatedwithX-linkedmentalretardation[56].
Mutationsoftheirortholog,WOC,causelarvallethalityinD.
melanogaster[31].
WOC/ZMYMgene-encodedproteinsareinvolvedinvariousprocesses,includingtranscription,DNArepairandsplicing.
WOCisatranscriptionalregulatorthatcoloca-lizeswiththeinitiatingformsofRNApolymeraseII[31,32].
TheWOCproteinsalsocolocalizewithalltelo-meres,andmutantsofWOCareassociatedwithfrequenttelomericfusions[31,32].
ZMYM2,ZMYM3andZMYM4arecomponentsofamultiproteincorepressorcomplex,includinghistonedeacetylase1(HDAC1)andHDAC2[47,48].
ZMYM2bindstovarioustranscriptionalregula-torsincludingSmadproteins[57].
Italsobindstoproteinsinvolvedinhomologousrecombination,suchasRAD18,HHR6AandHHR6B,whicharehumanorthologsoftheyeastRADproteins[58],andtospliceosomalcomponentsincludingSFPQ(splicingfactor,proline-andglutamine-rich)[59].
Interestingly,theSFPQgeneisacomponentofthesyn-tenicclusterofZMYM4(Figure7B).
TheparalogofSFPQintheclusterofZMYM3isNONO(non-POUdomain-containing,octamer-binding),whichisapartnerofSFPQinheteromers[60].
PSPC1(paraspecklecomponent1)presentintheclusterofZMYM2alsoshowssimilaritytoSFPQandNONOgenes.
Inadditiontotheirinvolvementinsplicing,theSFPQproteinscontributetoDNArepairbyinteractingwithRAD51[61].
TheyarealsorecruitingHDAC1totheSTAT6transcriptioncomplex[62].
There-fore,itislikelythatWOC/ZMYMandSFPQ/NONO/PSPC1proteinscooperativelyactintranscriptionregula-tion,splicingandDNArepair,andthattheyhavecoe-volvedbymaintainingtheirfunctionalrelationships.
TheirDUF3504domainsmaycontributetosomeoftheprotein-proteininteractions.
EvolutionofCryptonsTodateCryptonshavebeenidentifiedinalimitednum-beroffungiandanimalspecies.
HereinwereportthepresenceofCryptonsinnewspecies,butinformationregardingtheiroveralldistributioncontinuestobepat-chy(Table1).
CryptonFelementsarepresentinthreephylaoffungi(Ascomycota,BasidiomycotaandZygomycota)andtwoordersofoomycetes(Peronospor-alesandSaprolegniales).
Ourphylogeneticanalysissup-portsthehorizontaltransferofCryptonFelementsbetweenfungiandoomycetes(Figure2),whichisconsis-tentwithfrequenthorizontaltransferofgenesbetweenthem[63].
CryptonSelementsarealsopresentintwooomyceteordersandonespeciesofdiatoms.
Bothoomy-cetesanddiatomsarelineagesofstramenopiles,andtheoriginofCryptonSelementscoulddatebacktotheircommonancestor.
AnimalCryptons(CryptonAandCryptonI)werefoundinsixphyla:Chordata,Echinodermata,Hemichordata,Arthropoda,MolluscaandCnidaria.
CryptonIelementshavethesameoverallstructureasCryptonAelementsandwereobservedonlyininsectgenomes.
ItispossiblethatCryptonIelementsconstituteabranchofCryptonAbuttheyhaveevolvedmorerapidlyininsects.
Theover-alldistributioninfungi,oomycetesandanimalsindi-catesthatCryptonswerelongpresentinthesethreeeukaryoticgroups,probablywithsomecontributionofahorizontaltransfer.
ItislikelythatCryptonsoriginatedinthecommonancestorofthesethreegroups,althoughbecauseofthelowresolutionoftheYRphylogeny,wecannotruleoutthepossibilityoftheirindependentorigins.
TheidentificationofCryptonelementsinmedakaissurprising.
ThenucleotidediversityofCryptonsinthemedakagenomeclearlyshowsthatCryptonsweremain-tainedinthelineageleadingtomedakaforalongtime.
ItispossiblethatCryptonsinvadedthemedakapopula-tionafterthesplitofmedakafromthethreeactinopter-ygianfishspecies(Gasterosteus,TakifuguandTetraodon),whosegenomeshavebeensequenced.
TheverticaltransferofCryptonsinthelineageleadingtomedakaisapreferablescenariobecauseofthedomesti-cationofCryptoninthecommonancestorofbilateriananimals,whichledtotheoriginofWOCgenes.
Inmostidentifiedhostorganisms,Cryptonsarepreservedinverylowcopynumbers(Additionalfile1).
WefoundseveralfragmentsofCryptonsinvariousvertebrates,includingzebrafish(Table3).
TheoriginofCrypton-derivedgenestookplaceatdifferenttimesduringtheevolutionofvertebrates(Figure6).
ThisisconsistentwiththehypothesisthatCryptonscontinuedtomaintainverylowcopynumbersinthevertebrategenomesandwereoccasionallyamplifiedincertainlineages.
ConclusionsThisstudyhasrevealedthediversityofauniqueclassofDNAtransposons,Cryptons,andtheirrepeateddomesti-cationevents.
TheDUF3504domainsaredomesticatedYRsofanimalCryptonelements.
OurfindingsaddanewrepertoireofdomesticatedproteinsandprovideKojimaandJurkaMobileDNA2011,2:12http://www.
mobilednajournal.
com/content/2/1/12Page14of17furtherevidenceforanimportantroleoftransposableelementsasareservoirfornewcellularfunctions.
MethodsDatasourceGenomesequencesofvariousspecieswereobtainedmostlyfromGenBank,andsequencesofknownCryp-tons,DIRSandNgarowereobtainedfromRepbasehttp://www.
girinst.
org/repbase/.
AllcharacterizedCryp-tonshavebeendepositedinRepbase.
SequenceanalysisCharacterizationofnewCryptonswasachievedbyrepeatedBLAST[64]andCENSOR[65]searchesusinggenomesequencesofvariousspecieswithCryptonsasqueries.
Allanalysesweredonewithdefaultsettings.
TheconsensussequencesofelementswerederivedusingthemajorityruleappliedtothecorrespondingsetsofmultiplealignedcopiesofCryptons.
Alignmentgapsweremanuallyadjustedtomaximizesimilaritytootherrelatedelements.
CharacterizationofDUF3504geneswasperformedbyBLASTsearchesagainstbothproteinandnucleotidedatabaseswithknownDUF3504genesasqueries.
Wepredictedexon-intronboundarieswiththeaidofSoftBerryFGENESH:http://linux1.
softberry.
com/berry.
phtmltopic=fgen-esh&group=programs&subgroup=gfindandmanuallyadjustedthemthroughthecomparisontoorthologoussequencesinotherspecies.
SequencealignmentandphylogeneticanalysisWeusedMAFFT[66]withthelinsioptionorMUSCLE[67]withdefaultsettingstoalignbothnucleotideandpro-teinsequencesofvariousCryptonsandCrypton-derivedproteins.
WeconstructedmaximumlikelihoodtreesbyusingPhyML[68,69]with100bootstrapreplicates[70]fortheaminoacidsubstitutionmodelLG.
Wealsocon-structedtreeswithothersubstitutionmodels,WAG,RtREVandDCMut,andwiththeNeighbor-joiningmethod,buttheresolutiondidnotimprove.
ThetreetopologysearchmethodwasNearestNeighborInter-change(NNI),andtheinitialtreewasBIONJ.
Thephylo-genetictreesweredrawnwithFigTree1.
3.
1softwarehttp://tree.
bio.
ed.
ac.
uk/software/figtree/.
AdditionalmaterialAdditionalfile1:PDFfilelistingCryptonelementsfoundinthisstudy.
Additionalfile2:PDFfilelistingCrypton-derivedgenesinfungi.
Additionalfile3:PDFfileshowingalignmentofCryptonsandCrypton-derivedgenesinSaccharomycetaceaefungiinfastaformat.
Additionalfile4:PDFfileshowingalignmentofCryptonsandCrypton-derivedgenesinanimalsinfastaformat.
Additionalfile5:PDFfilelistingtheaccessionnumbersforDUF3504genes.
Additionalfile6:PDFfileshowingalignmentofKCTD1,KCTD15andrelatedproteinsequencesinfastaformat.
Abbreviationsbp:basepair;EN:endonuclease;MYA:millionyearsago;RT:reversetranscriptase;TE:transposableelement;TERT:telomerasereversetranscriptase;TIR:terminalinvertedrepeat;TSD:targetsiteduplication;YR:tyrosinerecombinase.
AcknowledgementsThisworkwassupportedbyNationalInstitutesofHealthgrant5P41LM006252.
ThecontentissolelytheresponsibilityoftheauthorsanddoesnotnecessarilyrepresenttheofficialviewsoftheNationalLibraryofMedicineortheNationalInstitutesofHealth.
Authors'contributionsKKKinitiatedtheresearch.
KKKandJJperformedtheresearchandwrotethemanuscript.
Bothauthorsreadandapprovedthefinalmanuscript.
CompetinginterestsTheauthorsdeclarethattheyhavenocompetinginterests.
Received:19August2011Accepted:19October2011Published:19October2011References1.
KapitonovVV,JurkaJ:AuniversalclassificationofeukaryotictransposableelementsimplementedinRepbase.
NatRevGenet2008,9:411-414.
2.
MillerWJ,HagemannS,ReiterE,PinskerW:P-elementhomologoussequencesaretandemlyrepeatedinthegenomeofDrosophilaguanche.
ProcNatlAcadSciUSA1992,89:4018-4022.
3.
GreiderCW,BlackburnEH:ThetelomereterminaltransferaseofTetrahymenaisaribonucleoproteinenzymewithtwokindsofprimerspecificity.
Cell1987,51:887-898.
4.
GladyshevEA,ArkhipovaIR:Telomere-associatedendonuclease-deficientPenelope-likeretroelementsindiverseeukaryotes.
ProcNatlAcadSciUSA2007,104:9352-9357.
5.
KapitonovVV,JurkaJ:RAG1coreandV(D)JrecombinationsignalsequenceswerederivedfromTransibtransposons.
PLoSBiol2005,3:e181.
6.
VolffJN:Turningjunkintogold:domesticationoftransposableelementsandthecreationofnewgenesineukaryotes.
Bioessays2006,28:913-922.
7.
CurcioMJ,DerbyshireKM:Theoutsandinsoftransposition:fromMutokangaroo.
NatRevMolCellBiol2003,4:865-877.
8.
AzizRK,BreitbartM,EdwardsRA:Transposasesarethemostabundant,mostubiquitousgenesinnature.
NucleicAcidsRes2010,38:4207-4217.
9.
KapitonovVV,JurkaJ:Self-synthesizingDNAtransposonsineukaryotes.
ProcNatlAcadSciUSA2006,103:4540-4545.
10.
BaoW,JurkaMG,KapitonovVV,JurkaJ:NewsuperfamiliesofeukaryoticDNAtransposonsandtheirinternaldivisions.
MolBiolEvol2009,26:983-993.
11.
BaoW,KapitonovVV,JurkaJ:GingerDNAtransposonsineukaryotesandtheirevolutionaryrelationshipswithlongterminalrepeatretrotransposons.
MobDNA2010,1:3.
12.
BundockP,HooykaasP:AnArabidopsishAT-liketransposaseisessentialforplantdevelopment.
Nature2005,436:282-284.
13.
TudorM,LobockaM,GoodellM,PettittJ,O'HareK:ThepogotransposableelementfamilyofDrosophilamelanogaster.
MolGenGenet1992,232:126-134.
14.
FengQ,MoranJV,KazazianHHJr,BoekeJD:HumanL1retrotransposonencodesaconservedendonucleaserequiredforretrotransposition.
Cell1996,87:905-916.
15.
KojimaKK,FujiwaraH:Anextraordinaryretrotransposonfamilyencodingdualendonucleases.
GenomeRes2005,15:1106-1117.
16.
YangJ,MalikHS,EickbushTH:IdentificationoftheendonucleasedomainencodedbyR2andothersite-specific,non-longterminalrepeatretrotransposableelements.
ProcNatlAcadSciUSA1999,96:7847-7852.
KojimaandJurkaMobileDNA2011,2:12http://www.
mobilednajournal.
com/content/2/1/12Page15of1717.
PyatkovKI,ArkhipovaIR,MalkovaNV,FinneganDJ,EvgenevMB:ReversetranscriptaseandendonucleaseactivitiesencodedbyPenelope-likeretroelements.
ProcNatlAcadSciUSA2004,101:14719-14724.
18.
Nunes-DübySE,KwonHJ,TirumalaiRS,EllenbergerT,LandyA:Similaritiesanddifferencesamong105membersoftheIntfamilyofsite-specificrecombinases.
NucleicAcidsRes1998,26:391-406.
19.
GoodwinTJ,ButlerMI,PoulterRT:Cryptons:agroupoftyrosine-recombinase-encodingDNAtransposonsfrompathogenicfungi.
Microbiology2003,149:3099-3109.
20.
BroachJR,HicksJB:Replicationandrecombinationfunctionsassociatedwiththeyeastplasmid,2μcircle.
Cell1980,21:501-508.
21.
DoakTG,WitherspoonDJ,JahnCL,HerrickG:SelectiononthegenesofEuplotescrassusTec1andTec2transposons:evolutionaryappearanceofaprogrammedframeshiftinaTec2geneencodingatyrosinefamilysite-specificrecombinase.
EukaryotCell2003,2:95-102.
22.
JacobsME,Sánchez-BlancoA,KatzLA,KlobutcherLA:Tec3,anewdevelopmentallyeliminatedDNAelementinEuplotescrassus.
EukaryotCell2003,2:103-114.
23.
GoodwinTJ,PoulterRT:Anewgroupoftyrosinerecombinase-encodingretrotransposons.
MolBiolEvol2004,21:746-759.
24.
GoodwinTJ,PoulterRT:TheDIRS1groupofretrotransposons.
MolBiolEvol2001,18:2067-2082.
25.
LorenziHA,RobledoG,LevinMJ:TheVIPERelementsoftrypanosomesconstituteanovelgroupoftyrosinerecombinase-encondingretrotransposons.
MolBiochemParasitol2006,145:184-194.
26.
GolicKG,LindquistS:TheFLPrecombinaseofyeastcatalyzessite-specificrecombinationintheDrosophilagenome.
Cell1989,59:499-509.
27.
vanDuyneGD:Astructuralviewoftyrosinerecombinasesite-speciicrecombination.
InMobileDNAII.
Editedby:CraigNL,CraigieR,GellertM,LambowitzAM.
Washington,DC:AmericanSocietyofMicrobiologyPress;2002:93-117.
28.
JurkaJ,KapitonovVV:FirstCryptonsfrominvertebrates.
RepbaseRep2008,8:232-233.
29.
JurkaJ:FirstCryptonsfrominsects.
RepbaseRep2009,9:468-480.
30.
HedgesSB,KumarS:TheTimeTreeofLifeNewYork:OxfordUniversityPress;2009.
31.
RaffaGD,CenciG,SiriacoG,GoldbergML,GattiM:TheputativeDrosophilatranscriptionfactorWocisrequiredtopreventtelomericfusions.
MolCell2005,20:821-831.
32.
Font-BurgadaJ,RossellD,AuerH,AzorínF:DrosophilaHP1cisoforminteractswiththezinc-fingerproteinsWOCandRelative-of-WOCtoregulategeneexpression.
GenesDev2008,22:3007-3023.
33.
MakinoT,McLysaghtA:Ohnologsinthehumangenomearedosagebalancedandfrequentlyassociatedwithdisease.
ProcNatlAcadSciUSA2010,107:9270-9274.
34.
KurakuS,MeyerA,KurataniS:Timingofgenomeduplicationsrelativetotheoriginofthevertebrates:didcyclostomesdivergebeforeorafterMolBiolEvol2009,26:47-59.
35.
OhnoS:EvolutionbyGeneDuplicationNewYork:Springer-Verlag;1970.
36.
DlakiM,MushegianA:Prp8,thepivotalproteinofthespliceosomalcatalyticcenter,evolvedfromaretroelement-encodedreversetranscriptase.
RNA2011,17:799-808.
37.
PancerZ,AmemiyaCT,EhrhardtGR,CeitlinJ,GartlandGL,CooperMD:Somaticdiversificationofvariablelymphocytereceptorsintheagnathansealamprey.
Nature2004,430:174-180.
38.
KapitonovVV,JurkaJ:HarbingertransposonsandanancientHARBI1genederivedfromatransposase.
DNACellBiol2004,23:311-324.
39.
SarkarA,SimC,HongYS,HoganJR,FraserMJ,RobertsonHM,CollinsFH:MolecularevolutionaryanalysisofthewidespreadpiggyBactransposonfamilyandrelated"domesticated"sequences.
MolGenetGenomics2003,270:173-180.
40.
CasolaC,HucksD,FeschotteC:Convergentdomesticationofpogo-liketransposasesintocentromere-bindingproteinsinfissionyeastandmammals.
MolBiolEvol2008,25:29-41.
41.
HammerSE,StrehlS,HagemannS:HomologsofDrosophilaPtransposonsweremobileinzebrafishbuthavebeendomesticatedinacommonancestorofchickenandhuman.
MolBiolEvol2005,22:833-844.
42.
GuoF,GopaulDN,vanDuyneGD:StructureofCrerecombinasecomplexedwithDNAinasite-specificrecombinationsynapse.
Nature1997,389:40-46.
43.
HollandMJ,YokoiT,HollandJP,MyamboK,InnisMA:TheGCR1geneencodesapositivetranscriptionalregulatoroftheenolaseandglyceraldehyde-3-phosphatedehydrogenasegenefamiliesinSaccharomycescerevisiae.
MolCellBiol1987,7:813-820.
44.
RepM,ReiserV,GartnerU,TheveleinJM,HohmannS,AmmererG,RuisH:Osmoticstress-inducedgeneexpressioninSaccharomycescerevisiaerequiresMsn1pandthenovelnuclearfactorHot1p.
MolCellBiol1999,19:5474-5485.
45.
LiuL,IshiharaK,IchimuraT,FujitaN,HinoS,TomitaS,WatanabeS,SaitohN,ItoT,NakaoM:MCAF1/AMisinvolvedinSp1-mediatedmaintenanceofcancer-associatedtelomeraseactivity.
JBiolChem2009,284:5165-5174.
46.
DingX,LuoC,ZhouJ,ZhongY,HuX,ZhouF,RenK,GanL,HeA,ZhuJ,GaoX,ZhangJ:TheinteractionofKCTD1withtranscriptionfactorAP-2αinhibitsitstransactivation.
JCellBiochem2009,106:285-295.
47.
HakimiMA,DongY,LaneWS,SpeicherDW,ShiekhattarR:AcandidateX-linkedmentalretardationgeneisacomponentofanewfamilyofhistonedeacetylase-containingcomplexes.
JBiolChem2003,278:7234-7239.
48.
GockeCB,YuH:ZNF198stabilizestheLSD1-CoREST-HDAC1complexonchromatinthroughitsMYM-typezincfingers.
PLoSOne2008,3:e3255.
49.
JiangW,LechnerJ,CarbonJ:Isolationandcharacterizationofagene(CBF2)specifyingaproteincomponentofthebuddingyeastkinetochore.
JCellBiol1993,121:513-519.
50.
GohPY,KilmartinJV:NDC10:ageneinvolvedinchromosomesegregationinSaccharomycescerevisiae.
JCellBiol1993,121:503-512.
51.
GilletteWK,MartinRG,RosnerJL:ProbingtheEscherichiacolitranscriptionalactivatorMarAusingalanine-scanningmutagenesis:residuesimportantforDNAbindingandactivation.
JMolBiol2000,299:1245-1255.
52.
MartinRG,GilletteWK,MartinNI,RosnerJL:ComplexformationbetweenactivatorandRNApolymeraseasthebasisfortranscriptionalactivationbyMarAandSoxSinEscherichiacoli.
MolMicrobiol2002,43:355-370.
53.
deCrozéN,MaczkowiakF,Monsoro-BurqAH:ReiterativeAP2aactivitycontrolssequentialstepsintheneuralcrestgeneregulatorynetwork.
ProcNatlAcadSciUSA2011,108:155-160.
54.
DuttaS,DawidIB:Kctd15inhibitsneuralcrestformationbyattenuatingWnt/β-cateninsignalingoutput.
Development2010,137:3013-3018.
55.
XiaoS,NalaboluSR,AsterJC,MaJ,AbruzzoL,JaffeES,StoneR,WeissmanSM,HudsonTJ,FletcherJA:FGFR1isfusedwithanovelzinc-fingergene,ZNF198,inthet(8;13)leukaemia/lymphomasyndrome.
NatGenet1998,18:84-87.
56.
vanderMaarelSM,ScholtenIHJM,HuberI,PhilippeC,SuijkerbuijkRF,GilgenkrantzS,KereJ,CremersFPM,RopersHH:CloningandcharacterizationofDXS6673E,acandidategeneforX-linkedmentalretardationinXq13.
1.
HumMolGenet1996,5:887-897.
57.
WarnerDR,RobertsEA,GreeneRM,PisanoMM:IdentificationofnovelSmadbindingproteins.
BiochemBiophysResCommun2003,312:1185-1190.
58.
KunapuliP,SomervilleR,StillIH,CowellJK:ZNF198protein,involvedinrearrangementinmyeloproliferativedisease,formscomplexeswiththeDNArepair-associatedHHR6A/6BandRAD18proteins.
Oncogene2003,22:3417-3423.
59.
KasyapaCS,KunapuliP,CowellJK:Massspectroscopyidentifiesthesplicing-associatedproteins,PSF,hnRNPH3,hnRNPA2/B1,andTLS/FUSasinteractingpartnersoftheZNF198proteinassociatedwithrearrangementinmyeloproliferativedisease.
ExpCellRes2005,309:78-85.
60.
Shav-TalY,ZiporiD:PSFandp54nrb/NonO:multi-functionalnuclearproteins.
FEBSLett2002,531:109-114.
61.
RajeshC,BakerDK,PierceAJ,PittmanDL:Thesplicing-factorrelatedproteinSFPQ/PSFinteractswithRAD51Dandisnecessaryforhomology-directedrepairandsisterchromatidcohesion.
NucleicAcidsRes2011,39:132-145.
62.
DongL,ZhangX,FuX,ZhangX,GaoX,ZhuM,WangX,YangZ,JensenON,SaarikettuJ,YaoZ,SilvennoinenO,YangJ:PTB-associatedsplicingfactor(PSF)functionsasarepressorofSTAT6-mediatedIgεgenetranscriptionbyrecruitmentofHDAC1.
JBiolChem2011,286:3451-3459.
63.
RichardsTA,DacksJB,JenkinsonJM,ThorntonCR,TalbotNJ:Evolutionoffilamentousplantpathogens:geneexchangeacrosseukaryotickingdoms.
CurrBiol2006,16:1857-1864.
KojimaandJurkaMobileDNA2011,2:12http://www.
mobilednajournal.
com/content/2/1/12Page16of1764.
AltschulSF,MaddenTL,SchfferAA,ZhangJ,ZhangZ,MillerW,LipmanDJ:GappedBLASTandPSI-BLAST:anewgenerationofproteindatabasesearchprograms.
NucleicAcidsRes1997,25:3389-3402.
65.
KohanyO,GentlesAJ,HankusL,JurkaJ:Annotation,submissionandscreeningofrepetitiveelementsinRepbase:RepbaseSubmitterandCensor.
BMCBioinformatics2006,7:474.
66.
KatohK,KumaK,TohH,MiyataT:MAFFTversion5:improvementinaccuracyofmultiplesequencealignment.
NucleicAcidsRes2005,33:511-518.
67.
EdgarRC:MUSCLE:multiplesequencealignmentwithhighaccuracyandhighthroughput.
NucleicAcidsRes2004,32:1792-1797.
68.
GuindonS,DufayardJF,LefortV,AnisimovaM,HordijkW,GascuelO:Newalgorithmsandmethodstoestimatemaximum-likelihoodphylogenies:assessingtheperformanceofPhyML3.
0.
SystBiol2010,59:307-321.
69.
GuindonS,LethiecF,DurouxP,GascuelO:PHYMLOnline:awebserverforfastmaximumlikelihood-basedphylogeneticinference.
NucleicAcidsRes2005,,33WebServer:W557-W559.
70.
AnisimovaM,GascuelO:Approximatelikelihood-ratiotestforbranches:afast,accurate,andpowerfulalternative.
SystBiol2006,55:539-552.
doi:10.
1186/1759-8753-2-12Citethisarticleas:KojimaandJurka:Cryptontransposons:identificationofnewdiversefamiliesandancientdomesticationevents.
MobileDNA20112:12.
SubmityournextmanuscripttoBioMedCentralandtakefulladvantageof:ConvenientonlinesubmissionThoroughpeerreviewNospaceconstraintsorcolorgurechargesImmediatepublicationonacceptanceInclusioninPubMed,CAS,ScopusandGoogleScholarResearchwhichisfreelyavailableforredistributionSubmityourmanuscriptatwww.
biomedcentral.
com/submitKojimaandJurkaMobileDNA2011,2:12http://www.
mobilednajournal.
com/content/2/1/12Page17of17