Analysis of CACTA transposases reveals intron loss as major factor influencing their exon/intron structure in monocotyledonous and eudicotyledonous hosts

RESEARCH Open Access Analysis o CACTA ansposases e eals in on loss as majo ac o in luencing hei exon/in on s uc u e in monoco yledonous and eudico yledonous hos s Jan P Buchmann 1,4* , A i Löy ynoja 1 , Thomas Wicke 2 and Alan H Schulman 1,3 Abs ac Backg ound: CACTA elemen s a e DNA ansposons and a e ound in nume ous o ganisms. Despi e hei low ac i i y, se e al housand copies can be iden i ied in many genomes. CACTA elemen s anspose using a ‘cu -and-pas e’ mechanism, which is acili a ed by a DDE ansposase. DDE ansposases om CACTA elemen s con ain, despi e hei conse ed unc ion, di e en exon numbe s among a ious CACTA amilies. While ea lie s udies analyzed he ances al his o y o he DDE ansposases, no s udies ha e examined exon loss and gain wi h a iew o mechanisms ha could d i e he changes. Resul s: We analyzed 64 ansposases om di e en CACTA amilies among monoco yledonous and eudico yledonous hos species. The anno a ion o he exon/in on bounda ies showed a ange om one o six exons. A obus mul iple sequence alignmen o he 64 ansposases based on hei p o ein sequences was c ea ed and used o phylogene ic analysis, which e ealed eigh di e en clades. We obse ed ha he exon numbe s in CACTA ansposases a e no speci ic o a hos genome. We ound ha ancien CACTA lineages di e ged be o e he di e gence o monoco yledons and eudico yledons. Mos exon/in on bounda ies we e ound in h ee dis inc egions among all he ansposases, g ouping 63 conse ed in on/exon bounda ies. Conclusions: We p opose a model o he ances al CACTA ansposase gene, which consis s o ou exons, ha p eda es he di e gence o he monoco yledons and eudico yledons. Based on his model, we p opose pa hways o in on loss o gain o explain he obse ed a ia ion in exon numbe s. While in on loss appea s o ha e p e ailed, a pu a i e case o in on gain was ne e heless obse ed. Keywo ds: T ansposases, In on loss, Molecula e olu ion, DNA ansposons, Plan s Backg ound CACTA elemen s a e DNA ansposons ound in genomes ac oss he phylogene ic spec um, om algae [1] o ascu- la plan s [2-6] o animals [7,8]. The i s CACTA elemen desc ibed a he molecula le el was En-1 in Zea mays [2]; since hen, hey ha e been well documen ed in he g asses. Al hough CACTA elemen s usually do no accoun o he la gegenomesizes ounding asses,CACTA amilies ne e heless can be highly abundan . In a ew cases, howe e , including Tpo1 in Lolium pe enne ( yeg ass) and Caspa in he T i iceae, CACTA elemen s a e known o ha e con ibu ed conside ably o he expan- sion o he genome size o hei hos [9-12]. Mo eo e , CACTAs can in luence he e olu ion o he hos gen- ome in o he ways [12]. In Glycine max (soybean), CACTA elemen s can a ec lowe colo and cap u e hos genes [13-16]. CACTA elemen s a e some imes as- socia ed wi h egula o y elemen s o genes, he e o e possibly in luencing gene exp ession [10,17]. Despi e hei p e alence and impac , e olu iona y s udies abou * Co espondence: [email p o ec ed] 1 Ins i u e o Bio echnology, Viikki Biocen e , Uni e si y o Helsinki, PO Box 65, FIN-00014 Helsinki, Finland 4 P esen add ess: Ma ie Bashi Ins i u e o In ec ious Diseases and Biosecu i y, Cha les Pe kins Cen e , Uni e si y o Sydney, Sydney NSW 2006, Aus alia Full lis o au ho in o ma ion is a ailable a he end o he a icle © 2014 Buchmann e al.; licensee BioMed Cen al L d. This is an Open Access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License (h p://c ea i ecommons.o g/licenses/by/4.0), which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly c edi ed. The C ea i e Commons Public Domain Dedica ion wai e (h p://c ea i ecommons.o g/publicdomain/ze o/1.0/) applies o he da a made a ailable in his a icle, unless o he wise s a ed. Buchmann e al. Mobile DNA 2014, 5:24 h p://www.mobilednajou nal.com/con en /5/1/24 CACTA elemen s, o DNA ansposons in gene al, a e sca ce. The CACTA supe amily belongs o he Class II o ansposable elemen s, p oli e a ing by a ‘cu and pas e’ mechanism. In con as o Class I elemen s, which ans- pose ia an RNA in e media e and he e o e copy he o iginal elemen , CACTAs anspose he o iginal elem- en i sel . CACTA elemen s cons i u e app oxima ely 2 o 5% o a g ass genome [16,18]. Howe e , only ew ac i e CACTA elemen s ha e been iden i ied in plan s [2-6,19]. In addi ion, only se en pu a i e ansc ibed ansposases ha e been iden i ied in he T i iceae [10]. A ull-leng h CACTA elemen consis s o wo e minal in e ed epea s (TIRs) bo de ing wo open eading ames(ORFs), one encoding a ansposase and he o he , called ORF2, a p o ein o unknown unc ion. The i s and las 5 bp o he TIRs consis o he highly conse ed CACTA and TAGTG mo i s, espec i ely, hence he name o he elemen . The unc ion o he ORF2 p o ein has been de e mined in speci ic CACTA amilies o sup- po excision and ansposi ion [20]. Howe e , he ans- posase is he key ansposi ion enzyme. I binds o he TIR du ing excision, c ea ing a 3-bp a ge si e duplica- ion (TSD) [21]. The ca aly ic cen e o he ansposase is he acidic iad known as he ‘DDD/E’mo i , which is highly conse ed [22]. The p esence o CACTA elemen s ac oss he phylo- gene ic spec um and he highly conse ed ca aly ic co e o hei ansposases indica e an ancien p esence. In e - es ingly, he numbe o exons in ansposases among CACTA ansposons di e s e en among he g asses. T ansposases in ice we e ound ha ha e ou exons [23], while s udies in maize epo ed up o ele en exons o CACTA ansposases [2,24]. In he ecen ly se- quenced g ass B achypodium dis achyon, he exon num- be o ansposases among CACTA supe amilies anges om one o h ee. The e o e, he analysis o he exon/ in on con igu a ion o CACTA ansposases o e s an excellen oppo uni y o s udy he e olu iona y mecha- nisms o in on gain and loss in DNA ansposons. In addi ion, analyzing exon numbe a ia ions in such a highly conse ed and ancien gene as he CACTA ans- posase can o e a pe spec i e on he ‘in on-ea ly’and ‘in on-la e’models [25,26]. The goal o his s udy was o analyze he di e ences in exon numbe s in CACTA ansposases in monoco yle- donous and eudico yledonous plan s and o iden i y an e olu iona y mechanism o explain hose di e ences. This was accomplished using phylogene ic and compa a- i e analyses, which equi ed a solid and obus mul iple sequence alignmen (MSA). We cons uc ed such an MSA based on p o ein consensus sequences o 64 ans- posases om CACTA amilies anno a ed in en mono- co yledonous and eudico yledonous species. Ou phylogene ic analysis e ealed ha ancien CACTA lineages di e ged be o e he di e gence o he monoco yle- dons and eudico yledons, suppo ing an in on-ea ly model o CACTA ansposases. The analysis o he MSA iden i- ied conse ed exon/in on bounda ies and pu a i e in on gain among he ansposases examined. Combining hese analyses lead o a model o a pu a i e ancien CACTA ansposase, in which in on loss was he main mechanism shaping he exon/in on con igu a ions o cu en anspo- sases ound in monoco yledonous and eudico yledonous plan s. Resul s We analyzed 64 au onomous CACTA ansposases om en di e en monoco yledonous and eudico yledonous species. All analyzed ansposases a e de i ed om con- sensus sequences om dis inc i e CACTA amilies. Be- cause amilies o ansposable elemen s (TEs) di e om each o he based on he 80-80-80 ule, hey we e consid- e ed o hologous [27]. The e o e, he name o he amily, o example, Cal in, will indica e he consensus se- quence o he ansposase and no he consensus o he whole elemen . We e e o he plan in which a CACTA amily and i s ansposase we e anno a ed as i s hos . Excep o ansposases iden i ied in B. dis achyon,we sea ched he PTREP [28] and Repbase [29] da abases o CACTA amilies wi h anno a ed ansposases (see Ma e ials and Me hods). The selec ion was based on wo c i e ia: i) he anno a ion had o clea ly s a e ‘ anspo- sase’, ha is anno a ions wi hou ORFs desc ibed as ans- posases we e omi ed because CACTA elemen s ha e wo ORFs, he ansposase and ORF2; ii) he p esence o wo ORFs was expec ed, he eby a oiding selec ion o anno a- ions ha ing a p edic ed ansposase ha spans mos o a consensus sequence, such as ATENSPM10 in Repbase, whe e he consensus is 8,272 bp and he p edic ed ansposase co e s posi ions 1,201 o 7,766. We se- lec ed nine ansposases om So ghum bicolo ,eigh ansposases om Z. mays, i e ansposases om T i icum aes i um,13 omO yza sa i a,and11 om B.dis achyon (Addi ional ile 1). This esul ed in a o al o 46 ansposases om monoco yledonous hos s. Fo he eudico yledonous da ase , we selec ed all anspo- sases om eudico yledonous hos s in Repbase i ing ou c i e ia, o aling in eigh een elemen s: se en ans- posases om elemen s anno a ed in A abidopsis hali- ana, i e omF aga ia esca, h ee omVi is ini e a, and one each om Pe unia hyb ida, Malus domes ica, and G. max (Addi ional ile 1). Anno a ion o exon/in on bounda ies on CACTA ansposases Fo simplici y, he e m ‘bounda y’will indica e exon/ in on bounda ies in his s udy. Excep o ansposases Buchmann e al. Mobile DNA 2014, 5:24 Page 2 o 15 h p://www.mobilednajou nal.com/con en /5/1/24 in B. dis achyon, bounda ies we e ex ac ed om he e- spec i e PTREP and Repbase en ies (Table 1, Ma e ial and Me hods). The ele en B achypodium dis achyon ansposases we e de i ed om consensus sequences o he au onomous amilies in his genome [18]. We manu- ally anno a ed he ansposases and bounda ies by align- ing o he mos simila BLASTX hi wi hin he PTREP da abase. Addi ional alignmen s agains ansc ip ion da- abases om ice and B. dis achyon did no inc ease he quali y o he bounda y p edic ions, because ansc ip- ome da a is sca ce o CACTA ansposases. De no o gene p edic ion did no e u n signi ican esul s. Ou inal da ase consis ed o 64 ansposases wi h 86 anno a ed bounda ies on he 40 ansposases ha con ained mo e han one exon (Table 1). Ou o he 64 anno a ed ansposases, 24 con ained only one exon and he e o e no bounda ies. On he emaining 40 ansposases, we anno a ed be ween wo and six exons (Addi ional ile 1). The leng h o he ansposases anged om 552 amino acids (amino acids; PSL, 1 exon) o 4,785 amino acids (EnSpm4_F es, 4 exons), and a e aged 1,163 amino acids. The six ansposases Isido , Ru us, Sand o, Radon, I an, and Isaac we e an- no a ed on he 3’end o he co esponding CACTA consensus sequence (Addi ional ile 1). Gene a ion o a obus mul iple sequence alignmen using con idence sco es Ou phylogene ic and compa a i e analyses we e based on an MSA de i ed om he selec ed 64 consensus anspo- sase p o ein sequences. Due o he possibly ancien o igin o ce ain CACTA ansposases and hei gene ally low ac- i i y, we assumed ha some pa s o sequences migh be mo e e olu iona ily di e ged han o he s. In addi ion, he o ma ion o consensus sequences can in oduce weak e- gions in o an MSA. A obus MSA is he e o e c ucial be- cause e o s o unce ain ies can in luence he downs eam analysis. In addi ion, iden i ying weakly aligned egions o posi ionsinanMSAand hen emo ing hemmayim- p o e downs eam phylogene ic analysis [30]. GUIDANCE is a me hod o in e un eliable egions in an MSA and emo e he po en ially e oneous signal om subsequen analyses ([31]; Ma e ials and Me hods). The inal MSA was 2,516 esidues long and con ained i e uns able egions placed be ween posi ions 120 o 186, 196 o 251, 381 o 416, 728 o 766, and in he 3’ end, s a ing om posi ion 1,665 (Addi ional ile 2). GUIDANCE sco es ange om 0 (low con idence) o 1 (high con idence) and a e calcula ed o single esidues as well as o whole columns. Because he e is no ec- ommended con idence sco e o esidues and columns in an MSA, a ade-o be ween sensi i i y and speci i- ci y is equi ed. High sensi i i y (low cu o alue) e- ains as many columns as possible while high speci ici y (high cu o alue) keeps only columns o e y high con idence. The de aul GUIDANCE cu o o 0.93 emo ed 638 columns (app oxima ely 25%) om he alignmen , in- cluding he badly aligned egions and 34 anno a ed bounda ies. Howe e , GUIDANCE kep columns wi h only one esidue, o example, mos o he badly aligned 3’end. To e ain as many bounda ies as possible o he analysis we applied ou own imming: we emo ed col- umns con aining only esidues wi h sco es below 0.804 (keeping bounda ies) and columns wi h only one esidue (no compa able and/o bad aligned). This app oach e- mo ed 1,398 columns (app oxima ely 44%): he badly aligned egions bu only 13 anno a ed bounda ies. This inal MSA was 1,118 esidues long and con ained 73 an- no a ed bounda ies in 64 ansposases (Figu e 1). Be- cause he i s bounda y is also he beginning o he i s in on, in ons we e named in he 5’ o 3’di ec ion and designa ed as subsc ip s o he name o he ansposase, o example, he i s in on and bounda y o ansposase Ba on is desc ibed as Ba on 1 . We mapped conse ed DDE mo i s [22] on o he MSA, which we e all in posi- ions wi h high con idence alues (Figu e 1). This MSA was used o all u he analysis. Exon numbe s in CACTA ansposases a e no speci ic o a hos genome RAxML [32] was used o calcula e he phylogene ic ee (Figu e 2). A maximum likelihood (ML) ee was gene - a ed based on 200 dis inc , andomized, maximum pa si- mony ees and i s obus ness assessed by using 1,000 boo s ap eplica es and by es ing he in luence o se - e al ou g oups (Addi ional ile 3, Ma e ial and Me hods). The esul ing ee shows he ela ion be ween indi idual ansposases bu no hei e olu ion o e ime; ha is he b anch leng hs do no indica e he ime when ans- posases di e ged om each o he bu how close hey a e on he molecula le el (Figu e 2). We iden i ied eigh clades, designa ed α o θ(Figu e 2). C ucially, he ans- posases g ouped p ima ily by hei exon numbe s a he han by hei hos s and he analysis o he clus e s ound no hos -speci ic exon numbe s o CACTA ansposases (Figu e 2). Ancien CACTA lineages di e ged be o e he di e gence o monoco yledons and eudico yledons We iden i ied h ee clades in which monoco yledonous and eudico yledonous ansposases clus e ed oge he . EnSpm2_Gmax om soybean g ouped in Clade αwi h ansposases om se e al monoco yledonous hos s, analo- gous o EnSpm3_F es and EnSpm4_F es om s awbe y in Clade ζ. Clade δg ouped ansposases om s awbe y, apple, and se e al g asses. The o he clades con ained only ansposases om ei he eudico yledonous o Buchmann e al. Mobile DNA 2014, 5:24 Page 3 o 15 h p://www.mobilednajou nal.com/con en /5/1/24 Table 1 Exon/in on bounda ies o he 34 analyzed CACTA ansposases wi h mo e han one exon. 12 3 4 5 EnSpm12_F es 462 | 564 G C 718 | 771 I EnSpm10_F es 826 | 846 Joey 842 | 893 II Janus 837 | 894 II F 846 | 894 II G 847 | 894 II No man 879 | 921 III En1 879 | 925 III Al ed 885 | 925 III H 838 | 972 EnSpm3_V in 821 | 894 II 856 | 925 III EnSpm8_Sbic 754 | 783 I 976 | 0 S o m 827 | 782 I 951 | 0 She man 831 | 782 I 954 | 0 J 495 | 521 750 | 885 EnSpm2_Mdom 755 | 782 I 886 | 893 II Baldu 731 | 782 I 837 | 895 II I 834 | 782 I 954 | 910 Isido 857 | 894 II 892 | 920 III Radon 841 | 894 II 877 | 921 III Ru us 851 | 894 II 887 | 921 III EnSpm13_V in 821 | 894 II 856 | 925 III EnSpm5_V in 824 | 894 II 859 | 925 III Isaac 861 | 894 II 900 | 925 III Sand o 744 | 782 I 851 | 928 III Balduin 850 | 895 II 890 | 930 III DOPPIA 843 | 894 II 890 | 936 III K 744 | 7,82I 850 | 936 III Ho ace 712 | 711 981 | 1,054 EnSpm4_F es 812 | 0 992 | 0 1,244 | 0 EnSpm3_F es 681 | 0 770 | 781 I 919 | 0 Seamus 730 | 782 I 833 | 892 II 878 | 925 III Da io 726 | 711 842 | 839 895 | 890 III A on 851 | 833 899 | 879 II 1,013 | 1,060 Ko bin 510 | 567 G 718 | 782 I 814 | 894 II 853 | 0 Ches e 520 | 563 G 728 | 777 I 823 | 889 II 858 | 920 III Ba on 522 | 568 G 730 | 781 I 825 | 893 II 861 | 925 III EnSpm8_F es 158 | 163 830 | 893 II 975 | 0 1,219 | 0 1,500 | 0 ATENSPM6_A hal 802 | 809 918 | 922 III 978 | 981 1,011 | 1,012 1,141 | 0 The posi ions a e ela i e o he beginning o he ansc ip ion s a and gi en as ollows: On he p o ein sequence | on he immed mul iple sequence alignmen (MSA). 0 and numbe s in i alic indica e bounda ies wi h GUIDANCE sco es below 0.804 and emo ed in he inal MSA. Supe sc ip s indica e Regions I o III and G clus e , espec i ely (Figu e 1). Buchmann e al. Mobile DNA 2014, 5:24 Page 4 o 15 h p://www.mobilednajou nal.com/con en /5/1/24 monoco yledonous hos s (Figu e 2). Despi e he long e o- lu iona y ime sepa a ing monoco yledonous and eudico- yledonous hos s, he p esence o mixed clades and he close ela ion o clades wi h only monoco yledonous o eudico yledonous hos s sugges s ha he CACTA anspo- sase phylogeny a he han he hos phylogeny is p ima y, ha is ha he main ansposase b anches di e ged al eady be o e he di e gence o monoco yledons and eudico yledons. Indeed, a close look a he phylogene ic ee e ealed ha ansposases wi hin clades end o ha e hesamenumbe o exons(Figu e2). The majo i y o CACTA ansposase bounda ies a e ound in h ee egions on he MSA To analyze he e olu ion o exon/in on a angemen s in CACTA ansposases, we compa ed he bounda ies om he 33 ansposases con aining 73 in ons ha we e no emo ed in he imming p ocess (Table 1, Figu e 1). We iden i ied 3 egions, labeled I o III, in he MSA, which con ain 63 ou o he 73 bounda ies (Figu e 1). Ou side hose egions, we iden i ied eigh bounda ies in- side he DDE mo i , ou bounda ies be ween Regions I and II, one bounda y be ween Regions II and III and i e bounda ies downs eam o Region III. Mos bounda ies a e close o each o he bu no in he same posi ion on he alignmen . This can be due o small e o s in o- duced by calcula ing he MSA o consensus sequences. The e o e, we analyzed he dis ances be ween bounda - ies o iden i y which we e sha ed among ansposases. We analyzed he bounda ies by clus e ing hem based on hei posi ions on he MSA. We se he maximal dis- ance be ween bounda ies s ill conside ed o be in he Figu e 1 Mul iple sequence alignmen based on p o ein sequences o he 64 analyzed CACTA ansposases. Colo ed boxes indica e amino acids, g ay boxes indica e esidues wi h a GUIDANCE sco e below 0.804, and whi e boxes indica e gaps in he mul iple sequence alignmen (MSA). The plo below he MSA shows GUIDANCE sco es o he co esponding posi ion in he MSA. Columns wi h a sco e below 0.804 a e indica ed in ligh blue while columns wi h a sco e o 0.804 and abo e in da k blue. Posi ions ela i e o he MSA and co esponding GUIDANCE sco e a e shown be ween he MSA and he plo . Highly conse ed DDE ansposase mo i s as desc ibed in [22] a e depic ed on op. In he phylogene ic ee, colo s indica e he hos as shown in he legend. Majo clades a e depic ed α o θ. Exon/in on bounda ies a e depic ed as blue ci cles i hei GUIDANCE sco e was abo e 0.804 and ed o he wise. The numbe in he bounda y indica es he bounda y numbe on he co esponding ansposase. Regions I o III a e indica ed by dashed lines and co esponding oman capi als. Posi ions o pu a i e in on gain a e depic ed as desc ibed in he legend. Buchmann e al. Mobile DNA 2014, 5:24 Page 5 o 15 h p://www.mobilednajou nal.com/con en /5/1/24 same egion o 16 esidues, which is hal he leng h o he sho es in on anno a ed (33 amino acids in ATENSP- M_A hal 3 ). Bounda ies ha we e close han 16 esidues o each o he we e g ouped oge he . No bounda ies wi hin a egion we e u he han 16 esidues apa (Tables 2, 3, Addi ional iles 4, 5, 6). The dis ances be ween he closes bounda ies o Regions I and II is 98 esidues (Addi ional ile 7), bu 30 esidues be ween Region II and III (Addi ional ile 7). The closes bounda y ups eam o Region I is 60 esidues away, whe eas he closes bounda y downs eam o Region III is 36 esidues away. This clus e ing con i med he p e iously iden i ied egions as clea ly dis inc . The ou bounda ies EnSpm10_F es 1 , Da io 2 ,A on 1 , and ATENSPM6_A hal 1 be ween Region I Figu e 2 Majo i y- ule based phylog am o he 64 analyzed CACTA ansposases. The phylogene ic ee is he same as in Figu e 1. Boo s ap alues ep esen he pe cen age ou o 1,000 boo s ap eplica es. Only boo s aps below 100% a e indica ed. T ansposase hos s a e colo ed as indica ed in he legend. Numbe s in pa en heses indica e he numbe o exons. Clades a e indica ed by dashed lines and labeled α o θ. Table 2 Dis ances be ween exon/in on bounda ies wi hin Region I Baldu 1 Ba on 2 1 Ba on 2 C 1 11 10 C 1 Ches e 2 5 4 6 Ches e 2 EnSpm2_Mdom 1 0 1 11 5 EnSpm2_Mdom 1 EnSpm3_F es 2 1 0 10 4 1 EnSpm3_F es 2 I 1 01115 0 1 I 1 K 1 0 1 11 5 0 1 0 K 1 Ko bin 2 0 1 11 5 0 1 0 0 Ko bin 2 Sand o 1 0 1 11 5 0 1 0 0 0 Sand o 1 Seamus 1 0 1 11 5 0 1 0 0 0 0 Seamus 1 She man 1 0 1 11 5 0 1 0 0 0 0 0 She man 1 S o m 1 0 1 11 5 0 1 0 0 0 0 0 0 Dis ances be ween exon/in on bounda ies in he MSA wi hin Region I (depic ed in Figu e 1). The dis ances a e gi en in esidues in he alignmen . Buchmann e al. Mobile DNA 2014, 5:24 Page 6 o 15 h p://www.mobilednajou nal.com/con en /5/1/24 and II, as well as I 2 be ween Region II and III could no be clus e ed in hose Regions. We iden i ied only one add- i ional clus e con aining ou bounda ies ou side Regions I o III. I g oups he i s in ons om all membe s o Clade γand was he e o e named Region G. Based on hese analyses o dis ances be ween all bounda - ies, we es ablished ha Regions I o III and G in he MSA we e clea ly sepa a ed om each o he as well as om all o he bounda ies. Gi en he dis inc ness o he ou bound- a y egions, we examined i he bounda ies hemsel es we e conse ed among he analyzed ansposases. Bounda ies in Regions I o III a e conse ed among mos ansposases while Region G ep esen s pu a i e in on gain Due o he p oximi y o bounda ies in Regions I o III and hei clea sepa a ion om o he bounda ies, we es ab- lished ha bounda ies wi hin a egion a e sha ed be ween he di e en ansposases. The clus e ing o bounda ies wi hin Regions I o III indica es ha he bounda ies a e conse ed among he analyzed ansposases. This is sup- po ed by he phylogene ic ee, in which pu ely mono- co yledonous o eudico yledonous clades sha e bounda ies (Figu e 1). Bounda ies in Region I a e on, o close o, he posi ion o he conse ed E om he DDE mo i , suppo - ing he claim ha Region I ep esen s conse ed bounda - ies among he ansposases (Figu e 1). The e o e, we conside ed he 63 bounda ies in Regions I o III as con- se ed wi hin each egion. All ansposases in Clade γ sha e hei i s in ons wi h a maximum dis ance o i e esidues (Figu e 1, Table 4). This is a unique clus e in he whole ee, indica ing in on gain since all membe s o Clade γsha e his in on bu none o i s ances o nodes and ansposases in o he clades. Only wo bounda ies om a monoco yledonous hos a e ound ou side Regions I o III We iden i ied 17 bounda ies ou side Regions I o III (Figu e 1). Only J 1 and H1 a e om a monoco yledonous hos , whe eas he emaining 15 bounda ies we e anno- a ed in ansposases om eudico yledonous hos s. Bounda ies I 1 and ATENSPM6 1,2,3 canno be clus e ed and he e o e we e no u he cha ac e ized. The ans- posases Ho ace, Da io, and A on ha e h ee sepa a e bounda ies which a e no a he apa han six esidues: Ho ace 1 and Da io 1 , Da on 2 and A on 1 , Ho ace 2 and A on 3 . While his appea s as ano he case o in on gain, hei ela ion in he phylogene ic ee is no p ope ly e- sol ed and does no suppo his in e p e a ion. Ou analysis o he bounda ies iden i ied 63 conse ed bounda ies and 4 cases o pu a i e in on gain in Region G. Mos conse ed in ons we e iden i ied in anspo- sases om monoco yledonous hos s. In con as , all unique bounda ies excep wo we e iden i ied in eudico- yledonous hos s. We decided o combine he esul s o he phylogene ic and bounda y analyses o de elop a model o unde s and how he obse ed exon/in on con- igu a ion e ol ed. De ining consensus exon numbe s o each phylogene ic clade A compa ison o he phylogene ic ee and he conse ed bounda ies e ealed a high consis ency be ween clades and bounda y posi ions. Based on he majo i y o exons pe clade, we cons uc ed a loose consensus o ep esen he exon numbe o ansposases in he co esponding clade. Fo example, Clade ζg oups oge he se en ans- posases o which ou , he majo i y, ha e wo exons. The e o e, a ep esen a i e ansposase om Clade ζhas wo exons and one consensus bounda y. We used his ap- p oach o each clade (Figu e 3). Ou app oach esul ed in ollowing exon numbe s o ep esen a i e ansposases: one exon o Clade α;Cladesβ,δ,andθ h ee exons each; Clade η ou exons; Clade γ i e exons. Designa ing con- sensus exon numbe s o each clade simpli ied u he he analysis o de elop a model o he loss and gain o bounda ies in CACTA ansposases. A model o loss and gain o exon/in on bounda ies in CACTA ansposases Because i had he la ges numbe o con i med exons, we compa ed all consensus bounda ies o Clade γ(Figu e 3). Clade αhas no anno a ed in ons. The second, hi d, and ou h in on o Clade γcan be ound h oughou he phylo- gene ic ee, whe eby he hi d in on o Clade γis he mos conse ed, ollowed by i s ou h and second in on. The ou h in on o Clade γis ound among Clades β,θ,ι,and in Isaac. The hi d in on is missing in he Clades EnSpm8, δ,andθ, bu o he wise is ound in all clades con aining in- ons. The second in on o Clade γis p esen in Clades δ, EnSpm8, and η. This compa ison indica es ha CACTA ansposases we e as a whole losing a he han gaining in- ons. Howe e , Clades γand ζha e in ons ha a e no ound in o he clades (Figu e 3), he i s in on in Clade γ ep esen ing an in on gain. The unique in ons in Clade ζ canno be classi ied as losses o gains because he phylogen- e ic ee does no allow a de ini i e classi ica ion. We p opose ha he consensus ansposase in Clade γ ep esen s he mos likely exon/in on con igu a ion o an ancien ansposase, con aining a leas ou exons and h ee in ons (Figu e 3). The h ee bounda ies co - espond o hose iden i ied in Regions I o III in he MSA (Figu es 1, 3). Using he pu a i e ances o model ansposase, we can in e he eme gence o he known ansposases h ough in on loss and gain (Figu e 3). Discussion In sum, we analyzed 64 CACTA ansposases om 11 monoco yledonous and eudico yledonous hos s. Ou Buchmann e al. Mobile DNA 2014, 5:24 Page 7 o 15 h p://www.mobilednajou nal.com/con en /5/1/24 Table 3 Dis ances be ween exon/in on bounda ies wi hin Region III Al ed 1 Balduin 2 5 Balduin 2 Ba on 4 0 5 Ba on 4 Ches e 4 5 10 5 Ches e 4 En1 1 0 5 0 5 En1 1 EnSpm13_V in 2 0 5 0 5 0 EnSpm13_V in 2 EnSpm3_V in 2 0 5 0 5 0 0 EnSpm3_V in 2 EnSpm5_V in 2 0 5 0 5 0 0 0 EnSpm5_V in 2 Isaac 2 0 5 0 5 0 0 0 0 Isaac 2 Isido 2 5 10 5 0 5 5 5 5 5 Isido 2 No man 1 4 9 4 1 4 4 4 4 4 1 No man 1 Radon 2 4 9 4 1 4 4 4 4 4 1 0 Radon 2 Ru us 2 4 9 4 1 4 4 4 4 4 1 0 0 Ru us 2 Sand o 2 3 2 3 8 3 3 3 3 3 8 7 7 7 Sand o 2 Seamus 3 05 05 00 0 0 054 443 Dis ances be ween exon/in on bounda ies in he MSA wi hin Region III (depic ed in Figu e 1). The dis ances a e gi en in esidues in he alignmen . Buchmann e al. Mobile DNA 2014, 5:24 Page 8 o 15 h p://www.mobilednajou nal.com/con en /5/1/24 phylogene ic analysis indica es di e gence o ancien CACTA lineages al eady be o e he di e gence o he monoco yledons and eudico yledons. The analysis o 73 bounda ies ac oss 33 ansposases wi h mo e han one exon iden i ied 55 conse ed exon/in on bounda ies and allowed us o econs uc he exon/in on con igu - a ion o a CACTA ansposase ep esen ing he ances al s a e be o e he di e gence o monoco yledonous and eudico yledonous plan s. The model consis s o a leas ou exons. We p opose a mechanism o he e olu ion o he ex an CACTA ansposases in which hey we e shaped mainly by in on loss, al hough one case o pu a- i e in on gain was ound. Po en ial o g ea e egula ion o CACTA elemen s in eudico yledons S udies o he PElemen in D osophila and Ac/Ds in maize ha e shown ha al e na i e splicing can egula e issue-speci ic ansposi ion o elemen s. Fo example, he Pelemen e ains i s hi d in on in soma ic cells, inhibi ing ansposi ion [33,34]. Should his occu wi h CATCA ansposases as well, ou da a sugges s ha ele- men s in dico yledonous hos s ha e mo e possibili ies o egula ion. In e es ingly, mos non-clus e ed bound- a ies and he pu a i e in on gain clus e we e ound in ansposases om dico yledonous hos s, whe eas he majo i y o bounda ies in Regions I o III we e ound in ansposases om monoco yledonous hos s. The num- be o ansposable elemen s in eudico yledonous ge- nomes is gene ally lowe han in monoco yledonous genomes, consis en wi h a igh e con ol o anspos- able elemen s in eudico yledonous hos s. The e o e, he la ge numbe o unique bounda ies ound ou side Re- gions I o III could be associa ed wi h mo e con ol o exp ession o CACTA elemen s in eudico yledons han in monoco yledons. Di e ences in in on gain and loss among TE ansposases P e iously, in on gain and loss in ansposases o DNA ansposable elemen s was s udied o Ma ine -like ele- men s in lowe ing plan s [35]. In ha s udy, degene a e p ime s we e used o ex ac agmen s o DDE anspo- sases om 54 plan species o phylogene ic analysis. The esul s we e consis en wi h e ical ansmission Table 4 Dis ances be ween exon/in on bounda ies wi hin Clus e G Ba on 1 Ches e 1 5 Ches e 1 EnSpm12_F es 1 4 1 EnSpm12_F es 1 Ko bin 1 14 3 Dis ances be ween exon/in on bounda ies in he MSA wi hin Clus e G (depic ed in Figu e 1). The dis ances a e gi en in esidues in he alignmen . Figu e 3 Model o he loss and gain o in ons in CACTA ansposases. Simpli ied phylogene ic ee based on he consensus exon numbe s pe clade as desc ibed in he ex . Below he ee he pu a i e ances o ansposase wi h ou exons is depic ed. Exons a e depic ed as g ay ec angles wi h in ons as colo ed lines. Blue, ed and g een depic in ons conse ed in Regions I o III, G indica es clus e G wi h he pu a i e in on gain. Conse ed in ons sha e he same colo band. In on loss is depic ed by i s co esponding colo and ci cled −, in on gain by an enci cled +. G ay balloons indica e how he obse ed con igu a ion a ose om he pu a i e ances o . Buchmann e al. Mobile DNA 2014, 5:24 Page 9 o 15 h p://www.mobilednajou nal.com/con en /5/1/24