International Journal o f Molecular Sciences Article Metallothionein Gene Family in the Sea Urchin Paracentrotus lividus: Gene Structure, Differential Expression and Phylogenetic Analysis MariaAntoniettaRagusa1,*,AldoNicosia2,SalvatoreCosta1,AngelaCuttitta2and FabrizioGianguzza1 1 DepartmentofBiological,Chemical,andPharmaceuticalSciencesandTechnologies,UniversityofPalermo, 90128Palermo,Italy;[email protected](S.C.);[email protected](F.G.) 2 LaboratoryofMolecularEcologyandBiotechnology,NationalResearchCouncil-InstituteforMarineand CoastalEnvironment(IAMC-CNR)DetachedUnitofCapoGranitola,TorrettaGranitola,91021Trapani, Italy;[email protected](A.N.);[email protected](A.C.) * Correspondence:[email protected];Tel.:+39-091-238-97401 AcademicEditor:MasatoshiMaki Received:6March2017;Accepted:5April2017;Published:12April2017 Abstract: Metallothioneins (MT) are small and cysteine-rich proteins that bind metal ions such as zinc, copper, cadmium, and nickel. In order to shed some light on MT gene structure and evolution,weclonedsevenParacentrotuslividusMTgenes,comparingthemtoEchinodermataand Chordatagenes. Moreover,weperformedaphylogeneticanalysisof32MTsfromdifferentclassesof echinodermsand13MTsfromthemostancientchordates,highlightingtherelationshipsbetween them. SinceMTshavemultiplerolesinthecells,weperformedRT-qPCRandinsituhybridization experiments to understand better MT functions in sea urchin embryos. Results showed that the expressionofMTsisregulatedthroughoutdevelopmentinacelltype-specificmannerandinresponse tovariousmetals. TheMT7transcriptisexpressedinalltissues,especiallyinthestomachandin theintestineofthelarva,butitislessmetal-responsive. Incontrast,MT8isectodermicandrises onlyatrelativelyhighmetaldoses. MT5andMT6expressionishighlystimulatedbymetalsinthe mesenchymecells. OurresultssuggestthattheP.lividusMTfamilyoriginatedafterthespeciation eventsbygeneduplications,evolvingdevelopmentalandenvironmentalsub-functionalization. Keywords:metallothionein;multigenefamilies;evolution;metal;echinoderms;embryonicdevelopment; geneexpression 1. Introduction Metallothioneins (MTs) represent a superfamily of widespread proteins existing of many organisms, ranging from prokaryotes to vertebrates. The superfamily consists of constitutive and stress-induciblemembers,withvariablemasses. Itisrichincysteine(Cys)residues(nearly30%oftheir aminoacidcomposition),andtheresiduesconstitutethemetal–thiolateclusters[1]. MTspossessgreat affinityforbothessential(zinc,copper,selenium)andxenobiotic(cadmium,lead,mercury)metals, bindingthemthroughspecificCys-CysandCys-Xxx-Cysmotifs. Usually,Cys-Cysmotifsarelocated intheC-terminalmoiety(alsoknownas↵-domain), andCys-Xxx-CysmotifsmaptheN-terminal half(or�-domain). MTsareknowntoexhibitaplethoraofbiologicalfunctions,includingprotection againstmetaltoxicity,controlofoxidativestressandregulationofphysiologicalhomeostasis[2–5]. Inadditiontotheircentralroleasmetalscavengers,MTsarealsoinvolvedinanumberofcellular activities,includingcellproliferation[6],differentiation[7,8],apoptosisandimmuneresponse[9,10]. Finally,MTshavealsogainedattentioninbiomedicalstudies,duetotheirproposedinvolvementin Int.J.Mol.Sci.2017,18,812;doi:10.3390/ijms18040812 www.mdpi.com/journal/ijms Int.J.Mol.Sci.2017,18,812 2of26 cancerandneurologicaldiseases[11–15]. MTshavebeenclassifiedintodifferentfamiliesonthebasis ofstructuralfeaturesandarrangementofCysmotifs[16,17]. Nevertheless,thelargenumberofMTs discoveredsofarclearlydemonstratestheexistenceofintermediateisoforms,makingtheclassification ofMTsverychallenging. Indeed, mechanisms of duplication, convergence and functional differentiation have created acomplexevolutionaryhistory,whichisdifficulttodefine[18–20]. Throughtheyears,severalstudiesexploringMTsindifferentorganisms(fromvertebrates,suchas humans,rodents,aves,andamphibians,toinvertebrates,suchasmolluscs,nematodesandinsects) havebeenreported[17]. Deuterostomesuperphylumdiversifiedaround510millionyearsago(Myr),intoechinoderms, hemichordates,tunicatesandvertebrates. TheEchinodermataphylumincludesseastars(Asteroidea), seaurchinsandsanddollars(Echinoidea),brittlestars(Ophiuroidea),seacucumbers(Holothuroidea) and sea lilies (Crinoidea). Thus, Echinoderms represent a very fascinating phylum since they are closelyrelatedtoChordates. AmongEchinodermata,seaurchinMThomologueshavebeenidentifiedinStrongylocentrotus purpuratus [21], Lytechinus pictus [22], Sterechinus neumayeri and Sphaerechinus granularis [23], revealing an unusual distribution of Cys motifs. Additionally, the three-dimensional structure analysisofS.purpuratusMTArevealedthatthisunusualCysmotifdistributioncausedaninverted architectureofthe↵-and�-domainswithrespecttovertebratestructure[24]. Previously,wereported the identification of five different MT homologues (PlMT4–8) from the Mediterranean sea urchin speciesParacentrotuslividus. Twofamilymembers,PlMT7andPlMT8,areconstitutivelyexpressed andupregulatedinresponsetocadmiumtreatment,whereasPlMT4,PlMT5andPlMT6appeartobe specificallyswitched-onaftercadmiumexposure[25]. Herein,withtheaimofbetterunderstandingtheevolutionaryrelationships,functionalvariety, andtheutilizationofMTsduringdevelopment,thegeneorganisationofP.lividusMTswasanalysed andtheirmRNAexpressionpatternswereunveiled.Particularly,wedeterminedtheexpressionprofiles andthespatialpatternsofP.lividusMTtranscriptsduringdevelopmentandaftermetaltreatments. Moreover,exploitingtheadvancesinhomologuesdetectionandhomologyproteinmodelling, theoretical structure calculation methods were applied. Evolutionary perspectives on MTs in deuterostomeswereaccomplishedcombiningphylogenyandgenefeatures. 2. Results 2.1. TheMetallothioneinGenesofP.lividus Theavailabilityoflarge-scaletranscriptionaldatasetsfortheMediterraneanseaurchinP.lividus allowed us to carry out a transcriptome survey for a comprehensive identification of the MT homologues. WeperformedBLASTNandTBLASTNsearchesusingMTcDNAsequencespreviously clonedasqueries[25]andaclusteringanalysisofMTexpressedsequencetags(EST)retrieved.NoMT4, MT5orMT6sequenceswerefoundinthedatabases,confirmingtheirlowexpression. Collectively, two MT7 transcript populations differing in length and three MT8 different populations were retrieved. Theiridentificationwascheckedmanuallyandthematchingsequenceswerereconfirmed bycomparativeanalysis. TheseresultssuggestthatMT7transcriptsmayderivefromasinglegeneby alternativesplicingormultiplepolyadenylationsignalsorevenfromtwodifferentgenes. Moreover, itispossibletohypothesisethepresenceofatleastthreeMT8genes. In order to identify and isolate the expressed MT genes, total genomic DNA from P. lividus spermwasextractedandamplifiedusingprimerpairsselectedasdescribedinMaterialsandMethods. Theamplifiedproductswerecloned,sequencedandanalysed. Fourgenomicclones,codingforMT4, MT5,MT6,MT7andthreediverseclonescorrespondingtoMT8(namedMT8a,b,c)wereobtained. ThecomparisonbetweencDNAandgenomicsequencesrevealedthatthetranscriptionunitsare composedbyfourexonsinterruptedbythreeintronsandaredifferentinlength. Thefirsttwointrons Int.J.Mol.Sci.2017,18,812 3of26 Int. J. Mol. Sci. 2017, 18, 812 3 of 25 interruptthecodingsequenceafterthefirstnucleotideofthecodon(phase-1),thelastintronislocated inthinet3er0rUupTtR t.hTe hcoedgineng esesqturuecntcuer aefsteorf ththe efiPrs.tl nivuidcluesotMidTe ogfe tnhees coadreonre (pprheasseen-1t)e,d thien laFsitg iunrtero1n. iAs llolcoaftetdh em possiens tshcea 3n′ oUnTicRa. lTshpel igceinnge sstirtuecst,uidreesn otiffi thede Pa.t l5iv-ideunsd MbTy gGeTneasn adrea rte3pr-eesnedntbeyd AinG Ficgounrsee 1n. sAulsl osef qthueemnc es. 0 0 Morpeoosvseers,sa cacnoomnpicaarla stpivliecianng asliytessi,s idbeentwtifeieedn aPt. 5li′v-eidnuds bayn GdTS a.npdu raptu 3r′a-etnuds MbyT AgGe nceosns(eSntrsouns gseyqlouceenncterso. tid Moreover, a comparative analysis between P. lividus and S. purpuratus MT genes (Strongylocentrotid diverged 35–50 Myr from the Parechinidae [26]) showed that all possess the same structure [27]. diverged 35–50 Myr from the Parechinidae [26]) showed that all possess the same structure [27]. Nevertheless,intronlengthsandsequencesaredifferentbetweenallhomologousgenes. Nevertheless, intron lengths and sequences are different between all homologous genes. Figure 1. Schematic gene structures of the Paracentrotus lividus and Strongylocentrotus purpuratus Figure 1. Schematic gene structures of the Paracentrotus lividus and Strongylocentrotus purpuratus metallothioneins (MTs; drawn to scale). The bent arrows indicate the putative transcription start sites metallothioneins(MTs;drawntoscale).Thebentarrowsindicatetheputativetranscriptionstartsites (TSS). Numbers under schemes indicate base pair numbers of exons and introns; boxes represent (TSS).Numbersunderschemesindicatebasepairnumbersofexonsandintrons; boxesrepresent exons: white boxes indicate untranslated regions, and coding regions are coloured. The grey diamond exons:whiteboxesindicateuntranslatedregions,andcodingregionsarecoloured.Thegreydiamond indicates one N stretch in the SpMTB1 intron. indicatesoneNstretchintheSpMTB1intron. In silico predictions showed two polyadenylation sites in the MT7 gene (score 0.876 and 0.898) wInhiscihli ccoouplrde deixcptiloanins sthhoew perdestewncoep oofl ytawdoe nMyTla7t iomnRsNiteAs isnpethcieesM dTi7ffegreennet (isnc olreeng0.t8h7 6duanridng0 .e8m98b)rwyoh ich coulddeevxeplolpaimnetnhte. presenceoftwoMT7mRNAspeciesdifferentinlengthduringembryodevelopment. MTM8bTa8nb danMdT M8cT8sch oshwoewdeda papprporxoixmimaatetelyly tthhee ssaammee lleennggtthh aanndd 9944%% ididenetnittyit. yB.oBtho tghegneesn ecosnctoanint ain 136a13d6d aitdiodnitaiolnbapl sbpins tinh ethfier fsitrsitn itnrtornona nandda a4 47766--bbpp ddeelleettiioonn iinn tthhee sesecocnodn dinitnrotrno wniwthi trhesrpeescpt etcot MtoTM8aT. 8a. In alignable sequences, the identity between MT8a and MT8b is 97%, higher than the identity with Inalignablesequences,theidentitybetweenMT8aandMT8bis97%,higherthantheidentitywith MT8c (93%). The MT8a and MT8b-expressed sequences show 98% of identity between them and 95% MT8c(93%). TheMT8aandMT8b-expressedsequencesshow98%ofidentitybetweenthemand95% with MT8c, differing for the presence of a simple AT-rich region in the 3′ UTR. withMT8c,differingforthepresenceofasimpleAT-richregioninthe30 UTR. 2.2. Predicted 3D Structural Model of P. lividus MTs 2.2. Predicted3DStructuralModelofP.lividusMTs On the basis of the computational analysis, we determined the key features of MT homologues Onthebasisofthecomputationalanalysis,wedeterminedthekeyfeaturesofMThomologues in P. lividus. As previously highlighted [25], only MT7 and MT8 follow the MT family 4 rule in P. lividus. As previously highlighted [25], only MT7 and MT8 follow the MT family 4 rule (Echinoidea: IPR001396; [17]): P-D-x-K-C-[V,F]-C-C-x(5)-C-x-C-x(4)-C-C-x(4)-C-C-x(4,6)-C-C located (Echinoidea:IPR001396;[17]): P-D-x-K-C-[V,F]-C-C-x(5)-C-x-C-x(4)-C-C-x(4)-C-C-x(4,6)-C-Clocated near the N terminus. Other isoforms instead have divergent amino acid sequences and in particular neartheNterminus. Otherisoformsinsteadhavedivergentaminoacidsequencesandinparticular MT4 and MT6 have a cysteine pattern that is slightly different. MT4andMT6haveacysteinepatternthatisslightlydifferent. Int.J.Mol.Sci.2017,18,812 4of26 Int. J. Mol. Sci. 2017, 18, 812 4 of 25 The MTs are organized in a N-domain and a C-domain bearing a pattern of conserved Cys The MTs are organized in a N-domain and a C-domain bearing a pattern of conserved Cys residues required for binding bivalent metal ions [17]. As occurred in S. purpuratus MTs, Cys-Cys residues required for binding bivalent metal ions [17]. As occurred in S. purpuratus MTs, Cys-Cys motifs,typicaloftheC-domaininmammals,arelocatedintheN-domain.Multiplesequencealignment motifInst,. Jt. yMpoli. cSacil. 2o01f7 , t1h8,e 8 1C2 - domain in mammals, are located in the N-domain. Multiple 4s oef q25u ence (MSA)analysisofP.lividusandS.purpuratusMTs(Figure2)showedthat,inadditiontoconserved alignment (MSA) analysis of P. lividus and S. purpuratus MTs (Figure 2) showed that, in addition to CyspatteTrnh,e tMheTas cacreep otergdaanmizeindo ina cai dNs-sduobmsatiintu atinodn sa dCo-dnoomtaailnw baeyasripngo sas epsasttseirmn iolaf rcponhsyesrivceadl cChyesm ical conserved Cys pattern, the accepted amino acids substitutions do not always possess similar physical featurreessi.dWuees arregquueirtehda ftosru bcihndchinagn bgievsamlenigt hmtentoalt isounpsp [o17rt].t hAes soacmcuerrsetdru icnt uSr. epuanrpdurcaotuusl dMaTltse, rCbyiso-cChyesm ical pchroepmemirctoaiteli fsfs,e, aatltyluoprwiecsain.l WgofMe taThreig nuCve-od tlohvmaetma isneu ncitnh i ncmhdaaminffgmeerasel snm,t aipgraeht thl onwcoaatty essdu. pinp otrhte thNe- dsoammaei ns.t rMucutlutirpel ea nsdeq cuoeunlcde alter biochaelimgnicmael nptr (oMpSeArt)i easn,a alyllsoisw oifn Pg. MlivTid uins vaonldv eSm. peunrpt uirna tduisf fMerTesn (tF pigautrhew 2a) yshs.o wed that, in addition to conserved Cys pattern, the accepted amino acids substitutions do not always possess similar physical chemical features. We argue that such changes might not support the same structure and could alter biochemical properties, allowing MT involvement in different pathways. FFiigguurree 22.. MuMltuipltliep sleeqsueeqnuceen acleignamligennmt (eMntSA(M) oSfA P). loivfidPu.s laivniddu Ss. paunrdpuSra.tupsu rMpuTr saetuqsueMncTess. eIdquenetnictieess. Iadnedn tciotinesseravnadtivcoe nssuebrsvtaittuivtieonsus basrtei tiunt iroends foanret. iRnerde dshfaodnint.g Rreepdressheandtisn igderneptirtyes aemntosnigd eanllt isteyquamenoc nesg. aDllostse Fqdiugeeunnroect ee2s .g .MaDpuolstt.is pNdleoe snne-oqctuoeenngscaeepr vasl.aigtNinvomen es-nuctob (nsMstieStruAvt)ai ootfin vPse . alsirvueib diusnts i btaulnatdcio kSn .f spounarrtpe.u irnatubsl aMckT fsoeqnut.ences. Identities and conservative substitutions are in red font. Red shading represents identity among all sequences. IInn oorrDddoeetrsr dttooen ooobtbett agaiainpns s.s oNommonee-c ioinnnddseiirccvaaatttiiiovoenn s ouonbns tssittturrutuioccnttusu rareree c cinhh abanlnagcgkee sfso,,n wwt. ee ccoommppuutteedd tthhee sseeccoonnddaarryy eelleemmeennttss and derived the 3D structures of P. lividus MTs (PlMTs). Different templates were selected to model andderivedthe3DstructuresofP.lividusMTs(PlMTs). Differenttemplateswereselectedtomodelthe the MediItne rorradneera tno osbetaa iunr scohmine iMndTicsa, toionn tohne sbtrauscitsu roef chhaenugreisst, iwcse tcoo mmpauxtiemd itshee csoecnofniddeanryc ee,l epmeernctesn tage MediterraneanseaurchinMTs,onthebasisofheuristicstomaximiseconfidence,percentageidentity and derived the 3D structures of P. lividus MTs (PlMTs). Different templates were selected to model identity and alignment coverage. If required, insertions were modelled ab initio. The generated and alignment coverage. If required, insertions were modelled ab initio. The generated models the Mediterranean sea urchin MTs, on the basis of heuristics to maximise confidence, percentage models were validated by assessing Ramachandran plot analysis and the percentage of residues in wereidveanlitditya taendd bayligansmseesnsti ncgovReraamgea. cIhf arneqduriarnedp, liontseartnioanlys swiseraen dmotdheellepde racbe nintaitgioe. oTfher egseidneureastedin the the favoured/allowed region ranged from 91% to 96%. In a manner similar to those described in favoumreodde/las lwloewree vdalridegatieodn brya ansgseesdsinfrgo Rmam91ac%hatnodr9a6n% p.lotI nanaalymsias nannder thsei mpeilracernttoagteh oofs reesdideuscesr iibne d in previous studies, PlMTs mainly consisted of coils and turns; while different numbers of helical previtohue sfasvtouudrieeds/,alPlolMweTds rmegaioinnl yrancogends ifsrtoemd o91f%c otiol s96a%n.d Intu ar nmsa;nwnehri lseimdiliaffre troe nthtonseu mdebsecrrisbeodf ihne lical ssttrruucctptuurrerevesiso wwusee rrseetu ccdooimmesp,p uPutlteMeddT (s(t thhmrreaeeien ↵lαy-- hhceoelnliicsceiesstse iidnn MoMf TTc4o4,i,l MsM aTTn66d aa ntnuddrn MMs;T Tw88,h, titwlwe ood aiafnfneddre ffnootuu nrru ↵αm--hhbeeelrliiscc eeossf iihnne MMlicTaTl55 aanndd MMTT77,,s rtrreeusscppteuecrctetisivv weelelyyr)e).. c omputed (three α-helices in MT4, MT6 and MT8, two and four α-helices in MT5 and MGeTn7e, rreaslplye,c ttihvee lgyl)o. bal structure of each protein resembled that of MTs from other organisms: two Generally, the global structure of each protein resembled that of MTs from other organisms: domains Gceonnenrealcltye, dth eb ygl oab afll esxtriubclteu rhei nofg eea. chH porwoteevine rr,e sMemTb4l eadn tdha tM oTf M5 Tpsr fortoemin ost haepr poergaarnedis mtos: tbweo more twodomainsconnectedbyaflexiblehinge. However,MT4andMT5proteinsappearedtobemore compdaocmt athinasn c MonTn6ec, t7e da nbdy 8a (flFeigxuibrlee 3h)i.n Tgeh.e H reolwateivveer ,s MolvT4e natn adc cMesTs5i bpilriottye i(nRsS aAp)p oeaf rtehde tCo ybse r emsoidreu es in compactthanMT6,7and8(Figure3). Therelativesolventaccessibility(RSA)oftheCysresiduesinthe compact than MT6, 7 and 8 (Figure 3). The relative solvent accessibility (RSA) of the Cys residues in the folded protein was also calculated. Based on RSA values, sea urchin MTs exhibited a similar foldedproteinwasalsocalculated. BasedonRSAvalues,seaurchinMTsexhibitedasimilarpattern the folded protein was also calculated. Based on RSA values, sea urchin MTs exhibited a similar pattern of solvent accessibility in the Cys, as such residues were found to mainly adopt a buried (RSA ofsolpvaetntetrna cocfe ssoslivbeinlitt ayccinestshibeilCityy sin, athses Cuychs, ares ssiudcuhe rseswideurees fwouerned fotuonmd taoi nmlyainaldyo apdtoaptb au bruierdied(R (RSSAA< 0.1) < 0.1) or intermediate (0.1 < RSA < 0.4) conformation. Conversely, exposed and highly exposed orint<e r0m.1)e doria itnete(0rm.1e<diaRtSe A(0<.1 0<. 4R)ScAo n<f o0r.4m) actoinofno.rmCaotniovne. rCseolnyv,eerxspeloys, eedxpaonsdedh aignhdl yhiegxhplyo seexdporseesdi dues residues (0.4 < RSA < 1 and RSA > 1 respectively) were found to be strongly underrepresented (Figure (0.4<reRsSidAue<s (10a.4n <d RRSASA < 1> a1ndre RspSAec >t i1v reelsyp)ewcteivreelyfo) uwnedre tfooubneds ttoro bneg sltyrounngldye urnredperrerespernetseednt(eFdi g(Fuirgeur3e) . 3). 3). Figure 3. Ribbon diagrams and surface representations of PlMTs. (A) General overview of the 5 PlMTs Figure3.RibbondiagramsandsurfacerepresentationsofPlMTs.(A)Generaloverviewofthe5PlMTs generated by homology modelling with cysteine residues labelled in orange. As far as possible, N- gFeignuerrae t3e.d Ribbybohno mdioalgorgaymsm aondde slluinrfgacwe irtehprceyssetneitnateiorness iodfu PelsMlTabs.e (lAle)d Giennoerraaln ogvee.rvAieswfa orf athsep 5o PsslMibTles, terminal is on the left and C-terminal is on the right; (B) Relative solvent accessibility (RSA) calculated Nge-nteerramas tipenedar lcbeisyno tahngoetmh oeof llceoofgrtryae nsmpdooCndd-etielnlrgimn cgiyn swateliiitnsheo crneystsihtdeeuirneiseg ihnrte ;es(aiBdch)u Rpesreo llataetbiinve.el lseodl vinen otraacncegses. ibAisli tfyar(R aSsA p)ocsaslicbullea, tNed- atesrpmeirncaeln itsa goen othfec olerfrte aspnodn Cd-itnegrmcyinstael iinse orne sthideu reigshint; e(Bac) hReplraottievien s.olvent accessibility (RSA) calculated as percentage of corresponding cysteine residues in each protein. Int.J.Mol.Sci.2017,18,812 5of26 Int. J. Mol. Sci. 2017, 18, 812 5 of 25 AccordingtothestructuralfeaturesofS.purpuratusMTs(SpMTs)[24]andinsilicoprotein–ligand bindiAngccosritdeinrge ctoo gtnhiet iostnr,ucptuorsasli bfleeatmureetsa lo–ft hSi.o plautrepucrlautustse rMTstsr u(SctpuMreTss)o [f24P]l ManTds inw esirleicop rperdoitcetiend–. Tlihgeanadb sbeinncdeinogf ssiptee crieficcogCnyistiorens, ipdousessibinle tmheetCal-–dtohmioalainte ocfluMstTer4 ,st5ruacntdur6esa osfw PelMll Tass twheerep rperseednicceteodf. aTnhea dadbisteionncea loCf yspserecisfiidc uCeyisn rtehseidNu-edso imn athine oCf-MdoTm6ariani soef tMheTp4o, s5s aibnidli t6y aths awtethlle asse thhoem porleosgeuncees wofe aren laikdedliytipornoanl eCtyosa crceespidtuvea riina ttiohne sNin-dmoemtaalinb inodf iMngTw6 irtahiosue ttmheo dpiofiscsaibtiiolintsy otfhtahte thtoetsael nhuommboelorgoufebso nwdeerde dliikveallye nptrocantei otnos .acEcelspetw vhaerriea,tiothnes Nin- manedtaCl -bdinodminaign swoitfhtohuets emhoodmifioclaotgiouness omf atyhel iktoetlayl ennucmombepra sosf dbiofnfedreednt mdievtaalle–ntht iocalattioencslu. sEtelsrse.where, the N- and C-domains of these homologues may likely encoAmspasshso dwifnfeirnenFti gmuerteal4–,thfiooulartea ncdlusthterrese. metal–thiolate clusters were identified in the N- and C-domAsa isnhroewspne citniv FeilgyuorfeM 4T, 7foaunrd aMndT 8t.hTrehee mcluetsatel–rtchoinonlaetcet icvliutisetserbse twweereen itdheenctyifsiteedin ien tthhioe laNte- garnodu pCs- adnodmtahien mreestpaelcitoivnesliyn otfh MeNT7- aanndd MC-Td8o.m Thaien csluosftMerT c7onannedctMivTit8iews ebreetwcoemenp tahree dcywstietihnteh tehiwolealtl-ed gerfionuepds SapnMd tThAe. mThetuasl, iaolngso irnit hthme sNto- asnudp eCr-imdopmosaeintsh oefp MolTy7p eapntdid MeTb8a cwkberoen ceosmfrpoamretdh ewseithh othmeo wloeglul-edsewfineerde aSpppMliTedAa. nTdhuans,a alolggooruisthsmpast itaol asurrpaenrgimempeonstes tohfeC pyoslryepseidputiedsea bmaocnkgbostnreusc tfurorems athnedsree lhaotemdomloegtuale–ss uwlfeurer calpupstleierds wanedre arnetarlioegvoeuds( Fsipgautirael 4a)r.rangements of Cys residues among structures and related metal– sulfur clusters were retrieved (Figure 4). Figure 4. Metal–thiolate cluster analysis and structural similarities between PlMT7, PlMT8 and Figure4.Metal–thiolateclusteranalysisandstructuralsimilaritiesbetweenPlMT7,PlMT8andSpMTA. SpMTA. (A) Cluster connectivities between the cysteine thiolate groups and the metal ions in the α- (A) Cluster connectivities between the cysteine thiolate groups and the metal ions in the ↵- and and β-domains of the sea urchin MTs. Cys residues are boxed in yellow, while the metal ions are �-domainsoftheseaurchinMTs.Cysresiduesareboxedinyellow,whilethemetalionsarespecifiedby specified by roman numerals; (B) Superposition of the 3D structures of PlMT7 and PlMT8 with romannumerals;(B)Superpositionofthe3DstructuresofPlMT7andPlMT8withSpMTA.Proteinsare SpMTA. Proteins are in ribbon representation. The Cys residues are in stick representation; a Gln inribbonrepresentation. TheCysresiduesareinstickrepresentation;aGlnresidueisalsoshown. TrehseidPu.lei visid aulssop rsohtoewinns.a Trehein Pi.v loivriyd,utsh per↵o-tdeionms aairne oinf SivpoMryT,A thies sαh-dowomnaiinnc oyfa SnpaMndTAth ies �sh-doowmna iinn cisyainn and the β-domain is in violet. Superposition was created and rendered using Chimera package. violet.SuperpositionwascreatedandrenderedusingChimerapackage. It should be noted that computational prediction of the metal binding sites in P. lividus as well ItshouldbenotedthatcomputationalpredictionofthemetalbindingsitesinP.lividusaswellas as in S. purpuratus MTs suggests alternative thiolate clusters, composed by Cys encompassing inS.purpuratusMTssuggestsalternativethiolateclusters,composedbyCysencompassingresidues residues from both the N- and C-domain. Thus, it could be hypothesised that a wide range of fromboththeN-andC-domain. Thus,itcouldbehypothesisedthatawiderangeofconformational conformational states and metal interactions between domains could be supported. statesandmetalinteractionsbetweendomainscouldbesupported. 2.3. Phylogenetic Analysis of Deuterostome MTs 2.3. PhylogeneticAnalysisofDeuterostomeMTs In order to study metallothionein evolution in deuterostomes, annotated protein sequences were Inordertostudymetallothioneinevolutionindeuterostomes,annotatedproteinsequenceswere retrieved from databases. Moreover, when not already annotated, MT sequences were searched by retrievedfromdatabases. Moreover,whennotalreadyannotated,MTsequencesweresearchedby similarity in transcriptome databases (Table 1). All available transcriptomes of Echinodermata similarityintranscriptomedatabases(Table1). AllavailabletranscriptomesofEchinodermataspecies species were considered, including Asteroidea (sea stars), Ophiuroidea (brittle stars), and were considered, including Asteroidea (sea stars), Ophiuroidea (brittle stars), and Holothuroidea Holothuroidea (sea cucumbers). Moreover, for understanding and gaining insight into the evolutionary trends of the MTs, phylogenetically interesting species belonging to non-vertebrate chordates as amphioxus and ascidians were considered [28–30]. Additionally, jawless vertebrates Int.J.Mol.Sci.2017,18,812 6of26 (seacucumbers). Moreover,forunderstandingandgaininginsightintotheevolutionarytrendsof theMTs,phylogeneticallyinterestingspeciesbelongingtonon-vertebratechordatesasamphioxus and ascidians were considered [28–30]. Additionally, jawless vertebrates (agnates/Cyclostomata), representativesofanancientvertebratelineage,andrelativelyearlydivergingteleostfishes,likeeels andcatfisheswereselected. Alsotwowell-studiedfishes,theJapanesetigerpufferandthezebra-fish, andahumanMTwerechosen[31,32]. Unfortunately, noacornworm(Hemichordata)norsealily (Crinoidea)MTsequenceswerefound. TheMSAbetweenMTsfromthesedifferentclassesisshown in Figure 5. The alignment was also used to construct the phylogenetic tree shown in Figure 6. Asanoutgroup,anMTfromtheciliateprotozoaTetrahymenapyriformiswasselected.TetrahymenaMTs constituteanexcellentexampleofanMTsubfamilywellstudiedintermsofmoleculargeneticsand proteinlevels[33–35]. Thissubfamilyisconsiderablydivergentfromthedeuterostomeparadigm, soitisunambiguouslyoutsidethecladeofinterestinthisphylogeneticstudy. Nevertheless,theycan be successfully aligned to sequences from the ingroup. Since the three MT isoforms identified in T. pyriformis are longer than deuterostome ones [36–38], we selected the shortest one (MT-2) as theoutgroup. The phylogenetic analysis of echinoderm MT sequences suggested that many events of gene duplicationhaveoccurredindependentlyindifferentspecies.Foreachspeciesthesequencedivergence isverydifferent,suggestingthatgeneduplicationoccurredafterspeciationsandatdifferenttimes. MTs of echinoderms exhibit high sequence heterogeneity, both among them and in relation to the vertebratepeptides,andPlMTsappearthemostdivergent. MTsofEucidaristribuloides,arepresentative cidaroid, branch apart from other sea urchin sequences, confirming cidaroids are the basal group within the class [39]. Echinarachnius parma, a representative clypeasteroida, results in the cidaroid sisterbranch. Even though the inter-relationships of the classic deuterostome phyla appear to be resolved, the intra-relationships within each of the phyla have been troublesome and not completely concluded[40].TheanalysisofthedifferentclassesofEchinodermataMTssuggeststhatophiuroidMTs weremorecloselyrelatedtoasteroidthantotheechinozoanMTsand,inturn,totheholoturianMTs. Asalreadyreported,theMTsequencesoftunicatesandamphioxusshowalowidentitywith otherdeuterostomeMTs,notbeingdirectlyalignabletotheisoformsofotherclasses[28,30]. Thus, BranchiostomaMTsappearinaseparatetreebranchwithrespecttothetwotunicatesandgroupwith theechinodermisoformsatagreaterdistanceofthevertebratecounterparts. TheseMTsarethemost similartoMTsofmolluscs(datanotshown). SeasquirtMTsaretheshortestMTssofaridentified indeuterostomesandareclearlyseparatedfromotherChordataandEchinodermataMTs. Finally, theMTsoflampreys, representativesofanancientvertebratelineagethatdivergedfromourown ~500Myr,asexpected,appearthenearestrelatedtoothervertebrataMTs. InordertoanalysealsoMTgenestructureevolution,newBLASTsearchesingenomicdatabases were performed using MT mRNA sequences as queries (Table 1). Genomic sequences of already annotatedMTgeneswerealsoretrievedfromdatabases. Novelannotatedgenesfromselectedgenome draftsequenceswerecomparedtoalreadyknownMTgenes. Int.J.Mol.Sci.2017,18,812 7of26 Table1.Species,protein,transcriptandgenesequenceIDscollectedforphylogeneticalandgenestructuralanalyses. Taxonomy Species Name IDProtein/Reference IDmRNA IDGene/Scaffold# EchinoideaEchinoida Paracentrotuslividus PlMT4/8 Ragusaetal.,2013 § EchinoideaEchinoida Sterechinusneumayeri SnMT P55953 SpMTA P04734.2 SPU_017989.3a Scaffold1168v3.1 Strongylocentrotus EchinoideaEchinoida SpMTB2 Q27287 SPU_017134.3c Scaffold1906v3.1 purpuratus SpMTB1 SPU_001866.3a Scaffold249v3.1 HtMT1 comp104953_c1_seq2⌃ EchinoideaEchinoida Heliocidaristuberculata § HtMT2 comp104953_c1_seq1⌃ SgMTA Q26497 EchinoideaTemnopleuroida Sphaerechinusgranularis SgMTB Q26496 EchinoideaTemnopleuroida Lytechinuspictus LpMT1 O02033 AlMT1 c31462_g1_i3¶ EchinoideaArbacoida Arbacialixula § AlMT2 c31386_g2_i2¶ Echinoidea Echinarachniusparma EpMT § GAVF01002251 Clypeasteroida EtMT1 JI315060 §JZLH010301553 EchinoideaCidaroida Eucidaristribuloides EtMT2 § GAZP01041405 EtMT3 JI295076 Asteroidea Asteriasforbesi AfMT § GAUS01062044.1 Asteroidea Asteriasrubens ArMT § GAUU01048766.1 Asteroidea Asteriasamurensis AamMT § GAVL01015559.1 Asteroidea Leptasteriassp.AR-2014 LsMT § GAVC01041863 OeMT1 GAUQ01108229 Ophiuroidea Ophiocomaechinata OeMT2 § GAUQ01027563 OeMT3 GAUQ01073893 HolothuroideaAspidochirotida Apostichopusjaponicus AjMT § GH551565 §MODV01037111+MODV01043468 §Scaffold4005KN886207.1+ HolothuroideaAspidochirotida Parasthicopusparvimensis PpMT § Locus_1_Transcript_31588/53328 Scaffold3285KN885487.1 HolothuroideaAspidochirotida Parastichopuscalifornicus PcMT1 § GAVO01014408.1 HolothuroideaDendrochirotida Sclerodactylabriareus SbMT § GAUT01018048.1 Int.J.Mol.Sci.2017,18,812 8of26 Table1.Cont. Taxonomy Species Name IDProtein/Reference IDmRNA IDGene/Scaffold# Crinoidea - Hemichordata - UrochordataEnterogona Cionaintestinalis CiMT1 ACN32211.2Franchietal.,2010 BW491384FJ217357 Scaffold186(JGIdatabase) UrochordataEnterogona Ascidiasydneiensissamea AsMT Yamaguchietal.,2004 Asy-sig-715andAsy-sig-997 HcMTA AY314949 UrochordataStolidobranchia Herdmaniacurvata § HcMTb AY314939.1 UrochordataStolidobranchia Herdmaniamomus HmMT § EL733027.1 BfMT1 BW764364 CephalochordataAmphioxiformes Branchiostomafloridae Guirolaetal.,2012 Guirolaetal.,2012 BfMT2_S FE561990.1 CephalochordataAmphioxiformes Branchiostomabelcheri BbMT20 XP_019631158.1 XM_019775599.1 NW_017803933(AYSS01018500) §APJL01076593+APJL01076594+ CraniataCyclostomata Lethenteroncamtschaticum LcMT § DC612982.1 APJL01076595 CraniataCyclostomata Petromyzonmarinus PmMT § CO548937.1 CraniataGnathostomataElopomorpha Anguillaanguilla AaMT ABF50549.1 DQ493910.1 §AZBK01727549 NM_001200077.1orAF087935 CraniataGnathostomataOtomorpha Ictaluruspunctatus IpMT O93571 NC_030419 andJT349175.1 CraniataGnathostomataOtomorpha Clariasmacrocephalus CmMT AGC79138.1 JX312865.1 DrMTBl ENSDARP00000131449 ENSDART00000170342 ENSDARG00000102051 CraniataGnathostomataOtomorpha Daniorerio (like) (CAA65933.1) ENSDARP00000061006 DrMT2 ENSDART00000061007 ENSDARG00000041623 (AAH49475.1) CraniataGnathostomata Takifugurubripes TrMT2 ENSTRUP00000022394 ENSTRUT00000022487 ENSTRUG00000008907.1 Euteleosteo-morpha CraniataGnathostomataPrimates Homosapiens HsMT1A NP_005937.2 NM_005946 GeneID:4489 Scaffold#:Scaffoldnumber.§Sequencesidentifiedinthiswork.⌃HeliocidarisT_Trinitydatabase.¶AvailableinEchinobase. Int.J.Mol.Sci.2017,18,812 9of26 Int. J. Mol. Sci. 2017, 18, 812 9 of 25 Figure 5. MSA of MT protein sequences. Conservation score higher than 0.5 is highlighted in yellow. Figure 5. MSA of MT protein sequences. Conservation score higher than 0.5 is highlighted in Unaligned cysteines are in orange. Cysteines aligned in all sequences are highlighted in red. Dots yellow. Unaligned cysteines are in orange. Cysteines aligned in all sequences are highlighted in denote gaps. Species acronyms, taxonomy and sequence IDs are indicated in Table 1. red.Dotsdenotegaps.Speciesacronyms,taxonomyandsequenceIDsareindicatedinTable1. Int.J.Mol.Sci.2017,18,812 10of26 Int. J. Mol. Sci. 2017, 18, 812 10 of 25 FigurFei g6u. rAem6.inoA maciindo seacqiudesnecqeu-benacsee-db anseeidghnbeoiguhrb-joouinr-ijnoign itnrege.t rSepe.eciSepse acciersonacyrmonsy amnsd asnedquseenqcuee nIDces are IDs are indicated in Table 1. The values at nodes indicate bootstrap support greater than 50%. indicated in Table 1. The values at nodes indicate bootstrap support greater than 50%. Tetrahymena Tetrahymenapyriformiswasusedasoutgrouptorootthetree.Branchlengthsaredrawntoascaleof pyriformis was used as outgroup to root the tree. Branch lengths are drawn to a scale of amino acid aminoacidsubstitutionspersite. substitutions per site. Forthefirsttime,MTgenesoftheslatepencilurchinoftheCidaroidaorder(E.tribuloides),twosea For the first time, MT genes of the slate pencil urchin of the Cidaroida order (E. tribuloides), two cucumbers(ApostichopusjaponicusandParasthicopusparvimensis),alamprey(Lethenteroncamtschaticum) sea cucumbers (Apostichopus japonicus and Parasthicopus parvimensis), a lamprey (Lethenteron andaneel(Anguillaanguilla)werecharacterized. GenestructuresaredepictedinFigure7andshow camtschaticum) and an eel (Anguilla anguilla) were characterized. Gene structures are depicted in clearlythehighvariabilityofintronlengths,thelongestbeingtheCidaroidaE.tribuloidesgeneand Figure 7 and show clearly the high variability of intron lengths, the longest being the Cidaroida E. theshortestthefishgenes. Alltheintronsthatinterruptthecodingregionsoftheselectedgenesare tribuloides gene and the shortest the fish genes. All the introns that interrupt the coding regions of the inphase-1andallthecodingregionswiththeexclusionoftheascidianone[28]areinterruptedby selecttewdo ginetnroesn sa.rIen tienr epshtiansgely-1, tahnedla malpl rtehye gceondeincogn rtaeignisonalss owtihteh itnhtreo enxicnluthseio3n UoTf Rthleik aeseccihdiinaond oenrme [28] 0 are inatnedrrauspcitdeida nbyo ntews.oA instarlorenasd. yInrteeproerstteind,gllayn, ctehleet l(aBmrapnrcehyio sgteonmea cfloornitdaaien)sg eanlseos ttrhuec tiunrterosnar ieni nthseo m3′ eUTR like wecahyinuondiqeurem,i nadneded aBsfMcidTi1alna ckosntehse. inAtsro nalirnetahdey3 0rUepTRoratnedd,t hleanBcfMelTet2 fi(Brsrtainncthroionsitsomloac atfelodriadtaoen) lygene struc4tucroedso anrsef rionm sothmeeA wTGay[ 3u0n].iqTuhue,s ,inaldsoeegde nBefMstrTu1ct luarcekssr etflheec tindtirvoenrg ienn ctheein 3c′ oUdTinRg asenqdu ethncee Bs.fMT2 first intron is located at only 4 codons from the ATG [30]. Thus, also gene structures reflect divergence in coding sequences.