Convergent evolution and structural adaptation to the deep ... · 1 Convergent evolution and...

1

Convergentevolutionandstructuraladaptationtothedeepocean1

intheproteinfoldingchaperoninCCTα2

3

AlexandraA.-T.Weber1,2,3*,AndrewF.Hugall1,TimothyD.O’Hara14

1Sciences,MuseumsVictoria,GPOBox666,MelbourneVIC3001,Australia5

2CentredeBretagne,REM/EEP,Ifremer,LaboratoireEnvironnementProfond,29280Plouzané,6 France7

3ZoologicalInstitute,UniversityofBasel,Vesalgasse1,4051Basel,Switzerland8

*AuthorforCorrespondence:AlexandraA.-T.Weber,Sciences,MuseumsVictoria,GPOBox666,9 MelbourneVIC3001,Australia,+61383417641,[email protected]

11

12

13

.CC-BY-NC-ND 4.0 International licenseunder acertified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available

The copyright holder for this preprint (which was notthis version posted July 9, 2020. ; https://doi.org/10.1101/775585doi: bioRxiv preprint

https://doi.org/10.1101/775585

http://creativecommons.org/licenses/by-nc-nd/4.0/

2

Abstract14

ThedeepoceanisthelargestbiomeonEarthandyetitisamongtheleaststudiedenvironments15 ofourplanet.Lifeatgreatdepthsrequiresseveralspecificadaptations,howevertheirmolecular16 mechanisms remain understudied. We examined patterns of positive selection in 416 genes17 from four brittle star (Ophiuroidea) families displaying replicated events of deep-sea18 colonization (288 individuals from216species).We foundconsistentsignaturesofmolecular19 convergence in functions related to protein biogenesis, including protein folding and20 translation. Five genes were recurrently positively selected, including CCTα (Chaperonin21 ContainingTCP-1subunitα),whichisessentialforproteinfolding.Molecularconvergencewas22 detected at the functional and gene levels but not at the amino-acid level. Pressure-adapted23 proteinsareexpected todisplayhigherstability tocounteract theeffectsofdenaturation.We24 thus examined in silico local protein stability of CCTα across the ophiuroid tree of life (96725 individualsfrom725species)inaphylogenetically-correctedcontextandfoundthatdeepsea-26 adaptedproteinsdisplayhigherstabilitywithinandnexttothesubstrate-bindingregion,which27 was confirmedby in silico global protein stability analyses. This suggests that CCTαnot only28 displaysstructuralbutalsofunctionaladaptationstodeepwaterconditions.TheCCTcomplexis29 involved in the folding of ~10% of newly synthesized proteins and has previously been30 categorizedas ‘cold-shock’protein innumerouseukaryotes.We thuspropose thatadaptation31 mechanisms to cold and deep-sea environments may be linked and highlight that efficient32 protein biogenesis, including protein folding and translation, are key metabolic deep-sea33 adaptations.34

Keywords35

Echinodermata;TCP-1;proteinfolding;positiveselection;proteinstability;pressureadaptation36

37



https://doi.org/10.1101/775585


3

Introduction38

Thedeepocean(>200m)coversapproximatelytwo-thirdsoftheglobalseafloorarea,yetitis39 among the least studied environments of our planet in terms of biodiversity, habitats and40 ecosystem functioning (Ramirez-Llodra et al. 2010). It harbors specific environmental41 conditionssuchashighhydrostaticpressure(onaverage~400atm),lowtemperatures(0-4°C),42 absence of light and scarcity of food. Life at great depths requires multiple metabolic43 adaptationsresultinginaphysiologicalbottleneck(Gross&Jaenicke1994)limitingthevertical44 distribution of species (Brown&Thatje 2014; Somero1992). It has also beenproposed that45 high pressure may drive speciation events (Brown & Thatje 2014; Gaither et al. 2016).46 Numerouscellularprocessessuchasenzymaticprocesses,protein folding,assemblyofmulti-47 subunit proteins and lipoprotein membranes are impacted by high pressure and low48 temperature (Carney 2005; Jaenicke 1991; Pradillon & Gaill 2007; Somero 1992). More49 specifically,cellularmechanismsofdeep-seaadaptationhavebeenrecentlyreviewed(Yancey50 2020)and includechanges inmembrane fluidity (highpressureand lowtemperaturerigidify51 membranes),extrinsicadaptations(i.e.changesinthecellularmilieusuchastheexpressionof52 molecular chaperones or increased concentration of piezolytes, i.e. stabilizing solutes), and53 intrinsic adaptations (i.e. changes in protein sequences). An essential intrinsic adaptation of54 deep-seaproteinsistheirincreasedstability(i.e.moreresistanttodenaturation)comparedto55 their shallow-water counterparts (Gross & Jaenicke 1994; Siebenaller 2010; Somero 1992,56 1990,2003).Thishasbeenshowninnumerousstudies,whichmainlyfocusedonahandfulof57 proteins(suchaslactatedehydrogenase)indeep-seafishes(Dahlhoff&Somero1991;Lemaire58 etal.2018;Morita2008,2003;Siebenaller&Somero1979,1978;Sukaetal.2019;Wakaietal.59 2014).Interestingly,patternsofproteinadaptationtotemperatureshowhigherflexibility(i.e.60 decreasedstability)withdecreasingtemperature(Fieldsetal.2015;Saarmanetal.2017). As61 pressureandtemperaturestronglyco-varyinthedeepsea-temperaturedecreasesaspressure62 increases - it therefore can be difficult to disentangle the respective combined or opposing63 effectsofthesefactorsonproteinstabilityevolution.64

Patternsofpositiveselectionhavebeeninvestigatedtouncovergenesunderlyingadaptationto65 specificenvironments,includingthedeep-sea,innon-modelspecies(Kober&Pogson2017;Lan66 etal.2018;Oliveretal.2010;Sunetal.2017;Zhangetal.2017;Weberetal.2017).Although67 valuable, thesestudies typically focusonasingleor fewshallow-deep transitions ina limited68 numberofspecies,andthuslackthecomparativepowertoseparateconfoundingeffects.With69 almost2100 species, brittle stars (Ophiuroidea) are a largeandancient classof echinoderms70 (Stöhr et al. 2017, 2012). These diverse marine invertebrates have colonized every marine71 habitat, highlighting their strong adaptive abilities. Furthermore, their phylogeny is well-72 resolved (O’Hara et al. 2017, 2014) and they represent a major component of the deep-sea73 faunaintermsofabundanceandspeciesdiversity(Christodoulouetal.2019a;Christodoulouet74 al.2019b),makingthemimportantmodelsformarinebiodiversityandbiogeography(O’Haraet75 al.2019;Woolleyetal.2016).Itisusuallyassumedthatdeep-seaorganismscolonizedthedeep-76 sea fromshallowwaters(Brown&Thatje2014);however,colonization fromdeeptoshallow77 waters has also been reported (Bribiesca-Contreras et al. 2017; Brown&Thatje 2014). Four78 large independent ophiuroid families (Amphiuridae, Ophiodermatidae, Ophiomyxidae and79 Ophiotrichidae)haveacommonancestorfromshallowwaterwithextantspeciesoccurringin80 deepandshallowseas(Bribiesca-Contrerasetal.2017).Duetotheserepeatedandindependent81 deep-seacolonizationevents,thesefourbrittlestarfamiliesprovideanidealframeworktotest82 forconvergentmolecularevolutiontothedeepsea.83



https://doi.org/10.1101/775585


4

MaterialsandMethods84

Phylogenomicdatagenerationandprocessing85

Thedatausedinthisstudyisanextension(436additionalsamples)ofapreviouslypublished86 exon-capturephylogenomicdatamatrixof1484exonsin416genesfor708individualophiuroid87 samples representative of thewhole class Ophiuroidea (O’Hara et al. 2019). The full dataset88 usedhereencompassed1144individualophiuroidsamplesaccountingfor826species.Details89 onspecimencollection,environmentalparametersandlistofspeciesareavailableinTableS1.90 Thesetof416single-copygeneswere firstdetermined inaphylogenomicanalysis toresolve91 thephylogenyofOphiuroidea(O’Haraetal.2014).Thesegeneswererigorouslyscreenedtobe92 one-to-oneorthologsin52ophiuroidtranscriptomes(representativeoftheentireclass),inthe93 seaurchinStronlgylocentrotuspurpuratus, thehemichordateSaccoglossus kowalevskii and the94 fish Danio rerio. The subsequent exon-capture system laboratory, bioinformatic and95 phylogenetic procedures are described in previous works (O’Hara et al. 2017; Hugall et al.96 2016) and summarized in dryad packages https://doi.org/10.5061/dryad.db339/10 and97 https://datadryad.org/stash/dataset/doi:10.5061/dryad.rb334. Briefly, base-calling used a98 minimumreadcoverageoffive.ExonboundarieswereinitiallybasedontheStrongylocentrotus99 purpuratus andDanio rerio genomes, and then revised using the exon-capture readmapping100 information. For all selection analyses, each codon immediately adjacent to exon boundaries101 wasignored.TheprimarydatahadIUPAC-codedheterozygoussites,whichwerethenrandomly102 resolved. However, these sites had little influence as both ambiguity-coded and randomly103 resolveddatasetsreturnedthesamepositiveselectiontestresults.Aglobalphylogenetictreeof104 all 1144 samples for 416 genes (273kb sites)was generated viaRAxML v.8.2.11 (Stamatakis105 2014)using a codonpositionpartitionmodel. First a fully-resolved all-compatible consensus106 topologywas generated from 200 RAxML fast bootstrap samples (the -f –d command), onto107 which branch lengths were then optimized using a codon position GTR-G model (the -f –e108 command). The tree was then rooted according to (O’Hara et al. 2017) defining the sister109 superordersOphintegridaandEuryophiurida.110

Fourbrittlestarfamilieswereinvestigatedthatincludedspeciesdisplayingindependentevents111 of deep-sea colonization from shallow-water (Bribiesca-Contreras et al. 2017): Amphiuridae112 (111individualsfrom95species,depthrange:-0.5mto-5193m;temperaturerange:-1.6°Cto113 28.8°C), Ophiodermatidae (60 individuals from 38 species, depth range: -0.5m to -1668m;114 temperature range: 2.6°C to 28.3°C), Ophiomyxidae (41 individuals from 29 species, depth115 range:-1.5mto-792m;temperaturerange:4.6°Cto28.7°C)andOphiotrichidae(78individuals116 from62species,depthrange:-1mto-405m;temperaturerange:10.5°Cto29.5°C)(Figure1A).117 Positive selection analyses were conducted separately per family. 1664 alignments were118 generated, representingeachgene(416) ineach family (4). Ineachalignment,amaximumof119 30%missing data per sequence was allowed. Alignments that lacked deep species (>200m)120 afterfilteringwerenotused.Asall thesefourfamiliesbelongtothesuperorderOphintegrida,121 sequences of Asteronyx loveni belonging to the sister superorder Euryophiurida122 (Asteronychidae)wereusedasoutgroups. After filtering,1649alignmentswereavailable for123 furtheranalyses.124

Phylogeneticreconstructionandpositiveselectionanalyses125

For each of the 1649 alignments, aMaximum-Likelihood phylogenywas reconstructed using126 RAxMLv.8.2.11withthefollowingparameters:-x12345-#100-fa-mGTRGAMMA-p12345127



https://doi.org/10.1101/775585


5

(Figure 1B). Deep (>200m) species (tips) and monophyletic groups of deep species (nodes)128 were labeled as “Foreground” branches for positive selection analyses (Amphiuridae: 46129 species; 11 independent events of deep-sea colonization; Ophiodermatidae: 7 species; 4130 independent events; Ophiomyxidae: 10 species; 4 independent events; Ophiotrichidae: 4131 species;4independentevents).Then,thepackageHyPhywasusedtoconductseveralpositive132 selectionanalyses(Figure1C):1)BUSTED(Branch-siteUnrestrictedStatisticalTestforEpisodic133 Diversification)(Murrelletal.2015)totestforgene-widepositiveselection(atleastonesiteon134 atleastonebranch);2)aBSREL(adaptiveBranch-SiteRandomEffectsLikelihood)(Smithetal.135 2015)todetectspecificbranchesevolvingunderepisodicpositiveselection;3)MEME(Mixed136 EffectsModelofEvolution)(Murrelletal.2012)to findsitesevolvingunderepisodicpositive137 selection. 4)MNMmethod; it has been shown recently thatmutations at adjacent sites often138 occur as a result of the same mutational event (i.e. multinucleotide mutations, MNMs) and139 thereforemay bias classical branch-site tests for positive selection (Venkat et al. 2018). The140 authors of that study developed a new model of positive selection detection incorporating141 MNMs,whichwealsousedondeeplineages(>200m).Foreachgene,p-valueswerecorrected142 formultipletestingusingtheHolmmethod(Holm1979).Thep-valuesignificancelevelusedfor143 allthepositiveselectiondetectionmethodswas0.05.Finally,weused:5)RELAX(Wertheimet144 al. 2014) to test for relaxation of selection, and exclude potential candidate genes displaying145 relaxationofselection.Foreachofthefourfamilies,positivelyselectedcandidategenesofeach146 method were overlapped on a Venn diagram (Figures 1C and S1). To be considered as a147 candidategeneforpositiveselectioninonefamilyandtominimizetheriskoffalsepositives,a148 genehadtodisplayasignificantsignalinatleastthreeoutoffourmethodsincludingMEMEand149 MNM(BUSTED,MEME,MNMorMEME,MNM,aBSREL)andnotdisplayrelaxationofselection150 (RELAX). This set of candidates was used for functional annotation. Final sets of positively151 selected genes per family were then compared among each other to test for convergent152 evolution. To confirm that positive selectionwas detected only in deep-sea lineages, positive153 selection was also tested in shallow-water lineages for the five genes displaying convergent154 positiveselectionsignatures(CCTα,PFD3, tkt, rpl34,rpl8;seeResults,Tables1,S3).Foreach155 family, the same number of shallow-water species as was used for deep-water species was156 randomlylabeledas‘Foreground’ineachgenetreeandpositiveselectiontestswereperformed157 forthefivemethodsasdescribedabove.158

GeneOntologyannotationsandamino-acidconvergenceanalyses159

To explore which functions may be involved in deep-sea adaptation, the representative160 sequence of each of the 416 genes was extracted from the sea urchin Strongylocentrotus161 purpuratus genomeandblastedagainst thenrdatabase fromNCBIusingBLAST+.WeusedS.162 purpuratusasreferencebecausesequenceannotationforthisspeciesisofhighquality(nohigh-163 quality brittle star reference genome is currently available) and to use a single complete164 representativesequenceforeachgene.Thetop50hitswereextractedandloadedinBLAST2GO165 v.4.1. for annotation (Conesa et al. 2005). Mapping, annotation and slim ontology (i.e. GO166 subsets of broader categories) were performed with BLAST2GO using default parameters,167 exceptfortheannotationcut-offparameter,whichwassetto45.GOcategoriesweredescribed168 usingthelevel3ofslimontology.169

CCTα,theonlycandidategenedisplayingpositiveselectionsignalinthreefamilies(seeResults)170 was further analyzed for signatures of convergent evolution. Specifically, amino-acid profiles171 were investigated forconvergentshiftsusingPCOC(Reyetal.2018).Thismethod,whichhas172



https://doi.org/10.1101/775585


6

beenshowntodisplayhighsensitivityandspecificity,detectsconvergentshifts inamino-acid173 preferences rather than convergent substitutions. The CCTα amino-acid alignment174 encompassing the four families and outgroups was used to generate a maximum-likelihood175 phylogenyaspreviouslydescribedbutthistimeusingthePROTGAMMAWAGproteinmodelof176 sequence evolution (Figure S2). For each family, positively selected branches resulting from177 aBSRELanalyseswere labeledas foregroundbranches (i.e. thebrancheswith the convergent178 phenotype in the nucleotide topology) in four different scenarios: i) Amphiuridae,179 Ophiodermatidae, Ophiomyxidae; ii) Amphiuridae, Ophiodermatidae; iii) Amphiuridae,180 Ophiomyxidae; iv) Ophiodermatidae, Ophiomyxidae. Detection of amino-acid convergence in181 thesefourscenarioswasthenperformedusingPCOCandadetectionthresholdof0.9(Reyetal.182 2018).183

Proteinstructuremodeling,proteinstabilityprofileandoverallproteinstability184

To infer thepositionofpositively selectedmutationsonCCTα, the correspondingamino-acid185 sequenceoftheindividualAmphiuraconstrictaMVF214041wasusedtoobtainthesecondary186 andtertiaryproteinstructuresofthisgene.Thisshallow-waterspecieswaschosenbecauseits187 CCTαsequencehadnomissingdata.ThesecondarystructurewasmodeledusingInterPro72.0188 webbrowser (https://www.ebi.ac.uk/interpro/). Theproteinmodelwas generatedusing the189 normalmodeoftheonlinePhyre2server(http://www.sbg.bio.ic.ac.uk/~phyre2/)(Kelleyetal.190 2015). The online server EzMol 1.22 was used for image visualization and production191 (http://www.sbg.bio.ic.ac.uk/ezmol/) (Reynolds et al. 2018). The online server NetSurfP-2.0192 (http://www.cbs.dtu.dk/services/NetSurfP/) was used to predict amino-acid surface193 accessibilityandsecondarystructures.194

WethenexaminedtheproteinstabilityprofilesofCCTαacrossthewholeophiuroidclass(967195 sequenceswithlessthan30%missingsites,representing725species)usingeScapev2.1(Gu&196 Hilser 2009, 2008). This algorithm calculates a per-site estimate of Gibbs free energy of197 stabilization based on a sliding window of 20 residues. More specifically, it models the198 contributionofeachresiduetothestabilityconstant,ametricthatrepresentstheequilibriumof199 thenatively foldedand themultipleunfolded statesof aprotein (D’Aquinoet al. 1996). Sites200 adapted to elevatedpressure (or high temperature at atmospheric pressure) are expected to201 displaystabilizingmutations(i.e.morenegativedeltaGvalues),whereassitesadaptedto low202 temperaturesatatmosphericpressureareexpected todisplaymutations increasing flexibility203 (i.e.decreasingstability, thusmorepositivedeltaGvalues) (Fieldsetal.2015;Saarmanetal.204 2017).Foreachsiteoftheapicaldomain(codons203-353),wecalculatedtheaveragedeltaG205 valueforall324shallow-waterspecies(424individuals)(0-200m)and401deep-waterspecies206 (543 individuals) (>200m) and their respective standard deviation. To test the difference207 between these average values in a phylogenetic context, we used phylogenetically-corrected208 ANOVA (R function phylANOVA of the phytools v.0.6-60 R package; 10,000 simulations). To209 correctforrelatednessamongspecies,weusedtheglobalRAxMLphylogenetictreeprunedto210 the 967 tips. To investigate regions rather than individual codons,we contrasted shallow vs.211 deepspeciesalongthewholegene,averagingdeltaGvaluesacross10residuesandperforming212 aphylogenetically-correctedANOVAaspreviouslydescribed.213

Finally,we tested how themutations at the positively selected sites, and the oneswith local214 stabilityvaluessignificantlydifferentbetweenshallowanddeepspeciesinthephylogenetically-215 corrected ANOVA, impacted the global protein stability. For each one of these sites, we216 identified‘shallow-water’and‘deep-water’mutations,andusedtheshallow-waterAmphiuridae217



https://doi.org/10.1101/775585


7

AmphiuraconstrictaMVF214041sequenceastemplatetoaddthese‘shallow-water’and‘deep-218 water’mutationsoneatthetime.Amodelstructureforeachoneofthesenewsequenceswas219 generated using the normal mode of the online Phyre2 server220 (http://www.sbg.bio.ic.ac.uk/~phyre2/)(Kelleyetal.2015).Theglobalstabilityofeachmodel221 structurewas then calculatedusing the ‘stability’ functionof theprogramFoldXwithdefault222 values(Schymkowitzetal.2005).Foreachsite,the DDGvaluebetween‘deep-water’(mutant)223 and‘shallow-water’(wild-type)mutationswascalculated,whereDDG>0.5kcal/molindicates224 stabilizingmutations,DDG<-0.5kcal/molindicatesdestabilizingmutation,and-0.5kcal/mol<225 DDG<0.5kcal/molindicatesneutralmutations(Khatunetal.2004;Khan&Vihinen2010).226

Results227

Five genes involved in protein biogenesis are recurrently positively selected in deep-sea brittle228 stars229

We used 416 single-copy orthologs from 216 species (288 individuals) of four brittle star230 families(Figure1A)toexaminepatternsofpositiveselectionindeep-seaspecies(>200m).For231 each gene of each family,we used four different positive selectionmethods and onemethod232 detectingrelaxationofselection(Figure1B-C).Given thatourdataset isnotrepresentativeof233 thewholegenome,wechoseastringentapproachtofocusonafewcandidategenesdisplaying234 strongsignalsofpositiveselection.Therefore,weonlykeptcandidategenesthathadsignificant235 signatureofpositiveselectioninatleastthreemethodsandwhichalsodidnotshowrelaxation236 of selection. We found 36 candidate genes in Amphiuridae, 9 in Ophiodermatidae, 6 in237 OphiomyxidaeandnoneinOphiotrichidae(TableS2;FigureS1).Fourgenesinvolvedinembryo238 development were positively selected in Amphiuridae. Common functions among candidate239 genes from different families include protein folding, translation (i.e. ribosomal proteins),240 catabolicprocessandcellproliferation(TableS2).241

Five genes were positively selected in at least two families, which are notably involved in242 protein foldingand in translation (Table1).Among these,CCTαwaspositively selected inall243 three families and significant in each one of the selection detection methods (Table 1). To244 confirm that positive selection was detected only in shallow-deep transition, we performed245 positive selection analyses on these five common candidate genes but this time labelling246 shallow-waterlineagesas‘Foreground’.Mostofthegenesdidnotdisplaysignaturesofpositive247 selection in shallow-water environments, except PFD3 in Amphiuridae and tkt in248 Ophiotrichidae,whichweresignificantintheBUSTED,aBSRELandMEMEmethods(TableS3).249 This suggests that these two genes are evolving faster not only in response to deep-sea250 conditionsbutalsopossiblyinresponsetoadditionalenvironmentalconditions.251



https://doi.org/10.1101/775585


8

252

253

254

255

256

257

258

Table1:commonpositivelyselectedcandidategenesinthreefamiliesandtheircharacteristics(3of4259 methods, not displaying relaxation of selection). *Positively selected in 4 of 4methods, not displaying260 relaxationofselection.italic:commonBiologicalProcessannotation.261

262

Genename

Description BlastReferencesequence

GOterms:BiologicalProcess Positivelyselectedin

CCTα*

chaperonincontaining TCP1complexsubunitα

XP_780270.1 proteinfolding Amphiuridae,Ophiodermatidae,Ophiomyxidae

PFD3 prefoldinsubunit3 XP_797937.1 macromolecularcomplexassembly;proteincomplexassembly;proteinfolding

Amphiuridae,Ophiodermatidae

tkt* transketolase 2isoformX2

NP_1229589.1 biologicalprocess Amphiuridae,Ophiodermatidae

rpl34 subunitribosomal XP_797232.1 ribosomebiogenesis;translation

Amphiuridae,Ophiomyxidae

rpl8 60SribosomalL8 XP_796001.1 Translation Amphiuridae,Ophiomyxidae



https://doi.org/10.1101/775585


9

263

Fig. 1: Workflow used in this study. A: Schematic representation of the phylogeny of Ophiuroidea264 (redrawn fromChristodoulou et al. (2019b)). Four families (288 individuals from216 species)with a265 shallow-water common ancestor and extant species in shallow (0-200m) and deep (>200m)266 environments are highlighted in different colors. The width of each triangle is proportional to the267 samplednumberofspeciesineachfamily.Ophiomyxidae(blue),Ophiodermatidae(green),Amphiuridae268 (pink)andOphiotrichidae(yellow).B:Foreachfamily(numberofspeciesusedinbrackets)andeachone269 ofthe416single-copyorthologs,MaximumLikelihood(ML)reconstructionswereperformed.Treesare270 drawn for illustration onlywith the branches of shallow-water (0-200m) species colored in black and271 deep-water (>200m) species in blue. C: For each resulting ML tree, deep species were labeled as272 foregroundbranchesandfourpositiveselectiondetectionmethodswereused(BUSTED,aBSREL,MEME,273 MNM). To detect and exclude candidate genes displaying relaxation of selection, i.e. accumulation of274 substitutions not due to increased selection pressure, the method RELAX was used. The final set of275 candidate genes for each family encompassed genes positively selected in at least 3methods and not276 displaying relaxationof selection.Convergent evolutionwas examinedbyoverlapping candidate genes277 per family. For the most interesting candidate gene, amino-acid convergence and tertiary structure278 analyseswereperformed,andproteinstabilityprofilesandglobalproteinstabilitywerecalculated.279



https://doi.org/10.1101/775585


10

280

Fig.2: Structureand functionof theCCTcomplex, selectionanalysesonCCTαandcomparisonof local281 stability values from CCTα apical domain between shallow and deep species. A: Model of tertiary282 structureoftheCCTαsubunit.Eachsubunitiscomposedofanapicaldomain(AD;green)containingthe283 substratebindingregions(PL:ProximalLoop;AH:ApicalHinge;H11:Helix11),anintermediatedomain284 (ID;blue)andanequatorialdomain (ED;pink) containing theATPbindingsitesandwherehydrolysis285 takes place. B: Model of the top view of the CCT complex, encompassing 8 paralogous subunits. C:286 QuaternarystructuremodeloftheCCTcomplexencompassingadoubleringof8paralogoussubunits.D:287 SimplifiedmodelofPrefoldin(PFD)-CCTinteractioninthefoldingofnewlysynthesizedactinortubulin.288 A-D: Adapted from Bueno-Carrasco & Cuellar, 2018, “Mechanism and Function of the Eukaryotic289 ChaperoninCCT”.E:LocalizationofthepositivelyselectedsitesonthetertiarystructureofCCTαinthe290 four ophiuroid families investigated. F: Average protein stability profiles and respective standard291



https://doi.org/10.1101/775585


11

deviations(verticalbars)foreachcodonoftheCCTαapicaldomainin324species(424individuals)from292 shallowwater(0-200m; inblack)and401species(543 individuals) fromdeepwater(>200m; inblue)293 representativeofthewholeophiuroidclass.Asmaller(i.e.morenegative)valueofdeltaGisindicativeof294 substitutionsincreasingstability.ThesubstratebindingregionsPL,AHandH11arehighlightedaswellas295 thepositivelyselectedsites.296

CCTα, the sole candidategenedetectedas significantlypositively selected in three families in297 the shallow-deep contrast, did not display signatures of positive selection in shallow-water298 environments (TableS3). It isa subunitof theoctamericChaperoninContainingTCP1(CCT)299 complex,acytosoliceukaryoticchaperoninhavingacentralroleinproteinfolding(Figure2A-300 D) (Bueno-Carrasco & Cuéllar 2019; Valpuesta et al. 2005). The three other CCT subunits301 present inourdataset,CCTd,CCTeandCCTh,didnotdisplayconsistentsignaturesofpositive302 selection;rather,whilepositiveselectionwasdetectedatthesitelevelinallsubunits,relaxation303 of selection was detected in all 3 subunits in Amphiuridae (Table S6), indicating that the304 accumulationofmutations inthesesubunits isnotclearlyduetonaturalselection.Therefore,305 CCTawastheonlyoneoutoffoursubunitstoshowconsistentsignaturesofpositiveselection306 acrossmethods.The remaining four subunitsof theCCTcomplexcouldnotbe testedas they307 werenotpresentinourdataset.308

Interestingly, PFD3, a subunit of the hexameric co-chaperone prefoldin interacting with CCT309 (Gestaut et al. 2019; Martín-Benito et al. 2002) was positively selected in two families310 (Amphiuridae: MNM, MEME, BUSTED; Ophiodermatidae: MNM, aBSREL, MEME, BUSTED)311 (Table 1; Tables S2, S6; Figure 2D), although itwas also positively selected in shallow-water312 Amphiuridae (Table S3). The two other prefoldin subunits present in our dataset (PFD1 and313 PFD5)didnotshowaconsistentsignatureofpositiveselection(TableS4),suggestingthatsub-314 units of this co-chaperone can evolve relatively independently from each other. Finally, two315 ribosomalproteins(Rpl8andRpl34)wererecurrentlypositivelyselectedinAmphiuridaeand316 Ophiomyxidae, while the ribosomal protein S6 was positively selected in Ophiodermatidae317 (Table1;TableS2).318

CCTαanddeep-seaadaptation:convergenceatthegenebutnotamino-acidlevel319

Whilepositiveselectionwasdetectedatthegenelevelinthreefamilies,positiveselectionatthe320 site level (MEME)was detected in all four families. The sites displaying positive selection in321 CCTαwerenotthesameamongthefourfamilies,exceptforsite26,whichwassharedbetween322 AmphiuridaeandOphiotrichidae(Figure2E;Table2;TableS5).Threesiteswerefoundinthe323 equatorialdomain,i.e.theATPbindingregion,twositeswerefoundintheintermediatedomain324 and three siteswere found in the apical domain, i.e. the substrate binding region (Figure 3)325 (Bueno-Carrasco&Cuéllar2019).Furthermore,wefoundthat7out8positivelyselectedsites326 increasedtheglobalproteinstability,suggestingthatthesesitesmaybeinvolvedinstructural327 adaptationofCCTa (Table2).This is corroboratedby the fact that6out8mutations involve328 changes towards hydrophobic amino-acids, which are known to increase protein stability329 (Kellis et al. 1988; Pace et al. 2011). Finally, convergent evolution at the site level was not330 detectedwhenexaminingamino-acidprofiles(PCOCposteriorprobabilitiesatallsites<0.9).331

332

333

334



https://doi.org/10.1101/775585


12

Table 2: Summary of the positively selected sites in each family and sites with local stability values335 significantlydifferentbetweenshallowanddeepspecies.Foreachsite, the ‘deep’and ‘shallow’amino-336 acids are reported, as well as their impact on global protein stability: DDG>0.5 kcal/mol: mutations337 increasingstability;DDG<-0.5kcal/mol:mutationsdecreasingstability;-0.5kcal/mol<DDG<0.5kcal/mol:338 neutral mutations in terms of stability. ED: equatorial domain; ID: intermediate domain; AD: apical339 domain;PS:positivelyselected.*:themostcommonamino-acidsbetweenshallowanddeepspeciesare340 reported.341

Site Tertiarystructure

Detectionmethod

Amino-acid‘deep’

Amino-acidproperty

Amino-acid

‘shallow’

Amino-acidproperty

DDG[kcal/mol]

26 ED PSAmphiuridae&PSOphiotrichidae A hydrophobic S Polar

uncharged 34.7

175 ID PSAmphiuridae Y hydrophobic S Polaruncharged 11.8

203* AD phylANOVA P specialcase Q Polaruncharged

-17.3

215* AD phylANOVA V hydrophobic P specialcase 2.1

219 AD PSOphiotrichidae I hydrophobic T Polaruncharged 3.7

263* AD phylANOVA H Positivelycharged Q Polar

uncharged 22.9

327* AD phylANOVA S,VPolar

uncharged,hydrophobic

A hydrophobic 11.9,30.1

328* AD phylANOVA T Polaruncharged S Polar

uncharged 27.6

329 AD PSAmphiuridae V hydrophobic M hydrophobic 15.7

350 AD PSOphiodermatidae I hydrophobic K Positively

charged 15.7

375 ID PSAmphiuridae V hydrophobic I hydrophobic 12.3462 ED PSOphiomyxidae C specialcase A hydrophobic -3.8

465 ED PSOphiodermatidae F hydrophobic K Positively

charged 22.9

342

Energeticlandscapesrevealstructuraladaptationwithinandnexttothesubstratebindingregion343

Next we calculated site-specific protein stability profiles of CCTα in 967 individuals of 725344 species representative of the whole Ophiuroidea class, to test the hypothesis that deep-sea345 adapted proteins are more stable than their shallow-water counterparts. For each site, we346 comparedtheaveragestabilitymeasureof324shallow-waterspecies(0-200mdepth)vs.401347 deep-water species (>200m depth), where lower DG values correspond to higher stability348 (Figures 2F, S3A).We focused on the apical domain as it encompasses the substrate binding349 region,whosepositionandstructurearehighlyconservedacrosseukaryotes(Joachimiaketal.350 2014).Thisregion iscomposedof theproximal loop(PL), theapicalhinge(AH)andHelix11351 (H11)(Figures2A,3).WhileAHandH11arealmostinvariantacrossallophiuroids,thestability352 measurewaslower(i.e.morestable)indeepcomparedtoshallowspecieswithinthePLandin353 two sites following the PL (codons 214-217), close to a positively selected site (codon 219)354 (Figures 2F, S3A). In contrast, three codons displayed significantly higher flexibility in deep355



https://doi.org/10.1101/775585


13

comparedtoshallowspecies(FigureS3A,B),suggestingthatincreasedflexibilitymayplayarole356 indeep-seaadaptationoutsidetheligandbindingregion,possiblyrelatedtolowtemperature.357 Nevertheless,whenaveragingDGvaluesacross10codons,onlythesignalclosetoPLremained358 significantinthephylogenetically-correctedANOVAcontrastingstabilityvaluesofshallowand359 deepspecies(FigureS4).This indicates thatsubstitutions towardsamorestablePLoccurred360 independentlyintheophiuroidtreeoflife.Finally,weexaminedhowthesemutationsimpacted361 theglobal(ratherthanlocal)stabilityvalueoftheprotein.Wefoundthat4outof5mutations362 increasedtheglobalproteinstability,includingthesitelocatedinthesubstratebindingregion363 (i.e.site215inthePL),confirmingtheresultsfoundwiththelocalstabilityvalues.364

365

Fig.3:PrimarystructureofCCTa.Ligandbindingsites (proximal loop,apicalhingeandhelix11),ATP366 bindingsites,positivelyselectedsites(methodMEME)andsitesdisplayingdifferentlocalstabilityvalues367 betweenshallowanddeepspecies(methodphylANOVA)arehighlighted.Thesequenceof theshallow-368 waterAmphiuridaeAmphiuraconstrictaisusedastemplate.369

Discussion370

Proteinbiogenesisisessentialfordeep-seaadaptation371

In this study,we examined patterns of positive selection in 416 genes from four brittle star372 families representing replicated instancesofdeep-sea colonization.We found that four genes373 involved in functions related to protein biogenesis including protein folding and protein374



https://doi.org/10.1101/775585


14

synthesis (i.e. ribosomal proteins)were convergently positively selected in three out of four375 families. Furthermore, an additional ribosomal protein was positively selected in376 Ophiodermatidae and two transcription factors were positively selected in Amphiuridae.377 Previousstudieshaveshownthatseveralgenesinvolvedingeneticinformationprocessingsuch378 asprotein folding, transcriptionor translation (including ribosomalproteins)werepositively379 selected in the hadal (>6000m depth) amphipod Hirondella gigas (Lan et al. 2017), in the380 bathyal(200-3000mdepth)fishAldrovandiaaffinis(Lanetal.2018)andinthebathyalbarnacle381 Glyptelasma gigas (Gan et al. 2020). Taken together, these and previous results indicate that382 adaptationsinproteinbiogenesisareessentialtodeep-sealife.383

At the individual family level, four genes related to embryo development were positively384 selected in Amphiuridae. Interestingly, nine genes related to the same function (embryo385 development;GO:0009790)weredetectedascandidategenesfordeep-seaadaptationbetween386 populationsofthedeep-seafishCoryphaenoidesrupestris livingatdifferentdepths(Gaitheret387 al.2018).Thissuggeststhattheremightbesimilarselectionpressuresondevelopmentatboth388 young (i.e. species level) and old (i.e. family level) evolutionary timescales for life at great389 depths.Finally,nocandidategenesfordeep-seaadaptationwerefoundinOphiotrichidaeacross390 differentpositiveselectionmethods.Thismightbeexplainedbythefactthatthisfamilyisthe391 youngest froma fossil andaphylogeneticpointof view (O’Haraet al. 2017). In that sense, it392 startedcolonizingthedeep-sealaterthantheotherfamilies,andthereforetherewaspossibly393 less time for natural selection to act. In supportwith this, themaximum sampling depth for394 Ophiotrichids was 405m, whereas it was deeper for the three other families (Amphiuridae:395 5193m;Ophiodermatidae:1668m;Ophiomyxidae:792m).396

Independentevolutionaryhistoriesofsubunitsfrommacromolecularcomplexes397

CCTaandprefoldin3(PFD3),thetwogenesrecurrentlypositivelyselectedinvolvedinprotein398 folding,aresubunitsofoctamericandhexamericcomplexes,respectively.TheCCTcomplex is399 estimated to fold ~10% of newly synthesized proteins, including actin and tubulin, and is400 involvedinnumerouscorecellularprocessessuchascytoskeletonformation,cellsignaling,cell401 recognition and protein degradation (Bueno-Carrasco & Cuéllar 2019), while the prefoldin402 complex is known to assist CCT in actin folding (Martín-Benito et al. 2002).We tested four403 subunits of the octameric CCT complex, yet CCTα was the only one to show consistent404 signatures of positive selection across different methods. Indeed, the other tested subunits405 displayedrelaxationofselectionornosignaturesofpositiveselection.Thismightbeduetothe406 differentdegreesofsubunitspecializationinCCT,asCCTαhasintermediatebindingproperties407 (i.e. neither high ATP affinity nor high substrate affinity) compared to the other subunits408 (Bueno-Carrasco&Cuéllar2019).Therefore, it ispossible thatCCTαcanevolvemorerapidly409 becauseitmayexperiencelessnegativeselectionthantheothersubunits.Inthesameway,we410 testedthreesubunitsofthehexamericprefoldincomplex,andPFD3wastheonlysubunittobe411 consistentlypositivelyselected.Whileweacknowledgethatweanalyzedanexistingdatasetand412 thusdidnottestallsubunitsofbothmacromolecularcomplexes(4/8inCCT;3/6inprefoldin),413 it appears surprising that individual subunits of macromolecular complexes can evolve414 relativelyindependentlyfromeachother.415

WenotethatapreviousstudyactuallyhighlightedevolutionaryindependenceofCCTsubunits416 acrossoldevolutionarytimescales(i.e.amongeukaryotes),showingthatCCTα,CCTγandCCTζ417 evolvedunderpositiveselectionafterduplicationeventswhich ledtosub-functionalization in418



https://doi.org/10.1101/775585


15

eukaryotes,mostlikelyinresponsetofoldingincreasinglycomplexcytosolicproteins(Fares&419 Wolfe2003). Furthermore, ithasbeenshownthat theexactsamemutation in theconserved420 ATP-bindingregionofalleightCCTsubunitshavedramaticallydifferentphenotypicoutcomes421 inyeast,suchasdifferentialtemperaturesensitivity,differentialgrowthinpresenceofactinand422 tubulinpolymerizationinhibitors,ordifferentialoverandunder-expressionofgenes(Amitetal.423 2010).Thisnotonlyindicatesthatthefunctionalactivityofeachsubunitisnotredundantbut424 alsopossiblyhighlights ‘moonlighting’proteins(i.e.otherfunctionsofindividualsubunitsthat425 arecarriedoutnotaspartofthefullcomplex)(Jeffery2009;Huberts&vanderKlei2010).In426 thatsense,thefunctionalcoherenceofseveralhetero-oligeromericproteincomplexesincluding427 CCT has been reviewed (Matalon et al. 2014). Interestingly, they found a lack of coherence428 amongsubunitsatvirtuallyallfunctionallevels,includingdifferentlevelsofsubunitabundance,429 differencesintranscriptionalregulation,differentialsubunitduplicationorloss,andfinallylack430 of coherence ingenetic interaction.Explanations for this lackof coherence includeenhancing431 complex assembly or subunit multi-functionality (Matalon et al. 2014). There is increasing432 evidenceforthelattercase,asmanyCCTmonomerfunctionshavebeenreportedandrecently433 reviewed (Vallin & Grantham 2019). Taken together, our results are consistent with the434 hypothesisthatCCTaandPFD3maybemoonlightingproteinsinvolvedindeep-seaadaptation.435

Convergentevolution,functionalandstructuraladaptationsinCCTa436

Our results suggest that convergent evolution was detected at the functional (i.e. protein437 folding)andatthegenelevels(i.e.CCTa,PFD3)indeep-seaadaptation.Yet,itwasnotdetected438 attheamino-acidlevelinCCTa,asthemajorityofpositivelyselectedsitesweredifferentamong439 families,exceptforonesitesharedbetweenAmphiuridaeandOphiotrichidae.Thismightbedue440 totheolddivergencetimeamongtheinvestigatedbrittlestarfamiliesandthebroadnatureof441 the selection regime (i.e. depth and temperature). It has been shown that rates ofmolecular442 convergencedecreasewith time (Storz2016) and the last commonancestorofAmphiuridae,443 Ophiodermatidae and Ophiomyxidae is estimated to be approximately 250million years old444 (O’Hara et al. 2017). Furthermore, convergence at the amino-acid level is often less common445 thanconvergenceathigherlevelsofbiologicalhierarchy(e.g.gene,pathwayorspecieslevels)446 (Bolnicketal.2018;Tenaillonetal.2016,2012).447

Our results indicate that CCTa evolvedunder positive selectionduring deep-sea colonization448 andspeciationevents.ThesemutationspossiblyrevealfunctionaladaptationsofCCTa,suchas449 adaptations in ATP or substrate binding. Indeed, it has previously been reported that the450 shallow groove created by the conservedHelix 11 and the flexible Proximal Loop allows the451 bindingofavarietyofsubstrates(Joachimiaketal.2014;Yametal.2008),suggestingthatthe452 detectedsubstitutionsinthePLandinadjacentaminoacidsallowefficientsubstratebindingin453 deep-seaspecies.Similarly, ina studyonmetabolicenzymes from37ctenophores,numerous454 sites associated with adaptation to depth, temperature or both were located close to the455 substratebindingregion(Winnikoffetal.2019).AmajorsubstrateoftheCCTcomplexandits456 co-chaperoneprefoldin is actin, an essential cytoskeletonprotein (Martín-Benito et al. 2002).457 While it should be noted that CCTa is not directly involved in actin folding (actin binds the458 subunitsCCTd,CCTbandCCTe)(Llorcaetal.1999,2000),examinationofcytoskeletalproteins459 is of interest due to their essential role in cells. Unfortunately, as actin and tubulinwerenot460 includedinourdataset,noconclusionsaboutcytoskeletalproteinsofdeep-seabrittlestarscan461 yetbedrawn.However,actinofdeep-sea fisheshasbeenextensivelystudied, showing that it462 displays structural adaptations to high-pressure enhancing stability (Morita 2003;Koyama&463



https://doi.org/10.1101/775585


16

Aizawa2008;Wakaietal.2014;Yancey2020).Furthermore,genomeanalysisofthehadalfish464 Pseudoliparis swirei revealed that a gene family related to cytoskeleton was expanded, and465 severalgenesfromthreegeneontologycategorieslinkedtocytoskeletonwereevolvingunder466 positiveselection(Wangetal.2019).Finally,13genes involvedincytoskeletonsystemswere467 positivelyselected inthedeep-seafishA.affinis (Lanetal.2018),suggestingthatcytoskeletal468 proteinsarepivotalfordeep-seaadaptation.469

Furthermore,wetestedthehypothesisthatCCTaisstructurallyadaptedtodeep-seaconditions,470 asacommonfeatureofhigh-pressureadaptedproteinsistodisplayincreasedstability(Somero471 2003;Ohmaeetal.2013;Yancey&Siebenaller2015).Weinvestigatedlocalandglobalstability472 in CCTa and found that four out of five sites with phylogenetically-corrected significantly473 different local stability showed increased global protein stability, suggesting that these474 mutations reflect structural deep-sea adaptation. Moreover, seven out of eight mutations at475 positivelyselectedsitesincreasedglobalproteinstability,andinvolvedhydrophobicamino-acid476 substitutions,whichareknowntoincreaseproteinstability(Kellisetal.1988;Paceetal.2011).477 It has also been shown that several proteins from the abyssal fish Coryphaenoides armatus478 displayed increased protein stability as an adaptation to hydrostatic pressure (Morita 2003,479 2008;Brindleyetal.2008;Lemaireetal.2018).Finally,severalcriticalmutations inboththe480 enzyme synthesizing trimethylamine oxide (TMAO: a heavily studied stabilizing cosolute or481 piezolyte(Yancey2020))andinHSP90,havebeenproposedtoenhanceoverallproteinstability482 inthehadalfishPseudoliparisswirei(Wangetal.2019).483

Roleofmolecularchaperonesindeep-seaadaptation484

Wehaveshownthatoverdeepevolutionarytimescales,CCTαdisplaysrecurrentsignaturesof485 acceleratedevolutionandstructuraladaptationintransitionfromshallowtodeep-seahabitats486 inbrittlestars.Thiscontrastswithprevious findingsonseaurchinswhere,CCTε-butnot the487 other subunits- showed signatures of positive section, yet not in the two deep-sea species488 included in the work (Kober & Pogson 2017). A functional equivalent of the eukaryotic489 chaperoninCCTinbacteriaisthechaperonincomplexGroEL/GroES(Gupta1990;Mayhewetal.490 1996).Thischaperoninhasbeenshown tobeconstitutivelyhighlyexpressed in thedeep-sea491 bacteria Shewanella, and that a major target of this chaperonin was a subunit of the 50S492 ribosome (Satoet al.2015).Furthermore, convergentevolution in chaperoneshaspreviously493 beenshown(Draceni&Pechmann2019),andthereisevidencethatothergroupsofmolecular494 chaperones such asHSP70 andHSP90 (Kim et al. 2013) are involved in deep-sea adaptation495 (Yancey2020).Forinstance,theHSP70genefamilywasexpandedinadeep-seamussel(Sunet496 al.2017)andconvergentamino-acidchangeshavebeenreportedinfouroutoffivecopiesofthe497 HSP90proteininahadalfish(Wangetal.2019).Inourdataset,twoHSP90geneswerepresent498 (HSP90b1 and HSP90ab1), yet neither showed signatures of accelerated evolution.499 Nevertheless,onlytheexaminationofwholegenomeswillgiveacompleteviewofHSPfamily500 evolutioninbrittlestars,whichshouldbeexploredinfurtherstudies.501

Similaritiesbetweendeep-seaandcoldadaptationmechanisms502

TherearenumerousstudiesontheroleofCCTatshorterevolutionarytimescales,whichreveal503 itsroleincold-stressresponse.Notably,CCThasbeencharacterizedasa‘cold-shock’proteinin504 several eukaryotes due to the overexpression of the investigated subunits when organisms505 wereexposedtocoldstress(Heetal.2017;Kayukawaetal.2005;Someretal.2002;Yinetal.506 2011).Furthermore,CCThasbeenshowntodisplayspecificstructural(Pucciarellietal.2006)507



https://doi.org/10.1101/775585


17

andfunctional(Cuellaretal.2014)adaptationstocoldenvironmentinAntarcticfish,inaddition508 tobeingalsooverexpressedinAntarcticfishexposedtoheatstress(Buckley&Somero2009).509 Moreover, there is evidence for a linkbetween cold-stress responseandhigh-pressure stress510 responseinbacteria(Welchetal.1993;Wemekamp-Kamphuisetal.2002),althoughithasalso511 beenshownthatheat-shockproteinsarealsoexpressedunderhigh-pressurestressinbacteria512 (Aertsen et al. 2004) or constitutively in the deep-sea bacteriaShewanella (Sato et al. 2015).513 Additionally,transcriptomeanalysesrevealedthatcold-inducibleproteinfamiliesareexpanded514 inahadalamphipod(Lanetal.2017),andseveralproteins involved incoldshockhavebeen515 showntoevolveunderpositiveselectionindeep-seaamphipodandfish(Lanetal.2018).Taken516 together,ourfindingssupportthehypothesisthatcoldshockproteinsplayanimportantrolein517 deep-sea adaptation (Brown & Thatje 2014), which was also proposed by others (Lan et al.518 2018).519

Limitationsandoutlook520

Weacknowledgethatourstudylacksfunctionalvalidationtodemonstratethatthechangesare521 truly adaptive (which would be experimentally demanding as CCT folds ~10% of newly522 synthesized proteins and we analyzed the sequences of more than 700 species), yet we523 minimized false inferences by applying stringent positive selection detection criteria.524 Furthermore,weusedaproxyoffunctionalvalidationbyinvestigatinginsilicolocalandglobal525 proteinstabilityinadatasetwithgreatcomparativepower,bothintermsofphylogeneticand526 environmental diversity. Finally, direct experimental testing on deep-sea organisms remains527 technicallychallenging,sowemadeuseofthepowerofmoleculardatatoindirectlyrevealnew528 insightsindeep-seaadaptation.Furtherstudiesshouldincludewholegenomestoobtainamore529 completeviewofdeep-seaadaptationhotspotsandpossiblemechanisms,suchasrecentstudies530 ondeep-seafishes(Gaitheretal.2018;Wangetal.2019),hydrothermalventmussel(Sunetal.531 2017)orcoldseepgastropod(Liuetal.2020).Also,whilewefocusedonintrinsicadaptations532 (i.e. at the genomic level), mechanisms of extrinsic adaptations (i.e. changes in the cellular533 milieu)through,forinstance,osmolyteconcentration(piezolytessuchasTMAO)ormembrane534 phospholipid composition changes, should not be overlooked (Yancey & Siebenaller 2015;535 Somero2003;Yancey2020).However, this isbeyondthescopeof thisstudy.With increasing536 interestsindeep-seabiodiversity,ecosystemsandresourcesinthelastdecades(Danovaroetal.537 2017,2014;Gloveretal.2018),theseareexcitingtimesfordivingdeeperintomechanismsof538 deep-seaadaptation.539

References540

AertsenAetal.2004.Heatshockprotein-mediatedresistancetohighhydrostaticpressureinEscherichia541 coli.ApplEnv.Microbiol.70:2660–2666.542

Amit M et al. 2010. Equivalent mutations in the eight subunits of the chaperonin CCT produce543 dramaticallydifferentcellularandgeneexpressionphenotypes.J.Mol.Biol.401:532–543.544

BolnickDI,BarrettRD,OkeKB,RennisonDJ,StuartYE.2018.(Non)parallelevolution.Annu.Rev.Ecol.545 Evol.Syst.49:303–330.546

Bribiesca-ContrerasG,VerbruggenH,HugallAF,O’HaraTD.2017.Theimportanceofoffshoreorigination547 revealedthroughophiuroidphylogenomics.In:Proc.R.Soc.B.Vol.284TheRoyalSocietyp.20170160.548



https://doi.org/10.1101/775585


18

Brindley AA et al. 2008. Enzyme sequence and its relationship to hyperbaric stability of artificial and549 naturalfishlactatedehydrogenases.PLoSOne.3.550

BrownA,ThatjeS.2014.Explainingbathymetricdiversitypatternsinmarinebenthicinvertebratesand551 demersalfishes:physiologicalcontributionstoadaptationoflifeatdepth.Biol.Rev.89:406–426.552

Buckley BA, Somero GN. 2009. cDNA microarray analysis reveals the capacity of the cold-adapted553 Antarctic fish Trematomus bernacchii to alter gene expression in response to heat stress. Polar Biol.554 32:403–415.555

Bueno-CarrascoMT,CuéllarJ.2019.MechanismandFunctionoftheEukaryoticChaperoninCCT.ELSJohn556 WileySonsLtdEd.1–9.doi:10.1002/9780470015902.a0028208.557

CarneyRS.2005.Zonationofdeepbiotaoncontinentalmargins. In:OceanographyandMarineBiology.558 CRCPresspp.221–288.559

Christodoulou M, O’Hara Timothy, et al. 2019a. Unexpected high abyssal ophiuroid diversity in560 polymetallic nodule fields of the Northeast Pacific Ocean, and implications for conservation.561 BiogeosciencesDiscuss.1–54.562

Christodoulou M, O’Hara Timothy D., Hugall AF, Arbizu PM. 2019b. Dark Ophiuroid Biodiversity in a563 ProspectiveAbyssalMineField.Curr.Biol.29:3909–3912.564

ConesaAetal.2005.Blast2GO:auniversal tool forannotation,visualizationandanalysis in functional565 genomicsresearch.Bioinformatics.21:3674–3676.doi:10.1093/bioinformatics/bti610.566

CuellarJetal.2014.Assistedproteinfoldingatlowtemperature:evolutionaryadaptationoftheAntarctic567 fishchaperoninCCTanditsclientproteins.Biol.Open.3:261–270.568

DahlhoffE,SomeroGN.1991.Pressureandtemperatureadaptationofcytosolicmalatedehydrogenases569 ofshallowanddeep-livingmarine invertebrates:evidenceforhighbodytemperatures inhydrothermal570 ventanimals.J.Exp.Biol.159:473–487.571

DanovaroR,CorinaldesiC,Dell’AnnoA,SnelgrovePV.2017.Thedeep-seaunderglobalchange.Curr.Biol.572 27:R461–R465.573

DanovaroR, SnelgrovePV,TylerP. 2014. Challenging theparadigmsof deep-sea ecology.TrendsEcol.574 Evol.29:465–475.575

D’Aquino JA et al. 1996. The magnitude of the backbone conformational entropy change in protein576 folding.ProteinsStruct.Funct.Bioinforma.25:143–156.577

DraceniY,PechmannS.2019.Pervasiveconvergentevolutionandextremephenotypesdefinechaperone578 requirementsofproteinhomeostasis.Proc.Natl.Acad.Sci.116:20009–20014.579

Fares MA, Wolfe KH. 2003. Positive selection and subfunctionalization of duplicated CCT chaperonin580 subunits.Mol.Biol.Evol.20:1588–1597.581

Fields PA, Dong Y, Meng X, Somero GN. 2015. Adaptations of protein structure and function to582 temperature:thereismorethanonewayto‘skinacat’.J.Exp.Biol.218:1801–1811.583



https://doi.org/10.1101/775585


19

Gaither MR et al. 2016. Depth as a driver of evolution in the deep sea: Insights from grenadiers584 (Gadiformes: Macrouridae) of the genus Coryphaenoides. Mol. Phylogenet. Evol. 104:73–82. doi:585 10.1016/j.ympev.2016.07.027.586

GaitherMRetal.2018.Genomicsofhabitatchoiceandadaptiveevolution inadeep-sea fish.Nat.Ecol.587 Evol.2:680.588

Gan Z et al. 2020. Comparative transcriptomic analysis of deep-and shallow-water barnacle species589 (Cirripedia, Poecilasmatidae) provides insights into deep-sea adaptation of sessile crustaceans. BMC590 Genomics.21:1–13.591

Gestaut D et al. 2019. The Chaperonin TRiC/CCT Associates with Prefoldin through a Conserved592 ElectrostaticInterfaceEssentialforCellularProteostasis.Cell.177:751–765.593

GloverAG,WiklundH,ChenC,DahlgrenTG.2018.PointofView:Managingasustainabledeep-sea‘blue594 economy’requiresknowledgeofwhatactuallylivesthere.eLife.7:e41319.595

Gross M, Jaenicke R. 1994. Proteins under pressure: the influence of high hydrostatic pressure on596 structure,functionandassemblyofproteinsandproteincomplexes.Eur.J.Biochem.221:617–630.597

GuJ,HilserVJ.2008.Predictingtheenergeticsofconformationalfluctuationsinproteinsfromsequence:a598 strategyforprofilingtheproteome.Structure.16:1627–1637.599

GuJ,HilserVJ.2009.Sequence-basedanalysisofproteinenergylandscapesrevealsnonuniformthermal600 adaptationwithintheproteome.Mol.Biol.Evol.26:2217–2227.601

Gupta RS. 1990. Sequence and structural homology between a mouse T-complex protein TCP-1 and602 the’chaperonin’family of bacterial (GroEL, 60-65 kDa heat shock antigen) and eukaryotic proteins.603 Biochem.Int.20:833–841.604

HeY,WangL,ZhuW,DongZ,LiuN.2017.EffectsofsalinityoncoldtoleranceofMalaysianredtilapia.605 Aquac.Int.25:777–792.606

HolmS.1979.Asimplesequentiallyrejectivemultipletestprocedure.Scand.J.Stat.65–70.607

HubertsDHEW,vanderKleiIJ.2010.Moonlightingproteins:Anintriguingmodeofmultitasking.Biochim.608 Biophys.ActaBBA-Mol.CellRes.1803:520–525.doi:10.1016/j.bbamcr.2010.01.022.609

HugallAF,O’HaraTD,HunjanS,NilsenR,MoussalliA.2015.Anexon-capturesystemfortheentireclass610 Ophiuroidea.Mol.Biol.Evol.33:281–294.611

JaenickeR.1991.Proteinstabilityandmolecularadaptationtoextremeconditions.In:EJBReviews1991.612 Springerpp.291–304.613

JefferyCJ.2009.Moonlightingproteins—anupdate.Mol.Biosyst.5:345–350.614

Joachimiak LA,Walzthoeni T, Liu CW,AebersoldR, Frydman J. 2014. The structural basis of substrate615 recognitionbytheeukaryoticchaperoninTRiC/CCT.Cell.159:1042–1055.616

KayukawaTetal.2005.ExpressionofmRNAforthet;chcomplexpolypeptide–1,asubunitofchaperonin617 CCT,isupregulatedinassociationwithincreasedcoldhardinessinDeliaantiqua.CellStressChaperones.618 10:204.619



https://doi.org/10.1101/775585


20

Kelley LA, Mezulis S, Yates CM, Wass MN, Sternberg MJ. 2015. The Phyre2 web portal for protein620 modeling,predictionandanalysis.Nat.Protoc.10:845.621

KellisJT,NybergK,S˘ailD,FershtAR.1988.Contributionofhydrophobicinteractionstoproteinstability.622 Nature.333:784–786.doi:10.1038/333784a0.623

KhanS,VihinenM.2010.Performanceofproteinstabilitypredictors.Hum.Mutat.31:675–684.624

KhatunJ,KhareSD,DokholyanNV.2004.Cancontactpotentialsreliablypredictstabilityofproteins? J.625 Mol.Biol.336:1223–1238.626

Kim YE, Hipp MS, Bracher A, Hayer-Hartl M, Ulrich Hartl F. 2013. Molecular chaperone functions in627 proteinfoldingandproteostasis.Annu.Rev.Biochem.82:323–355.628

KoberKM,PogsonGH.2017.Genome-widesignalsofpositiveselectioninstrongylocentrotidseaurchins.629 BMCGenomics.18:555.630

Koyama S, AizawaM. 2008. Piezotolerance of the cytoskeletal structure in cultureddeep-sea fish cells631 usingDNAtransfectionandproteinintroductiontechniques.Cytotechnology.56:19–26.632

LanYetal.2018.Denovotranscriptomeassemblyandpositiveselectionanalysisofanindividualdeep-633 seafish.BMCGenomics.19:394.634

LanYetal.2017.Molecularadaptationintheworld’sdeepest-livinganimal:Insightsfromtranscriptome635 sequencingofthehadalamphipodHirondelleagigas.Mol.Ecol.26:3732–3743.636

Lemaire B et al. 2018. Molecular adaptation to high pressure in cytochrome P450 1A and aryl637 hydrocarbonreceptorsystemsofthedeep-seafishCoryphaenoidesarmatus.Biochim.Biophys.ActaBBA-638 ProteinsProteomics.1866:155–165.639

LiuRetal.2020.DenovogenomeassemblyoflimpetBathyacmaealactea(Gastropoda:Pectinodontidae),640 the first reference genome of a deep-sea gastropod endemic to cold seeps. Genome Biol. Evol. doi:641 10.1093/gbe/evaa100.642

LlorcaOetal.2000.EukaryoticchaperoninCCTstabilizesactinandtubulinfoldingintermediatesinopen643 quasi-nativeconformations.EMBOJ.19:5971–5979.644

LlorcaO et al. 1999. Eukaryotic type II chaperonin CCT interactswith actin through specific subunits.645 Nature.402:693–696.646

Martín-BenitoJetal.2002.Structureofeukaryoticprefoldinandofitscomplexeswithunfoldedactinand647 thecytosolicchaperoninCCT.EMBOJ.21:6377–6386.648

MatalonO,HorovitzA,LevyED.2014.Different subunitsbelonging to the sameprotein complexoften649 exhibitdiscordantexpressionlevelsandevolutionaryproperties.Curr.Opin.Struct.Biol.26:113–120.650

MayhewM et al. 1996. Protein folding in the central cavity of the GroEL–GroES chaperonin complex.651 Nature.379:420–426.652

MoritaT.2008.Comparativesequenceanalysisofmyosinheavychainproteinsfromcongenericshallow-653 anddeep-livingrattailfish(genusCoryphaenoides).J.Exp.Biol.211:1362–1367.654



https://doi.org/10.1101/775585


21

MoritaT.2003.Structure-basedanalysisofhighpressureadaptationofα-actin.J.Biol.Chem.278:28060–655 28066.656

Murrell B et al. 2012. Detecting individual sites subject to episodic diversifying selection. PLoS Genet.657 8:e1002764.658

MurrellBetal.2015.Gene-wideidentificationofepisodicselection.Mol.Biol.Evol.32:1365–1371.659

O’Hara TD, Hugall AF, Thuy B, Moussalli A. 2014. Phylogenomic resolution of the class Ophiuroidea660 unlocksaglobalmicrofossilrecord.Curr.Biol.24:1874–1879.661

O’HaraTD,HugallAF,ThuyB,StöhrS,MartynovAV.2017.Restructuringhighertaxonomyusingbroad-662 scalephylogenomics:ThelivingOphiuroidea.Mol.Phylogenet.Evol.107:415–430.663

O’Hara TD, Hugall AF, Woolley SN, Bribiesca-Contreras G, Bax NJ. 2019. Contrasting processes drive664 ophiuroidphylodiversityacrossshallowanddeepseafloors.Nature.565:636.665

OhmaeE,MiyashitaY,KatoC.2013.Thermodynamicandfunctionalcharacteristicsofdeep-seaenzymes666 revealedbypressureeffects.Extremophiles.17:701–709.doi:10.1007/s00792-013-0556-2.667

OliverTAetal.2010.Whole-genomepositiveselectionandhabitat-drivenevolution inashallowanda668 deep-seaurchin.GenomeBiol.Evol.2:800–814.669

PaceCNetal.2011.ContributionofHydrophobicInteractionstoProteinStability.J.Mol.Biol.408:514–670 528.doi:10.1016/j.jmb.2011.02.053.671

PradillonF,GaillF.2007.Adaptationtodeep-seahydrothermalvents:somemolecularanddevelopmental672 aspects.J.Mar.Sci.Technol.37–53.673

Pucciarelli S, Parker SK, Detrich HW, Melki R. 2006. Characterization of the cytoplasmic chaperonin674 containingTCP-1fromtheAntarcticfishNototheniacoriiceps.Extremophiles.10:537–549.675

Ramirez-Llodra EZ et al. 2010. Deep, diverse and definitely different: unique attributes of theworld’s676 largestecosystem.Biogeosciences.7:2851–2899.doi:10.5194/bg-7-2851-2010.677

Rey C, Guéguen L, SémonM, Boussau B. 2018. Accurate detection of convergent amino-acid evolution678 withPCOC.Mol.Biol.Evol.35:2296–2306.679

ReynoldsCR, IslamSA,SternbergMJ.2018.EzMol:Awebserverwizardfortherapidvisualizationand680 imageproductionofproteinandnucleicacidstructures.J.Mol.Biol.430:2244–2248.681

SaarmanNP,KoberKM,SimisonWB,PogsonGH.2017.Sequence-basedanalysisof thermaladaptation682 andproteinenergylandscapesinaninvasivebluemussel(Mytilusgalloprovincialis).GenomeBiol.Evol.683 9:2739–2751.684

SatoH,NakasoneK, YoshidaT,Kato C,MaruyamaT. 2015. Increases of heat shockproteins and their685 mRNAs at high hydrostatic pressure in a deep-sea piezophilic bacterium, Shewanella violacea.686 Extremophiles.19:751–762.687

Schymkowitz J et al. 2005. The FoldXweb server: an online force field. Nucleic Acids Res. 33:W382–688 W388.689



https://doi.org/10.1101/775585


22

Siebenaller J, SomeroGN. 1978. Pressure-adaptivedifferences in lactatedehydrogenases of congeneric690 fisheslivingatdifferentdepths.Science.201:255–257.691

SiebenallerJF.2010.Effectsofthedeep-seaenvironmentoninvertebrates.Comp.HighPress.Biol.319–692 341.693

SiebenallerJF,SomeroGN.1979.Pressure-adaptivedifferencesinthebindingandcatalyticpropertiesof694 muscle-type (M 4) lactate dehydrogenases of shallow-and deep-livingmarine fishes. J. Comp. Physiol.695 129:295–300.696

SmithMDetal.2015.Lessismore:anadaptivebranch-siterandomeffectsmodelforefficientdetectionof697 episodicdiversifyingselection.Mol.Biol.Evol.32:1342–1353.698

SomerL,ShmulmanO,DrorT,HashmueliS,KashiY.2002.TheeukaryotechaperoninCCTisacoldshock699 proteininSaccharomycescerevisiae.CellStressChaperones.7:47.700

SomeroGN.1992.Adaptationstohighhydrostaticpressure.Annu.Rev.Physiol.54:557–577.701

SomeroGN. 1990. Life at low volume change: hydrostatic pressure as a selective factor in the aquatic702 environment.Am.Zool.30:123–135.703

SomeroGN. 2003. Protein adaptations to temperature and pressure: complementary roles of adaptive704 changes in amino acid sequence and internal milieu. Comp. Biochem. Physiol. B Biochem. Mol. Biol.705 136:577–591.doi:10.1016/S1096-4959(03)00215-X.706

Stöhr S,O’HaraTD,ThuyB. 2012.Globaldiversityof brittle stars (Echinodermata:Ophiuroidea). PLoS707 ONE.7:e31940.doi:10.1371/journal.pone.0031940.708

Stöhr S, O’Hara TD, Thuy B. 2017. World Ophiuroidea Database. Accessed at709 http://www.marinespecies.org/ophiuroideaon2017-07-27.710

Storz JF. 2016. Causes ofmolecular convergence and parallelism in protein evolution.Nat. Rev. Genet.711 17:239–250.712

Suka A et al. 2019. Stability of cytochromes c′ from psychrophilic and piezophilic Shewanella species:713 implications for complex multiple adaptation to low temperature and high hydrostatic pressure.714 Extremophiles.23:239–248.715

SunJetal.2017.Adaptationtodeep-seachemosyntheticenvironmentsasrevealedbymusselgenomes.716 Nat.Ecol.Evol.1:0121.717

TenaillonOetal.2016.Tempoandmodeofgenomeevolutionina50,000-generationexperiment.Nature.718 536:165.719

TenaillonOetal.2012.Themoleculardiversityofadaptiveconvergence.Science.335:457–461.720

VallinJ,GranthamJ.2019.TheroleofthemolecularchaperoneCCTinproteinfoldingandmediationof721 cytoskeleton-associatedprocesses:implicationsforcancercellbiology.CellStressChaperones.24:17–27.722

Valpuesta JM,CarrascosaJL,WillisonKR.2005.Structureandfunctionof thecytosolicchaperoninCCT.723 ProteinFold.Handb.725–755.724



https://doi.org/10.1101/775585


23

Venkat A, HahnMW, Thornton JW. 2018.Multinucleotidemutations cause false inferences of lineage-725 specificpositiveselection.Nat.Ecol.Evol.2:1280.726

WakaiN,TakemuraK,MoritaT,KitaoA. 2014.Mechanismofdeep-sea fishα-actinpressure tolerance727 investigatedbymoleculardynamicssimulations.PloSOne.9:e85852.728

WangKetal.2019.MorphologyandgenomeofasnailfishfromtheMarianaTrenchprovideinsightsinto729 deep-seaadaptation.Nat.Ecol.Evol.3:823–833.730

WeberAA-Tetal.2017.Positiveselectiononspermionchannelsinabroodingbrittlestar:consequence731 oflife-historytraitsevolution.Mol.Ecol.26:3744–3759.732

Welch TJ, Farewell A, Neidhardt FC, Bartlett DH. 1993. Stress response ofEscherichia coli to elevated733 hydrostaticpressure.J.Bacteriol.175:7170–7177.734

Wemekamp-Kamphuis HH, Karatzas AK, Wouters JA, Abee T. 2002. Enhanced levels of cold shock735 proteins in Listeria monocytogenes LO28 upon exposure to low temperature and high hydrostatic736 pressure.ApplEnv.Microbiol.68:456–463.737

Wertheim JO, Murrell B, SmithMD, Kosakovsky Pond SL, Scheffler K. 2014. RELAX: detecting relaxed738 selectioninaphylogeneticframework.Mol.Biol.Evol.32:820–832.739

WinnikoffJR,FrancisWR,ThuesenEV,HaddockSHD.2019.Combingtranscriptomesforsecretsofdeep-740 seasurvival:Environmentaldiversitydrivespatternsofproteinevolution.Integr.Comp.Biol.741

WoolleySNetal.2016.Deep-seadiversitypatternsareshapedbyenergyavailability.Nature.533:393–742 410.743

YamAYetal.2008.DefiningtheTRiC/CCTinteractomelinkschaperoninfunctiontostabilizationofnewly744 madeproteinswithcomplextopologies.Nat.Struct.Mol.Biol.15:1255.745

YanceyPH. 2020. Cellular responses inmarine animals to hydrostatic pressure. J. Exp. Zool. Part Ecol.746 Integr.Physiol.earlyonlineversion.doi:10.1002/jez.2354.747

Yancey PH, Siebenaller JF. 2015. Co-evolution of proteins and solutions: protein adaptation versus748 cytoprotectivemicromoleculesandtheirrolesinmarineorganisms.J.Exp.Biol.218:1880–1896.749

Yin Q et al. 2011. Molecular cloning of Litopenaeus vannamei TCP-1-eta gene and analysis on its750 relationshipwithcoldtolerance.YiChuanHered.33:168–174.751

Zhang Yanjie et al. 2017. Adaptation and evolution of deep-sea scale worms (Annelida: Polynoidae):752 insightsfromtranscriptomecomparisonwithashallow-waterspecies.Sci.Rep.7:46205.753

754

Acknowledgements755

We are grateful to W. Salzburger for providing access to the HPC sciCORE cluster and to J.756 SarrazinandS.Arnaud-Haondforcommentsonapreviousversionofthismanuscript.Wethank757 L.Bribiesca-ContrerasforhelpwithRanalyses.Wethankthreeanonymousreviewersandthe758 editorial team for valuable comments on a previous version of thismanuscript. Calculations759 were performed at sciCORE (http://scicore.unibas.ch/) scientific computing center at the760 University of Basel, Switzerland and at DATARMOR (http://www.ifremer.fr/pcdm) scientific761



https://doi.org/10.1101/775585


24

computingcenteratthePôledeCalculetdeDonnéesMarines(PCDM),Ifremer,Brest,France.762 ThisworkwassupportedbyanEndeavourPostdoctoralFellowshipawardedbytheAustralian763 Department of Education and Training [grant number 6534_2018 to AATW] and a Marie764 Sklodowska-CurieGlobalFellowshipawardedbytheEuropeanUnion’sHorizon2020research765 andinnovationprogramme[grantnumber797326toAATW].766

Datadeposition767

Phylogenomicdata(includingrawreads)andscriptsfordatasetgenerationareavailablein768 NCBIBioprojectPRJNA311384anddryadpackages:https://doi.org/10.5061/dryad.db339/10769 andhttps://datadryad.org/stash/dataset/doi:10.5061/dryad.rb334.770

Conflictofinterest771

None772

Supplementaryfigurecaptions773

Fig. S1:Overlapof positively selected candidate genes among threebrittle star families.A:Numberof774 candidategenesperfamilypositivelyselectedinatleastthreepositiveselectiondetectionmethodsand775 notdisplayingrelaxationofselection.B:Numberofcandidategenesperfamilypositivelyselectedinall776 fourpositiveselectiondetectionmethodsandnotdisplayingrelaxationofselection.Inbothconditionsof777 AandB,nogenewaspositivelyselectedinthefamilyOphiotrichidae.778

Fig.S2:Maximum-likelihoodreconstructionofCCTαusingamino-acidsequences.Thefourfocalfamilies779 (Amphiuridae, Ophiotrichae, Ophiomyxidae and Ophiodermatidae) and the outgroup (Asteronychidae)780 arelabelled.Positivelyselectedlineages(aBSRELmethod)ofeachfamilyarehighlightedinred.781

Fig.S3:ComparisonoflocalstabilityvaluesfromCCTαapicaldomainbetweenshallowanddeepspecies.782 A:Logtransformedp-valuesofthephylogenetically-correctedANOVAperformedbetweenaveragedelta783 G values of shallow vs. deep species at each codon of the CCTα apical domain. The substrate binding784 regionsPL,AHandH11arehighlightedinlightorange.Thepositivelyselectedsitesarehighlightedwitha785 red star. P-value level corresponding to 0.05 is highlighted in red. The most significant codons in a786 “significance peak” (204, 214, 265 and 324) are highlighted. B: Beanplots of delta G values between787 shallow (0-200m) and deep (>200m) species for each one of the most significant codons in the788 phylogenetically-correctedANOVA.Horizontalbarrepresentstheaveragevalueofthedataset.789

Fig.S4:Comparisonoflocalstabilityvaluesover10-codonwindowsonthecompleteCCTαgenebetween790 shallowanddeepspecies.A:Averageproteinstabilityprofilesover10codonwindowsforthecomplete791 CCTα gene in 324 species (424 individuals) from shallow water (0-200m) and 401 species (543792 individuals)fromdeepwater(>200m)representativeofthewholeophiuroidclass.Asmaller(i.e.more793 negative)valueofdeltaG is indicativeofsubstitutions increasingstability.Thepositivelyselectedsites794 are highlightedwith a red star. B: Log transformed p-values of the phylogenetically-corrected ANOVA795 performed between average delta G values over 10 codon windows of shallow vs. deep species. The796 positivelyselectedsitesarehighlightedwitharedstar.P-valuelevelcorrespondingto0.05ishighlighted797 inred.C:BeanplotsofdeltaGvaluesbetweenshallow(0-200m)anddeep(>200m)speciesforbothofthe798 mostsignificant10codonwindowsinthephylogenetically-correctedANOVA.Horizontalbarrepresents799 theaveragevalueofthedataset.800

Supplementarytables(inseparateexcelfile)801

Table S1: List of species used in this study, GPS coordinates and environmental parameters at their802 samplinglocations.Emptycellsindicatemissingdata.803



https://doi.org/10.1101/775585


25

Table S2: Positively selected candidate genes per family and their Gene Ontology annotation (P:804 Biological Process; F:Molecular Function; C: Cellular Component). Genes positively selected in several805 familiesarehighlightedinbold. 806

Table S3: Results of positive selection tests in shallow-water lineages for the five candidate genes for807 deep-seaadaptation.Significancelevel:0.05.Significantresultsareinbold.NS:notsignificant 808

TableS4:Resultsofpositiveselectiontestsforthethreeprefoldinsubunitsforeachfamily.Significance809 level:0.05.Significantresultsareinbold.NS:notsignificant 810

Table S5: Sites displaying episodic positive selection in CCTα.Methodused:MEME. Significance level:811 0.05.Significantsitesareinbold. 812

TableS6:ResultsofpositiveselectiontestsforthefourCCTsubunitsforeachfamily.Significantresults813 areinbold.Significancelevel:0.05.NS:notsignificant814



https://doi.org/10.1101/775585


Convergent evolution and structural adaptation to the deep ... · 1 Convergent evolution and...

Documents

Transcript of Convergent evolution and structural adaptation to the deep ... · 1 Convergent evolution and...