文档下载 文档下载导航
设为首页 | 加入收藏
搜索 请输入内容:  
 导航当前位置: 文档下载 > 所有分类 > 自然科学 > 生物学 > Putative cis-regulatory drivers in colorectal cancer
侵权投诉

Putative cis-regulatory drivers in colorectal cancer

非编码DNA在癌症生物学中也占一席之地

Putative cis-regulatory drivers in colorectal cancer

doi:10.1038/nature13602

Putativecis-regulatorydriversincolorectalcancer

HalitOngen1,2,3,ClausL.Andersen4,JesperB.Bramsen4,BodilOster4,MadsH.Rasmussen4,PedroG.Ferreira1,2,3,JuanSandoval5,EnriqueVidal5,NicolaWhiffin6,AlexandraPlanchon1,2,3,IsmaelPadioleau1,2,3,DeborahBielser1,2,3,LucianaRomano1,2,3,IanTomlinson7,RichardS.Houlston6,ManelEsteller5,8,9,TorbenF.Orntoft4&EmmanouilT.Dermitzakis1,2,3

Thecis-regulatoryeffectsresponsibleforcancerdevelopmenthavenotbeenasextensivelystudiedastheperturbationsoftheproteincodinggenomeintumorigenesis1,2.Tobettercharacterizecolorectalcancer(CRC)developmentweconductedanRNA-sequencingexperi-mentof103matchedtumourandnormalcolonmucosasamplesfromDanishCRCpatients,90ofwhichweregermline-genotyped.Byinvestigatingallele-specificexpression(ASE)weshowthatthegermlinegenotypesremainimportantdeterminantsofallelicgeneexpressionintumours.UsingthechangesinASEinmatchedpairsofsampleswediscover71geneswithexcessofsomaticcis-regulatoryeffectsinCRC,suggestingacancerdriverrole.Wecorrelategeno-typesandgeneexpressiontoidentifyexpressionquantitativetraitloci(eQTLs)andfind1,693and948eQTLsinnormalsamplesandtumours,respectively.Weestimatethat36%ofthetumoureQTLsareexclusivetoCRCandshowthatthisspecificityispartiallydrivenbyincreasedexpressionofspecifictranscriptionfactorsandchangesinmethylationpatterns.Weshowthattumour-specificeQTLsaremoreenrichedforlowCRCgenome-wideassociationstudy(GWAS)PvaluesthansharedeQTLs,whichsuggeststhatsomeoftheGWASvariantsaretumourspecificregulatoryvariants.Importantly,tumour-specificeQTLgenesalsoaccumulatemoresomaticmutationswhencomparedtothesharedeQTLgenes,raisingthepossibilitythattheyconstitutegermline-derivedcancerregulatorydrivers.Collectivelytheintegrationofgenomeandthetranscriptomerevealsasubstan-tialnumberofputativesomaticandgermlinecis-regulatorycancerchangesthatmayhavearoleintumorigenesis.

Thenon-codinggenomehassofarbeenoverlookedinthesearchfordriversincancer,exceptforsomeisolatedexamples1,2.Thegenomecontainsregulatorynon-codinggermlinevariationaffectinggeneexpres-sion,namelyeQTLs,whicharemajorcomponentsincomplexdiseasepredisposition3,andthesehavenotbeenexaminedwithinthecontextoftumorigenesis.EpistasisbetweeneQTLsandcodingmutationsingeneshasaroleindisease4;thereforeitislikelythatinteractionsbetweensomaticregulatoryvariantsoreQTLsandcodingvariationareimport-antintumorigenesis.ToexaminethisweperformedanRNAsequen-cinganalysisof103matchedtumourandnormalcolonmucosaCRCsamples(SupplementaryFig.2)oftheSYSCOLconsortium(Supplemen-taryTable1).Ninetysamplesweregenotypedfortheirgermlinegenomeandimputedto1000Genomesphase1release5(SupplementaryMethods).WealsoRNA-sequenced20referencetissuestocreateamorecompre-hensivecontroltranscriptomeforCRC.AgeneraloverviewandmainfindingsaresummarizedinSupplementaryFig.3.

Weexaminedtranscriptomeperturbationsduringtumorigenesisbyidentifyingdifferentiallyexpressedgenes(DEGs)andfind1,676DEGs(falsediscoveryrate(FDR)55%,foldchange$2)(SupplementaryTable2andSupplementaryFig.4).ThefunctionaltermsenrichedinDEGsaregiveninSupplementaryTable3,andpathwaysimpacted,whichincludeknownCRCpathways,aregiveninSupplementaryTable4.There

1

isnosignificantdifferenceofdifferentialexpressionbetweenknownCRCdrivergenes6,7orTheCancerGenomeAtlas(TCGA)pan-cancergenes7,8,andexpressionmatchedrandomgenes(SupplementaryFig.5).Inaddition,wefind213differentiallysplicedgenes(FDR55%)(Sup-plementaryTable5,SupplementaryFigs6and7).AmongtheseareTGFBI(ref.9),RASA1(ref.10),SLC26A3(ref.11)andSLC39A14(ref.12),whichhavebeenpreviouslyimplicatedinCRC.Theenrichedfunc-tionaltermsaregiveninSupplementaryTable6.

Wetestedtherelationshipofsamples’geneexpressionandsearchedforregionsofthegenomeinwhichCRCmodifiesthecorrelationofpro-ximalgeneexpression.Weobservethatthetumoursamples,normalsamples,andthe20referencetissuesamplesformthreedistinctclus-ters(SupplementaryFig.8),andthevarianceoftheCRCtranscriptomeisincreasedwhencomparedtothenormalsamples(SupplementaryFig.9).InallchromosomesthereareblocksofproximalgeneswhosecorrelationsarealteredinCRC(SupplementaryFigs10and11,andSup-plementaryTable7)pointingtocoordinatedregulationlikelyduetoconsistentepigeneticeffectsandlocuscontrolregions.TheresultsabovesuggestarewiringoftheregulatorynetworksinCRC.

Tofurtherdissectcis-regulatoryeffectsinCRCweexaminedvariationinregulationofgeneexpression.Onepowerfulmethodfordiscoveringcis-regulatoryvariabilityisASEanalysis,whichwasconductedasdescribedpreviously13.Wefindthatapproximately10%ofcodingheterozygoussitesexhibitsignificantASE(FDR51%,P,0.001).TheproportionofsiteswithASEissignificantlyhigherintumourscomparedtonormalsamples(Fig.1a)andapproximately34%ofASEisCRC-specific.Thisexcessislikelytobeanindicationofcopynumberalterations(CNAs)thathavebeendescribedpreviously6,14,15.Furthermore,inapproximately10%ofASEsitessignificantinbothmatchedsamples,thedirectionoftheASEeffectisreversed(Fig.1b),alsoindicatinggenesfoundinpotentialCNAregionsorgeneswherethecis-regulatorylandscapehasbeenaltered.

Weproposethatthesomaticeventsintheregulatoryregionsofgenesarereflectedinthedifferenceofallelicratiosbetweenmatchedpairsatexpressedheterozygoussites.WedefineASEsomaticevents(Supplemen-taryMethods)andobserveasignificantcorrelation(Spearman’srho50.05,P58.831027)betweenASEsomaticeventratesandcodingsomaticmutationrates16ofgenes,indicatingthatsomaticASEeventsmaybemorelikelytobeselectedingenesimplicatedinCRC(Fig.2aandSupplementaryFig.12).Todetermineascoreofsomaticdysregu-lationwecomparedtheASEsomaticeventrateofeachgenetoallothergenes(SupplementaryMethods).Thescoreisdefinedastheenrich-mentoflowPvalues(p1)17.Weobserveabimodaldistributionofp1valueswheremostgenesaresimilartoothergenesbutasmallfractionofgenesdemonstratehighp1,representinggeneswithdysregulation(SupplementaryFig.13andSupplementaryTable8).KnownCRCdrivergenes6,7andTCGApan-cancergenes7,8havesignificantlyhigherp1comparedtoothergenes(SupplementaryFig.14).Therefore,thep1

DepartmentofGeneticMedicineandDevelopment,UniversityofGenevaMedicalSchool,1211Geneva,Switzerland.2InstituteforGeneticsandGenomicsinGeneva(iGE3),UniversityofGeneva,1211Geneva,Switzerland.3SwissInstituteofBioinformatics,1211Geneva,Switzerland.4DepartmentofMolecularMedicine,AarhusUniversityHospital,8000Aarhus,Denmark.5CancerEpigeneticsandBiologyProgram(PEBC),BellvitgeBiomedicalResearchInstitute(IDIBELL),08908Barcelona,Catalonia,Spain.6DivisionofGeneticsandEpidemiology,TheInstituteofCancerResearch,Sutton,SurreySM25NG,UK.7NuffieldDepartmentofClinicalMedicineandOxfordNIHRComprehensiveBiomedicalResearchCentre,WellcomeTrustCentreforHumanGenetics,RooseveltDrive,OxfordOX37BN,UK.8

´CatalanadeRecercaiEstudisAvançats(ICREA),08010Barcelona,DepartmentofPhysiologicalSciencesII,SchoolofMedicine,UniversityofBarcelona,08007Barcelona,Barcelona,Spain.9Institucio

Spain.

00MONTH2014|VOL000|NATURE|1

©2014Macmillan Publishers Limited. All rights reserved

第1页

你可能喜欢

热门文档

相关文档

站点地图 | 文档上传 | 侵权投诉 | 手机版
本站所有资源均来自互联网,本站只负责收集和整理,均不承担任何法律责任,如有侵权等其它行为请联系我们.
文档下载 Copyright 2013 doc.guandang.net All Rights Reserved.  -  email