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 > Genome sequencing of normal cells reveals developmental lineages and mutational processes
侵权投诉

Genome sequencing of normal cells reveals developmental lineages and mutational processes

RESEARCHLETTER

METHODS

Organoidgeneration.TwohomozygousC57Bl6malemice(mouse1andmouse2)withLgr5-kiwereeuthanizedat116and98weeksofage,respectively.Thesemicewerechosenforourexperimentsbecausetheyhaveanormalphenotype,arenotknowntobehypermutableandarelikelytobehighlyinbredbeingderivedfromacolonyestablishedin2006/2007.Clonalorganoidlineswereestablishedfromsingleglands(stomach)13,17,27,singlecrypts(smallintestine16,17andcolon15,17)orsinglecells(prostate;W.K.andH.C.,manuscriptinpreparation)andexpandedinvitrofor3–6weeks.Singlestomachglands13,17,27,smallintestine16,17andcoloniccrypts15,17areknowntobeclonalandarederivedfromasinglestemcell.Prostateorganoidlineswereestablishedfromsingleprostatecellsasfollows:theprostatewasdividedintothreelobepairsandeachlobewasplacedin5mgml21collagenasetypeII(Gibco)inadvancedDMEM/F12mediumanddigestedfor1–2hat37uC.Glandularstruc-tureswerewashedwithADMEM/F12andcentrifugedat100g.Structuresweresubsequentlydigestedin5mlofTriplE(Gibco)withtheadditionof10mMY-27632(Sigma)for15minat37uC.Trypsin-treatedcellswerewashedandsinglecellswereseededingrowth-factor-reducedMatrigel(BDBiosciences)underprostateorga-noidconditions.

Whole-genomesequencing.DNAwasextractedfrom25organoidlinesandthetailofeachmouse,andwhole-genomesequencingwasperformedasdescribedpreviously26.Readswerealignedtothereferencemousegenome(NCBI37)byusingBurrows–WheelerAligner(BWA)ondefaultsettings.ReadsthatwereunmappedorPCR-derivedduplicateswereexcludedfromtheanalysis.

Somaticvariantcallingandvalidation.Tocallsomaticsingle-nucleotidesubsti-tutionsthatwereuniquetoeachorganoid,theCaVEMan(cancervariantsthroughexpectationmaximization)algorithm26wasusedbycomparingeachorganoidwiththetailofthesamemouse.Wesoughtheterozygousvariants,toavoidcontami-natingthedatasetwithsubclonalmutationslikelytohaveariseninvitro.Inaddi-tiontopreviouslyreportedpost-processingfiltering26,weappliedthefollowingcriteriaforvariantcalling:depthofatleast5Xinthevariantlocus;variantallelefractionofatleast25%;presenceofthevariantinatleastoneforwardandonereverseread;variantnotlocatedinasingle-nucleotidepolymorphismlocus(asperdbSNP;doc.guandang.net);variantfellwithinahomozygousregion.Putativesomaticvariantsweresubjectedtomanualinspection;artefactswereexcludedfromthedataset.Tovalidatethefinalcatalogueofsubstitutionsuniquetoeachorganoid,randomlyselectedsubstitutions(n5743variants;11%)wereexperimentallyvalidated.DNAoftheorganoidsandtailtissuefromsamemousewasamplifiedbyPCR(primersequencesavailableonrequest)targetingthevariantlocation.PCRampliconsweresequencedonanIlluminaMiSeq(readlength150basepairs)toamediandepthof15,000X.Theprecisionofthefinalcatalogueofsubstitutionswasthusdeterminedtobeatleast92%.

Measurementofsensitivity.Tomeasurethesensitivityofthesequencingandanalysispipelinetodetectheterozygousmutations,weextracted1,000randomlyselectedheterozygousSNPsfromtailsequencesofeachmouse.Asameasureofsensitivity,wedeterminedhowmanyofthese1,000SNPswewereabletocallineachorganoid,usingthesamestringentpost-processingfilteringcriteriaasusedininitialvariantcalling,namelyminimumdepthoffivereads,anallelefractionofmorethan25%,andvariantpresentsinatleastonereadineachdirection.TheheterozygousSNPsweredeterminedbyinterrogating,1.5millionpotentialSNPpositions(asindbSNP)ineachtailandselectingthoseSNPsinwhichthetwomajoralleleswerepresentwithanallelefractionof45–55%.Aminimumdepththresholdof10Xwaschosen(tailsequencecoveragebeing,35X)toavoidbiasingforSNPpositionswitharelativelyhighdepth.OfthiscatalogueofheterozygousSNPs,1,000SNPswererandomlyselected.

Embryonicvariantcallingandvalidation.Anembryonicmutationwasdefinedasamutationthatwaspresentinatleasttwoorganoidsoroneorganoidandthetail.Inaddition,themutationhadtobeabsentfromatleastoneorganoidorthetail.Toderiveunbiasedcataloguesofputativeembryonicvariants,weusedCaVEManinacomparisonofeachorganoidagainstthetailoftheothermouse.Fromtheresultinglistofvariantsweremovedvariantsthatwerepresentinallorganoidsofthesamemouse,variantsinlociofsingle-nucleotidepolymorphisms,andvariantsthatdidnotfallintoahomozygousregion.Weappliednumericalrulestoexcludelikelygermlinevariantsthatwerenotidentifiedassuchasaresultofinsufficientcoverage.Artefactswerethenremovedbymanualinspection.Everypotentialvariantthusdetectedwassubjectedtovalidationbycapillarysequencingineveryorganoidandcontroltissue.Afterlineagetreeconstructionfromunbiasedcalls,furthersup-portingvariantsweresought.Thefinalcatalogueofembryonicmutationscom-prised35mutations.Everymutationwascapillarysequencedineveryorganoid,yielding437outof475informativereads.Onthebasisofthese,theprecisionofthecatalogueofembryonicvariantswas100%.

Lineagetreeconstruction.Wefirstconstructedhypotheticaltreesforeachmousetodeterminetheminimumnumberofcelldivisionsrequiredtoresolvethetreesfully.Ontheassumptionthattwobifurcatinglineagetreeshadledtoallthecellsin

thetwomice,weconstructedtwohypotheticaltreestoincludetheminimumnumberofbifurcationsrequiredtoplaceall25organoidlinessinglyattheendofthebranchesofeachtree:13organoidsinmouse1and12inmouse2.Togetherthetwotreescomprised23celldivisions(ExtendedDataFig.4).

Wethenconstructedlineagetreesmanually,andcomputationallyusingmaxi-mumparsimony(asimplementedinthePHYLIPDNAPARSsoftware,withsettingsrandomizeinputorderandoutgrouproot)29,30fromvariantsthatwereconfirmedtobegenuine,non-germline.Inperformingthecomputationalanalyses,therootofeachlineagewasrepresentedbywild-typesequences.Auniquemost-parsimonioussolutionwasfoundforeachmouseintowhichallembryonicvariantsfittedwithnohomoplasy.

Thesephylogeneticlineagetreesmaycorrespondtothetreeofcelldivisionsduringembryogenesis.However,itispossiblethatsomecelldivisionsweremissed,eitherbecausemultiplemutationsatalineagebranchpointhadactuallyarisenatmorethanonecelldivisionand/orbecausecertaincelldivisionshadlackedmuta-tionsand/orbecausesomedaughtercellshaddiedduringdevelopmentandhadnotcontributedtotheadultmouse.Toevaluatewhetherthelineagetreesmightcor-respondtocelldivisions,twotestswereappliedusingthereadcountsofmutationsinthetailsample(seealsoExtendedDataFig.2).

First,whenmultiplemutationsdefinedanembryonicprecursor,wetestedwhethertherewasasignificantdifferenceinthetailallelefrequenciesofthemutations(Fisher’sexacttest).Significantdifferencesbetweenmutationsdefiningthesameprecursorwouldhaveindicatedthatthemutationsdidnotoccurduringthesamecelldivi-sionbutsuccessively,andanothercellgenerationwouldhavetobeadded.Withineachprecursorcell,wedidnotfindmutationswithconsistentsignificantdifferencesintailallelefrequency.Inthesecircumstancesweclassifiedsuchbranchesassingle-celldivisions.However,becausesequencecoverageandhencereadcountsinthetailarelimited,wecannotexcludethepossibilitythatrealdifferencesinreadcountswerenotdetectedbecauseoflackofstatisticalpower.

Second,itispossiblethat,betweengenerations,‘silent’celldivisions(inwhichnomutationtookplace)occurredthatwedidnotcapturebecauseofthelackofdetectablemutations.Theexistenceof‘silent’celldivisionscanbetestedbycompar-ingthetailallelefrequenciesofaprecursorcellwiththesumoftheallelefrequen-ciesofthetwoderivativecells,whichshouldbesimilar.Accordingtothistest,itispossibletoderivethetreesthatweproposewithouthavingtoinvoke‘silent’celldivisions.Again,however,alackofstatisticalpowerthroughlimitedcoveragemayhavelimitedourabilitytodetectthesedifferences.Moreover,ifacelldivisionresultedinonedaughterwithnomutationsandtheotherdaughterdiedanddidnotcontri-butetotheadult,wewouldbeunabletodetecttheexistenceofthiscelldivision.Copynumber.Wederivedcopynumberdatabyaveragingwhole-genomecove-ragedatawithindefinedsegments.Organoidswerecomparedwithtailsofthesamemousetoidentifychangesinsomaticcopynumber.

Mutationalsignatures.Mutationalsignaturesweredeterminedusingnon-negativematrixfactorization,asdescribedpreviously2,24.

Analysisofallelefrequencies.WeappliedtheDirichletprocess,asdescribedpreviously26,toassessthereadcountfrequencyofmutationsineachorganoid.Analysisofassociationofmutationswithchromatinorganization.Tostudywhetherthesomaticmutationsuniquetoeachorganoidwereassociatedwithdistinctchromatinstateswetestedwhethermutationswereenrichedinregionsofrepressedversusactivechromatin.Tothisend,fileswiththelocationofhistonemodificationpeaksfromChIP-seqdataweredownloadedfromUCSCforfivedifferentmodifi-cations(H3K14me1,H3K4me3,H3K9me3,H3K27me3andH3K36me3),profiledacrosstwomousecelllines(CH12andG1E)andtwoprimarysources(erythro-blastsandmegakaryocytes).Toobtainaconservativesegmentationofthegenomeintodifferentchromatinregions,asegmentwasannotatedasmarkedbyahistonemodificationonlywhensupportedbyallfourcelltypes.Wethencalculatedthedensityofmutationsineachregion.Toavoidthepossibleconfoundingeffectofsequencecomposition,asecondanalysiswasperformedcalculatingthemutationdensityasthenumberofGRNorCRNmutationsperG/Cnucleotideineverychro-matinregion,yieldinganalogousresults.Non-homozygousareasoftheorganoidgenomes,whichwereexcludedfrommutationcalling,werenotconsideredinthisanalysis.Statisticalsignificancewasassessedforeverypairwisecomparison(n510)usinganExactPoissontestfortheratiobetweentworateparameters(r51inthenullhypothesis)andadjustedformultipletestingusingBenjamini–Hochberg’sFalseDiscoveryRate.

Analysisofassociationofmutationswithreplicationtiming.Toevaluatetheassociationofthemutationdensitywithreplicationtiming,weused16Repli-chipdatasetsfromENCODE/FSU:wavelet-smoothedvaluesofmeanearly/lateS-phaseratiosfromUCSCforfourmousecelllines(CH12,J185a,L1210andMEL);fiveprimarysources(epidermalstemcells-EpiSC-5,EpiSC-7-,embryonicstemcells-ES-46C,ES-D3-andembryonicfibroblast-MEF-).Wenoticedagenerallygoodagreementbetweentechnicalreplicatesbutsubstantialvariationbetweencelltypes,particularlyatcertaingenomicregions.Togenerateaconservativesegmentation

上一页第5页

你可能喜欢

热门文档

相关文档

  • GenomeStudio分析软件流程

    GenomeStudio分析软件流程_基础医学_医药卫生_专业资料。GenomeStudio 分析软件使用简单流程一、导入数据简单分析 双击 GenomeStudio 软件图标 ,打开软件,进入如下界面,...

  • gene&genome I(07-3)

    gene&genome I(07-3)_生物学_自然科学_专业资料。Chapter 2 From Genes to Genomes Terminology and Basic Concepts Gene Allele, Polymorphism Genome G...

  • 教程UCSC Genome Browser

    The Genome Browser ? 搜索实例: human, Mar 2006 assembly, BRCA1 中国测序论坛www.seq.cn 4 中国测序论坛www.seq.cn Genome Browser 结果显示页面基因组注释...

  • Genome Walking Kit (10次量)

    Genome Walking Kit 特异性引物设计示意图特异性引物设计原则: 根据经过验证的已知序列区(最好不少于 500 bp)设计三个特异性引物(见上图) :设计方向为需要扩 增...

  • Genome Studio分析基因表达并用MEV制作热图

    Genome Studio分析基因表达并用MEV制作热图_生物学_自然科学_专业资料。Genome Studio分析基因表达并用MEV制作热图Genome Studio Illumina The Contents Introduction· ...

  • Genome Shuffling技术原理及其在菌种改良中的应用

    该 实验首次证 实了Genome Sufn 可应用于真核微生物 ,为后续的 hfigl 研究开辟 了道路。 嚏壤 、 i: 页臻得 良好 的成绩 。因此 ,在体育运 动钟 ,有 ...

  • 1.Human genome project

    Size of the Human Genome Per haploid genome New estimate 32,000 genes Functional Genomics 利用结构基因组学提供的信息,以高通量,大规模实验 方法及统计与...

  • GENOME+SHUFFLING技术优化及其应用进展

    GENOME+SHUFFLING技术优化及其应用进展_理学_高等教育_教育专区。基因组重组2010年4月-_—_—_-_?_——_——_—--?—————_—_——_?一I 第4期(总第...

  • 【论文】Genome shuffling技术在微生物育种中的应用

    介绍了Genome shuffling技术的特点及作用,...

  • Gene and genome

    一、基因组(Genome)一般概念 基因组(Genome)一般概念 (Genome) 一套完整单体的遗传物质总和。可以指一个细胞、一 个个体、一个种群。 基因组与基因 对人类而言...

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