HowardHughesMedicalInstituteandSectionofMolecularCellandDevelopmentalBIOLOGy,UniversityofTexas,Austin,TX78712,USA 1Presentaddress:DepartmentofMolecularandCellBiology,UniversityofCalifornia,Berkeley,CA94720,USA. 2Presentaddress:CenterforCellDynamics,JohnsHopkinsUniversitySchoolofMedicine,Baltimore,MD21205,USA. 3Correspondingauthor(wallingford@mail.utexas.edu). INTRODUCTION EmbryosofthefrogXenopuslaevisareanidealmodelsystemforinvivoimagingofdynamicbiologicalprocesses,fromtheinnerworkingsofindividualcellstothereshapingoftissuesduringembryogenesis.Theirexternallydevelopingembryosaremoreamenabletoinvivoanalysisthaninternallydevelopingmammalianembryos,andthelargesizeoftheembryosmakethemparticularlysuitablefortime-lapseanalysisoftissue-levelmorphogeneticevents.Inaddition,individualcellsinXenopusembryosarelargerthanthoseinothervertebratemodels,makingthemidealforimagingcellbehaviorandsubcellularprocesses(e.g.,followingthedynamicsoffluorescentfusionproteinsinlivingorfixedcellsandtissues).Xenopusembryosareamenabletosimplemanipulationsofgenefunction,includingknockdownandmisexpression,andthelargenumberofembryosavailableallowsevenaninexperiencedresearchertoperformhundredsofsuchmanipulationsperday.Transgenesisisquiteeffectiveaswell.Finally,becausethefatemapofXenopusembryosisstereotypical,simpletargetedmicroinjectionscanreliablydeliverreagentsintospecifictissuesandcelltypesforgenemanipulationorforimaging.Althoughyolkopacitycanhinderdeepimaginginintactembryos,almostanycellintheearlyembryocanbeplacedintoorganotypicculture,suchthatthecellsofinterestaredirectlyapposedtothecoverglass.FurThermore,liveimagingtechniquescanbecomplementedwithimmunostainingandinsituhybridizationapproachesinfixedtissues.Thisprotocoldescribesmethodsforlabelingandhigh-magnificationtime-lapseimagingofcellbiologicalanddevelopmentalprocessesinXenopusembryosbyconfocalmicroscopy. RELATEDINFORMATION ProtocolsforLow-MagnificationLiveImagingofXenopusEmbryosforCellandDevelopmentalBiology(Wallingford2010a)andPreparationofFixedXenopusEmbryosforConfocalImaging(Wallingford2010b)arealsoavailable,asaredetailsonperformingknockdownormisexpressionstudiesinXenopusembryos(Guille1999;Siveetal.2000).InformationisalsoavailableonEmbryoDissectionandMicromanipulationTools(Siveetal.2007). ExamplesofconfocalimagingofliveXenopusembryoscanbeseeninFigures1and2.Themethodsdescribedherehavealsobeenusedtomonitortissue-levelmorphogeneticevents,suchasgastrulationandneuraltubeclosure(WallingfordandHarland2002;Haigoetal.2003;Ewaldetal.2004).Imagingcanalsobeperformedsimultaneouslywithmeasurementoftheforcesgeneratedbymovingtissuesduringdevelopment(Zhouetal.2009). MATERIALS Reagents Agarose(2%,preparedin1/3XMMR)(forimagingembryosthroughtheneurulastage) Agarose,low-melt(0.8%)(forimagingtailbud/tadpole-stageembryos) Plasmidsencodinggreenfluorescentprotein(GFP)orredfluorescentprotein(RFP)fusionproteins ThereareavarietyoffluorescentproteinssuitableforimaginginXenopusembryos.EnhancedGFP(eGFP)andmonomericRFP(mRFP),inparticular,offerexcellentperformancewhenbalancingbrightnessversusphotobleaching.Generally,makingfusionstoXenopusproteinsispreferable,becausethesearemorereliable.However,fluorescentfusionstomammalianproteinsexpressedinXenopuscanalsobeused.ForexpressioninXenopus,vectorsoftheCS2family(CS2+,CS107,etc.)arerecommended.BecausemanyGFPfusionproteinsaregeneratedusingtheClontecheGFPvectors,wehavecreatedausefulCSfamilyvector(CS10R)designedforeasyshuttlingfromtheClontechvectors.Thisvectorisavailableuponrequest.ManyofourplasmidsaredepositedwiththeEuropeanXenopusResourceCentre(http://port.ac.uk/research/exrc/). Xenopusembryosofthestageofinterest Equipment Comb Combscanbepreparedbycarefulmeltingofaplastichaircomb. Computer,equippedwithimageprocessingsoftware TheprotocoldescribedhereusesApplecomputersequippedwithAdobePhotoshopandQuickTimePro,althoughnumeroussoftwarepackagesareavailable,withImageJbeingperhapsthemostcommonlyused. Coverglass Careshouldbetakenwhenselectingcoverslipsforhigh-resolutionimagingexperiments.Differentmicroscopemanufacturerscalibratetheirobjectivelensesforslightlydifferentthicknessesofglass.Additionally,differentcoverslipmanufacturersuseslightlydifferentglasscompositionsandmaketheircoverslipsinavarietyofthicknesses.Consultyourmicroscopeservicerepresentativefortheidealcoverslipthicknessandglasscomposition.Somehigh-resolutionlensesalsoprovidecorrectioncollarstoenablefinescalematchingtothecoverslipthickness. Embryodissectionequipment(e.g.,forceps,hairloops,hairknives[Keller1991],razorblades) Foradditionalinformation,seeEmbryoDissectionandMicromanipulationTools(Siveetal.2007). EquipmentforinjectingXenopusembryos Microscope,inverted,equippedwithvideo-recordingcapABIlities(e.g.,ZeissLSM5PASCALandLSM5LIVEconfocalmicroscopes),equippedwithPlan-NeofluarandPlan-Apochromatobjectives(10X,numericalaperture[NA]=0.3;20X,NA=0.5;40Xoil,NA=1.3;63Xoil,NA=1.4). Dependingontheapplication,awidevarietyofconfigurationsmightbenecessary.Ingeneral,itisessentialtoinvestasignificantamountoftimeintrialanderrortooptimizeeachimagingapplication. Xenopusimagingchambers Commercialchambersareavailable,butreusableimagingchambersareeasytomanufactureandeasytouse.Forimaginginaqueoussolutions(e.g.,alltheliveimagingapplicationsdescribedhere),asimplePetridishviewingchamberwithacoverglassbottomisusedforimagingoninvertedmicroscopes.Theassemblyconsistsofathreadedplasticinsertandacounterthreadedmetalbase.Theselockacircularcoverglassintoplacewithano-ring.PhotographsofthecomponentsareshowninFigure3andaschematicofthechamberisshowninFigure4.Manymachineshopscanfabricatethesedishesbasedoncomputer-aideddesign(CAD)drawings(madeusingSolidWorks),availablefromtheauthoruponrequest. METHOD DISCUSSION Xenopusembryosprovideanexcellentplatformforliveimagingacrossawiderangeofsizescales.Forexample,wholeembryoscanbefilmedwithastereomicroscopetoaskquestionsabouttissuemorphogenesis(seeLow-MagnificationLiveImagingofXenopusEmbryosforCellandDevelopmentalBiology[Wallingford2010a]).Theverysameembryocanthenberemovedfromthestereoscopestageandmountedintactonaconfocalmicroscopeforthree-dimensionaltime-lapseimagingofmigratorycellmovements(Fig.2),cytoskeletalorganization,orthedynamicsofGolgistructure(Fig.1C,D),tonamejustafewapplications.ItisimportantalsotonotethatXenopusembryoscanbeeffectivelyimagedacrossawidevarietyoftimescales.Intactembryoshavebeenusedtomake4Dconfocalmoviesofmyeloidcellmigrationthatspan15h,andthebeatingofciliahasbeenimagedat>370frames/sec(Fig.2). Variouscell-membrane-targetedGFPfusionsexistforexaminingcellmorphologyinlivingtissues,includingGFPstargetedtothemembranebyadditionoftransmembranedomainsorbyadditionoffarnesylationsequences.Evenathighlevelsofexpression,littleGFPisdetectedinthecytoplasm.Itshouldbenoted,however,thatintracellularvesiclesarelabeledwithhigherdosesofmRNA.Thesereagentsprovideverybrightlabelingofcellmembranes,allowingvisualizationnotonlyofthecellbody,butalsomembranefeaturessuchasexocyticpitsandfilopodia/lamellipodia(Wallingfordetal.2000;Hayesetal.2007;Davidsonetal.2008). Similarly,imagingofproteinlocalizationwithfluorescentfusionproteinsishighlyeffectiveinXenopus.Wehaveimagedawidevarietyofproteins,makingeffectiveuseofGFPandRFPfusionproteinstovisualizecytoskeletalstructures,suchasmicrotubules(Kiesermanetal.2008)andsubcellularcompartments,suchastheGolgi(Fig.1C,D).Finally,avarietyoffluorescentsensorscanbeappliedinlivingXenopusembryos,includingGFP-basedsensorsofGTPaseactivity(BeninkandBement2005)andfluorescentcalciumindicators(Wallingfordetal.2001).Becauseoverexpressioncanleadtoectopicorabnormalproteinlocalization,careshouldbetakentoinjectthelowestdoseofmRNAthatallowsvisualizationoftheconstruct.ThelargesizeofthecellsinXenopusembryosallowsresolutiontobesacrificedforphotons. Thegenerationofmosaicembryos,inwhichmanipulatedandunmanipulatedcellscanbevisualizedwithinasingletissueofasingleembryo,isanextremelypowerfulexperimentalapproach.Xenopusembryosareverywell-suitedformosaicanalysesusingsimpletargetedmicroinjectionapproaches(Fig.6).Forexample,reagentsformanipulatinggenefunction(e.g.,antisensemorpholino-oligonucleotides,mRNAs)caneasilybedeliveredspecificallytothedesiredtissuesbytargetedinjection.Targetingiseasilyachievedbecauseofthewell-knownpigmentationpatternsofearlyembryosandthestereotypicfatemapsforthefour-,eight-,16-,or32-cellstagesofXenopus(DaleandSlack1987;Moody1987;MoodyandKline1990).Basictargetedmicroinjectionisperformedbyconsultingfatemapstodeterminetheblastomereoriginofthetissueofinterestandinjectingthedesiredreagents(mRNAsormorpholino-oligonucleotides)accordingly.NotethattheanimalpigmentationofXenopusembryosisgenerallyasymmetric,andtwomore-darklypigmentedblastomerescanusuallybeidentifiedinfour-cellembryos.Thesetwodarkerblastomereswillformtheventraltissues,andthelighteroneswillformthedorsaltissues(Fig.6).Thus,injectionsintothelighterblastomereswilltargettissuessuchastheneuralplateornotochord,andinjectingthedarkerblastomereswilltargettheepidermis.Injectionsatearlystages(fourtoeightcells)canbeusedtotargetlargertissuedomains(germlayers),whereasinjectionsatlaterstagescanbeusedtotargetmorespecifictissues,suchasthegutorkidney(Wallingfordetal.1998;Lietal.2008).Alwaysrememberthatfatemapsarepredictive,notaguarantee.Injectedembryosshouldbescreenedatgastrulaorneurulastagestoensureexpressioninonlytheintendedtissues(WallingfordandHarland2002;KiesermanandWallingford2009). Liveimagingofintactembryosisanexcellentapproach,butislimitedtosurfacetissuesbytheopacityofamphibianembryos.Tocircumventthisproblem,cellsandsubcellularstructureswithintheembryocanbevisualizedquiteeasilythroughmicrosurgeryandsubsequentcultureoftissueexplants.Withpractice,nearlyanytissuecanbemicrosurgicallyisolatedfromanyregionoftheembryouptothelatetadpolestages(Keller1991).Briefly,explantedtissuesareheldinplacegentlyunderafragmentofcoverglasssecuredinplacewithmodelingclayorhigh-vacuumsiliconegrease.Whenculturedinanappropriatemedium(e.g.,DFAorSteinberg’s,whicharespeciallyformulatedtomatchthelow-chloride,high-pHconditionsoftheinterstitialfluidoftheearlyXenopusembryo),cellswithinexplantscontinueinvivoprogramsofcellmotilityandcellrearrangement(Kelleretal.1985).Ingeneral,cellswithinphysicallyisolatedexplantsautonomouslyrecapitulatemovementswithintheembryo.Forexample,imagingofdorsalmarginalzoneexplantsisusedcommonlytoexamineconvergentextension.Asisthecaseforwholeembryos,Xenopusexplantshavebeenusedtoimagecellprotrusiveactivity,proteinlocalization,extracellularmatrixorganization,andevencalciumdynamics(Wallingfordetal.2000,2001;DavidsonandWallingford2005;Davidsonetal.2008).CulturedXenopusanimalcaps,oftenusedforgeneexpressionstudies,arealsoveryamenableforimaging,havingbeenusedtoexamineproteinlocalization(Axelrodetal.1998;Parketal.2005),nucleocytoplasmicshuttling(Batutetal.2007),andmorphogendiffusion(Williamsetal.2004). Withthecombinationofexternaldevelopment,largecellsize,andawell-definedfate-mapfortargetedmicroinjectionandgenerationofmosaics,Xenopusembryosprovideanunparalleledsystemfortheinvivostudyofthecellbiologyofvertebratedevelopment.Examinationofcellsincultureisanimmenselypowerfulapproach,butmanybiologicalprocessesthathavebeenwell-studiedinculturehavesubsequentlybeenfoundtobesubjecttosubstantialdevelopmentalcontrolinembryosorintacttissues.Thatis,cellularprocessessuchasmigrationanddivisionhavefrequentlybeenfoundtodifferdramaticallyevenindifferentcelltypeswithinthesameembryo.Thus,extensiveinvivostudywillbeessentialbeforeacomprehensiveunderstandingwillemerge.Moreover,becausebasiccellbiologicalprocessesunderlieallofdevelopmentalbiology,fromsignalingtomorphogenesis,acomprehensiveunderstandingofdevelopmentwillrequirethatcellbiologicalapproachesbeappliedmoreandmoretodevelopmentalproblems.ThecellbiologystudiesthatcanbemosteffectivelycarriedoutinXenopusembryoswillcomplementthegeneticstudiesofmouseandzebrafishaswemoveforward. ACKNOWLEDGMENTS ThemethodsdescribedherehavebeendevelopedandoptimizedbyallthemembersoftheWallingfordlaboratory.WethankHye-JiChafortheGolgiimage.ThankstoAndrewEwaldfortheoriginalconstructionoftheXenopusimagingchambers.WealsothankAndrewEwaldfordiscussions.WorkintheWallingfordlaboratoryhasbeenfundedbytheNationalInstitutesofHealth/NationalInstituteofGeneralMedicalSciences(NIH/NIGMS),theMarchofDimes,TheBurroughsWellcomeFund,andtheSandlerProgramforAsthmaResearch. REFERENCES
Viewlargerversion(89K):Figure1.ImagingofmorphogenesisinliveXenopusembryos.(A)Stillframesfromatime-lapsemovieofneuraltubeclosureinXenopustakenwithastereomicroscope.MountingforthisapplicationisdescribedinLow-MagnificationLiveImagingofXenopusEmbryosforCellandDevelopmentalBiology(Wallingford2010a).(B)TheimagesshowninthiscolumncorrespondtostagessimilartothoseshowninA,butathighermagnificationtoshowcellsoutlinedwithmembraneGFP(memGFP)intheregionindicatedbytheyellowboxinA;seealsoLeeetal.(2007).(C)Single3Dprojectionofmucus-secretingcellsontheepidermisofaXenopusembryo.GolgistructurescanbelocalizedwithGalT-RFP(Nicholsetal.2001)andapicalexocyticvesiclesarehighlightedusingmemGFP(Hayesetal.2007).(D)StillframesofamovieofadividingcellintheneuralepitheliumofXenopus,showingthemicrotubules(
-GFP;KwanandKirschner2005)andthecellmembrane(memRFP).
Viewlargerversion(26K):Figure2.Invivotime-lapseimagingofXenopusembryoscanbeperformedacrossawidevarietyofsizeandtimescales.(Top)Amovietakenat~370frames/secandspanningonly~35msecshowsthebeatingof20-mm-longmotileciliaonasinglemulticiliatedcellintheXenopusepidermis(Parketal.2008).(Bottom)Confocalstackscollectedevery5minandspanning~12hshowthedispersalofindividualfluorescentmyeloidcellsthroughoutaXenopusembryofromtheirorigininsurgicallytransplantedventralbloodislands.Eachindividualcellis~30µmacrossandtheentireembryoshownhereis~1mmlong.MountingforbothimagingapplicationswasasdescribedinSteps2.iv-2.vi(Fig.7B).
Marc’smodifiedRinger’s(MMR)(1X)
Tricaine(0.15%)(optional;seeStep5)
Viewlargerversion(115K):Figure3.Componentsofthecustom-madechamberforimaginginaqueoussolutionswithaninvertedmicroscope.SeeFigure4forassembly.
Viewlargerversion(29K):Figure4.Schematicofacustom-madechamberforimagingembryosinaqueoussolutions.CADdrawingsofthesecomponentsareavailablefromtheauthoruponrequest.
Viewlargerversion(32K):Figure5.GenerationofmosaicXenopusembryosbytargetedmicroinjectionofplasmidDNAs(Vizeetal.1991).(A)mRNAencodingafluorescentmarkerisinjectedfirsttolabelthetissueuniformly.(B)AsecondinjectionofplasmidDNAismadesubsequently.(C)MosaicismisobservableinembryosinjectedwithmRNA(red)andplasmidDNAthatsegregatesandtranscribesheterogeneously(green).
Viewlargerversion(56K):Figure6.GenerationofmosaicXenopusembryosbytargetedmicroinjection.(A)Tissue-levelmosaicscanbemadebysequentialinjectionsatthefour-andeight-cellstages(Kiesermanetal.2008).(B)Cell-levelmosaicscanbegeneratedbysequentialinjectionsatthefour-cellandthe16-or32-cellstages(Grayetal.2009).
Viewlargerversion(26K):Figure7.Mountingofembryosforinvivotime-lapseimaging.(A)Mountingforblastula,gastrula,andneurulastages(seeSteps2.i-2.iii).(B)Mountingfortailbudstages(seeSteps2.iv-2.vi).
Thesizeofstacks,theamountofoverlapbetweenthesliceswithinstacks(z-resolution),andthecollectiontimes(time-resolution)mustbeoptimizedempiricallyforyourspecimen,fluorophore,andapplication(see,e.g.,Kiesermanetal.2008,SupplementalTable2).
Viewlargerversion(32K):Figure8.Schematicprotocolforcollectionof4DdatasetsfromXenopusembryos.(A)Protocolforimagingneuraltubeclosure.(B)Protocolforimagingtheepidermis.
Adjustments
Levels).ThisfunctionismoreusefulthantheBrightness/Contrastfunction.
Sharpen
UnsharpMask).
Noise
Median)orDust&Scratches(Filter
Noise
Dust&Scratches)filter.
ImageSize)andcompress(File
SaveAs)individualtifffiles.ThiscanbedoneforalltheimagesusingtheFile
Automate
Batchfunction.Alternatively,themoviecanberesizedafterassembly(seeStep9).Tinymoviesareusefulfore-mailingtocollaborators,butthelargest,highestqualitymoviesthatarepracticalshouldbesubmittedwithapaperforreview.
OpenImageSequence).
Export
OptionsfunctioninQuickTimePro.
-tubulindistributionandmicrotubulearchitectureduringepithelialcellshapechange.Development134:1431–1441.