yellowsoftmod

Cooperativecouplingofcell-matrixandcell–celladhesionsincardiacmuscleMeganL.
McCaina,HyungsukLeea,1,YvonneAratyn-Schausa,AndréG.
Kléberb,andKevinKitParkera,2aDiseaseBiophysicsGroup,WyssInstituteforBiologicallyInspiredEngineering,SchoolofEngineeringandAppliedSciences,HarvardUniversity,Cambridge,MA02138;andbDepartmentofPathology,BethIsraelDeaconessMedicalCenter,HarvardMedicalSchool,Boston,MA02215EditedbyRobertLanger,MassachusettsInstituteofTechnology,Cambridge,MA,andapprovedMay1,2012(receivedforreviewFebruary21,2012)Adhesionbetweencardiacmyocytesisessentialforthehearttofunctionasanelectromechanicalsyncytium.
Althoughcell-matrixandcell–celladhesionsreorganizeduringdevelopmentanddis-ease,thehierarchicalcooperationbetweenthesesubcellularstruc-turesispoorlyunderstood.
Wereasonedthat,duringcardiacdevelopment,focaladhesionsmechanicallystabilizecellsandtis-suesduringmyofibrillogenesisandintercalateddiscassembly.
Astheintercalateddiscmatures,wepostulatedthatfocaladhesionsdisassembleassystolicstressesaretransmittedintercellularly.
Finally,wehypothesizedthatpathologicalremodelingofcardiacmicroenvironmentsinducesexcessivemechanicalloadingofinter-calateddiscs,leadingtoassemblyofstabilizingfocaladhesionsadjacenttothejunction.
Totestourmodel,weengineeredμtissuescomposedoftwoventricularmyocytesondeformablesubstratesoftunableelasticitytomeasurethedynamicorganizationandfunctionalremodelingofmyofibrils,focaladhesions,andinterca-lateddiscsascooperativeensembles.
Maturingμtissuesincreasedsystolicforcewhilesimultaneouslydevelopingintoanelectrome-chanicalsyncytiumbydisassemblingfocaladhesionsatthecell–cellinterfaceandformingmatureintercalateddiscsthattransmittedthesystolicload.
Wefoundthatengineeringthemicroenvironmenttomimicfibrosisresultedinfocaladhesionformationadjacenttothecell–cellinterface,suggestingthattheintercalateddiscre-quiredmechanicalreinforcement.
Inthesepathologicalmicroenvir-onments,μtissuesexhibitedfurtherevidenceofmaladaptiveremo-deling,includinglowerworkefficiency,longercontractioncycleduration,andweakenedrelationshipsbetweencytoskeletalorga-nizationandforcegeneration.
Theseresultssuggestthatthecoop-erativebalancebetweencell-matrixandcell–celladhesionsintheheartisguidedbyanarchitecturalandfunctionalhierarchyestab-lishedduringdevelopmentanddisruptedduringdisease.
adherensjunctions∣sarcomere∣mechanotransduction∣extracellularmatrixIntheadultheart,healthyventriculartissueischaracterizedbyspatialsegregationofcell-matrixadhesionstotransversemyocyteborders(1)andcell–celladhesionstolongitudinalmyo-cyteborders(2).
Manycardiomyopathiesarecharacterizedbylateralizationofcell–celladhesions(3–5),increasedexpressionofcell-matrixadhesions(6,7),andarrhythmogenesis(8),suggest-ingthatspatiallyorganizedadhesionisessentialforeffectiveeletromechanicalcoupling(9).
Althoughinvivostudieshaveshownthatlocalizationofcell-matrixandcell–celladhesionsisdevelopmentallyregulated(10–13),thefactorsthatregulatetheirassemblyanddisassemblyarenotwellunderstood.
Cell-matrixadhesionsareespeciallyimportantduringorgano-genesis,anchoringdevelopingcellsandtissuestotheextracellu-larmatrix(ECM)(14)anddirectingcellmigration(15–17).
Asdevelopmentprogresses,cell-matrixadhesionsmustselectivelydisassemblesothatcell–celladhesionscanform(18).
Intheheart,thisexchangeofadhesionsitesoccurswhilemyocytesarecontractile(19,20),indicatingthatmyofibrils,cell–cell,andcell-matrixadhesionsmustcooperativelyremodelwithoutcompro-misingcardiacoutput.
Becausecell-matrix(21–24)andcell–celladhesions(25,26)aremechanosensitive,mechanicalforceslikelyserveascuesforremodelingadhesionsandassemblingtissues.
Invitrostudieshavedemonstratedthatcytoskeletaltension(27)andexogenouscyclicstrain(28,29)promotecell–celladhesionandtissueassemblyinmanycelltypes.
Culturingnoncardiaccellsonstiffsubstratestipsthebalanceofadhesioninfavoroffocaladhesionsandawayfromcell–celladhesions(30–32),suggestingthatmechanicalforcescanmodulatetheassemblyordisassemblyoftissues.
Manycardiomyopathiesarecharacterizedbyincreasedfibrosisandtissuestiffening(33,34),suggestingthatcardiacmyo-cytesmightbesimilarlysensitivetotissuecomplianceandfavorfocaladhesionformationovercell–celljunctionformationinafibroticmicroenvironment.
Wehypothesizedthatfocaladhesionsmechanicallystabilizemyocytesduringtissueassembly,whenmyofibrilsarecontractilebuttheintercalateddiscisnotyetfullyassembled.
Wepostulatedthatfocaladhesionsdisassembleasintercalateddiscsmaturesuchthatsystolicforcesaretransmittedintercellularly.
Further-more,wereasonedthatstiffmicroenvironmentspotentiateexces-siveintrinsicloadingthatdestabilizestheintercalateddiscandinducesfocaladhesionformationnearthecell–celljunction.
Wemodeleddeveloping,healthy,anddiseasedcardiactissuebyengineeringμtissuescomposedoftwoneonatalratventricularmyocytesondeformablesubstratesoftunableelasticity.
Tochar-acterizeprogressivestagesoftissueassembly,weculturedμtis-suesonphysiologicalsubstratesandmeasuredstructuralandfunctionaloutputsovertime.
μtissuesgraduallyincreasedsystolicforcegeneration,disassembledfocaladhesionsnearthecell–cellinterface,andformedintercalateddiscstotransmitthesystolicload.
Tomimicdiseasedmyocardium,weculturedμtissuesonstiffsubstratesandobservedfurtherincreasesinforcegeneration,maladaptivefunctionalremodeling,andfocaladhesionforma-tionadjacenttothecell–cellinterface.
Thesedatasuggestthatregulationofcell-matrixandcell–celladhesionsduringcardiacdevelopmentisguidedbyanintrinsichierarchythatisalteredindiseaseduetomechanicalremodelingofthemicroenvir-onment.
ResultsECMRegulatesCooperativeIntra-andIntercellularAssembly.
Wehy-pothesizedthatfocaladhesionattachmentstotheECMareimportantindevelopmentfordirectingmigrationandmyofibril-logenesisandstabilizingmyocytesastheybecomecontractile.
However,wesuspectedthatfocaladhesionsadjacenttothecell–celljunctiondisassembleoncetheintercalateddisciscapableoftransmittingsystolicloadsintercellularly.
Totestthismodel,weAuthorcontributions:M.
L.
M.
,H.
L.
,Y.
A.
-S.
,A.
G.
K.
,andK.
K.
P.
designedresearch;M.
L.
M.
performedresearch;M.
L.
M.
,H.
L.
,andY.
A.
-S.
contributednewreagents/analytictools;M.
L.
M.
,H.
L.
,andY.
A.
-S.
analyzeddata;andM.
L.
M.
andK.
K.
P.
wrotethepaper.
Theauthorsdeclarenoconflictofinterest.
ThisarticleisaPNASDirectSubmission.
1Presentaddress:SchoolofMechanicalEngineering,YonseiUniversity,Seoul120-749,Korea.
2Towhomcorrespondenceshouldbeaddressed.
E-mail:kkparker@seas.
harvard.
edu.
Thisarticlecontainssupportinginformationonlineatwww.
pnas.
org/lookup/suppl/doi:10.
1073/pnas.
1203007109/-/DCSupplemental.
www.
pnas.
org/cgi/doi/10.
1073/pnas.
1203007109PNAS∣June19,2012∣vol.
109∣no.
25∣9881–9886BIOPHYSICSANDCOMPUTATIONALBIOLOGYDownloadedbyguestonJanuary22,2021engineeredrectangularμtissuesconsistingoftwoneonatalratventricularmyocytesonmicropatternedfibronectin(FN)islandsandmonitoredthemovertime.
Eighthoursafterseeding,spreadingmyocytestypicallyprotrudedalongalateraledgeoftheisland,independentofwhetherone(Fig.
1A)ortwomyocytes(Fig.
1B)adheredtotheisland.
After12h,μtissuesfilledtheislandandstriatedmyofibrilsandsigmoid-likecell–cellinterfaceswereobservable(Fig.
1CandD).
Incontrast,myocytesonsub-stratescoatedwithuniformFNextendedrandomlyorientedpro-trusionsbeforeformingconfluent,isotropicmonolayers(Fig.
S1A–C).
Cell–celljunctionswererelativelylinear(Fig.
S1D)andperpendiculartomyofibrils,characteristicofjunctionsinvivo(2).
Theseresultsindicatethatcell-ECMinteractionsguidecellspreadingandmyofibrillogenesisearlyintissueassembly,influ-encingcell–celljunctionformationbyregulatingmyocyteshape.
Althoughfocaladhesionsarecriticalearlyintissueassembly,wereasonedthatfocaladhesionsnearthecell–cellinterfacedis-assembleasadjacentintercalateddiscsestablishintercellularmechanicalcontinuity.
Toinvestigatethis,weimmunostainedforpaxillintomonitordifferencesinfocaladhesionlocalizationbetweenμtissuesculturedonmicropatternedpolyacrylamidegelswithphysiologicalstiffness(13kPa)(35)for1d(Day1)and4d(Day4)torepresentimmatureandmaturetissues,respectively.
AtDay1(Fig.
1E),paxillinsignal,althoughrelativelydiffuse,wasdetectedinplaquesatthelongitudinalendsoftheμtissue,colo-calizedwithterminatingmyofibrils.
Paxillinplaqueswerealsodetectednearthecell–cellinterface,suggestiveoffocaladhesionformation.
AtDay4(Fig.
1F),paxillinsignalwaslocalizedtothelongitudinalendsoftheμtissueinlargerandmoredistinctpla-ques.
Nearthecell–cellinterface,paxillinsignalappearedrela-tivelyweak,suggestingthatfocaladhesionsnearthecell–cellinterfacedisassembledoverthecourseoftissuematuration.
Becausebothcell-matrixandcell–celladhesionsanchorthecytoskeleton,wereasonedthatcell–celljunctionmorphologyisconstrainedbycell-matrixadhesionsandregulatedbythestageoftissueassembly.
InDay1and4μtissues,thecell–celljunctionhadasigmoid-likecontour,similartoyin-yanginterfacesob-servedinmigratingendothelialcellsconstrainedtoFNislands(17)anddiagonalinterfacesinpatternedmyocytepairs(36).
Toquantifythis,wemeasuredjunctiongeometrywhilevaryingμtissuelengthtowidthratio(μLW)(Fig.
1G–J).
Wedefinedthelongitudinalandtransverseaxisoftheμtissueasthexandyaxis,respectively,andfitthejunctiontothelogisticfunctionforasigmoidcurve,fJa1ebx.
Thequantitya·brepresentsslopeattheμtissuecenter,wherex0(Fig.
1K).
WegroupedtogetherμLW2–8and8–12andfoundthat,atbothDay1and4,slopewaslowerforμLW8–12comparedtoμLW2–8(Fig.
1L).
JunctionslopedecreasedbetweenDay1and4forμLW8–12(Fig.
1L),likelybecausemyocyteswerestillremodel-ingtheiradhesionsandmyofibrilsatthisearlytimepoint.
Wequantifiedtime-dependentchangesincytoskeletalarchitecturebycalculatingtheactinorientationalorderparameter(OOP),whichrangesfrom0forisotropicto1forperfectlyalignedsys-tems(37),andfoundaslight,butnotstatistical,increaseinactinOOPbetweenDay1and4forμLW2–8and8–12(Fig.
1M).
AtbothDay1and4,theOOPforμLW8–12washigherthanμLW2–8duetoincreasedmyofibrilalignmentathigherμLW.
Thesedatasuggestthatmyofibrillogenesisandremodelingofcell-matrixandcell–celladhesionsarecooperativeprocessesincar-diacdevelopment.
SystolicForceOutputIncreasesasTissuesAssemble.
Basedonourstructuraldata,wereasonedthatthemagnitudeofpeaksystolictractionforcetransmittedtotheECMnearthecell–cellinterfacedecreasedasμtissuesdisassembledfocaladhesions,formedmaturecell–celljunctions,andtransmittedsystolicforcesinter-cellularly.
Togeneratetractionstressvectormapsofthesub-strate,weculturedμtissuesonmicropatternedpolyacrylamidegels(13kPa)dopedwithfluorescentmicrobeadsfortractionforcemicroscopy(TFM)(38).
Witheachspontaneouscontrac-tion,μtissuesdeformedthegelanddisplacedthefluorescentbeads,whichwascapturedwithaCCDcameraat40Hz.
Beaddisplacementwasmeasured,yieldingdisplacementvectormapsusedtocalculatetractionstressvectormapsduringcontractionandrelaxation(37,38).
FollowingTFMmeasurements,μtissueswereimmunostainedforactin,β-catenin,andnucleitoexaminecytoskeletalstructureandidentifythecell–celljunction.
AtDay1(Fig.
2AandB),wedetectedpeaksystolictractionstressesatthelongitudinalendsoftheμtissueandnearthecell–cellinter-face,consistentwiththepresenceoffocaladhesions.
AtDay4(Fig.
2CandD),peaksystolictractionstresseswerereduced,ornotdetected,nearthecell–cellinterface,suggestingthatcontractileforcewastransmittedprimarilythroughthecell–cellFig.
1.
μTissueformationonmicropatternedFNislands.
Single(A)andpaired(B)myocytesculturedfor8honFNislandstypicallyprotruded(whitearrows)alongedgesoftheFNpattern(yellowdashedlines).
(C)Twelvehoursafterseeding,μtissuescoveredFNislandsandstriatedmyofibrilswereappar-ent(white:actin,blue:nuclei).
(Scalebars:10μm.
)(D)Time-dependentmea-surementsofislandcoverageforμtissueswithlengthtowidthratios(μLW)of3.
3,5,and6.
7indicatethatmyocytespreadingwascompleteafter12h.
Barsstandarderror;n≥4ateachpoint.
RepresentativeDay1(E)and4(F)μtissuesstainedforactin(red),paxillin(green),andnuclei(blue).
(Scalebars:10μm.
)β-Cateninimmunostains(green:β-catenin,red:actin,blue:nu-clei)forrepresentativeDay4μtissuesculturedonpolydimethylsiloxanecover-slipswithμLW2(G),4(H),8(I),and12(J)demonstratethatcell–celljunctionshadsigmoid-likecontours.
(Scalebars:10μm.
)(K)Cell–celljunctioncontourswerefittothelogisticfunctionforasigmoidcurve,definedasfJ.
TheaveragefJforeachμLWisplotted.
Average(L)a·boffJand(M)actinOOP,whichrangesfrom0forisotropicto1forperfectlyalignedsystems,areplottedforDay1and4μtissuesgroupedbyμLW2–8and8–12.
Barsstandarderror;*p<0.
05relativetoμLW8–12,Day4group;n≥8foreachgroup.
9882∣www.
pnas.
org/cgi/doi/10.
1073/pnas.
1203007109McCainetal.
DownloadedbyguestonJanuary22,2021junctioninsteadoffocaladhesions.
Basedonthelocationoftheβ-cateninimmunosignal,wedeterminedtractionstressvectorscontributedbyeachmyocyte,calculatedtractionforcevectors(~Tn)bymultiplyingtractionstressateachgridpointnbythegridunitsurfacearea,andcalculatedtheaveragelongitudinaltensionthroughacross-sectionofeachmyocyte(Nxx;cell)(Fig.
2E).
WeplottedNxx;cell1andNxx;cell2throughasinglecontractioncyclefortheDay1(Fig.
2F)and4(Fig.
2G)μtissuesshowninFig.
2AandBandFig.
2CandD,respectively.
IntheDay1μtissue,myocyteswereslightlyoutofphaseandreachedpeaksystoleatdifferenttimepoints,suggestiveofweakelectricalcoupling(MovieS1).
Conversely,intheDay4μtissue,myocytesreachedpeaksystolesimultaneously(MovieS2),implyingelectricalcontinuityacrossthecell–celljunction.
ToconfirmelectricalcouplingonDay4,wemeasuredintercellularelectricalconductancebetweenmyocyteswithdualvoltageclamp(Fig.
S2A–C)anddetectedCx43immu-nosignalinpunctateclustersalongthecell–cellinterface(Fig.
S2DandE),demonstratingthatμtissueshadelectricalcontinuityviagapjunctionchannels(39).
Theseresultssuggestthat,betweenDay1and4,myocytesbecomesynchronousandtransmitforceacrossthecell–celljunctioninsteadoftotheECMnearthecell–cellinterface,indicativeoffocaladhesiondisassemblyandinter-calateddiscmaturation.
Tocorrelateadhesionremodelingtosystolicforcegeneration,wecalculatedtheaveragelongitudinaltensionthroughacross-sectioninDay1and4μtissues(Nxx;μtissue)andadaptedapre-viouslypublishedmethod(40)tocalculateforcesexertedagainstthejunctionbycells1(~FJ;cell1)and2(~FJ;cell2)(Fig.
2E).
Becauseeachmyocytemustbeinmechanicalequilibrium,thevectorialsumof~Tnassociatedwitheachmyocyteindicatestheforceexertedagainstthecell–celljunction(~FJ;Cell∑Cell~Tn)(Fig.
2E).
Themagnitudeofthex-componentof~FJ;Cellindicatesthelongitudinalcell–celljunctionforce(Fx;J;Cellj∑CellTx;nj).
AsshowninFig.
2HandI,Nxx;μtissue,Fx;J;cell1,andFx;J;cell2didnotchangebetweenDay1and4forμLW2–8.
However,Nxx;μtissue,Fx;J;cell1,andFx;J;cell2increasedsignificantlybetweenDay1and4forμLW8–12,likelybecausemyofibrilsarelongerathigherμLWandrequiremoretimetomature.
Thus,inμtissueswithhighμLW,systolicforcegenerationincreaseswhileadjacentfocaladhesionsdisassembleandintercalateddiscsmature,indi-catingthatforcegenerationandtissueformationoccuronsimilartimescalesandarepotentiallylinked.
StiffMicroenvironmentsInduceMaladaptiveRemodelingofContrac-tileFunction.
Cardiacmyocytesremodelinresponsetochangesinthephysicalmicroenvironment.
Forexample,theelasticmodulusofembryoniccardiactissuerangesfrom1–6kPaandincreasesto10–15kPaintheadultheart,whichisreportedtoincreasebeatingrate,myofibrilmaturation,andworkoutputinculturedmyocytes(35,41).
Increasesintissuestiffnessbeyond50kPaarecharac-teristicofmanycardiomyopathiesduetoexcessivefibrosis(33),whichisreportedtoinhibitcontractionandpromotestressfiberformationinvitro(35,41)andexceedstheelasticmodulusofisolated,relaxedmyofibrils(42).
Wehypothesizedthatincreasedstiffnessofthecardiacmicroenvironmentinducesmaladaptiveremodelingofcontractilefunctionandaltersthebalancebetweencell-matrixandcell–celladhesion.
Totestthis,weperformedTFMandpaxillinimmunostainingonμtissuesculturedongelswithlow(1kPa),moderate(13kPa),andstiff(90kPa)elasticmodulitocharacterizestiffness-dependentchangestocontractilefunctionandfocaladhesionformation.
Withincreasingelasticmodulus,peaksystolicdisplacementdecreased,forcegenerationincreased,andfocaladhesionsizeatthelongitudinalendsincreased(Fig.
3A–C).
Stiffness-dependentchangestodisplace-mentandstressgenerationareillustratedbytheslopesoftracesrelatingthemaximumtractionstressvector(max:jσx;μtissuej)tothemaximumdisplacementvector(max:jΔxμtissuej)inthelongi-tudinaldirection(Fig.
4A).
ByplottingpeaksystolicNxx;μtissue,Fx;J;cell1,andFx;J;cell2foreachstiffness(Fig.
4B),wefoundarelationshipanalogoustotheFrank–Starlinglaw,whereforcegenerationincreasesinresponsetoincreasedload(43)orme-chanicalstretch(44).
Wealsocalculatedaveragepeaksystolicworkasafunctionofstiffness,whichisanimportantparameterforevaluatingcardiacfunctionbecauseeffectivepumpingisFig.
2.
Remodelingofforcegenerationandtransmissionindevelopingcardiacμtissues.
(A)Day1μtissuestainedforactin(red),β-catenin(green),andnuclei(blue).
Thecorrespondingpeaksystolictractionstressvectormap(B)revealedthatcell2generatedsubstantialtractionagainstthesubstratenearthecell–cellinterface(whitearrows).
Conversely,aDay4μtissue(C,samestainingprotocolasA)showednegligibletractionstressnearthecell–cellinterfaceatpeaksystole(D).
(BandD)Solidwhiteline:μtissueoutline;dashedwhiteline:junctionoutline.
(Scalebars:10μm.
)(E,i)Schematic(notdrawntoscale)and(ii)equationsusedtocalculateNxx;cell1andNxx;cell2,(iii)Nxx;μtissue,and(iv)~FJ;cell1and~FJ;cell2.
PlotsofNxx;cell1andNxx;cell2forthe(F)Day1and(G)Day4μtissuesshowninAandBandCandD,respectively,duringasinglecontractioncycle.
Asterisksindicatepeaksystoleforeachmyocyte,whichoccuratdifferenttimepointsinF,butnotG.
Average(H)Nxx;μtissueand(I)Fx;J;cell1andFx;J;cell2asafunctionofdayincultureandμLW.
Barsstandarderror;*p<0.
05relativetoμLW8–12,Day4group;n≥6μtissuespergroup.
McCainetal.
PNAS∣June19,2012∣vol.
109∣no.
25∣9883BIOPHYSICSANDCOMPUTATIONALBIOLOGYDownloadedbyguestonJanuary22,2021dependentonforcegenerationandshortening.
Wefoundthatpeaksystolicworkincreasedfromsofttomoderatesubstratesbutdidnotfurtherincreasefrommoderatetostiffsubstrates(Fig.
4B).
Thus,μtissuesonstiffsubstratesgeneratesignificantlymoreforcetoproducesimilarmagnitudesofworkasonmoder-atesubstrates.
Previousreportsobservedthatcontractioncycledurationisprolongedinagingmyocardium(45,46)andthatforce-frequencyrelationshipsaredefectiveinheartfailuremodels(47).
Weaskedifmicroenvironmentalelasticityregulatescontractioncycledy-namics.
Thedurationofcontraction,averagedforallμLWoneachstiffness,increasedasthesubstratestiffened,althoughonlysoftandstiffsubstrateswerestatisticallydifferent(Fig.
4C).
Furthermore,μtissuesonstiffsubstrateshadprolongedpeaksystole(Fig.
S3),similartopreviousreports(35).
Inadditiontochangesintissuecompliance,cardiacdiseaseisassociatedwithremodelingofmyocyteshapeandcytoskeletalarchitecture(48),whichisthoughttobeacompensatoryresponsetomechanicaloverload(49).
Weaskedifsubstratestiffnessaffectsrelationshipsbetweentissuegeometryandforcegenera-tiontodetermineifthephysicalmicroenvironmentaffectsintrin-siccompensatorymechanisms.
WegroupedtogetherμLW2–4,4–8,and8–12andfoundthat,onsoftandmoderatesubstrates,peaksystolicNxx;μtissuewashighestatμLW8–12(Fig.
4D),likelybecausemyofibrillengthandalignmentaremaximized.
Thesam-plesizeforμLW8–12onsoftsubstratesislow(n2)becausetheseμtissuestypicallygeneratedenoughforcetodeformthehighlycompliantsubstrateoutofthemicroscopefocalplane,hin-deringtwo-dimensionalTFM.
Ourdataarelikelytheweakestofthepopulation,explainingthelackofsignificancebetweenμLW4–8and8–12.
Interestingly,peaksystolicNxx;μtissuewasnotdependentonμLWforstiffsubstrates(Fig.
4D),suggestingthatsubstratestiffnessdominatesovertissuearchitecturewhentheelasticmodulusisveryhigh.
WealsoquantifiedactinOOPandcell–celljunctionslopeforμtissuesoneachsubstratetoidentifyanystructuralvariabilitythatcouldcontributetothesefunctionaldifferencesandfoundthatbothparameterswereindependentofstiffness(Fig.
S4).
Collectively,thesedatashowthat,althoughstiffmicroenvironmentsincreaseforcegeneration,theoveralleffectismaladaptive,asdemonstratedbydecreasedworkeffi-ciency,increasedcontractioncycleduration,andweakenedrela-tionshipbetweencytoskeletalorganizationandforcegeneration.
FocalAdhesionsReinforceCell–CellJunctionsinStiffMicroenviron-ments.
Diseasedcardiactissueischaracterizedbyremodelingofcell-ECMandcell–celladhesions(3–7)anddecreasedtissuecompliance(33).
Weaskedifstiffmicroenvironmentsaffectthebalancebetweencell–cellandcell-matrixadhesions,asreportedfornoncardiaccells(30–32).
Onstiffsubstrates,weoftenob-servedtractionforceandfocaladhesionsnearthecell–cellinter-face(Fig.
3C),similartoDay1μtissues(Fig.
2B),suggestingthatμtissuesonstiffsubstrateshaveincreasedfocaladhesionforma-tionnearthecell–cellinterface.
Toquantifydifferencesinforcetransmissiontothecell–celljunctionrelativetotheECMforagivenμtissue,wecalculatedtheratioofFx;J;celltoNxx;cellatpeaksystoleforeachmyocyteandaveragedtogetherFx;J;CellNxx;Cellfrombothmyocytes(Fx;J;cellNxx;cellFx;J;Cell1Nxx;Cell1Fx;J;Cell2Nxx;Cell2∕2).
Onmoderatesubstrates,peaksystolicFx;J;cellNxx;cellsignificantlyincreasedbetweenμLW2–8and8–12(Fig.
4E),indicatingthatμtissueswithhighμLWtransmitrelativelymoreforcetothecell–celljunctionthantheECM.
ForallμLW,peaksystolicFx;J;cellNxx;cellwasnotstatisticallydifferentbetweensoftandmoderatesubstrates.
However,peaksystolicFx;J;cellNxx;cellwassignificantlyloweronstiffcomparedtomoderatesub-stratesatμLW8–12,indicatingdecreasedforcetransmissiontothecell–celljunctionrelativetotheECM.
Onstiffsubstrates,peaksystolicFx;J;cellNxx;cellwassimilarbetweenμLW2–8and8–12,indicatingthatthisparameterisnotregulatedbyμLWonstiffsubstrates.
Ourresultssuggestthat,instiffmicroenvironments,cardiacmyocytesgeneratemoreforce,whichexcessivelyloadstheintercalateddisc.
ThistriggersfocaladhesionassemblyFig.
3.
Regulationofdisplacement,tractionstress,andfocaladhesionformationbysubstratestiffness.
Displacementvectormaps(i),tractionstressvectormaps(ii),andpaxillinimmunostains(iii)forrepresentativeμtissueswithμLW8on(A)soft(1kPa),(B)moderate(13kPa),and(C)stiff(90kPa)substrates(C).
ForA–C,iandiiarefromthesameμtissueandiiiisfromadifferentμtissue.
Blue:nuclei.
(Scalebars:10μm.
).
0.
030.
33SoftModerateStiffLW246810120369Max.
|σx,tissue|(kPa)Max.
|xtissue|(m)StiffModerateSoftSoftMod.
StiffDurationofContraction(msec)ABCD0.
51.
10.
90.
7Nxx,tissueFx,J,cell1Fx,J,cell2WorkPeakSyst.
Force(N)PeakSyst.
Work(pJ)00.
51.
01.
5PeakSyst.
Nxx,tissue(N)PeakSyst.
Fx,J,cell(N)/Nxx,cell(N)E2-88-12LWSoftModerateStiffSoftMod.
Stiff01002001505000.
40.
81.
2051015Fig.
4.
Substratestiffnessregulatescontractilefunctionandbalanceofcell–cellandcell-matrixadhesion.
(A)Magnitudesofmaximumstressvectorsinthex-direction(max:jσx;μtissuej)plottedagainstmagnitudesofmaximumdis-placementvectorsinthex-direction(max:jΔxμtissuej)formultipleμtissuesdur-ingcontractioncycles.
(B)AveragepeaksystolicNxx;μtissue,Fx;J;cell1,Fx;J;cell2,andworkplottedforμtissuesoneachstiffness.
Barsstandarderror;*p<0.
05relativetomoderatesubstrates;n≥11μtissuesforeachstiffness.
(C)Averagedurationofcontractionforμtissuesoneachstiffness.
Barsstandarderror;*p<0.
05;n≥10foreachstiffness.
(D)PeaksystolicNxx;μtissueplottedasafunctionofμLWforμtissuesgroupedbyμLW2–4(soft:n4,moderate:n5,stiff:n4),4–8(soft:n4,moderate:n6,stiff:n3),and8–12(soft:n2,moderate:n9,stiff:n5).
Barsstandarderror;*p<0.
05relativetoμLW8–12forthesamestiffness.
(E)AveragepeaksystolicFx;J;cell∕Nxx;cellforμtissuesgroupedbyμLW2–8(soft:n8,moderate:n11,stiff:n7)and8–12(soft:n2,moderate:n9,stiff:n6).
Barsstandarderror;*p<0.
05.
9884∣www.
pnas.
org/cgi/doi/10.
1073/pnas.
1203007109McCainetal.
DownloadedbyguestonJanuary22,2021adjacenttothecell–cellinterfaceasamaladaptivemeansofstructurallyreinforcingthecell–celljunction,withthepotentialtodisruptmechanotransductionwithinthetissue.
DiscussionRemodelingofcell-matrixandcell–celladhesionsisimportantduringcardiacdevelopmentanddisease.
Here,weengineeredcardiacμtissuesinvitrotocharacterizethedynamicorganizationandfunctionalremodelingofmyofibrils,cell–cell,andcell-matrixadhesionsduringmorphogenesisandpathogenesis.
Ourhypoth-esisfordevelopmentalandpathologicalregulationofadhesionsisillustratedinFig.
5.
Atearlystagesoftissueformation,weob-servedfocaladhesiondisassemblynearthecell–cellinterfacecoincidentwithmaturationoftheintercalateddiscandtransmis-sionofsystolicforcesacrossthecell–cellboundary,suggestingthatmature,healthycardiactissuespreferentiallyformandmaintaincell–celljunctionsinsteadoffocaladhesionsattheirlongitudinalborders.
Onstiffsubstrates,weobservedfocaladhe-sionformationadjacenttothecell–cellinterface,suggestingthatfibroticmicroenvironmentsdestabilizecell–celladhesionduetoincreasedforcegeneration,whichstimulatesfocaladhesionfor-mation.
Intheheart,cell-matrixadhesionsanchormaturingcellsandtissuestotheECM(14)anddirectcellmigrationandmorpho-genesis(15–17);however,robustcell–celladhesioniscriticaltotheprimaryfunctionofmanytissues,includingbarrierfunctionofbloodvesselendothelium(50)andformationofanelectromecha-nicalsyncytiumintheheart(9).
Ourmodelsystemrecapitulatesinvivoobservationsthatfocaladhesionsdecreaseascell–celljunctionsincreaseatlongitudinalmyocytebordersduringcardiacdevelopment(10–13).
Ourdatacombinedwithpreviousreportssuggeststhatcell-matrixadhesionsarecriticalearlyindevelop-mentbeforecell–celladhesionsarefullymature;however,asde-velopmentprogresses,cell–celladhesionsareultimatelyfavoredatlongitudinalmyocyteborders,suggestiveofadevelopmentallyregulatedhierarchyofadhesionthatdrivestheassemblyandmaintenanceofwell-coupledtissues.
Innoncardiacpreparations,increasesinsubstratestiffnesspromotecell–celldecouplingandfocaladhesionformation(30–32).
Similarly,manycardiomyopathies,suchasmyocardialinfarction,arecharacterizedbyincreasedfibrosisandtissuestif-fening(33,34),increasedintegrinexpression(6,7),anduncou-plingbetweenmyocytes(3–5),indicatingthatmyocytesincreasematrixadhesionwhiledecreasingcell–celladhesion.
Ourresultssuggestthatstiffness-mediatedincreasesinforcegenerationnecessitatecompensatoryassemblyoffocaladhesionsadjacenttotheintercalateddisctostructurallyreinforcethejunction,whichisconsistentwithreportsthatcadherin–cadherinadhesionshavelowerfailurestrengthsthanintegrin-ECMadhesions(51–53).
Onstiffsubstrates,severalfunctionalparametersofμtissuesdeviatedfromthoseonphysiologicalsubstrates,providingnewinsightintostiffness-dependentmaladaptiveremodeling.
First,becauseμtissuesonstiffsubstratesgeneratedmoreforcetoattainsimilarmagnitudesofworkasonmoderatesubstrates,energydemandsinstiffmicroenvironmentsarelikelygreater,whichcouldcontributetometabolicremodelingthataccompaniescar-diomyopathiessuchaspathologicalhypertrophy(54).
Second,contractioncycledurationincreasedonstiffsubstrates,similartoobservationsinagingmyocardium(45,46),suggestingthatdecreasedtissuecomplianceissufficienttodisruptcontractioncycledynamicsandcouldcontributetoalteredforce-frequencyrelationships(47).
Finally,theaveragepeaksystoliclongitudinaltensiongeneratedbyμtissueswasuncorrelatedtotissuegeometryonlyonstiffsubstrates,suggestingthatstructuralremodelingofmyocyteshape,whichisanimportantcompensatorymechanism,isadaptiveonlyincellularmicroenvironmentswithmechanicalcompliancescorrespondingtothehealthyheart.
Insimilarstudies,thecell–celljunctionforceexertedbynonmyocytesrangedfrom20to200nN(27,40).
Wemeasuredpeaksystolicjunctionforcesupto1,000nNincardiacμtissuesonphysiologicalsubstrates,suggestingthatcardiacintercalateddiscsareexposedtosubstantiallyhigherforcesthancell–celljunctionsinotherorgans,potentiallyexplainingwhytheheartisespeciallyvulnerabletomutationsofcell–celladhesionproteins,suchasdesmoplakin(55–57).
Beyondplayingaroleintissueassembly,mechanicalforcesactivatemechanotransductionandsignalingatadhesionsites(22–26),indicatingthatcellularphysiologycouldbeaffectedbyalteredforcetransmission.
Furthermore,becausecell–celljunctionsarecriticalforgapjunctionformation(58),stiffness-dependentremodelingofcell–celladhesioncouldcon-tributetogapjunctionredistributionandarrhythmogenesisasso-ciatedwithmanycardiomyopathies(3–5,8).
Insummary,theseresultssuggestthatcell-matrixadhesionsareimportantfordirect-ingmyofibrillogenesisduringearlydevelopmentbutareprogres-sivelyreplacedwithcell–celladhesionsatlongitudinalmyocytebordersasthetissuematuresandpreferentiallytransmitssystolicloadsintercellularly.
Thisfunctionalhierarchycanbeperturbedindiseaseduetoincreasedstiffnessofthemicroenvironment,withfocaladhesionassemblynearthecell–cellinterfaceindica-tiveofacompensatoryeffortbythemyocytetomaintainthestructuralintegrityofthetissue.
MaterialsandMethodsExperimentalmethodsaredescribedindetailintheSIMaterialsandMethods;abriefdescriptionisincludedhere.
MicropatterningofPolyacrylamideGels.
Fibronectinwascross-linkedwithbio-tinusingSulfo-NHS-LC-Biotin(Pierce).
Polyacrylamidegelsubstrateswerefabricatedwiththefollowingelasticmoduliandacrylamide/bisconcentra-tions:1kPa:5∕0.
1%;13kPa:7.
5∕0.
3%;90kPa:12∕0.
6%(59).
Elasticmoduliwereverifiedwithatomicforcemicroscopy(60)(Fig.
S5AandB).
Streptavi-din-acrylamideand200nmfluorescentbeadswereaddedtothegelsolutionforafinalconcentrationof1∶5and1∶100,respectively,byvolume.
Polyacry-lamidegelswerecuredonactivated25mmcoverslipsandmicrocontactprinted(61)withbiotinylatedFN(62)aftercarefullydryingthegelsurfacebyincubatingat37°Cfor10min(Fig.
S5C).
CellCulture.
AllprocedureswereconductedaccordingtotheguidelinesoftheHarvardUniversityAnimalCareandUseCommittee.
Ventricularmyo-cytesfrom2-d-oldSprague–Dawleyratheartswereisolatedandculturedusingpreviouslydescribedprotocols(37,39).
15;000–50;000cells∕cm2wereseededonsubstrates.
Epinephrine(0.
2μM)wasaddedforthefirstandlast24hinculturetomaintainspontaneousbeating.
TractionForceMicroscopy.
High-resolutionTFMwasusedtoimagebeaddis-placementinspontaneouslycontractingmyocytesculturedonmicropat-ternedpolyacrylamidegelsubstrates(37,38).
Followingtheexperiment,Fig.
5.
Hypothesizedschematicofcooperativecouplingofcell-matrixandcell–celladhesionsincardiacmuscle.
Duringdevelopment,focaladhesionsguidemigrationandcell–celljunctionsareassembling.
Inhealth,cell–celljunctionsdominateoverfocaladhesionsnearthecell–cellinterfaceandforcesaretransmittedintercellularly.
Indisease,whenthemicroenvironmentisstiff,focaladhesionsreassemblenearthecell–cellinterfacetostabilizeexcessivelyloadedcell–celljunctions.
McCainetal.
PNAS∣June19,2012∣vol.
109∣no.
25∣9885BIOPHYSICSANDCOMPUTATIONALBIOLOGYDownloadedbyguestonJanuary22,2021coverslipswerefixed,immunostainedforβ-catenin,andimagedtoidentifythecell–celljunction.
Methodsusedtoacquiredisplacementandtractionstressvectorsfrombeaddisplacementimageshavebeenpreviouslydescribed(37,38).
ThetractionstressfieldwascalculatedfromthedisplacementmapusingFouriertransformtractioncytometrymethods.
ThetechniquesusedtocalculateNxx;cell,~FJ;cell,andNxx;μtissuearedescribedindetailintheSIMaterialsandMethods.
Statistics.
Dataaredisplayedasmeanstandarderror.
Datawerestatisti-callyanalyzedusingstudent'st-test,withp<0.
05consideredsignificant.
ACKNOWLEDGMENTS.
TheauthorswishtothankProfessorAnnaGrosbergforhercommentsonthemanuscript,athoughtfulreviewerforasuggestionregardingtheanalysisofourdata,andtheHarvardCenterforNanoscaleSystemsforuseofcleanroomfacilities.
ThisworkwasfundedbytheAmericanHeartAssociationPredoctoralFellowship(0815729D),NationalInstitutesofHealth(1R01HL079126),NanoscaleScienceandEngineeringCentersupportedbytheNationalScienceFoundation(PHY-0117795),HarvardMaterialsResearchScienceandEngineeringCentersupportedbytheNationalScienceFoundation(DMR-0213805),WyssInstituteforBiologicallyInspiredEngineering,andHarvardSchoolofEngineeringandAppliedSciences.
1.
AnastasiG,etal.
(2009)Dystrophin-glycoproteincomplexandvinculin-talin-integrinsysteminhumanadultcardiacmuscle.
IntJMolMed23:149–159.
2.
GourdieRG,GreenCR,SeversNJ(1991)Gapjunctiondistributioninadultmammalianmyocardiumrevealedbyananti-peptideantibodyandlaserscanningconfocalmicro-scopy.
JCellSci99:41–55.
3.
MatsushitaT,etal.
(1999)Remodelingofcell–cellandcell–extracellularmatrixinter-actionsattheborderzoneofratmyocardialinfarcts.
CircRes85:1046–1055.
4.
KostinS,etal.
(2003)Gapjunctionremodelingandalteredconnexin43expressioninthefailinghumanheart.
MolCellBiochem242:135–144.
5.
SmithJH,GreenCR,PetersNS,RotheryS,SeversNJ(1991)Alteredpatternsofgapjunctiondistributioninischemicheartdisease.
Animmunohistochemicalstudyofhumanmyocardiumusinglaserscanningconfocalmicroscopy.
AmJPathol139:801–821.
6.
TerracioL,etal.
(1991)Expressionofcollagenbindingintegrinsduringcardiacdevel-opmentandhypertrophy.
CircRes68:734–744.
7.
HelingA,etal.
(2000)Increasedexpressionofcytoskeletal,linkage,andextracellularproteinsinfailinghumanmyocardium.
CircRes86:846–853.
8.
LiJ,PatelVV,RadiceGL(2006)Dysregulationofcelladhesionproteinsandcardiacarrhythmogenesis.
ClinMedRes4:42–52.
9.
KleberAG,RudyY(2004)Basicmechanismsofcardiacimpulsepropagationandassociatedarrhythmias.
PhysiolRev84:431–488.
10.
AngstBD,etal.
(1997)Dissociatedspatialpatterningofgapjunctionsandcelladhe-sionjunctionsduringpostnataldifferentiationofventricularmyocardium.
CircRes80:88–94.
11.
PetersNS,etal.
(1994)Spatiotemporalrelationbetweengapjunctionsandfasciaadherensjunctionsduringpostnataldevelopmentofhumanventricularmyocardium.
Circulation90:713–725.
12.
HirschyA,SchatzmannF,EhlerE,PerriardJC(2006)Establishmentofcardiaccytoarch-itectureinthedevelopingmouseheart.
DevBiol289:430–441.
13.
WuJC,SungHC,ChungTH,DePhilipRM(2002)RoleofN-cadherin-andintegrin-basedcostameresinthedevelopmentofratcardiomyocytes.
JCellBiochem84:717–724.
14.
BakerEL,ZamanMH(2010)Thebiomechanicalintegrin.
JBiomech43:38–44.
15.
ParkerKK,etal.
(2002)Directionalcontroloflamellipodiaextensionbyconstrainingcellshapeandorientingcelltractionalforces.
FASEBJ16:1195–1204.
16.
RidleyAJ,etal.
(2003)Cellmigration:Integratingsignalsfromfronttoback.
Science302:1704–1709.
17.
HuangS,BrangwynneCP,ParkerKK,IngberDE(2005)Symmetry-breakinginmam-maliancellcohortmigrationduringtissuepatternformation:Roleofrandom-walkpersistence.
CellMotilCytoskeleton61:201–213.
18.
HarrisTJ,TepassU(2010)Adherensjunctions:Frommoleculestomorphogenesis.
NatRevMolCellBiol11:502–514.
19.
WoodsJR,Jr,DandavinoA,BrinkmanCR,3rd,NuwayhidB,AssaliNS(1978)Cardiacoutputchangesduringneonatalgrowth.
AmJPhysiol234:H520–524.
20.
GrantDA(1999)Ventricularconstraintinthefetusandnewborn.
CanJCardiol15:95–104.
21.
GeigerB,SpatzJP,BershadskyAD(2009)Environmentalsensingthroughfocaladhe-sions.
NatRevMolCellBiol10:21–33.
22.
WangN,ButlerJP,IngberDE(1993)Mechanotransductionacrossthecellsurfaceandthroughthecytoskeleton.
Science260:1124–1127.
23.
IngberD(1991)Integrinsasmechanochemicaltransducers.
CurrOpinCellBiol3:841–848.
24.
BellinRM,etal.
(2009)Definingtheroleofsyndecan-4inmechanotransductionusingsurface-modificationapproaches.
ProcNatlAcadSciUSA106:22102–22107.
25.
ChopraA,TabdanovE,PatelH,JanmeyPA,KreshJY(2011)Cardiacmyocyteremodel-ingmediatedbyN-cadherin-dependentmechanosensing.
AmJPhysiolHeartCircPhy-siol300:H1252–1266.
26.
leDucQ,etal.
(2010)VinculinpotentiatesE-cadherinmechanosensingandisrecruitedtoactin-anchoredsiteswithinadherensjunctionsinamyosinII-dependentmanner.
JCellBiol189:1107–1115.
27.
LiuZ,etal.
(2010)Mechanicaltuggingforceregulatesthesizeofcell–celljunctions.
ProcNatlAcadSciUSA107:9944–9949.
28.
ZhuangJ,YamadaKA,SaffitzJE,KleberAG(2000)Pulsatilestretchremodelscell-to-cellcommunicationinculturedmyocytes.
CircRes87:316–322.
29.
BirukovKG,etal.
(2003)Magnitude-dependentregulationofpulmonaryendothelialcellbarrierfunctionbycyclicstretch.
AmJPhysiolLungCellMolPhysiol285:L785–797.
30.
KrishnanR,etal.
(2011)Substratestiffeningpromotesendothelialmonolayerdisrup-tionthroughenhancedphysicalforces.
AmJPhysiolCellPhysiol300:C146–154.
31.
KimJH,AsthagiriAR(2011)MatrixstiffeningsensitizesepithelialcellstoEGFandenablesthelossofcontactinhibitionofproliferation.
JCellSci124:1280–1287.
32.
GuoWH,FreyMT,BurnhamNA,WangYL(2006)Substraterigidityregulatestheformationandmaintenanceoftissues.
BiophysJ90:2213–2220.
33.
BerryMF,etal.
(2006)Mesenchymalstemcellinjectionaftermyocardialinfarctionimprovesmyocardialcompliance.
AmJPhysiolHeartCircPhysiol290:H2196–2203.
34.
FarhadianF,etal.
(1995)Fibronectinexpressionduringphysiologicalandpathologicalcardiacgrowth.
JMolCellCardiol27:981–990.
35.
EnglerAJ,etal.
(2008)Embryoniccardiomyocytesbeatbestonamatrixwithheart-likeelasticity:Scar-likerigidityinhibitsbeating.
JCellSci121:3794–3802.
36.
PedrottyDM,etal.
(2008)Structuralcouplingofcardiomyocytesandnoncardiomyo-cytes:Quantitativecomparisonsusinganovelmicropatternedcellpairassay.
AmJPhy-siolHeartCircPhysiol295:H390–400.
37.
GrosbergA,etal.
(2011)Self-organizationofmusclecellstructureandfunction.
PLoSComputBiol7:e1001088.
38.
ButlerJP,Tolic-NorrelykkeIM,FabryB,FredbergJJ(2002)Tractionfields,moments,andstrainenergythatcellsexertontheirsurroundings.
AmJPhysiolCellPhysiol282:C595–605.
39.
McCainML,etal.
(2012)Cell-to-cellcouplinginengineeredpairsofratventricularcardiomyocytes:RelationbetweenCx43immunofluorescenceandintercellularelectri-calconductance.
AmJPhysiolHeartCircPhysiol302:H443–450.
40.
MaruthamuthuV,SabassB,SchwarzUS,GardelML(2011)Cell-ECMtractionforcemodulatesendogenoustensionatcell–cellcontacts.
ProcNatlAcadSciUSA108:4708–4713.
41.
JacotJG,McCullochAD,OmensJH(2008)Substratestiffnessaffectsthefunctionalmaturationofneonatalratventricularmyocytes.
BiophysJ95:3479–3487.
42.
ZhuJ,SabharwalT,KalyanasundaramA,GuoL,WangG(2009)Topographicmappingandcompressionelasticityanalysisofskinnedcardiacmusclefibersinvitrowithatomicforcemicroscopyandnanoindentation.
JBiomech42:2143–2150.
43.
ShielsHA,WhiteE(2008)TheFrank–Starlingmechanisminvertebratecardiacmyo-cytes.
JExpBiol211:2005–2013.
44.
ProsserBL,WardCW,LedererWJ(2011)X-ROSsignaling:Rapidmechano-chemotrans-ductioninheart.
Science333:1440–1445.
45.
HarrisonTR,DixonK,RussellRO,Jr,BidwaiPS,ColemanHN(1964)Therelationofagetothedurationofcontraction,ejection,andrelaxationofthenormalhumanheart.
AmHeartJ67:189–199.
46.
LakattaEG,GerstenblithG,AngellCS,ShockNW,WeisfeldtML(1975)Prolongedcon-tractiondurationinagedmyocardium.
JClinInvest55:61–68.
47.
KushnirA,ShanJ,BetzenhauserMJ,ReikenS,MarksAR(2010)RoleofCaMKIIdeltaphosphorylationofthecardiacryanodinereceptorintheforcefrequencyrelationshipandheartfailure.
ProcNatlAcadSciUSA107:10274–10279.
48.
GerdesAM(2002)Cardiacmyocyteremodelinginhypertrophyandprogressiontofailure.
JCardFail8:S264–S268.
49.
McCainML,ParkerKK(2011)Mechanotransduction:Theroleofmechanicalstress,myocyteshape,andcytoskeletalarchitectureoncardiacfunction.
PflugersArch462:89–104.
50.
DejanaE,Tournier-LasserveE,WeinsteinBM(2009)Thecontrolofvascularintegritybyendothelialcelljunctions:Molecularbasisandpathologicalimplications.
DevCell16:209–221.
51.
BaumgartnerW,etal.
(2000)Cadherininteractionprobedbyatomicforcemicroscopy.
ProcNatlAcadSciUSA97:4005–4010.
52.
KokkoliE,OchsenhirtSE,TirrellM(2004)Collectiveandsingle-moleculeinteractionsofalpha5beta1integrins.
Langmuir20:2397–2404.
53.
ZhangY,SivasankarS,NelsonWJ,ChuS(2009)Resolvingcadherininteractionsandbindingcooperativityatthesingle-moleculelevel.
ProcNatlAcadSciUSA106:109–114.
54.
SheehySP,HuangS,ParkerKK(2009)Time-warpedcomparisonofgeneexpressioninadaptiveandmaladaptivecardiachypertrophy.
CircCardiovascGenet2:116–124.
55.
OlsonTM,etal.
(2002)Metavinculinmutationsalteractininteractionindilatedcardiomyopathy.
Circulation105:431–437.
56.
DelmarM,McKennaWJ(2010)Thecardiacdesmosomeandarrhythmogeniccardio-myopathies:Fromgenetodisease.
CircRes107:700–714.
57.
AsimakiA,etal.
(2009)Anewdiagnostictestforarrhythmogenicrightventricularcardiomyopathy.
NEnglJMed360:1075–1084.
58.
KostetskiiI,etal.
(2005)InduceddeletionoftheN-cadheringeneintheheartleadstodissolutionoftheintercalateddiscstructure.
CircRes96:346–354.
59.
Aratyn-SchausY,OakesPW,StrickerJ,WinterSP,GardelML(2010)Preparationofcomplaintmatricesforquantifyingcellularcontraction.
JVisExp46:e2173.
60.
RadmacherM,FritzM,HansmaPK(1995)Imagingsoftsampleswiththeatomicforcemicroscope:Gelatininwaterandpropanol.
BiophysJ69:264–270.
61.
ChenCS,MrksichM,HuangS,WhitesidesGM,IngberDE(1997)Geometriccontrolofcelllifeanddeath.
Science276:1425–1428.
62.
HyndMR,FramptonJP,Dowell-MesfinN,TurnerJN,ShainW(2007)Directedcellgrowthonprotein-functionalizedhydrogelsurfaces.
JNeurosciMethods162:255–263.
9886∣www.
pnas.
org/cgi/doi/10.
1073/pnas.
1203007109McCainetal.
DownloadedbyguestonJanuary22,2021