A wireless centrifuge force microscope (CFM) enables ... · A wireless centrifuge force microscope (CFM) enables multiplexed single ... module is properly balanced, 2) ... enables

Awirelesscentrifugeforcemicroscope(CFM)enablesmultiplexed

single-moleculeexperimentsinacommercialcentrifuge

TonyHoang1,DhruvS.Patel2,KenHalvorsen3,*

1DepartmentofChemistry,2DepartmentofBiology,&3TheRNAInstitute,UniversityatAlbany,StateUniversityofNewYork,1400WashingtonAvenue,Albany,NewYork,12222,

UnitedStates

*Correspondence:[email protected]:Thecentrifugeforcemicroscope(CFM)wasrecentlyintroducedasaplatformformassivelyparallelsingle-moleculemanipulationandanalysis.Herewedevelopedalow-costandself-containedCFMmodulethatworksdirectlywithinacommercialcentrifuge,greatlyimprovingaccessibilityandeaseofuse.Ourinstrumentincorporatesresearchgradevideomicroscopy,apowersource,acomputer,andwirelesstransmissioncapabilitytosimultaneouslymonitormanyindividuallytetheredmicrospheres.Wevalidatedtheinstrumentbyperformingsingle-moleculeforceshearingofshortDNAduplexes.Fora7bpduplex,wecollectedover1000statisticsofforcedependentshearingkineticsfrom2pNto12pNwithdissociationtimesintherangeof10-100seconds.Weextendedthemeasurementtoa10bpduplex,applyinga12pNforceclampanddirectlyobservingsingle-moleculedissociationoveran85minuteexperiment.OurnewCFMmodulefacilitatessimpleandinexpensiveexperimentsthatdramaticallyimproveaccesstosingle-moleculeanalysis.Keywords:single-molecule,biophysics,centrifugeforcemicroscope,DNA

.CC-BY-NC-ND 4.0 International licensepeer-reviewed) is the author/funder. It is made available under aThe copyright holder for this preprint (which was not. http://dx.doi.org/10.1101/060269doi: bioRxiv preprint first posted online Jun. 22, 2016;

Introduction:Thepowerofsingle-moleculeexperimentshasbeenwellestablishedoverthelast

coupleofdecades,withvariousexperimentalplatforms(opticalandmagnetictweezers,AFM,andothers)enablingunparalleledinsightsintonanoscalebiologyandchemistry1-3.Unfortunately,singlemoleculesensitivityoftencomesatahighpriceintermsoftime,money,andexpertise.Thevastmajorityofsingle-moleculepullingexperimentsareperformedonemoleculeatatime(withsomenotableexceptions4-6),onsophisticatedequipmentthatoftenexceeds$100,000.Thesefactorsconspiretolowertheaccessibilityofsingle-moleculetechniquestothebiomedicalresearchersthatcouldarguablybenefitmostfromtheiruse.

Inpreviouswork,wepresentedanapproachtosolvethesechallengesbyusingcentrifugalforcetoperformhundredstothousandsofsimultaneoussingle-moleculepullingexperiments7.Inaninstrumentwecallthecentrifugeforcemicroscope(CFM),amicroscopeisrotatedsuchthattheopticalaxisisalignedwiththeforce,enablingobservationofmanymicrospherestetheredtoacoverglass.Ourfirstprototypewasessentiallyacustommadeopen-aircentrifugesupportingamicroscopeimagingarm(Figure1a).Whileimportantforestablishingproof-of-concept,theprototypehadshortcomingsincludingsafety,easeofuse,anddynamicrange.Morerecently,wedevelopedanewCFMthatintegratesintoamodifiedbenchtopcentrifuge(Figure1b),improvingsafety,reducingcost,andaddingfeatures8.

Inthiswork,weintroduceafullywirelessandself-containedCFMmodulethatenablesuseincommercialbenchtopcentrifuge(Figure1c).Thisadvancedramaticallyincreasesaccessibilitybyeliminatingtheneedforcentrifugemodification(andthusadedicatedcentrifuge),loweringthecost(below$2,000),andsimplifyingoperation.Wehaveengineeredhardwareandsoftwarenecessarytoremotelyperformvideomicroscopywithinarotatingcentrifuge.Toillustrateapplicationinsingle-moleculebiophysics,weappliedshearingforcestoshortDNAduplexesandobservedforceandbasepairdependentdissociation.Itisourhopethatthisworkfurtherdemocratizessingle-moleculeexperiments,enablingusebyabroadrangeofscientificresearchers.Furthermore,withcosteffectiveandstraightforwardreproductionofourCFMmodule,theinstrumentmayfindusesbeyondsingle-moleculebiophysicsinareassuchascolloidalphysics,biologicalseparations,andcelladhesion.

.CC-BY-NC-ND 4.0 International licensepeer-reviewed) is the author/funder. It is made available under aThe copyright holder for this preprint (which was not. http://dx.doi.org/10.1101/060269doi: bioRxiv preprint first posted online Jun. 22, 2016;

Figure1:EvolutionoftheCFM.a)Thefirstprototypeusedatablemountedrotarystagewithamicroscopeimagingarmandcounterweight.b)Thesecondgenerationinstrumentintegratedintoaswingingbucketcentrifugeandcommunicatedwithanexternalcomputerviaaconnectedfiberoptic,c)ThewirelessCFMpresentedhereiscompletelyselfcontainedwithonboardpower,computing,storage,andwirelessdatatransmission.ExperimentalSection:Instrument:TheCFMmoduleisdesignedtofitwithinacommercialswingingbucketcentrifuge(SorvallX1R)with400mLbuckets.Totheextentpossible,weusedcommerciallyavailablepartstofacilitatereproductionbyotherlabs.LabeledimagesandacompletepartslistaregivenintheSupportingInformation(FigureS1andTableS1).Briefly,theopticalassemblyismadefromaluminumlenstubes(Thorlabs)supportingthecamera(PointGreyFlea3),infinity-correctedobjective(OlympusPlanAchromat–seeNoteS1),tubelens(Thorlabsachromaticdoublet),turningmirrors(Thorlabs),sample,andlightdiffuser(Thorlabs).Thisopticalassemblyiscontainedwithinthe3Dprintedhousing(seeFileS1fordrawings),whichalsocontainstheLED,alithiumpolymerbattery,singleboardcomputer,andsupportingcircuitry(wiringdetailsinFigureS2).Themoduleiscounterbalancedbyquartersarrangedin3tiersfroma3Dprintedhousing(seeFileS1fordrawing).ThemassoftheCFMmoduleisbalancedbyusingtheappropriatenumberofquarters,whilethecenterofmassisbalancedbyrearrangingthepositioningofthequartersamong12possiblelocations.Sampleandmicrospherepreparation:DNAhandleswerepreparedbyhybridizing122oligos(IntegratedDNATechnologies)toagenomicsingle-strandedDNA,asdescribedinpreviouswork8,9.DNAhandleD-1wasconstructedwitha10nt3’overhanganda5’doublebiotin.DNAhandlesD-2,D-3,andD-4contained3’overhangs(10nt,7nt,and6nt,respectively)withreversecomplementaritytoDNAhandle1andalsocontaineda5’doublebiotin.AlistofoligosispresentedinTable

.CC-BY-NC-ND 4.0 International licensepeer-reviewed) is the author/funder. It is made available under aThe copyright holder for this preprint (which was not. http://dx.doi.org/10.1101/060269doi: bioRxiv preprint first posted online Jun. 22, 2016;

S-2,wherethe122oligosforeachDNAhandleiscomprisedof109“backbone”oligos,12“variable”oligos,and1“overhang”oligo.TopreparetheDNAcoatedbeads,wetook40µlofstreptavidinmicrospheres(ThermoFisherDynabeadsM-2702.8µmdiameter)andwashedtwicewith100µlofPBS+0.1%tween20usingamagnet(ThermoFisherDynaMagSpinMagnet).Followingthewash,thebeadswerebroughttoa10µlvolume,and10µlofthebiotinlabeledDNAD-1(~300pM)wasaddedandreactedfor1houronavortexertopreventsettlingofthebeads.Chamberassemblyandpreparation:Thesamplechamberwasmadebytwocoverglasses(WarnerInstruments)separatedbydoublesticktape(Kaptontape.com)andsupportedonaflatdiscwithacentralopening.TheprocessofbuildingthechamberisillustratedinFigureS3.Thepreparedchamberholds1-10µL.Topreparethechamberforsinglemoleculeexperiments,wefirstflowedin0.1mg/mLofStreptavidin(Amresco)inPBS,andallowedtoadsorbtothesurfacefor~1minute.ThechamberwasthenwashedwithPBS+0.1%tweenbyflowingthroughfourvolumesof50µLin30secondintervals.Next,5µLoftheDNAD-2,D-3,orD-4(~1.5nM)wasincubatedfor5minutestoallowbindingofthebiotinlabelstothestreptavidinsurface.Another4roundsof50µLPBS+0.1%tweenwasrepeatedtoremoveunreactedDNA.AtthispointtheDNAcoateddynabeadswereadded(typically5µL).Finally,thechamberwassealedwithhighvacuumgreasetopreventliquidescape.ExperimentalprotocolThepreparedandsealedchamberwassecurelyscrewedintothelowerlenstubeandthenscrewedintotherestoftheopticalassembly.Withthecameraoperatingfromthehostcomputer,wemanuallyadjustedthefocusbythreadingtowardtheobjectiveandlockingintoplaceusingalenstubecoupler.Oncesecure,the3DhousingwasclosedandthebatteryandthemicroUSBplugwereconnected.Atthispoint,theCFMmodulewaslightlytappedtoremovemicrospheresfromthecoverglassandtheninvertedtoallowcontrolofthemicrospheretouchtimeagainstthecoverglass.Followingthetouchtime,theCFMmodulewasinvertedback,andthenloadedintothecentrifugebucket.Therunwasstartedbyacommandlinescriptfromthehostcomputer(seeNoteS2andFileS2),andthecentrifugewascontrollednormallyfromthefrontpaneltocontrolspeed.Wetypicallystartedimagerecordingbeforestartingthecentrifuge.ThecentrifugespeedisrelatedtotheforceonthetethersbyF=mω2R,wheremisthemassofthebead(minusthemassofthemediumdisplaced),ωisthemagnitudeoftheangularvelocity,andRisthedistanceofthemicrospheresfromtheaxisofrotation.Themassofeachmicrosphereis6.9x10-12gaspreviouslydetermined7,8,andthedistanceRwasmeasuredtobe133mm.Usingthesevalues,weusedthefollowingspeeds:2pN=445RPM,6pN=770RPM,8pN=890RPM,10pN=995RPM,and12pN=1090RPM.Dataanalysis:Imageswereanalyzedbyfirstidentifyingtetheredbeadsinthefirstmovieframefollowingacceleration.Tetheredmicrosphereswerevisuallyselectedbasedonfocus(seeFigure3a).Doubletethers(thoughrare)wereeliminatedastheywereobserved,typicallyseenasa

.CC-BY-NC-ND 4.0 International licensepeer-reviewed) is the author/funder. It is made available under aThe copyright holder for this preprint (which was not. http://dx.doi.org/10.1101/060269doi: bioRxiv preprint first posted online Jun. 22, 2016;

lateralshiftwhenthefirsttetherbreaks.Tetheredmicrospheresweretrackedthroughoutsubsequentframesandthenumberofmicrospheresbeingreleasedateachframewasrecorded.Safetyconsiderations:Theseexperimentsinvolveatypicaluseofacentrifuge,whichcaninvolveenormouskineticenergy.Whilecommercialcentrifugesaregenerallydesignedtocontainanycatastrophicfailure,greatcaremustbetakentoensurethat:1)theCFMmoduleisproperlybalanced,2)thatthemassesorspeedsuseddonotexceedthoserecommendedbythemanufacturer,3)thatallcomponentsaresecuredwithinthebuckets.ResultsandDiscussion: WedevelopedawirelessCFMmoduletoimprovetheaccessibilityandreducethecostofmultiplexedsinglemoleculeexperiments.Themodulewasdesignedtobefullycompatiblewithexistingcommercialcentrifuges,andweusedtheSorvallX1Rswingingbucketcentrifugeasourmodelsystem(Figure2a).Followingasimilaropticaldesignasprevious8,weadditionallyimplementedanonboardcomputinganddatatransmissionstrategytoallowtheCFMtofitwithinasinglebucket,andtoenableusewithoutmodificationofthecentrifuge(Figure2band2c).WechosetousetheIntelEdisonasouronboardcomputerduetoitssmallsize,lowcost,andintegratedWiFicapabilities.WeinterfacedtheonboardcomputerwithasmallUSBmachinevisioncamera(PointGreyFlea3)anddevelopeda3Dprintedhousingtosecurethevariousopticalandelectricalcomponents(seeFileS1).TheresultingCFMmodulecanbebuiltfor~$1,000-$2,000dependingonspecificcameraandobjectivechoice(seeTableS1andNoteS1).Allpartsarecommerciallyavailable,withtheexceptionofonecustommachinedthreadadapterforwhichwehaveofferedalternatives(seeNoteS1),andthe3Dprintedhousing.However,wehavemadeavailablethedrawingfilesforthe3Dprinting,andcapable3Dprintersarenowwidelyavailable,evenatmanypubliclibraries. WeadditionallydevelopedsoftwareandprocedurestofacilitatethefabricationanduseoftheCFM(SeeSupplementalInformation).Briefly,theonboardcomputerwasloadedwithaLinuxbasedoperatingsystem,modifiedtoenableUSBtransmissionoflargefiles,andloadedwiththePointGreycameradrivers.Controloftheonboardcomputerwasthenperformedremotelyfromahostcomputer(SeeNoteS2).ThecommunicationschemefornormaloperationisoutlinedinFigure2d,requiringonly2stepsfromtheuser:1)connectingtheCFMwithitsbatterypowersource,2)runningaprogramfromthehostcomputercommandlinedictatingthenumberofimagesandframerate.TheseinstructionswerewirelesslyreceivedbytheCFMandexecuted.Thecentrifugewascontrolledthroughthefrontpaneltocontrolspeed,acceleration,temperature,andtime.Attheendoftherun,theimagefileswerewirelesslytransmittedbacktothehostmachineanderasedfromtheonboardmemory. Usingthissetup,werecordedimagesof2080x1552pixelstoaccommodate>2000microspheres(2.8µmdiameter)witha40xobjectiveand>8000fora20xobjective(Figure2e).Weusedapixeldepthof8bits,yieldinganimagesize~3.2MBwiththefirst4pixelsencodingthecamerageneratedtimestamp.Theonboardcomputercontains4GBofbuiltin(eMMC)storage,whichleaves1-2GB,orabout300-600frames,offreememoryforfilestorage.WefoundthatsavingimagestoRAMandlatersavingthemtotheeMMCresulted

.CC-BY-NC-ND 4.0 International licensepeer-reviewed) is the author/funder. It is made available under aThe copyright holder for this preprint (which was not. http://dx.doi.org/10.1101/060269doi: bioRxiv preprint first posted online Jun. 22, 2016;

infastermaximumframerates(upto10fpsatfullresolutionwithoutskippedframes)duetolesscomputationaloverhead,butreducedthetotalnumberofframesto~250duetothelowerstoragecapacity(1GB).Forframeratesslowerthan1fps,wegrabbed1imageatatimewithapauseinsoftware,achievingruntimesapproaching1.5hoursonasinglebattery.ThepostexperimenttransferislargelydictatedbyWiFispeeds,andinourhandswefoundatransferspeed~5MB/s(1.4fpsonaverage).

Figure2:Instrumentdesignandvalidation.a)Ourmodelbenchtopcentrifuge(SorvallX1series),bandc)TheCFMmoduleisdesignedtofitinaswingingbucket,andcontainsaminiaturemicroscope,abattery,andacomputerwithwificapabilities,d)FlowchartforCFMcontrolsoftware,e)260x194pixelimagesfromtheCFMmoduleat40xand20xmagnification,representing1/64thofthefullframesize.

Todemonstratesingle-moleculemanipulationonthewirelessCFMmodule,we

performedexperimentstoshearshortDNAduplexes.WeconstructeddoublebiotinlabeledDNAhandleswithshortcomplementaryoverhangs.ThetwoDNAhandleswereseparatelyreactedwithstreptavidinbeadsandastreptavidincoatedcoverglasstoenablesinglemoleculetethersconnectedbyashortduplex(Figure3a).WemadebeadswithvariousdilutionsoftheDNAuntilwefoundconditionsthatconsistentlyproduced<15%attachmentfrequency,whichisdesiredtominimizemultipletethers.UsingthisarrangementweincubatedDNAloadedbeadsagainstthesurfacefor~1minutetoenableformationoftheduplex,andthenstartedthecentrifugetoattainaconstantdesiredvelocity.

Wefirstperformedconstantforcepullingexperimentsona7bpduplex.Asexpected,wefoundthedistributionofrupturetimestofollowanexponentialdecay(Figure3b).Withthelargenumberofstatistics,wewereadditionallyabletoindependentlyanalyzemultiplerunsunderthesameconditionstoassessrepeatability(Figure3binset).Fromfittingfiveindividualdistributions,wefoundameandissociationtimeconstantfor12pNof14.8secondswithastandarddeviationof1.8seconds.Asacomparison,fittingthecumulativedistributionofthecombineddatagaveatimeconstantof14.9±0.1seconds.

Weexpandedthedatabyrepeatingthemeasurementatdifferentforcelevelsovertherangefrom2pNto12pN,withtimeconstantsincreasingforsmallerforces(Figure3c).

.CC-BY-NC-ND 4.0 International licensepeer-reviewed) is the author/funder. It is made available under aThe copyright holder for this preprint (which was not. http://dx.doi.org/10.1101/060269doi: bioRxiv preprint first posted online Jun. 22, 2016;

WefittheseresultswiththeBell-Evansmodelforforceactivateddissociation10,11wheretheforcedependentoffrateisdescribedaskoff(f)=koffexp(f/fβ),wherekoffisthethermaloff-rate,fistheappliedforce,andfβisthecharacteristicforcescaleoftheinteraction.fβcanequivalentlybeexpressedaskBT/x,wherekBistheBoltzmannconstant,Tistheabsolutetemperature,andxisthelengthscaledescribingtheseparationoftheboundandunboundstates.Usingthismodel,wefoundathermaloff-ratekoffof0.004±0.002s-1andaforcescalefβof4.5±1.0pN.

Figure3:ForceshearingofshortDNAduplexes.a)TheexperimentalsetupconsistsofmicrospherestetheredtoacoverglassbybiotinlabeledDNAhandles,withashortcentralduplexdissociatingunderappliedcentrifugalforce.b)Forthe7bpduplexundera12pNforceclamp,wecollectedover600statisticsandfoundadissociationtimeconstantof14.9seconds.Byindependentlyanalyzingthefiveindependentrunscomprisingthedata,wefoundgoodrepeatabilityandestimatedanuncertaintyof~15%inthetimeconstant(inset).c)Forcedependentdissociationofthe7bpduplex(squares)showedincreasingoff-rateasafunctionofforce,withafittothedata(line)providingasolutionoffrateof0.004s-1andaforcescaleof4.5pN.Acomparisonwitha10bpduplexshowsadramaticallysloweroff-rate(triangle).Individualdataforeachforceisshownintheinset.Errorbarsweredeterminedbasedonthestandarddeviationofmultipleindependentexperiments.

Weadditionallyinvestigateda6bpduplexanda10bpduplexataconstantforceof12pN.Asexpected,weobservedmarkedlyslowerdissociationforthe10bpduplex,observingdissociationoveran85minuteexperimenttodetermineatimeconstantof28±1minutes.Inattemptingtomeasurethe6bpduplex,wedidnotobservesignificanttethersoverseveralruns,presumablybecausethedissociationtimescaledecreasedtoodramatically.Overthisrelativelynarrowrangeoflengths,theforcescaledoesnotchangedramatically12,sothatour10bpoffratecanreasonablybeextrapolatedtoestimatethermaloff-rate.Basedonthis,weestimate~4x10-5s-1,whichcorrespondstoadissociationtimeconstantof~7hours.

Incomparingourresultswithothers,wenotethatmostifnotallDNAshearingexperimentshaveusedduplexes10bporlarger12-14,andonehasevenpredictedthat8bporlesswillnotbemeasureable15.Thesequenceofour10bpduplexwaschosentomatchpreviouswork12,whiletheshorterduplexesweretruncatedfromthatsequence.Incomparisontothatwork,wefoundaforcescaleforthe7bpduplexthatagreeswithinerror,butouroff-ratesarenotablyslowerthanthosepredictedinthatwork.Forthe7bpand10bpduplexes,theypredictoffratesof0.3s-1and0.01s-1,respectively,whichareapproximatelytwoordersofmagnitudefasterthanourmeasurements.

Thereareafewfactorsthatlikelycontributetothisdifference.Our10bpduplex,whileofidenticalsequence,hasflankingbasesonbothsidesthatcontributebasestackinginteractions.CalculationsusingMfold16estimatethatthesestackinginteractionscontribute

.CC-BY-NC-ND 4.0 International licensepeer-reviewed) is the author/funder. It is made available under aThe copyright holder for this preprint (which was not. http://dx.doi.org/10.1101/060269doi: bioRxiv preprint first posted online Jun. 22, 2016;

anadditional1.5to2.7kcal/moltothestabilityofthe10bpduplex,suggestinganoff-ratethatshouldbeonetotwoordersofmagnitudeslower.Evenusingtheconservative(1.5kcal/mol)adjustmentwouldputourmeasurementswithinerroroftheirs12.Similarcorrectionswouldapplyforourshorterduplexesaswell.Another(lesssignificant)contributiontotheobserveddifferencemaybeduetoanoverestimationoftheoff-rateinref[12].Theyapproximatedthattheforce-loadingratewasdictatedbythetetherelasticityatrupture,whichoverestimatestheloadingrate(andthustheoff-rate)sincemuchoftheexperimentaltimeisspentreachingthatloadingrate.

Conclusions: Aswehaveshownhere,thewirelessCFMmoduleiscapableofmultiplexedsingle-moleculeexperimentswiththroughputthatcomparesfavorablywithsomeofthehighestthroughputmethods4,5.Ourmodulardesignallowsunprecedentedaccessibilitytosingle-moleculeexperimentsbyleveragingcommonbenchtopcentrifugeswithbucketsof400mLorlarger.Thecostofourmoduleisonetotwoordersofmagnitudelessexpensivethanmanycompetingsingle-moleculetechnologies,andupto100,000timescheaperifoneconsiderscostperpullcomparedtoopticaltweezers,forexample.ThesimplicityofoperatingourCFMmodulemakesitamenabletousebyawiderangeofresearchersincludingundergraduates(suchasoneauthorofthispaper).Herewehaveprovidedallofthenecessarydesigns,instructions,andsoftwaretomakereproductionofthissystemstraightforward. Thecentrifugeforcemicroscopeisstillaninstrumentinitsinfancy,buthasalreadyundergonesignificantimprovementhereandinpreviouswork.Still,furtherdevelopmentworkisneededtoimprovesomeofitscurrentlimitations.Integrationofmotorizedcomponentstocontrolstagepositioningin3-axeswouldbenefitsamplesetup,focusing,andtracking.Flowcontrolwouldenablechangingbuffersandreagentsduringexperimentsformorecomplexexperiments.Furtherimprovementsinbatteries,singlechipcomputers,andWiFiprotocolswillenablelongerexperiments,fasterdatatransmission,andreal-timedataanalysis. WhiletheCFMisunlikelytosupplantexistingsingle-moleculetechnologies,webelieveitoccupiesanimportantnicheformultiplexedsinglemoleculeexperimentswhereperhapslowcostandeaseofusearemoreimportantthansub-nanometerandsub-millisecondresolutions.Withthiswork,itisourhopethatthisCFMmodulewillencouragemorewidespreadsingle-moleculeexperimentationbybiomedicalresearchers,andleadtonewandexcitingavenuesofresearch.AcknowledgementsTheauthorsthankAlanChenandPanT.X.LifortheirfinancialsupportandmentorshiproleforT.H.WethankEricWarnkeforvitalsoftware,hardware,andprogrammingsupport,ArunChandrasekaranforpreparingDNAhandles,andAndreasKarlandThermoFisherScientificforprovidingmaterialsandsupportforthecentrifuge.WeacknowledgetheSUNYSTEMResearchPassportprogramforfinancialsupportofD.P.,andCombinedFundingApplication(UniversityatAlbany–RNAInstituteCapitalProject#X765)fromtheStateofNewYork.WethankWesleyWong,DarrenYang,MariaBasantaSanchez,andSriRanganathanforusefuldiscussions.

.CC-BY-NC-ND 4.0 International licensepeer-reviewed) is the author/funder. It is made available under aThe copyright holder for this preprint (which was not. http://dx.doi.org/10.1101/060269doi: bioRxiv preprint first posted online Jun. 22, 2016;

AuthorContributionsTHdesigned,built,andprogrammedtheinstrumentandperformedexperiments.DSPdevelopedprotocolsandperformedexperiments.KHconceivedtheproject,analyzedthedata,andwrotethepaper.CompetingFinancialInterestsKHandTHhavepatentspendingonaspectsofthiswork.References:1. NeumanKC,NagyA.Single-moleculeforcespectroscopy:opticaltweezers,magnetictweezersandatomicforcemicroscopy.NatMethods5,491-505(2008).2. RitortF.Single-moleculeexperimentsinbiologicalphysics:methodsandapplications.JPhys-CondensMat18,R531-R583(2006).3. Deniz,A.A.,S.Mukhopadhyay,andE.A.Lemke.2008.Single-moleculebiophysics:attheinterfaceofbiology,physicsandchemistry.J.R.Soc.Interface.5:15–45.4. SittersG,KamsmaD,ThalhammerG,Ritsch-MarteM,PetermanEJG,WuiteGJL.Acousticforcespectroscopy.NatMethods12,47-50(2015).5. DeVlaminckI,etal.HighlyParallelMagneticTweezersbyTargetedDNATethering.NanoLetters11,5489-5493(2011).6. FazioT,VisnapuuML,WindS,GreeneEC.DNAcurtainsandnanoscalecurtainrods:High-throughputtoolsforsinglemoleculeimaging.Langmuir24,10524-10531(2008).7. HalvorsenK,WongWP.Massivelyparallelsingle-moleculemanipulationusingcentrifugalforce.Biophysicaljournal98,L53-L55(2010).8. YangD,WardA,HalvorsenK,WongWP.Multiplexedsingle-moleculeforcespectroscopyusingacentrifuge.Nat.Commun.7:11026doi:10.1038/ncomms11026(2016).9. HalvorsenK,SchaakD,WongWP.Nanoengineeringasingle-moleculeswitchusingDNAself-assembly.Nanotechnology22:494005(2011).10. Bell,GI.Modelsforthespecificadhesionofcellstocells.Science200:618–627(1978).11. EvansE,RitchieK.Dynamicstrengthofmolecularadhesionbonds.Biophys.J.72:1541–1555(1997).12. StrunzT,OroszlanK,SchäferR,GüntherodtHJ.DynamicforcespectroscopyofsingleDNAmolecules.ProceedingsoftheNationalAcademyofSciencesoftheUnitedStatesofAmerica,96(20),11277–11282(1999).

.CC-BY-NC-ND 4.0 International licensepeer-reviewed) is the author/funder. It is made available under aThe copyright holder for this preprint (which was not. http://dx.doi.org/10.1101/060269doi: bioRxiv preprint first posted online Jun. 22, 2016;

13. PopeLH,DaviesMC,LaughtonCA,RobertsCJ,TendlerSJ,WilliamsPM.Force-inducedmeltingofashortDNAdoublehelix.Eur.Biophys.J.,30,53–62(2001).14. SchumakovitchI,GrangeW,StrunzT,BertonciniP,GuntherodtHJ,HegnerM.TemperaturedependenceofunbindingforcesbetweencomplementaryDNAstrands.Biophys.J.,82,517–521(2002).15. SattinBD,PellingAE,GohMC.DNAbasepairresolutionbysinglemoleculeforcespectroscopy.NucleicAcidsResearch,32,4876-4883(2004).16. ZukerM.Mfoldwebserverfornucleicacidfoldingandhybridizationprediction.NucleicAcidsResearch,31,3406-3415(2003).

.CC-BY-NC-ND 4.0 International licensepeer-reviewed) is the author/funder. It is made available under aThe copyright holder for this preprint (which was not. http://dx.doi.org/10.1101/060269doi: bioRxiv preprint first posted online Jun. 22, 2016;