transport efficiency of membrane-anchored kinesin-1 … · however, the collective transport by...

1

Classification:Biologicalscience,BiophysicsandcomputationalbiologyTitle:Transportefficiencyofmembrane-anchoredkinesin-1motorsdependsonmotordensityanddiffusivityRahulGrovera,b,JanineFischerb,FriedrichW.Schwarza,c,WilhelmJ.Waltera,d,PetraSchwillee,StefanDieza,b,c,1aCenterforMolecularBioengineering(BCUBE),TechnischeUniversitätDresden,01069Dresden,GermanybMaxPlanckInstituteofMolecularCellBiologyandGenetics,01307Dresden,Germanyc Center for Advancing Electronics Dresden (cfaed), Technische Universität Dresden, 01069Dresden,GermanydCurrentaddress:BiozentrumKleinFlottbek,UniversitätHamburg,22609Hamburg,GermanyeMaxPlanckInstituteofBiochemistry,DepartmentofCellularandMolecularBiophysics,82152Martinsried,GermanyFootnotesAuthorcontributions:RG,PS,andSDdesignedresearch;RGandJFperformedresearch;RGandFSanalyzeddata;WJWcontributedreagents;RGandSDwrotethepaper.Theauthorsdeclarenoconflictofinterest.1Towhomthecorrespondenceshouldbeaddressed.Email:[email protected]:molecularmotors,microtubules,supportedlipidbilayer,efficiencyofcargotransport

not peer-reviewed) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was. http://dx.doi.org/10.1101/064246doi: bioRxiv preprint first posted online Jul. 15, 2016;

http://dx.doi.org/10.1101/064246

2

AbstractIn eukaryotic cells, membranous vesicles and organelles are transported byensembles ofmotor proteins. Thesemotors, such as kinesin-1, have beenwellcharacterized in vitro as single molecules or as ensembles rigidly attached tonon-biological substrates. However, the collective transport by membrane-anchored motors, i.e. motors attached to a fluid lipid bilayer, is poorlyunderstood. Here, we investigate the influence ofmotors anchorage to a lipidbilayeronthecollectivetransportcharacteristics.Wereconstituted‘membrane-anchored’ gliding motility assays using truncated kinesin-1 motors with astreptavidin-binding-peptide tag that can attach to streptavidin-loaded,supportedlipidbilayers.Wefoundthatthediffusingkinesin-1motorspropelledthemicrotubules inpresenceofATP.Notably,we found the gliding velocity ofthemicrotubules tobe stronglydependenton thenumberofmotorsand theirdiffusivity in the lipid bilayer. Themicrotubule gliding velocity increasedwithincreasingmotordensityandmembraneviscosity, reachingup to the steppingvelocity of single-motors. This finding is in contrast to conventional glidingmotilityassayswherethedensityofsurface-immobilizedkinesin-1motorsdoesnot influence themicrotubule velocity over awide range.We reason, that thetransportefficiencyofmembrane-anchoredmotors is reducedbecauseof theirslippage in the lipidbilayer,aneffectwhichwedirectlyobservedusingsingle-molecule fluorescencemicroscopy.Our results illustrate the importance of themotor-cargocoupling,whichpotentiallyprovidescellswithanadditionalmeansofregulatingtheefficiencyofcargotransport.SignificanceMolecularmotors, suchaskinesin-1, areessentialmolecules involved inactiveintracellulartransport.Currentmechanisticinsightsontransportbymotorsaremainlybasedoninvitrostudieswherethemotorsareboundtorigidsubstrates.However,whentransportingmembranouscargounderphysiologicalconditions,multiplemotorsareoftenonly looselycoupledviaa lipidbilayer. Inthisstudy,weinvestigatehowthemotors’transportefficiencyisaffectedwhenboundtoalipid bilayer. In our reconstituted gliding motility assays, we show thatmembrane-anchoredmotorsexhibitreducedtransportefficiencyduetoslippageinthelipidbilayer.Notably,theefficiencyincreasesathighermotordensityandreduced membrane diffusivity, providing cells with an additional means ofregulatingtheefficiencyofcargotransport.


http://dx.doi.org/10.1101/064246

3

Intracellulartransportofmembrane-boundvesiclesandorganelles isaprocessfundamental tomanycellular functions includingcellmorphogenesis, signalingandgrowth(1–4).Activecargotransportinsidetheeukaryoticcellsismediatedby ensembles of motor proteins, such as kinesins and dynein, walking onmicrotubule tracks (5), and myosins walking on actin filaments (6). Gainingmechanistic insight into the functioning of these motors inside the complexenvironmentofcells ischallenging.Severalstudieshavethusemployed invitroapproachestoinvestigatethetransportmediatedbygroupsofsameordifferentmotors attached to cargos such as silica beads (7), quantum dots (8), glasscoverslips(9,10)orDNAscaffolds(11,12).Although,theseapproachesprovideuswithknowledgeaboutthecollectivedynamicsofmulti-motortransport,akeyanomaly in these in vitro systems is the use of non-physiological rigid cargo.Vesicularcargotransportbymolecularmotorsentails theirattachmentto fluidlipidbilayereitherdirectlyorviadifferentadaptormolecules.Theanchoringofmotors in a diffusive lipid environment induce loose inter-motor coupling andthemotorscandiffusewithinthelipidbilayer,therebyincreasingtheflexibilityofthesystem.Theeffectofmembrane-motorcouplingonthetransportbehaviorofmotors ispoorlyunderstood. Inaddition,notmuch isknownabouthowthetransportefficiencyofthemotors isaffectedbytheirdensityanddiffusivityonthemembranouscargo.In recent past, a few in vitro studies have been performed with membranouscargo such as liposomes (13) or supported lipid bilayers (SLBs) (14) toinvestigatethetransportcharacteristicsofactinfilamentsbasedmotorproteinssuch as myosin Va and myosin 1c. Moreover, a recent study from our groupreportedthelong-rangetransportofgiantvesicles(1-4μm)drivenbykinesin-1motors as aproofof concept thatmodelmembrane systems canbyutilized tostudy microtubule-based motors such as kinesin-1 (15). However, due to thespherical geometry of GUVs or liposomes, it is difficult to determineunequivocally the number of motors transporting the cargo. Therefore,investigating the cooperative effects in cargo transport driven by membrane-anchoredmotors in suchmembrane-model systems is difficult. To circumventthis problem, we reconstituted the transport driven by multiple-motorsanchored to lipid bilayer in an inverse geometry. We anchored recombinanttruncatedkinesin-1motorswithstreptavidinbindingpeptidetag(SBP)toaflatbiotinylated supported lipid bilayer (SLB) via streptavidin. These membrane-anchoredmotorswerefoundtopropelthemicrotubulesinthepresenceofATP.We examined the transport efficiency of such motors (ratio of observedmicrotubule gliding velocity and maximal microtubule gliding velocity) atdifferent motor density and diffusivity. The flat geometry of the lipid bilayerallowed us to directly determine the density and diffusivity of GFP-taggedkinesin-1 using total internal reflection fluorescence (TIRF) microscopy. Weobservedthatthetransportefficiencyofmembrane-anchoredkinesin-1motorsis reduced due to motor slippage in the membrane. However, the transport


http://dx.doi.org/10.1101/064246

4

efficiencyisincreasedathigherdensityandreduceddiffusivityofmotorsinthemembrane.ResultsDiffusivity of kinesin-1 motors anchored to SLB is determined by thediffusivityoflipidsinSLB.SLBsareoneofthemostsuitablemodelmembranesystemstoquantifyprotein-lipidinteractions(14,16,17).Theycanbesubjectedto single molecule imaging techniques such as TIRF microscopy, due to theirflatness.Thisisnotpossiblewithothermembrane-modelsystemssuchasgiantunilamellar vesicles or liposomes due to their spherical geometry. Using TIRFmicroscopy we could quantify the density of motors as well as the motordiffusivity on SLBs with high accuracy. SLBs with 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC) and 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[biotinyl(polyethyleneglycol)-2000] (DSPE-PEG-2000-Biotin) in molar ratio99:1werepreparedbyfusionofsmallunilamellarvesiclesonhydrophilicglasscoverslip.This lipidmixturewasdopedwith0.05%1,2-dioleoyl-sn-glycero-3-phosphoethanolamine-Atto647n (DOPE-ATTO647n) as a fluorescent marker.Homogenous SLBs were obtained (see Methods) as indicated by the uniformspreadingof lipidmarker, visibleunder the fluorescencemicroscope (Fig. 1A).Fluorescence recovery after photobleaching (FRAP) experiments wereperformed to determine the fluidity of the lipids in SLBs (see Methods). ThediffusioncoefficientoflipidsinSLBwasobtainedtobeDLipid=3.0±0.3µm2/s(N= 16 SLBs, four independent experiments, Fig. 1B) by fitting the fluorescencerecoverycurves,accordingtothemethodologydescribedinRef.(18).Weexpressedandpurifiedratkinesin-1heavychainisoformKIF5C,truncatedto430a.a(19),withSBPtagatthetail(rKin430-SBP)(20),aswellaswithSBPandGFP-tag at the tail (rKin430-SBP-GFP). rKin430-SBP was anchored tobiotinylated SLB via streptavidin. Thereby, biotinylated SLBs were incubatedwithasaturatingconcentrationof streptavidin (0.5µM),100 foldshigher thanthenumberofbiotinylatedlipids.Thus,itwasensuredthatthesurfacedensityofmotors on the streptavidin loaded biotinylated SLBs is regulated by the bulkconcentrationofmotorsappliedtothereactionchamberandnotbythenumberofavailablebindingsites.TomeasurethediffusivityofmotorsanchoredtoSLBwe addedGFP-labeled rKin430-SBP at low concentration (<10nM) alongwithrKin430-SBP (0.25 µM) to the streptavidin-coated biotinylated SLBs. ThediffusingsinglemoleculesofrKin430-SBP-GFPwereexcitedwitha488-nmlaserand imaged using TIRF microscopy. Single-particles were tracked using theFluorescence Image Evaluation Software for Tracking and Analysis (FIESTA)(21) toobtain the singlemolecule trajectories (Fig.1C).Theensembleaveragediffusion coefficient of rKin430-SBP-GFP was obtained to be DKin = 3.1 ± 0.2µm2/s(mean±95%confidenceinterval,N=40molecules)ascharacterizedbylinearfittingtothemeansquaredisplacement(MSD)asafunctionoftime(Fig.1D,seeMethods).ThediffusioncoefficientofrKin430-SBPobtainedfromsingle-particle tracking is consistentwith the fluidity of lipids in SLBs obtained from


http://dx.doi.org/10.1101/064246

5

FRAPanalysis.Thus,ourresultsvalidatethatthediffusivityofmotorsattachedtoaSLBisgovernedbythefluidityoflipidanchorsintheSLB.Furthermore,thesingle-particle trackingstudiesconfirmthat therKin430-SBP is freelydiffusinguponanchoringtoSLBs.Microtubule gliding velocity increasedwith increasing surface density ofmembrane-anchored motors. It has been previously reported that themicrotubule gliding velocity propelled by surface-immobilized kinesin-1, isindependent of the motor density over a wide range (22, 23). But are thetransport characteristics of loosely coupled membrane-anchored motorsdifferent from the surface-immobilized motors? To address this question, wereconstituted microtubule gliding driven by kinesin-1 anchored to a SLB.Experimental set-upof themembrane-anchoredglidingassay is shown in (Fig.2A). Rhodamine-labeled, double-stabilized microtubules (see Methods) wereappliedtothemembrane-anchoredrKin430-SBP,inabuffersolutionwith1mMATP. Themotors anchored to the SLB propelled themicrotubules, confirmingthat the motors were functional upon binding to biotinylated SLBs. Wesystematically varied the surface density ofmotors on the SLBs by incubatingstreptavidin-bound biotinylated SLBswith different concentration of rKin430-SBPrangingfrom(0.07µM–0.63µM).Weobservedsmoothmicrotubuleglidingat high motor densities, however, microtubule translocation got slower andwigglieratlowermotordensities(MovieS1).Tocalculatethemicrotubuleglidingvelocity,translocationofmicrotubuleswasdividedintotwocomponents(i)translationalcomponentduetoactivetransportby themotors and (ii) diffusional component due to attachment to a diffusivelipidbilayerviamotors.Translationalanddiffusivecomponentweredeterminedby fitting the MSD of the microtubule center over time with the followingequationasdescribedin(24)𝑀𝑆𝐷 𝑡 = 𝑣!" ∙ 𝑡 ! + 4𝐷!" ∙ 𝑡 + 𝑐 [1]here𝑣!"is the translational velocity of a microtubule,𝐷!"is the diffusionalcomponentand𝑐istheoffsetaccountingforthelocalizationuncertaintyandthedynamicerrorduetofinitecameraacquisitiontime(25).Wevalidatedthatthemicrotubulesdrivenbymembrane-anchoredmotorshavebothtranslationalanddiffusional component by performing motility assay in presence of 1 mMadenylyl imidodiphosphate tetralithium(AMP-PNP),anon-hydrolysableanalogofATP.Thisensuredthatthemotorswereboundtothemicrotubulesinarigorstate,thusthemicrotubulesonlyhavediffusionalcomponent.Infact,thisiswhatwe observed (Movie S1), where the MSD of the microtubules in AMP-PNPincreasedonlylinearlywithtime(Fig.2B).


http://dx.doi.org/10.1101/064246

6

TheMSDdataof theglidingmicrotubules fittedwell to theEq.1 for allmotorconcentrations (Fig. 2B). The slope of theMSD curves vs. time increasedwithincreasing motor concentration. To resolve whether the microtubule glidingvelocity was dependent on the number of motors or the motor density, webinnedthemicrotubulesover1µmlengthintervalsanddeterminedtheaveragemicrotubule velocities. We observed that the gliding velocities were largelyindependent of themicrotubule lengths for allmotor concentrations (Fig. 2C).Thus, the transport efficiency of themembrane-anchoredmotors propelling amicrotubuleissetbythenumberofmotorsperunitlengthofamicrotubuleandnotbytheabsolutenumberofmotors.Membrane anchored motors slip backward in lipid bilayer whilepropelling a microtubule forward. One of the striking observations in ourmembrane-anchored gliding motility assays was that the microtubules uponcollision did not cross each other (Fig. 3A upper panel, Movie S2). Taxol-stabilized GTP-grown microtubules always aligned with passing microtubulesupon collision. Stiffer, double-stabilized microtubules aligned with passingmicrotubuleswhencollidingatshallowangles,buttemporarilystalleduntilthepassingmicrotubulesglidedawaywhencollidingatsteepangles.Thisbehavioris in contrast to gliding motility assays with surface-immobilized kinesin-1,where the microtubules cross over each other upon collision, without anynoticeablehindrance(Fig.3Alowerpanel,MovieS2,andRef.(26)).Wethereforehypothesizethattheforceoutputofmembrane-anchoredmotorsisloweredduetotheirmembrane-anchoredtailsslippinginthelipidbilayer.Todirectlyimagewhetherthemotorsslipinthelipidbilayerwhilepropellingamicrotubule, we performedmembrane-anchored gliding assays with rKin430-SBP spiked with low concentration (<25 nM) of rKin430-SBP-GFP. Spikingexperimentsenabledustoresolvesinglemotorspropellingamicrotubuleforawide range of motor concentrations.We observed that at high motor density(whenthemicrotubulesmovedfast)themembrane-anchoredmotorspropellingit slipped backward slowly. However, at low motor density (when themicrotubulesmoved slowly) themotors slipped backward at higher velocities(Fig. 3B). The relation between slipping velocity ofmotors in the lipid bilayerandmicrotubuleglidingvelocityinourexperimentscanbeclearlyseenwhenafastmovingmicrotubuleencountersanobstacle(anotherpassingmicrotubule)briefly, thus causing the microtubule to stall. When a moving microtubule isstalled, the motors underneath the microtubule started slipping backward athigh speed, as seen in the kymograph (Fig. 3C). These results suggest that themotorslippingvelocityandthemicrotubuleglidingvelocityareaddinguptothesteppingvelocityofasinglerKin430-SBPmotor.Frictional forces on themembrane-anchoredmotors and onmicrotubuledeterminethetransportefficiency.Toquantitativelydescribeourobservationof increasing transport efficiency with increasing membrane-anchored motor


http://dx.doi.org/10.1101/064246

7

density,wedevelopeda theoretical descriptionofmembrane-anchoredglidingmotility.Thenanoscopicset-upwiththephysicalparametersusedtodevelopthemathematical model is illustrated in (Fig. 4A). We resolved the dynamics ofmembrane-anchoredglidingmotilitybyconsideringthevelocitiesandfrictionalforces of membrane-anchored motors and microtubules. (1) Velocities:Membrane-anchored motors step on a microtubule with a velocity𝑣!"#$thuspropelling a microtubule with a velocity 𝑣!" relative to the substrate, in adirectionoppositetoitssteppingdirection.Asobservedintheexperiments,themotors drag their anchors in the lipid bilayer, under the microtubule, in thesteppingdirectionwithavelocity 𝑣!"#-!"#$relativetothesubstrate,asthemotoris not rigidly bound but anchored to a fluid bilayer. Thus, the relationshipbetweenthedifferentvelocitiescanbeformulatedas: 𝑣!" = −(𝑣!"#$ − 𝑣!"#-!"#$)or|𝑣!"#$| = −𝑣!"#-!"#$ + 𝑣!" [2]

The stepping velocity of rKin430-SBP-GFP was determined by performingstandard steppingmotility assays,where themovement ofmotors on surface-immobilized microtubules in the assay buffer was recorded and analyzed(Methods, Fig. S1) with𝑣!"#$ = 0.67 ± 0.14 µm/s (mean ± SD, N = 545molecules).(2)Forces:Understeadystate,thenetforceactingonthesystemiszero, i.e. the forces on themotors balance the forces on themicrotubule. Thefrictional force acting on a microtubule 𝐹!" can be estimated by thehydrodynamic drag, due to its motility in the aqueous solution. The frictionalforceonakinesinmotor𝐹!"#canbeestimatedbythedraginthefluidbilayer.Atany instance there would be several motors𝑁interacting with a microtubule.These𝑁 motors stepping, with the same velocity𝑣!"#$, in an uncorrelatedmanneronamicrotubulewouldexperienceequaldragforce.Theforcebalanceequationthenyields:

𝑁 ∙ 𝐹!"# + 𝐹!" = 0[3]

Thefrictionalforceonamicrotubulecanbeestimatedbyconsideringitasalongrigid cylinderwith its lengthmuch larger than its radiusLMT >> rMT. The dragcoefficientforcylindricalobjectsmovingparalleltothesurfaceisgivenby(22)

𝛾 =2𝜋𝜂𝐿!"

ln (2ℎ 𝑟!") [4]

Themagnitudeof thedrag forceon themicrotubule canbedeterminedby theStokes’law:


http://dx.doi.org/10.1101/064246

8

𝐹!" = 𝛾 ∙ 𝑣!" =2𝜋𝜂𝐿!"

ln (2ℎ 𝑟!") ∙ 𝑣!" [5]

A microtubule would experience maximum drag force when its moving athighest velocity which is 𝑣!"!"# = 0.67 µm/s calculated from equation (2)considering 𝑣!"#!!"#$ = 0. Thus for𝜂 =10-3 Pa∙s (water),𝐿!" =10 µm,ℎ =50nmand𝑟!" =12.5nm,themaximumforceonamicrotubulewouldbe 𝐹!"!"# =20fN. The magnitude of maximum drag force on a microtubule in aqueousenvironmentis200-foldslessthanthestallforceofasinglekinesin-1motor5-7pN(27,28).Thus,evenasinglesurface-immobilizedkinesin-1motorcanpropela microtubule at maximum velocity in aqueous environment. Therefore, themicrotubule gliding velocity is independent of motor density for surface-immobilizedkinesin-1overawiderange.

Frictional forceonamembrane-anchoredmotorcanbeestimatedbyusingtheEinstein-Smoluchowski-relation where the drag coefficient is related to thediffusivityDKinofamoleculeby

𝛾 =𝑘!𝑇𝐷!"#

[6]

Themagnitudeoftheforcethencanbedeterminedby

𝐹!"# = 𝛾 ∙ 𝑣Kin-slip =𝑘!𝑇𝐷!"#

∙ 𝑣!"#-!"#$ [7]

Single membrane-anchored kinesins would experience a maximum drag forcewhentheyareslippingunderastationarymicrotubuleatitsmaximumsteppingvelocity,whichis𝑣!"#-!"#$

!"# =0.67µm/scalculatedfromequation(2)considering𝑣!" = 0.Wheninteractingwithamicrotubule,themobilityofamotorisreducedto only one dimension since the rKin430-SBP head walk on a singleprotofilamentofamicrotubule.Thereby,thediffusioncoefficientofmicrotubule-interactingmotorsinthelipidbilayerwillbereducedtohalf.For 𝑘! =1.38x10-23 J/K,𝑇 =295 K and𝐷!"#-!" = 𝐷!"# 2 =1.56 µm2/s, the maximum force on

membrane-anchored rKin430-SBPpropelling amicrotubulewould be 𝐹!"#!"# =1.7fN. This value is 250 folds smaller than the stall force of a single kinesin-1motor.Thus,kinesin-1motorsunderthelowloadconditionofourset-upwouldstepon amicrotubule at theirmaximumstepping velocity.By substituting theexpressionof forces formicrotubuleandkinesinmotor in theequation (3)weobtain


http://dx.doi.org/10.1101/064246

9

𝑁 ∙𝑘!𝑇

𝐷!"#-!" ∙ 𝑣!"#-!"#$ +

2𝜋𝜂𝐿!"ln (2ℎ 𝑟!")

∙ 𝑣!" = 0 [8]

whichcanbesimplifiedto

𝑣!" = −𝑁 ∙

𝑘!.𝑇𝐷!"#-!"2𝜋𝜂𝐿!"

ln (2ℎ 𝑟!") ∙ 𝑣!"#-!"#$

[9]

𝑣!"#-!"#$issubstitutedintheaboveequationas𝑣!"#$ − 𝑣!"fromequation(2),toobtaintherelation

𝑣!"𝑣!"#$

= −1

1+ 𝑐 ∙ 𝐷!"#-!"𝜌

[10]

where !!"!!"#$

is the transport efficiency,𝜌 = 𝑁𝐿!" is the number of motors

interactingwithamicrotubuleperunitlengthi.e.thelinearmotordensity,andcisaconstantwhichdependsonthephysicalparametersgivenby

𝑐 = 2𝜋𝜂

ln 2ℎ 𝑟!" ∙ 𝑘!𝑇= 0.72 µm-3 ∙ 𝑠 [11]

Hereby,wehavederiveda simplemodel thatdescribes thedependenceof thetransport efficiencyofmembrane-anchoredmotorson their lineardensity andtheirdiffusivity.Themodelpredictsthatthetransportefficiencyofmembrane-anchoredmotorsincreaseswithincreasingmotordensitybutisindependentofthemicrotubule lengthwhich iswhatweobserved inourmembrane-anchoredgliding assays (Fig. 2B-C). Formotors immobilized on the glass substrate, theapparentDKin-1Diszero.SubstitutingintoEq.10,weobtainthatthemicrotubulegliding velocity is as high as the single motor stepping velocity and isindependentof themotordensity.Thisresulthasbeenreportedpreviously forkinesin-1 (23,29)andwhatweobserve for rKin430-SBPoverawide rangeofmotordensity.However,forthemembrane-anchoredmotorsthemodelpredictshighertransportefficiencyathighmotordensityand/oratlowdiffusivityofthemotors’lipidanchors.

Theoretical model describes the experimental findings well. Weexperimentally tested the predictions of our theoretical model, i.e. thedependenceof transport efficiencyofmembrane-anchoredmotorson i)motor


http://dx.doi.org/10.1101/064246

10

densityonSLBsandii)diffusivityofmotorslipid-anchorinSLB.Tomeasurethesurface density of motors on the SLBs, we incubated the streptavidin-coatedbiotinylated SLBs with rKin430-SBP along with a low concentration of GFP-labeledrKin430-SBP(~20nM),inafixedmolarratioof150:1.Thesampleswereexcitedat488nmandtheimagesofGFP-labeledsinglemotorsdiffusingontheSLBswererecorded,usingTIRFmicroscopy.Toavoiderrorsincountingduetophotobleaching,weonlyconsideredthefirstfiveframesofthemoviestreamtodetermine thenumberofmotorsanchored toaSLB (seeMethods).Themotorsurface density was determined by averaging the total number of diffusingparticles, per unit area of the field of view. Using this methodology we coulddirectly measure the surface density of motors. Subsequently, membrane-anchoredglidingmotilityassayswereperformedonthesamesamplestoobtainthe microtubule gliding velocity, for a measured motor surface density. Tocompareourexperimentaldatawiththetheoreticalmodel,weassumedalinearrelationshipbetweenthesurfacemotordensity 𝜎andthelinearmotordensity𝜌,givenby(30) 𝜌 = 𝜎 ∙ 𝜔 [12]

where𝜔istheinteractionreachofthemotortobindtoamicrotubulefilament.Thustheequation[10]canbewrittenas

𝑣!"𝑣!"#$

= −1

1+ 0.72 · 𝐷!"#-!"𝜎 · 𝜔 [13]

Theexperimentswereperformedoverawiderangeofsurfacemotordensitytoquantifythedependenceoftransportefficiencyonmotordensity(Fig.4B).Our theoretical model predicts that at a fixed motor density, the transportefficiencyofmembrane-anchoredmotors increaseswith increasingviscosityofthelipidanchors.Totestthisprediction,wevariedthediffusivityoftheSLBstowhichthemotorswereanchored,byaddingcholesterol. Ithasbeenpreviouslyshown that the addition of cholesterol increases the packing of lipids withunsaturatedacyl chains suchasDOPCand thus reduces thediffusivityof SLBs(31). We added cholesterol to the lipid mixture in two different molar ratios(20% and 60% of the total lipid concentration) with the compositionDOPC:Cholesterol:DSPE-PEG-2000-Biotin 79:20:1 (20% CH) and DOPC:Cholesterol:DSPE-PEG-2000-Biotin39:60:1(60%CH),alongwithatraceamountof fluorescent lipid-marker DOPE-ATTO647n to check the fluidity andhomogeneity of SLBs. We performed FRAP experiments and found that thediffusivityoftheSLBsreducedafteradditionofcholesterol,with60%CHbeingless diffusive than 20% CH (Movie S3). Subsequently, we measured thediffusivity of membrane-anchored motors (rKin430-SBP along with a lowconcentrationofrKin430-SBP-GFP)intheCHSLBs(Methods)andobtained1.0±


http://dx.doi.org/10.1101/064246

11

0.2 µm2/s; (mean± 95% confidence interval;N = 48molecules) and0.5 ± 0.1µm2/s(mean±95%confidenceinterval,N=58molecules)for20%and60%CHSLBs, respectively.Thus,withsinglemolecule trackingof rKin430-SBP-GFPweconfirmed that the addition of cholesterol in our experimental set-up reducedthediffusivityofthemembrane-anchoredmotors.Subsequently, membrane-anchored gliding motility assays at different motordensitieswereperformedon theCHSLBs.Weobserved that,on20%CHSLBsmicrotubulesdidnotcrosseachotheruponcollision,whereason60%CHSLBsthey did (Movie S4). Thus, the motors anchored to 60% CH SLBs were lessslippery than the motors anchored to 20% CH SLBs and therefore producedenough force output to propel microtubules over each other. The transportefficiencyofmotorsanchoredtoCHSLBsalso increasedwith increasingmotordensity,similartoourfindingson0%CHSLBs(Fig.4B).Furthermore,foragivenmotordensity,thetransportefficiencywashighestfor60%CHSLBsfollowedby20% CH and least for 0% CH SLBs.We fitted our experimental data with thetheoretical model (Eq. 13) using fitting routines in MATLAB (Methods). Ourtheoretical model with only one free parameter (namely𝜔) fitted well to theexperimentaldata fordifferentmotordensitiesaswellasdifferentdiffusivities(Fig.4B).Fromthefitsweobtainedthevaluesof𝜔tobe0.31±0.07µm,0.24±0.01µm,and0.22±0.02µm,(mean±95%confidenceinterval)for0%CH,20%CHand60%CHrespectively.Previousstudieshaveestimatedthevalueof𝜔forkinesin-1motors immobilizedon a solid substrate tobe around0.02µm (30).However,forourexperimentalset-upweexpectahigherreachasthekinesin-1motorsarediffusingfreelyonalipidbilayer.Forexampleamembrane-anchoredkinesin-1diffusingat3.1µm2/scanexploreacircleofradius0.45µmin50ms.Thus,theobtainedvalueof𝜔isreasonableforourmembrane-anchoredglidingmotility assays. Our experimental data is consistentwith the predictions fromour theoretical model, describing the increase in transport efficiency ofmembrane-anchored motors with increasing motor density and increasingviscosity.


http://dx.doi.org/10.1101/064246

12

DiscussionIn this study, we established membrane-anchored gliding motility assays, toinvestigate the transport characteristics of kinesin-1 motors that are looselycoupledviaalipidbilayer.Forkinesin-1tightlycoupledtoasubstrateithasbeenpreviouslyreported that themicrotubuleglidingvelocity iseither independentof themotor density or reduced at highermotor density (23, 32). Our resultsshow that the gliding velocity of microtubules driven bymembrane-anchoredkinesin-1 is reduced due to slipping of themotor anchors in the lipid bilayer.However, the gliding velocity increases with increasing motor density orincreasingviscosity.Invivo,cargotransportvelocitiesdeterminedbytrackingvesiclesororganellesinvariouscelltypesexhibitsignificantspreadsandthevelocityhistogramsoftencontain multiple peaks (33–36). In the complex environment of a cell, therecould be many factors that influence cargo transport, such as motor activity,membrane-motorbindingpartners,themovementofcytoskeletalfilaments(37)etc.toproducefasterorslowermotility.Onewayinwhichthetransportcanberegulatedinsidecellsisbyvaryingthelipidcompositionofmembranouscargos.For example, if the motors are segregated into lipid-microdomains, both thefrictional forces on the motor anchors as well as the motor density wouldincrease. This results in less slippage of the motors and thus an increasedtransport efficiency. Along these lines, a recent study has reported thatcytoplasmic dynein clusters into lipid-microdomains on phagosomes to driverapid transport towards lysosomes (38).The importanceofmotordensityandmembrane rigidity was also highlighted previously when tubular transportintermediates between organelles were reconstituted by extracting nano-vesicular tubes fromGUVswithkinesin-1motors(39,40).There itwasshownthataminimummotordensitywas required topull themembrane tubes fromGUVsdependingonthelipidcomposition.To the best of our knowledge, this is the first reconstitution of microtubuleglidingmotility onmembrane-anchoredmotor proteins.We believe, our assayprovidesausefultooltostudytheregulationofcargotransportbyavarietyofmolecularmotorssuchaskinesin-3(KIF16B,KIF1A),whichcandirectlybindtoamembranouscargo.Furthermore,theseassayscanbeusedtogainmechanisticinsight into the recruitmentof variousmotors to their specific cargoand theirtransport characteristics. So far, our experiments were performed on flatmembranes and do not yet account for varying cargo geometries such asspherical vesicles (as in secretory granules) and small tubules (as in theEndoplasmic reticulum). Most likely, this geometry additionally influences thetransportefficiencyandwillbeasubjectofourfurtherstudies.


http://dx.doi.org/10.1101/064246

13

MethodsSupportedlipidbilayer(SLB)preparation.Glasscoverslipswerepreparedby15 min sonication in 5% Mucasol solution, extensive rinsing with nanopurewater, 15 min sonication in 100% ethanol, and again extensive rinsing withnanopure water. Coverslips were dried with pressured nitrogen and plasmacleaned (Diener electronic) for 10 min. Multi-lamellar vesicles (MLVs) of thedesiredlipidmixturewerepreparedbytaking7.5µgoftotallipidsinaglassvial(Sigma) in the requiredmolar ratio and evaporating the solvent (chloroform)under a constant stream of nitrogen. Any residual solvent was removed bykeepingtheglassvialundervacuumovernight.ThelipidswerethenrehydratedinH20S75buffer(20mMHEPES,75mMNaCl,pH7.2)toaconcentrationof0.2mg/mlandsonicatedfor20mintoformsmallunilamellarvesicles(SUVs).TheSUVdispersionwasaddedtotheexperimentalchambers,formedbyattachingacut200µlEppendorf tubeusingUVadhesive(NOA83,Norlandproducts) toaplasmacleanedglasscoverslip.CaCl2wasaddedtoafinalconcentrationof3mMtoinducefusionofSUVsandformationofanSLB.After45minofincubationatroomtemperature,thesamplewaswashedwith1mlofH20S75bufferinstepsof 50μl to removeunfused vesicles. SLBswerepreparedwith1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC, Avanti) and 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[biotinyl(polyethylene-glycol)-2000](DSPE-PEG-2000-Biotin, Avanti) in molar ratio 99:1 with 0.05% 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine-Atto647n(DOPE-ATTO647n,Atto-Tecfluorescentlabels)as a fluorescent lipid marker. The composition of SLBs with cholesterol wasDOPC:Cholesterol(Avanti):DSPE-PEG-2000-Biotin = 79:20:1 (20% CH) andDOPC:Cholesterol:DSPE-PEG-2000-Biotin=39:60:1(60%CH).rKin430-SBP purification and attachment to SLBs.AcodonoptimizedDNAsequenceofkinesin-1containing theN-terminal430aminoacidsof theRattusnorvegicuskinesin-1 isoformkif5c,with the tags8xHis,mCherry andSBP,waspurchasedfromInvitrogen(GeneArt,Invtirogen).Tworestrictionsites,PacIandAscI,wereintroducedinpET24dvector(#69752-3,Addgene)andtherKin430-mCherry-SBPsequencewasinsertedinthevectorafterdigestingitwithPacIandAscI restriction enzymes (NewEngland Bioloabs). ThemCherry sequencewascutoutusingtherestrictionenzymeNgoMIVandthecutplasmidwasligatedtoobtain the rKin430-SBP plasmid. The rKin430-SBP-GFP DNA sequence waspreparedbyinsertingamultifunctionalGFP(mfGFP)tag(20)having8xHis,SBP,and c-Myc tag, in tandem in a loop of the GFP sequence. The sequence wasinserted into rKin430 plasmid (19) in pET17 vector (# 69663-3, Addgene).Briefly, PCR with primers GGTACCGTGAGCAAGGGCGAGGAGCTGTTC andCAATTGTTACTTGTACAGCTCGTCCATGCCGAGAGTG was used to amplify themfGFPsequenceandrestrictionsitesKpnIandMfeIwereaddedatthe5’and3’endofthecomplementarysequence,respectively.ThemfGFPsequencewastheninserted into rKin430plasmid inpET17vectorusing restrictionenzymesKpnIand MfeI (New England Bioloabs). Both constructs were expressed in BL21(DE3) E . coli cells. Expression with isopropyl β-D-1-thiogalactopyranoside


http://dx.doi.org/10.1101/064246

14

(IPTG)andaffinitypurificationwith theHisTrapColumn(GEHealthCare)wascarried out as described previously (19). rKin430-SBP and rKin430-SBP-GFPconcentrationswerequantifiedbycomparisonwiththeknownGFPstandardsonaCoomassie-stained4-12%Bis-TrisNuPAGEgel(ThermoFisherScientific).rKin430-SBP and rKin430-SBP-GFPwere attached to SLBs, by first incubatingthebiotinylatedSLBswith0.5µMstreptavidin(Sigma)in100µltotalvolumefor10minfollowedbywashingwith1mlofH20S75buffertoremovetheunboundstreptavidin. Before adding the motors to the streptavidin-loaded SLBs, theexperimental chambers were equilibrated by exchanging H20S75 buffer withmotorbuffer(20mMHEPES,75mMNaCl,1mMATP,1mMMgCl2,1mMDTTand20mMD-glucose). 50µl of thebuffer in the experimental chamberwerethenreplacedwith50µlofmotorsolution,consistingoftherequiredamountofrKin430-SBPor rKin430-SBP-GFP inmotorbuffer.After incubating themotorson thestreptavidin-loadedSLBs for6min,unboundmotorswerewashedwith200µlofassaybuffer(20mMHEPES,75mMNaCl,1mMATP,1mMMgCl2,1mMDTT, 40mMD-glucose, 0.02mg/ml glucose oxidase, 0.01mg/ml catalaseand1µMtaxol).Stepping and gliding motility assays. Experiments to obtain the steppingvelocity of single rKin430-SBP-GFP motors were performed in flow channelsmadeof silanized coverslips, asdescribed in (41).Briefly, first a solutionofβ-tubulinantibodies(0.5%SAP.4G5,ThermoFisherScientific)diluted inH20S75bufferwas flushed into a flow channel and incubated for 5min, followedby awashingstepwithH20S75buffer.Theflowchannelwasthenincubatedwith1%Pluronic F127 in H20S75 for 45minutes to block the surface from unspecificbinding of proteins. Subsequently, the flow channelwaswashedwith 80 μl ofH20S75supplementedwith10μMtaxol(H20S75T).Fluorescent,taxol-stabilizedrhodaminemicrotubules(usedinthesteppingmotilityassaysaswellasforthedata in Fig. 3A and B) were prepared as described previously (41). Taxol-stabilized GMP-CPP microtubules (also referred to as ‘double-stabilizedmicrotubules’, used throughout our work unless mentioned otherwise) wereprepared by polymerizing 2.5 µM tubulinmix (1:3 rhodamine labeled tubulin:unlabeledtubulin)inBRB80buffer(80mMPIPES,1mMMgCl2,1mMEGTA,pH6.9) supplemented with 1.25 mM GMP-CPP and 1.25 mM MgCl2. Thepolymerizationwascarriedoutat37°Cfor2hoursin80µlBRB80,afterwhich120µlofBRB80with15µMtaxolisadded.Microtubuleswerealwayspreparedfreshly before each experiment and free tubulin was removed byultracentrifugation in an Airfuge (Beckman Coulter) at 100,000 x g for 10minutes. Thepelletwas re-suspended in 200µlH20S75Tbuffer. A solution ofmicrotubules was flushed in and allowed to attach to the antibodies on thecoverslip for 5 minutes. Unbound microtubules were removed from the flowchannel bywashingwith40μl ofH20S75T. Finally, 20μl of 100pM rKin430-SBP-GFPinassaybufferwasflushedintotheflowchannel.


http://dx.doi.org/10.1101/064246

15

Imageacquisitionforsingle-moleculeTIRFandFRAPexperiments.ImageswereobtainedusingaNikonmicroscope(NikonEclipseTiequippedwithPerfectFocusSystem(PFS)andaFRAPmodule,PlanApo100xoil immersionobjectivelens, NA 1.49) with an electron multiplying charge-couple device (EMCCD)camera(iXonultraEMCCD,DU-897U,Andor) inconjunctionwithNIS-Elements(Nikon) software. Single-molecule imaging for stepping assays, and single-particle tracking of rKin430-SBP-GFP on SLBs were performed using TIRFmicroscopyandamonolithiclasercombiner(AgilentMLC400)whichhasadualoutput forFRAPand fluorescence imaging.SLBswere imagedwithaCy5 filterset(exc:642/20.DichroicLP647,em:700/75)andrKin430-SBP-GFPmoleculeswereimagedwithanotherfilterset(exc:475/35.DichroicLP491,em:525/45).Imageswereacquiredincontinuousstreamingmodewith100msexposureforsteppingassaysand50msexposure forsingle-particle tracking, to localize thediffusingmotorsonSLBs.ForestimatingthemotordensityonSLBs,GFP-labeledmotorswere imaged for150 frames (256x256pixels)with100msexposuretimeeach.Forphotobleachingexperiments512x512pixel imagesof theSLBswere captured at 0.1 s interval for 1 s using the 647-nm laser line, followingwhicha150x150pixelregioninthecenteroffieldofviewwasbleachedusing647-nmlaseratfullpowerusingtheFRAPmodulefor4.2s(5scaniterations).Time-lapseimageswerethenrecordedatanintervalof0.5sfor250framestomonitortherecoveredfluorescenceinthebleachedarea.Imagesofrhodamine-labeledmicrotubulesinmembraneglidingmotilityassayswereobservedbyepi-fluorescence where microtubules were excited with a metal arc lamp(Intensilight, Nikon) in conjunction with a rhodamine filter set (exc: 555/25.DichroicLP561,em:609/54).Dataanalysis forFRAP, single-particle trackingandmotility assays.FRAPimageswereanalyzedtoestimatethediffusioncoefficientofthelipidsinaSLBusinganalgorithmdescribed inRef. (18).TheMATLABscriptwasmodified tocorrect for the fixedpatternnoisearisingdue tonon-uniform illuminationandTIRFimaging.Tocorrectforthefixednoise,alltheimagesbeforebleachingwereaveraged.Themeanofthe1%ofthetotalpixelswiththelowestintensitieswascalculated as a normalization factor and all the images in the stackwere thencorrected by multiplication with the normalization factor and dividing by themeanimagesothattheoverall intensitiesofallpixels in imageswereuniform.After background correction, the centers of the bleached regions and anappropriate unbleached reference were manually selected. The fluorescencerecoveryovertimewasfittedwiththemathematicalsolutiondescribedin(18)inordertocalculatethediffusioncoefficientsofthelipidsintheSLBs.Themeanofallindividualrecoverycurvesfromdifferentregionsofinterestwasthenfittedagaintoreducetheeffectofrandomfluctuations.Asaresult,themeandiffusioncoefficient for lipids in a SLB was obtained. The error was estimated bycalculatingthestandarderrorofthemeanofallindividualfits.


http://dx.doi.org/10.1101/064246

16

Fordeterminingthediffusivityofmembrane-anchoredrKin430-SBP-GFP,singlemolecules were tracked using Fluorescence Image Evaluation Software forTracking and Analysis (FIESTA) (21). Displacement data for all the singlemolecule trajectories were calculated for discrete time points and thedisplacement data were cumulated to calculate the average mean-squaredisplacement (MSD) foreverydiscrete timepoint.The first8points (basedontheestimationpublishedin(25))oftheMSDthusobtainedwerethenfittedwitha linearcurveusingtheerrorbarsasweightstoget thediffusioncoefficientofrKin430-SBP-GFPonSLBs.For stepping motility assays, the mean velocity for individual motors wasdeterminedby calculating the slopeof the trajectories in a kymograph (space-timeplotofintensityoveraspecifiedarea,spacedimensionwaschosenbyalinedrawnoveramicrotubule).KymographevaluationwasperformedusingFIESTAsoftware.For membrane-anchored gliding motility assays, microtubules were trackedusing FIESTA. All the connected tracks obtained from the software werevisualized to manually exclude erroneous tracks from further analysis.Erroneous tracking could be due to, for example, microtubules crossing eachother, sample drift or microtubule fragmentation. Therefore, the microtubulelength was used as a control parameter for post-processing of tracks as thelengthisexpectedtoremainconstantoverthedurationoftheexperiment.Theensemble-averagemicrotubulevelocityforaparticularsurfacemotordensityordiffusivitywasobtainedbycalculatingtheMSDofthemicrotubulecentersasafunction of time. Due to the imaging of discrete frames, the time 𝑡 is given asmultiples 𝑛 of the acquisition time interval such that 𝑡 = 𝑛∆𝑡. The MSD iscalculatedforthenon-overlappingtimeintervalsusingtheformalismdescribedin (25). The mean translational velocity and the diffusion coefficient for anindividualmicrotubulewerecalculatedbyfittingthefirst25pointsoftheMSDtimeplotwithequation[1].TheMSDdataforthefitwasweightedbytheinverseoftheerror.Onlymicrotubules,whichwereinthefieldofviewformorethan30frames were analyzed. To calculate the mean velocity of an ensemble ofmicrotubulesataparticularmotordensity,wecalculatedthecumulatedMSDforall themicrotubules in an image stack. The cumulatedMSDwas calculated bycumulating the displacement data of all the individual microtubule for eachdiscretetimepoint.Thefirst25datapointsoftheMSDdatathusobtainedwerefittedwithequation[1],togetanaveragemicrotubuleglidingvelocity.TheMSDdata for the fitwasweightedby the inverseof error.Theerror for the fitwasestimated by performing a boot strapping analysis (42). The MSD data fordifferent microtubules were randomly picked (with replacement) keeping thetotalnumberofmicrotubule tracks analyzed the sameas in the initialdataset.The standard deviation of the mean velocity obtained from the bootstrapanalysis,gavethestandarderrorofthemeanoftheinitialdataset.Microtubulegliding velocities as function ofmotor densitywere fitted to equation [11] for


http://dx.doi.org/10.1101/064246

17

different diffusivities of kinesin-1 on SLBs using the MATLAB curve fittingtoolbox(nonlinearleastsquarefittingwithLavenberg-Marquardtmethod).Thefitswereweightedbytheinverseoftheensemblemicrotubulevelocityerror.Determination of the motor surface density on SLBs. With the single-molecule sensitivity of our TIRF set-up wewere able to determine the actualnumber of diffusingmotors directly by counting. As compared to the indirectmeasurementbasedonthetotalfluorescenceintensityofGFPmolecules,whichcan be skewed by several factors such as TIRF angle, optical aberrations, GFPclustersinthesampleetc.;thisdirectmeasurementismorerobust.Specifically,the kinesin-1 motor densities on biotinylated SLBs (at different bulk motorconcentrations) were determined by incubating the SLBs with rKin430-SBPspiked with rKin430-SBP-GFP in a molar ratio 150:1. To avoid aberrations incounting due to photobleaching, the sample was focused by imaging the Atto647ndopedSLBbeforeactivatingtheperfect-focusmechanismoftheNikonTE2000 Eclipse microscope. Movie streams with 50 frames at 100 ms exposuretimeswerethenrecordedatfivedifferentfieldsofviews,byexcitingthesamplewitha488nmlaser.ThenumberofdiffusingrKin430-SBP-GFPinthefirstthreeframesof the imagestackswere thencountedusing thecell counterplug-inofthe image processing and analysis software FiJi. The means and standarddeviationsofthemeasurementswerethencalculated.Glidingmotilityassays,torecord themicrotubule gliding velocity at that particularmotor density, werethenperformed.AcknowledgementsWearegratefultoC.BräuerfortechnicalassistanceaswellasC.Herold,L.MuandM.StrempelforhelpwithinitialSLBexperiments.WethankM.BraunandallothermembersoftheDiezLabforfruitfuldiscussions.Weacknowledgesupportfrom the Volkswagen Foundation (Grant I/84 087), the European ResearchCouncil (Starting Grant 242933 to SD), the Dresden International GraduateSchool for Biomedicine and Bioengineering (stipend to RG) as well as theGerman Research Foundation within the Cluster of Excellence Center forAdvancingElectronicsDresden.References1. Hirokawa N, Takemura R (2005) Molecular motors and mechanisms of

directionaltransportinneurons.NatRevNeurosci6(3):201–214.2. Hirokawa N, Noda Y (2008) Intracellular Transport and Kinesin Superfamily

Proteins,KIFs:Structure,Function,andDynamics.PhysiolRev88(3):1089–1118.3. Vale RD (2003) The molecular motor toolbox for intracellular transport. Cell

112(4):467–80.4. BarlanK,RossowMJ,GelfandVI(2013)Thejourneyoftheorganelle:Teamwork

andregulationinintracellulartransport.CurrOpinCellBiol25(4):483–488.5. ValeRD(1987)Intracellulartransportusingmicrotubules-basedmotors.AnnRev

CellBiol3(1):347–378.6. HartmanMA, Spudich JA (2012) Themyosin superfamily at a glance. J Cell Sci


http://dx.doi.org/10.1101/064246

18

125(Pt7):1627–32.7. Rogers AR, Driver JW, Constantinou PE, Kenneth Jamison D, Diehl MR (2009)

Negative interference dominates collective transport of kinesin motors in theabsenceofload.PhysChemChemPhys11(24):4882–9.

8. Ali MY, Lu H, Bookwalter CS, Warshaw DM, Trybus KM (2008) Myosin V andKinesin act as tethers to enhance each others’ processivity. ProcNatl Acad Sci105(12):4691–4696.

9. LeducC,RuhnowF,HowardJ,DiezS(2007)Detectionoffractionalstepsincargomovement by the collective operation of kinesin-1 motors. Proc Natl Acad Sci104(26):10847–10852.

10. Gagliano J,WalbM,BlakerB,Macosko JC,HolzwarthG (2010)Kinesin velocityincreases with the number of motors pulling against viscoelastic drag. EurBiophysJ39(5):801–13.

11. DerrND,etal. (2012)Tug-of-War inMotorProteinEnsemblesRevealedwithaProgrammableDNAOrigamiScaffold.Science338(6107):662–665.

12. Furuta K, et al. (2013) Measuring collective transport by defined numbers ofprocessiveandnonprocessivekinesinmotors.ProcNatlAcadSci110(2):501–6.

13. Nelson SR, Trybus KM, Warshaw DM (2014) Motor coupling through lipidmembranesenhancestransportvelocitiesforensemblesofmyosinVa.ProcNatlAcadSci111(38):E3986–E3995.

14. PyrpassopoulosS,FeeserEa.,MazerikJN,TyskaMJ,OstapEM(2012)Membrane-bound Myo1c powers asymmetric motility of actin filaments. Curr Biol22(18):1688–1692.

15. Herold C, Leduc C, Stock R, Diez S, Schwille P (2012) Long-range transport ofgiantvesiclesalongmicrotubulenetworks.ChemPhysChem13(4):1001–1006.

16. Vogel SK,PetrasekZ,HeinemannF, SchwilleP (2013)Myosinmotors fragmentandcompactmembrane-boundactinfilaments.Elife2013(2):1–18.

17. BridgesAA,etal.(2014)Septinassembliesformbydiffusion-drivenannealingonmembranes.ProcNatlAcadSci111(6):2146–2151.

18. Goehring NW, Chowdhury D, Hyman A a., Grill SW (2010) FRAP analysis ofmembrane-associated proteins: Lateral diffusion and membrane-cytoplasmicexchange.BiophysJ99(8):2443–2452.

19. RogersKR,etal.(2001)KIF1Disafastnon-processivekinesinthatdemonstratesnovelK-loop-dependentmechanochemistry.EMBOJ20(18):5101–13.

20. KobayashiT,etal.(2008)Engineeringanovelmultifunctionalgreenfluorescentproteintagforawidevarietyofproteinresearch.PLoSOne3(12):e3822.

21. Ruhnow F, Zwicker D, Diez S (2011) Tracking single particles and elongatedfilamentswithnanometerprecision.BiophysJ100(11):2820–8.

22. HuntAJ,GittesF,HowardJ(1994)Theforceexertedbyasinglekinesinmoleculeagainstaviscousload.BiophysJ67(2):766–781.

23. Howard J, Hudspeth AJ, Vale RD (1989) Movement of microtubules by singlekinesinmolecules.Nature342(6246):154–158.

24. QianH,SheetzMP,ElsonEL(1991)SingleparticleTracking.BiophysJ60(4):910–921.

25. Michalet X (2010) Mean square displacement analysis of single-particletrajectories with localization error: Brownian motion in an isotropic medium.PhysRevE82(4):041914.

26. Kerssemakers J, et al. (2009)3Dnanometer trackingofmotilemicrotubulesonreflectivesurfaces.Small5(15):1732–1737.

27. Svoboda K, Schmidt CF, Schnapp BJ, Block SM (1993) Direct observation ofkinesinsteppingbyopticaltrappinginterferometry.Nature365(6448):721–7.

28. CarterNJ,CrossR(2005)Mechanicsofthekinesinstep.Nature435(7040):308–12.

29. Hancock WO, Howard J (1998) Processivity of the Motor Protein KinesinRequiresTwoHeads.JCellBiol140(6):1395–1405.

30. Duke T, Holy TE, Leibler S (1995) “Gliding assays” for motor proteins: A


http://dx.doi.org/10.1101/064246

19

theoreticalanalysis.PhysRevLett74(2):330–333.31. Simonsson L, Höök F (2012) Formation and diffusivity characterization of

supported lipid bilayers with complex lipid compositions. Langmuir28(28):10528–33.

32. BielingP,TelleyIA,Piehler J,SurreyT(2008)Processivekinesinsrequire loosemechanicalcouplingforefficientcollectivemotility.EMBORep9(11):1121–7.

33. Hill DB, PlazaMJ, Bonin K, Holzwarth G (2004) Fast vesicle transport in PC12neurites:velocitiesandforces.EurBiophysJ33(7):623–32.

34. Zahn TR, et al. (2004) Dense core vesicle dynamics in Caenorhabditis elegansneuronsandtheroleofkinesinUNC-104.Traffic5(7):544–59.

35. KuralC,etal.(2005)Kinesinanddyneinmoveaperoxisomeinvivo:atug-of-warorcoordinatedmovement?Science308(5727):1469–72.

36. Levi V, SerpinskayaAS, GrattonE, GelfandVI (2006)Organelle transport alongmicrotubules in Xenopus melanophores: evidence for cooperation betweenmultiplemotors.BiophysJ90(1):318–27.

37. Kulic IM, et al. (2008) The role of microtubule movement in bidirectionalorganelletransport.ProcNatlAcadSci105(29):10011–10016.

38. RaiA,etal. (2016)DyneinClusters intoLipidMicrodomainsonPhagosomes toDriveRapidTransporttowardLysosomes.Cell164(4):722–734.

39. Leduc C, et al. (2004) Cooperative extraction of membrane nanotubes bymolecularmotors.ProcNatlAcadSci101(49):17096–101.

40. Campàs O, et al. (2008) Coordination of Kinesin motors pulling on fluidmembranes.BiophysJ94(12):5009–17.

41. Schneider R, Glaser T, Berndt M, Diez S (2013) Using a quartz paraboloid forversatile wide-field TIR microscopy with sub-nanometer localization accuracy.OptExpress21(3):3523.

42. Blainey PC, et al. (2009) Nonspecifically bound proteins spin while diffusingalongDNA.NatStructMolBiol16(12):1224–9.


http://dx.doi.org/10.1101/064246

20

Figure1.Membrane-anchoredkinesin-1motorsdiffusewithsimilardiffusivityasthelipids.(A)Time-lapseseriesoffluorescencerecoveryafterphotobleaching(FRAP)imagesforaSLBwithamolar composition of DOPC:DSPE-PEG-2000-Biotin:DOPE-Atto647n = 99:1:0.05 (Scale bar, 10µm).(B)Representativecurveforthenormalizedfluorescenceintensityafterphotobleachingvs.time (black,N =4SLBs,mean±SD).Thediffusivityof the lipidbilayerwasdetermined tobeDLipid=3.0±0.3µm2/s(mean±SD,N=4independentexperiments,dashedredline,fitaccordingtoRef.(18)).(C)Single-moleculetrajectoriesoffreelydiffusingrKin430-SBP-GFPanchoredonabiotinylatedSLBviastreptavidin(Scalebar,10µm).(D)MSDdata(black,mean±SD)ofdiffusingrKin430-SBP-GFP molecules. The diffusion coefficient was determined to be DKin = 3.1 ± 0.2µm2/s(mean±95%confidenceinterval,N=40trackedmolecules,redline,linearfittothefirsteightpoints).


http://dx.doi.org/10.1101/064246

21

Figure2. In-vitroreconstitutionofamembrane-anchoredglidingmotilityassay.(A)Schematicdrawing (not drawn to scale) of the experimental setup: truncated rat kinesin-1 withstreptavidin-binding-tag (rKin430-SBP) is attached, via streptavidin, to abiotinylatedSLB.Themotorsdiffusively-anchoredon the SLBpropel themicrotubules. (B)Representative ensembleMSD data for the center positions of themicrotubules (mean ± SEM,N≥ 40microtubules) atdifferentmotorconcentrationsin1mMATPand1mMAMP-PNP(onlyfor0.13µMrKin430-SBP).The red arrow indicates increasingmotor concentration. To calculate the linear translocationcomponents(i.e. themicrotubulevelocitiesv) thedatawas fitbyEq.1. (C)Ensemble-averagedmicrotubule gliding velocities for differentmicrotubule lengths, binned into 1 µm intervals, atdifferentmotorconcentrations(mean±95%confidence interval,N≥15microtubules foreachdatapoint).


http://dx.doi.org/10.1101/064246

22

Figure 3. Membrane-anchored kinesin-1 motors slip in the lipid bilayer, while propelling amicrotubule. (A) Time-lapse images for microtubules driven by membrane-anchored motors(upperpanel)andsurface-immobilizedmotors (lowerpanel)withschematicexperimental set-ups on the right (Scale bar, 5 µm). The arrows indicate the transport directions of the glidingmicrotubules.Microtubulespropelledbymembrane-anchoredmotorsdonotcrosseachotherincontrast to microtubules driven by surface-immobilized motors. (B-C) Representativekymographs(invertedcontrast)showingthemovementof individualrKin430-SBP-GFPmotors(dark signals)while propellingmicrotubules at lowmotordensity (B) andhighmotordensity(C).Theredlinesmarkthetrailingendsofthemicrotubulesasguidestotheeye.In(C)aglidingmicrotubulecollideswithanotherpassingmicrotubuleandistemporarilystalled(alsoshownintheschematicsontheright)untiltheothermicrotubuleglidesaway.


http://dx.doi.org/10.1101/064246

23

Figure 4. Theoretical description of the membrane-anchored gliding motility fits well to theexperimental observations. (A) Nanoscopic view of the experimental set-up illustrating thevelocities and frictional forces of microtubules and motors, used to derive the mathematicalmodel.Thedepictedmicrotubulesglideto the leftwithrespect to thesubstrateatvelocityvMT.ThemotorstepsonthemicrotubulewithvelocityvStepandtherebymovesitsanchorintheSLBtothe right at velocity vKin-slip with respect to the substrate. The frictional forces act on themicrotubulesandmotors in thedirectionsopposite to theirmotion. (B)Averagedmicrotubuletransportefficiency(mean±95%confidenceinterval,N≥40microtubulesforeachdatapoint)forvariouskinesin-1surfacedensities(mean±SD,N=5regionsof interest)anddifferentSLBcompositions (0% CH, 20% CH and 60% CH). The data were fitted (solid lines) to themathematicalmodel(Eq.13),withonefreeparameterω(reachofadiffusingkinesin-1motortobindtoamicrotubule,c=0.72µm-3⋅s,R2≥0.95).


http://dx.doi.org/10.1101/064246

24

Supporting Information Transport efficiency of membrane-anchored kinesin-1 motors depends

on motor density and diffusivity

Rahul Grover, Janine Fischer, Friedrich W. Schwarz, Wilhelm J. Walter, Petra Schwille, Stefan Diez

Supporting Figure

Figure S1. Single-molecule stepping velocity of rKin430-SBP-GFP. A) Representative kymograph of single rKin430-SBP-GFP molecules moving on a surface-immobilized microtu-bule. Time is progressing from top to bottom, while the motors (dark signals) move along a microtubule from left to right. B) Histogram of single-molecule velocities with an ensemble average velocity of 0.67 ± 0.14 µm/s (mean ± SD, N = 545 molecules). Schematics shown on top of the histogram.

Supporting Movies Movie S1. Membrane-anchored gliding motility at different motor densities with bulk rKin430-SBP concentrations 0.63 µM (top left panel, S1.1), 0.13 µM (top right panel, S1.2), 0.07 µM (bottom left panel, S1.3), and 0.13 µM in non-hydrolysable ATP analog AMP-PNP (bottom right panel, S1.4; Scale bar, 10 µm). Microtubule gliding velocity decreased with decreasing motor concentration and the respective trajectories became wigglier. In the presence of AMP-PNP, the microtubules anchored to SLBs via rigor motor-binding were diffusing only, without any linear translocation (S1.4). Movie S2. rKin430-SBP gliding motility assay on different substrates: Motors anchored to SLB (left panel, S2.1) and motors immobilized on glass substrate (right panel, S2.2; Scale bar, 10 µm). Movie S3. Addition of cholesterol to the DOPC SLBs reduced the diffusivity of lipids. Time-lapse movie showing FRAP for two different lipid compositions: DOPC:Cholesterol:DSPE-PEG-2000-Biotin = 79:20:1 (20% CH, left panel, S3.1) and DOPC:Cholesterol:DSPE-PEG-2000-Biotin = 39:60:1 (60% CH, right panel, S3.2). Fluorescence recovery of the photobleached region is faster in the SLB with 20% CH as compared to 60% CH (Scale bar, 10 µm). Movie S4. Membrane-anchored gliding motility at different SLB diffusivities: 20 % CH (left panel, S4.1) and 60 % CH (right panel, S4.2). While microtubules do not cross each other when the membrane viscosity is low (left panel) they can cross each other at higher membrane viscosity (right panel, Scale bar, 10 µm).


http://dx.doi.org/10.1101/064246

transport efficiency of membrane-anchored kinesin-1 … · however, the collective transport by...

Documents