• Nenhum resultado encontrado

Behavior of pyrene as a polarity probe in palmitoylsphingomyelin and palmitoylsphingomyelin/cholesterol bilayers: A molecular dynamics simulation study

N/A
N/A
Protected

Academic year: 2021

Share "Behavior of pyrene as a polarity probe in palmitoylsphingomyelin and palmitoylsphingomyelin/cholesterol bilayers: A molecular dynamics simulation study"

Copied!
11
0
0

Texto

(1)

ContentslistsavailableatScienceDirect

Colloids

and

Surfaces

A:

Physicochemical

and

Engineering

Aspects

jo u r n al ho me p ag e :w w w . e l s e v i e r . c o m / l o c a t e / c o l s u r f a

Behavior

of

pyrene

as

a

polarity

probe

in

palmitoylsphingomyelin

and

palmitoylsphingomyelin/cholesterol

bilayers:

A

molecular

dynamics

simulation

study

António

M.T.M.

do

Canto

a

,

Patrícia

D.

Santos

a

,

Jorge

Martins

b

,

Luís

M.S.

Loura

c,d,∗

aCentrodeQuímicadeÉvora,DepartamentodeQuímica,EscoladeCiênciaseTecnologia,UniversidadedeÉvora,RuaRomãoRamalho,59,7000-671Évora, Portugal

bIBB-CBMEandDCBB-FCT,UniversidadedoAlgarve,CampusdeGambelas,8005-139Faro,Portugal

cFaculdadedeFarmácia,UniversidadedeCoimbra,PólodasCiênciasdaSaúde,AzinhagadeSantaComba,3000-548Coimbra,Portugal dCentrodeQuímicadeCoimbra,UniversidadedeCoimbra,RuaLarga,3004-535Coimbra,Portugal

h

i

g

h

l

i

g

h

t

s

•N-Palmitoylsphingomyelin/cholesterol bilayersaresimulatedinpresenceof pyrene.

•The presence of pyrene does not affectthehostbilayerproperties sig-nificantly.

•Pyrene dynamics are considerably slowed down by the addition of cholesterol.

•Pyrene location inside the bilayer hydrocarbonregionisnotaffectedby cholesterol.

•However, pyrene hydration is reduced upon increasing the cholesterolconcentration.

g

r

a

p

h

i

c

a

l

a

b

s

t

r

a

c

t

a

r

t

i

c

l

e

i

n

f

o

Articlehistory: Received25June2014

Receivedinrevisedform5December2014 Accepted7December2014

Availableonline16December2014 Keywords: Cholesterol Lipidbilayer Moleculardynamics Polarity Pyrene Sphingomyelin

a

b

s

t

r

a

c

t

Pyreneisapolycyclicaromatichydrocarbonnotedforitsremarkableopticalspectroscopicproperties. Amongitsusesasafluorescentprobe,measurementoflipidbilayer’sequivalentpolaritythroughthe pyreneHameffectstandsout.Tothiseffect,theratiooftheintensitiesofthefirstandthirdvibronicbands (I1/I3)initsemissionspectrumofpyreneismeasured.However,issuesconcerningthepreciselocation

ofbilayer-insertedpyreneandthepossibilityofprobe-inducedperturbationofhostbilayerproperties arepotentialsourcesofconcerninthisregard.Atomisticmoleculardynamicssimulationsconstitutea usefulmethodforthecharacterizationoflipidmembranesystems,and,inparticular,tounderstandthe behavioroffluorescenceprobesuponincorporationinlipidbilayers.Inthisreport,wepresentadetailed characterizationofthebehaviorofpyreneinfluidN-palmitoylsphingomyelin(PSM)andPSM/cholesterol membranes,withemphasisonthedegreeofproximitybetweentheprobeandwatermoleculesinside bilayers,relatedtotheuseofpyrenetomeasureequivalentlipidbilayerpolarity.Itisconcludedthat pyreneexertsminoreffectsonbilayerproperties,withslightlocaldisorderingbeingapparentforhigh

Abbreviations:ACF,autocorrelationfunction;Chol,cholesterol;DPPC,dipalmitoylphosphatidylcholine;MD,moleculardynamics;MSD,meansquareddisplacement;PC, phosphatidylcholine;POPC,1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine;PSM,N-palmitoylsphingomyelin;SM,sphingomyelin.

∗ Correspondingauthorat:FaculdadedeFarmácia,UniversidadedeCoimbra,PólodasCiênciasdaSaúde,AzinhagadeSantaComba,3000-548Coimbra,Portugal. Tel.:+351239488485;fax:+351239488503.

E-mailaddress:lloura@ff.uc.pt(L.M.S.Loura).

http://dx.doi.org/10.1016/j.colsurfa.2014.12.012

(2)

cholesterolcontent.Whereasrotationandlateraldiffusionofpyrenearegreatlyslowedin cholesterol-richsystems,itsrelativetransverselocationisnotsignificantlyaffected.WhilehydrationofPSMbilayers, assensedbypyrene,isalreadylowcomparedtothatoffluidphosphatidylcholine,itbecomesevensmaller forhighcholesterolmolefractionatthestudiedtemperature.

©2014ElsevierB.V.Allrightsreserved.

1. Introduction

Phospholipidbilayersarethebasicstructuralunitsof biologi-calmembranes,demarcatingtheinteriorandexteriorofcellsand theirorganelles.Asmodelsofbiologicalmembranes,lipidbilayers havebeenamajorresearchtopicinbiophysicalchemistry[1,2]. Oneoftheirmostimportantpropertiesistheirhydrationorpolarity profile,whichshapesthehydrophobicmembranebarrier[3],and therebyconstitutesanenergeticdeterminantforinsertionof pep-tides,proteinsandamphiphilicmolecules,aswellasfortransport ofbothpolarandapolarsolutesacrossthebilayer[4].

Several techniques based on fluorescence spectroscopy are availablefortheestimationofpolarityinlipidbilayers,exploiting the environmentalsensitivity of extrinsic probes, suchas pro-dan,laurdan,dansyl,oranthroyllabeledprobes(e.g.[5,6]).Among thesemethodologies,onethathasbeensurprisinglyseldomused inmembranesystemsisthatbasedontheHameffectofpyrene[7], theso-calledPyscaleofequivalentpolarity[8,9].Pyreneisa poly-cyclicaromatichydrocarbonnotedforitsremarkablefluorescence properties,suchasanunusuallylongexcited-statelifetime(>100ns inavarietyofaeratedsolventsandmicellarormodelmembrane systems),emissionspectrumhighlysensitivetosolventpolarity andconcentration-dependentand/orviscosity-influencedexcimer formation[10].ThePyscaleofpolarityisbasedonthe measure-mentoftheratioofthefluorescenceintensitiesofthefirstand thirdvibronicbands(I1/I3)inthespectraofpyrene.Previously,we demonstratedtheapplicationofthisscaletolipidbilayers com-posedofphosphatidylcholine(PC),bothinpresenceandabsence ofcholesterol(Chol)[11].

However,therearetwocrucialissuesregardinguseofextrinsic membrane probes,and pyrene inparticular: what is their pre-ciselocationwithinthebilayer(andthuswhatregiontheyreport on),andwhethertheyinducesignificantperturbationonthehost bilayer properties. For this purpose, molecular dynamics (MD) simulationshavebeenusefultocharacterizesimultaneously mem-braneprobebehaviorandprobe-inducedperturbationwithatomic detail, for several classes of fluorescent lipophilic probes (see [12,13]forreviews), includingfree pyrene[14,15].Recently,we carriedoutMDsimulationsofpyrenein 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine(POPC)bilayers,withvaryingamounts ofChol[16].Thisworkwasvaluableintheclarificationoftheuse ofthepyreneHamEffecttoeffectivelymeasureequivalentpolarity inmixedlipidbilayerscontainingunsaturatedPCandChol.

Inthisreport,weextendourpreviousMDstudybyapplying this methodology to the study of pyrene inside N-palmitoyl-sphingomyelin(PSM)/Cholbilayers(structuresareshowninFig.1, togetherwiththatofpyrene).AlongsidePCandChol, sphingosine-basedsphingomyelin(SM)isamajorcomponentofmammalian plasmamembranes,andisparticularlyrelevantas(togetherwith Chol)themain constituent oflipidrafts,specialized membrane domainsimplicatedinprocessesincludingcellularsignalingand trafficking[17–19].SMchainstypicallydisplayalargerdegreeof saturationand highermaintransitiontemperaturecomparedto otherlipidclasses[2].Forthisreason,andbecauseoftheirabilityto formanetworkofintermolecularhydrogenbondsinvolvingtheir amideand/orhydroxylgroups,SM-enrichedmembranedomains arehighlyordered.Thisalsoappliestorafts,inwhichthe order-ing effects of Chol contribute to create a very tight molecular

packing.To investigate ifpyrene is wellaccommodated within SM/Cholbilayers,itspreciselocationandeffectonmembrane prop-erties,andtogaininsightsonthepolaritysensedbytheprobein thesesystems,wecarriedoutextensive(300–400ns)simulations ofPSM/Cholbilayers(0mol%,5mol%,20mol%,40mol%,45mol% and50mol%ofthelattercomponent).Thedeterminationofbilayer polaritycorrespondingtoraftsenrichedinSM/Cholisrelevantfor theunderstandingofthesolvationpropertiesoflipidbilayers,that mayinfluencethestructureandfunctioningofproteinsassociated todomain-limitedbiochemicalreactionsandprocesses.Inbilayer hydrationterms,itisexpectedthatbilayerdomainsenrichedin PSM,butwithlowCholcontent(≤5mol%)areofsignificantlyhigher polaritythan thosedomains containingupper Cholproportions (20mol%≤Cholcontent≤50mol%).Whenchanging the propor-tionofCholinPSMbilayers,subtlechangesinbilayerhydration are revealed,togetherwiththe effects ofpyrene in thebilayer orderinganddynamicsofthelipidiccomponents,studiedatthe atomicresolution.Besidestheintrinsiclessfluidcharacteristicsof bilayerdomainsenrichedinPSM/Cholmixtures,thereare unequiv-ocalvariationsinbilayerhydrationthatmustbetakenintoaccount whenanalyzingmembranephenomenainvolvingthespecialized raftdomains.

2. Methods

MDsimulationsandanalysisoftrajectorieswerecarriedout usingtheGROMACS4.6.3package[20–22].PSMunited-atom struc-tureandtopology,asusedbyNiemeläetal.[23],wereobtained throughtheLipidbookwebpage[24].Cholunitedatomstructure andtopologywereadaptedfromthoseofHöltjeetal.[25] (avail-ablefordownload attheGROMACS webpage[26])bychanging themoleculetypesfromCH2/CH3toLP2/LP3,toavoid overcon-densationofthebilayer,asdescribedandtestedelsewhere[27]. Pyrenewasparameterizedasdescribed byHoff[14,28]. Bilayer modelswithvaryingnumbers(showninTable1)ofPSMandChol moleculeswereassembledusingGROMACS4.6.3tools,andfully hydratedwithSPCwater[29].Forthesimulationswith incorpo-ratedprobe,initialstructureswith2or4pyrenemolecules(one ortwoin eachleaflet,respectively)werethenobtainedby ran-domlyinsertingprobemoleculesinsideaPSMorPSM/Cholbilayer withoutreplacementoflipids.

(3)

Table1

Compositionofthesystemsstudiedinthepresentwork.

Cholmolefraction nPSM nChol nwater/(nPSM+nChol)

0 144 – 32.9 5 152 8 37.7 20 120 30 32.2 40 90 60 31.4 45 72 60 34.8 50 72 72 32.9

Prior to the full MD simulation, all systems underwent a

steepest-descentenergyminimizationofthestructure,followed

byasmallMDrun(100ps)toproperlyallowthesolventmolecules

toadjustand/orrelaxaroundthemembrane.Extensivemolecular

dynamicssimulations(300nsfor0,5,and20mol%Chol,400nsfor

40,45,and50mol%Chol)werethencarriedoutunderconstant

numberofparticles,pressureandtemperature,andusingperiodic

boundaryconditions.Pressure(semi-isotropic, 1bar,1.0ps

cou-plingtime)andtemperature(333K,0.1pscouplingtime)control

wascarriedoutusingtheweak-couplingBerendsenschemes[30].

Thetemperature,wellabovethereportedvalueforthemain transi-tionofPSM(∼314K[2]),waschosentoensurethatbilayerswerein aproperliquid-crystallinestate.Allbondlengthswereconstrained totheirequilibriumvalues using theSETTLEalgorithm [31] for waterandtheLINCSalgorithm[32]forallotherbonds.An inte-grationtimestepof2fswasused.VanderWaalsinteractionswere cutoffat1.0nm.Coulombinteractionswerecalculatedusingthe ParticleMeshEwaldmethod[33],withacut-offof1.0nmforthe realspacecomponent.Forvisualizationofstructures/trajectories, VisualMolecularDynamics software(Universityof Illinois) was used[34].

3. Resultsanddiscussion 3.1. Molecularareas

Fig.2depictstheevolutionofthemolecularareasofPSMand Cholforallstudiedsystems(calculatedusingthemethodofHofsäß etal.[35]),whileFig.3illustratesthefinalconfigurationsineach

Fig.2.TimevariationsofthemolecularareasofPSM(inthe0.45–0.6nm2range)andChol(inthe0.2–0.3nm2range)forthesystemswith0,2,and4insertedpyrene molecules(blue,redandgreenlines,respectively).FromAtoF,Cholmolefractionis0,0.05,0.20,0.40,0.45and0.50.(Forinterpretationofthereferencestocolorinthis figurelegend,thereaderisreferredtothewebversionofthisarticle.)

(4)

Fig.3. Final4-pyrenesimulationstructures.Water,PSMandCholmoleculesaredepictedascyan,greenandredlines,respectively.Pyrenemoleculesareshownasvander Waalsrepresentations.FromAtoF,Cholmolefractionis0,0.05,0.20,0.40,0.45and0.50.(Forinterpretationofthereferencestocolorinthisfigurelegend,thereaderis referredtothewebversionofthisarticle.)

case.Acommonfeaturetoallsimulationsisthatconvergenceof theseparametersisachievedrapidly,indicatingthatequilibrium conditionsaremaintainedthroughouttheruns.

Comparisonwithexperimentalvaluesofarea/lipidisvery dif-ficult,becauseoftheextremelylimitednumberofexperimental studies.Toourknowledge,novalues at allhavebeenreported forPSM/Cholbilayers,andasinglevalueof0.464nm2 hasbeen reportedforPSMfromX-raydiffractionat55◦C [36].However, asnotedbytheauthorsofthiswork,thisexperimentalarea/lipid valuewasactuallycalculatedfromthemeasuredbilayer period-icities,assumingapartialspecificvolumeof1.012mL/gat55◦C, correspondingtothatofdipalmitoylphosphatidylcholine(DPPC). However,itiswidelyknownthatfluidDPPCandPSMbilayershave verydifferentphysicalproperties,andtherefore,this “experimen-tal”area/lipid valueis probablyincorrect.Indeed, experimental valuesforothersaturatedSMpurebilayersareconsiderablylarger. Forexample,0.55nm2hasbeenreportedforN-stearoyl-SM[37]. Thesedifficultiesincomparisonwithexperimentalarea/lipid val-ueswerealsofoundand discussedinrecent simulationstudies [38,39].Forthesereasons,weresorttocomparingourmolecular areavalueswithresultsobtainedfromprevioussimulationstudies. The molecular area of pure PSM, averaged for t>50ns, is (0.530±0.014)nm2,whichagreeswellwiththevaluesatT=323K ofNiemeläetal.[23],who obtained(0.52±0.01)nm2 (withthe sameforce-field description),Metcalfand Pandit[38],who cal-culated(0.533±0.004)nm2 (usingslightlydifferentunited-atom parameters)andof Jämbeckand Lyubartsev[39],whoreported (0.541±0.004)nm2 (withanall-atomforcefield).Incorporation

ofeither 2or4 pyrenemoleculesin purePSM leadstono sig-nificantchangesinPSMmoleculararea((0.530±0.013)nm2and (0.529±0.015)nm2areobtained,respectively;seeFig.2A).

Uponloading thePSM bilayerwithincreasingChol content, two effects becomeapparent. First, themolecular areaof PSM decreasesslightly,reachingaminimumof(0.493±0.012)nm2for 20mol%Chol,andincreasesslightlyforhigherCholmolefractions, upto(0.535±0.013)nm2for50mol%Chol.Onepossible explana-tiontothis non-monotonic variation isthat, whereasrelatively low amounts ofChol arecapable ofordering toa small extent purefluidPSMbilayers(similarlytothewell-knowneffectonfluid PCbilayers),toomuchCholwouldactuallycausedisruptionthe organizationofPSM(possiblybyreducinginter-PSMHbonding). However,thisresult isbestinterpretedcautiouslyatthis stage, becauseartificialincreasesinmolecularareaofPClipidshavebeen observedatveryhighCholfractionswhenusingthearea parti-tioningschemeemployedhere,duetoitsassumptionthatthearea fractionsofthetwocomponentsareequaltotheirvolumefractions [40].Inanycase,itisclearfromthisanalysisthatthesystemsare convenientlyequilibratedearlyoninthesimulationsandthatthe CholcondensingeffectinPSMbilayersismuchlesspronounced thanintheirPCcounterparts.

Second, while pyrene incorporation does not affect signifi-cantlylipidmolecularareasforlowCholcontent,itisclearthat, for higher Cholmole fractions (≥40mol%), inclusion of pyrene inducesanincrease, whichismostvisible forthesystemswith four inserted probe molecules. For example, for 50mol% Chol, (0.535±0.013)nm2,(0.539±0.013)nm2and(0.545±0.013)nm2

(5)

areobtainedfromthesimulationswith0,2and4pyrenemolecules, respectively. Thisis a first indication that pyrenemay actually reducetheorderofthesesystems,similarlytowhatweobserved forPOPC/Chol[16].Althoughthesevariationsarestillwithinthe statisticaluncertainty associatedwiththecalculations,itshould bestressedthatthesevaluesareaveragedoverallPSMmolecules (notonlythoseneartheprobemolecules),andlocalprobe disor-deringeffectsaremostprobablymoresignificant.Thisissuewill beaddressedindetailinconnectiontothelipidacylchainorder parameters(Section3.2).

3.2. Massdensityprofiles

Fig.4showsmassdensityprofilesalongthebilayernormalof thedifferentcomponentsofthe4-pyrenesystems.Thedistance betweendensity peaks increases from 4.0nm for purePSM to 4.4nmfor PSM/50mol% Chol.Thisagreeswellwiththe experi-mentalpeak separationsin theelectron densityprofiles, which varybetween4.2nmfor purePSMand 4.5nmforPSM/50mol% Chol[36].Webelievethatthiscomparisonismoremeaningfulthan thatwiththemolecularareasreportedinthisexperimentalstudy, because,unlikethelatterparameter,theexperimentaldensitypeak separationdoesnotdependonfurtherassumptions.

PurePSMbilayersareremarkablycompactforfluidmembranes, because of their characteristic strong intermolecular hydrogen bondingproperties(e.g.[23]andreferencestherein).Therefore,it isnotsurprisingthat,similarlytothemodestvariationsin molecu-larareaoftheprevioussection,changesintheshapeofthedensity profilesuponincreasingCholcontentaresubtle.Inparticular,the

distancebetweenthecenterofthebilayerandthelipid/water inter-face(operationallydefinedastheplaneforwhichlipidandwater massdensitiesareequal)isremarkablyinvariant(mostlyinthe z∼2.4–2.5nmrange)forallsystemsshowninFig.4.Thisagrees witharecentsynchrotronX-raydiffractionstudy,whichreports thatthethicknessofliquid-orderedstructuresofsphingomyelin andcholesteroldoesnotdiffersignificantlyfromfluidphasesof bilayersformedbythepurephospholipid[41].Similarly,pyrene transversedistributionisvirtuallyunaffectedbythepresenceof Chol,anditsmaximumisintherangeofz∼1.2–1.3nmforall com-positions.SmallchangesacrossthepanelsofFig.4(e.g.,deeper penetrationintothebilayercenterforsomepeaks)areprobablynot significantandreflectthebehaviorofindividualpyrenemolecules. Theasymmetricaldensityprofileofpyreneinthe20mol% simula-tionistheresultofararetranslocationeventbetweenmonolayers duringequilibration,whichledtothefinalconfigurationofFig.3C with3(1)moleculesintheupper(lower)leaflet,aswellas3:1peak ratiointhepyrenedistributioninFig.4C.

Payingcloseattention,onecanobservethatthewaterprofile extendstoslightlydeeperlocationsinthemembraneinthe Chol-poorsystems(0and5mol%Chol),comparedtothosewith20mol% or more of this component. This is most important regarding pyrene-reportedpolarity,anditisreflectedinthedegreeofoverlap ofthepyreneandwaterdistributions.Awaytoquantifythiseffect istocalculateanaveragewaterdensitysensedbypyrene,usingthe equation



(water)



pyrene=



pyrene(z)water(z)dz



pyrene(z)dz (1) 0 200 400 600 800 1000 1200 ρ/kgm-3 0 200 400 600 800 1000 1200 ρ/kgm-3 0 200 400 600 800 1000 1200 -4 -2 0 2 4 ρ/kgm-3 z/nm -4 -2 0 2 z/nm 4

A

B

C

D

E F

Fig.4.MassdensityprofilesofPSM(red),Chol(magenta),pyrene(multipliedby10forbettervisualization;green),andwater(blue)inthesystemswith4insertedpyrene molecules.FromAtoF,Cholmolefractionis0,0.05,0.20,0.40,0.45and0.50.(Forinterpretationofthereferencestocolorinthisfigurelegend,thereaderisreferredtothe webversionofthisarticle.)

(6)

0 2 4 6 8 10 12 0 10 20 30 40 50 Chol mol% <ρ (w at er )> py re n e / kg m -3

Fig.5.Variationofthewaterdensityassensedbypyrene(squares;calculatedusing Eq.(1)),forvaryingbilayercomposition.Thedottedlineisjustaguidetotheeye.

ThisisillustratedinFig.5,whichdemonstratesthatloadingthe bilayerwith≥20mol%Cholreducesthreefoldtheaveragedensity ofwatersensedbypyrene,comparedtotheChol-freeand5mol% Cholsystems.Thedottedlineshowninthefigureisreminiscentof thevariationexpectedforliquidordered/liquiddisorderedphase separationinthe∼6–22mol%compositionrange,asdescribedfrom variationsinlipidlateraldiffusioncoefficientsmeasuredbyNMR [42]. No particular significance should begiven to this fact, as twofewcompositionsareaddressedinthisstudy,andespecially becausemeso-orlarge-scalephaseseparationcannotbe appreci-atedinatomisticsimulationsofbilayersmadeupof100–200lipid molecules.

3.3. PSMorderparametersandCholtiltinthebilayers

For our united atomforcefield, deuterium order parameters (SCD)forsaturated(SCDsat)andunsaturated(SunsatCD )carbonsare deter-mined fromthe order tensor elements Sab usingthe following relations[43]: −Ssat CD= 2 3Sxx+ 1 3Syy (2) −Sunsat CD = 1 4Szz+ 3 4Syy+ √ 3 2 Sxy (3) where,inturn

Sab= 123cosacosb−ıab a,b=x,y,z (4) Inthelatterequation,a(orb)istheanglemadebyath(or bth)molecularaxiswiththebilayernormalandıabistheKronecker delta( denotesbothensembleandtimeaveraging).Themolecular axesforthenthCH2are[44]:

z:vectorfromCn−1toCn+1;

y:vectorperpendiculartoz,lyingintheplanedefinedbyCn1,Cn, andCn+1,pointingfromCnawayfrom1/2(Cn+1+Cn−1);

x:vectorperpendiculartobothzandy.

Figs.6and7showtheorderparameterscalculatedforthe sph-ingosineandpalmitoylchains(respectively)ofPSM.Theprofiles

0.0 0.1 0.2 0.3 0.4 3 5 7 9 11 13 15 17 -SCD n

A

0.0 0.1 0.2 0.3 0.4 3 5 7 9 11 13 15 17 -SCD n

B

0.0 0.1 0.2 0.3 0.4 3 5 7 9 11 13 15 17 -SCD n

C

0.0 0.1 0.2 0.3 0.4 3 5 7 9 11 13 15 17 -SCD n

D

0.0 0.1 0.2 0.3 0.4 3 5 7 9 11 13 15 17 -SCD n

E

0.0 0.1 0.2 0.3 0.4 3 5 7 9 11 13 15 17 -SCD n

F

Fig.6.Deuteriumorderparameter−SCDofthesphingosinechainofPSM,forthesystemswith0,2,and4insertedpyrenemolecules(blue,redandgreenlines,respectively). FromAtoF,Cholmolefractionis0,0.05,0.20,0.40,0.45and0.50.(Forinterpretationofthereferencestocolorinthisfigurelegend,thereaderisreferredtothewebversion ofthisarticle.)

(7)

0.0 0.1 0.2 0.3 0.4 1 3 5 7 9 11 13 15 -SCD n

A

0.0 0.1 0.2 0.3 0.4 1 3 5 7 9 11 13 15 -SCD n

B

0.0 0.1 0.2 0.3 0.4 1 3 5 7 9 11 13 15 -SCD n

C

0.0 0.1 0.2 0.3 0.4 1 3 5 7 9 11 13 15 -SCD n

D

0.0 0.1 0.2 0.3 0.4 1 3 5 7 9 11 13 15 -SCD n

E

0.0 0.1 0.2 0.3 0.4 1 3 5 7 9 11 13 15 -SCD n

F

Fig.7. Deuteriumorderparameter−SCDofthepalmitoylchainofPSM,forthesystemswith0,2,and4insertedpyrenemolecules(blue,redandgreenlines,respectively). FromAtoF,Cholmolefractionis0,0.05,0.20,0.40,0.45and0.50.(Forinterpretationofthereferencestocolorinthisfigurelegend,thereaderisreferredtothewebversion ofthisarticle.)

obtainedhereforPSMandPSM/20mol%Cholaresomewhathigher thantheexperimentalprofilesobtainedby1HNMRforPSMwith perdeuteratedpalmitoylchain[45], butagree verycloselywith thosereportedinothersimulationworks([23]forPSM;[39]for bothsystems).UponincreasingthemolefractionofCholinthe sys-temfrom0to0.40,theorderparametersforeachsegmentincrease steadily.HigherproportionsofCholleadtonodiscerniblechanges inthe−SCDprofiles.Thisconfirmsthatthenon-monotonic vari-ationofthePSMmolecularareawasindeedartifactual,andthat thereisnoevidenceforamaximuminorderforintermediate com-positions.

Lookingnowattheeffectofinsertingpyrenemoleculesonthe orderparameterprofiles,it appearsthat,for smalltomoderate Cholcontent(upto20mol%),pyrenemayinduceanincreasein −SCD,mostvisiblyintheintermediateregionofthechains(where pyrenepredominantly resides)and inthe4-pyrenesimulations. For40mol%Chol,nosignificantdifferencesare observed,while for higher Chol concentration, loadingthe bilayerwith pyrene moleculeleadstoareductionintheorderparameter,mostnotably inthelowerend of thechains(where pyreneisrarelylocated, thereforeleadingtotheappearanceofvoidsunderneaththeprobe molecules).Again,whilethiseffectismodestoverall,itcouldbe significantinthevicinityoftheprobemolecules.Toaddressthis possibility,we calculatedacylchain orderparameters for vary-ingdistancetothenearestpyreneprobe.Theresults,illustrated inFig.8forpalmitoylchainsinthe2-pyrenesimulations,show thatwhereastheeffectofpyrenemoleculesontheorderofnearby

lipidmoleculesissmallforpurePSMand5mol%Chol,a signifi-cantreductionisobservedforthenearestlipidneighbors(located at<0.6nm)athigherCholconcentrations.Thisisrelevantforthe interpretationoffluorescencemeasurements,becausetheprobes onlyreportontheirlocalphysical-chemicalproperties.Thus,itcan beinferredthatpyreneisanespeciallyadequateprobeofPSM/Chol membranepropertiesforlowCholfractions,butreportsalower localorderforhighCholcontent,comparedtothebulkbilayer.

Wealsoexaminedthepossibilityofaneffectofpyreneonthe tiltofthelongaxisofChol.Weverifiedthatverysmallchangesin theangulardistributionofthisvectorrelativetothebilayer nor-malwereproduceduponloadingeachsystemwith2or4pyrene molecules. While adding pyrene to a PSM/5mol% Chol bilayer decreasestheaveragetiltfrom27.3◦(nopyrene)to26.0◦(4pyrene molecules),smallervariations(<0.5◦inallcases)areobtainedinthe othersystems,withvisuallyindistinguishabledistributions cen-teredaround∼24◦for20mol%Choland∼22for≥40mol%Chol (datanotshown).

3.4. Pyreneorientationandrotationaldynamics

Asillustratedin thesnapshotsofFig.3,pyrenetendsto ori-entitslongaxisroughly alongthebilayernormal.Fig.9shows theangulardistributionsofthelongaxistilt.Whileasmall frac-tionofconformationsinthelessorderedsystems(0and5mol% Chol)displayslargetiltangles(visibleinsomeoftheprobes’ ori-entationsinFig.3AandB),thesearenotveryfrequentmolecular

(8)

0.0 0.1 0.2 0.3 0.4 1 3 5 7 9 11 13 15 -SCD n

A

0.0 0.1 0.2 0.3 0.4 1 3 5 7 9 11 13 15 -SCD n

B

0.0 0.1 0.2 0.3 0.4 1 3 5 7 9 11 13 15 -SCD n

C

0.0 0.1 0.2 0.3 0.4 1 3 5 7 9 11 13 15 -SCD n

D

0.0 0.1 0.2 0.3 0.4 1 3 5 7 9 11 13 15 -SCD n

E

0.0 0.1 0.2 0.3 0.4 1 3 5 7 9 11 13 15 -SCD n

F

Fig.8.Deuteriumorderparameter−SCDofthepalmitoylchainofPSM,forthesystemswith2insertedpyrenemolecules.Theblue,redandgreenlinesaretheorderparameter profilesaveragedforpalmitoylchainslocatedatR<0.6nm,0.6nm<R<1.2nm,andR>1.2nmfromthenearestpyrenemolecule,respectively.FromAtoF,Cholmolefraction is0,0.05,0.20,0.40,0.45and0.50.(Forinterpretationofthereferencestocolorinthisfigurelegend,thereaderisreferredtothewebversionofthisarticle.)

orientations. In accordance, the average tilt angles of the dis-tributions shown in Fig. 9 (35–36◦ for 0–5mol% Chol) are considerablysmallerthanthose weobtainedin POPC(48◦)and evenPOPC/20mol%Chol(41◦),andalmostcomparabletothatin POPC/40mol%Chol(32◦)[16].ForhigherCholcontent,the distribu-tionsshiftprogressivelytolowertiltangles,albeittoasmallextent (averagesare30◦,29◦,27◦and26◦for20,40,45and50mol%Chol, respectively).In theselatter compositions, themost significant featureis thathighlytiltedconfigurations(>60◦)becomehighly improbable. 0.00 0.01 0.02 0.03 0 20 40 60θ/deg80 P(θ)

Fig.9. Angulardistributionsofthepyrenelongaxistilt,relativetothebilayer nor-mal,forthe4pyrenesimulationswith0(black),5(blue),20(red),40(magenta), 45(green)and50(cyan)mol%Chol.(Forinterpretationofthereferencestocolorin thisfigurelegend,thereaderisreferredtothewebversionofthisarticle.)

Relatedtothequestionoftheangulardistributionsisthe ampli-tudeand dynamicsof pyrenerotation. Fig.10shows rotational autocorrelationfunctions(ACFs)ofthelongaxisofpyreneforthe differentstudiedsystems,definedas

C(t)=P2(cos()) (5) 0.0 0.2 0.4 0.6 0.8 1.0 0 5 10 15 20 25 ACF(t) t/ns

Fig.10.AveragerotationalACFsofthepyrenelongaxisinthesimulationswith 4insertedpyrenemoleculesofthesystemswith0(black),5mol%(blue),20mol% (red),40mol%(magenta),45mol%(green)and50mol%(cyan)ofChol.(For inter-pretationofthereferencestocolorinthisfigurelegend,thereaderisreferredtothe webversionofthisarticle.)

(9)

where()istheanglebetweentheprobelongaxisattimes andt+,andP2(x)=(3x21)/2isthesecondLegendre polyno-mial.Averagingisperformedover,whichassumingasufficiently ergodictrajectory,isanapproximationoftheensembleaverage.

Fig.10 showsonly therotational ACFsfor the4-pyrene sys-tems,beingthe2-pyrenecurvessimilar(datanotshown).Because averagingiscarriedoutoverasmallnumberofprobemolecules, theyaremostlikelyaffectedbylimitedsampling.Still,itis pos-sibletoobservethat,inallcases,theACFsdropalmostinstantly to∼0.4–0.6,probablybecauseofveryfastmotions.Followingthis verysteepdecrease,furtherdecayisslowandlimitedinextent, andafiniteresidualvalueofACFisobtainedinallcases.The rota-tionalmotionofpyrenecanbedescribedbya“wobbling-in-cone” model[46].Inthismodel,theprobehasconsiderableorientational freedomforconfigurationswherethelongaxisliesinsideacone ofagivensemi-anglemax(withconeaxisnormaltothebilayer plane),hencetheverysteepinitialdecrease.However,verytilted configurations(>max)areforbidden(asthetiltangular distri-butionsofFig.9show,configurationswithtiltangles>60◦ are indeedhighlyimprobable),andhencetheACFsdonotdecayto zeroevenatlongtimes.HigherACF residualvaluescorrespond tonarrowerdistributions,andareobservedforthemoreordered Chol-richbilayers.Fromoursimulations,thereislittle(probably non-significant)differencebetweentheACFsfor0and5mol%Chol, aswellasbetweenthoseofthesystemswith45and50mol%Chol, withtheothersystemsdisplayinganintermediatebehavior.This steadilyslowerandmorerestrictedrotationalmotionoftheprobe agreeswiththeincreasedorderparameterprofilesobservedforthe mostChol-enrichedsystems.

3.5. Pyrenelateraldiffusion

LateraldiffusioncoefficientsDwerecalculatedfromthe two-dimensionalmeansquareddisplacement(MSD),usingtheEinstein relation D=1 4t→∞lim dMSD(t) dt (6) Inturn,MSDisdefinedby MSD(t)=|| ri(t+t0)− ri(t0)||2 (7) where riisthe(x,y)positionofthecenterofmassofmoleculei ofagivenspecies,theaveragingiscarriedoutoverallmoleculesof thiskindandtimeoriginst0.Toeliminatenoiseduetofluctuations inthecenterofmassofeachmonolayer,allMSDanalyseswere car-riedoutusingtrajectorieswithfixedcenterofmassofoneofthe monolayers,andthefinalresultisaveragedoverthetwoleaflets. Fig.11showsMSDforpyreneandthehostlipidsforvarying con-centrationofChol,whilethecorrespondingDvaluesaregivenin Table2.

ThesignificanceofMSDplotsandaccuratecalculationoflateral diffusioninmembranesremains,toagreatextent,acontroversial problem.Itdependslargelyontheavailabletimewindow[47,48]. Samplingproblemsaremoreimportantinlateraldiffusionthan insomeotherproperties,becauseitinvolveslarge-scalemotions ofwholemoleculesratherthanlimitedrange/segmentalmotions (likethoseinvolvedinlipidacylchainorprobelongaxis orienta-tion).Forrelativelyshorttimes,lipiddiffusion(asperceivedbyMSD variation)ismainlyduetoconformationalchangesofthe hydro-carbonchainsratherthandiffusionoftheentiremolecule[47],and thereforeitsmeaninganditsrelationshiptoexperimental observ-ablesaresomewhatquestionable.Asthesimulationtimescaleis extended,thepossibilityofobservingtruediffusionalmotionalong thebilayerplaneisincreased.Oursimulationsspan300ns,which isclearlyenoughfordescribinglateraldiffusioninthemore dis-orderedsystems,butpossiblyinsufficientforChol-richbilayers,

0 1 2 3 4 5 6 MSD/nm2

A

0.1 0.2 0.3 0.4 0.5

B

0.0 0.1 0.2 0.3 0.4 0.5 0 5 10 t/ns 15

C

Fig.11.Meansquaredisplacementsofpyrene(A;systemswith4pyrenemolecules), PSM(B;systemswithnoprobe)andChol(C;systemswithnoprobe)for0(black), 5mol%(blue),20mol%(red),40mol%(magenta),and45mol%(green)ofChol.(For interpretationofthereferencestocolorinthisfigurelegend,thereaderisreferred tothewebversionofthisarticle.)

wheredynamicsisconsiderablysloweddown.Probablyowingto this,MSDsofpyreneinthe50mol%Cholsystemwereunphysical, andthedatashownexcludethiscomposition.

As expected, the diffusion coefficients of the three species decreasewithincreasingCholcontentandconcomitantorderingof thebilayer.ThevaluesobtainedforPSMarecomparable(∼2-fold lower)totheexperimentaldataofFilippovetal.[42]),obtained ineggyolksphingomyelin/Chol.Thereasonsforthequantitative differencecouldbenon-equivalenceinbilayerpropertiesarising fromthedifferenceincomposition(accordingtothevendor[49], chicken eggyolk sphingomyelin hasat least 3%of unsaturated fattyacids,whichcouldaccountforincreasedfluidity),theabove mentionedsamplingdifficulties,and/orsmallforcefield parame-terissues.Accuratelateraldiffusioncoefficientsareverydifficultto obtain,becausesmallchangesinthebilayerorder/fluidityleadto largevariationsintheDvalues.Forexample,thevaluesmeasured byFilippovetal.[42]forT=323Kareactuallyquantitativelyvery similartoourscalculatedat333K.Mostimportantly,theseauthors reportthatloadingwith∼40mol%CholreducesD(SM)bya fac-torof2–3,whichisconsistentwithourobservationthatD(PSM) isreducedbylessthananorderofmagnitudeasCholcontentis increasedto45mol%(asexpectedfor Chol-richbilayers,which, althoughmoreordered,retaintheirlateraldiffusion-related fluid-ity[50]).Accordingtoourdata,asimilarvariationoccursforChol, whoseDvaluesinfactmatchcloselythoseofPSMforagivenbilayer composition.

Thiscontrastswiththebehaviorofpyrene,forwhichincreasing Cholmolefractionleadstoadramaticdecreaseoftwofullorders ofmagnitudeinthelateraldiffusioncoefficient.Althoughprecise estimationofD(pyrene)isdifficultbecauseofthesmallnumberof diffusingmoleculesusedforaveraging,evenfromvisualinspection ofFig.11itisclearthatChol-inducedorderingofthebilayeraffects

(10)

Table2

Lateraldiffusioncoefficientsofpyrene(systemswith4pyrenemolecules),PSM,andChol(systemswithnoprobe).

Species D/(10−8cm2/s)

0mol%Chol 5mol%Chol 20mol%Chol 40mol%Chol 45mol%Chol

Pyrene 111±11 54±11 41±9 1.2±0.2 1.1±0.2

PSM 5.4±1.6 5.5±1.5 6.3±0.3 2.6±1.0 1.4±1.0

Chol – 5.1±0.5 4.2±0.4 2.5±0.4 1.6±0.3

diffusionofpyrenetoamuchlargerextentthanthoseofPSMor

Chol.ThevaluesofDofthethreespeciesareinfactsimilarforvery

highCholconcentration,andthe“anomaly”isthefactthatpyrene

exhibitsveryfastdiffusioninpurePSM(∼20timesfasterthanPSM),

whichisreadilyreduced uponloadingofthebilayerwithChol.

Pyreneisasmallerapolarmolecule(202.25g/mol),whichresides

inthemoredisordered,hydrophobicregionofthebilayer.Because

itisunabletoestablishstrongelectrostatic/Hbondinginteractions

withwater/lipidcomponents,unlikePSMandChol,itpossesses

muchmorediffusionalfreedomthanthelatterforlow-Chol

sys-tems.However,increasingthemolefractionofCholleadstohigher

orderparametervalues,mostnotablyinthemiddleregionofthe

lipidchains,where pyreneis predominantlylocated.Thisisthe

probablereasonwhydiffusionofpyreneismoreseverelyreduced

thanthatofthelipidspeciesuponincreasingCholmolefraction.

4. Concludingremarks

4.1. Behaviorofpyreneasamembraneprobe

From thebilayer propertiesaddressed in this study,pyrene

exertsminoreffectsonPSMbilayerproperties,anddoesnotcause

significantoverallperturbation.Itcausesaslight,mainlylocal

dis-orderingofChol-rich(≥20mol%)bilayers,whichdoesnotpreclude

theuseofpyreneinmembranestudies.Ontheotherhand,although

thecondensingeffectofCholismuchlesssignificantinPSM(very

smallareadecrease/−SCD increase,aspurePSMalreadydisplays

lowmoleculararea/high−SCDprofile)thaninPCbilayers,pyrene

dynamicsisstillgreatlyhindereduponaddingChol.Thisis

prob-ablybecausetheorderingeffectofCholismostnotablyfeltinthe

middleregionofthelipidchains,wherepyreneispredominantly

located.

4.2. Polaritysensedbypyrene

Hydration sensed by pyrene in pure PSM bilayers is

sub-stantially low. For comparison, we calculated (pyrene) water,

as defined in Eq. (1), for our data obtained in POPC and

POPC/Chol[16].ThepresentvaluesforPSMandPSM/5mol%Chol (∼10–11kgm−3)are wellbelow those calculated in purePOPC (∼45–50kgm−3) and even in POPC/20mol% Chol (∼25kgm−3). Remarkably,(pyrene) waterinPSM/Choldropstoevenlower val-ues(∼3–4kgm−3)forhighCholmolefraction(≥20mol%).Thisisan importantcontributionintheunderstandingofthephysical prop-ertiesoflipidrafts,whicharemainlycomposedofSMandChol.The apolarnatureofthesedomainsmayberelevant,e.g.fortheirrole inthesortingofmembraneproteins.

ThislowhydrationagreeswithpreliminaryI1/I3valuesobtained inPOPC/CholandeggSM/Cholmultilamellarvesicles(0–45mol% Chol)ofvaryinglipidiccomposition[51].Thevaluesobtainedin POPC/Cholat25◦C(whichliebetween1.17±0.01forpurePOPC, and 1.11±0.01 for POPC/45mol% Chol) are consistently higher thanthosemeasuredforpurePSMinthefluidphase(whichlie between1.08±0.01at45◦Cand1.06±0.01at55◦C),inaccordance withthesimulationresults.However,theexperimentalvariation ofI1/I3 withPSM/Chol compositionat fixedtemperatureis not paralleltothatofthesimulated(pyrene) water.Fortheclosest

temperature addressed experimentally, 55◦C, I1/I3 shows little variation for ≤25mol% Chol (while (pyrene) water decreases markedly),andforhigherCholcontentitactuallyincreases(while (pyrene) water showslittlechange).Whilethesevariationsare notdivergent,theypointtothepossibilitythatalterationsinI1/I3 mayreflectothercomposition-influencedfactors(thatwould,on theirown,ledtoanincreaseinI1/I3withCholcontent)inadditionto waterpenetrationalone.Regardingthis,itisnoteworthythat,upon increasingCholmolefraction,SMbilayersundergoavery signifi-cantincreaseindipolepotential[52].Itistemptingtohypothesize thatmembranedipolepotentialcouldplay aninfluenceonI1/I3 similartothatofthesolventdipolemomentinisotropicmedia, especially for suchdehydrated environmentsas the interior of PSM/Cholbilayers.Inthisscenario,acontinuousincreaseindipole potential,combined withthevariation in hydration feltby the probeasdescribedhere,couldresultintheexperimentally deter-mineddependenceofI1/I3onChol-concentration.

Acknowledgements

L.M.S.L.,A.M.T.M.C,andP.D.S.acknowledgefundingbyFEDER, throughtheCOMPETEprogram,andbyFCT(Fundac¸ãoparaa Ciên-ciaea Tecnologia,Portugal),projectreference FCOMP-01-0124-FEDER-010787(FCTPTDC/QUI-QUI/098198/2008).P.D.S. acknowl-edgesagrantunderthissameproject.L.M.S.L.acknowledges addi-tionalfundingbyFCT,projectreferencePEst-OE/QUI/UI0313/2014. J.M. acknowledges thesubsidy by national Portuguese funding throughFCT–Fundac¸ãoparaaCiênciaeaTecnologia,projectsref. PEst-OE/EQB/LA0023/2013,andPTDC/QUI-BIQ/112943/2009.

References

[1]G.Cevc,D.Marsh,PhospholipidBilayers:PhysicalPrinciplesandModels, Wiley-Interscience,NewYork,1987.

[2]D.Marsh,HandbookofLipidBilayers,seconded.,CRCPress,BocaRaton,FL, 2013.

[3]O.H.Griffith,P.J.Dehlinger,S.P.Van,Shapeofthehydrophobicbarrierof phos-pholipidbilayers.Evidenceforwaterpenetrationinbiologicalmembranes,J. Membr.Biol.15(1974)159–192.

[4]D.Marsh,Membranewater-penetrationprofilesfromspinlabels,Eur.Biophys. J.31(2002)559–562.

[5]C.D.Stubbs,C.Ho,S.J.Slater,Fluorescencetechniquesforprobingwater pene-trationintolipidbilayers,J.Fluoresc.5(1995)19–28.

[6]T.Parasassi,E.Gratton,Membranelipiddomainsanddynamicsasdetectedby laurdanfluorescence,J.Fluoresc.5(1995)59–69.

[7]A.Nakajima,Solventeffectonthevibrationalstructureofthefluorescenceand absorptionspectraofpyrene,Bull.Chem.Soc.Jpn.44(1971)3272–3277.

[8]D.C.Dong,M.A.Winnik,ThePyscaleofsolventpolarity.Solventeffectsonthe vibronicfinestructureofpyrenefluorescenceandempiricalcorrelationswith ETandYvalues,Photochem.Photobiol.35(1982)17–21.

[9]D.C.Dong,M.A.Winnik,ThePyscaleofsolventpolarities,Can.J.Chem.62 (1984)2560–2665.

[10]E.Melo,J.Martins,Kineticsofbimolecularreactionsinmodelbilayersand biologicalmembranes.Acriticalreview,Biophys.Chem.123(2006)77–94.

[11]D.Arrais,J.Martins,Bilayerpolarityanditsthermaldependencyintheoandd

phasesofbinaryphosphatidylcholine/cholesterolmixtures,Biochim.Biophys. Acta1768(2007)2914–2922.

[12]L.M.S.Loura,J.P.PratesRamalho,Fluorescentmembraneprobes’behaviorin lipidbilayers:insightsfrommoleculardynamicssimulations,Biophys.Rev.1 (2009)141–148.

[13]L.M.S.Loura,J.P.PratesRamalho,Recentdevelopmentsinmoleculardynamics simulationsoffluorescentmembraneprobes,Molecules16(2011)5437–5452.

(11)

[14]B.Hoff,E.Strandberg,A.S.Ulrich,D.P.Tieleman,C.Posten,2H-NMRstudyand

moleculardynamicssimulationofthelocation,alignment,andmobilityof pyreneinPOPCbilayers,Biophys.J.88(2005)1818–1827.

[15]J. ˇCurdová,P. ˇCapková,J.Pláˇsek,J.Repáková,I.Vattulainen,Freepyreneprobes ingelandfluidmembranes:perspectivethroughatomisticsimulations,J.Phys. Chem.B111(2007)3640–3650.

[16]L.M.S.Loura,A.M.T.MartinsdoCanto,J.Martins,Sensinghydrationand behav-iorofpyreneinPOPCandPOPC/cholesterolbilayers:amoleculardynamics study,Biochim.Biophys.Acta1828(2013)1094–1101.

[17]K.Simons,E.Ikonen,Functionalraftsincellmembranes,Nature387(1997) 569–572.

[18]K.Simons,D.Toomre,Lipidraftsandsignaltransduction,Nat.Rev.Mol.Cell Biol.1(2000)31–39.

[19]D.Lingwood,K.Simons,Lipidraftsasamembrane-organizingprinciple,Science 327(2010)46–50.

[20]H.J.C.Berendsen,D.vanderSpoel,R.vanDrunen,GROMACS:a message-passingparallelmoleculardynamicsimplementation,Comput.Phys.Commun. 91(1995)43–56.

[21]D.vanderSpoel,E.Lindhal,B.Hess,G.Groenhof,A.E.Mark,H.J.C.Berendsen, GROMACS:fast,flexibleandfree,J.Comput.Chem.26(2005)1701–1718.

[22]B.Hess,C.Kutzner,D.vanderSpoel,E.Lindahl,GROMACS4algorithmsfor highlyefficient,load-balanced,andscalablemolecularsimulation,J.Chem. TheoryComput.4(2008)435–447.

[23]P. Niemelä, M.T. Hyvönen, I. Vattulainen, Structure and dynamics of sphingomyelinbilayer:insightgainedthroughsystematiccomparisonto phos-phatidylcholine,Biophys.J.87(2004)2976–2989.

[24]J.Doma ´nski,P.Stansfeld,M.S.P.Sansom,O.Beckstein,Lipidbook:apublic repos-itoryforforcefieldparametersusedinmembranesimulations,J.Membr.Biol. 236(2010)255–258.

[25]M.Holtje,T.Forster,B.Brandt,T.Engels,W.vonRybinski,H.D.Holtje, Molec-ulardynamicssimulationsofstratumcorneumlipidmodels:fattyacidsand cholesterol,Biochim.Biophys.Acta1511(2001)156–167.

[26]http://www.gromacs.org/@api/deki/files/29/=cholesterol.tgz (accessed 21.06.14).

[27]J.R.Robalo,J.P.PratesRamalho,L.M.S.Loura,NBD-labeledcholesterolanalogues inphospholipidbilayers:insightsfrommoleculardynamics,J.Phys.Chem.B 117(2013)13731–13742.

[28]B.Hoff,AromateninPhospholipid-Doppelschichten:Molekulardynamische SimulationenundExperimentelleValidierung(Ph.D.thesis),Universitätsverlag Karlsruhe,Karlsruhe,Germany,2005.

[29]H.J.C.Berendsen,J.P.M.Postma,W.F.vanGunsteren,J.Hermans,Interaction modelsforwaterinrelationtoproteinhydration,in:B.Pullman(Ed.), Inter-molecularForces,Reidel,Dordrecht,TheNetherlands,1981,pp.331–342.

[30]H.J.C.Berendsen,J.P.M.Postma,A.DiNola,J.R.Haak,Moleculardynamicswith couplingtoanexternalbath,J.Chem.Phys.81(1984)3684–3690.

[31]S.Miyamoto,P.A.Kollman,SETTLE:ananalyticalversionoftheSHAKEand RATTLE algorithmsfor rigid water models,J. Comput.Chem. 13 (1992) 952–962.

[32]B.Hess,H.Bekker,H.J.C.Berendsen,J.G.E.M.Fraaije,LINCS:alinearconstraint solverformolecularsimulations,J.Comput.Chem.18(1997)1463–1472.

[33]U.Essman,L.Perela,M.L.Berkowitz,T.Darden,H.Lee,L.G.Pedersen,Asmooth particlemeshEwaldmethod,J.Chem.Phys.103(1995)8577–8593.

[34]W.Humphrey,A.Dalke,K.Schulten,VMD:visualmoleculardynamics,J.Mol. Graph.14(1996)33–38.

[35]C.Hofsäß,E.Lindahl,O.Edholm,Moleculardynamicssimulationsof phospho-lipidbilayerswithcholesterol,Biophys.J.84(2003)2192–2206.

[36]P.R.Maulik,G.G.Shipley,N-palmitoylsphingomyelinbilayers:structureand interactionswithcholesterolanddipalmitoylphosphatidylcholine, Biochem-istry35(1996)8025–8034.

[37]P.R.Maulik,P.K.Sripada,G.G.Shipley,Structureandthermotropicproperties ofhydratedN-stearoylsphingomyelinbilayermembranes,Biochim.Biophys. Acta1062(1991)211–219.

[38]R.Metcalf,S.A.Pandit,Mixingpropertiesofsphingomyelinceramidebilayers: asimulationstudy,J.Phys.Chem.B116(2012)4500–4509.

[39]J.P.M.Jämbeck,A.P.Lyubartsev,Anotherpieceofthemembranepuzzle: extend-ingslipidsfurther,J.Chem.TheoryComput.9(2013)774–784.

[40]M.Alwarawrah,J.A. Dai,J.Y.Huang, Amolecularviewofthecholesterol condensing effect in DOPC lipid bilayers, J. Phys. Chem. B 114 (2010) 7516–7523.

[41]P.J.Quinn,Structureofsphingomyelinbilayersandcomplexeswithcholesterol formingmembranerafts,Langmuir29(2013)9447–9456.

[42]A.Filippov,G.Orädd,G.Lindblom,Theeffectofcholesterolonthelateral dif-fusionofphospholipidsinorientedbilayers,Biophys.J.84(2003)3079–3086.

[43]J.P. Douliez,A.Leonard,E.J. Dufourc,Restatementoforderparametersin biomembranes:calculationofCCbondorderparametersfromCDquadrupolar splittings,Biophys.J.68(1995)1727–1739.

[44]D.P.Tieleman,D.vanderSpoel,H.J.C.Berendsen,Moleculardynamics sim-ulations of dodecylphosphocholine micelles at three different aggregate sizes:micellarstructureandchainrelaxation,J.Phys.Chem.B104(2000) 6380–6388.

[45]T.Bartels,R.S.Lankalapalli,R.Bittman,K.Beyer,M.F.Brown,Raftlike mix-turesofsphingomyelinandcholesterolinvestigatedbysolid-state2H-NMR

spectroscopy,J.Am.Chem.Soc.130(2008)14521–14532.

[46]K.KinositaJr.,S.Kawato,A.Ikegami,Atheoryoffluorescencepolarizationdecay inmembranes,Biophys.J.20(1977)289–305.

[47]A.P.Lyubartsev,A.L.Rabinovich,Recentdevelopmentincomputersimulations oflipidbilayers,SoftMatter7(2011)25–39.

[48]M.Javanainen,H.Hammaren,L. Monticelli,J.-H. Jeon,M.S. Miettinen,H. Martinez-Seara,R.Metzler,I.Vattulainen,Anomalousandnormaldiffusionof proteinsandlipidsincrowdedlipidmembranes,FaradayDiscuss.161(2013) 397–417.

[49]http://www.avantilipids.com/index.php?option=comcontent&view=article& id=437&Itemid=277&catnumber=860061(accessed21.06.14).

[50]O.G.Mouritsen,Life–AsaMatterofFat,Springer,NewYork,2005.

[51]D.Arrais,J.Martins,Phospholipid/cholesterolbinarymixtures:polarity varia-tionswithcompositionandtemperature,Eur.Biophys.J.40(Suppl.1)(2011) 195.

[52]T.J.McIntosh,S.A.Simon,D.Needham,C.H.Huang,Interbilayerinteractions betweensphingomyelinandsphingomyelincholesterolbilayers,Biochemistry 31(1992)2020–2024.

Imagem

Fig. 1. Structures of PSM (A), Chol (B), and pyrene (C).
Fig. 2 depicts the evolution of the molecular areas of PSM and Chol for all studied systems (calculated using the method of Hofsäß et al
Fig. 3. Final 4-pyrene simulation structures. Water, PSM and Chol molecules are depicted as cyan, green and red lines, respectively
Fig. 4 shows mass density profiles along the bilayer normal of the different components of the 4-pyrene systems
+5

Referências

Documentos relacionados

CONCLUSÕES E PROPOSTAS PARA TRABALHOS FUTUROS Nesta dissertação, foi apresentada uma avaliação do desempenho das funções de proteção aplicadas em transformadores de potência,

Ao ser atingida a data limite para a entrega das autorizações, definimos o horário em que cada aluno (consoante a turma e respectivo número de alunos autorizados) se deslocaria

Os e-marketplaces trazem significativa redução de custos ao favorecer aspetos como a maior transparência de preços, mecanismos eficientes de mercado e preços, comparabilidade de

O estudo de caso foi desenvolvido com usuários que trabalham em sistemas integrados, através de metodologia desenvolvida, para descrever os resultados obtidos na

Poder ter assistido a diferentes aulas de piano, desde Outubro de 2013 a Maio de 2014, de diferentes alunos, incluindo vários que acabaram por não vir a participar no

Then, this work follows the procedures, train of thought and decision making done during the development process of the image acquisition system, deep learning

No Capítulo 2 são caracterizados os museus de Portugal, de acordo com os dados estatísticos disponíveis desde 2000 até 2015, que incluem a evolução do número de museus,