Citation
w-Conotoxins: A look at the relationship between conformational equilibria and calcium channel specificity

Material Information

Title:
w-Conotoxins: A look at the relationship between conformational equilibria and calcium channel specificity
Creator:
Arrar, Mehrnoosh
Publication Date:
Language:
English

Subjects

Subjects / Keywords:
Atoms ( jstor )
Calcium channels ( jstor )
Crystal structure ( jstor )
Disulfides ( jstor )
Electric potential ( jstor )
Isomerization ( jstor )
Molecular dynamics ( jstor )
Pain ( jstor )
Simulations ( jstor )
Trajectories ( jstor )
Molecular dynamics
Molecular dynamics--Computer simulation
Genre:
Undergraduate Honors Thesis

Notes

Abstract:
The omega conotoxins are disulfide-rich peptides isolated from the venom of marine cone snails in the genus Conus. Most of the conotoxins in this family range from 24 to 27 amino acid residues and have a compact four-loop structure maintained by three disulfide bridges. The conotoxins' unprecedented specificity and potency as voltage sensitive calcium channel (VSCC) inhibitors enables the cone snails to paralyze their prey. Effectively inhibiting the neuronal (N-type) subtype VSCC is considered a novel form of pain management. Inhibition of the P/Q-type VSCCs in mammals results in adverse physiological effects. Molecular dynamics (MD) simulations of three omega conotoxins MVIIA, GVIA, and MVIIC with varying specificities were performed in order to characterize structural differences in sub-type specificity. A position 10 mutation of N-type inhibitors (GVIA and MVIIA) created a new specificty for an N-type VSCC variant, and so we examine the effect of the this mutation on the conformational equilibria of the peptides. ( en )
General Note:
Awarded Bachelor of Science; Graduated May 4, 2010 summa cum laude. Major: Chemistry
General Note:
Advisor: Dr. Adrian E. Roitberg
General Note:
College/School: College of Liberal Arts and Sciences

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Mehrnoosh Arrar. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

! -Conotoxins:Alookattherelationshipbetween conformationalequilibriaandcalciumchannel specicity MehrnooshArrar DepartmentofChemistry UniversityofFlorida Gainesville,FL April2010

PAGE 2

Abstract Theomegaconotoxinsaredisulde-richpeptidesisolatedf romthevenomofmarine conesnailsinthegenus Conus .Mostoftheconotoxinsinthisfamilyrangefrom24to 27aminoacidresiduesandhaveacompactfour-loopstructur emaintainedbythree disuldebridges.Theconotoxins'unprecedentedspecici tyandpotencyasvoltage sensitivecalciumchannel(VSCC)inhibitorsenablestheco nesnailstoparalyzetheir prey.Eectivelyinhibitingtheneuronal(N-type)subtype VSCCisconsideredanovel formofpainmanagement.InhibitionoftheP/Q-typeVSCCsin mammalsresultsin adversephysiologicaleects.Moleculardynamics(MD)simulationsofthree conotoxinsMVIIA,GVIA,and MVIICwithvaryingspecicitieswereperformedinordertoc haracterizestructural dierencesinsub-typespecicity.Aposition10mutationo fN-typeinhibitors(GVIA andMVIIA)createdanewspecictyforanN-typeVSCCvariant ,andsoweexamine theeectofthethismutationontheconformationalequilib riaofthepeptides. ii

PAGE 3

Acknowledgements IoweabigthankstomyresearchadvisorAdrianRoitberg,who hastaughtme agreatdealthelastfouryears.ThefacultyintheDepartmen tofChemistryand mypeers,particularyMalloryGeraceandJamesCrooks,have beenasourceofdiscussionandcollaborationinthisprojectandinothersIhav eworkedonduringmy undergraduatecareer.Ofcoursenoneofthiswouldbepossib lewithoutmyfamily, whosesupporthasalwaysbeeninvaluable.Thankyou. iii

PAGE 4

ContentsAbstract ii Acknowledgements iii 1Introduction 1 1.1BiologicalSignicance..........................11.2Structuralcharacteristicsof -conotoxins................3 1.3BindingtoVoltageGatedCalciumChannels............ ..4 2Methods 8 2.1MolecularDynamics...........................82.2Clustering.................................9 3ResultsandDiscussion 11 3.1Structuralstability............................1 1 3.2LoopDynamics..............................123.3ComparisontoexperimentalNOEstructuraldata....... ....14 3.3.1ViolationsofNOEdistanceupperbounds...........1 5 3.3.1.1GVIAandGVIA[K10].................15 iv

PAGE 5

3.3.1.2MVIAandMVIA[K10].................173.3.1.3MVIIC.........................17 3.4Disuldebondisomerization....................... 18 3.5ShiftinLoopConformations.......................2 0 4Conclusions 23 Bibliography 25 v

PAGE 6

Chapter1Introduction1.1BiologicalSignicanceTheconotoxinsareafamilyofdisulde-richpeptidesisola tedfromthevenomof predatorymarineconesnailsofgenus Conus .Theiranalgesicpropertieshavealready beenexploredandtheyareregardedasapowerfulmeanstound erstandingthemechanismofpaininhibitioninmammalsviablockingvoltagesen sitivecalciumchannels (VSCC)[1,2,3].Currently,oneconotoxinMVIIAisbeinguse dasapainreliever commerciallyknownasZiconotide[4].TheConusfamilyrepresentsadiversenumberofspecies,and anomenclaturefor distinguishingthevarioustypesofconotoxinshasbeenest ablished[5,6].Theconotoxinshavebeendividedintosuperfamilies,distinguishi ngthetypeofdisuldefold, andeachconotoxinisassignedaRomannumeralinitsnametha tcorrespondstoits superfamily.ConotoxinsarealsoassignedaGreekletterco rrespondingtothetypeof voltagesensitivechanneltheytarget.Thenameofaconotox inhasathirdcomponent 1

PAGE 7

thatidentiesitsspecies,andafourthidentifyingtheord erinwhichitwasfound; -GVIAforexampleisaconotoxinisolatedfrom C.geographus ,withaframework characteristicoftheO-superfamily(VIandVII),thattarg etscalciumchannels,and istherstpeptideofthistypeidentied(A).Withinthe -conotoxins,theidentiedsequencesarehypervaried,how everthecharacteristicfour-loopfoldisconservedamongallconotoxin s,aswellasaTyr-13residue inbindingloop2.Thisso-calledcysteineknotisbelievedt obekeyinmolecularrecognitionbyVSCCs[7].The -conotoxinstargetasubsetofneuronal(N-type)VSCCs withunprecedentedspecicity,makingthempopularmodels inthedesignofmore potentandnonaddictiveanalgesicdrugs.Untilrecently,c onotoxinswerebelievedto berigidpeptides,howevertheirdistinctspecicitiesand potenciesasinhibitorsofNtypeVSCCmaybeattributedtoconformationalrexibilityth atplaysasignicantrole inbinding[8].N-typeVSCCsarefoundthroughoutthecentra lnervoussystem,and theN-typecurrentisresponsibleforthereleaseofneurotr ansmitters,thusproducing ananalgesiceectinmammals.Calciumchannelswiththe 1Bsubunitproduce theseN-typecurrents.Whiletheconotoxinsgenerallydobl ockN-typeVSCCs,some ofthesepeptidesalsoinhibitothertypesofVSCCs,suchast heP/Q-type,whichare calciumchannelswiththe 1asubunit.InhibitionoftheseVSCCscanbelethal,or causeadverseeectsduetotoxicity,howeverligandsthatb indtoP/Q-typeVSCCs ingeneralarealsousedasdiagnostictoolsforneuraldegen erativesyndromessuchas Lambert-Eatonmyasthenicsyndrome[9].Thusthecharacter izationofthestructure anddynamicsofconotoxinsiskeytoamorecontrolledapproa chtomanytherapies. 2

PAGE 8

1.2Structuralcharacteristicsof -conotoxins Agreatdealofprogresshasbeenmadeduringthelasttwodeca desincharacterizingthestructuresoftheomegaconotoxinsandstructurally similarcompoundsusing manyspectroscopictechniquesandthex-raycrystalstruct ures.InFigure1.1welist thestructurestodatethathavebeendepositedintheprotei ndatabank(PDB).The structuralmotifcommoninallofthecharacterizationsofo megaconotoxinsisthe triple-strandedbetasheetcross-linkedbythethreedisul debridges[8,10].Recent studiesonMVIIAhaveraisednewquestionsaboutthestabili tyofconotoxins,particularlyinreferencetotheideaofthepeptidesas"rigids caolds".Inastructural analysisMVIIAwasobservedhavetwopopulationsofChidihe dralanglesofthe Cys8-Cys20disuldebond,andthehypothesisofdisuldebo ndisomerizationbeing necessaryforthebindingofthepeptidetoVSCCshasyettobe conrmedasoneof theconformationshasnotyetbeenisolated.Theideaofconf ormationalexchange beingthecauseofbroadresonanceswasmentionedasearlyas 1996byNielsenet al[10].Thesebroadresonancesarepredominantlylocatedo nLoop2,comprisedof residues9through14,andhaveasignicantimpactontheint erpretationofspectroscopydata[8].Whiletheomegaconotoxinshaveaconsensusfour-loopstruc turewithatriplestrandedbetasheet,therelativeorientationsoftheloops ,especiallyloops2and 4,varyamongthecharacterizedcrystalstructures[10].Th eseorientationshavenot yetbeenlinkedtoanyVSCCsubtypespecicity,althoughith asbeenobservedthat amostspecicconotoxinCVIDhasauniqueloop4curvedtowar dsloop2,whilein GVIAloop4pointsoutwards[11].Thenatureofresiduesidec hainsinconotoxins 3

PAGE 9

alsomakesitdiculttopredictthespecicity.TheGVIAbin dingloop(loop2) forexample,hasmanyhydroxylsidechains,whereasMVIIAha spositivelycharged sidechains;yetbothbindtoN-typeVSCCs.Whilemanyresidu e-replacementsand functionalassayshaveaddedtremendouslytoourcurrentkn owledgeofimportant residuesintheconotoxins,thesestudiesoftenlackcorres pondingstructuralanalysis, withtheassumptionthatthemutationdoesnotsignicantly aecttheconotoxin structure(seeRef[11]foracomprehensivereview).Thusac hangeinpIC 50 values couldbeduetoadirectinteractionwiththeVSCC,orduetoan entropiceectif thatresidueservestostabilizetheoverallstructure.Moleculardynamics(MD)wasusedtosampletheconformation alspaceofthreeconotoxins:MVIIA,MVIIC,andGVIA.MVIIAandCareisolatedfrom C.magnus and haveahighsequencehomologybuttargetNandP/QtypeVSCCs, respectively. GVIAisfromthesamesuperfamilyastheothertwoconotoxins ,butisolatedfrom C. geographus ,aspreviouslymentioned.AlthoughitalsotargetstheN-ty peVSCC,it doessowithlesspotencythanMVIIA,when 125 I-MVIIAisusedasthedisplacement ligand[11].Figure1.2liststheconotoxinsofthisstudywi ththeirsequencesdivided intotherespectivefourloops,andtheirVSCCspecicities 1 1.3BindingtoVoltageGatedCalciumChannelsSystematicAlaninescansofconotoxinshaveestablishedth attheTyr13hydroxyl groupiscrucialtobindingtoanyVSCC[7].Atwo-pointbindi ngmechanismof 1 ThePDBidsmarkedwithoneasteriskwerethestartingstruct uresforMD.**ThisPDBhas constraintsondisuldebridges 4

PAGE 10

Figure1.1: Theavailableexperimentalstructuraldatatodate 5

PAGE 11

conotoxinstoN-typecalciumchannels,withabasicresidue onloop1andahydrogen bondinginteractionwithTyr13hasalsobeenproposed[7,11 ,12].Thehighsequence identityofMVIIAandMVIICoersaninterestingcasestudyf orthespecicityof conotoxinstoVSCCsubtypes.Radioligandbindingassayson MVIIA-Canalogs,with N-andC-terminalhalvesofthepeptidesswappedhavesugges tedthattheimportant residuesformolecularrecognitionbytheN-typeVSCCmustl ieontheN-terminalhalf oftheconotoxin,whereasnodistinctionhasyetbeenmadere gardingasequence-local setofresidueskeytobindingtoP/Q-typeVSCCs[13].Incont rast,anotherstudy inwhichallloophybridsofMVIIAandMVIICwereevaluatedfo rtheiranityand potencyasN-orP/Q-typeblockershasindicatedthatloops2 and4actconcertedlyin thebindingmechanismofconotoxins,asbothACACandCACAwe remoreeective blockersforP/Q-andN-typeVSCCsrespectively,thantheot herhybridpeptidesin whichloops2and4werefrompeptideswithdierentN-orP/Qtypespecicities [11]. ArecentstudyidentiedaVSCCresistanttomanyN-typeVSCC inhibitors,and Figure1.2: Conotoxinsequencesanddisuldebonds 6

PAGE 12

identiedanomegaconotoxinCVIDthateectivelyinhibits thisN-typevariant.The samestudyshowedthatconotoxinswithaLysatposition10bi ndedtothisN-type VSCCmoreeectively,despitethelackofanynoticeablecha ngeinoverallstructure [14].ThustheMDsimulationsofthethreeconotoxinsandtwo mutantpeptides willaordamuchneededevaluationoftheconformationsoft heseinhibitors.We characterizethedynamicsofconotoxinsandhowtheypertai nto(i)specicitiesfor N-versusP/Q-typeVSCCs,and(ii)newN-typevariantVSCCsp ecicityafterthe K10mutation. 7

PAGE 13

Chapter2Methods2.1MolecularDynamicsThecysteineknotscaoldoftheconotoxinsmakesthemanint erestingsystemfor classicalmoleculardynamics(MD).Usingtheclassicalequ ationsofmotion,andan empiricalforceeld,MDsimulationsprovideatomictrajec torieswhichcanbeused toextractinformationabouttheinternalmotionsofaprote in[15].AcommonproblemencounteredinMDsimulationshoweverpertainstotheis sueofsampling[16,17]. WhileMDisausefulwaytosimulatephysicalmotionsofaprot einorruid,one trajectoryisunlikelytotraversemanybarriersintheener gylandscapeofaprotein, andthusmanytrajectories,orenhancedsamplingtechnique sareoftenrequiredto obtainstatisticallysignicantinformationabouttheacc essibleminimaintheconformationalspaceoftheprotein[18,19].Asalloftheconotoxi nshavethreedisulde bondsrestrainingthefoldofthepeptide,thiscommonissue ofcompletesamplingis mitigatedinthesesimulations.TheAmber99SBforceelda simplementedinAm8

PAGE 14

ber10wasusedforallsimulations[20,21].Fiveseparate80 nsMDsimulationswere performedontheconotoxins,usingexperimentallycharact erizedstructuresfromthe PDB.Thecurrentavailableexperimentalstructurestodate withPDBIDsforthe conotoxinsinthisstudyareoutlinedinFigure1.1.TheconotoxinsweresimulatedinanexplicitTIP3Pwateroct ahedron.Thesystem wasrstheatedto300Katconstantvolume,withaweak(10kca l/mol)restraint onthepeptidefor50ps,witha2fstimestep.Alangevintherm ostatwasusedwith acollisionfrequencyof1ps 1 .Thesystemwasthenequilibratedfor100pswith constantpressureandperiodicboundaries,againwithawea krestraintonthepeptide.TheproductionMDrunswereatconstanttemperaturean dpressure,witha10 Angstromcutofornonbondedatompairelectrostaticcalcu lations,andtheSHAKE algorithmwasusedtorestrainnon-polarhydrogens.Partic almeshewald(PME)was usedforthecalculationoflongrangeelectrostaticsforth eproductionMDrun.The equilibrationwasdoneusingtheAmbermoduleSander.2.2ClusteringForpopulationanalysisoftheMDtrajectories,weclustere dthelast75nsofeach trajectoryusingtheMeansalgorithm,asimplementedinthe AmberToolsmodule ptraj[20].Tendierentalgorithmsweretestedonasmaller sampletrajectory,and clusteringwasevaluatedbasedonvecommonlyusedcriteri a[13]:thepseudoFstatistic(pSF);theDavies-Bouldinindex(DBI);theratio ofexplainedvariation(sum ofsquaresregression)tothetotalsumofsquares(SSR/SST) ;populationsizes;visual inspectionofclusters[22]. 9

PAGE 15

Figure2.1: IntialMDstructures:Peptidebackbonesareshownasribbon s(GVIAblue;GVIA[K10] red;MVIIAorange;MVIIA[K10]green;MVIICpurple).Loops2 and4areontheupperrightand left,anddisuldebondsareshowninyellow. 10

PAGE 16

Chapter3ResultsandDiscussion3.1StructuralstabilityTherootmeansquareddeviation(RMSD)ofthepositionsofth ebackboneatoms duringdynamicsfromtheirinitialpositionsareshowninFi gure3.1.AstheRMSD overtimedoesrerectastableconformationineachoftheMDr uns,animportant distinctionamongtheconotoxinsistheaverageRMSDfromth einitialstructure. Forbothmutantpeptides,thestablestructurefromtheMDis closertotheinitial structurethantheirrespectivewild-typeMDruns.MVIIAin particularhasan averageRMSDof4.85 0.54 AfromitsinitialstructurewhileMVIIA[K10]has anaverageRMSDof1.51 0.27 A.AlthoughbothMVIIAandGVIAsampled conformationsconsistentlybetween3to4 Afromtheirinitialstructures,MVIIC structureswerealwayswithin2 Aoftheinitialstructure.TheMDresultssuggest thatbothMVIIAandGVIAhaveagreaterconformationalrexib ilityascompared totheirrespectivemutantstructures,andtotheP/Q-typeV SCCtargetingpeptide 11

PAGE 17

MVIIC. Figure3.1: EvolutionofRMSDofpeptidebackboneatomsfromtheinitial structuresduringMD 3.2LoopDynamicsTheaverageatomicructuationsforeachresidueoftheconot oxinsareshowninFigure3.2.Theside-by-sidecomparisonshowsthatthemutatio ndoesnotchangethe globaldynamicsofthepeptides,thatistosaythattheovera lltrendsinlooprexibilitydonotchange.ForGVIA,themutationstabilizesthepept ideratheruniformly, ascanbeseenfromtheparticularlydecreasedructuationso ftheterminiandthe residuesofloop2(9to14).Thereisalsoaninterestingshif tinthemostmobile residueofloop2afterthemutation,fromArg10intheGVIAsi mulationtoTyr13 12

PAGE 18

inGVIA[K10],whichistheresiduekeyinbinding,asmention edpreviously. TheR10KmutationinMVIIAappearstohaveparticularlystab ilizedtheresiduesof loop1toabouthalftheaverageructuationsinthewild-type simulation.Theeect ofthemutationonloop2inMVIIAisunlikethatofGVIA.Rathe rthantheentire loopbecomingmorestable,Lys10,Met12andTyr13ructuates lightlymoreinthe mutantstructure.Theseareidentiedaskeyresiduesinmut agenisisstudies[7,11]. Itshouldalsobenotedthatforbothsimulationsallthecyst eineresidues,except20 and25inMVIIA[K10],ructuatelessaftertheR10Kmutation. Thiswillbediscussed ingreaterdetailinregardstotheideaofdisuldebondisom erizationinconotoxins. ThedynamicsofMVIICloopsdiermostfrombothGVIAandMVII Aatloop3 Figure3.2: Averageructuationsofthebackboneatomsbyresidueforthe conotoxinsduringMD. (residues17,18,19).WhilebothN-typetargetingpeptides showminimalructuations inloop3,theP/Q-typetargetingpeptidehaspeakructuatio nsinthisregion.This isparticularlyinterestingconsideringthehighsequence identitybetweenMVIICand MVIIA.Itseemsthataparticularloopsequencehaslessofan impactonitsdynamics,asloop3isidenticalinMVIIAandMVIIC,exceptforposi tion17(athreoninevs. 13

PAGE 19

serine),butratherthatinteractionswiththeotherloopsd eterminethesedynamics. Thisisalsothecasewhenconsideringthatasinglemutation inloop2oftheother twoconotoxinsalteredthedynamicsofallfourloops.3.3ComparisontoexperimentalNOEstructural data AsthecalculatedRMSDoftheconotoxinsthroughoutthedyna micsrepresentsasignicantdeviationfromtheinitialPDBstructure,particul arlyforMVIIAandGVIA (roughly3and4 A,respectively),weusedtheoriginaldistancerestraints fromnuclearoverhausereect(NOE)signals,asreportedinthePDB foreachconotoxin (asindicatedinFigure1.1)asanadditionalcomparisontoe xperimentaldata.For allthebackbonehydrogendistancesinthePDB,thecorrespo ndingdistanceswere measuredoverthe80nsofMD.ItshouldbenotedthatfromtheR MSDplotsin Figure3.1weconsidertherst5nsofthetrajectorytobeani nitialequilibration period,andforthisreasonweevaluatedonlythestructures fromthelast75ns.The NOEsignalintensitiesvarywith r 6 ,where r istheinterprotondistance.Thus,asa semi-quantitavemeasureoftheagreementbetweenexperime ntalpredictionandthe structuressampledinourdynamics,wecalculatedthecorre spondingensembleaverage andcomparedtheinversesixthrootofthisvaluetotheexper imentally determineddistance.Althoughspectroscopicmeasuresvar yintheirsignal-to-noise ratios,generallytheinterprotonupper-bounddistancesc anbepredictedforseparationsofupto5 Aandwith 0.5 Aaccuracy[23].Thesecomparisonsarereported 14

PAGE 20

inFigure3.3.InthisguretheNOEupperboundsfrompdbdata arealongthe x-axisandalongthey-axisaretheMDdistances.The y = x linerepresentsanexact correlationbetweenthetwomethods,andtheerrorbarsshow thestandarddeviation ofthedistanceduringtheMDtrajectory.Weidentifythosed istancesgreaterthan 0.5 AoftheexperimentalupperboundstobeNOEviolations.This analysisisnot exact,astheNOEdistanceupperboundsdependonaclassica tionofsignalstrength asweak,medium,orstrong,anddierentspectroscopistsus edierentrangeswhen depositingthedataintothePDB.Inanycase,byanalyzingth eseapparentviolations, wenoticesomegeneraltrendsinthedegreetowhichthesimul ationagreeswiththe experimentalresults.3.3.1ViolationsofNOEdistanceupperbounds3.3.1.1GVIAandGVIA[K10]ForGVIA,43NH-NHdistanceswerereportedinthePDB,6ofwhi chwereviolatedby ourMDaverages.RecallingthesequenceofGVIA,wehighligh tthefactthatallofthe NOEviolationsinvolveeitheroneoftheCysresidues,oraCy s-adjacentresidue.After measuringthebackbonedihedralanglesfortheseresidues, weidentifymorethanone distinctclusteronaRamachandranplotforoneorbothofthe residuesoftheseamide pairs.FortheGVIA[K10]simulation,oftheavailable28NHNHdistances,wefound onlyoneviolation.Thisagaininvoledacysteineresidue(C ys8-Tyr27),anditshould benotedthattheexperimentalpredictionforthisdistance was7.0 A,corresponding toaweakNOEsignal.Allofthepreviouslymentionedresidue sinvolvedinthewild15

PAGE 21

Figure3.3: ComparisonofNOEdistances(alongthex-axes)andthecorre spondingMDdistances (alongthey-axes),asdiscussedinthetext.They=xlinethu scorrespondstoexperimentalupper boundsforthedistances,andtheerrorbarsshowthestandar ddeviationofthedistancesduring eachMDtrajectory,forthelast75ns.(A)GVIA;(B)GVIA[K10 ];(C)MVIIA;(D)MVIIA[K10]; (E)MVIIC 16

PAGE 22

typeNOEviolationswithmultipleclustersonRamachandran plotshaveoneclear clusterofphi-psidihedralsinthemutantsimulation.This supportsourobservation thattherexibilityinconformationsfoundforGVIAisrestr ictedwiththeO10K mutation.3.3.1.2MVIAandMVIA[K10]AsimilartrendinthenumberofNOEviolationswasobservedf ortheMVIIAwildtypeandmutantsimulations.Inthewild-typesimulation,w eidentifyfromthe28 availabledistances8ofwhichviolatetheexperimentalpre diction.Allbutoneofthese involveeitheroneofthecysteineresidues,oraCys-adjace ntresidue.Theoutlierin thistrendwasthedistancebetweenArg10andLeu11,forwhic hwecalculated 4.53 0.13,whereastheNOEpredictionwas4.0 A.ForMVIIA[K10]therewere43 interamidedistancesinthePDB,onlytwoofwhichwereviola tedbythesimulation. BothdistancesinvolvedCys25.3.3.1.3MVIICNomutantstructuresofMVIICweresimulated,asthepurpose ofexploringthe structureofthisconotoxinisasamodelforP/Q-typeVSCCbl ockersasopposedto N-type.Itisinterestingtonote,howeverthatbasedonboth theRMSDovertime andthenumberofNOEviolationsforMVIICitsconformationa lrexibilityappears morecloselyrelatedtothatofthemutantN-typeVSCCblocke rsGVIAandMVIIA. Ofthe33experimentaldistancesreported,only3wereviola tedand,aswasthecase forMVIIA[K10],alldistancesinvolvedthelastcysteinere sidue.Takencollectively, thevecomparisonstoexperimentaldatasuggestadiscrepa ncyinthedisuldebond 17

PAGE 23

dynamics.Thewild-type/mutantcomparisonsindicatethat thebackbonedihedrals nearthedisuldebondssamplearelocalizedinasingleconf ormationaftertheposition 10mutation.3.4DisuldebondisomerizationInFigure3.4weshowthedihedralangles(CB-S-S'-CB')ofth ethreedisuldebonds duringthelast75nsofeachMDtrajectory.Thetypicaldihed ralangleforadisulde bondis+90degrees.Theconotoxinshavetwoouterdisuldeb onds,aslabeledin Figure3.5,whichcorrespondtotherstandsecondcysteine residues.Thedisuride bondingnetworkisi+3,withibeingthecardinalnumberofth ecysteineresidue.The bondinthecenteroftheknotisbetweenthesecondandfthcy steineresidues,and thisbondisgenerally-90whiletheouterdisuldebondsare +90degrees.Fromthe gureitisclearthatthedisuldebondbetweenthesecondan dfthcysteineshas noisomerizationeventsinanyoftheconotoxins.Themutati onatposition10tends tostabilizeoneconformationforthedisuldebonds.ThereisageneralconsensusintheliteratureofthebroadNO EpeaksforMVIIA andGVIAatloop2[8,11,10].Disuldebondisomerizationha sbeenproposedby Atkinsonetal.in2000asapossiblecauseofconformational exchange,particularly atloop2.Aninterestingidearegardingthedynamicsofthec onotoxinswouldbethat thedihedralanglesofthedisuldebondswereinsomewayrel atedtotheoverallloop orientationsoftheconotoxins.Ascanbeseeninthegureso ftheaveragestructures, thereisarangeofopentoclosedconformations,withrespec ttotheorientationof loops2and4.Weclusteredeachtrajectorytoanalyzethedis uldebondtrendsin 18

PAGE 24

Figure3.4: Dihedralanglesforthethreedisuldebondsintheconotoxi ns.Fromlefttoright,the disuldebondsarebetween(i)rstandfourth;(ii)seconda ndfth;and(ii)thirdandsixthcysteine residues.Duetotheknotfoldoftheconotoxins,disuldebo nd(iii)isbetween(i)and(ii).The panelsfromtoptobottomare(A)GVIA,(B)GVIA[K10],(C)MVI IIA,(D)MVIIA[K10],(E)MVIIC. 19

PAGE 25

eachconformation,buttherewasnodirectcorrelation.Weo bservedisuldebond rippinginboththeclosedandopenclusters.Instead,thein creasedructuationof thedihedralanglesofthedisuldebondsisagoodmarkeroft heconformational rexibilityofthepeptide. Figure3.5: Averagestructuresfrom75nsMDdisplayedinaribbonrepres entation,coloredbyloops ((A)GVIA,(B)GVIA[K10](C)MVIIA,(D)MVIIA[K10],(E)MVII C).Loopcolorsare:(1)orange, (2)blue,(3)green,and(4)red.EachN-andC-terminusislab eledaccordingly,andthedisulde bondsareshowninyellowandnumberedinboxA.3.5ShiftinLoopConformationsTofurtherexploretherangeofconformationssampleddurin gtheMDtrajectories, weattemptedtocharacterizethestructuresbasedonloop24orientations.Because thecriticalbindingresidueshavebeenidentiedtobeonth esetwoloops,oneproposedbindingmechanismfortheconotoxinstotheVSCCshasb eenthetwo-pronged mechanism[7,11,12].Wewouldexpect,then,thatdierenti albindingtoVSCC sub-typeswouldoccurwhentheconotoxinshavedistinctloo p2-4distances.The initalpdbstructures,superimposedinFigure2.1,donotsh owsuchatremendous 20

PAGE 26

dierenceintheloop2-4orientationsofthepeptides.Usin gtheMDtrajectories, wechosetworesiduesaslabelpointsoneachloop,thecarbon alphasonresidues10 and21,andthenanalyzedthedistributionofthisdistancei neachsimulation.The results,showninFigure3.6indicateashiftintheconforma tionalensembleofthetwo N-typeblockersMVIIAandGVIAafterthemutationatpositio n10.Whilethereis aslightoverlapintheloop2-4distancesforthemutantandw ild-typehistograms,it isnotthecasethattherearetwostableconformationsandon eismoreenergetically favorableafterthemutation,butinsteadanewminimaisfou ndintheconformationallandscape.Indeed,thedataforGVIAsuggestthatthe rearetwoloop2-4 conformationsinthewild-type,butthatneitheristhenews tructureobservedafter theK10mutation.Supposingthetwo-prongedbindingmechan ism,thissuggeststhat theN-typeVSCCvariant,whichonlyMVIIA[K10]andGVIA[K10 ]selectivelybind to,hasadierentstructure,likelywithtwobindingpocket sclosertogetherthanin the 1bsubunit,whichistargetedbythewildtypeconotoxinswhi chhaveamore openconformationwithrespecttoloops2and4. 21

PAGE 27

Figure3.6: Histogramsofthedistancesbetweencarbonalphasonresidu es10and21inMVIIA (top),GVIA(middle)andMVIIC(bottom).Theredlinesindic atewildtypepeptides,andblue linesrepresenttheK10mutants. 22

PAGE 28

Chapter4ConclusionsWithconservedfour-loopfoldsandhypervariablesequence s,theconotoxinsserveas aninterestingclassofpeptidesthatselectivelybindtova rioussubtypesofVSCCs. Fromexperimentalbindingandmutationalstudies,itwasdi scoveredthataLys atposition10forbothGVIAandMVIIA(N-typeVSCCblockers) resultedina newspecicityforanN-typevariant.Asthismutationinvol vedaswitchfroma hydroxyprolineinGVIAandanarginineinMVIIA,tolysine,w ewerecuriousabout thephysicalbasisforthisnewspecicity.Usingvariousan alysesofallatommolecular dynamicssimulations,wehaveidentiedtheeectofthemut ationtobeashiftinthe loop2-4orientationofthepeptides,fromanopentoclosedc onformation.Wealso foundthattherewasadecreaseinconformationalrexibilit yinbothpeptidesafterthe K10mutation,mostnotablyaccompaniedbythestabilizatio nofthedisuldebonds inthepeptides,correlatinghigherspecicitytoanarrowe rconformationalminimum. TheoverallbackbonedynamicsoftheN-typeblockersbecame moresimilartothe P/Q-typeVSCCblockerMVIICaftertheK10mutation.Asthese mutantpeptides 23

PAGE 29

didnotshowanyincreaseinbindingtoP/Q-typeVSCCs,wecan inferthatthe specicityoftheconotoxinwithinaparticularsubtypeisr elatedtoitsbackbone conformationalrexibility,butthatthesidechainslargel ydeterminethegeneralsubtypetarget. 24

PAGE 30

Bibliography [1]LJCruzandBMOlivera.Calciumchannelantagonists.ome ga-conotoxindenes anewhighanitysite. JournalofBiologicalChemistry ,261(14):6230{6233,May 1986. [2]BaldomeroM.Olivera,JeanRivier,CraigClark,Cecilia A.Ramilo,GloriaP. Corpuz,FeC.Abogadie,E.EdwardMena,ScottR.Woodward,Da vidR. Hillyard,andLourdesJ.Cruz.Diversityofconusneuropept ides. Science 249(4966):257{263,July1990. [3]GeraldW.Zamponi,RichardJ.Lewis,SlobodanM.Todorov ic,StephenP. Arneric,andTerranceP.Snutch.Roleofvoltage-gatedcalc iumchannelsinascendingpainpathways. BrainResearchReviews ,60(1):84{89,2009.ADecade ofPainResearch:NewApproaches,NewTargets. [4]StefanMenzler,JackA.Bikker,NirmalaSuman-Chauhan, andDavidC.Horwell. Designandbiologicalevaluationofnon-peptideanalogues ofomega-conotoxin MVIIA. Bioorganic&MedicinalChemistryLetters ,10(4):345{347,February 2000. 25

PAGE 31

[5]BaldomeroM.OliveraandLourdesJ.Cruz.Conotoxins,in retrospect. Toxicon 39(1):7{14,2001. [6]ChristopherJArmishawandPaulFAlewood.Conotoxinsas researchtoolsand drugleads. CurrentProtein&PeptideScience ,6(3):221{240,June2005. [7]JIKim,MTakahashi,AOgura,TKohno,YKudo,andKSato.Hy droxylgroup oftyr13isessentialfortheactivityofomega-conotoxinGV IA,apeptidetoxin forn-typecalciumchannel. J.Biol.Chem. ,269(39):23876{23878,September 1994. [8]R.AndrewAtkinson,BrunoKieer,AnnickDejaegere,Fin tonSirockin,and Jean-FrancoisLefevre.Structuralanddynamiccharacteri zationof-Conotoxin MVIIA:thebindingloopexhibitsslowconformationalexcha nge,. Biochemistry 39(14):3908,April2000. [9]VALennonandEHLambert.Autoantibodiesbindsolubiliz edcalciumchannelomega-conotoxincomplexesfromsmallcelllungcarcinoma: adiagnosticaidfor Lambert-Eatonmyasthenicsyndrome. MayoClinicProceedings.MayoClinic 64(12):1498{1504,December1989. [10]NielsenK.J.,ThomasL.,LewisR.J.,AlewoodP.F.,andC raikD.J.Aconsensus structureforomega-Conotoxinswithdierentselectiviti esforvoltage-sensitive calciumchannelsubtypes:ComparisonofMVIIA,SVIBandSNX -202. Journal ofMolecularBiology ,263:297{310,October1996. 26

PAGE 32

[11]KatherineJ.Nielsen,TinaSchroeder,andRichardLewi s.Structure-activity relationshipsofomega-conotoxinsatn-typevoltage-sens itivecalciumchannels. JournalofMolecularRecognition ,13(2):55{70,2000. [12]GlennE.Kirsch,JohnA.Drewe,HaliA.Hartmann,Mauriz ioTaglialatela, MarielladeBiasi,ArthurM.Brown,andRolfH.Johot.Diere ncesbetween thedeepporesofk+channelsdeterminedbyaninteractingpa irofnonpolar aminoacids. Neuron ,8(3):499{505,March1992. [13]KazukiSato,CecileRaymond,NicoleMartin-Moutot,To ruSasaki,AkiraOmori, AtsukoOhtake,JaeIlKim,ToshiyukiKohno,MasamiTakahash i,andMichael Seagar.Bindingofchimericanalogsof[omega]-conotoxinM VIIAandMVIICto then-andP/Q-typecalciumchannels. FEBSLetters ,414(2):480{484,September 1997. [14]DavidJ.Adams,AmandaB.Smith,ChristinaI.Schroeder ,TakahiroYasuda, andRichardJ.Lewis.omega-ConotoxinCVIDinhibitsapharm acologicallydistinctvoltage-sensitivecalciumchannelassociatedwitht ransmitterreleasefrom preganglionicnerveterminals. J.Biol.Chem. ,278(6):4057{4062,2003. [15]J.AndrewMcCammon,BruceR.Gelin,andMartinKarplus. Dynamicsoffolded proteins. Nature ,267(5612):585{590,June1977. [16]JBClarage,TRomo,BKAndrews,BMPettitt,andGNPhilli ps.Asampling probleminmoleculardynamicssimulationsofmacromolecul es. Proceedingsofthe NationalAcademyofSciencesoftheUnitedStatesofAmerica ,92(8):3288{3292, April1995. 27

PAGE 33

[17]WilliamL.JorgensenandJulianTirado-Rives.Monteca rlovsmolecular dynamicsforconformationalsampling. TheJournalofPhysicalChemistry 100(34):14508{14513,1996. [18]YujiSugitaandYukoOkamoto.Replica-exchangemolecu lardynamicsmethod forproteinfolding. ChemicalPhysicsLetters ,314(1-2):141{151,November1999. [19]DonaldHamelberg,JohnMongan,andJ.AndrewMcCammon. Acceleratedmoleculardynamics:Apromisingandecientsimulatio nmethodfor biomolecules. TheJournalofChemicalPhysics ,120(24):11919,2004. [20]D.A.Case,T.A.Darden,T.E.CheathamIII,C.L.Simmerl ing,J.Wang,R.E. Duke,R.Luo,M.Crowley,R.C.Walker,W.Zhang,K.M.Merz,B. Wang, S.Hayik,A.Roitberg,G.Seabra,I.Kolossvary,K.F.Wong,F .Paesani,J.Vanicek,X.Wu,S.R.Brozell,T.Steinbrecher,H.Gohlke,L.Yan g,C.Tan,J.Mongan,V.Hornak,G.Cui,D.H.Mathews,M.G.Seetin,C.Sagui,V .Babin,and P.A.Kollman.Amber10. UniversityofCalifornia,SanFrancisco ,2008. [21]ViktorHornak,RobertAbel,AsimOkur,BentleyStrockb ine,AdrianRoitberg, andCarlosSimmerling.Comparisonofmultipleamberforce eldsanddevelopmentofimprovedproteinbackboneparameters. Proteins:Structure,Function, andBioinformatics ,65(3):712{725,2006. [22]JianyinShao,StephenW.Tanner,NephiThompson,andTh omasE.Cheatham. Clusteringmoleculardynamicstrajectories:1.character izingtheperformanceof dierentclusteringalgorithms. JournalofChemicalTheoryandComputation 3(6):2312{2334,November2007. 28

PAGE 34

[23]JoeW.KeepersandThomasL.James.Atheoreticalstudyo fdistancedeterminationsfromNMR.two-dimensionalnuclearoverhausereec tspectra. Journal ofMagneticResonance(1969) ,57(3):404{426,May1984. 29