A-Structure-for-Deoxyribose-Nucleic-Acid

AStructureforDeoxyriboseNucleicAcidJ.D.WatsonandF.H.C.Crick(1)April25,1953(2),Nature(3),171,737-738Wewishtosuggestastructureforthesaltofdeoxyribosenucleicacid(D.N.A.).Thisstructurehasnovelfeatureswhichareofconsiderablebiologicalinterest.AstructurefornucleicacidhasalreadybeenproposedbyPauling(4)andCorey1.Theykindlymadetheirmanuscriptavailabletousinadvanceofpublication.Theirmodelconsistsofthreeintertwinedchains,withthephosphatesnearthefibreaxis,andthebasesontheoutside.Inouropinion,thisstructureisunsatisfactoryfortworeasons:(1)WebelievethatthematerialwhichgivestheX-raydiagramsisthesalt,notthefreeacid.Withouttheacidichydrogenatomsitisnotclearwhatforceswouldholdthestructuretogether,especiallyasthenegativelychargedphosphatesneartheaxiswillrepeleachother.(2)SomeofthevanderWaalsdistancesappeartobetoosmall.Anotherthree-chainstructurehasalsobeensuggestedbyFraser(inthepress).Inhismodelthephosphatesareontheoutsideandthebasesontheinside,linkedtogetherbyhydrogenbonds.Thisstructureasdescribedisratherill-defined,andforthisreasonweshallnotcommentonit.Wewishtoputforwardaradicallydifferentstructureforthesaltofdeoxyribosenucleicacid(5).Thisstructurehastwohelicalchainseachcoiledroundthesameaxis(seediagram).Wehavemadetheusualchemicalassumptions,namely,thateachchainconsistsofphosphatediestergroupsjoiningbeta-D-deoxyribofuranoseresidueswith3',5'linkages.Thetwochains(butnottheirbases)arerelatedbyadyadperpendiculartothefibreaxis.Bothchainsfollowright-handedhelices,butowingtothedyadthesequencesoftheatomsinthetwochainsruninoppositedirections(6).EachchainlooselyresemblesFurberg's2modelNo.1(7);thatis,thebasesareontheinsideofthehelixandthephosphatesontheoutside.TheconfigurationofthesugarandtheatomsnearitisclosetoFurberg'sstandardconfiguration,thesugarbeingroughlyperpendiculartotheattachedbase.Thereisaresidueoneachevery3.4A.inthez-direction.Wehaveassumedanangleof36°betweenadjacentresiduesinthesamechain,sothatthestructurerepeatsafter10residuesoneachchain,thatis,after34A.Thedistanceofaphosphorusatomfromthefibreaxisis10A.Asthephosphatesareontheoutside,cationshaveeasyaccesstothem.Thestructureisanopenone,anditswatercontentisratherhigh.Atlowerwatercontentswewouldexpectthebasestotiltsothatthestructurecouldbecomemorecompact.Thenovelfeatureofthestructureisthemannerinwhichthetwochainsareheldtogetherbythepurineandpyrimidinebases.Theplanesofthebasesareperpendiculartothefibreaxis.Theyarejoinedtogetherinpairs,asinglebasefromonechainbeinghydroden-bondedtoasinglebasefromtheotherchain,sothatthetwoliesidebysidewithidenticalz-coordinates.Oneofthepairmustbeapurineandtheotherapyrimidineforbondingtooccur.Thehydrogenbondsaremadeasfollows:purineposition1topyrimidineposition1;purineposition6topyrimidineposition6.Ifitisassumedthatthebasesonlyoccurinthestructureinthemostplausibletautomericforms(thatis,withtheketoratherthantheenolconfigurations)itisfoundthatonlyspecificpairsofbasescanbondtogether.Thesepairsare:adenine(purine)withthymine(pyrimidine),andguanine(purine)withcytosine(pyrimidine)(9).Inotherwords,ifanadenineformsonememberofapair,oneitherchain,thenontheseassumptionstheothermembermustbethymine;similarlyforguanineandcytosine.Thesequenceofbasesonasinglechaindoesnotappeartoberestrictedinanyway.However,ifonlyspecificpairsofbasescanbeformed,itfollowsthatifthesequenceofbasesononechainisgiven,thenthesequenceontheotherchainisautomaticallydetermined.Ithasbeenfoundexperimentally3,4thattheratiooftheamountsofadeninetothymine,andtheratioofguaninetocytosine,arealwaysveryclosetounityfordeoxyribosenucleicacid.Itisprobablyimpossibletobuildthisstructurewitharibosesugarinplaceofthedeoxyribose,astheextraoxygenatomwouldmaketoocloseavanderWaalscontact.ThepreviouslypublishedX-raydata5,6ondeoxyribosenucleicacidareinsufficientforarigoroustestofourstructure.Sofaraswecantell,itisroughlycompatiblewiththeexperimentaldata,butitmustberegardedasunproveduntilithasbeencheckedagainstmoreexactresults.Someofthesearegiveninthefollowingcommunications(10).Wewerenotawareofthedetailsoftheresultspresentedtherewhenwedevisedourstructure(11),whichrestsmainlythoughnotentirelyonpublishedexperimentaldataandstereochemicalarguments.Ithasnotescapedournotice(12)thatthespecificpairingwehavepostulatedFigure1Thisfigureispurelydiagrammatic(8).Thetworibbonssymbolizethetwophophate-sugarchains,andthehorizonalrodsthepairsofbasesholdingthechainstogether.Theverticallinemarksthefibreaxis.immediatelysuggestsapossiblecopyingmechanismforthegeneticmaterial.Fulldetailsofthestructure,includingtheconditionsassumedinbuildingit,togetherwithasetofcoordinatesfortheatoms,willbepublishedelsewhere(13).WearemuchindebtedtoDr.JerryDonohueforconstantadviceandcriticism,especiallyoninteratomicdistances.WehavealsobeenstimulatedbyaknowledgeofthegeneralnatureoftheunpublishedexperimentalresultsandideasofDr.M.H.F.Wilkins,Dr.R.E.Franklinandtheirco-workersatKing’sCollege,London.Oneofus(J.D.W.)hasbeenaidedbyafellowshipfromtheNationalFoundationforInfantileParalysis.1Pauling,L.,andCorey,R.B.,Nature,171,346(1953);Proc.U.S.Nat.Acad.Sci.,39,84(1953).2Fu