搜索
CombinatorialChemistryandNewDrugsTofightdisease，theimmunesystemgeneratesproteinsknownasantibodiesthatbindtoinvadingorganisms.Thebodycanmakeaboutatrilliondifferentantibodies，producedbyshufflingandreshufflingtheirconstituentparts.Buttheimmunesystemisnotequippedtocraftaspecializedantibodyeachtimeitisfacedwithanewpathogen.Insteadthebodyselectivelydeploysonlythoseexistingantibodiesthatwillworkmosteffectivelyagainstaparticularfoe.Theimmunesystemdoesthis，ineffect，bymassscreeningofitsantibodyrepertoire，identifyingtheonesthatworkbest.Inaprocesscalledcombinatorialchemistry，wegeneratealargenumberofrelatedcompoundsandscreenthecollectionfortheonesthatcouldhavemedicinalvalue.Thisapproachdiffersfromthemostcommonwaypharmaceuticalmakersdiscovernewdrugs.Theytypicallybeginbylookingforsignsofadesiredactivityinalmostanythingtheycanfind，suchasdiversecollectionsofsyntheticcompoundsorofchemicalsderivedfrombacteria，plantsorothernaturalsources.Oncetheyidentifyapromisingsubstance（knowninthefieldasaleadcompound），theylaboriouslymakemanyone-at-a-timemodificationstothestructure.，testingaftereachsteptodeterminehowthechangesaffectedthecompound`schemicalandbiologicalproperties.Oftentheseproceduresyieldacompoundhavingacceptablepotencyandsafety.Foreverynewdrugthatmakesittomarketinthisway，however，researchersquitelikelytinkeredwithandabandonedthousandsofothercompoundsenroute.Theentireprocedureistime-consumingandexpensive：ittakesmanyyearsandhundredsofmillionsofdollarstomovefromaleadcompoundinthelaboratorytoabottleofmedicineontheshelfofyourlocalpharmacy.Theclassicalapproachhasbeenimprovedbyscreeningteststhatworkmorerapidlyandreliablythaninthepastandbyburgeoningknowledgeabouthowvariousmodificationsarelikelytochangeamolecule`sbiologicalactivity.Butasmedicalsciencehasadvanced，demandfordrugsthatcaninterveneindiseaseprocesshasescalated.Tofindthosedrugs，researchersneedmanymorecompoundstoscreenaswellasawaytofindleadcompoundsthatrequiredlessmodification.Drugmakershavetwobasiccombinatorialtechniquesattheirdisposal.Thefirst，knownasparallelsynthesis，wasinventedinthemid1980sbyH.MarioGeysen，nowatGlaxoWellcome.Heinitiallyusedparallelsynthesisasaquickwaytoidentifywhichsmallsegmentofanygivenlargeproteinboundtoanantibody.Geysengeneratedavarietyofshortproteinfragments，orpeptides，bycombiningmultipleaminoacids（thebuildingblocksofpeptidesandproteins）indifferentpermutations.Byperformingdozensorsometimeshundredsofreactionsatthesametimeandthentestingtoseewhethertheresultingpeptideswouldbindtotheparticularantibodyofinterest，herapidlyfoundtheactivepeptidesfromalargeuniverseofpossiblemolecules.Inaparallelsynthesis，alltheproductsareassembledseparatelyintheirownreactionvessels.Tocarryouttheprocedure，chemistsoftenuseaso-calledmicrotitreplate-----asheetofmoldedplasticthattypicallycontainseightrowsand12columnsoftinywells，eachofwhichholdsafewmillilitersoftheliquidinwhichthereactionswilloccur.Thearrayofrowsandcolumnsenablesworkerstoorganizethebuildingblockstheywanttocombineandprovidesareadymeanstoidentifythecompoundinaparticularwell.Chemistsoftenstartsacombinatorialsynthesisbyattachingthefirstsetofbuildingblockstoinert，microscopicbeadsmadeofpolystyrene（oftenreferredtoassolidsupport.）.Aftereachreaction，researcherswashawayanyunreactedmaterial，leavingbehindonlythedesiredproducts，whicharestilltetheredtothebeads.Althoughthechemicalreactionsrequiredtolinkcompoundstothebeadsandlatertodetachthemintroducecomplicationstothesynthesisprocess，theeaseofpurificationcanoverweightheseproblems.Inmanylaboratoriestoday，robotsassistwiththeroutineworkofparallelsynthesis，suchasdeliveringsmallamountsofreactivemoleculesintotheappropriatewells.Inthisway，theprocessbecomesmoreaccurateandlesstedious.Thesecondtechniqueforgeneratingacombinatoriallibrary，knownasasplit-and-mixsynthesis，waspioneeredinthelate1980sbyArpadFurka，nowatAdvancedChemTechinLouisville，Ky.Incontrasttoparallelsynthesis，inwhicheachcompoundremainsinitsowncontainer，asplit-and-mixsynthesisproducesamixtureofrelatedcompoundsinthesamereactionvessel.Thismethodsubstantiallyreducesthenumberofcontainersrequiredandraisesthenumberofcompoundsthatcanbemadeintothemillions.Thetrade-off，however，isthatkeepingtrackofsuchlargenumbersofcompoundsandthentestingthemforbiologicalactivitycanbecomequitecomplicated.Aswementionedearlier，oneoftheproblemswithasplit-and-mixsynthesisisidentifyingthecompositionofareactivecompoundwithalargemixture.KitLamoftheUniversityofArizonahasdevelopedawaytoovercomethisobstacle.Henotedthatattheendofasplit-and-mixsynthesis，allthemoleculesattachedtoasinglebeadareofthesamestructure.Scientistscanpulloutfromthemixturethebeadsthatbearbiologicallyactivemoleculesandthen，usingsensitivedetectiontechniques，determinethemolecularmakeupofthecompoundattached.Unfortunately，thetechniquewillworkonlyforcertaincompounds，suchaspeptidesorsmallsegmentsofDNA.Nevertheless，becauseofdifficultiesofidentifyingcompoundsmadeinasplit-and-mixsynthesis，mostpharmaceuticalcompaniestodaycontinuetorelyonparallelsynthesis.