企业QQ
官方微信
byMichaelKoelle,1/6/96PrimerDesignGraphicTheidentificationofnovelmembersofgenefamiliesbyPCRusingdegenerateprimershasbeenconsideredmoreofanartthanascience,somuchsothatthemethodsbooksIvecomeacrosshavebeentootimidtodiscusstheconsiderationsthatgointothedesignofthisexperiment,muchlessgiveaprotocolforitsexecution.Attheriskofleadingmyreadersonwildgoosechases,Imcommittingmymethodstopaper.Thefollowingisbasedonmyreadingoftherecentliterature(e.g.BuckandAxel,Cell65:175-187,1991;Riddleetal.,Cell75:1401-1416,1993;Kraussetal.,Cell75:1431-1444,1993),discussionswithseveralothersuccessfulpractitionersoftheart,andmyownexperienceisolatingvertebratehomologsoftheC.elegansegl-10gene(KoelleandHorvitz,Cell84:1-20,1996).PrimerdesignThisisthemostimportantfactorinthesuccessoftheexperiment,anddeservescarefuldeliberation.Isuggestdiagrammingoutanalignmentoftheexistingmembersofyourgenefamily,highlightingconservedresidues,andlabelingeachimportantpositioninthealignmentwiththenumberofcodonsthatencodetheaminoacid(s)atthatposition.Anexamplebasedontheoriginalmembersoftheegl-10genefamilyisincludedbelow,andcitedinthefollowingdiscussion.Primerdegeneracy
IntheearlydaysofdegeneratePCRsomenovelgenesweresuccessfullyamplifiedusingprimerpoolsthatwereover1000-folddegenerate.However,primersofsuchhighdegeneracyappearnottohavebeengenerallysuccessful(withsomeexceptions,e.g.Giovaneetal.,GenesDev.8:1502-1513,1994),andmostrecentsuccesseshavecomeusingprimerpoolsof100-folddegeneracyorless.Fivemethodsforreducingthedegeneracyoftheprimersarediscussedbelow:1)judiciousselectionoftheprimersitesThepositioningoftheprimersisacompromisebetweenplacingthematthecodonsforthemostconservedaminoacids,andplacingthematthecodonsforlessconservedaminoacidswhosedegeneracymaybelower.Considerthecaseofthe3primers("3Tand3A")intheexampleshownbelowfortheegl-10genefamily.Atfirstitmightseemmoresensibletoplacetheseprimers3codonstotheright,wherethereisablockof5outof6absolutelyconservedaminoacids:SY
(P/Q)RFL
.Unfortunately,thisblockofaminoacidsisencodedby5184differentDNAsequences.Theactualprimersusedwereplaced3codonstotheleft.Atthispositiononly3outof6aminoacidsareabsolutelyconserved:M
(E/K)(K/N)(D/N)SY
.However,thisblockofaminoacidsisencodedbyonly768differentDNAsequences.2)theuseofinosineasa"neutral"baseInosineisapurine(whichoccursnaturallyintRNAs)thatcanformbasepairswithcytidine,thymidine,andadenosine(althoughtheinosine:adenosinepairingpresumablydoesntfitquitecorrectlyindoublestrandedDNA,sotheremaybeanenergeticpenaltytopaywhenthehelixbulgesoutatthispurine:purinepairing).Recently,mostpeoplehavebeenusinginosineintheirprimersatpositionswhereanyofthefourbasesmightberequired.Eachuseofinosinethusreducesthedegeneracyoftheprimerpool4-fold.However,yourisktheoccurrenceofI:Gmismatches,andthereforemustassumethatexactbasepairingatotherpositionsintheprimerwillovercomesuchaproblem.Mostoligosynthesisfacilitieswillmakeinosine-containingoligos,noproblem.Ihadexcellentluckwithinosine-containingprimerswiththeegl-10genefamily,exceptinthecaseoftheprimer"5out",a20mercontaining5inosines,(including2nearthe3endoftheprimer)whichfailedtoamplifyproductsevenfromaclonedegl-10cDNA.So,perhaps5outof20inosinesistoomany.UsinginosineintheprimersrequiresthattheDNApolymeraseusedinthePCRreactionbecapableofsynthesizingDNAoveraninosine-containingtemplate.Taqpolymeraseiscapableofdoingthis,butsomeothers(e.g.Vent)appearnottobeableto.3)usingmultipleseparateoligopoolsatasinglepositionInanefforttouseprimerpoolswiththelowestpossibledegeneracy,itissometimesusefultosynthesizeprimersoveraparticularstretchofcodonsastwoormoreseparatepools,eachofwhichwillhavelowerdegeneracythanyouwouldgetbysynthesizingasinglepoolincludingallofthesamecodons.ThepoolsarethenusedseparatelytocarryoutPCRreactions.Forexample,theprimerpools"3T"and"3A"intheegl-10examplebelowareidentical,exceptattheirserinecodons.Sadly,serineisencodedby6differentcodons,TC(A/G/C/T)andAG(T/C).Synthesizingasinglepoolcoveringallthesepossibilitiesmightrequireahighdegeneracyandwouldnecessarilyincludesomenon-serinecodons.Bysplittingintotwopools(onenondegeneratecontaininganinosine,theother2-folddegenerate)Iwasabletokeepthedegeneracylow,andavoidallnon-serinecodons.Anotherexampleisshowninthecaseofprimers"5inE"and"5inR",whichagainareidenticalexceptatonecodon.4)includingpartialcodonsattheendsoftheprimersThevariouscodonsencodinganaminoacidorasetofsimilaraminoacidsareoftenidenticalattheirfirst(andmaybesecond)positions,butdifferentattheirthirdposition.Youcantakeadvantageofthisbysynthesizingonlythefirstorfirstandsecondpositionsofthe3mostcodoncoveredbyyourprimerpools,thusgivingyouoneortwoextrapositionsofexactmatchbasepairswithoutaddinganydegeneracy.Intheegl-10example,theprimerpools"3T"and"3A"coverastretchinwhichthelastcodonmustencodeprolineorglutamine.ThecodonsforthesetwoaminoacidsallstartwithC,buttheirlasttwopositionsaredegenerate.Therefore,onlythenondegenerateCwasincludedintheprimerpools.5)useofcodonbiasSomeorganismshavestrongbiasesforusingparticularcodonstoencodecertainaminoacids.Intheoryyoucouldreducethedegeneracyofaprimerpoolbyonlyincludingthesemostcommoncodons,andtakingtheriskthatthegene(s)youarelookingforwillfollowtheorganismsgeneralcodonbiasenoughtoallowsuchprimerpoolstowork.IhaventheardofanyoneactuallyusingthiscodonbiasstrategyinasuccessfuldegeneratePCRexperiment,butyoumighttryitifyouredesperate.Otherconsiderationsinprimerdesign
1)primerlengthIntheexamplebelowtheshortstretchesofsequencesimilarityamongtheegl-10familymembersforcedmetouseprimersonly19-21baseslong.TheseareshorterthantheprimersIhaveheardofpeopleusinginothersuccessfulexperiments.Forexample,LindaBucksprimerswere31-33mers.2)3endPeopleItalkedtoemphasizedthespecialimportanceofhavinganexactmatchbetweentheprimerandtemplatenearthe3endoftheprimer,althoughImnotawareofspecificdatasupportingthisidea.Foregl-10Itriedtoavoidhavinganyinosinesnearthe3endsoftheprimers(exceptforprimer"5out",whichinfactfailedtogiveanyproducts),andalsoanchoredtheprimerswhenpossiblewithanondegeneratecodonattheir3ends,sothat100%oftheprimersinthepoolwouldbeabletopairperfectlywiththecorrecttemplateovertheselastfewbases.3)nestedprimersIfthesequencesimilarityinyourgenefamilypermits,itisagoodideatomakenestedsetsofPCRprimers.ThatwayoneroundofPCRcanbeperformedusingtheoutsideprimers,andindividualproducts(orthewholemix)canthenbereamplifiedusingtheinsideprimers.Productsamplifiedthroughbothroundsaremorelikelytobethedesirednewgenefamilymembers,andlesslikelytobespuriousproductsfromsequencesthathappentocontainacoupleofprimerannealingsitesbychance.DeterminingoptimalreactionconditionsAnumberofparameterscanbevariedtooptimizereactionconditionsfordegeneratePCR.Theseinclude:primerconcentration,magnesiumconcentration,templateconcentration,numberofcyclesofamplification,andthetemperaturesandtimesofeachstepintheamplificationcycle.Ifeachoftheseparametersistobeindependentlyvaried,thenumberofpossibilitiesquicklyreachesmindbogglingproportions.Myphilosophyhasbeentofixalmostalltheseparametersatthestandardlevelsthathavebeensuccessfulforotherpeople,andtovaryonlytheoneparameterthatIthinkisthemostcrucial:thetemperatureoftheannealingstepduringamplification.MystandardPCRreactionsareasfollows:
Inpractice,thereactionsaresetupbyplacingtheprimersandtemplateintoa0.5mltube,thenaddingtwodropsofmineraloilfromabluetip,andaddingontopoftheoil38.5lofapremixcontainingalltheothercomponents.Inthisway,itiseasytosetupmanydifferentprimer/templatecombinationsatonce.Thetubesarethenbrieflyspuninamicrofugetocombinethetwoaqueousphases,andthetubesareimmediatelyplacedinthePCRblockpreheatedto95fora"hotstart".Myamplificationprogram:95X3min.(hotstart)??X1min.(thisannealingtemperatureisvariedtooptimizetheamplification)72X2min.94X45sec.40cyclesoftheabove3steps72X5min.holdat4Thistakes~4.5hourstorunonanMJResearchmachine.Totesttheprimersandoptimizetheconditions,Idoaseriesofamplificationrunsstartingwithanannealingtemperatureof25,andincreasingin5incrementsuntilamplificationfailstooccur.Typicallyforeachprimerpairbeingtested,ateachtemperature,Irun3reactionscontainingdifferenttemplates:1)apositivecontrolcontaining2ngofaclonedmemberofthegenefamilyofinterestastemplate.2)anegativecontrolcontainingnotemplate(thisisveryimportant--youdontwanttogetfooledbycontaminants).3)anexperimentalreactioncontainingacomplextemplatesuchasgenomicDNAortotalcDNA.FortotalC.elegansgenomicDNAIvebeenusing300ngasatemplate.UsingratbraincDNAasatemplateIamplifiedoffofonly2ng.However,thiswasonlybecauseIdidnthaveverymuchcDNA.Ifpossible,itwouldbebettertouse~200ngofcDNAasatemplate,asLindaBuckdidtoamplifytheodorantreceptors.ChoiceoftemplateGenomicDNAhastheadvantagethatallmembersofyourgenefamilyarepresentinequimolaramounts,andgenomicDNAisprobablyreadilyavailable.Theobviousdisadvantageisthatintronsmaydisrupttheprimersites,ormaycausetheamplificationproducttobesolongthatitisnotamplifiedefficiently.cDNAtemplates,thoughhardertoobtain,overcomethisproblem.AbigadvantageofcDNAisthatthedesiredamplificationproductsshouldbeofaknownsize,andyoucanthereforeeasilypickthemoutfromamongspuriousproductsofothersizes.Rememberthatthe"correct"sizedbandamplifiedoffacDNAtemplatemaybeacomplexmixtureofproductsfrommanygenefamilymembers,soyoumayhavetoanalyzemanyclonesgeneratedfromsuchabandtoassessitscomplexity.LindaBuckusedrandomprimedcDNAforhertemplate,presumablytoavoidbiasingthecDNAtowardsthe3endsoftranscripts.Inmycase,IknewthattheregionIwasamplifyingshouldbeattheextreme3endofthecodingsequence,soIusedoligo-dTprimedcDNA.Forthelazyandrich,Clontechsellsoligo-dTprimedcDNApreparedfromvarioustissuesofmanyspeciestouseastemplatesforPCR.AnalysisofPCRproductsAfteramplificationrun20lofeachreactionoutonanagarosegel.Iuse2%3:1Nusieve:SeaKemLEagarose(youcanbuythispremixedfromFMC)in1XTAE.Thisgelisnotverylowmelting,andthusisntverysuitableforcloningdirectlyfromthegel,butitgivesveryniceresolution.Iusethe123bpladderfromGibcoasasizestandard.Obviouslyyouexpecttogetproductsoffthepositivecontrol,andnottogetthemoffthenegativecontrol.Usingthecomplextemplate,youwillprobablygetasmearatthelowerannealingtemperatures,whichwillresolveintoasmallnumberofbandsastheannealingtemperaturerises.Ipickanannealingtemperaturethatgivesamodestnumberofbands,andthencloneallthesebandsandsequencethem.Ifnoproductsareevidentintheexperimentalsamples,agoodtricktotryistouse2loftheapparentlyfailedreactionasatemplate,andreamplifyunderthesameconditions.Thisoftengivesvisibleproducts.Ifyouwanttocloneproductsthatareonlybarelyvisible,youcangetmoreofthembyjustreamplifyingtheoriginalreactionasdescribedabove.Anotherwaytoamplifyindividualproductsseparatelyistocutthebandsoutofthegelthatwasusedtoanalyzetheoriginalreactions(actuallyItakeaboreoutofthegelwithaPasteurpipette),melttheDNAcontainingagarose,anduse2lofitasatemplatetoreamplifyunderthesameconditions.Theabovedescribedmethodforreamplifyingspecificbandscan(andshould)beusedtotestamplifiedproductstoseeiftheyaresingleprimerartefacts.Use2lofanagarosegelboretosetupeachofthreePCRreactions,containingeitherindividualprimerorbothtogether.Obviously,youreonlyinterestedinproductsthatrequirebothprimersinordertobeamplified.IclonePCRproductsbyrunningthePCRreactionoutonalowmeltagarosegel(2%Nusieveagarosein1XTAE).Icutthedesiredbandout,meltitat70,mixwellbypipettingup/down,anduse5lofthemeltedagarosedirectlyinaligationreactionwithadTtailedvector.ThisvectorDNAispreparedasfollows:cut1gbluescriptSKwithEcoRVina20lreaction.Add20l1XPCRbuffer,and2l2mMdTTP.Add0.5l"ampliTaq"polymerase(2.45U),andincubateat~72for20min.RuntheDNAoutona0.8%Seaplaqueagarosegelin1XTAE,cutouttheband,meltat70,mixwell,anduse5linaligationreaction.Itturnsoutthatonlyabout50%ofthecoloniesobtainedaftertransformationofthistypeofreactionmayhaveinserts;therestarevectorreclosures.However,ifblue/whiteselectionisused,virtuallyallthewhitecolonieshaveinserts.