A modular steroid-inducible gene expression system for use in ...

Chemically inducible systems generally consist of chimeric transcription factors and cognate promoters that are derived from genetic elements of ... Skiptomaincontent Advertisement SearchallBMCarticles Search Amodularsteroid-induciblegeneexpressionsystemforuseinrice DownloadPDF DownloadPDF Methodologyarticle OpenAccess Published:15October2019 Amodularsteroid-induciblegeneexpressionsystemforuseinrice DanielaVlad1,BaselAbu-Jamous2,PengWang3&JaneA.Langdale  ORCID:orcid.org/0000-0001-7648-39241  BMCPlantBiology volume 19,Article number: 426(2019) Citethisarticle 4612Accesses 9Citations Metricsdetails AbstractBackgroundChemicallyinduciblesystemsthatprovidebothspatialandtemporalcontrolofgeneexpressionareessentialtools,withmanyapplicationsinplantbiology,yettheyhavenotbeenextensivelytestedinmonocotyledonousspecies.ResultsUsingGoldenGatemodularcloning,wehavecreatedamonocot-optimizeddexamethasone(DEX)-induciblepOp6/LhGRsystemandtesteditsefficacyinriceusingthereporterenzymeβ-glucuronidase(GUS).ThesystemistightlyregulatedandhighlysensitivetoDEXapplication,with6 hofinductionsufficienttoinducehighlevelsofGUSactivityintransgeniccallus.Inseedlings,GUSactivitywasdetectableintherootafterinvitroapplicationofjust0.01 μMDEX.However,transgenicplantsmanifestedseveredevelopmentalperturbationswhengrownonhigherconcentrationsofDEX.Thedirectcauseofthesegrowthdefectsisnotknown,butthericegenomecontainssequenceswithhighsimilaritytotheLhGRtargetsequencelacO,suggestingnon-specificactivationofendogenousgenesbyDEXinduction.Theseoff-targeteffectscanbeminimizedbyquenchingwithisopropylβ-D-1-thiogalactopyranoside(IPTG).ConclusionsOurresultsdemonstratethatthesystemissuitableforgeneraluseinrice,whenthemethodofDEXapplicationandrelevantcontrolsaretailoredappropriatelyforeachspecificapplication. BackgroundThestudyofgenefunctioninplantsreliesonreversegenetictoolsthatfacilitateoverexpressionoftransgenesorsuppressionofendogenousgeneexpression.Forsomegenes,neitherapproachisfeasibleduetothedevelopmentofnon-viablephenotypes.Toenablefunctionalanalysisinsuchcases,severalchemicallyinduciblesystemshavebeenestablishedtocontroltransgeneexpression(forreviewssee[1,2,3,4]).Chemicallyinduciblesystemsgenerallyconsistofchimerictranscriptionfactorsandcognatepromotersthatarederivedfromgeneticelementsofheterologousorganisms(toavoidinterferencewithexpressionofendogenousgenes)[2].Thedexamethasone(DEX)-induciblepOp6/LhGRgeneexpressionsystem[5,6]isbasedonamodifiedEscherichiacolilac-repressorsystem[7].ThepOp6promotercontainsaconcatemerizedbindingsitecomprisedofsixdirectrepeatsofthe18basepairlacoperator(lacO)sequence[5].ThissiteisboundbythechimerictranscriptionfactorLhGR,whichisafusionbetweenthehighaffinityDNA-bindingdomainofthemutantlacrepressor,lacIHis17,theGal4transcription-activation-domain-IIandtheDEX-bindingdomainoftheratglucocorticoidreceptor(GR)fusedattheN-terminus.AdditionofDEXtotransgenicplantscontainingbothLhGRandpOp6leadstorelocationofLhGRfromthecytoplasmtothenucleus,andconsequenttranscriptionalactivationoftransgenesequenceslinkedtopOp6.Chemicallyinduciblegeneexpressionsystemshavebeenpredominantlytestedindicotyledonousspecies[1,2,3,4].However,twosteroid-induciblesystemshavebeentestedinrice-theDEX-inducibleGal4/GVG[8]andtheestrogen-induciblepLex/XVE[9,10,11,12].ActivationofGal4/GVGinriceledtotightlycontrolledtransgeneexpressionbutseverelyperturbedplantgrowthafterexposuretomoderateconcentrationsofinducer.Similardetrimentaleffectswerereportedinotherspecies[13,14,15].ThepLex/XVEsystemwasusedmoresuccessfullytoinduceexpressioninricecallus,rootsandleavesbutdifferentapplicationmethodswererequiredineachcaseduetoinefficientestradioluptakeviatheroots.ThemorerecentlydevelopedpOp6/LhGRsystemhasnotyetbeentestedinmonocotsbutbecausethepOppromoterisnotactivatedbyendogenousfactors,atleastnotinmaize[2,16],itmayprovideasuitablealternativetopLex/XVE.Tocreateaversatileinduciblesystemfortransgeneexpressioninmonocots,wedevelopedaversionofpOp6/LhGRusingtheGoldenGatemodularcloningsystem[17].TheGoldenGatesystemisdesignedtoprovidearapid,modularandscar-lessassemblyoflargeconstructs,offeringaflexiblechoiceofpromotersandselectionmodules.Testsofthesysteminstabletransgenicricelinesrevealedtighttemporalcontrolovertransgeneexpressionandthereliabilityofarangeofapproachesforinduction.ResultsConstructdesignTocreateinducibleconstructs,thepOp6induciblepromoterandthesequenceencodingthecorrespondingchimerictranscriptionalactivator(LhGR),weresynthesizedasGoldenGatecompatiblelevel0modules,‘PU’and‘SC’[17]respectively.LhGRwascodonoptimizedforuseinrice(rcoLhGR)andfurther‘domesticated’toremoveallrecognitionsitesfortypeIIrestrictionenzymesusedinGoldenGatecloning:BsaI,BpiI,Esp3IandDraIII(Additional file 1).Promoters,openreadingframesandterminatorsequencessynthesizedaslevel0moduleswerecombinedintostandardGoldenGatelevel1vectorstogenerateindividualtranscriptionalunitsforantibioticselection,transcriptionalactivationandreportergeneexpression.Threelevel2versionsoftheinducibleGoldenGateconstructweregenerated,differinginthechoiceofselectionmoduleandinthereportergenesequenceused.ThethreeversionsweredesignedtotestthepotentialofthefulllengthCaMV35Spromoter(p35S)and/orofanyintronspresentinthereportergenesequence,toactivateexpressionfromthepOp6promoter.Inallcases,fourlevel1moduleswerecombinedintolevel2constructsforricetransformation.Allthreelevel2constructscontainedadsRedreportergenedrivenbytheconstitutivericeactinpromoter(pOsACT)inposition2(toallowidentificationoftransgenicseedonthebasisoffluorescence),andthercoLhGRgeneexpressedunderthecontrolofamaizeubiquitinpromoterthatcontainsanintron(pZmUbi)[18]inposition3.Inposition1,thefirstconstruct(17203)containedp35Sdrivingexpressionofthehygromycin-resistanceselectablemarkergenehygromycinphosphotransferase(HYG),whereasconstructs17610and17613hadpOsACTdrivingexpressionofHYG.Inposition4,theE.coliuidAgeneencodingtheenzymeβ-glucuronidase(GUS)wasusedasareporterforpOp6promoteractivity.Inconstructs17203and17610,theuidAsequencecontainedtwointronsfromtheGFA1(At1g06220)geneinArabidopsis,whereasthesequenceinconstruct17613wasintron-less(Fig. 1;Additional files 2, 3, 4). Fig.1Schematicillustrationoftheconstructsgeneratedinthisstudy.Plasmids17203,17610and17613werecreatedusingtheGoldenGatemodularcloningsystem(seeAdditionalfiles 2, 3, 4forcompletesequenceofeach).HYG:hygromycinphosphotransferase;p35S:CaMV35Spromoter;pOsAct:riceactinpromoter;pZmUbi:maizeubiquitinpromoter;GUS:uidAgeneencodingβ-glucuronidase.Theinduciblesystemconsistsoftheactivator,i.e.aricecodonoptimized(rco)versionofLhGRplusitscognatepOp6promoter.Ineachconstruct,allfourpromoter:codingsequencemodulescontainthenosterminator(notrepresented)FullsizeimageThepOp6/rcoLhGRsystemisfunctionalinriceTotestthepOp6/rcoLhGRsysteminamonocotspecies,allconstructsweretransformedintorice(OryzasativasppjaponicacvKitaake)toobtainstabletransgeniclines.PositiveT0transformantswerefirstvalidatedbyhistochemicalGUSdetectionafterDEXinductionofdetachedleaffragments(datanotshown),andthentransgenecopynumberwasassessedbyDNAgelblotanalysis(Additional file 5).Lineswithoneortwocopiesofthetransgeneplusline17613_11,forwhichinsertioncopywasnotdeterminedinT0plants,werepropagatedintotheT1generation.AtleastfourindependentT1lineswereobtainedforeachofthethreeconstructs,withtransgenecopynumberrangingfromonetothree(Additional file 6).TheefficacyofthepOp6/rcoLhGRsysteminricewasfirstassessedbymeasuringGUSactivityusinganextractableenzymaticassay.LeavesweredetachedfromindividualsofatleastfourindependentT1linesperconstruct,treatedwithDEXfor24 handthenassayedforGUSactivityusingthesubstrate4-methylumbelliferylß-D-glucuronide(MUG)(Additional file 7).AlinethatconstitutivelyexpressesasyntheticGUSvariantfromStaphylococcus(GUSPlus)underthecontrolofthemaizeubiquitinpromoter(pZmUbi:GUS)wasusedasapositivecontrolforthefluorometricMUGassay.Figure 2showsthatDEXinductionoftenresultedinhigherlevelsofGUSactivityinT1individualstransformedwiththeinducibleconstructsthaninnon-inducedindividualsoftheconstitutivepZmUbi:GUSline(compareFig. 2a-c‘induced‘with2D‘mock’).GiventhatGUSPlusisreportedtobeten-foldmoreactivethantheenzymeencodedbyE.coliuidAthatwasusedintheinducibleconstructs[19,20],thisobservationsuggeststhatactivationofthepOp6promoterbythercoLHGRisveryeffective.Curiously,GUSactivityinthepZmUbi:GUSlinewassuppressedinthepresenceofDEX(Fig. 2d)suggestingapossibleinhibitoryinteractionbetweenthesteroidandthesyntheticGUSenzymeinthefluorometricassay.Importantly,inallbutoneofthe14induciblelinestested,GUSactivitywassignificantlyhigherafterDEXapplication,withverylittleactivitydetectedintheabsenceoftheinducer(Fig. 2a-c).Evenintheexceptionalline(17613–6),highlevelsofbackgroundactivitywereonlyobservedinoneofthetwoindividualsexamined(17613-6A,Fig. 2c).TherewasnocorrelationbetweentransgenecopynumberandlevelsofGUSactivitybeforeorafterinduction.Forexample,line17203_10wassegregatingforasingleT-DNAinsertionandline17203_7fortwo(linked)insertions(Additionalfile 5),yetlevelsofGUSactivityafterDEXinductionweresimilarinbothlines(Fig. 2a).VariationinactivitybetweenindividualswithinT1familiesmaybeexplainedbyvariablezygosityresultingfromsegregationofthetransgene.TogethertheseresultsindicatethatthepOp6/rcoLhGRsystemisfunctional,sensitive,andtightlyregulatedinrice. Fig.2GUSactivityisinducedbyDEXin14independentT1pOp6/rcoLhGRlines.GUSenzymaticactivitymeasuredintotalleafproteinextractsfromindividualssegregatinginT1linestransformedwithconstruct17203(a,e),17610(B,F)or17613(c,g).IndividualsfromapZmUbi:GUSlinewereusedaspositivecontrols,andwild-typeKitaakericeplants(WT)wereusedasnegativecontrol(d).Theassayused4-methylumbelliferylβ-D-glucuronide(4-MUG)asasubstratetodetectactivityintheabsence(greybars)orpresence(blackbars)of10 μMDEX.Barsrepresentthemeanvaluefrom3technicalreplicatesperindividual±SD.*P  0.05(Additionalfile 7).Assuch,thepresenceofthep35Spromoter(andofanyenhancerswithinit)didnotinfluencereportergeneexpression.Thisconclusionwasvalidatedbytheobservationthatverylittle‘leaky’reportergeneexpressionwasobservedintheabsenceofinducer(Fig. 2e).Itremainspossible,however,thatconstructdesignsthatbringthepOp6induciblepromoterclosertop35SmightresultintransgeneactivationintheabsenceofDEX,forexampleasinYooetal.[21].SimilarcomparisonsbetweenGUSactivitylevelsinlinestransformedwithconstruct17610(nop35S;GUS + introns)(Fig. 2b,f)andthosewith17613(nop35S;GUSnointrons)(Fig. 2c,g)revealedthatthepresenceofintronsinthereportergenehadanenhancingeffectonactivitylevelsbothintheabsenceandpresenceofinducer.Inthepresenceofinducer,thisdifferencewasstatisticallysignificantinaWilcoxon’sranktestatP