THEJOURNALOFBIOLOGICALCHEMISTRY VOL.287,NO.20,pp.16267–16275,May11,2012 ©2012byTheAmericanSocietyforBiochemistryandMolecularBiology,Inc. PublishedintheU.S.A. Deciphering Modern Glucocorticoid Cross-pharmacology Using Ancestral Corticosteroid Receptors□S Receivedforpublication,January25,2012,andinrevisedform,February21,2012Published,JBCPapersinPress,March21,2012,DOI10.1074/jbc.M112.346411 JeffreyA.Kohn,KirtiDeshpande,andEricA.Ortlund1 FromtheDepartmentofBiochemistryandtheDiscoveryandDevelopmentalTherapeuticsProgram,WinshipCancerInstitute, EmoryUniversitySchoolofMedicine,Atlanta,Georgia30322 Background:Drugsthattargetsteroidreceptorsarenotoriouslypromiscuous,causinganarrayofoff-targetsideeffects. Results:ReversalofthehistoricalmutationH853Rinthemineralocorticoidreceptor(MR)fullyrestoresagonistactivityby mometasonefuroate,anMRantagonist. Conclusion: A single residue outside of the ligand-binding pocket toggles agonism versus antagonism response by MR to syntheticligands. Significance:Ancestralproteinsareidealtoolstoelucidatethemechanismsofdrugselectivity. Steroid receptors (SRs) are the largest family of metazoan reproductivediseases,andcancer.Overtime,avastrepertoire transcription factors and control genes involved in develop- ofreceptorshasevolvedtorespondtosmallchemicalstimuli, ment,endocrinesignaling,reproduction,immunity,andcancer. making them attractive pharmacological targets. However, The entire hormone receptor system is driven by a molecular becausemostreceptorsbelongtolargeclassesofevolutionary- switchtriggeredbythebindingofsmalllipophilicligands.This relatedproteinsthatshowhighstructuralsimilarity,targetinga makes the SRs ideal pharmaceutical targets, yet even the best single receptor subtype is a major challenge. Poor selectivity clinically approved synthetic steroidal agonists are prone to can cause serious off-target side effects, as seen in the treat- cross-reactivity and off-target pharmacology. The mechanism mentsofmajordepression(3),heartdisease(4,5),asthma(4,5), underlying this promiscuity is derived from the fact that SRs andallergies(6). sharecommonstructuralfeaturesderivedfromtheirevolution- To fully understand the mechanisms supporting receptor- aryrelationship.Moreoftenthannot,rationalattemptstoprobe ligandrecognition,selectivity,andactivation,robuststructure- SRdrugselectivityviamutagenesisfailevenwhenhighquality functionrelationshipsmustbebuiltfromextensivemutational structuralandfunctionaldataareavailableduetothefactthat analysisandliganddesign.Thisanalysisishinderedforseveral important mutations often result in nonfunctional receptors. reasons.First,aminoacidresiduesconferringproteinfunction ThishighlightsthefactthatSRssufferfrominstability,prevent- and ligand specificity between homologous receptors can be ingin-depthmutationalanalysisandhamperingcrystallization difficult to identify among the vastly more prevalent neutral of key receptor-ligand complexes. We have taken a unique mutations that accumulate over time (7). Second, restrictive approachtoaddressthisproblembyusingaresurrectedances- mutations that are not directly related to the protein-ligand tralproteintodeterminethestructureofapreviouslyintracta- interactioncanaccumulateinextantproteins,preventingthe blecomplexandidentifiedthestructuralmechanismsthatcon- tolerance of function-shifting mutations (8). Third, many fer activation and selectivity for a widely used glucocorticoid, mutationsaredestabilizingandresultinlossofproteinfunc- mometasonefuroate.Moreover,wehaveidentifiedasingleres- tion,complicatingthedistinctionbetweenaneffectthatisspe- iduelocatedoutsideoftheligand-bindingpocketthatcontrols cific to the protein-ligand interaction versus an effect that is mometasonefuroateantagonismversusagonisminthehuman globallyinactivatingtotheprotein(7).Althoughmostconclu- mineralocorticoidreceptor. sionsarecurrentlydrawnfromfunction-killingmutations,the insightneededtounderstandligandselectivityamongaclassof homologous proteins would be better drawn from function- Complexlifedependsonintra-andintercellularcommuni- shiftingmutationsthatpreservereceptoractivation. cation, whereby secreted messengers are detected by specific Theseproblemshavehinderedthedesignofselectivedrugs receptorstoregulatemetabolism,reproduction,cellcycles,and thattargethumansteroidreceptors(SRs).2SRsareafamilyof more.Thiscoordinationtightlycontrolscellularactivitywithin ligand-regulated transcription factors that control genes thehigherorganism.Poorcoordinationoftheseprocessescan involved in development, endocrine signaling, reproduction, resultinmanyhealthconcerns,includingmetabolicdisorders, immunity,andcancer(9).Thismakesthemattractivepharma- ceutical targets. Although SRs show exquisite selectivity for □S ThisarticlecontainssupplementalFigs.S1andS2. Theatomiccoordinatesandstructurefactors(code4E2J)havebeendepositedin 2Theabbreviationsusedare:SR,steroidreceptor;PR,progesteronereceptor; theProteinDataBank,ResearchCollaboratoryforStructuralBioinformatics, AR,androgenreceptor;MR,mineralocorticoidreceptor;GR,glucocorticoid RutgersUniversity,NewBrunswick,NJ(http://www.rcsb.org/). receptor; AncSR, ancestral SR; AncGR, ancestral GR; TIF2, transcription 1To whom correspondence should be addressed: Dept. of Biochemistry, intermediaryfactor2;MOF,mometasonefuroate;ANOVA,analysisofvari- EmoryUniversitySchoolofMedicine,Atlanta,GA30322.Tel.:404-727- ance;HSD,honestlysignificantdifference;AncCR,ancestralcorticosteroid 5014;Fax:404-727-2738;E-mail:[email protected]. receptor;LBP,ligand-bindingpocket. MAY11,2012•VOLUME287•NUMBER20 JOURNALOFBIOLOGICALCHEMISTRY 16267 This is an Open Access article under the CC BY license. DecipheringGlucocorticoidCross-pharmacologyUsingAncSRs theirendogenoushormones,SR-targetingdrugstendtobepro- transformed into Escherichia coli strain BL21(DE3) and miscuousandcausemanyoff-targetsideeffects(10,11).Thisis expressed as a maltose-binding protein-His fusion. Cultures becauseSRs,consistingoftheestrogenreceptor,progesterone (1.3litersinTerrificBroth)weregrowntoanA of0.6–0.7 600 receptor (PR), androgen receptor (AR), mineralocorticoid andinducedwith400(cid:1)Misopropyl(cid:2)-D-thiogalactopyranoside receptor (MR), and glucocorticoid receptor (GR), descended and 50 (cid:1)M MOF at 30°C for 4 h. Cell mass was collected by fromacommonancestor(cid:1)500millionyearsago(seeFig.1A) centrifugation at 4000 (cid:2) g for 15 min, lysed, and purified by and show high structural similarity (9, 12). In the absence of nickelaffinitychromatography.Themaltose-bindingprotein- ligand, SRs remain partially unfolded and associate with heat Histagwascleavedbytobaccoetchvirusproteaseat4°Cover- shockproteins(13,14).Thisinstabilityisnecessarytopermit nightwithsimultaneousdialysisintobuffercontaining300mM theconformationalchangesthatdrivereceptoractivationupon NaCl, 20 mM Tris (pH 7.4), 5% glycerol and purified to ligandbinding(1,2,7,14–16),butithaslimitedourabilityto homogeneity by nickel affinity, followed by gel filtration probe receptor-ligand interactions via mutagenesis, as many chromatography. mutationsofinterestdisabletheproteinentirely. Crystallization, Data Collection, and Structural Refine- Withtheadvancementofwhole-genesynthesisandpioneer- ment—PureAncGR2wasconcentratedto3–5mg/mlinbuffer ingeffortsmadeincomputationalandevolutionarybiology,it containing300mMNaCl,20mMTris(pH7.4),5%glycerol,50 is now possible to predict and “resurrect” ancestral genes. (cid:1)MCHAPS,and50(cid:1)MMOF.Crystalsweregrownviahanging Ancestral gene reconstruction is used to study the molecular dropvapordiffusionat4°Cfromsolutionscontaining0.75(cid:1)lof evolutionofabiologicalsystem(12,17–19)butshowspromis- AncGR2-TIF2-MOF solution, 0.75 (cid:1)l of 1.5–3 M ammonium ing applications to the process of drug design. By comparing formate,andadodecapeptidederivedfromtheGRcoactivator twoancestralproteinsfromnodesonanevolutionarytree,we human TIF2 ((cid:3)H N-ENALLRYLLDKD-CO(cid:4), SynBioSci 3 2 are provided with a smaller subset of possible amino acid Corp.).Crystalswerecryoprotectedbyimmersionin1.5–3M replacementstodissectbetweenrelatedproteinsthathavedif- ammoniumformatecontaining25%glycerolandflash-frozen ferentligandspecificities.Oureffortscanbefocusedonfewer inliquidnitrogen.Datatoaresolutionof2.5Åwerecollectedat residues when probing structure-function relationships than the Southeast Regional Collaborative Access Team at the whenlookingonlyatextantproteins.Thisapproachtherefore AdvancedPhotonSource(Argonne,IL)(Table1).Thestruc- allows us to avoid interference from neutral and restrictive tureoftheAncGR2-MOF-TIF2complexwassolvedbymolec- mutations that have accumulated over time. Furthermore, ular replacement using PHASER in the CCP4 software suite. unlikemanyextantproteins,ancestralproteinsshowremarka- ModelbuildingandrefinementwereperformedusingRefmac ble tolerance toward changes in function-shifting residues, and Coot. Cavity volumes were calculated using CASTp, and makingthemmorestableunderlaboratoryconditions. figuresweregeneratedinPyMOL.TherefinedAncGR2-MOF Wehypothesizedthatonecouldexploittheseancestralpro- structure has been deposited in the Protein Data Bank (code teinstounderstandcross-pharmacologyinhumanSRs.Ances- 4E2J). tral SRs (AncSRs) are more tolerant to mutation than their Mutagenesis—Wild-type AncGR1 (GenBankTM accession extant descendants (20), and their molecular and structural number JF896321.1) and AncGR2 were subcloned into a evolutionhasalreadybeencharacterized(8,12,20–23).Anc- pMCSG7-maltose-bindingprotein-Hisexpressionvector,and SRsthereforemakeaneffectivemodeltostudythestructural the following mutations were created from these constructs: mechanisms of SR pharmacology. To achieve this goal, we AncGR1-S106P, AncGR1-S106P/L111Q, AncGR1-S106P/L111A, determinedthestructureoftheancestralglucocorticoidrecep- AncGR2-P106S, AncGR2-P106S/Q111L, and AncGR2-P106S/ tor2(AncGR2)ligand-bindingdomainincomplexwithafrag- Q111A.MutagenesiswasperformedusingaQuikChangeIIXL mentofhumantranscriptionintermediaryfactor2(TIF2)and kit(Stratagene). mometasone furoate (MOF). We draw upon functionally Ligand Binding Assays—Wild-type or mutant AncGR1 or importanthistoricalaminoacidsubstitutionstoelucidatethe AncGR2wasexpressedasdescribedaboveandassayedpriorto mechanismsdrivingGRactivationforthiswidelyusedgluco- tobaccoetchvirusproteasecleavageaspurifiedmaltose-bind- corticoid. Furthermore, we use a combination of structural ing protein fusion proteins. All fluorescence polarization analysisandfunctionalassaystoexplaintheselectivityofthis experiments were performed in buffer containing 150 mM drugagainstMRandARandstrongcross-reactivitywithPR. NaCl, 10 mM HEPES (pH 7.4), 5 mM DTT, 3 mM EDTA, and 0.005%Tween20.Bindingaffinityfordexamethasone-fluores- EXPERIMENTALPROCEDURES ceinwasmeasuredwithaconstantconcentrationof12nMdex- Chemicals and Reagents—Chemicals were purchased from amethasone and a variable protein concentration of 10(cid:4)10– SigmaorFisher.ThepMALCH10TandpMCSG7-basedpro- 10(cid:4)5 M. Competition assays were performed at a protein teinexpressionvectorsweregiftsfromJ.Tesmer(Universityof concentration1.2timesitsbindingaffinityfordexamethasone Texas, Austin) and J. Sondek (University of North Carolina, in the presence of 12 nM dexamethasone and 10(cid:4)10–10(cid:4)5 M ChapelHill),respectively.TheemptypSG5-basedmammalian competingligand.DatawereprocessedwithGraphPadPrism5. expression vector and human MR and GR pSG5 constructs, Statisticalsignificancewasdeterminedbytwo-factoranalysisof pFRlucreporter,andphRLtkreporterweregiftsfromJ.Thorn- variance(ANOVA),andindividualcomparisonsweremadewith ton(UniversityofOregon,Eugene). Tukey’shonestlysignificantdifference(HSD)posthoctests. ProteinExpressionandPurification—AncGR2(GenBankTM In-cellActivationAssays—Allancestral,andmutantligand- accessionnumberEF631976.1)inapMALCH10Tvectorwas binding domains were cloned into a pSG5 expression vector 16268 JOURNALOFBIOLOGICALCHEMISTRY VOLUME287•NUMBER20•MAY11,2012 DecipheringGlucocorticoidCross-pharmacologyUsingAncSRs TABLE1 Datacollectionandrefinementstatistics AU,asymmetricunit;r.m.s.d.,rootmeansquaredeviation;PDB,ProteinDataBank. Resolution(highestshell;Å) 2.50(2.59–2.50) Spacegroup P6 Unitcelldimensions a(cid:5)1104.4,b(cid:5)104.4,c(cid:5)143.9Å;(cid:3)(cid:5)(cid:2)(cid:5)90.0°,(cid:4)(cid:5)120.0° No.ofreflections 30,710 R (highestshell)a 7.7%(42.2%) sym Completeness(highestshell) 99.90%(98.96%) Averageredundancy(highestshell) 8.0(7.9) I/(cid:5) 29.3(5.3) Monomers/AU 2 No.ofproteinatoms/AU 4195 No.ofligandatoms/AU 85 No.ofwaters/AU 151 R b(R )c 20.5(25.5) working free AverageB-factors(Å2) Protein 45.0 Ligand 53.5 Water 45.5 r.m.s.d. Bondlengths(Å) 0.005 Bondangles 1.078° PDBcode 4E2J a R (cid:5)(cid:6)(cid:1)I(cid:4)(cid:7)I(cid:8)(cid:1)/(cid:6)(cid:1)I(cid:1),whereIistheobservedintensity,and(cid:7)I(cid:8)istheaverageintensityofseveralsymmetry-relatedobservations. bRsym (cid:5)(cid:6)(cid:2)F(cid:1)(cid:4)(cid:1)F(cid:2)/(cid:6)(cid:1)F,whereF andF aretheobservedandcalculatedstructurefactors,respectively. cRwork(cid:5)ing(cid:6)(cid:2)F(cid:1)(cid:4)o(cid:1)F(cid:2)/(cid:6)c(cid:1)F foro7%oftheodatancotusedatanystageofthestructuralrefinement. free o c o immediatelyfollowingaGal4DNA-bindingdomainandaGR hingesequence.CHO-K1cellsweregrownandmaintainedin phenolred-freecomplete(cid:3)-minimalessentialmedium(Invit- rogen)supplementedwith10%charcoal/dextran-strippedFBS (Invitrogen) and penicillin/streptomycin. Cells grown in 96-well assay plates were transfected at 70–90% confluence with1ngofreceptor,100ngofupstreamactivatorsequence- drivenfireflyluciferasereporter(pFRluc),and0.1ngofconsti- tutiveRenillaluciferasereporter(phRLtk)for4husingLipo- fectamine 2000 in Opti-MEM I (Invitrogen). Transfections wereendedbyreplacementwithcomplete(cid:3)-minimalessential medium, and cells were allowed to recover overnight. After recovery, cells were treated in triplicate with 10(cid:4)12–10(cid:4)6 M ligand or vehicle (Me SO) in complete (cid:3)-minimal essential 2 mediumfor24h(finalworkingMe SOof1%)andthenassayed 2 with Dual-Glo luciferase substrate (Promega). Firefly activity was normalized to Renilla activity, and the -fold increase in activationwascalculatedrelativetothevehiclecontrol.Dose- responsecurvesweregeneratedinGraphPadPrism5.Statisti- cal significance was determined by two-factor ANOVA, and individualcomparisonsweremadewithTukey’sHSDposthoc tests. RESULTS FIGURE1.Evolutionaryhistoryofcorticosteroidreceptorsandstructure AncGR2-TIF2-MOF Crystal Structure—MOF is a powerful ofAncGR2-TIF2-MOF.A,simplifiedphylogenetictreedepictingtheevolu- topicalanti-inflammatorydrugfortheskinandairwaysandis tionofcorticosteroidreceptors.Activatinghormonesarelistedontheright. ER,estrogenreceptor;ERR,estrogenreceptor-relatedreceptor.B,structureof the active ingredient of Nasonex, Asmanex, and Elocon (24). AncGR2(red)incomplexwithhumanTIF2(green)andMOF(cyan).C,AncGR2 Although MOF has been in clinical use for over 24 years, it LBPresidues(red)withMOFshown(cyan). suffersfromseverecross-pharmacology,resultinginunwanted side effects and limiting its use to topical applications. MOF bindingrequiresarearrangementofthehelixH6-H7regionof strongly activates GR, cross-reacts with PR, and is selective the receptor to accommodate the large 17(cid:3)-furoate moiety, against AR and MR. The ternary AncGR2-TIF2-MOF crystal inducing a 200-Å3 (1.3-fold) increase in the volume of the structurerevealsthestructuralbasisforMOFbindingtoverte- ligand-binding pocket relative to dexamethasone; this high- brate GRs (Fig. 1, B and C, and supplemental Fig. S1). The lightstheabilityofSRstoexpandtheirligand-bindingpockets hydrogenbondnetworkthatisrequiredforactivationofcorti- to accommodate exogenous ligands (26, 27). The AncGR2- coidreceptors(25)isintactandisstabilizedbyadipole-dipole TIF2-MOFstructurealsorevealsthatastrongH-bondisnot interactionbetweenMOFC21-ClandAncGR2Asn-33.MOF possiblebetweenMOFandGln-111ofGR(Fig.1C),aninter- MAY11,2012•VOLUME287•NUMBER20 JOURNALOFBIOLOGICALCHEMISTRY 16269 DecipheringGlucocorticoidCross-pharmacologyUsingAncSRs Structural and Evolutionary Basis for Selectivity against MR—Ourstructurealso suggestsamechanismfortheselec- tivityofMOFagainstMRandAR.H6andH7,whichborderthe 17(cid:3)-bindingarea,arepartiallyunwoundandstabilizedbyPro- 637/Pro-106inGR/AncGR2,accommodatingthefuroatemoi- ety(Fig.3A,left);MR,AR,andAncGR1haveaserineatthissite that caps H7, positioning the helix within 2.5 Å of where the furoate would rest, creating a steric incompatibility (Fig. 3A, right).Wehaveshowninpreviousworkthat,duringtheevolu- tionofGR,S106PandL111Qsubstitutionswerecriticalinboth reshapingtheH6-H7regionofthereceptorandgeneratinga new H-bond with the 17-OH moiety of cortisol, the endoge- nousglucocorticoid(8,22).Totesttheeffectofreversingthese criticalsubstitutionswithrespecttoMOFbindingaffinity,we generatedtwoAncGR2mutants,P106SandP106S/Q111L,and measuredtheirbindingaffinitiesforcortisol,dexamethasone, and MOF using fluorescence polarization competition assays against dexamethasone-fluorescein. The P106S reversal reducedtheaffinityofAncGR2forallthreeligandsbyonlyan orderofmagnitude(Fig.3B).Thisresultwassurprisingbecause theP106Smutationwasidentifiedasthedrivingforcebehind theH6-H7rearrangementrequiredtoopenspaceintherecep- torforspecificrecognitionofhormoneswithC17substituents (13, 22). Because MOF binding requires this structural rear- rangement(Fig.3A),Pro-106likelyplaysaroleinstabilizingthe H6-H7loopinaproductivebindingmodebutisnotabsolutely required to induce this structural change. AncGR2-P106S/ Q111L, which is known to be inactive to endogenous ligands (8), did not bind dexamethasone-fluorescein (supplemental Fig.S2).Thispreventedcompetitionassaysonthismutantbut FIGURE2.Structuralbasisforoff-targetactivationofPR.A,thehumanPR (hPR)-MOFcomplex(white;ProteinDataBankcode1SR7)superimposedon suggestedthatH7indeedrepositionstoplaceLeu-111incon- theAncGR2-MOFcomplex(red).PRresiduePhe-794maintainsspaceforthe tactwithC17ofthesteroid.Thisgeneratesapolarincompati- 17(cid:3)-furoatemoietyandcontributesahydrophobicinteractionviathearom- bility with C17-OH-containing steroids, such as cortisol and atizedsidechain(15).PRresidues791ESSF794onH7appeartoplayakeyrolein allowingstrongMOFbindingbypositioningtheH6-H7loopandH7viaa dexamethasone, and introduces a steric clash with the bulky conservedGlu-791–Ser-793H-bond.B,thismotifisstrictlyconservedamong furoatesubstituentofMOF.Totestthishypothesis,wegener- extantPRsbutisnotpresentinAncSR2. ated a P106S/L111A mutant, designed to alleviate this steric clash in the Pro-106 background, which restored binding to action that plays a critical role in the specific recognition of MOFanddexamethasone(Fig.3B).Asexpected,cortisolbind- 17(cid:3)-OH-substitutedligandsandisabsolutelyrequiredforcor- ingwasonlymarginallyrestoredbecausecortisoldoesnotcon- tisol activation (8). Instead, hydrophobic interactions replace taintheadditionalbulkyhydrophobicgrouppresentonMOF thisinteractioninafashionanalogoustothestructureofthe tostabilizethecoreofthereceptorintheabsenceofthecritical GR-fluticasonefuroatecomplex(28). Gln-111–17-OHH-bond.Interestingly,dexamethasonebind- Structural and Evolutionary Basis for PR Cross-reactivity— ingwasmorefullyrestoredthancortisolbinding,presumably TheAncGR2-TIF2-MOFstructureallowsforthedirectstructural duetoadditionalinteractionsonitsmodifiedbackbone.Thus, comparison of GR-MOF and PR-MOF complexes and reveals theH6-H7regionofthereceptorcanadoptanexpandedcon- how additional space in the H6-H7 loop region is created to formationintheabsenceofPro-106,suggestingthattheH6-H7 accommodatethefuroatemoiety.PRresidues791ESSF794onH7 region of GRs is inherently flexible, allowing it to adapt to appear to play a key role in allowing strong MOF binding by ligand-inducedperturbation.Thisreshapesourunderstanding expandingthepocketviaaconservedGlu-791–Ser-793H-bond of the role of the H6-H7 region within the ligand-binding betweentheH6-H7loopandH7(Fig.2A).Stericbulkprovidedby domainintherecognitionofsyntheticglucocorticoids. PRPhe-794betweenH7andH3maintainsspaceforthe17(cid:3)- WehaveshownpreviouslythatAncGR1,whichprecededthe furoate moiety and contributes a hydrophobic interaction evolutionofAncGR2,isalowsensitivityMR-likereceptorwith viathearomatizedsidechain(26).Thismotifisstrictlycon- activationbybothmineralocorticoidsandglucocorticoids(22). served among PRs but is not present in AncSR2 (the com- BecauseMOFisselectiveagainstMR,wereasonedthatMOF mon ancestor of all 3-keto-SRs) or AncSR3 (the common woulddisplaysimilarselectivityagainstAncGR1.Surprisingly, ancestor of PR and AR) (Figs. 1A and 2B). Therefore, PR MOFboundAncGR1withanaffinitycomparablewithdexam- responsetoMOFwasprobablyalateevolutionaryderivation ethasoneandcortisol(Fig.3C),indicatingthatAncGR1H7had resultinginthiscross-reactivity. already acquired the plasticity needed to accommodate the 16270 JOURNALOFBIOLOGICALCHEMISTRY VOLUME287•NUMBER20•MAY11,2012 DecipheringGlucocorticoidCross-pharmacologyUsingAncSRs FIGURE3.BindingofMOFbymodernandancestralSRsrequiresexpansionoftheLBP.A,structureofthe17(cid:3)-bindingpocket.LikePRandAncGR2,human GR(hGR)hasanextendedH6-H7loopconformation(left);MR,AR,andAncGR1haveatightenedH6-H7thatwouldcreateastericincompatibilitywiththe furoate(right).BandC,thebindingaffinitiesofAncGR2(B)andAncGR1(C)mutantsfortheindicatedligandweremeasuredbyfluorescencepolarization competitionwithdexamethasone-fluorescein(Dex).AncGR2-P106S/Q111L(AncGR2-SL)didnotbinddexamethasone-fluorescein,andcompetitionexperi- mentscouldnotbeperformedforthisreceptor.Statisticalanalyseswereperformedusingtwo-factorANOVA,withTukey’sHSDposthoctestsusedfor individualcomparisons.Comparisonsfoundtobestatisticallysignificanttop(cid:9)0.05aremarked.*,comparedwiththesameligandbindingforthewild-type receptor;#,comparedwithdexamethasonebindingforthesamemutant;†,comparedwithMOFbindingforthesamemutant.AncGR1-SA,AncGR2-P106S/ Q111A;AncGR1-PQ,AncGR1-S106P/L111Q;AncGR1-PA,AncGR1-S106P/L111A.DandE,receptoractivationforGR-like(D)andMR-like(E)receptorswas measuredbyDual-LuciferasereportergeneactivationintransientlytransfectedCHO-K1cellcultures.Themean(cid:10)S.E.isshown(n(cid:5)3).Statisticalanalyseswere performedusingtwo-factorANOVA,withTukey’sHSDposthoctestsusedforindividualcomparisons.Comparisonsfoundtobestatisticallysignificanttop(cid:9) 0.05aremarked.*,comparedwithactivationofthesamereceptorbycortisol;#,comparisonsmadeasindicatedonthefigure.X,nobindingoractivation observed. bulkyfuroatemoiety.TheforwardmutationsAncGR1-S106P A Single Residue Controls MOF Selectivity and Transcrip- andAncGR1-S106P/L111Qhadnoeffectsspecifictoapartic- tional Activity—To determine the structural differences ular ligand, but the AncGR1-S106P/L111A mutation selec- betweenGRandMRthatgovernMOFrecognition,wecharac- tively reduced cortisol binding while leaving dexamethasone terizedtheabilityofMOFtodriveluciferasereportergeneacti- and MOF binding unaffected. This is presumably due to the vation across the entire ancestral corticosteroid phylogeny. removalofa17(cid:3)-interaction.Thesedatashowthatreceptor- Although MOF only very weakly activated MR (Fig. 4A), it ligand interactions at the 17(cid:3)-site are important for effective strongly activated AncGR1 and the ancestral corticosteroid ligand binding, although poor interactions here can be sur- receptor (AncCR) with a subnanomolar potency, comparable mounted by stronger interactions elsewhere along the ligand withthestrongactivationseeninAncGR2andGR(Fig.4,Aand scaffold. B).Thisprovidesfurtherevidencethatthecorticoidreceptors MAY11,2012•VOLUME287•NUMBER20 JOURNALOFBIOLOGICALCHEMISTRY 16271 DecipheringGlucocorticoidCross-pharmacologyUsingAncSRs is unable to stabilize an active receptor conformation (25). Thus,MRmusthaveaccumulatedepistaticchangesthatpro- hibit activation from this drug. We therefore examined the importanceofresiduesthatchangedonthelineageleadingto MRwithrespecttoMOFactivation.Mutationofresiduesinor near the ligand-binding pocket had no significant impact on MOF activation without affecting receptor activation toward cortisolanddexamethasone,consistentwithourfluorescence polarization competition assays (Fig. 3, D and E) (data not shown).Wethereforelookedforchangesoutsideoftheligand- binding pocket and activation function surface. Arg-116 and Gln-120inAncCR,correspondingtoHis-853andLeu-857in MR,respectively,arelocatedonthesolvent-exposedfaceofH7 andinteractwiththemainchainoftheloopbetweenH5and (cid:2)1, at AncCR residues Gly-87 and Met-89 (MR residues Ser- 824andPhe-826)(Fig.5A).Theseresiduesare(cid:11)14–18Åfrom the furoate moiety of MOF and closest to the B ring of the steroid(10–16Å),yetreversalofallfourresiduesinMRtotheir ancestralstates(MR-GMRQ)completelyrestoredMOFactiva- tion(Fig.5B).Wenarroweddownthecauseofthiseffect,first tothoseresiduesonhelix7(hMR-RQ),andfurthertothesingle residueatMRsite853.Reversalofthissiteviathesubstitution H853R conferred full MOF activation (Fig. 5B). In wild-type MR,His-853interactswiththemainchainatomsintheH5-(cid:2)1 loopandappearstostabilizetheMR-likeconfigurationofH7, whichmustunwindtosupportactivationbyligandswithbulky C17(cid:3)substituents. The stronger interaction provided by an argininesubstitutionatthissitestabilizedMR-H853Rtoenable MOFactivation(Figs.5Aand6).Importantly,thesechangesare neutralwithrespecttoactivationbycortisol:neithertheEC 50 nor activation of cortisol was affected by the MOF selectivity mutations(Fig.5,BandC),indicatingthatthestructuraldeter- minantsofMOFactivationareuniquefromthosethatsupport theendogenousligandrecognition.Introducingtheequivalent forwardsubstitutionsinAncCR(AncCR-SFHL)failedtoabro- gate MOF response (Fig. 5B), which is in line with the more promiscuousphenotypeoftheancestralprotein.Intriguingly, makingtheequivalentsitemutationshorizontallybetweenMR and GR (MR-H853L/L857S and GR-L647H/S651L) not only failedtoenableMOFactivationinMRbutalsoabrogatedacti- vationbycortisolanddexamethasoneinGR(Fig.5,DandE). MutationsattheseresiduesduringtheevolutionofGRswere previouslyidentifiedtobedestabilizingtoGRs,contributingto FIGURE4.Activationofmodernandancestralcorticosteroidreceptorsby thelowaffinitybuthighselectivityofmodernGRsforendoge- syntheticglucocorticoids.A,corticoidreceptorphylogenywithresponseto syntheticglucocorticoids(dexamethasone(Dex)/MOF)isshown.Fullago- nous glucocorticoids (20). Here, disruption of this site in GR nismisshowningreen,andweakornoagonismisshowninred.HumanGR fullydestabilizedtheactivereceptorduringcortisolanddexa- (hGR)andMR(hMR)wereusedtorepresentextantmammalianGRandMR. methasone binding. In contrast, MOF expanded the ligand- The-foldactivation(B)andpotency(C)ofcorticosteroidreceptorligand- bindingdomainsweremeasuredviaDual-Luciferasereportergeneactiva- bindingpocket(LBP)tomakeadditionalhydrophobiccontacts tionintransientlytransfectedCHO-K1cells.Themean(cid:10)S.E.isshown(n(cid:5)3). offeredbythefuroatering(Fig.6)andwasabletostabilizethe Forthepurposeofthisresearch,activationbelow10-foldoverthecontrol(B, red line) was considered weak agonism/antagonism, whereas activation active conformation, albeit at a much lower potency than in abovethisthresholdwasconsideredfullagonism. wild-typeGR(Fig.5,DandE). Weanticipatethatthefindingsproducedbythisstudywillbe fromAncCRtoAncGR2areabletounwindH6-H7toaccom- applicabletoligandsthatprotrudeintoextrasteroidalbinding modate the MOF 17(cid:3)-furoate moiety without requiring the regions within the LBP. Furthermore, the finding that ligand S106Psubstitution. specificityisstronglyinfluencedbystructuralfeaturesthatlie Furthermore,wehaveshownthatMOFexhibitsitsselectiv- welloutsideoftheLBPmustbetakenintoconsiderationduring ityforGRoverMRnotviaadifferenceinpotencybutratherin thedevelopmentoffuturedrugs.Thefactthatthesesitescould efficacy:althoughMOFbindsMR,withan(cid:11)100nMpotency,it notbeidentifiedusingextantproteinshighlightsthepowerof 16272 JOURNALOFBIOLOGICALCHEMISTRY VOLUME287•NUMBER20•MAY11,2012 DecipheringGlucocorticoidCross-pharmacologyUsingAncSRs FIGURE5.Distalresiduescontrolcorticosteroidspecificity.A,keyresiduespreceding(cid:2)-sheet1andH7inAncCRandhumanMR(hMR)werecross-mutated. The-foldactivation(B)andpotency(C)weremeasuredviaDual-LuciferasereportergeneactivationintransientlytransfectedCHO-K1cells.Dex,dexametha- sone.ThesameresiduesinGRandMRwerecross-mutated,andthe-foldactivation(D)andpotency(E)weremeasuredviaDual-Luciferasereportergene activationintransientlytransfectedCHO-K1cells.Themean(cid:10)S.E.isshown(n(cid:5)3).Statisticalanalyseswereperformedusingtwo-factorANOVA,withTukey’s HSDposthoctestsusedforindividualcomparisons.Comparisonsfoundtobestatisticallysignificanttop(cid:9)0.05aremarked(*).hGR-HL,humanGR-L647H/ S651L;hMR-LS,humanMR-H853L/L857S.X,nobindingoractivationobserved. usingancestralgenereconstructiontoidentifytheobscurecon- as we are left with mesophilic proteins to use for structure- served structural mechanisms that support activation via functionanalysis.ThisisexemplifiedintheSRfamilyand,in endogenousversussyntheticligandsthatmaybeexploitedby particular,withmodernGRs,whicharenotoriouslydifficultto selectivetherapeutics. manipulateunderlaboratoryconditions(23,29).Furthermore, modernproteinshaveaccumulatedmillionsofyearsofneutral DISCUSSION mutationsthatmakeitdifficulttoidentifyfunctionallyimpor- Wehavesuccessfullyadaptedancestralgenereconstruction tantaminoacidresidues,aswellasrestrictivemutationsthat toshedlightonthestructuralmechanismsofdrugselectivity canfurtherprohibitmutationalanalysis. for SRs. Our approach combines structural and evolutionary Workaroundstotheseproblemsarelimitedandfrequently biology to overcome many of the obstacles that frequently involve the incorporation of stabilizing mutations. Although hinderproteinresearchusingmodernproteins.Itiswellknown this approach does improve the stability of modern proteins, thatfunction-shiftingaminoacidchangesarenottoleratedwell including GR (23, 29), mutations such as these may alter the inmodernproteinsbecausemostproteinsareonlymoderately wayligandsinteractwiththeirtargetreceptors.Asaresult,the stable (7, 15, 16). They display a narrow thermal window of behaviorofthesemutantsmaynotaccuratelymirrorthebehav- activitydictatedbytheeffectsofnaturalselectiononbothther- iorofthewild-typeproteins.Incontrast,ancestralproteinsare malandkineticstability(15)andbytheaccumulationofneutral subjectedtorigoroustestingduringthereconstructionprocess mutationsoverevolutionarytime(7).Afinebalanceisneces- toensuretheirbehaviorisconsistentwiththebehaviorofother sarytoallowsmallperturbationsorsignals,suchasligandbind- ing, to functionally alter protein structure: although too little proteinswithintheirphylogeny(e.g.thatthestructuralmech- stability prevents proper protein folding, too much stability anisms for activation are conserved). Ancestral proteins are prevents a receptor from adopting an active conformation in inherently more tolerant to mutation and may serve as ideal responsetostimuliwithinthehostorganism.Wearetherefore models in which to study structure-activity relationships for limitedbytheeffectsofbothnaturalselectionandneutraldrift, moderatelystableeukaryoticproteins(8,22,23).Evenwhenthe MAY11,2012•VOLUME287•NUMBER20 JOURNALOFBIOLOGICALCHEMISTRY 16273 DecipheringGlucocorticoidCross-pharmacologyUsingAncSRs 3. Remick,R.A.(1988)Anticholinergicsideeffectsoftricyclicantidepres- santsandtheirmanagement.Prog.Neuropsychopharmacol.Biol.Psychia- try12,225–231 4. Decalmer,P.B.,Chatterjee,S.S.,Cruickshank,J.M.,Benson,M.K., andSterling,G.M.(1978)Beta-blockersandasthma.Br.HeartJ.40, 184–189 5. Hanania,N.A.,Singh,S.,El-Wali,R.,Flashner,M.,Franklin,A.E.,Garner, W.J.,Dickey,B.F.,Parra,S.,Ruoss,S.,Shardonofsky,F.,O’Connor,B.J., Page,C.,andBond,R.A.(2008)Thesafetyandeffectsofthebeta-blocker, nadolol,inmildasthma:anopen-labelpilotstudy.Pulm.Pharmacol.Ther. 21,134–141 6. Roy,M.,Dumaine,R.,andBrown,A.M.(1996)HERG,aprimaryhuman ventriculartargetofthenonsedatingantihistamineterfenadine.Circula- tion94,817–823 7. Taverna,D.M.,andGoldstein,R.A.(2002)Whyareproteinsmarginally stable?Proteins46,105–109 8. Bridgham,J.T.,Ortlund,E.A.,andThornton,J.W.(2009)Anepistatic ratchetconstrainsthedirectionofglucocorticoidreceptorevolution.Na- ture461,515–519 9. Carson-Jurica,M.A.,Schrader,W.T.,andO’Malley,B.W.(1990)Steroid receptorfamily:structureandfunctions.Endocr.Rev.11,201–220 10. Stahn,C.,Löwenberg,M.,Hommes,D.W.,andButtgereit,F.(2007)Mo- lecularmechanismsofglucocorticoidactionandselectiveglucocorticoid receptoragonists.Mol.Cell.Endocrinol.275,71–78 11. Madauss,K.P.,Stewart,E.L.,andWilliams,S.P.(2007)Theevolutionof progesteronereceptorligands.Med.Res.Rev.27,374–400 FIGURE6.SchematicsummarizingrelevantfeaturesofSRLBPthatdic- 12. Thornton,J.W.(2001)Evolutionofvertebratesteroidreceptorsfroman tateMOFactivation.Activationoccurswhenaligand(e.g.MOF(cyan))binds ancestralestrogenreceptorbyligandexploitationandserialgenomeex- totheLBP,stabilizingtheAF-2helix(darkred)toallowforcoactivatorbinding pansions.Proc.Natl.Acad.Sci.U.S.A.98,5671–5676 (e.g.TIF2(darkgreen))andsubsequenttranscriptionalcontrol.InGRsandPRs, 13. Bledsoe,R.K.,Montana,V.G.,Stanley,T.B.,Delves,C.J.,Apolito,C.J., theLBPcanexpandtoaccommodatesteroidsthataresubstitutedatC3(blue) orC17(cid:3)(lightgreen).InMR,weidentifiedasinglesiteoutsideoftheLBPthat McKee,D.D.,Consler,T.G.,Parks,D.J.,Stewart,E.L.,Willson,T.M., cantoggleMOFagonismversusantagonism,ostensiblybyformingabridge Lambert,M.H.,Moore,J.T.,Pearce,K.H.,andXu,H.E.(2002)Crystal betweenH7(bluedot)andtheH5-(cid:2)1loop(reddot).Inwild-typeMR,His-853 structureoftheglucocorticoidreceptorligand-bindingdomainrevealsa makesaweakhydrogenbondthatcannotsupportMOFactivation(redH);the novelmodeofreceptordimerizationandcoactivatorrecognition.Cell historicalsubstitutiontoapositivelychargedarginine(greenR)strengthens 110,93–105 thisinteraction,restoringactivation. 14. Gee,A.C.,andKatzenellenbogen,J.A.(2001)Probingconformational changesintheestrogenreceptor:evidenceforapartiallyunfoldedinter- resurrectionofanentireproteinisnotfeasible,theinsertionof mediate facilitating ligand binding and release. Mol. Endocrinol. 15, ancestralresiduesinmodernproteinscanincreasestabilityand 421–428 enhanceadaptabilityandtolerancetomutations(30).Inaddi- 15. Godoy-Ruiz,R.,Ariza,F.,Rodriguez-Larrea,D.,Perez-Jimenez,R.,Ibarra- tion, we have found that ancestral proteins tend to be more Molero,B.,andSanchez-Ruiz,J.M.(2006)Naturalselectionforkinetic promiscuoustosyntheticligandsordrugactivation,especially stabilityisalikelyoriginofcorrelationsbetweenmutationaleffectson incasesinwhichtheancestralproteinsdisplayamorepromis- proteinenergeticsandfrequenciesofaminoacidoccurrencesinsequence alignments.J.Mol.Biol.362,966–978 cuous phenotype than the extant proteins for endogenous 16. DePristo,M.A.,Weinreich,D.M.,andHartl,D.L.(2005)Missenseme- ligands.Thus,resurrectedproteinsmaypermitthecrystalliza- anderingsinsequencespace:abiophysicalviewofproteinevolution.Nat. tionandfunctionalanalysisofpreviouslyintractablecomplexes Rev.Genet6,678–687 duetotheirenhancedstabilityandpromiscuity. 17. Gaucher,E.A.,Thomson,J.M.,Burgan,M.F.,andBenner,S.A.(2003) We have shown that, by mirroring what has been done in Inferringthepaleoenvironmentofancientbacteriaonthebasisofresur- evolutionarystudiesaimedatdiscoveringthestructuralmech- rectedproteins.Nature425,285–288 18. Jermann,T.M.,Opitz,J.G.,Stackhouse,J.,andBenner,S.A.(1995)Re- anism that conferred hormone selectivity, ancestral proteins constructingtheevolutionaryhistoryoftheartiodactylribonucleasesu- maybeusedtoexaminecross-pharmacologyamonghomolo- perfamily.Nature374,57–59 gous proteins. The advantages of using ancestral proteins to 19. Wouters,M.A.,Liu,K.,Riek,P.,andHusain,A.(2003)Adespecialization study the structural mechanisms of drug promiscuity lie not stepunderlyingevolutionofafamilyofserineproteases.Mol.Cell12, onlyintheirenhancedstabilitybutalsoinlocatingthestruc- 343–354 20. Carroll,S.M.,Ortlund,E.A.,andThornton,J.W.(2011)Mechanismsfor turalfeaturesthatcontributetodifferencesinligandrecogni- theevolutionofaderivedfunctionintheancestralglucocorticoidrecep- tion.Ancestralgenereconstructionthereforeprovidesanele- tor.PLoSGenet.7,e1002117 gantsolutiontosomeofthetroublingproblemsthatcurrently 21. Carroll,S.M.,Bridgham,J.T.,andThornton,J.W.(2008)Evolutionof interferewiththeprocessofdrugdesign. hormonesignalinginelasmobranchsbyexploitationofpromiscuousre- ceptors.Mol.Biol.Evol.25,2643–2652 REFERENCES 22. Ortlund,E.A.,Bridgham,J.T.,Redinbo,M.R.,andThornton,J.W.(2007) 1. Nagy,L.,andSchwabe,J.W.(2004)Mechanismofthenuclearreceptor Crystalstructureofanancientprotein:evolutionbyconformationalepis- molecularswitch.TrendsBiochem.Sci.29,317–324 tasis.Science317,1544–1548 2. Kallenberger,B.C.,Love,J.D.,Chatterjee,V.K.,andSchwabe,J.W.(2003) 23. Bridgham,J.T.,Carroll,S.M.,andThornton,J.W.(2006)Evolutionof Adynamicmechanismofnuclearreceptoractivationanditsperturbation hormone-receptor complexity by molecular exploitation. Science 312, inahumandisease.Nat.Struct.Biol.10,136–140 97–101 16274 JOURNALOFBIOLOGICALCHEMISTRY VOLUME287•NUMBER20•MAY11,2012 DecipheringGlucocorticoidCross-pharmacologyUsingAncSRs 24. Tan,R.A.,andCorren,J.(2008)Mometasonefuroateinthemanagement A.,andGreene,G.L.(2007)Structuralplasticityintheestrogenreceptor ofasthma:areview.Ther.Clin.RiskManag.4,1201–1208 ligand-bindingdomain.EMBORep.8,563–568 25. Bledsoe,R.K.,Madauss,K.P.,Holt,J.A.,Apolito,C.J.,Lambert,M.H., 28. Biggadike, K., Bledsoe, R. K., Hassell, A. M., Kirk, B. E., McLay, I. M., Pearce,K.H.,Stanley,T.B.,Stewart,E.L.,Trump,R.P.,Willson,T.M., Shewchuk,L.M.,andStewart,E.L.(2008)X-raycrystalstructureofthe andWilliams,S.P.(2005)Aligand-mediatedhydrogenbondnetwork novelenhancedaffinityglucocorticoidagonistfluticasonefuroateinthe required for the activation of the mineralocorticoid receptor. J. Biol. glucocorticoid receptor ligand-binding domain. J. Med. Chem. 51, Chem.280,31283–31293 3349–3352 26. Madauss, K. P., Deng, S. J., Austin, R. J., Lambert, M. H., McLay, I., 29. Seitz,T.,Thoma,R.,Schoch,G.A.,Stihle,M.,Benz,J.,D’Arcy,B.,Wiget, Pritchard,J.,Short,S.A.,Stewart,E.L.,Uings,I.J.,andWilliams,S.P. A.,Ruf,A.,Hennig,M.,andSterner,R.(2010)Enhancingthestabilityand (2004) Progesterone receptor ligand-binding pocket flexibility: crystal solubilityoftheglucocorticoidreceptorligand-bindingdomainbyhigh structures of the norethindrone and mometasone furoate complexes. throughputlibraryscreening.J.Mol.Biol.403,562–577 J.Med.Chem.47,3381–3387 30. Yamashiro,K.,Yokobori,S.,Koikeda,S.,andYamagishi,A.(2010)Im- 27. Nettles,K.W.,Bruning,J.B.,Gil,G.,O’Neill,E.E.,Nowak,J.,Guo,Y., provementofBacilluscirculans(cid:2)-amylaseactivityattainedusingthean- Hughs,A.,Kim,Y.,DeSombre,E.R.,Dilis,R.,Hanson,R.N.,Joachimiak, cestralmutationmethod.ProteinEng.Des.Sel.23,519–528 MAY11,2012•VOLUME287•NUMBER20 JOURNALOFBIOLOGICALCHEMISTRY 16275
Description: