- Email: [email protected]
A comparison of 11β-chloromethlylestradiol and tamoxifen aziridine as affinity labeling reagents for estrogen receptors
A COMPARISON OF llfl-CHLOROMETHLYLESTRADIOL AND TAMOXIFEN AZIRIDINE AS AFFINITY LABELING REAGENTS FOR ESTROGEN RECEPTORS Thomas Ratajczak, Vincent Atrache, and Roland HBhnel
Peter Antes, Mark Comber,
The Department of Obstetrics and Gynecology, University of Western Australia, King Edward Memorial Hospital for Women, Subiaco, Western Australia 6008, Australia Corresponding author: Roland Received Revised
December August
H~hnel
31, 1986
17. 1987
The tritium-labeled from of lll3-chloromethylestradiol was preparedby metalhydridereductionof the 17-ketoderivative. Affinitylabelingexperiments were carriedout using [)Hlllj+chloromethylestradiol and [aHltamoxifen aziridine with estrogenreceptorfrom crude,calf uterinecytosoland partiallypurified(heparin-sepharose chromatography) preparations.aothconpoundsformedhighlystablereceptor complexes.Estrogenspecific,covalentbinding,however,was indicatedonly for [*Hltanoxifen aziridine. An equilibrium dissociation constantof 2.8 x lo-loM was determinedfor the receptor-ISHI 11l3-chloron&hylestradiol of hormonedissociation kineticsat interaction.Measurement 30°C revealeda slow,singlephasedissociation of 118~chloromathylestradiol from the receptor(dissociation rate constant, 1.3 x lo-' min-1). This contrastedwith the normalbiphasic dissociation patternof estradiolin whichthe dissociation rate constantfor the slowercomponentwas 16.7 x lo-$min-l. The resultsindicatethat ll~-chloromethylestradiol readilyconverts the estrogenreceptorto a high affinitybindingform and suggestthat the radiolabeled hormonemay proveusefulfor studiesof estrogenaction.
STEROIDS 51/5-6
May-J une 1988 (499-518)
499
500
Ratajczak etal
The detailed#ysico-chemical and biochemical
characterization of the estrogenreceptoris a majorobjective covalent of this laboratory(l-3). Bindingsite-directed Labelingof the receptorwith alkylating affinityreagentsis one approachto the formationof a permanentlyidentifiable proteinsuitablefor structural analysisunderstrongly denaturingand di~ggre~t~ng conditions(41. TamoxifenaziridinePI!&), an analogueof the nonsteroidal antiestro+ntaxifen, is ca@rbleof efficientand selective irreversible reactionwith estrogenreceptors(5). As a receptorlabelTA has provento be a usefulprobe for studies into receptorstructureand function(5-71, Although antiestrogens bind to the estrogenreceptorfS-101the pssibilityof allostericratherthan directinteraction of thesebeds
at the bindingsite (8)sake the use of an
estrogen-based affinitylabelmore desirablefor studiesinto receptorbindingsite structure.The compoundllg-chloromethylestradiol (Org4333)is a potentestrogen(11). Using non-radiolabeled Org 4333,a numberof laboratories have rewrted on the strongand app%ently irreversible bindingof the ligandto estrogenreceptor(4, 11-13). fn this paperwe describethe preprationof Org 4333 in tritium-&&&A form and
STEROIDS
51/5-6
May-June1988
AFFINITY
LABELS
FOR ESTROGEN
RECEPTORS
compareE3H10rg4333,and 13HlTAas potentialaffinity labelingreagentsfor the calf uterineestrogenreceptor.
CZ,4,6~7-3~l~tradiol-l7~ (93Cilxsol),Sodium boroIPHlhydride (64Cilmrrol). [z4C17-globulin (0.016 mCi/mg),and I1*Clovalbumin (0.01mCi/mg)were purchasedfrom New EnglandNuclear,13Hltamoxifen aziridine(23.4Ci/mrol) fromAmersham,estradiol,diethylstilbestrol from Sigma,Org 4333 was donatedby Dr. F. J. Zeelenof Organon. Standard proteins,ovalbumin,and bovineserumalbumin were obtained from Serva,myosin,B-galactosidase, phoqhorylaseb, and transfer& ware purchasedfrom Sigma. charcoalWrit A) was obtainedfrom Pfanstiehl and dextranT40 from Pharmacia. Buffer% PMD buffer: 10 ~Mpotassiumdihydrogen orthophosphate pH 7.4 at 22V containing20 mN sodiumnrQbdate and 1 no dithiothreitol. Tris buffer:10 rrMtris(h~o~~yl)~methane-HCl buffer,pD 7.4.Thebufferswere storedat 4% J17a-'HIllI3-Chloror&hvl~adiol(I3Ef.lOra 43332 A solutionof 1Wchloromtthylestradiol (5 mg) in acetone (1 ml) was cooledin an ice bath and treatedwith Jones reagent (14),dropwise,untilthe colorpersisted.After 5 min stirring atO°Cmethanolwas addedandthesolventmixturewas evaporated undera nitrogenstream. The residuewas partition&between dichloromethane and waterand the organiclayerwas washedwith saturatedNail.and dried INa,sO,).P@oval of the dichloromethane undernitrogengave llj3-chloromethylestrone as a whitepowder. The ketone (2 mg) in dry tetrahydrofuran (driedover NaH) (1 ml) was addedto a glassvial containinga smallstirrerbar and 500 mci sodiumboro13Hlhydride. The glassvial was sealedand the reactionmixturestirredat room temperature for 6 h. Waterwas then addeddropise and stirringcontinuedfor 1 h to ensuredecomposition of excessreducingagent. The mixturewas pouredinto a separatingfunneland extracted3 timeswith dichloromethane. The conbinedorganicextractwas washedonce with waterand once with saturatedNaCl,driedand evaporatedundera nitrogenstream. Analysisof the productby thin layer~~o~t~ra~y (silicagel 60, 3:l ~~orofo~:e~yl acetate)showada singlepeak of radioactivity, which co-chromatographed with authenticllg-chloromethylestradiol.
STEROIDS
51/5-6
May-June 1988
501
502
Ratajczak etal
The radiolabeled nmterialwas purifiedby HPLC Warian MicropakCR-lo,4 x 300 mn, ethylacetate) and storedin ethanol HPI.C at -2OY. IJHlOrg4333 was shownby quantitative [conditions as above1to have a specificactivityof 7.7 Cilnn-ol. CvtosolPreparation Calf uterinetissuewas frozenin liguidnitrogenand pulverizedto a powderusinga Braunmicr&Iismembr&or.The tissuepowderwas quicklytransfered to a cold centrifugetubs and mixed (vortex) with 4 volumesof ice-coldFMD or Tris buffer. The cytosolwas isolatedafterhigh-speed centrifugation (96,000g) at 4OC for 30 min in a BecknmnLS-70 centrifuge.&here indicatedreceptorextractswere partially purifiedfrom cytosolby heparin-sepharose chr~tog~a~y (1). Scatchardm Dilutedcytosol(1 in 5 with Tris buffer)(0.2ml) was subjectedto saturation analysisby overnightincubation at 4OC with increasing amountsof ZSHlestradiol or 13H10rg4333 (0.05,0.1, 0.2, 0.25,0.5 and 1-O pool). Dextran-coated charcoalsuspension(0.5%wiv charcoal,0.05%w/v dextranin Tris buffer)(0.5mlf was addedand the contentswere mixed briefly(vortex)and kept at 4OC for 20 min. The charcoalwas pelletedat 2200g for 10 min and 0.5 ml of the supernatant was fluid G3eckmanEp) transfered to countingvials. Scintillation (3 ml) was addedand the radioactivity was measured. Nonspecific bindingof the tritiatedhormonesat each ligand dose was determined after incubation in the presenceof a loo-foldexcessof diethylstilbestrol. The equilibrium dissociation constantsr for estrogenreceptorcomplexes with estradioland Org KS 4 43 were estimatedby scatchardplot. Assamof t3~tradiol and C3H10ra4333 Dissociation from &trocrenRecentor Dissociation of estrogensfrom receptorwas carriedout usinga nrzdification of the methoddescribedby Wei&nan and Notides(15). Calf uterinecytosol.preparedin Tris buffer, was incubated with 30 nM E3Hlestradiol or [3H10rg4333 overnightat 4OC. Estradiol(5 riM>was addedto the 13Hlestrogen/cytosol mixtureswhichwere then incubatedat 30°C for 60 min. At the noted intermediate timesduplicate 0.1 ml alig,uots were transferedto ice-watercooledglasstubes containing0.1 ml hydroxyapatite in 2 ml of phosphatebuffer. The gel was resuspended periodically over 60 min and on completion of samplingincubation was continuedfor an additional30 min at 4*C to ensurecoqlete absorptionof receptor-estrogen complexes.Free steroidwas removedby
STEROIDS
51/5-6
May-j une 1988
AFFINITY
LABELS FOR ESTROGEN
RECEPTORS
repsated washing wl te buf&?r(4 x 2 ml adapts) in a procedure invorving ion in washbuffer, ~~t~ifu~t~o~ LI500g fur 2 tin>anddecanting of ths ~~~~t~t* The pellets werethentransferad to counting vialswith scintillation fluidGSe&mm ET?) fortrItium estimation. The inactivation of the ~~~o~-r3~lestrog~complexes in the presence of 30 #I of the corresponding E3Hlsteroid was monitored withouttheaddition of unlabeled estradiol in a parallel incubation at 3OT. In a similar way the nonspecific binding of the [3~~~~r~~~ was nettsured by intieion with a X00-fuld excessof ~~~~ estradiol.Dissociation rate constants weresteak as previously described (151fromthe slopesof log (% 13Hlligand bouud)vm incubation time,
Is werepreparsd as ~r~v~o~~~described f3lwith separating gelshavtng7.5%fw/vIacry&mide.Stacking gels contained 4% acryltideand ~~sa~~~~de was usedas crosslinker in all gels. Receptor sainples werepmpared in 100 ~1 of electr rxsis sam@e buffer(O&&I Tris,ECl, O.OOlMEKPA,pH 8.0containZng 20% (v/v)glycerol, 5% Iv/v)~~~~~t~~, 1% Wvf SDS)and heatedat 65Yzfor I5 aXin beforeloading ontothegel. ~~~o~reti~ separations wereperformed according to ~~~ (161. MyosinN = 210,000(171, B-gafactosidase lI6,OOO (181, phos~ryiase b 97,400Wf, transferrin 79,500(201, bovineserum albumin67,000(211,ovalbmin43,000(221, carbonic anhydrase 29,000(231,soybean try@& ~ibit~~ 20,100 (241,and a--la&albumin 14,400CZStwereusedas mfecu~ar weightmarkersto establish a ~e~at~o~h~~ betweenrelative mobility and LogM . Gels werefix& with12% fw/vItrichloroacetic acidand stained withCoornassie blueas previou&ly described $3). &fiter destaining with5% aceticacidthegels~~~~~~~g [VPl!&-labeled proteins werefrozenon dry iceand out into 2.1m slices,Eachslicewas placedin a counting vial,dried at 8WC for 2 h, and dissoWedin 250 rtlof 3096
503
504
Ratajczak
et al
glasstest tubes (20 x 150 run).With frequentagitation (vortex) the discswere washedseguentially with two 10 ml aliquots of the followingsolvents:boiling95% ethanolfor 15 min, 1:l ether,95% ethanolat room tem@rature,and ether. The discswere driedand countedfor tritiumcontent. G&yg;atb 01 it Glycerolgradientswere preparedand ultracentrifugal analysisperformedas describedby Atracheet (3). Badioa&ivitvMeasurements Radioactivity was measuredin a Ls 5800Beckmanliquid scintillation counter. Samplesof up to 0.5 ml were countedin BeckmanEP scintillation fluid (3 ml). C&enchcorrection was automaticafter externalstandardization usingthe H-number method.
We have previouslyreportedon the stronginteraction of Org 4333 with the uterineestrogenreceptorsof humanand calf (4). This compound.a potentialaffinitylabelfor the receptor,was preparedin tritium-labeled form by metalhydridereductionof the 17-ketoanalogue.A reactiveelectrophilic derivativeof tamoxifen(7), ['HITA,has been shownto couplecovalentlyto estrogenreceptorsfrom rat uterus (51and humanIKF-~cells (6). We have examinedthe utilityof both fSHIOrg4333 and FHITA in studiesdirectedat the locationand structural chemistryof the estradiolbindingsite. Characteristics of estroaenrecentorinteraction with LsHI llB-chloromethvlestradiol Incubation of cytosoliccalf uterineestrogenreceptorwith ['HlOrg4333 and [aHlestradiol. undersaturating
STEROIDS
51/5-6
May-June
1988
AFFINITY
LABELS
FOR ESTROGEN
RECEPTORS
.
,
--E2
6.66
3.5s
3
I
L
10
20 FRACTION
30
40
I 50
NUMBER
Figure1, ~ltrac~tr~fu~l analysisof cytosolestradiol receptorboundto I*Hlestradiol and I)HlOrg4333. Estradiolreceptorin FMD bufferedcalf uterinecytosolWas labeledwith 30 nM EJHlestradiol B.A. 93 Ci/mnol)(01or L3H10rg4333 @.A. 7.7 &i/ml) (0)for 4 h at 4T. unbound steroidsware renmed with dextran-coated charcoaland 0.2 I& aliquotsof the labeledextractsxere layeredonto 12-3296 v/v glycerolgradientspreparedin FWD buffer. The san'@eswere sedimented for 17 h at 237,OOO%g. ~14Cly~lo~lin,6.6s (32) and Cz~Clovalbumin, 3.55 (33)were used as internalmarkers.
STEROIDS
51/5-6
May-J une 1988
505
506
Ratajczak etal
conditions, producedan equivalentnumberof labeledbinding in Figure1, which sitesfor each hormone. This is demonstrated showsthe ultracentrifugal patternsfor receptorlabeledwith eithersteroidto be almostsuperimposable if specificactivity differences are taken intoaccount. In both casesa singlepeak in the 9s regionwas observed. The receptor-[3H10rg 4333 complexwas remarkablystable to treatmentwith dextran-coated charcoalat elevated temperature.Nearly75% of the conplexformedby 13H10rg 4333-labeling of a partiallypurified(heparin-se&arose chromatography) receptorextractwas stillmeasurableafter 30 min at 45*C (Figure2). In contrast,[3Hlestradiol-labeled receptorfrom the same extractwas completely dissociated under the same conditions(Figure2). The receptor-[3H10rg 4333 complex,however,failedto survivemore rigoroustestsof covalentbindingand was foundto be sensitiveto hot solvents used in a filterdisc-solvent extractionassay. The complexwas also
unstable
to denaturing
gel
electrophoresis. Increased
temperature, exclusionof dithiothreitol and n&ybdate from receptorpreparations, and emsure of cytosolicreceptorto r3H10rg4333 in bufferswith varyingpH (pH 7.4 - 8.5)failed to inducecovalentattachrtmnt. An indication of the affinityof Org 4333 for the receptor was obtainedby Scatchardplot analysisof data derivedfrom
STEROIDS
51/S-6
May-j une 1988
AFFINITY
10
LABELS
FOR ESTROGEN
RECEPTORS
20 TIME
Imin)
Figure2. Time+ependentstabilityof receptor-13Hlestradiol and receptor-['HIORG 4333 corrplexes to dextran-coated charcoaltreatmentat 45OC. Estradiolreceptor,partially purifiedby heparin-se&arose chromatography, was labeledby overnightincubation at 4OC with 30 nM 13Hlestradiol (0)or f3H10rg4333 f(r) in the absenceor presenceof 3000nM of the corresponding non-radioactive steroid. Aliguots (400&I fromthe labeledextractswere mixedat 4OC in tubes containingcharcoalpelletsfrom 1 mL aliguotsof dextran-coated charcoalsuspension(0.5%w/v charcoal,0.05%dextran). After 20 min the sampleswere furtherincubatedat 45OCwith occasional mixing. At selectedtimesduplicatesamplesware centrifuged and 200 ti aliguotsof the sup3rnatants were withdrawnfor radioactivity estimation.Receptor-bound radioactivity was calculatedaftercorrectionfor nonspecific bindingand expressedas a percentageof the controlvalue (=lOO%)observedpriorto 4J°C incubation.
STEROIDS
51/5-6
May-j une 1988
507
508
Ratajczak etal
incubation of uterinecytosolaliquotswith increasing concentrations of the radioactive ligand. The eguilibrium dissociation constant,I$.,, for Org 4333 was 2.8 + 1.8 x 1O-1oM (n = 2). Usingthe same extractsthe I$, valuedeterminedfor estradiolwas 3.1 + 0.4 x 1O-1oM (n = 2). The natureof receptor-Crg 4333 interaction was further characterized by comparingthe dissociation propertiesof 13Hlestradiol and 13H10rg4333 from the receptoras measuredby displacement of the labeledestrogenswith unlabeled we observedthe well estradiolat 30cC. For f3Hlestradiol documented(15,26. 27) biphasicor tm wmlpnent. firstorder dissociation curve (Figure3) generatedby low and high affinity bindingformsof the receptor.The rate constantof the more rapidwpnent
was k_, = 113.0+ 12.6 x 1O-3min-1
(n = 3). whilethat of the slowerdissociating wvnent
was
k-Z = 16.7 + 7.5 x 1O-3min-l (n = 3). Underthe same conditionsf3H10rg4333 dissociated only slowlyfrom the receptorand followadmonophasickineticswith a rate wnstant k = 1.3 + 0.2 x lo-'min-l (n = 3) (Figure31. Characteristics of estrouenreceatorinteraction with 13Hltamoxifen aziridin!& Receptorfrom crudecytosolwas incubatedwith ['HITA underconditions previouslyfoundto be optimalfor the covalent labelingof rat uterineestrogenreceptor(5). Specific,
STEROIDS
51/5-6
May-1 une 1988
AFFINITY
LABELS
30 TIME
FOR ESTROGEN
RECEPTORS
60 hn,
Figure3. Dissociation of LfHlestradiol and I)HlOrg4333 from the estrogenreceptor. Calf uterinecytosolin Tris buffer was equilibrated overnightat 4OC with 30 nM [aHlestradiol (0)or IaHIOrg4333 (0). After the additionof 5 PM unlabeledestradiolthe extractswere incubatedat 30°Cand aliquots(0.1 ml) were assayed for dissociation.Inactivation of the receptor-[:3Hlestrogen con@exes was monitoredunder the same conditions withoutadditionof unlabeledestradiol. All dissociation rreasurements were correctedfor nonspecific steroidbindingand receptorinactivation.The dissociation ratesware: Estradiol,k_, = 128 x lo-'min-I,k_, = 24 x lo-)min-I,Org 4333,k = 0.9 x lo-'mini.
STEROIDS
51/5-6
May-June 1988
509
510
Ratajczak etal
Figure4. Ultracentrifugal analysisof heparin-sepharose chromatography purifiede&radio1 receptorboundto [3Hlestradiol and 13Hltamoxifen aziridine.Estradiol receptorwas labeledfor 3 h at 4OC with 30 nM [3Hlestradiol (A)or 15 nM raHITAin the absenceor presenceof 3000nM unlabeledestradiol.All samplescontained7% UW. After removalof unboundligands(dextran-coated charcoal)the samples were subjectedto ultracentrifugal analysison 12-32%(V/V) glycerolgradientsas describedin the Figure1 legend. Fractionsfrom IfHlTA-labeled gradientswere dividedand measuredfor radioactivity (B)and covalentlabeling(C) using the ethanoldisc-extraction assaydescribedin Materialsand Methods. Catalase,11.3s (34)was used as an external sedimentation standard. Internalmarkerswere IX4C17-globulin, 6.65 (32)and [14Clovalbumin, 3.55 (33). covalentlylabeledreceptor,as determinedby the disc-extraction assay,represented 25% of the [fHlestradiol bindingsitesin the extract. Ultracentrifugal analysisof crudecytosolicreceptorlabeledwith [3~1~~revealeda major of radioactivity at 4.7s.whichwas only weaklyinhibited with unlabeledestradiol(notshown). Tanoxifenaziridinebinds -Y
proteins,includingalbumin,nonspecifically (5).and
STEROIDS
51/5-6
May-June 1988
AFFINITY
further
LABELS
FOR ESTROGEN
RECEPTORS
preliminary studieswith this ligandwere carriedout on
receptorextractspartiallypurifiedby heparin-sepharose chromatography.[)HITAlabelingwas more selectivefor receptorin theseextractsand the percentageof covalently coupledreceptorwas markedlyincreased(50%). In glycerolgradientspartiallypurifiedestrogenreceptor labeledwith ['Hlestradiol sedimented at 10s (Figure4A). The sedimentation profileof the [fHITA-labeled extract differedconsiderably and indicatedspecifically bound radioactivity at 9.6S,6.6S,and 3.5s (Figure4B). A high level of nonspecific bindingat 3.5swas also noted (Figure4B). When fractionscontaining bindingactivitywere assessedfor covalent attachment by the disc-extraction assay ['HITAwas foundto be boundcovalentlyto proteinsin all threesedimentation regions(Figure40. SDS-PAGEanalysisof a heparin-sepharose treatedreceptor extract,labeledwith [J~I~~,revealedtwo peaksof covalently bound (estradiolcompetitive)
radioactivity
that
correspondedto labeledcomponentswith M, 67,000and 50,000 L
(notshown).
DISCUSSION In this
reportwe have describedthe preparation of
formand llg-chloromethylestradiol (Org4333)in tritium-labeled
STEROIDS
51/5-6
May-June 1988
511
512
Ratajczak et al
have studied the properties of this compoundas an affinity labeling reagent for the estrogen receptor. confirrm?dearlier reports
14,
Our studies have
11-131, based on experiments with
unlabeled Org 4333, that the steroid interacts strongly with the receptor.
Our results indicate, however, that the
receptor-E3H10rg 4333 complex can be disrupted under denaturing conditions. in Org 4333 iqarts
Thus, although the ~loro~thyl~
considerable stability
group
to the interaction
with receptor, the chlorine atom appears not to undergo nucleophilic substitution at the binding site (13). of this interaction warrants further study. greatly facilitated
by the availability
The nature
This should be
of tritium-labeled Org
4333
l
The man ~i~~~i~
dissociation
constant (2.8 x IO-lo MI,
derived for the interaction of receptor with 13H30rg4333 comparedfavourably with values of 1.2 x 1O-*o M, and 2.2 x 1O-1o M previously obtained by competitive Scatchard analysis (4) with the unlabeled ligand in calf and humanuterine cytosols,
respectively.
!Fheslow dissociation
rate observed for tfHl0rg 4333 from
receptor at 30%~was consistent with the low levels of exchange previously achieved, under similar conditions, with E3Hlestradiol for receptor binding sites presaturated with radioinert Org 4333 (4).
Weichmanand E&tides (28) have
STEROIDS51/S-6
May-June1988
AFFINITYLABELS FOR ESTROGEN RECEPTORS
proposedthat the dissociation kineticsof an estrogenfrom receptormay providean indexfor biological activityby reflecting the capacityof the hormoneto activateand maintain the receptorin a high affinitybindingstate. As previously described(15,26-28)the fast and slowlydissociating formsof receptorcomplexeswith e&radio1 may representnon-activeand activatedstates,respectively.With Org 4333only the slow dissociating form was evident,suggesting that the steroidmay readilyactivateall the receptorsto a form capableof high affinitybinding. In an earlierreport(29)receptor-estradiol complexes transformed at 37% to a nondissociable, thermallystableform wereproposedas the functional units requiredfor genomic activation.Pihether the receptor*rg4333 corrplex is equivalent to this stabilizedform of the receptormay be answeredby comparative studiesof the influencethat estradioland Org 4333 have on the interaction affinityof receptorwith non-specific DNh sequencesas well as estrogenrespnsive DJ& elements involvedin the controlof estrogenregulatedgenes (30,31). Our sedimentation experiments with the rolybdate-stabilized estrogenreceptorof calf uterushave consistently shown I*Hlestradiol-labeled receptorto sedimentas a single, well-defined peak at 9-10s (3). ~~u~as~~~
b~d~g~~
(9.3s)was observedfor receptorcovalentlylabeledwith
STEROIDS 5-l/5-6May-June
1988
513
514
Ratajczak &al
[fHlTA,the presenceof snmllercovalentlylabeledcomponents at 6.6sand 3.5s resultedin a markedlydifferentsedimentation profile. In contrast.Katzenellenbogen et al (51have demonstrated similarultracentrifugal patternsfor cytosolicrat uterineestrogenreceptorlabeledwith 13Hlestradiol. and 13HITA. Using low salt buffersthe same groupobtained sedimentation coefficients of 8s for both receptorcomplexesand showedthat receptorboundcovalentlyto EfHITAgave a similargradientprofile(5). The labeledcomponents(M, = 67,000,50,000)observedafter exposureof partiallypurified,calf uterinereceptorextracts to f3HlTAcorrespndedcloselyin molecularsize to those reprted previouslyby Katzenellenbogen et al (5) for I'HITA-labeled estrogenreceptorfrom rat uterus (M, = 60,000- 66,000,50,000). Unitsof similarsize
(Mr = 63,000)have also been reportedfor 13~l~~-labeled cytosoland nuclearestrogenreceptorsfrom humanMCF-7 cells (6). The possibility existsthat the smaller50K unit,observed in our studyto be covalentlylabeledby ['HITA,is a proteolytic fragmentof the 67K component.Monsma&t
(61
have describeda labeledreceptorform of similarsize (Mr =
53,000)afterproteolysis, by endogeneous enzymesof the
ML.= 63,000 t3H1TA-labeled estrogenreceptorin cytosolsof
STEROIDS
51/5-6
May-June 1988
AFFINITY
LABELS
FOR ESTROGEN
RECEPTORS
Mm-7 cells. The
resultsof this studyindicatethat to date tamoxifen
aziridineremainsthe best availableaffinitylabelingreagent for estrogenreceptorsand studiesusingthis ligandshould continueto providevaluablestructuralinformation on the protein. AlthoughOrg 4333 is not an affinitylabelfor the estrogenreceptorthe tritium-labeled compoundmay prove useful for the studyof receptorundernondenaturing conditionsand for studiesof estrogenaction.
This work was supportedbytheNationa1 Health andMedical ResearchCouncilof Australia.
REFERENCES
of methods 1. RatajczakT, and H&e1 R (1980). Investigation of estradiolreceptorpurification priorto affinity chromatography. J. SIERGIDBIOCI-NM. 13:439-444. 2. GschwendtM, H&nel R, and RatajczakT (1983). Purification to homogeneity of the nuclearestrogenreceptorfrom chick liver. BIGCHIMBIOPHYSAC!l'A 760:238-245. 3. AtracheV, RatajczakT, SenafiS, and H&e1 R (19851. Purification of the molybdate-stabilized 9-10sestradiol receptorfrom calf uterus. J BIOL CHEM 260:5936-5941. 4. RatajczakT, SheppardPN, caponRJ, and H&nel R (1981). The synthesisand studyof some potentialaffinitylabeling reagentsfor estrogenreceptors.STEROIDS38~537-555. 5. Katzenellenbogen JA, CarlsonKE, HeimanDF, RobertsonDW, Wei IL, and Katzenellenbogen BS (1983). Efficientand highlyselectivecovalentlabelingof the estrogenreceptor with IsHltamoxifen aziridine.J BIOL CH@l 258:3487-3495.
STEROIDS
51/5-6
May-June 1988
515
516
Ratajczak
et al
MonsmaFJ, Jr, Katzenellenlmgen BS. MillerMA, ZieglerYS. and Katzenellenbogen JA (1984). Characterization of the estrogenreceptorand its dynamicsin MCF-7hwnanbreast cancercellsusinga covalentlyattachingantiestrogen. ENmCRINDLIx;y 115:143-153. BS 1. MillerMA, MullickA, GreeneGL, and Katzenellenbogen (1985).Characterization of the subunitnatureof nuclear estrogenreceptorsby chemicalcross-linking and denseamino 117:515-522. acid labeling. ENDCCRINX.GGY T (1973). The influence 8. H&nel R, TwaddleE, and Ratajczak of syntheticanti-estrogens on the bindingof tritiated estradiol-178 by cytosolsof humanuterusand humanbreast BIGCHEM4:687-695. carcinoma.J SI'ERGID H (1982). Tamxifen and 9. Coezy E, BorgnaJL, and Rochefort nratabolites in MCH cells:correlation betweenbindingto estrogenreceptorand inhibition of cell growth. CANCERRES 42:317-323. 10. EckertRL, and Katzenellenbogen BS (1982). Physical properties of estrogenreceptorcomplexesin MCF-7human breastcancercells. J BIOL CHEM 237:8840-8846. 11. Van den BroekAJ, LeemhuisJ, De WinterMS, and ZeelenF (1983).Org 4333,a potent,irreversibly bindingestrogen agonist. PHARMWIEKBL LSCII 5:182-183. 12. LeclercqC, Devleeschouwer N, and HeusonJC (1983). Guide-lines in the designof new antiestrogens and cytotoxic-linked estrogensfor the treatmentof breast cancer. J STEROIDBIOCHEM19:75-85. 13. ReinerGCA, Katzenellentmgen BS, BindalRD, and Katzenellenbogen JA (1984). Biological activityand receptorbindingof a stronglyinteracting estrogenin human breastcancercells. CANCERRES 44:2302-2308. 14. BowersA, Halsalll'G,JonesERH, and LeminAJ (1953). The chemistryof the triterpenes and relatedcompounds.Part of the structureof polyporenic acid C. XVIII. Elucidation J CHEN Sot 2548-2560. 15. WeichmanBM, and NotidesAC (1977). Estradiol-binding kineticsof the activatedand nonactivated estrogen receptor. J BIOL CHFX 252:8856-8862. 16. LaermliUK (1970). Cleavageof structural proteinsduring the asser&lyof the head of bacteriophage T4. NATDRE227: 6.
680-685. 17.
Maize1JV, Jr. Polyacrylamide gel electrophoresis of viral proteins. In: Methodsin Viroloav@laranmuschK and KaprowskiH, eds),AcademicPress,New York (1971). V:pp 179-246. 18. FowlerAV, and Z&in I (1977). The aminoacid sequenceof 8-galactosidase of &cherichia&. PXICNATLACADSCIUSA 74:1507-1510.
STEROIDS
51/5-6
May-j une 1988
AFFINITY
LABELS
FOR ESTROGEN
RECEPTORS
19. TitaniH, KoideA, HermannJ. EricssonLH, KumarS, Wade RD. WalshKA, NeurathH. and FischerEH (1977).Completeamino acid sequenceof rabbitmuscleglycogenphosphorylase.PRGC NATL ACAD SC1 USA 74:4762-4766. 20. McGillivray WA, MendeZE, SinhaSK, SuttonMR, Lineback-Zins J, and Brew K (1982). The completeaminoacid sequenceof human serumtransferrin.PfiDcNATLACADSCIUSA 79:2504-2508. 21. SquireI?$ Moser P, and O'KonskiCT (1968). The hydrodynamicpropertiesof bovineserumalbuminmonomerand 7~4261-4272. dimer. BIOCHEMISTRY 22. NisbetAD, SaundryRH, Moir AJG, mthergill LA, and mthergill JE (1981). The completeamino-acidsequenceof hen ovalbumin.EUR J BIGCHEM115:335-345. 23. ReynaudJ, LuccioniF, BouthierM. !&varyJ, and DerrienY (1971). Etudehydrodynamigue compareedes anhydrases carbonigues erythrocytaires bovinesA et B. BIOCHIMIE53: 1095-1098. 24. KoideT, and IkenakaT (1973). Studieson soybeantrypsin inhibitors.I. Fragmentation of soybeantrypsininhibitor (KImits) by limitedproteolysis and by chemicalcleavage. EUR J BIGCHEM32:401-407. 25. Brew K, VanamanTC, and Hill RL (1967). Comparison of the aminoacid sequenceof bovinea-lactalbminand hens egg white lysozyme. J BIOL CHEM 242:3747-3749. 26. SakaiD, and GorskiJ (1984).Estrogenreceptortransformationto a high-affinity statewithoutsubunit-subunit 23:3541-3547. interactions.BIGCHEMISTRY 27. de Boer W, Ab G, and GruberM (1985). Estrogenreceptorin chickenoviduct:receptordissociation kineticsand transformation. J Sl‘Et-QID BIOCHEM23:9-18. 28. ~7eichman BM, and NotidesAC (1980). Estrogenreceptor activation and the dissociation kineticsof estradiol, 106:434-439. estrioland estrone. ENmcRrNDII)(;y 29. FishmanJH, and Fishmn J (1985). The formation at 37OCof a nondissociable receptor-estradiol complex. BIOCHEM BIOPHYSRES CXMUN 131:58-63. 30. SkafarDF, and NotidesAC (1985). Modulationof the estrogenreceptor's affinityfor DNA by estradiol.J BIOL CHEM 260:12208-12213. 31. Klein-Hitpass L, SchorppM, WagnerU, and RyffelGU (1986). An estrogen-responsive elementderivedfrom the 5' flanking regionof the Xenopusvitellogenin gene functionsin s 3-1061. transfected humancells. CELL 46:lO 32. CalvanicoNJ, and Tomsi TB, Jr, (1971). Fragmentsof human yG imnunoglobulins producedby porcinepancreatic esterase. ARCHBIOCHEMBIOPHYS144:269-280.
STEROIDS
51/5-6
May-June 1988
517
518
Ratajczak etal
33. CastellinoFJ, and BarkerR (1968). Examination of the dissociation of multichainproteinsin guanidine hydrochloride by merrbrane osnmetry. BIOCHEMISTRY 1:2201-2217. 34. SumnerJB, and GralenN (1938). !rhemolecularweightof crystalline catalase. J BIOL CHEM 125:33-36.
STEROIDS
51/5-6
May-June 1988
Peter Antes, Mark Comber,
The Department of Obstetrics and Gynecology, University of Western Australia, King Edward Memorial Hospital for Women, Subiaco, Western Australia 6008, Australia Corresponding author: Roland Received Revised
December August
H~hnel
31, 1986
17. 1987
The tritium-labeled from of lll3-chloromethylestradiol was preparedby metalhydridereductionof the 17-ketoderivative. Affinitylabelingexperiments were carriedout using [)Hlllj+chloromethylestradiol and [aHltamoxifen aziridine with estrogenreceptorfrom crude,calf uterinecytosoland partiallypurified(heparin-sepharose chromatography) preparations.aothconpoundsformedhighlystablereceptor complexes.Estrogenspecific,covalentbinding,however,was indicatedonly for [*Hltanoxifen aziridine. An equilibrium dissociation constantof 2.8 x lo-loM was determinedfor the receptor-ISHI 11l3-chloron&hylestradiol of hormonedissociation kineticsat interaction.Measurement 30°C revealeda slow,singlephasedissociation of 118~chloromathylestradiol from the receptor(dissociation rate constant, 1.3 x lo-' min-1). This contrastedwith the normalbiphasic dissociation patternof estradiolin whichthe dissociation rate constantfor the slowercomponentwas 16.7 x lo-$min-l. The resultsindicatethat ll~-chloromethylestradiol readilyconverts the estrogenreceptorto a high affinitybindingform and suggestthat the radiolabeled hormonemay proveusefulfor studiesof estrogenaction.
STEROIDS 51/5-6
May-J une 1988 (499-518)
499
500
Ratajczak etal
The detailed#ysico-chemical and biochemical
characterization of the estrogenreceptoris a majorobjective covalent of this laboratory(l-3). Bindingsite-directed Labelingof the receptorwith alkylating affinityreagentsis one approachto the formationof a permanentlyidentifiable proteinsuitablefor structural analysisunderstrongly denaturingand di~ggre~t~ng conditions(41. TamoxifenaziridinePI!&), an analogueof the nonsteroidal antiestro+ntaxifen, is ca@rbleof efficientand selective irreversible reactionwith estrogenreceptors(5). As a receptorlabelTA has provento be a usefulprobe for studies into receptorstructureand function(5-71, Although antiestrogens bind to the estrogenreceptorfS-101the pssibilityof allostericratherthan directinteraction of thesebeds
at the bindingsite (8)sake the use of an
estrogen-based affinitylabelmore desirablefor studiesinto receptorbindingsite structure.The compoundllg-chloromethylestradiol (Org4333)is a potentestrogen(11). Using non-radiolabeled Org 4333,a numberof laboratories have rewrted on the strongand app%ently irreversible bindingof the ligandto estrogenreceptor(4, 11-13). fn this paperwe describethe preprationof Org 4333 in tritium-&&&A form and
STEROIDS
51/5-6
May-June1988
AFFINITY
LABELS
FOR ESTROGEN
RECEPTORS
compareE3H10rg4333,and 13HlTAas potentialaffinity labelingreagentsfor the calf uterineestrogenreceptor.
CZ,4,6~7-3~l~tradiol-l7~ (93Cilxsol),Sodium boroIPHlhydride (64Cilmrrol). [z4C17-globulin (0.016 mCi/mg),and I1*Clovalbumin (0.01mCi/mg)were purchasedfrom New EnglandNuclear,13Hltamoxifen aziridine(23.4Ci/mrol) fromAmersham,estradiol,diethylstilbestrol from Sigma,Org 4333 was donatedby Dr. F. J. Zeelenof Organon. Standard proteins,ovalbumin,and bovineserumalbumin were obtained from Serva,myosin,B-galactosidase, phoqhorylaseb, and transfer& ware purchasedfrom Sigma. charcoalWrit A) was obtainedfrom Pfanstiehl and dextranT40 from Pharmacia. Buffer% PMD buffer: 10 ~Mpotassiumdihydrogen orthophosphate pH 7.4 at 22V containing20 mN sodiumnrQbdate and 1 no dithiothreitol. Tris buffer:10 rrMtris(h~o~~yl)~methane-HCl buffer,pD 7.4.Thebufferswere storedat 4% J17a-'HIllI3-Chloror&hvl~adiol(I3Ef.lOra 43332 A solutionof 1Wchloromtthylestradiol (5 mg) in acetone (1 ml) was cooledin an ice bath and treatedwith Jones reagent (14),dropwise,untilthe colorpersisted.After 5 min stirring atO°Cmethanolwas addedandthesolventmixturewas evaporated undera nitrogenstream. The residuewas partition&between dichloromethane and waterand the organiclayerwas washedwith saturatedNail.and dried INa,sO,).P@oval of the dichloromethane undernitrogengave llj3-chloromethylestrone as a whitepowder. The ketone (2 mg) in dry tetrahydrofuran (driedover NaH) (1 ml) was addedto a glassvial containinga smallstirrerbar and 500 mci sodiumboro13Hlhydride. The glassvial was sealedand the reactionmixturestirredat room temperature for 6 h. Waterwas then addeddropise and stirringcontinuedfor 1 h to ensuredecomposition of excessreducingagent. The mixturewas pouredinto a separatingfunneland extracted3 timeswith dichloromethane. The conbinedorganicextractwas washedonce with waterand once with saturatedNaCl,driedand evaporatedundera nitrogenstream. Analysisof the productby thin layer~~o~t~ra~y (silicagel 60, 3:l ~~orofo~:e~yl acetate)showada singlepeak of radioactivity, which co-chromatographed with authenticllg-chloromethylestradiol.
STEROIDS
51/5-6
May-June 1988
501
502
Ratajczak etal
The radiolabeled nmterialwas purifiedby HPLC Warian MicropakCR-lo,4 x 300 mn, ethylacetate) and storedin ethanol HPI.C at -2OY. IJHlOrg4333 was shownby quantitative [conditions as above1to have a specificactivityof 7.7 Cilnn-ol. CvtosolPreparation Calf uterinetissuewas frozenin liguidnitrogenand pulverizedto a powderusinga Braunmicr&Iismembr&or.The tissuepowderwas quicklytransfered to a cold centrifugetubs and mixed (vortex) with 4 volumesof ice-coldFMD or Tris buffer. The cytosolwas isolatedafterhigh-speed centrifugation (96,000g) at 4OC for 30 min in a BecknmnLS-70 centrifuge.&here indicatedreceptorextractswere partially purifiedfrom cytosolby heparin-sepharose chr~tog~a~y (1). Scatchardm Dilutedcytosol(1 in 5 with Tris buffer)(0.2ml) was subjectedto saturation analysisby overnightincubation at 4OC with increasing amountsof ZSHlestradiol or 13H10rg4333 (0.05,0.1, 0.2, 0.25,0.5 and 1-O pool). Dextran-coated charcoalsuspension(0.5%wiv charcoal,0.05%w/v dextranin Tris buffer)(0.5mlf was addedand the contentswere mixed briefly(vortex)and kept at 4OC for 20 min. The charcoalwas pelletedat 2200g for 10 min and 0.5 ml of the supernatant was fluid G3eckmanEp) transfered to countingvials. Scintillation (3 ml) was addedand the radioactivity was measured. Nonspecific bindingof the tritiatedhormonesat each ligand dose was determined after incubation in the presenceof a loo-foldexcessof diethylstilbestrol. The equilibrium dissociation constantsr for estrogenreceptorcomplexes with estradioland Org KS 4 43 were estimatedby scatchardplot. Assamof t3~tradiol and C3H10ra4333 Dissociation from &trocrenRecentor Dissociation of estrogensfrom receptorwas carriedout usinga nrzdification of the methoddescribedby Wei&nan and Notides(15). Calf uterinecytosol.preparedin Tris buffer, was incubated with 30 nM E3Hlestradiol or [3H10rg4333 overnightat 4OC. Estradiol(5 riM>was addedto the 13Hlestrogen/cytosol mixtureswhichwere then incubatedat 30°C for 60 min. At the noted intermediate timesduplicate 0.1 ml alig,uots were transferedto ice-watercooledglasstubes containing0.1 ml hydroxyapatite in 2 ml of phosphatebuffer. The gel was resuspended periodically over 60 min and on completion of samplingincubation was continuedfor an additional30 min at 4*C to ensurecoqlete absorptionof receptor-estrogen complexes.Free steroidwas removedby
STEROIDS
51/5-6
May-j une 1988
AFFINITY
LABELS FOR ESTROGEN
RECEPTORS
repsated washing wl te buf&?r(4 x 2 ml adapts) in a procedure invorving ion in washbuffer, ~~t~ifu~t~o~ LI500g fur 2 tin>anddecanting of ths ~~~~t~t* The pellets werethentransferad to counting vialswith scintillation fluidGSe&mm ET?) fortrItium estimation. The inactivation of the ~~~o~-r3~lestrog~complexes in the presence of 30 #I of the corresponding E3Hlsteroid was monitored withouttheaddition of unlabeled estradiol in a parallel incubation at 3OT. In a similar way the nonspecific binding of the [3~~~~r~~~ was nettsured by intieion with a X00-fuld excessof ~~~~ estradiol.Dissociation rate constants weresteak as previously described (151fromthe slopesof log (% 13Hlligand bouud)vm incubation time,
Is werepreparsd as ~r~v~o~~~described f3lwith separating gelshavtng7.5%fw/vIacry&mide.Stacking gels contained 4% acryltideand ~~sa~~~~de was usedas crosslinker in all gels. Receptor sainples werepmpared in 100 ~1 of electr rxsis sam@e buffer(O&&I Tris,ECl, O.OOlMEKPA,pH 8.0containZng 20% (v/v)glycerol, 5% Iv/v)~~~~~t~~, 1% Wvf SDS)and heatedat 65Yzfor I5 aXin beforeloading ontothegel. ~~~o~reti~ separations wereperformed according to ~~~ (161. MyosinN = 210,000(171, B-gafactosidase lI6,OOO (181, phos~ryiase b 97,400Wf, transferrin 79,500(201, bovineserum albumin67,000(211,ovalbmin43,000(221, carbonic anhydrase 29,000(231,soybean try@& ~ibit~~ 20,100 (241,and a--la&albumin 14,400CZStwereusedas mfecu~ar weightmarkersto establish a ~e~at~o~h~~ betweenrelative mobility and LogM . Gels werefix& with12% fw/vItrichloroacetic acidand stained withCoornassie blueas previou&ly described $3). &fiter destaining with5% aceticacidthegels~~~~~~~g [VPl!&-labeled proteins werefrozenon dry iceand out into 2.1m slices,Eachslicewas placedin a counting vial,dried at 8WC for 2 h, and dissoWedin 250 rtlof 3096
503
504
Ratajczak
et al
glasstest tubes (20 x 150 run).With frequentagitation (vortex) the discswere washedseguentially with two 10 ml aliquots of the followingsolvents:boiling95% ethanolfor 15 min, 1:l ether,95% ethanolat room tem@rature,and ether. The discswere driedand countedfor tritiumcontent. G&yg;atb 01 it Glycerolgradientswere preparedand ultracentrifugal analysisperformedas describedby Atracheet (3). Badioa&ivitvMeasurements Radioactivity was measuredin a Ls 5800Beckmanliquid scintillation counter. Samplesof up to 0.5 ml were countedin BeckmanEP scintillation fluid (3 ml). C&enchcorrection was automaticafter externalstandardization usingthe H-number method.
We have previouslyreportedon the stronginteraction of Org 4333 with the uterineestrogenreceptorsof humanand calf (4). This compound.a potentialaffinitylabelfor the receptor,was preparedin tritium-labeled form by metalhydridereductionof the 17-ketoanalogue.A reactiveelectrophilic derivativeof tamoxifen(7), ['HITA,has been shownto couplecovalentlyto estrogenreceptorsfrom rat uterus (51and humanIKF-~cells (6). We have examinedthe utilityof both fSHIOrg4333 and FHITA in studiesdirectedat the locationand structural chemistryof the estradiolbindingsite. Characteristics of estroaenrecentorinteraction with LsHI llB-chloromethvlestradiol Incubation of cytosoliccalf uterineestrogenreceptorwith ['HlOrg4333 and [aHlestradiol. undersaturating
STEROIDS
51/5-6
May-June
1988
AFFINITY
LABELS
FOR ESTROGEN
RECEPTORS
.
,
--E2
6.66
3.5s
3
I
L
10
20 FRACTION
30
40
I 50
NUMBER
Figure1, ~ltrac~tr~fu~l analysisof cytosolestradiol receptorboundto I*Hlestradiol and I)HlOrg4333. Estradiolreceptorin FMD bufferedcalf uterinecytosolWas labeledwith 30 nM EJHlestradiol B.A. 93 Ci/mnol)(01or L3H10rg4333 @.A. 7.7 &i/ml) (0)for 4 h at 4T. unbound steroidsware renmed with dextran-coated charcoaland 0.2 I& aliquotsof the labeledextractsxere layeredonto 12-3296 v/v glycerolgradientspreparedin FWD buffer. The san'@eswere sedimented for 17 h at 237,OOO%g. ~14Cly~lo~lin,6.6s (32) and Cz~Clovalbumin, 3.55 (33)were used as internalmarkers.
STEROIDS
51/5-6
May-J une 1988
505
506
Ratajczak etal
conditions, producedan equivalentnumberof labeledbinding in Figure1, which sitesfor each hormone. This is demonstrated showsthe ultracentrifugal patternsfor receptorlabeledwith eithersteroidto be almostsuperimposable if specificactivity differences are taken intoaccount. In both casesa singlepeak in the 9s regionwas observed. The receptor-[3H10rg 4333 complexwas remarkablystable to treatmentwith dextran-coated charcoalat elevated temperature.Nearly75% of the conplexformedby 13H10rg 4333-labeling of a partiallypurified(heparin-se&arose chromatography) receptorextractwas stillmeasurableafter 30 min at 45*C (Figure2). In contrast,[3Hlestradiol-labeled receptorfrom the same extractwas completely dissociated under the same conditions(Figure2). The receptor-[3H10rg 4333 complex,however,failedto survivemore rigoroustestsof covalentbindingand was foundto be sensitiveto hot solvents used in a filterdisc-solvent extractionassay. The complexwas also
unstable
to denaturing
gel
electrophoresis. Increased
temperature, exclusionof dithiothreitol and n&ybdate from receptorpreparations, and emsure of cytosolicreceptorto r3H10rg4333 in bufferswith varyingpH (pH 7.4 - 8.5)failed to inducecovalentattachrtmnt. An indication of the affinityof Org 4333 for the receptor was obtainedby Scatchardplot analysisof data derivedfrom
STEROIDS
51/S-6
May-j une 1988
AFFINITY
10
LABELS
FOR ESTROGEN
RECEPTORS
20 TIME
Imin)
Figure2. Time+ependentstabilityof receptor-13Hlestradiol and receptor-['HIORG 4333 corrplexes to dextran-coated charcoaltreatmentat 45OC. Estradiolreceptor,partially purifiedby heparin-se&arose chromatography, was labeledby overnightincubation at 4OC with 30 nM 13Hlestradiol (0)or f3H10rg4333 f(r) in the absenceor presenceof 3000nM of the corresponding non-radioactive steroid. Aliguots (400&I fromthe labeledextractswere mixedat 4OC in tubes containingcharcoalpelletsfrom 1 mL aliguotsof dextran-coated charcoalsuspension(0.5%w/v charcoal,0.05%dextran). After 20 min the sampleswere furtherincubatedat 45OCwith occasional mixing. At selectedtimesduplicatesamplesware centrifuged and 200 ti aliguotsof the sup3rnatants were withdrawnfor radioactivity estimation.Receptor-bound radioactivity was calculatedaftercorrectionfor nonspecific bindingand expressedas a percentageof the controlvalue (=lOO%)observedpriorto 4J°C incubation.
STEROIDS
51/5-6
May-j une 1988
507
508
Ratajczak etal
incubation of uterinecytosolaliquotswith increasing concentrations of the radioactive ligand. The eguilibrium dissociation constant,I$.,, for Org 4333 was 2.8 + 1.8 x 1O-1oM (n = 2). Usingthe same extractsthe I$, valuedeterminedfor estradiolwas 3.1 + 0.4 x 1O-1oM (n = 2). The natureof receptor-Crg 4333 interaction was further characterized by comparingthe dissociation propertiesof 13Hlestradiol and 13H10rg4333 from the receptoras measuredby displacement of the labeledestrogenswith unlabeled we observedthe well estradiolat 30cC. For f3Hlestradiol documented(15,26. 27) biphasicor tm wmlpnent. firstorder dissociation curve (Figure3) generatedby low and high affinity bindingformsof the receptor.The rate constantof the more rapidwpnent
was k_, = 113.0+ 12.6 x 1O-3min-1
(n = 3). whilethat of the slowerdissociating wvnent
was
k-Z = 16.7 + 7.5 x 1O-3min-l (n = 3). Underthe same conditionsf3H10rg4333 dissociated only slowlyfrom the receptorand followadmonophasickineticswith a rate wnstant k = 1.3 + 0.2 x lo-'min-l (n = 3) (Figure31. Characteristics of estrouenreceatorinteraction with 13Hltamoxifen aziridin!& Receptorfrom crudecytosolwas incubatedwith ['HITA underconditions previouslyfoundto be optimalfor the covalent labelingof rat uterineestrogenreceptor(5). Specific,
STEROIDS
51/5-6
May-1 une 1988
AFFINITY
LABELS
30 TIME
FOR ESTROGEN
RECEPTORS
60 hn,
Figure3. Dissociation of LfHlestradiol and I)HlOrg4333 from the estrogenreceptor. Calf uterinecytosolin Tris buffer was equilibrated overnightat 4OC with 30 nM [aHlestradiol (0)or IaHIOrg4333 (0). After the additionof 5 PM unlabeledestradiolthe extractswere incubatedat 30°Cand aliquots(0.1 ml) were assayed for dissociation.Inactivation of the receptor-[:3Hlestrogen con@exes was monitoredunder the same conditions withoutadditionof unlabeledestradiol. All dissociation rreasurements were correctedfor nonspecific steroidbindingand receptorinactivation.The dissociation ratesware: Estradiol,k_, = 128 x lo-'min-I,k_, = 24 x lo-)min-I,Org 4333,k = 0.9 x lo-'mini.
STEROIDS
51/5-6
May-June 1988
509
510
Ratajczak etal
Figure4. Ultracentrifugal analysisof heparin-sepharose chromatography purifiede&radio1 receptorboundto [3Hlestradiol and 13Hltamoxifen aziridine.Estradiol receptorwas labeledfor 3 h at 4OC with 30 nM [3Hlestradiol (A)or 15 nM raHITAin the absenceor presenceof 3000nM unlabeledestradiol.All samplescontained7% UW. After removalof unboundligands(dextran-coated charcoal)the samples were subjectedto ultracentrifugal analysison 12-32%(V/V) glycerolgradientsas describedin the Figure1 legend. Fractionsfrom IfHlTA-labeled gradientswere dividedand measuredfor radioactivity (B)and covalentlabeling(C) using the ethanoldisc-extraction assaydescribedin Materialsand Methods. Catalase,11.3s (34)was used as an external sedimentation standard. Internalmarkerswere IX4C17-globulin, 6.65 (32)and [14Clovalbumin, 3.55 (33). covalentlylabeledreceptor,as determinedby the disc-extraction assay,represented 25% of the [fHlestradiol bindingsitesin the extract. Ultracentrifugal analysisof crudecytosolicreceptorlabeledwith [3~1~~revealeda major of radioactivity at 4.7s.whichwas only weaklyinhibited with unlabeledestradiol(notshown). Tanoxifenaziridinebinds -Y
proteins,includingalbumin,nonspecifically (5).and
STEROIDS
51/5-6
May-June 1988
AFFINITY
further
LABELS
FOR ESTROGEN
RECEPTORS
preliminary studieswith this ligandwere carriedout on
receptorextractspartiallypurifiedby heparin-sepharose chromatography.[)HITAlabelingwas more selectivefor receptorin theseextractsand the percentageof covalently coupledreceptorwas markedlyincreased(50%). In glycerolgradientspartiallypurifiedestrogenreceptor labeledwith ['Hlestradiol sedimented at 10s (Figure4A). The sedimentation profileof the [fHITA-labeled extract differedconsiderably and indicatedspecifically bound radioactivity at 9.6S,6.6S,and 3.5s (Figure4B). A high level of nonspecific bindingat 3.5swas also noted (Figure4B). When fractionscontaining bindingactivitywere assessedfor covalent attachment by the disc-extraction assay ['HITAwas foundto be boundcovalentlyto proteinsin all threesedimentation regions(Figure40. SDS-PAGEanalysisof a heparin-sepharose treatedreceptor extract,labeledwith [J~I~~,revealedtwo peaksof covalently bound (estradiolcompetitive)
radioactivity
that
correspondedto labeledcomponentswith M, 67,000and 50,000 L
(notshown).
DISCUSSION In this
reportwe have describedthe preparation of
formand llg-chloromethylestradiol (Org4333)in tritium-labeled
STEROIDS
51/5-6
May-June 1988
511
512
Ratajczak et al
have studied the properties of this compoundas an affinity labeling reagent for the estrogen receptor. confirrm?dearlier reports
14,
Our studies have
11-131, based on experiments with
unlabeled Org 4333, that the steroid interacts strongly with the receptor.
Our results indicate, however, that the
receptor-E3H10rg 4333 complex can be disrupted under denaturing conditions. in Org 4333 iqarts
Thus, although the ~loro~thyl~
considerable stability
group
to the interaction
with receptor, the chlorine atom appears not to undergo nucleophilic substitution at the binding site (13). of this interaction warrants further study. greatly facilitated
by the availability
The nature
This should be
of tritium-labeled Org
4333
l
The man ~i~~~i~
dissociation
constant (2.8 x IO-lo MI,
derived for the interaction of receptor with 13H30rg4333 comparedfavourably with values of 1.2 x 1O-*o M, and 2.2 x 1O-1o M previously obtained by competitive Scatchard analysis (4) with the unlabeled ligand in calf and humanuterine cytosols,
respectively.
!Fheslow dissociation
rate observed for tfHl0rg 4333 from
receptor at 30%~was consistent with the low levels of exchange previously achieved, under similar conditions, with E3Hlestradiol for receptor binding sites presaturated with radioinert Org 4333 (4).
Weichmanand E&tides (28) have
STEROIDS51/S-6
May-June1988
AFFINITYLABELS FOR ESTROGEN RECEPTORS
proposedthat the dissociation kineticsof an estrogenfrom receptormay providean indexfor biological activityby reflecting the capacityof the hormoneto activateand maintain the receptorin a high affinitybindingstate. As previously described(15,26-28)the fast and slowlydissociating formsof receptorcomplexeswith e&radio1 may representnon-activeand activatedstates,respectively.With Org 4333only the slow dissociating form was evident,suggesting that the steroidmay readilyactivateall the receptorsto a form capableof high affinitybinding. In an earlierreport(29)receptor-estradiol complexes transformed at 37% to a nondissociable, thermallystableform wereproposedas the functional units requiredfor genomic activation.Pihether the receptor*rg4333 corrplex is equivalent to this stabilizedform of the receptormay be answeredby comparative studiesof the influencethat estradioland Org 4333 have on the interaction affinityof receptorwith non-specific DNh sequencesas well as estrogenrespnsive DJ& elements involvedin the controlof estrogenregulatedgenes (30,31). Our sedimentation experiments with the rolybdate-stabilized estrogenreceptorof calf uterushave consistently shown I*Hlestradiol-labeled receptorto sedimentas a single, well-defined peak at 9-10s (3). ~~u~as~~~
b~d~g~~
(9.3s)was observedfor receptorcovalentlylabeledwith
STEROIDS 5-l/5-6May-June
1988
513
514
Ratajczak &al
[fHlTA,the presenceof snmllercovalentlylabeledcomponents at 6.6sand 3.5s resultedin a markedlydifferentsedimentation profile. In contrast.Katzenellenbogen et al (51have demonstrated similarultracentrifugal patternsfor cytosolicrat uterineestrogenreceptorlabeledwith 13Hlestradiol. and 13HITA. Using low salt buffersthe same groupobtained sedimentation coefficients of 8s for both receptorcomplexesand showedthat receptorboundcovalentlyto EfHITAgave a similargradientprofile(5). The labeledcomponents(M, = 67,000,50,000)observedafter exposureof partiallypurified,calf uterinereceptorextracts to f3HlTAcorrespndedcloselyin molecularsize to those reprted previouslyby Katzenellenbogen et al (5) for I'HITA-labeled estrogenreceptorfrom rat uterus (M, = 60,000- 66,000,50,000). Unitsof similarsize
(Mr = 63,000)have also been reportedfor 13~l~~-labeled cytosoland nuclearestrogenreceptorsfrom humanMCF-7 cells (6). The possibility existsthat the smaller50K unit,observed in our studyto be covalentlylabeledby ['HITA,is a proteolytic fragmentof the 67K component.Monsma&t
(61
have describeda labeledreceptorform of similarsize (Mr =
53,000)afterproteolysis, by endogeneous enzymesof the
ML.= 63,000 t3H1TA-labeled estrogenreceptorin cytosolsof
STEROIDS
51/5-6
May-June 1988
AFFINITY
LABELS
FOR ESTROGEN
RECEPTORS
Mm-7 cells. The
resultsof this studyindicatethat to date tamoxifen
aziridineremainsthe best availableaffinitylabelingreagent for estrogenreceptorsand studiesusingthis ligandshould continueto providevaluablestructuralinformation on the protein. AlthoughOrg 4333 is not an affinitylabelfor the estrogenreceptorthe tritium-labeled compoundmay prove useful for the studyof receptorundernondenaturing conditionsand for studiesof estrogenaction.
This work was supportedbytheNationa1 Health andMedical ResearchCouncilof Australia.
REFERENCES
of methods 1. RatajczakT, and H&e1 R (1980). Investigation of estradiolreceptorpurification priorto affinity chromatography. J. SIERGIDBIOCI-NM. 13:439-444. 2. GschwendtM, H&nel R, and RatajczakT (1983). Purification to homogeneity of the nuclearestrogenreceptorfrom chick liver. BIGCHIMBIOPHYSAC!l'A 760:238-245. 3. AtracheV, RatajczakT, SenafiS, and H&e1 R (19851. Purification of the molybdate-stabilized 9-10sestradiol receptorfrom calf uterus. J BIOL CHEM 260:5936-5941. 4. RatajczakT, SheppardPN, caponRJ, and H&nel R (1981). The synthesisand studyof some potentialaffinitylabeling reagentsfor estrogenreceptors.STEROIDS38~537-555. 5. Katzenellenbogen JA, CarlsonKE, HeimanDF, RobertsonDW, Wei IL, and Katzenellenbogen BS (1983). Efficientand highlyselectivecovalentlabelingof the estrogenreceptor with IsHltamoxifen aziridine.J BIOL CH@l 258:3487-3495.
STEROIDS
51/5-6
May-June 1988
515
516
Ratajczak
et al
MonsmaFJ, Jr, Katzenellenlmgen BS. MillerMA, ZieglerYS. and Katzenellenbogen JA (1984). Characterization of the estrogenreceptorand its dynamicsin MCF-7hwnanbreast cancercellsusinga covalentlyattachingantiestrogen. ENmCRINDLIx;y 115:143-153. BS 1. MillerMA, MullickA, GreeneGL, and Katzenellenbogen (1985).Characterization of the subunitnatureof nuclear estrogenreceptorsby chemicalcross-linking and denseamino 117:515-522. acid labeling. ENDCCRINX.GGY T (1973). The influence 8. H&nel R, TwaddleE, and Ratajczak of syntheticanti-estrogens on the bindingof tritiated estradiol-178 by cytosolsof humanuterusand humanbreast BIGCHEM4:687-695. carcinoma.J SI'ERGID H (1982). Tamxifen and 9. Coezy E, BorgnaJL, and Rochefort nratabolites in MCH cells:correlation betweenbindingto estrogenreceptorand inhibition of cell growth. CANCERRES 42:317-323. 10. EckertRL, and Katzenellenbogen BS (1982). Physical properties of estrogenreceptorcomplexesin MCF-7human breastcancercells. J BIOL CHEM 237:8840-8846. 11. Van den BroekAJ, LeemhuisJ, De WinterMS, and ZeelenF (1983).Org 4333,a potent,irreversibly bindingestrogen agonist. PHARMWIEKBL LSCII 5:182-183. 12. LeclercqC, Devleeschouwer N, and HeusonJC (1983). Guide-lines in the designof new antiestrogens and cytotoxic-linked estrogensfor the treatmentof breast cancer. J STEROIDBIOCHEM19:75-85. 13. ReinerGCA, Katzenellentmgen BS, BindalRD, and Katzenellenbogen JA (1984). Biological activityand receptorbindingof a stronglyinteracting estrogenin human breastcancercells. CANCERRES 44:2302-2308. 14. BowersA, Halsalll'G,JonesERH, and LeminAJ (1953). The chemistryof the triterpenes and relatedcompounds.Part of the structureof polyporenic acid C. XVIII. Elucidation J CHEN Sot 2548-2560. 15. WeichmanBM, and NotidesAC (1977). Estradiol-binding kineticsof the activatedand nonactivated estrogen receptor. J BIOL CHFX 252:8856-8862. 16. LaermliUK (1970). Cleavageof structural proteinsduring the asser&lyof the head of bacteriophage T4. NATDRE227: 6.
680-685. 17.
Maize1JV, Jr. Polyacrylamide gel electrophoresis of viral proteins. In: Methodsin Viroloav@laranmuschK and KaprowskiH, eds),AcademicPress,New York (1971). V:pp 179-246. 18. FowlerAV, and Z&in I (1977). The aminoacid sequenceof 8-galactosidase of &cherichia&. PXICNATLACADSCIUSA 74:1507-1510.
STEROIDS
51/5-6
May-j une 1988
AFFINITY
LABELS
FOR ESTROGEN
RECEPTORS
19. TitaniH, KoideA, HermannJ. EricssonLH, KumarS, Wade RD. WalshKA, NeurathH. and FischerEH (1977).Completeamino acid sequenceof rabbitmuscleglycogenphosphorylase.PRGC NATL ACAD SC1 USA 74:4762-4766. 20. McGillivray WA, MendeZE, SinhaSK, SuttonMR, Lineback-Zins J, and Brew K (1982). The completeaminoacid sequenceof human serumtransferrin.PfiDcNATLACADSCIUSA 79:2504-2508. 21. SquireI?$ Moser P, and O'KonskiCT (1968). The hydrodynamicpropertiesof bovineserumalbuminmonomerand 7~4261-4272. dimer. BIOCHEMISTRY 22. NisbetAD, SaundryRH, Moir AJG, mthergill LA, and mthergill JE (1981). The completeamino-acidsequenceof hen ovalbumin.EUR J BIGCHEM115:335-345. 23. ReynaudJ, LuccioniF, BouthierM. !&varyJ, and DerrienY (1971). Etudehydrodynamigue compareedes anhydrases carbonigues erythrocytaires bovinesA et B. BIOCHIMIE53: 1095-1098. 24. KoideT, and IkenakaT (1973). Studieson soybeantrypsin inhibitors.I. Fragmentation of soybeantrypsininhibitor (KImits) by limitedproteolysis and by chemicalcleavage. EUR J BIGCHEM32:401-407. 25. Brew K, VanamanTC, and Hill RL (1967). Comparison of the aminoacid sequenceof bovinea-lactalbminand hens egg white lysozyme. J BIOL CHEM 242:3747-3749. 26. SakaiD, and GorskiJ (1984).Estrogenreceptortransformationto a high-affinity statewithoutsubunit-subunit 23:3541-3547. interactions.BIGCHEMISTRY 27. de Boer W, Ab G, and GruberM (1985). Estrogenreceptorin chickenoviduct:receptordissociation kineticsand transformation. J Sl‘Et-QID BIOCHEM23:9-18. 28. ~7eichman BM, and NotidesAC (1980). Estrogenreceptor activation and the dissociation kineticsof estradiol, 106:434-439. estrioland estrone. ENmcRrNDII)(;y 29. FishmanJH, and Fishmn J (1985). The formation at 37OCof a nondissociable receptor-estradiol complex. BIOCHEM BIOPHYSRES CXMUN 131:58-63. 30. SkafarDF, and NotidesAC (1985). Modulationof the estrogenreceptor's affinityfor DNA by estradiol.J BIOL CHEM 260:12208-12213. 31. Klein-Hitpass L, SchorppM, WagnerU, and RyffelGU (1986). An estrogen-responsive elementderivedfrom the 5' flanking regionof the Xenopusvitellogenin gene functionsin s 3-1061. transfected humancells. CELL 46:lO 32. CalvanicoNJ, and Tomsi TB, Jr, (1971). Fragmentsof human yG imnunoglobulins producedby porcinepancreatic esterase. ARCHBIOCHEMBIOPHYS144:269-280.
STEROIDS
51/5-6
May-June 1988
517
518
Ratajczak etal
33. CastellinoFJ, and BarkerR (1968). Examination of the dissociation of multichainproteinsin guanidine hydrochloride by merrbrane osnmetry. BIOCHEMISTRY 1:2201-2217. 34. SumnerJB, and GralenN (1938). !rhemolecularweightof crystalline catalase. J BIOL CHEM 125:33-36.
STEROIDS
51/5-6
May-June 1988