Binding of RU486 and deacylcortivazol to the glucocorticoid receptor is insensitive to sulfhydryl-modifying agents

J. Steroid Biochem. Molec. Biol. Vol.44, No. 3, pp. 217-225,1993

0960-0760/93$6.00+ 0.00 PergamonPressLtd

Printedin GreatBritain

BINDING OF RU486 AND DEACYLCORTIVAZOL TO THE GLUCOCORTICOID RECEPTOR IS INSENSITIVE TO SULFHYDRYL-MODIFYING AGENTS T. BUROLLAUD,P. M. DANZI~,N. TBARKA,P. FORMSTECHER*and M. DAUTREVAUX Laboratoire de BiochimieStructurale, Facult~de M&lecine, 1, Placede Verdun, 59045 LilleCodex,France (Received 23 September 1992; accepted 30 October 1992)

Smnmary--The differential sensitivity of the rat liver glucocorticoid receptor (GR) to sulfhydryl group modifying agents when bound to various agonist and antagonist ligands was studied. [3H]Triamcinolone acetonide (TA) binding was completely abolished by previous treatment of the unbound receptor with various N-alkylmaleimides. On the contrary, [3H]RU486 binding was only slightly affected by treatment with N-ethylmaleimide (NEM) and more significantly decreased with maleimides bearing bulky substituents. Ligand exchange experiments demonstrated that, unlike the agonist TA, the antiglucocorticoid RU486 was unable to protect the GR binding site from the effect of NEM. This lack of protection would seem to be due to the presence of the bulky 11/~-substituent in RU486 since RU26988 and RU28362, two 1lfl hydroxylated glucocorticoids bearing the same 17~t-propynyl side chain as RU486 but lacking the 1l/~-substituent could protect GR against NEM. The ability of a GR ligand to prevent NEM inactivation of TA binding appeared unrelated to its agonist or antagonist nature: deacylcortivazol, a potent agonist, afforded no protection whereas antagonists of the 17fl-carboxamide series did. These data strongly suggest that compounds bearing bulky substituents on the steroid A and/or C rings, like deacylcortivazol and RU486, are positioned differently from canonical glucocorticoids in the steroid binding groove of the GR.

INTRODUCTION

A critical event in the action of hormonal steroids is their interaction with their intracellular receptors which are ligand-inducible transcription factors. The ligand binding domain of the steroid receptors is encoded by an approx. 250 amino acid sequence, localized in the C-terminal portion of the protein [1, 2]. Very little is known about the molecular details of the steroid-receptor interaction. One of the most investigated models in this respect is the glucocorticoid receptor. The requirement of intact thiol groups for steroid binding is a well-documented characteristic of the unbound glucocorticoid receptor [3, 4]. The presence of bound steroid protects the receptor against inactivation by sulfhydryl-modifying reagents whereas sulfhydryl-protecting agents, such as fl-mercaptoethanol or dithiothreitol, stabilize the unbound receptor and reverse the effects of a mild oxidation. The kinetics of the inactivation o f the binding activity by various N-alkylmaleimides has suggested that the thiol groups essential for steroid binding lie in *To whom correspondenceshould be addressed.

an hydrophobic environment [5]. The use of other sulfhydryl-modifying reagents like methyl methanethiolsulfonate, hydrogen peroxide, arsenite and selenite has implied the involvement of several thiols in steroid binding which could be reversibly inhibited by the formation of an intramolecular disulfide bridge involving a putative vicinal dithiol [6-10]. These data are strongly supported by the direct identification of two different cysteines in the steroid binding domain of the receptor by the thiolspecific affinity label dexamethasone 21mesylate[ll, 12] and by photoaffinity labeling using triamcinolone acetonide (TA)[13]. A point should nevertheless be stressed: all these results were obtained with steroid ligands (cortisol, corticosterone, TA, dexamethasone and its 21-mesylate derivative) bearing most of the structural features of a canonical glucocorticoid, i.e. a pregnene skeleton with a C3-carbonyl, an ll/~-hydroxyl and a 17//hydroxyacetone side chain [14]. However, several steroids characterized by the presence of a modified 17 side chain (DXB, RU26988, RU28362), or by the replacement of the l l/~hydroxyl or 3-keto groups by a very bulky substituent (RU486 and deacyicortivazol) still

217

218

T. BUROLLAUDet

display very good affinity for the glucocorticoid receptor (Fig. 1). Since the 17//-hydroxyacetone side chain and the 3-keto group have been each supposed to interact with one of the two cysteines identified in the binding domain by affinity labeling [12, 13], the involvement of thiol groups in the binding of steroids devoid of these structural characteristics needed to be explored. Interestingly, DXB and RU486 are both antiglucocorticoids[15, 17], whereas RU26988, RU28362 and deacylcortivazol are potent glucocorticoid agonists [18, 19]. Therefore the short

al.

series of steroids chosen here afford th~ oppor: tunity to seek an eventual correlation between the agonist or antagonist nature of a ligand and its requirement of unmodified thiol groups for receptor binding. In the present work we compared the ability of the various ligands depicted in Fig. 1 to bind to the rat liver glucocorticoid receptor after chemical modification of the receptor by various maleimides and their ability to protect the receptor against the effects of N-ethylmaleimide (NEM).

COCH2OH o.

DXB

Dexamethasone

COCH20H

COCH20H

0 "~"

~

V

Deacylcortivazol

Triamcinolone acetonide

/ OH 0

~

~

~N~T~" ~

C"==CCH3

0

f

v

OH C==,CCH

v

R RU 486 R = H: RU 26988 R = CH~:RU 28362 Fig. 1. Chemicalstructuresof the steroidsused.

3

219

RU486, deacylcortivazol and SH groups of GCR MATERIALS AND METHODS

Chemicals

[1,2,4(n)-‘HITA (29 Ci/mmol) was purchased from Amersham Int. (Amersham, England). Unlabeled TA was from Serva (Heidelberg, Germany). [6,7-3H]RU486: 1l/3(4-N,N-dimethylaminophenyl)-17/?-hydroxy17a -( 1-propynyl)-estra-4,9( l O)-dien-3-one (44 Ci/ mmol) and unlabeled RU486, RU28362: 6-methyl- 1lb, 17/&dihydroxy- 17c+(1-propynyl)androsta- 1,4,6-triene-3-one, RU26988: 1I/?, 17/Idihydroxy- 17a-( 1-propynyl)-androsta- 1,4,6-triene-3-one and deacylcortivazol: 11/?,17a,2 1-trihydroxy-6,16a-dimethyl-2’-phenyl-2’H-pregna2,4,6-trieno[3,2-clpyrazol-20-one, were obtained from Roussel-Uclaf (Romainville, France). DXB: N-benzyl9a-tluoro-16a-methyl-11/3,17adihydroxy-3-oxo1,4-androstadiene- 17/?-carboxamide was synthesized as described previously [20]. Acetonitrile was from Farmitalia Carol Erba (Paris, France). NEM was purchased from Pierce Chemical Co. (Rockford, IL, U.S.A.). N-propyl to N-nonyl and benzylmaleimides were prepared from maleic anhydride and the appropriate amine through a two-step synthetic procedure [5,21]. Other chemicals were of analytical commercial grade. Bufirs

The following buffers were used: Bu&r A : 20 mM Tes (N-tris-hydroxymethylmethyl-Zaminoethane sulfonic acid), 1 mM EDTA, 60mM NaCl, 10% (v/v) glycerol, 10mM Na,MoO,; pH 7.4; 4°C. Bufir B: buffer A supplemented with 1 mM PMSF (phenylmethane sulfonic fluoride). Bufir C: buffer A with 0.5 or 1 mM DTT (dithiothreitol). Preparation of cytosol

Adrenalectomized male Wistar rats (250 g body weight) were killed by decapitation. The livers were removed and successively perfused with 0.9% NaCl (wt/vol) and ice-cold buffer A or C. They were homogenized in buffer B (1.5 ml/g of tissue) using a Teflon-glass Potter homogenizer and centrifuged at 320,OOOg for 60 min at 4°C. The supernatant or cytosol was then recovered by aspiration and the pH was readjusted to 7.4. [-‘HISteroid-receptor binding

Cytosol samples were incubated in duplicate overnight (16-18 h) at 4°C in the presence of

15 nM rHJTA. Bound mined in duplicate by coal (DCC) adsorption binding was measured samples in the presence non-radioactive TA.

radioactivity was detera dextran-coated charassay 1221.Non-specific by incubating parallel of a lOOO-foldexcess of

Modification maleimides

receptor

of

the

by

N-alkyl-

Five hundred millimolar solutions of the various N-alkylmaleimides were extemporaneously prepared in ethanol. Maleimides were then added to the cytosol at a 3 or 5 mM final concentration. Alkylation was performed for varying times at 4°C and ended by addition of /3-mercaptoethanol (20 mM final concentration). The alkylation samples were submitted to gel-filtration on an Ultrogel ACA 202 Column (IBF) equilibrated in buffer A or C. Exchange procedure [23]

Cytosol samples were incubated with the first steroid for 4 or 16 h at 4°C. In samples treated with NEM, alkylation was stopped by adding fi-mercaptoethanol. The excess of free ligand and NEM was removed by charcoal adsorption or by gel-filtration. The samples were then reincubated with the second steroid for exchange in the presence of 8% acetonitrile. Exchange varied from 0 to 60 min. Bound radioactivity was determined by DCC assay. General methodr

A Coomassie blue (G250) assay was used for protein determination [24]. Sulfhydryl groups were assayed using dithiobisnitrobenzoate [25]. Radioactivity was measured in an LKB 1214 Rack beta spectrometer using Aqualyte@’ (Baker Chemicals, Deventer, The Netherlands) as scintillation cocktail. RESULTS

Differential sensitivity of f3H]RU486 and t3HJTA binding activity to N-alkylmaleimides

In the rat liver there is no progesterone receptor [26] and both RU486 and TA bind to the same steroid receptor, the glucocorticoid receptor. There was no significant degradation of the test ligands in rat liver cytosol under the experimental conditions used (data not shown). When the unbound liver glucocorticoid receptor was alkylated with various N-alkylmaleimides, an irreversible inactivation of rH]TA binding

220

T. BUROLLAUD et al.

activity was observed, whatever the size of the substituent of the maleimide ring [Fig. 2(A)]. The same result has already been observed for [3H]dexamethasone binding activity [5]. On the contrary, [3H]RU486 binding activity appeared far less sensitive to maleimide pretreatment [Fig. 2(B)]. 80% of the binding activity was still present after alkylation with 3 mM NEM and more than 60% with 5 mM (data not shown). The same result was obtained with N-propylmaleimide. However, when longer chain N-alkylmaleimides or the bulky N-benzyl maleimide were used, significant inactivation was observed, but to a far lesser extent than for [‘HITA binding. Therefore, if as generally sup-

.-.-.. 1 Panel

A

1

8

0

Panel

8

0

15

10

5

Incubation

time

(h)

I

I

I

10

15

time

120

$ D P

100

’ &I

z

e

I-\

l\*_-----

80

t\

0

/

I 20 Incubation

.-.-•

1

.--

2

I 40 time

I 60 (min)

Fig. 3. Kinetics of steroid exchange in the presence of acetonitrile. After overnight incubation with 15 nM labeled (curves 1 and 3) or unlabeled (curve 2) TA or RU486, cytosol samples were submitted to the exchange procedure with 60 nM labeled (curves 1 and 2) or unlabeled (curve 3) steroids in the presence of 8% acetonitrile. At the times shown, specific binding activity was determined in duplicate and expressed as a percentage of the initial bound radioactivity (curves I and 3). Data for a TA/TA exchange are represented. Very similar curves were obtained for TA/RU486, RU486/RU486 and RU486/TA exchanges.

posed, NEM modifies a thiol essential for [‘HITA or [3H]dexamethasone binding, this thiol does not appear to be essential to [3H]RU486 binding. However its modification by a bulky maleimide substituent resulted in decreased binding activity, suggesting that RU486 binding to the receptor nevertheless occurs fairly close to this thiol group. RU486-receptor complexes are not protected against the e&c& of’ NEM as regards their steroid binding activity

B

S

Incubation

2

(h)

Fig. 2. Kinetics of [‘HITA and [‘H]RU486 binding to the glucocorticoid receptor after alkylation with various N-substituted maleimides. Duplicate samples of cytosol were pretreated with different maieimides at a 3 mM concentration for 5min. Reaction was stopped by P-mercaptoethanol (20 mM final concentration) and labeled steroid was added at a 1OnM concentration. At times shown, specific binding activity was assayed in duplicate and expressed as percentage of a control run with the same tritiated steroid in the absence of maleimide. (A) [3H]TA binding after treatment with N-ethyl (curve 2), N-nonyl (curve 3) or N-benzyl (curve 4) maleimide. (B) [‘H]RU486 after treatment with N-ethyl, N-propyl, N-hexyl, N-nonyl and N-benzyl maleimide (curves 2 to 6, respectively). In both A and B curve I was the control not treated with maleimide.

In order to study the ability of various steroid ligands to protect the receptor against the effects of treatment by NEM, we resorted to our previously described steroid receptor exchange assay in the presence of acetonitrile [23]. Each exchange experiment included three parallel assays: an association assay where an unlabeled steroid was exchanged with a radiolabeled steroid, a dissociation assay where a radiolabeled ligand was exchanged with an unlabeled one, and a control assay where the radiolabeled steroid was exchanged with itself and which enabled to check the stability of the total binding sites. Figure 3 shows typical results obtained in the absence of NEM treatment. TA or RU486 could be easily exchanged with themselves or with each other in 30 to 60 min (more than 80% steroid reassociation after 60 min). When the steroid-receptor complexes were treated with NEM, different results were obtained according to the nature of the ligands

RU486, deacylcortivazol and SH groups of G C R

used. TA-receptor complexes could be exchanged with either TA or RU486 [Fig. 4(A)], indicating a complete protection of the thiol group essential for TA binding against alkylPanel A

"O es O

120

loo o

80

2

o

40

3

°~

u 0,,

~

0

I

I

I

20

40

60

Panel B 120 lO0 1

~3

60

2

40 "~o

3

20 0

20

40

60

I n c u b a t i o n time ( m i n )

Panel C 120 O J~

.~ 100

~ 8o o

"-"

60

~

40

~

2o

N

0

3

20

ation by NEM. This group could lie in the steroid binding groove and be masked by TA in the TA-receptor complexes. On the contrary, alkylated RU486-receptor complexes could be exchanged with RU486, but not with TA [Fig. 4 (B and C, respectively)]. This strongly suggests that in these complexes the thiol group essential for TA binding was still accessible to covalent modification by NEM. In agreement with the direct alkylation experiments reported in Fig. 2, this resulted in an only slightly reduced exchange with RU486 [60% reassociation, curve 2, Fig. 4(B)], contrasting with the almost complete inhibition of TA binding [less than 15% reassociation, Fig. 4(C)]. The lack of protection of the thiol group by RU486 is not due to the absence of a 17[$hydroxyacetone substituent

I n c u b a t i o n time (rain)

80

221

I

f

40

60

I n c u b a t i o n time (rain) Fig. 4. Protection against N E M inactivation o f the receptor by previous TA or RU486 binding. After incubation overnight at 4°C with 15 nM TA or RU486, labeled (curves 1 and 3) or not (curve 2), cytosol samples were alkylated with N E M for 5 rain. NEM concentration was adjusted in function of the total sulfhydryl groups content of the cytosol samples which was in the 2.6-3.5 m M range. 4.0-5.0 m M N E M concentrations were required to allow a rapid and complete alkylation which was ended by adding 20raM 18-mercaptoethanol. After gel filtration on small ACA 202 columns, samples were submitted to exchange with 60 nM labeled (curves 1 and 2) or unlabeled (curve 3) steroids. At times shown specific binding activity was assayed and expressed as percentage o f the initial control value. (A) TA/TA or TA/RU486 exchanges. (B) RU486/RU486 exchange. (C) RU486/TA exchange.

Two main features distinguish the structure of RU486 from that of typical glucocorticoid agonists like TA or dexamethasone: the presence of a bulky p-dimethylaminophenyl group at the 11/~ position and the absence of the natural hydroxyacetone 171ffside chain, which is replaced by a 17~-propynyl substituent. This latter peculiar feature could be responsible for the lack of thiol protection since the 17/I-hydroxyacetone side chain was supposed to interact with the thiol group identified in the steroid binding site of the receptor by covalent labeling with dexamethasone mesylate [11, 12, 27]. In order to see whether the 17~-propynyl side chain could protect the receptor from alkylation by NEM, further experiments were performed with RU26988 and RU28362, two steroids bearing an 11/~ hydroxy group and the same 17 substituents as RU486. These compounds, unlike RU486, are very potent and highly specific glucocorticoid agonists[19]. RU28362 and RU26988 were able to protect the TA binding activity of the receptor against NEM inactivation (Table 1). However this protection was Table I. After overnight incubation in the presence of 15nM of the unlabeled steroid to be tested Protecting steroid TA RU486 RU26988 RU28362 DXB

[3H]'rA exchange yield (% of control) 83 15 49 5I 65

Cytosol samples were alkylated with NEM for 5 min and submitted to exchange with 60nM [3HrrA for 60min. Bound radioactivity was assayed and expressed as percentage of a non-alkylated control.

T. BUROLLAffD et al.

222

partial, in the 50% range, whereas TA gave 83% and RU486 15% or less protection in the same experiment. The striking difference between the protection levels afforded by RU486, RU28362 and RU26988 suggested that RU28362 and RU26988 were interacting with the steroid binding groove like classical agonists, whereas RU486 was positioned differently, with its 17 substituents more remote from the essential tbiol. This could result from a primary interaction of RU486 with the receptor involving the "11# hydropbobic pocket" postulated by Teutsch [17]. The ability of a steroid to prevent or not the effect o f N E M on TA binding activity is unrelated to its agonist or antagonist nature

In contrast to the lack of thiol protection afforded by RU486, a high level of exchange with [3H]TA was obtained for receptors complexed with DXB (Table 1), another antiglucocorticoid but of the 17fl-carboxamide series [15]. In this instance, the modification of the 17fl side chain by the introduction of a bulky benzyl substituent was apparently still compatible with thiol protection. We have previously reported that 17fl-carboxamide bearing a small alkyl substituent is also able to protect the binding site [28]. On the other hand, deacylcortivazol, one of the most potent glucocorticoid agonists available[18], characterized by a very bulky phenylpyrazolo substituent at position 3 and bearing the same 17fl side chain as TA, was unable to protect the TA binding activity of the receptor from the effects of NEM, whereas

80

1

.~. 6o 4o U

,~ 20 @

_

0

20

I

I

40

60

I n c u b a t i o n time (rain) Fig, 5. Lack of protection by deacylcortivazol against maleimide dependent receptor inactivation. Cytosol samples were incubated overnight in the presence of 15nM unlabeled deacylcortivazol,alkylated(curves 2 and 4) or not

(curves 1 and 3) by NEM and submittedto exchangewith 60 nM 13HrrA(curves 1 and 2) or 13H]RU486(curves3 and 4). Specificbinding is expressed as percentage of the total [JH]TA or [3H]RU486bindingassayedin a control sample.

m

120 1001

o

80

~. -

n •

2 t. 3

60_ o

40 20 0

--

I

#

I

I

20

40

60

80

Steroid

) .;, 100

I 1000

c o n c e n t r a t i o n (nM)

Fig. 6. Competition for steroid binding to the alkylated glucocorticoid receptor. Cytosol samples were treated (curves 2, 3 and 4) or not (curve 1) with 4.0 mM NEM for 5 min at 4°C and then incubated overnight in the presence of 10nM [3H]RU486and variable concentrations of TA (curves 1 and 2), DXB (curve 3) or deacylcortivazol (curve 4). Specificbindingactivitywas determinedin duplicate and expressed as percentage of a control run with [3HlRU486 alone. exchange with RU486 was unaffected (Fig. 5). This result strongly supports the hypothesis that deacylcortivazol does not interact with the steroid binding site in the same way as other glucocorticoid agonists. Deacylcortivazol binding to the receptor could involve a peculiar interaction of the bulky phenyl pyrazol group with the steroid binding groove in the A ring region, resulting in a modified position of the 17/1 side chain in the binding site, away from the thiol group protected by TA. Interestingly, deacylcortivazol 21-mesylate has been reported to be unable (contrary to dexamethasone 21mesylate), to covalently label the glucocorticoid receptor [29]. Finally, competition experiments demonstrated that alkylation of the unbound receptor with NEM resulted in complete loss of DXB binding, whereas deacylcortivazol binding was still possible (Fig. 6). Since neither DXB nor deacylcortivazol were available in the radiolabeled form, their ability to bind the alkylated receptor was deduced from their ability to compete with [3I-1]RU486 binding. DISCUSSION

The data presented here demonstrate that RU486 binding to the glucocorticoid receptor is far less sensitive to thiol modifying agents than the binding of classical glucocorticoid agonists. A similar differential sensitivity to sulfhydrylmodifying agents has already been described for the progesterone receptor when bound to the antiprogesterone RU486 or to the agonist

RU486, deacylcortivazoland SH groups of GCR

223

Another argument corroborating the hypothR5020 [30]. Whereas pretreatment of the receptor with NEM or iodoacetamide completely esis of a differential positioning in the binding abolished R5020 binding, it did not affect groove of steroids according to the steric bulk RU486 binding (iodoacetamide) or reduced it of their ring substituents was afforded by by only 40% (NEM). However, the insensitivity the comparison of the respective sizes of the of RU486 receptor binding to thiol modification ligand binding domain of the receptor and is not a general feature of antiprogestative and of its ligands. The steroid binding domain of antiglucocorticoid activities. Since, the binding glucocorticoid receptors encompasses the 250 of other antiglucocorticoids, like DXB, ap- carboxyterminal amino acids of the protein, peared sensitive to thiol modification, whereas according to various deletions and mutation binding of a potent glucocorticoid like deacyl- studies[33-36]. This defines a 31-34kDa cortivazol remained more or less unaffected. domain. However, using partial trypsin proteolIn our opinion, the differential sensitivity to ysis of the native receptor, a 16 kDa steroid maleimide alkylation of the glucocorticoid binding core was obtained, corresponding to receptor when bound to TA, RU486 or deacyl- about half the size of the canonical steroid cortivazol results from differential positioning binding domain and located in its N-terminal of the corresponding steroid in the steroid part [37]. This fragment was supposed to conbinding site. All the ligands tested compete rain most of the amino acids required for the for binding to the receptor with synthetic ligand binding function since it bound dexa(dexamethasone and triamcinolone acetonide) methasone with a rather high affinity and disand natural (corticosterone) glucocorticoid ago- played the same steroid binding specificity as the nists[15-19] and are therefore considered to intact receptor [27]. In particular the relative interact with the same core site. However, the binding affinities with respect to dexamethasone binding sites of the various natural and syn- of RU486, RU28352 and deacylcortivazol were thetic analogs are expected to only overlap and conserved. Owing to the size of a typical glucothe bulky substituents borne by RU486 and corticoid, as measured by radiocrystallography deacylcortivazol probably interact with other (C3-C,7 and O3-O21 distances of 8.5 and 11 A, parts of the ligand binding domain which are respectively), the distance between amino not involved, or only partly involved, in the groups facing the C3" carbonyl group and the binding of a ligand devoid of these substituents. 178-hydroxyacetone side chain in the steroid This peculiar interaction, which could even binding site was estimated to be in the 13 A become the primary interaction of these steroids range [38]. Moreover the depth of the hydrowith the receptor, could result in an unusual phobic 118 pocket explored with RU486 positioning of the steroid skeleton in the binding analogs was close to 10-12A according to groove. This hypothesis is supported by the the convincing data of Teutsch [19], whereas observed loss, in the case of RU486 of the structure-activity relationship studies of 178partial protection against the effects of NEM carboxamide derivatives revealed that the afforded by the 17~-propynyl side chain in nature of the 178 side chain affected the receptor RU26988 and RU28362. In RU486 the bulky binding affinity up to a 10 A distance from dimethylaminophenyl group added in 118 could CI7 [39] and the extension of the phenylpyrazolo modify the positioning of the steroid to fit into substituent in deacylcortivazol could be estithe 118 hydrophobic pocket not involved in mated between 5 and 6 A away from ring A. "normal" glucocorticoid binding[17]. On the It is difficult to imagine, with a rigid model of other hand, the presence of a classical 178- the structure of the ligand binding domain hydroxyacetone side chain on some RU486 of the receptor and assuming a unique kind analogs could be incompatible with a correct of positioning of the ligand, how a 16kDa positioning of the dimethylaminophenyl sub- polypeptide, i.e. with a Stokes radius very probstituent in the 118 pocket. This could explain, ably smaller than 20 A [40], could accommodate as suggested by Benhamou e t al.[31], why all the steric features of so different and so these compounds behave as progesterone ago- bulky substituted steroid molecules. To explain nists and display a 8- to 24-fold lower affinity our data and the surprising conservation of the for glucocorticoid receptor binding than binding specificity of the 16 kDa receptor fragRU486 [32]. It would be extremely interesting to ment a differential positioning of the various investigate the sensitivity of the binding of such steroids in the binding groove seems obviously compounds to thiol reagents. required and could be helped (or caused) by

224

T. BUROLLAUDet aL

the p u t a t i v e c o n f o r m a t i o n a l flexibility o f the s t e r o i d b i n d i n g c a v i t y p o s t u l a t e d b y several a u t h o r s [27, 40]. F i n a l l y , t h e g l u c o c o r t i c o i d r e c e p t o r is m a i n r a i n e d in a high-affinity s t e r o i d b i n d i n g c o n f o r m a t i o n w h e n b o u n d to H S P 9 0 [41], a p r o t e i n w h i c h c o u l d stabilize t h e 1 6 k D a f r a g m e n t , w h i c h e n c o m p a s s e s the H S P 9 0 b i n d i n g r e g i o n o f t h e r e c e p t o r [42], in a v e r y o p e n e x t e n d e d s t r u c t u r e a n d p a r t i c i p a t e in the f o r m a t i o n o f a f u n c t i o n a l b i n d i n g site. Acknowledgements--This work was supported by INSERM (CRE 910707), ARC and the University of Lille II. We are indebted to Drs Philibert and Teutsch from Roussel Uclaf for providing deacylcortivazol, RU486, RU28362 and RU26988. B. Masselot and S. Tournay are also acknowledged for technical and secretarial assistance.

REFERENCES 1. Evans R. M.: The steroid and thyroid hormone receptor superfamily. Science 240 (1988) 889-895. 2. Green S. and Chambon P.: Nuclear receptors enhance our understanding of transcription regulation. Trends Genet. 4 (1988) 309-314. 3. Rees A. M. and Bell P. A.: The involvement of receptor sulfhydryl groups in the binding of steroids to the cytoplasmic giucocorticoid receptor from rat thymus. Biochim. Biophys. Acta 411 0975) 121-132. 4. Young H. A., Parks W. P. and Scolnick E. M.: Effect of chemical inactivating agents on giucocorticoid receptor proteins in mouse and hamster cells. Proc. Natn. Acad. Sci. U.S.A. 72 (1975) 3060-3064. 5. Formstecher P., Dumur V., Idziorek T., Danze P. M., Sablonniere B. and Dautrevaux M.: Inactivation of unbound rat liver giucocorticoid receptor by N-alkylmaleimides at sub-zero temperatures. Biochim. Biophys. Acta 802 (1984) 306-313. 6. Miller N. R. and Simons S. S.: Steroid binding to hepatoma tissue culture cell 81ucocorticoid receptors involves at least two sulfhydryl groups. J. Biol. Chem. 263 0988) 15217-15225. 7. Bresnick E. H., Sanchez E. R., Harrison R. W. and Pratt W. B.: Hydrogen peroxide stabilizes the steroidbinding state of rat liver giucocorticoid receptors by promoting disulfide bond formation. Biochemistry 27 (1988) 2866-2872. 8. Tashima Y., Terui M., Itoh H., Mizunuma H., Kobayashi R. and Marumo F.: Effect of selenite on giucocorticoid receptor. J. Biochem. 105 (1989) 358-361. 9. Lopez S., Miyashita Y. and Simons S. S.: Structurally based, selective interaction of arsenite with steroid receptors. J. Biol. Chem. 205 (1990) 16039-16042. 10. Simons S. S., Chakraborti P. K. and Cavanaugh A. H.: Arsenite and cadmium (II) as probes of giucocorticoid receptor structure and function. J. Biol. Chem. 265 (1990) 1938-1945. II. Simons S. S. and Thompson E. B.: Dexamethasone 21-mesylate: an affinity label of giucocorticoid receptors from rat hepatoma tissue culture cells. Proc. Natn. Acad. Sci. U.S.A. 78 (1981) 3541-3545. 12. Simons S. S., Pumphrey J. G., Rudikoff S. and Eisen H. H.: Identification of cysteine 656 as the amino acid of hepatoma tissue culture cell glucocorticoid receptors that is covalently labeled by dexamethasone 21-mesylate. J. Biol. Chem. 262 (1987) 9676-9680.

13. Carlstedt-Duke J., Str6mstedt P . E . , B., Cederlund E., Gustafsson J. A. and Jornvall H.: Identification of hormone-interacting amino acid residues within the steroid-binding domain of the glucocorticoid receptor in relation to other steroid hormone receptors. J. Biol. Chem. 263 (1988) 6842-6846. 14. Rousseau G. G. and Schmit J. P.: Structure-activity relationships for glucocorticoids. (I) determination of receptor binding and biological activity. J. Steroid Biochem. 8 (1977) 911-919. 15. Rousseau G. G., Kirchoff J., Formstecher P. and Lustenberger P.: 17fl-carboxamide steroids are a new class of giucocorticoid antagonists. Nature 279 (1979) 158-160. 16. Philibert D.: RU38486: an original multifaceted antihormone/n vivo. In Adrenal Steroid Antagonism (Edited by M. K. Agarwall). Walter de Gruyter & Co, Berlin (1984) pp. 77-101. 17. Teutsch G.: 11 Beta-substituted 19-norsteroids: at the crossroads between hormone agonists and antagonists. In Adrenal Steroid Antagonism (Edited by M. K. Agarwal). Walter de Gruyter & Co, Berlin (1984) pp. 43-75. 18. Simons S. S., Thompson E. B. and Johnson D. F.: Anti-inflammatory pyrazoio-steroids: potent giucocorticoids containing bulky A-ring substituents and no C3-carbonyl. Biochem. Biophys. Res. Commun. 86 (1979) 793-800. 19. Teutsch G. and Costerousse G.: 17,~-alkynyl-llfl,17dihydroxyandrostane derivatives: a new class of potent giucocorticoids. Steroids 38 (1981) 651--665. 20. Formstecher P., Lustenberger P. and Dautrevaux M.: Synthesis of steroidal 17fl-carboxamide derivatives. Steroids 35 (1980) 265--272. 21. Blicq S., Danze P. M., Dumur V., Formstecher P. and Dautrevaux M.: Inhibition of giucocorticoid receptor transformation, subunit dissociation, and temperaturedependent inactivation by various N-substituted maleimides. Biochemistry 27 0988) 8436--8442. 22. Rousseau G. G., Baxter J. D. and Tomkins G. M.: Glucocorticoid receptors: relation between steroid structure and biological effects. J. Molec. Biol. 14 0972) 99-I05. 23. Danze P. M., Richard C., Formstecher P. and Dautrevaux M.: Steroid receptor exchange assay in the presence of acetonitrile: application to the study of giucocorticoid and antiglucocorticoid-receptor complexes. Steroids 55 (1990) 10-16. 24. Read S. M. and Northcote D. M.: Minimization of variation in the response to different proteins of the Coomassie blue G Dye assay for protein. Analyt. Biochem. 116 (1981) 53--64. 25. Ellman G. L.: Tissue sulthydryl groups. Archs Biochem. Biophys. 82 (1959) 70-77. 26. Str/ihle U., Boshart M., Klock G., Stewart F. and Schiitz G.: Glucocorticoid and progesterone--specific effects are determined by differential expression of the respective hormone receptors. Nature 339 (1989) 629-632. 27. Chakraborti P. K., Garabedian M. J., Yamamoto K. R. and Simons S. S.: Role of cysteines 640, 656, and 661 in steroid binding to rat giucocorticoid receptors. J. Biol. Chem. 267 (1992) 11366-11373. 28. Idziorek T., Sablonnitre B., Formstecher P., Dumur V. and Dautrevaux M.: Covalent chromatography by thiol-disulfide interchange of the highly-purified nontransformed rat liver giucocorticoid receptor. J. Steroid Biochem. 23 (1985) 593-597. 29. Simons S. S. and Miller P. A.: Affinity-labeling steroids as biologically active probes of antiglucocorticoid hormone action. J. Steroid Biochem. 24 (1986) 25-32. 30. Moudgil V. K., Anter M. J. and Hurd C.: Mammalian progesterone receptor shows differential sensitivity to

RU486, deacylcortivazol and SH groups of G-CR

31.

32.

33.

34. 35.

36.

sulfhydryl group modifying agents when bound to agonist and antagonist ligands. Y. Biol.i Chem. 264 (1989) 2203-221 I. Benhamou B., Garcia T., Lerouge T., Vergezac A., Gofflo D., Bigogne C., Chambon P. and Gronemeyer H.: A single amino acid that determines the sensitivity of progesterone receptors to RU486. Science 255 (1992) 206-209. Kloosterboer H. J., Deckers G. H. J., van der Heuvel M. 3. and Loozen H. J. J.: Screening of anti-progestagens by receptor studies and bioassays. J. Steroid Biochem. 31 (1988) 567-571. Weinberger C., Hollenberg S. M., Rosenfeld M. G. and Evans R. M.: Domain structure of human giucocorticoid receptor and its relationship to the v-erb-A oncogene product. Nature 318 0985) 670-672. Gigu~re V., HoUenberg S. M., Rosenfeld M. (3. and Evans R. M.: Functional domains of the human glucocorticoid receptor. Cell 46 (1986) 645-652. Rusconi S. and Yamamoto K. R.: Functional dissection of the hormone and DNA binding activities of the giucocorticoid receptor. EMBO J. 6 (1987) 1309-1315. Danielsen M., Northrop J. P., Jonklaas and Ringold G. M.: Domains of the glucocorticoid receptor involved in specific and nonspecific deoxyribonucleic acid binding,

37.

38.

39. 40.

41.

42.

225

hormone activation, and transcriptional enhancement. Molec. Endocr. 1 (1987) 816-822. Simons S. S., Sistare F. D. and Chakraborti P. K.: Steroid binding activity is retained in a 16-kDa fragment of the steroid binding domain of rat giucocorticoid receptors. J. Biol. Chem. 264 (1989) 14493-14497. Duax W. L., Weeks C. M. and Rohrer D. C.: Crystal structure of steroids: molecular conformation and biological function. Recent Prog. Horm. Res. 32 (1976) 81-116. Formstecher P., Lefebvre P. and Burollaud T.: Hormones et antihormones. Le module st~roidien. J. Pharm. Belg. 46 (1991) 37-48. Wrange (3. and Gustafsson J.-A.: Separation of the hormone- and DNA-binding sites of the hepatic glucocorticoid receptor by means of proteolysis. J. Biol. Chem. 253 (1978) 856-865. Scherrer L. C., Hutchison K. A., Sanchez E. R., Randall S. K. and Pratt W. B." A heat shock protein complex isolated from rabbit reticulocyte lysate can reconstitute a functional glucocorticoid receptor-Hsp90 complex. Biochemistry 31 (1992) 7325-7329. Chakraborti P. K. and Simons S. S.: Association of heat shock protein 90 with the 16 kDa steroid binding core fragment of rat giucocorticoid receptors. Biochem. Biophys. Res. Commun. 176 (1991) 1338-1344.