Characterization of the partially purified, ligand-free glucocorticoid receptor

Biochimica et Biophysica Acta 883 (1986) 215-224

215

Elsevier BBA 22421

Characterization of the partially purified, ligand-free glucocorticoid receptor P&er Krajcsi and P&er Arfinyi * Second Institute of Biochemistry, Semmelweis University Medical School, H-1444 Budapest (Hungary)

(ReceivedJanuary 27th, 1986) (Revised manuscriptreceivedMay 26th, 1986)

Key words: Steroidspecificity;Receptoractivation; Stokesradius; (Chick thymus) A new method was developed to synthesize a cortexolone-substitated affinity matrix, based on the fast, mild and quantitative reaction between a-ketomesylates and thiols. The resulting cortexolone-Sepharose adsorbed easily the cytosolic chick thymus glucocorticoid receptor. Owing to the relatively fast dissociation of the glucocorticoid receptor-cortexolone complex, glucocorticoid receptor could be eluted with cortexolone as well as with triamcinolone acetonide from the affinity gel with similarly good yields. We obtained 75-150-fold purification factors (yield: 20-30%) using this column procedure. The partially purified glucocorticoid receptor was obtained in non-activated form. It had a Stokes radius of 6.2 + 0.1 nm. It could be activated to DNA-cellulose binding form by heat or 0.3 M KCI. KCI treatment activated 30-50% of the partially purified glucocorticoid receptor. Heat activation, however, was rather poor. Cortexolone-complexed, partially purified ghicocorticoid receptor dissociated easily, and partially purified free glucocorticoid receptor, capable of steroid binding, could be obtained. Binding properties of the partially purified glucocorticoid receptor were then analyzed using different steroids. Dissociation rate constants were similar to those of the cytosolic glucocorticoid complexes. Association rate constants were consistently smaller than in the case of cytosolic glucocorticoid receptor, but the relative order of rates for different steroids was basically the same for glucocorticoid receptor in the two studied systems.

Introduction

The first obligatory step in the mechanism of glucocorticoid hormone action is the formation of a complex between the hormone and its receptor protein [1-3]. Parameters of this interaction are essential in the characterization of glucocorticoid receptor. Although our knowledge regarding the steroid specificity of glucocorticoid receptor is likely to be correct with regard to its general features, it has mostly been derived from studies

* To whom correspondenceshould be addressed. Abbreviations: PPO, 2,5-diphenyloxazole;POPOP, 1,4-bis(2(5-phenyloxazolyl))benzene.

on very crude preparations. An understanding of the detailed mechanisms of action of these hormones awaits studies with a system that is reconstituted with purified components. Up till now, steroid binding kinetics of purified glucocorticoid receptor have not been studied due to the fact that cytosolic glucocorticoid receptor, because of its lability, has only been purified complexed with dexamethasone or triamcinolone acetonide, and it could not be recovered from these complexes in a form that was still capable of binding steroids. Here we report a method of partial purification of nonactivated glucocorticoid receptor, which allows recovery of the aporeceptor with a substantial yield, and kinetic studies with this partially purified glucocorticoid receptor.

0304-4165/86/$03.50 © 1986 Elsevier Science Publishers B.V. (BiomedicalDivision)

216 Materials and Methods

Animals 6-12-week-old 'Hunnia hybrid' chickens were kept and fed as described previously [4]. Chemicals [1,2,4(n)-3H]Triamcinolone acetonide (26 C i / mmol) was purchased from the Radiochemical Centre (Amersham International, U.K.). Nonlabeled steroids were obtained from G. Richter Ltd. (Budapest, Hungary). Purity of these steroids, both chemical and radiochemical, was checked by thin-layer chromatography. Charcoal (Norit A) was purchased from Serva (Heidelberg, F.R.G.). CF-11 cellulose was from Whatman (Springfield Mill, U.K.). Sephacryl S-300 superfine, Sepharose 4B and proteins for molecular weight calibration were from Pharmacia (Uppsala, Sweden)• 1,4Butanediol diglycidyl ether was purchased from Ega-Chemie (Steinheim/Altbuch, F.R.G.). All other chemicals were of reagent grade and obtained from Reanal (Budapest, Hungary). Buffers Buffer A: 0.01 M Tris-HC1/1.5 mM EDTA (pH 7•4); buffer B: buffer A ( + 2 mM dithiothreitol); buffer C: buffer B (+20 mM Na2MoO4); buffer D: buffer A (+20 mM Na2MoO4). Preparation of the cytosol Chickens were killed by decapitation, and the thymuses were rinsed with physiological saline and weighed. All subsequent procedures were performed at 0 ° C. Thymuses were homogenized with 1.5 vol. buffer C in a motor-driven Teflon-glass homogenizer. The homogenate was centrifuged at 100000 x g for 45 min and the supernatant was used as cytosol. Preparation of cortexolone-21-mesylate To 363 mg cortexolone (1.05 mmol) in 8 ml pyridine at 0°C was added 100 /~1 methanesulfonyl chloride (1.3 mmol) with stirring. After 2 h of stirring at 0°C the reaction mixture was poured onto 200 g iced water. The precipitate was collected by filtration, washed with a total of 200 ml water at 0°C. Crude mesylate (405 mg, 92% yield) was obtained upon drying on air. It pro-

duced one spot on TLC (R F = 0.3, benzene/ethyl acetate 3:2). Recrystallization from dioxane/ heptane gave an analytically pure product (380 mg, 86%; m.p. 180-182°C; 1H-NMR (DMSO-d6) 0.63 (3H, S, 18-CH3), 1.2 (3H, S, 19-CH3), 3.25 (3H, S, SO2-CH3). NMR spectra measurements were carried out using a Bruker WM-250 FTNMR-spectrometer.

Preparation of affinity matrix 1,4-Butanediol diglycidyl ether was coupled to Sepharose 4B by the method of Grandics [5]. The epoxy-activated gel was converted to a matrix, containing SH-functions, according to Carlsson et al. [6]. Parameters of these reactions are given in Table I. 5 g of such freshly prepared suction-dried thioalkyl-Sepharose 4B was washed with ice-cold water/acetone 3:1 (20 ml), water/acetone 1:1 (20 ml), water/acetone 1 : 3 (20 ml) solutions and finally with acetone (5 x 20 ml). At last the gel was suspended in 8 ml dry acetone, and 352 mg (0.83 mmol) cortexolone-21-mesylate was added to the suspension at 0°C with stirring. After 10 min of stirring at 0°C, 50 ~tl triethylamine were added to the suspension. After stirring for another 3 h the gel was washed with acetone (10 x 20 ml), water/acetone 1:3 (20 ml), water/acetone 1:1 (20 ml), water/acetone 3:1 (20 ml) and water (10 x 20 ml). The gel is stable to storage in distilled water (0.02% NAN3). Yields of the coupling of the steroid to the matrix were 55-70%, determined by measuring the thiol contents• Determination of the thiol content 100 mg samples of the gel were reduced according to Carlsson et al. [6]. The thiol content was determined by the method of Brocklehurst et al. [7] by means of 5,5'-dithiobis(2-nitrobenzoic acid) (DTNB) in 0.01 M Tris-HC1/1.5 mM EDTA buffer (pH 8). Charcoal treatment Method A. Samples were incubated with 0.2 vol. 5% charcoal-0.5% dextran T-500, suspended in the same buffer as the samples, for 3 rain at 0 ° C, then centrifuged at 1500 x g for 5 rain, and the supernatants were used in further experiments. Method B. Samples were incubated with 2 vol.

217

0.5% charcoal-0.05% dextran T-500, suspended in the same buffer as the samples, for 3 min at 0°C, then centrifuged at 1500 x g for 5 rain, and given volumes of the clear supernatants were used for radioactivity determinations. Affinity chromatography Affinity chromatography was performed either as a column or as a batch procedure. Pilot experiments performed in order to optimize the yield and purification factor led to the following standard procedures. Batch 50 mg suction-dried affinity gel was equilibrated with buffer C or B as indicated. 1 ml cytosol was added and the resulting slurry was stirred with a stirring rod for 3 h at 0°C. After incubation, the suspension was washed with 1 ml of the indicated buffer. Cortexolone, 1.7- 10 - 4 M, or [3H]triamcinolone acetonide, 77 nM, was used for elution in 1 ml of the buffer used for washing by stirring for 3 h at 0°C. Eluate was separated from the gel by centrifugation (1500 x g, 2 min). Column Gel was packed into a column of 1.1 ml volume (8 x 22 mm) and equilibrated with buffer B or C. Cytosol (6 ml) was loaded on the column at a flow rte of 10 ml/h. Loading was followed by washing with buffer at the indicated flow rate (15-20 ml/h). The buffer used for elution of glucocorticoid receptor was supplemented with 1.7. 10 -4 M cortexolone or 77 nM [3H]triamcinolone acetonide. To determine the [3H]triamcinolone acetonide binding capacity of the eluates, in the case of batch as well as column procedures, the glucocorticoid receptor eluted with cortexolone was treated with charcoal-dextran by method A and the [3H]triamcinolone acetonide binding assay was performed thereafter. Also, glucocorticoid receptor eluted with [3H]triamcinolone acetonide was charcoal-dextran-treated by method B and protein-bound [3H]triamcinolone acetonide was determined directly from the supernatant. Further data about buffers used are given in the legend to Table II. [31-1]Triamcinolone acetonide binding assay This was performed by the charcoal method [8].

50 btl samples were incubated with an equal volume of [3H]triamcinolone acetonide diluted with buffer B or C to a final concentration of 38.5 nM, in the presence and absence of 3.9.10 -6 M unlabeled triamcinolone acetonide at 0-4°C for 210 rain. At the end of the incubation period samples were treated with charcoal-dextran (method B) and 250 /~1 of supernatants were transferred into vials and radioactivity was measured. The difference between the radioactivity of samples incubated with and without unlabeled triamcinolone acetonide gave the specific binding. Determination of denaturation rate constants Method A. Affinity chromatography was done by method A in 'batch' or in column. Affinity eluate was incubated at 0 or 25 °C. At indicated times 180 /~1 aliquots were withdrawn. Free cortexolone was removed by addition of charcoal-dextran (method A) and the clear supernatant was used for [3H]triamcinolone acetonide binding assay. Method B. Affinity chromatography was performed by one of the methods. Eluate was mixed with charcoal-dextran (method A) and clear supernatant was incubated at 0 or 25°C. At indicated times 50 #1 aliquots were withdrawn and used for [3H]triamcinolone acetonide binding assay. The remaining activities were plotted semilogarithmically vs. time. Parameters of the straight lines were calculated by the method of least squares. Sephacryl S-300 chromatography Affinity chromatography was done by method B. Gel filtration was performed at 0-4°C. A 2 X 33 cm column, equilibrated with buffer C (0.3 M KC1) was used. It was calibrated with dextran blue, thyroglobulin (8.5 nm), ferritin (6.1 nm), bovine serum albumin (3.55 nm), ovalbumin (3.05 nm) and 3H20. Flow rate was 22.5 ml/h. The column was loaded with 1.5 ml of eluate. 1.5 ml fractions were collected and assayed for radioactivity. Stokes radius of the purified glucocorticoid receptor-[3H]triamcinolone acetonide complex was determined from a calibration plot by the linear dependence of Kla/3 o n Stokes radius. Kav was calculated according to formula V~- V o / V i V0, where V~ is the elution volume of the sample, V0 is the exclusion volume of the column as

218

determined by blue dextran and V~ is the elution volume of the 3H20.

hormone-receptor complex formed during the first t min of incubation.

[ 3H] Triamcinolone acetonide-glucocorticoid receptor binding to DNA-cellulose Affinity chromatography was performed by method C batchwise or in a column. Activation and the binding of [3H]triamcinolone acetonidepurified glucocorticoid receptor complex to DNA-cellulose was determined as described previously [9], except that affinity eluate rather than cytosol was used.

Determination of protein content Protein content was determined by the Coomassie brilliant blue method [12] using bovine serum albumin as standard.

Determination of association rate constants Affinity chromatography was done as a batch or column procedure by method A. Affinity eluate was treated with charcoal-dextran (method A). Affinity eluate prepared in this way was used for determination of the association between purified glucocorticoid receptor, and [3H]triamcinolone acetonide or nonlabeled steroids were determined as described previously [10,11]. Determination of dissociation rate constants Affinity chromatography was performed by method A as a batch procedure but different steroids were used instead of cortexolone to recover bound receptor from the affinity matrix. The excess of steroid was removed by treatment of charcoal-dextran (method A). Dissociation rate constants were determined as described previously [10], but total concentration of purified glucocorticoid receptor was determined by incubation of the charcoal-treated affinity eluate with [3H]triamcinolone acetonide for a sufficiently long time (14-16 h). Column chromatography was done by method A. Free cortexolone was removed by treatment of charcoal-dextran (method A). The clear supernatant, obtained by centrifugation at 0°C, was incubated at 0°C for 2 h in the presence of different nonlabeled steroids at a concentration of 10 -5 M. After repeated charcoal treatment the clear supernatant was used for the dissociation experiment as described before. The dissociation rate constants were calculated from semilogarithmic plots of ln(R v - Bt) vs. time by the least squares method. R v is the total concentration of receptor and Br is the concentration of labeled

Radioactioity determinations Liquid scintillation counting was performed in a Beckman LS 350 radiospectrofluorimeter. Composition of the scintillation liquid was toluene/ PPO/POPOP (100 ml:4 g:0.05 g); counting efficiency was 36%. For determination of DNA-bound radioactivity Triton X-100 was mixed with 2 vol. of the scintillation cocktail described before and used as scintillation liquid. In this case counting efficiency was 27%. Results

Synthesis of the affinity matrix It was known that 21-mesylates of natural and synthetic glucocorticoids are thiol-specific reagents [19]. The reaction of mesylates with thiols is especially fast in the case of 17a-hydroxymesylates [13]. On the other hand, the preparation of thioalkyl-agarose gel was known [6]. It seemed attractive to couple steroid ligand to the matrix through a stable thioether linkage. Cortexolone-21-mesylate was synthesized and coupled to the thioalkyl-Sepharose 4B (Fig. 1). The coupling reaction was mild, fast and gave high yields (55-70%) and good reproducibility in a relatively wide range of concentration of reagents (Table I). Another advantage of the coupling method is that unreacted thiol groups do not cause any side reaction, so this procedure does not need any blocking reaction like other methods known for preparing glucocorticoid receptor specific affinity matrix. The matrix was used for several (20-40) subsequent experiments without any sign of leakage of the steroid from the gel. The affinity matrix bound glucocorticoid receptor with relatively low affinity, so bound glucocorticoid receptor could be recovered by a solution of cortexolone. This way, the procedure afforded 10-30-fold purification when it was per-

219

1.4- Butanedioldiglycidyl

TABLE I

ether

2

Cn~o

NeON

MODIFICATION OF SEPHAROSE 4B 5 g Sepharose 4B were epoxy-activated by means of indicated amounts of 1,4-butanediol diglycidyl ether, then further modified into thioalkyl agarose. Thereafter, cortexolone-21-mesylate was coupled to this gel. The SH-content of the gel was measured before and after the coupling reaction. Yields of the coupling reaction and steroid content of the affinity gel were calculated based on the decrease in SH-content.

I

OH 25"C 6 h ~Ne2S203 OH

Pydrnidine /~CISO2CH3 0°C 2h ~ Q'LOSO2CH3

OH 1,4-Butanediol mol SH-group. 106/ reel cortexolone. 106/ diglycidyl ether ml settled gel ml settled gel

30°~C[ DTT 25

~ i l o - -

(rag)

OH

te

I

°

0*C 1Acetone 3h Triethyl-

OH

°

H~,

0.13 0.36 1.20 2.52

0.08 0.23 0.65 1.38

OH OH

H

25 75 225 675

0 OH

0 Fig. 1. Scheme for synthesis of affinity matrix. DTT, dithiothreitol.

formed batchwise, and 50-150-fold purification of glucocorticoid receptor could be achieved by column chromatography (Table II). Yields decreased markedly with washing and fell into a range of 10-30%. Optimal concentration of gel-bound steroid was in the range 1-0.3 /~mol/ml settled

TABLE II PURIFICATION OF CHICK THYMUS GLUCOCORTICOID RECEPTOR One binding site per receptor molecule was assumed A: buffer C and cortexolone were used; B: buffer C and [3H]triamcinolone acetonide were used; C: buffer B and [3H]triamcinolone acetonide were used for purification. Charcoal-dextran was suspended in buffer A rather than in buffer D. Each entry in the table is the result of one representative experiment. 10-30 purifications were performed by each method. Method

Step

Total glucocorticoid receptor (pmol)

Total protein (mg)

Purification (fold)

Yield (%)

A; column

cytosoi affinity eluate

32.4 6.1

53.850 0.095

1.0 106.7

100.0 18.8

B; column

cytosol affinity eluate

21.5 5.7

50.610 0.093

1.0 144.2

100 26.5

C; column

cytosol affinity eluate

23.1 4.5

48.190 0.122

1.0 77.0

100 19.5

A; batch

cytosol affinity eluat¢

10.3 2.1

19.800 0.186

1.0 21.3

100 19.8

B; batch

cytosol affinity eluate

9.4 3.1

20.115 0.234

1.0 28.4

100 33.0

C; batch

cytosol affinity eluate

7.7 1.8

18.150 0.155

1.0 27.3

100 23.4

220 gel. Higher degree of substitution resulted in poorer recovery, lower degree of substitution resuited in poor binding of glucocorticoid receptor (data not shown).

Heat inactivation The heat stability of the purified glucocorticoid receptor was investigated at 0 and 25°C in the presence or absence of cortexolone. It could be established that the higher purity, achieved by column chromatography, resulted in a decrease of heat stability of the purified glucocorticoid recep-

Q. o

m

tor. Cortexolone had a slight stabilizing effect, in contrast with its destabilizing effect observed in the denaturation of the cytosolic glucocorticoid receptor [14]. Denaturation of the purified receptor followed first-order kinetics determined by semilogarithmic plot of the specific binding vs. time (Fig. 2). The purified glucocorticoid receptor was more labile at 25 °C than the cytosolic glucocorticoid receptor.

Sephacryl S-300 chromatography The Stokes radius of the purified glucocorticoid receptor-[3H]triamcinolone acetonide complex was 6.2 + 0.1 nm, in agreement with the literary data for nonactivated receptor [15] and with our results with cytosolic glucocorticoid receptor[ 3H]triamcinolone acetonide complex (Fig. 3). It is noteworthy that, in contrast to the cytosolic glucocorticoid receptor, in the absence of a reducing agent, purified glucocorticoid r e c e p t o r - t r i amcinolone acetonide complex was denatured during the gel filtration procedure, indicating again that partially purified receptor was less stable than cytosolic glucocorticoid receptor. t (b)

OB

AC F

T

2-

0 Time ( min )

5o-(b)

:ot

20

oo2~20

'7,

"'. . . . 4. . . 6. . . . 8

2

Stokes Radius (nm)

10

lo-~

x

al

0 v

0

3'0

9'0

150

Time (rain)

Fig. 2. Heat inactivation of purified glucocorticoid receptor. Affinity chromatographywas done by method A batchwise or in a column. Eluates of batchwise (©, m) and column chromatography (e, [3) were incubated in the presence (©, n) or absence (e~m) of cortexolone (1.71.10-4 M) at 25°C (a) or 0°C (b). Remain'.mg specific binding was determined after various incubation times.

30

40

50

60 70 BO 90 Elution Volume (mL)

100

110

120

Fig. 3. (a) Sephacryl S-300 chromatography of the purified glucocorticoid receptor-[3H]triamcinolone acetonide complex. Affinity chromatography was performed by method B in a column. Gel filtration was done at 4-6°C. The column (2 x 33 cm) was loaded with 1.5 ml eluate ( × ) or cytosol equilibrated with [3H]triamcinolone acetonide (O). 1.5 ml fractions were collected and measured for radioactivity. (b) Calibration plot of Sephacryl S-300 column. The calibration proteins used were: thyroglobulin (T: Stokes radius 8.5 nm), ferritin (F: 6.] nm), catalase(C: 5.22 nm), aldolase (A: 4.81 nm), albumin (B: 3.55 nm) and ova]bumin (©: 3.06 nm).

221

DNA-cellulose binding The purified glucocorticoid receptor-[aH]tri amcinolone acetonide complex was activated at 25°C in the presence or absence of 0.3 M KC1 and at 0°C in the presence of 0.3 M KC1. The activated complex was subsequently bound to DNA-cellulose. In each case maximum-type binding curves were observed (Fig. 4). When the activation was performed at 25°C in the absence of 0.3 M KCI, the maximum binding was achieved after 30 min of activation similarly to the activation profile of cytosolic glucocorticoid receptor [9], but it was much lower (9%) in the case of purified glucocorticoid receptor. The low initial binding (5%) proved that purified glucocorticoid receptor was in the nonactivated form. The activation was much more effective in the presence of 0.3 M KC1 either at 25 or at 0°C. The shapes of the activation curves were similar to the activation plots of cytosolic glucocorticoid receptor [9] but the initial binding of purified glucocorticoid receptor-[3H]triamcinolone acetonide

10" 8-

6 ~4

8'0

200

400

600

800

Concentration of Non[abe[ed Steroid (nM)

Fig. 5. Association kinetics of purified glucocorticoid receptor with nonlabeled steroids. Affinity chromatography was done by method A in a colunm. Excess of cortexolone was removed by charcoal from the eluate. After 2 h of incubation at 0 ° C it was added to an equal volume of a solution of labeled triamcinolone acetonide (77 nM) and competing nonlabeled hydrocortisone (e) or 11-deoxycorticosterone ( O ) or progesterone (11) at various concentrations. After 45 min of incubation protein-bound [3H]triamcinolone acetonide was determined. Incubation was done in the presence of nonlabeled triamcinolone acetonide (3.9.10 -6 M) to determine nonspecific binding. Total amount of purified glucocorticoid receptor was determined in the absence of the studied nonlabeled steroid.

30

complex was 4-times higher than the corresponding parameter of cytosolic glucocorticoid receptor-[SH]triamcinolone acetonide complex. The

2S

-

g

20

T A B L E III

15

RATE CONSTANTS OF INTERACTION OF THE PURIFIED CHICK THYMUS GLUCOCORTICOID RECEPTOR W I T H N O N L A B E L E D STEROIDS

E r~

'

Means and standard deviations calculated from 3 - 7 independent experiments are given. Purification was performed according to either of the A methods.

10

7

0

20

40

60

80

100

120

140

Activation Time (rain)

Fig. 4. Binding of purified glucocorticoid receptor-[3HJtriamcinolone acetonide complex to DNA-ceUulose. Affinity chromatography was performed in a column using method C. Eluate was activated at 2 5 ° C ( × , O) or at 0 ° C (12) in the presence ( O , 12) or absence ( × ) of 0.3 M KCi. After various times aliquots were taken and DNA-binding was determined. [ ) N A - b o u n d radioactivity was expressed as a percentage of zero time total complex.

Steroid

k2 (105 M -1 .rain - 1 )

k_2 (10 -3 min -~)

Estradiol Testosterone Cortexolone 11-Deoxycorticosterone Progesterone 11-Epiprednisolone Hydrocortisoue Corticosterone Prednisolone

_ a - " - a

18.7+ 3.2 24.1 5:2.8 17.9+ 1.9

1.47 + 0.70 0.67 + 0.23 0.16 ± 0.11 3.61 + 1.58 2.20 + 1.1 1.40 ± 0.61

3.69 ± 0.39 2.99 ± 0.45 3.82 ± 0.50 4.02 + 1.58 2.56 5:0.24 2.90 ± 0.27

Ka (nM)

25.1 44.6 239.0 11.1 11.6 20.7

" Could not be determined because of their fast dissociation.

222

maximum binding was less, 33% and 22%, respectively, and was reached earlier in the case of purified glucocorticoid receptor.

10"

Association rate constants

The association rate constant, determined by means of [3H]triamcinolone acetonide, of purified glucocorticoid receptor for triamcinolone acetonide was lower (6.6-103 M -t .min -1) at 0°C than in the case of cytosol glucocorticoid receptor (1.9.10 6 M - 1 " m i n - 1 ) [10]. Rate constants of association of purified glucocorticoid receptor with nonlabeled steroids, determined in competition experiments (Fig. 5), were significantly lower (Table III) than the corresponding parameters of cytosol glucocorticoid receptor [10]. However, the relative order of rate constants khydrocortisone m kprednisolone kcorticosterone > kll.deoxycorticosterone kprogesterone > kepiprednisolone

was similar in the two studied systems. There was one exception: epiprednisolone exhibited a relatively low association rate for purified glucocorticoid receptor. Dissociation rate constants

Dissociation rate constants of purified glucocorticoid receptor-nonlabeled steroid complexes were determined in competition experiments as described in Materials and Methods. Semilogarithmic plots of [3H]triamcinolone acetonide vs. time gave straight lines in accordance with firstorder kinetics (Fig. 6). There was no difference between dissociation rates determined using 10fold or 100-fold purified receptor. Average rates of dissociation (Table III) were in good agreement with the corresponding data observed for cytosol receptor [10]. Discussion

To study steroid binding properties of the purified glucocorticoid receptor the requirement of an uncomplexed, purified receptor preparation was essential. Several different purification meth-

5.

a. u

m~ l-n-

0-7

~o

12o

i~o

~4~o

3bo

Time (rain)

Fig. 6. Dissociation of purified glucocorticoid receptor-ligand complexes. Affinity chromatography was done by method A in a column. Free cortexolone was removed by treatment with charcoal. The supernatant was incubated at 0 ° C in the presence of ll-deoxycorticosterone (@), hydrocortisone (11) or estradiol ( O ) at a concentration of 10 - s M. After 2 h of incubation excess steroid was removed by a second charcoal treatment and an equal volume of radioactive triamcinolone acetonide (39 nM) was added for determination of the rate constant.

ods [3,16-28] of glucocorticoid receptor are known, but at the end of these procedures one could obtain a purified glucocorticoid receptor in a form complexed with a steroid [3H]triamcinolone acetonide, [3H]dexamethasone which dissociates slowly from the receptor. To reach our goal, we had to prepare a new affinity matrix from which the partially purified glucocorticoid receptor could be recovered in a form that was still capable of binding steroidal ligands. To this end, Sepharose 4B was modified to contain mercapto functions on a long arm and cortexolone was coupled to this matrix through a thioether linkage. This method of derivatization was significantly simpler and faster than those described in the literature. The coupling reaction was mild, fast, gave high yields and good reproducibility. Another advantage of the method is that unreacted, free thiol groups do not interfere with either binding or elution in the reductive environment used. The method could be general for a-ketomesylates. In contrast to the steroid affinity gels described earlier, there is no

223 proteinase-labile linkage (ester, amide) between the ligand and the support, so the matrix could be used with crude cytosol many times without any leakage of the ligand from the gel. Our affinity gel bound the receptor protein with a relatively low affinity so that it could be recovered by the use of cortexolone, a steroid that dissociated rapidly from the binding site. Steroid specificity of the partially purified uncomplexed receptor could be studied this way. Since cortexolone is bound not only by the glucocorticoid receptor, one cannot exclude the possibility of copurification of other receptor proteins or serum proteins. However, the radiolabeled ligand used to visualize the results, [3H]triamcinolone acetonide, is very specific for glucocorticoid receptor. Therefore, no interference of binding sites other than glucocorticoid receptor should be taken into account. Purification was 10-30-fold when it was performed as a batch procedure and 75-150-fold when the affinity gel was used in a column. The yields of purification fell into a range of 10-30%. Stokes radius of the purified glucocorticoid receptor-[aH]triamcinolone acetonide complex was in accordance with data determined by S-300 chromatography for rat thymus glucocorticoid receptor [15] and with our results obtained with cytosolic chick thymus glucocorticoid receptor (Fig. 3). The purified glucocorticoid r e c e p t o r - t r i amcinolone acetonide complex could be activated by either heat or salt treatment. The shapes of the DNA-binding-activation time curves are of maximum type, similarly to those of the cytosol glucocorticoid receptor-triamcinolone acetonide complexes. However, there are striking differences, because of the higher initial binding and lower maximum binding of purified glucocorticoid receptor-triamcinolone acetonide complexes and the especially small effect of heat activation of the purified glucocorticoid receptor-triamcinolone complex. This low level of-DNA-binding capacity of heat-activated purified receptor-steroid complexes are in good accordance with the data of Grandics et al. [29]. Heat stability of the purified glucocorticoid receptor is lower than that of the cytosolic glucocorticoid receptor. However, there was no signifi-

cant loss of the steroid binding of the purified glucocorticoid receptor at 0 ° C for 3 h. Thus, major features of our partially purified preparation corresponded to those of the purified, nonactivated glucocorticoid receptor preparations [22]. Binding properties of the partially purified glucocorticoid receptor could also be studied in our system. Dissociation rate constants of purified receptor-steroid complexes were similar to those of the complexes of the corresponding steroids and the cytosolic glucocorticoid receptor. Rates of association of purified glucocorticoid receptor ligands were significantly lower than that of cytosolic glucocorticoid receptor. The relative order of association rates of purified glucocorticoid receptor for the different steroids was basically similar to those of crude receptor. On the basis of the similar steroid specificity one can claim that affinity chromatography, at least when immobilized ligand is of low affinity, does not alter significantly the steroid binding domain of the glucocorticoid receptor. Thus, partial purification by the methods reported here establishes a link between experiments on cytosolic and on highly purified glucocorticoid receptors.

Acknowledgements This work was supported by the Central Research Fund of the Hungarian Academy of Sciences. The authors are indebted to Professor I. Horv~th for his continuous interest and criticism. Thanks are due to Ms. Olga Losonczi for expert technical assistance.

References 1 Munck, A. and Leung, K. (1977) in Receptors and Mechanism of Action of Steroid Hormones(Pasqualini, J.R., ed.), Part 2, pp. 311-398, Marcel Dekker, New York 2 Baxter, J.D. and Rousseau, G.G. (1979) in Glucocorticoid Hormone Action (Baxter, J.R. and Rousseau, G.G., eds.), pp. 1-24, Springer-Verlag,Berlin 3 Failla, D., Tomkins, G.M. and Santi, V. (1975) Proc. Natl. Acad. Sci. USA 72, 3849-3852 4 Ar~nyi, P. and N&ray, A. (1980) J. Steroid Biochem. 12, 267-272 5 Grandics, P. (1981) Acta Biochim. Biophys. Acad. Sci. Hung. 16, 51-55 6 Carlsson, J., Ax6n, R. and Unge, T. (1975) Eur. J. Biochem. 59, 567-572

224 7 Brocklehurst, K., Carlsson, J., Kierstan, M.P.J. and Crock, E.M. (1973) Biochem. J. 133, 573-584 8 Baxter, J.D. and Tomkins, G.M. (1971) Proc. Natl. Acad. Sci. USA 86, 923-937 9 Ar~myi, P. (1983) Eur. J. Biochem. 129, 549-554 10 Arhnyi, P. (1982)'J. Steroid Biochem. 17, 137-141 11 Arhnyi, P. and Quiroga, v. (1980) J. Steroid Biochem. 13, 1167-1172 12 Spector, T. (1978) Anal. Biochem. 86, 142-146 13 Simons, S.S., Jr., Pons, M. and Johnson, D.F. (1980) J. Org. Chem. 15, 3084-3088 14 T6th, K. and Arhnyi, P. (1983) Biochim. Biophys. Acta 761, 196-203 15 Weatherill, P.J. and Bell, P.A. (1985) Biochim. Biophys. Acta 839, 228-232 16 Santi, d.V., Washtien, W. and Pogolotti, A.L., Jr. (1979) in Glucocorticoid Hormone Action (Baxter, J.P. and Rousseau, G.G., eds.), pp. 109-122, Springer-Verlag, Berlin 17 Colman, P.D. and Feigelson, P. (1976) Mol. Cell. Endocrinol. 5, 33-40 18 Wrange, O., Carlstedt-Duke, J. and Gustafsson, J.A. (1979) J. Biol. Chem. 254, 9284-9290

19 Eisen, H.J. and Glindsmann, W.H. (1978) Biochem. J. 171, 177b1183 20 Blanchardie, P., Lustenberger, P., Orsonneau, J.L. and Bernard, S. (1984) Bioehemie 66, 505-511 21 Climent, H., Bugany, H. and Beato, M. (1976) FEBS Lett. 66, 317-321 22 Lustemberger, P., Formstecher, P. and Dautrevaux, M. (1981) J. Steroid Biochem. 14, 697-703 23 Manz, B., Heubner, A., K6hler, I., Grill, H.J. and Pollow, K. (1983) Eur. J. Biochem. 131, 333-338 24 De Kloet, E.R. and Burbach, P. (1978) J. Neurochem. 30, 1505 - 1507 25 Westphal, H.M. and Beato, M. (1980) Eur. J. Biochem. 106, 395-403 26 Govindan, M.V. and Manz, B. (1980) Eur. J. Biochem. 108, 47-53 27 Govindan, M.V. (1979) J. Steroid Biochem. 11, 323-332 28 Govindan, M.V. and Gronemeyer, H. (1984) J. Biol. Chem. 259, 12915-12924 29 Grandics, P., Miller, A., Schmidt, T.J., Mittman, D. and Litwack, G. (1984) J. Biol. Chem. 259, 3173-3180