Biochimica et Biophysica Acta, 434 (1976) 483-489
© Elsevier ScientificPublishing Company, Amsterdam -- Printed in The Netherlands BBA 37362 CHARACTERISTICS OF THE CORTICOSTERONE-RECEPTOR COMPLEX IN RAT LIVER CYTOSOL
ORJAN WRANGE Department of Chemistry, Karolinska Instituter, S-104 Ol Stockholm 60 (Sweden)
(Received December 4th, 1975)
SUMMARY Using gel filtration on Sephadex G-150 in low ionic strength, it was possible to separate a corticosterone-binding protein in rat liver cytosol from corticosteroidbinding globulin after incubation of cytosol with [3H]corticosterone. The corticosterone-protein complex ("a-Complex") had a sedimentation coefficient of 8-9 S in low ionic strength. In high ionic strength, the a-Complex rapidly dissociated with a half-life of 15 h, compared to a half-life of 31 h tbr the hepatic dexamethasonereceptor complex under identical conditions (0 °C). The a-Complex was saturable with an excess of unlabelled corticosterone or dexamethasone and was sensitive to heat and protease digestion. It is stressed that quantitation of the corticosteronereceptor complex must include separation of the receptor from corticosteroid-binding globulin as this protein binds corticosterone with high affinity and with a saturable amount of binding sites.
INTRODUCTION In the liver, glucocorticoid hormones are specifically bound to a soluble receptor protein [1, 2, 3]. This process is an essential step in the glucocorticoid stimulation of certain liver enzyme activities. Previous investigations on glucocorticoid-receptor interactions in rat liver have mainly been carried out witx synthetic steroids, especially dexamethasone. Recent studies suggest that the physical characteristics of the glucocorticoid-receptor complex vary depending on the nature of the ligand ([2, 4], Carlstedt-Duke, J., Gustafsson, J.-A., Gustafsson, S. A. and Wrange, 0., unpublished). It cannot be excluded that the somewhat different biological effects of natural and synthetic glucocorticoids may be due to different behavior of the respective steroid-receptor complexes in the cell [4]. In view of this, it seemed important to investigate some physical characteristics of the corticosterone-receptor complex in rat liver cytosol and compare these to the corresponding characteristics of the dexamethasone-receptor complex. MATERIALS AND METHODS Preparation o f c y t o s o l . Male Sprague-Dawley rats, 8 weeks old, were adrenal-
484 ectomized 2-4 days betore the experiment and were given a pellet diet and physiological saline ad libitum. Anima!s were killed with a blow on the head and livers were perfused in situ via the portal vein and inferior vena cava with 100-200 ml ot ice-cold Buffer A (0.02 M Tris. HC1 pH 7.4/10 mM MgC12/0.1 mM dithioerythritol/ 1 0 ~ (v/v) glycerol). The livers were taken out, minced and homogenized in 14 ml of Buffer A in a teflon-glass Potter-Elvehjem homogenizer using three strokes twice with an intervening cooling period of 1 min. The liver homogenate was centrifuged at 25 000 × g for 10 rain in an MSE High-speed 18 centrifuge and the supernatant was then centrifuged at 105 000 × g for 1 h at + 2 °C in a Beckman Model L3-50 ultracentrifuge. The floating lipid layer was carefully removed and the supernatant used as cytosol. The cytosol was incubated with radioactive steroids as described below. In some experiments, incubation with radioactive steroids was performed with the 25 000 × g supernatant and cytosol was then prepared as described above. In a third type of experiment, the liver homogenate was centrifuged at 175 000 × g for 45 min in an SW 50.1 rotor and the supernatant used as cytosol. Radioactive steroids. [1,2,6,7-3H]Corticosterone (specific radioactivity, 102 Ci/ mmol) was purchased from The Radiochemical Centre, Amersham, England and [1,2,4-3H]dexamethasone (specific radioactivity, 21.5 Ci/mmol) from New England Nuclear, Boston, Mass. The 3H-labelled steroids were purified prior to use as described previously (Carlstedt-Duke, J., Gustafsson, J.-,~., Gustafsson, S. A. and Wrange, ~)., unpublished). After purification these compounds were more than 9 9 ~ pure as checked by thin-layer chromatography. 3H-labelled steroids were stored at + 2 °C in Buffer A and 1 0 ~ (v/v) ethanol. This solution was also used to add steroid to the incubation mixtures. Receptor labelling was performed at 0 °C for 2 h using a steroid concentration of 8-30 nM. No qualitative differences were observed between the steroid-receptor complexes labelled using the different methods described above. However, a somewhat higher yield of labelled receptor was obtained following incubation with the 25 000 × g supernatant and with the cytosol prepared by direct centrifugation of the crude homogenate than following incubation with the 105 000 × g supernatant. Protein binding assays. These were carried out on 10 ml Sephadex G-25 columns equilibrated with Buffer B (0.01 M Tris.HC1 pH 7.4/1 mM disodium E D T A / 1 0 ~ (v/v) glycerol). In some experiments 0.2 mg/ml of hemoglobin (beef blood, crystallized twice, dialyzed and lyophilized, Sigma) was added to the sample to simplify collection of the protein-bound fraction (void volume). Gelfiltration. This was performed on Sephadex G-100 and G-150 (Pharmacia, Uppsala, Sweden) in columns with an inner diameter of 2.5 cm and a total volume of 300-350 ml. The columns were equiliblated with Buffer C (0.02 M Tris-HC1 pH 7.4/1 m M disodium E D T A / 1 0 ~ (v/v) glycerol/0.1 mM dithioerythritol, with or without 0.4 M KC1). Fractions, 3.5 ml, were collected using a flow rate of 2.5 ml/cm 2. h. Protein was quantitated according to the method of Lowry et al. [6]. Radioactivity was determined in a Packard Tri-Carb Model 2425 Liquid Scintillation spectrometer. Instagel ® (Packard Instrument Co., Inc., Warrenville, Downers Grove, Ill.) was used as scintillator fluid. Correction for quenching was made using the external standard technique.
485 RESULTS AND DISCUSSION When rat liver cytosol is labelled in vitro with 3H-corticosterone and is subsequently chromatographed on Sephadex G-100 in conditions of high ionic strength (0.4 M KCI), two labelled protein complexes are eluted (Carlstedt-Duke, J., Gustafsson, J.-.~., Gustafsson, S. A. and Wrange, ~)., unpublished). The first complex ("a-Complex"), which is eluted close to the exclusion volume, dissociates rapidly in high ionic strength; the protein has been characterized as a glucocorticoid receptor. The protein participating in the second complex ("/%Complex") has been identified as corticosteroid-binding globulin (Carlstedt-Duke, J., Gustafsson, J.-/~., Gustafsson, S. A. and Wrange, O., unpublished). The characterization of the a-Complex has been hampered by the lability of the complex in high ionic strength. When chromatographed on Sephadex G-150 in low ionic strength, however, the a-Complex was considerably more stable; it was eluted in the exclusion volume and was well separated from the E-Complex (Fig. 1). The a-Complex was present in an amount of about 10 fmol per mg of cytosol protein assuming one ligand binding site per receptor molecule. This amount was not changed even after extensive perfusion of the liver, whereas the amount of E-Complex varied between 0.3 and 180 fmol per mg protein depending on the extent of perfusion. Competition experiments showed that unlabelled corticosterone competed with [3H]corticosterone for the binding sites of both the a- and/~-proteins whereas unlabelled dexamethasone was capable of competing with [3H]corticosterone for the a-protein binding site only (Fig. 1). The protein nature of the a-Complex was indicated by its sensitivity to protease as compared to nucleases and by its low resistance to heat (Table I). The relatively high heat stability of the E-Complex has been described before (Carlstedt-Duke, J., Gustafsson, J.-•., Gustafsson, S. A. and Wrange, O., unpublished). Sedimentation analysis on glycerol gradients showed that the E-Complex sedimented at 4 S independent of ionic strength. The a-Complex sedimented at 8-9 S in low ionic strength (Fig. 3). When crude cytosol labelled in vitro with [3H]corticosterone was analyzed on glycerol gradients, it was not possible to detect any 8-9 S complex, probably due to insufficient separation of this complex from the/~-Complex sedimenting at 4 S. Treatment of the labelled cytosol with dextran-coated charcoal according to the method of Beato et al. [2] only resulted in partial removal of the /~-Complex from the cytosol. When the a-Complex was analyzed on glycerol gradients in conditions of high ionic strength, the radioactivity dissociated rapidly from the protein and was found in the upper part of the gradient. Even in low ionic strength, the a-Complex was found to be relatively labile and it was essential to perform the sedimentation analysis immediately after the gel filtration. Attempts to stabilize the a-Complex by the addition of [3H]corticosterone to the gradients, by changing the pH and the content of divalent ions or by adding thiol-protecting agents were unsuccessful. In some experiments the characteristics of the a-Complex were compared to those of the dexamethasone-receptor complex. Cytosol labelled in vitro with [3H]dexamethasone was chromatographed on Sephadex G-150 in low ionic strength. The protein-bound radioactivity was found in the exclusion volume. This complex was analyzed by glycerol gradient centrifugation and was found to sediment at 7-8 S
486
7.5
--
5.0-
FI u
>
zs
a
20
ab
4b
FRACTION
2Sl
o
i
- io -
2'0
3'0 ~b FRACTION
s'o
s'o
7'o
-
Fig. 1. Chromatography on Sephadex G-150 (lower ionic strength) of rat liver cytosol incubated with [3H]corticosterone alone (O---O) or in the presence of a 1000-fold excess of unlabelled corticosterone ( O - - O , Fig. 1A) or dexamethasone (O---O, Fig. 1B). V0 = void volume.
487 TABLE I E N Z Y M A T I C D I G E S T I O N A N D HEAT I N A C T I V A T I O N OF THE a-COMPLEX (a) The preparation of a-Complex was incubated with the following enzymes in a total volume of 1 ml of Buffer C (low ionic strength): DNAase (220 pg in 5 mM MgCl~; Deoxyribonuclease I from bovine pancreas, RNAase-free, Sigma), RNAase (220/tg; Ribonuclease " A " from bovine pancreas, protease-frce, Sigma) and Protease (1.7 rag; from Streptomyces griseus, Type VI, Sigma). Incubation was performed at 0 °C for 10 h. Protein binding was assayed by gel filtration on Sephadex G-25. In the heat inactivation experiments (b) the preparation of a-Complex in Buffer C (low ionic strength) was incubated for 5 rain at the indicated temperatures. Protein binding was assayed by gel filtration on SephadexG-25.
(a) Enzyme treatment Control Control, in 5 m M MgClz DNAase RNAase Protease
Protein binding as ~ of control 100 100 I 11 49 8
(b) Temperature (°C) 0 (Control) 25 40 55
Protein binding as ~ of control 100 85 15 9
ADH
200 c
.o
}. _~100
Botto l'r~
10
FRACTION
2'0
r
Fig. 2. Glycerol gradient centrifugation of a-Complex in low ionic strength. The analysis was performed in a 12-25 ~ glycerol gradient in Buffer C at 220 000 × g for 7 h (SW 50.1 rotor). The standard protein (yesat alcohol dehydrogenase, Sigma) was centrifuged in a separate tube. Yeast alcohol dehydrogenase (sedimentation coefficient 7.4 S according to Siegel and Monty [7]) was assayed according to the method of Dalziel [8]. S values were calculated as described by Martin and Ames [11].
488 Hb
OAM KCL ~, NO SALT
00]
~
l
A
B
/
SO0,
Catalase
400.
E~400
>. 300.
>. 300
p-
F-
u
200
200
O CI
Q
100.
01, Bottom
Cyt C
~
100
lb 2b FRACTION
o Bot orn
;b 2b FRACTION
Fig. 3. Glycerol gradient centrifugation of [3H]dexamethasone-labelled cytosol after chromatography on Sephadex G-150 in low ionic strength conditions. The analysis was performed in low ionic strength conditions at 125 000 × g for 16 h (Fig. 3A) or in high ionic strength conditions at 220 000 × g for 17 h (Fig. 3B), using an SW 50.1 rotor. The standard proteins used were horse liver catalase (11.2 S; see ref. 9) and cytochrome c (1.9 S; see ref. 7), both from Sigma, and hemoglobin (4.1 S; see ref. 9), taken fresh from human red blood cells lysed in hypotonic medium and diluted with Buffer C. Standards were run in separate tubes and were determined as previously described.
in low ionic strength and at 3-4 S in high ionic strength (Fig. 3). The dexamethasonereceptor complex showed a considerably greater stability towards dissociation than the a-Complex, both in high and low ionic strengths. The half-lives of the a-Complex at 0 °C were found to be 35 h and 15 h in low and high ionic strengths, respectively, whereas the corresponding figures for the dexamethasone-receptor complex were 46 h and 31 h (Fig. 4). Identical results were obtained whether or not an excess of unlabelled steroid is added at time zero to the buffer containing the labelled steroidreceptor complex. In conclusion, the present investigation has demonstrated the pronounced lability of the hepatic corticosterone-receptor complex, especially in buffers with high ionic strength. This finding explains the often documented difficulties involved in studies on interactions between natural corticosteroids and the hepatic glucocorticoid receptor ([2-4, 7], Carlstedt-Duke, J., Gustafsson, J.-A., Gustafsson, S. A. and Wrange, 0., unpublished). The competition experiments indicate that corticosterone and dexamethasone interact with the same binding site on the receptor. It may be suggested that electrostatic forces are of some importance Jn the interaction between the ligand and the receptor: being less polar than dexamethasone, corticosterone interacts with the protein with weaker electrostatic bonds and the corticosteronereceptor complex is therefore more affected by an increase in the ionic strength. Another important implication of the present work is that characterization
489
CORTICOSTERONE
A
DEXAMETHASONE
I00:
~
100',
o=
w
50'
50'
:46h
=35h 0 u
"-..t :31h
\ %
F
B
100
2'0
",t 1/2=15h
4"0 TIME
hours
I00
20
40
hour~"
TIME
Fig. 4. Dissociation of corticosterone-receptor complex (a-Complex) (Fig. 4A) and dexamethasonereceptor complex (Fig. 4B) in high (O---Q) and low (0---0) ionic strength. Cytosol labelled in vitro with [3H]corticosterone or [3Hldexamethasone was chromatographed on Sephadex G-150 in low ionic strength. Half of the a-Complex and the dexamethasone-receptor complex so obtained were made 0.4 M with respect to KCI. The resulting four samples were kept on ice, aliquots were taken off at indicated times and the amount of protein-bound radioactivity was determined by gel filtration on Sephadex G-25. The protein-bound radioactivity at the various times was plotted as per cent of the protein-bound radioactivity at time 0 in a semilogarithmic plot and the half-life of the steroid-receptor complex was determined from the plot assuming that the dissociation of the complex followed firstorder kinetics. a n d q u a n t i t a t i o n of c o r t i c o s t e r o n e - b i n d i n g sites in terms of the glucocorticoid receptor m u s t include some m e t h o d of separating the receptor from corticosteroid-binding globulin. The lability of the a - C o m p l e x a n d the insufficient r e m o v a l of the corticosterone complex by dextran-coated charcoal t r e a t m e n t may explain earlier reports [2, 4] o n different b e h a v i o r of d e x a m e t h a s o n e - a n d cortisol-receptor complexes on sucrose gradients in high a n d low ionic strengths. ACKNOWLEDGEMENTS This work was supported by grants from the Swedish Medical Research Council No. (03X-2819) a n d from LEO Research F o u n d a t i o n . 1 a m greatly indebted to Drs. Jan Carlstedt-Duke and J a n - A k e Gustafsson for valuable discussions d u r i n g the course of this study. REFERENCES 1 Beato, M., Bieswig, D., Braendle, W. and Sekeris, C. E. (1969) Biochim. Biophys. Acta 192, 494-507 2 Beato, M. and Feigelson, P. (1972) J. Biol. Chem. 247, 7890-7896 3 Koblinsky, M., Beato, M., Kalimi, M. and Feigelson, F. (1972) J. Biol. Chem. 247, 7898-7904 4 Giannopoulos, G. (1973) J. Biol. Chem. 248, 3876-3883 5 Reference deleted. 6 Lowry, O. H., Rosebrough, N. J., Farr, A. L. and Randall, R. J. (1951) J. Biol. Chem. 193, 265-275 7 Litwack, G., Filler, R., Rosenfield, S. A., Lichtasch, N., Wishman, C. A. and Singer, S. (1973) J. Biol. Chem. 248, 7481-7486 8 Siegel, L. M. and Monty, K. I. (1966) Biochim. Biophys. Acta 112, 346-362 9 Dalziel, K. (1957) Acta Chem. Stand. 11, 397-398 10 Handbook of Biochemistry, Selected Data for Molecular Biology (1970) (Sober, H. A., ed.), p. C-15, 2nd edn., The Chemical Rubber Co., Cleveland, Ohio I1 Martin, R. G. and Ames, B. N. (1961) J. Biol. Chem. 236, 1372-1379
© Elsevier ScientificPublishing Company, Amsterdam -- Printed in The Netherlands BBA 37362 CHARACTERISTICS OF THE CORTICOSTERONE-RECEPTOR COMPLEX IN RAT LIVER CYTOSOL
ORJAN WRANGE Department of Chemistry, Karolinska Instituter, S-104 Ol Stockholm 60 (Sweden)
(Received December 4th, 1975)
SUMMARY Using gel filtration on Sephadex G-150 in low ionic strength, it was possible to separate a corticosterone-binding protein in rat liver cytosol from corticosteroidbinding globulin after incubation of cytosol with [3H]corticosterone. The corticosterone-protein complex ("a-Complex") had a sedimentation coefficient of 8-9 S in low ionic strength. In high ionic strength, the a-Complex rapidly dissociated with a half-life of 15 h, compared to a half-life of 31 h tbr the hepatic dexamethasonereceptor complex under identical conditions (0 °C). The a-Complex was saturable with an excess of unlabelled corticosterone or dexamethasone and was sensitive to heat and protease digestion. It is stressed that quantitation of the corticosteronereceptor complex must include separation of the receptor from corticosteroid-binding globulin as this protein binds corticosterone with high affinity and with a saturable amount of binding sites.
INTRODUCTION In the liver, glucocorticoid hormones are specifically bound to a soluble receptor protein [1, 2, 3]. This process is an essential step in the glucocorticoid stimulation of certain liver enzyme activities. Previous investigations on glucocorticoid-receptor interactions in rat liver have mainly been carried out witx synthetic steroids, especially dexamethasone. Recent studies suggest that the physical characteristics of the glucocorticoid-receptor complex vary depending on the nature of the ligand ([2, 4], Carlstedt-Duke, J., Gustafsson, J.-A., Gustafsson, S. A. and Wrange, 0., unpublished). It cannot be excluded that the somewhat different biological effects of natural and synthetic glucocorticoids may be due to different behavior of the respective steroid-receptor complexes in the cell [4]. In view of this, it seemed important to investigate some physical characteristics of the corticosterone-receptor complex in rat liver cytosol and compare these to the corresponding characteristics of the dexamethasone-receptor complex. MATERIALS AND METHODS Preparation o f c y t o s o l . Male Sprague-Dawley rats, 8 weeks old, were adrenal-
484 ectomized 2-4 days betore the experiment and were given a pellet diet and physiological saline ad libitum. Anima!s were killed with a blow on the head and livers were perfused in situ via the portal vein and inferior vena cava with 100-200 ml ot ice-cold Buffer A (0.02 M Tris. HC1 pH 7.4/10 mM MgC12/0.1 mM dithioerythritol/ 1 0 ~ (v/v) glycerol). The livers were taken out, minced and homogenized in 14 ml of Buffer A in a teflon-glass Potter-Elvehjem homogenizer using three strokes twice with an intervening cooling period of 1 min. The liver homogenate was centrifuged at 25 000 × g for 10 rain in an MSE High-speed 18 centrifuge and the supernatant was then centrifuged at 105 000 × g for 1 h at + 2 °C in a Beckman Model L3-50 ultracentrifuge. The floating lipid layer was carefully removed and the supernatant used as cytosol. The cytosol was incubated with radioactive steroids as described below. In some experiments, incubation with radioactive steroids was performed with the 25 000 × g supernatant and cytosol was then prepared as described above. In a third type of experiment, the liver homogenate was centrifuged at 175 000 × g for 45 min in an SW 50.1 rotor and the supernatant used as cytosol. Radioactive steroids. [1,2,6,7-3H]Corticosterone (specific radioactivity, 102 Ci/ mmol) was purchased from The Radiochemical Centre, Amersham, England and [1,2,4-3H]dexamethasone (specific radioactivity, 21.5 Ci/mmol) from New England Nuclear, Boston, Mass. The 3H-labelled steroids were purified prior to use as described previously (Carlstedt-Duke, J., Gustafsson, J.-,~., Gustafsson, S. A. and Wrange, ~)., unpublished). After purification these compounds were more than 9 9 ~ pure as checked by thin-layer chromatography. 3H-labelled steroids were stored at + 2 °C in Buffer A and 1 0 ~ (v/v) ethanol. This solution was also used to add steroid to the incubation mixtures. Receptor labelling was performed at 0 °C for 2 h using a steroid concentration of 8-30 nM. No qualitative differences were observed between the steroid-receptor complexes labelled using the different methods described above. However, a somewhat higher yield of labelled receptor was obtained following incubation with the 25 000 × g supernatant and with the cytosol prepared by direct centrifugation of the crude homogenate than following incubation with the 105 000 × g supernatant. Protein binding assays. These were carried out on 10 ml Sephadex G-25 columns equilibrated with Buffer B (0.01 M Tris.HC1 pH 7.4/1 mM disodium E D T A / 1 0 ~ (v/v) glycerol). In some experiments 0.2 mg/ml of hemoglobin (beef blood, crystallized twice, dialyzed and lyophilized, Sigma) was added to the sample to simplify collection of the protein-bound fraction (void volume). Gelfiltration. This was performed on Sephadex G-100 and G-150 (Pharmacia, Uppsala, Sweden) in columns with an inner diameter of 2.5 cm and a total volume of 300-350 ml. The columns were equiliblated with Buffer C (0.02 M Tris-HC1 pH 7.4/1 m M disodium E D T A / 1 0 ~ (v/v) glycerol/0.1 mM dithioerythritol, with or without 0.4 M KC1). Fractions, 3.5 ml, were collected using a flow rate of 2.5 ml/cm 2. h. Protein was quantitated according to the method of Lowry et al. [6]. Radioactivity was determined in a Packard Tri-Carb Model 2425 Liquid Scintillation spectrometer. Instagel ® (Packard Instrument Co., Inc., Warrenville, Downers Grove, Ill.) was used as scintillator fluid. Correction for quenching was made using the external standard technique.
485 RESULTS AND DISCUSSION When rat liver cytosol is labelled in vitro with 3H-corticosterone and is subsequently chromatographed on Sephadex G-100 in conditions of high ionic strength (0.4 M KCI), two labelled protein complexes are eluted (Carlstedt-Duke, J., Gustafsson, J.-.~., Gustafsson, S. A. and Wrange, ~)., unpublished). The first complex ("a-Complex"), which is eluted close to the exclusion volume, dissociates rapidly in high ionic strength; the protein has been characterized as a glucocorticoid receptor. The protein participating in the second complex ("/%Complex") has been identified as corticosteroid-binding globulin (Carlstedt-Duke, J., Gustafsson, J.-/~., Gustafsson, S. A. and Wrange, O., unpublished). The characterization of the a-Complex has been hampered by the lability of the complex in high ionic strength. When chromatographed on Sephadex G-150 in low ionic strength, however, the a-Complex was considerably more stable; it was eluted in the exclusion volume and was well separated from the E-Complex (Fig. 1). The a-Complex was present in an amount of about 10 fmol per mg of cytosol protein assuming one ligand binding site per receptor molecule. This amount was not changed even after extensive perfusion of the liver, whereas the amount of E-Complex varied between 0.3 and 180 fmol per mg protein depending on the extent of perfusion. Competition experiments showed that unlabelled corticosterone competed with [3H]corticosterone for the binding sites of both the a- and/~-proteins whereas unlabelled dexamethasone was capable of competing with [3H]corticosterone for the a-protein binding site only (Fig. 1). The protein nature of the a-Complex was indicated by its sensitivity to protease as compared to nucleases and by its low resistance to heat (Table I). The relatively high heat stability of the E-Complex has been described before (Carlstedt-Duke, J., Gustafsson, J.-•., Gustafsson, S. A. and Wrange, O., unpublished). Sedimentation analysis on glycerol gradients showed that the E-Complex sedimented at 4 S independent of ionic strength. The a-Complex sedimented at 8-9 S in low ionic strength (Fig. 3). When crude cytosol labelled in vitro with [3H]corticosterone was analyzed on glycerol gradients, it was not possible to detect any 8-9 S complex, probably due to insufficient separation of this complex from the/~-Complex sedimenting at 4 S. Treatment of the labelled cytosol with dextran-coated charcoal according to the method of Beato et al. [2] only resulted in partial removal of the /~-Complex from the cytosol. When the a-Complex was analyzed on glycerol gradients in conditions of high ionic strength, the radioactivity dissociated rapidly from the protein and was found in the upper part of the gradient. Even in low ionic strength, the a-Complex was found to be relatively labile and it was essential to perform the sedimentation analysis immediately after the gel filtration. Attempts to stabilize the a-Complex by the addition of [3H]corticosterone to the gradients, by changing the pH and the content of divalent ions or by adding thiol-protecting agents were unsuccessful. In some experiments the characteristics of the a-Complex were compared to those of the dexamethasone-receptor complex. Cytosol labelled in vitro with [3H]dexamethasone was chromatographed on Sephadex G-150 in low ionic strength. The protein-bound radioactivity was found in the exclusion volume. This complex was analyzed by glycerol gradient centrifugation and was found to sediment at 7-8 S
486
7.5
--
5.0-
FI u
>
zs
a
20
ab
4b
FRACTION
2Sl
o
i
- io -
2'0
3'0 ~b FRACTION
s'o
s'o
7'o
-
Fig. 1. Chromatography on Sephadex G-150 (lower ionic strength) of rat liver cytosol incubated with [3H]corticosterone alone (O---O) or in the presence of a 1000-fold excess of unlabelled corticosterone ( O - - O , Fig. 1A) or dexamethasone (O---O, Fig. 1B). V0 = void volume.
487 TABLE I E N Z Y M A T I C D I G E S T I O N A N D HEAT I N A C T I V A T I O N OF THE a-COMPLEX (a) The preparation of a-Complex was incubated with the following enzymes in a total volume of 1 ml of Buffer C (low ionic strength): DNAase (220 pg in 5 mM MgCl~; Deoxyribonuclease I from bovine pancreas, RNAase-free, Sigma), RNAase (220/tg; Ribonuclease " A " from bovine pancreas, protease-frce, Sigma) and Protease (1.7 rag; from Streptomyces griseus, Type VI, Sigma). Incubation was performed at 0 °C for 10 h. Protein binding was assayed by gel filtration on Sephadex G-25. In the heat inactivation experiments (b) the preparation of a-Complex in Buffer C (low ionic strength) was incubated for 5 rain at the indicated temperatures. Protein binding was assayed by gel filtration on SephadexG-25.
(a) Enzyme treatment Control Control, in 5 m M MgClz DNAase RNAase Protease
Protein binding as ~ of control 100 100 I 11 49 8
(b) Temperature (°C) 0 (Control) 25 40 55
Protein binding as ~ of control 100 85 15 9
ADH
200 c
.o
}. _~100
Botto l'r~
10
FRACTION
2'0
r
Fig. 2. Glycerol gradient centrifugation of a-Complex in low ionic strength. The analysis was performed in a 12-25 ~ glycerol gradient in Buffer C at 220 000 × g for 7 h (SW 50.1 rotor). The standard protein (yesat alcohol dehydrogenase, Sigma) was centrifuged in a separate tube. Yeast alcohol dehydrogenase (sedimentation coefficient 7.4 S according to Siegel and Monty [7]) was assayed according to the method of Dalziel [8]. S values were calculated as described by Martin and Ames [11].
488 Hb
OAM KCL ~, NO SALT
00]
~
l
A
B
/
SO0,
Catalase
400.
E~400
>. 300.
>. 300
p-
F-
u
200
200
O CI
Q
100.
01, Bottom
Cyt C
~
100
lb 2b FRACTION
o Bot orn
;b 2b FRACTION
Fig. 3. Glycerol gradient centrifugation of [3H]dexamethasone-labelled cytosol after chromatography on Sephadex G-150 in low ionic strength conditions. The analysis was performed in low ionic strength conditions at 125 000 × g for 16 h (Fig. 3A) or in high ionic strength conditions at 220 000 × g for 17 h (Fig. 3B), using an SW 50.1 rotor. The standard proteins used were horse liver catalase (11.2 S; see ref. 9) and cytochrome c (1.9 S; see ref. 7), both from Sigma, and hemoglobin (4.1 S; see ref. 9), taken fresh from human red blood cells lysed in hypotonic medium and diluted with Buffer C. Standards were run in separate tubes and were determined as previously described.
in low ionic strength and at 3-4 S in high ionic strength (Fig. 3). The dexamethasonereceptor complex showed a considerably greater stability towards dissociation than the a-Complex, both in high and low ionic strengths. The half-lives of the a-Complex at 0 °C were found to be 35 h and 15 h in low and high ionic strengths, respectively, whereas the corresponding figures for the dexamethasone-receptor complex were 46 h and 31 h (Fig. 4). Identical results were obtained whether or not an excess of unlabelled steroid is added at time zero to the buffer containing the labelled steroidreceptor complex. In conclusion, the present investigation has demonstrated the pronounced lability of the hepatic corticosterone-receptor complex, especially in buffers with high ionic strength. This finding explains the often documented difficulties involved in studies on interactions between natural corticosteroids and the hepatic glucocorticoid receptor ([2-4, 7], Carlstedt-Duke, J., Gustafsson, J.-A., Gustafsson, S. A. and Wrange, 0., unpublished). The competition experiments indicate that corticosterone and dexamethasone interact with the same binding site on the receptor. It may be suggested that electrostatic forces are of some importance Jn the interaction between the ligand and the receptor: being less polar than dexamethasone, corticosterone interacts with the protein with weaker electrostatic bonds and the corticosteronereceptor complex is therefore more affected by an increase in the ionic strength. Another important implication of the present work is that characterization
489
CORTICOSTERONE
A
DEXAMETHASONE
I00:
~
100',
o=
w
50'
50'
:46h
=35h 0 u
"-..t :31h
\ %
F
B
100
2'0
",t 1/2=15h
4"0 TIME
hours
I00
20
40
hour~"
TIME
Fig. 4. Dissociation of corticosterone-receptor complex (a-Complex) (Fig. 4A) and dexamethasonereceptor complex (Fig. 4B) in high (O---Q) and low (0---0) ionic strength. Cytosol labelled in vitro with [3H]corticosterone or [3Hldexamethasone was chromatographed on Sephadex G-150 in low ionic strength. Half of the a-Complex and the dexamethasone-receptor complex so obtained were made 0.4 M with respect to KCI. The resulting four samples were kept on ice, aliquots were taken off at indicated times and the amount of protein-bound radioactivity was determined by gel filtration on Sephadex G-25. The protein-bound radioactivity at the various times was plotted as per cent of the protein-bound radioactivity at time 0 in a semilogarithmic plot and the half-life of the steroid-receptor complex was determined from the plot assuming that the dissociation of the complex followed firstorder kinetics. a n d q u a n t i t a t i o n of c o r t i c o s t e r o n e - b i n d i n g sites in terms of the glucocorticoid receptor m u s t include some m e t h o d of separating the receptor from corticosteroid-binding globulin. The lability of the a - C o m p l e x a n d the insufficient r e m o v a l of the corticosterone complex by dextran-coated charcoal t r e a t m e n t may explain earlier reports [2, 4] o n different b e h a v i o r of d e x a m e t h a s o n e - a n d cortisol-receptor complexes on sucrose gradients in high a n d low ionic strengths. ACKNOWLEDGEMENTS This work was supported by grants from the Swedish Medical Research Council No. (03X-2819) a n d from LEO Research F o u n d a t i o n . 1 a m greatly indebted to Drs. Jan Carlstedt-Duke and J a n - A k e Gustafsson for valuable discussions d u r i n g the course of this study. REFERENCES 1 Beato, M., Bieswig, D., Braendle, W. and Sekeris, C. E. (1969) Biochim. Biophys. Acta 192, 494-507 2 Beato, M. and Feigelson, P. (1972) J. Biol. Chem. 247, 7890-7896 3 Koblinsky, M., Beato, M., Kalimi, M. and Feigelson, F. (1972) J. Biol. Chem. 247, 7898-7904 4 Giannopoulos, G. (1973) J. Biol. Chem. 248, 3876-3883 5 Reference deleted. 6 Lowry, O. H., Rosebrough, N. J., Farr, A. L. and Randall, R. J. (1951) J. Biol. Chem. 193, 265-275 7 Litwack, G., Filler, R., Rosenfield, S. A., Lichtasch, N., Wishman, C. A. and Singer, S. (1973) J. Biol. Chem. 248, 7481-7486 8 Siegel, L. M. and Monty, K. I. (1966) Biochim. Biophys. Acta 112, 346-362 9 Dalziel, K. (1957) Acta Chem. Stand. 11, 397-398 10 Handbook of Biochemistry, Selected Data for Molecular Biology (1970) (Sober, H. A., ed.), p. C-15, 2nd edn., The Chemical Rubber Co., Cleveland, Ohio I1 Martin, R. G. and Ames, B. N. (1961) J. Biol. Chem. 236, 1372-1379
