Specific Binding Sites for Natural Glucocorticoids in Plasma Membranes of Rat Liver1 TAKASHI SUYEMITSU AND HIROSHI TERAYAMA Zoological Institute, Faculty of Science, University of Tokyo, Tokyo, Japan x 10"9M at OC) in PM,. The discrepancy in the cortisol binding parameters obtained by the two different methods seems to be due mainly to the lability of some binding sites, especially in PM,. The glucocorticoid-binding sites in the plasma membranes of rat liver appear to have the highest affinity of corticosterone, followed by cortisol and cortisone. A synthetic glucocorticoid [3H]dexamethasone, did not show any specific binding to the liver plasma membranes. Neither dexamethasone nor nonglucocorticoids such as estradiol given simultaneously affected [3H]cortisol binding to the plasma membranes. (Endocrinology 96: 1499, 1975)

ABSTRACT. The presence of sites specifically binding natural glucocorticoids in plasma membrane (PM) preparations (PM0> density = 1.13 - 1.16; PM,, density = 1.16 - 1.18) from rat liver was elucidated by equilibrium dialysis as well as by centrifugal methods. Equilibrium dialysis showed the presence of binding sites having a higher affinity for [3H]cortisol (Kd = 1.4 x 10"9M at 4 C) in PM0, and that of the binding sites having a lower affinity for [3H]cortisol (Kd = 1.3 x 10"8M at 4 C) in PM,, while centrifugal analysis showed the presence of higher affinity binding sites (Kd = 1.5 - 1.9 x 10~9M at 0 C) in both PM0 and PM,, and also of intermediate affinity binding sites (Kd = 4.1

H

ORMONES seem to exhibit their biological effects through binding to specific sites or receptors in the target cells. For glucocorticoids, as for other steroid hormones, the presence of soluble receptors in the cytosol as well as in the nucleus has been well documented (1-3). However, specific binding sites for steroid hormones in plasma membranes have scarcely been investigated, in contrast to many reports devoted to the membraneassociated receptors for nonsteroid hormones such as epinephrine, glucagon and so on (4-6). In the present paper we have investigated the binding of [3H]cortisol and [3H]-dexamethasone2 to plasma membrane preparations isolated from the liver of adReceived August 6, 1974. 1 This work was supported in part by the Grant-inAid for Scientific Researches from the Ministry of Education, Japan and by a Ford Foundation Grant, No. 740-0403. 2 The chemical names of non-standard steroids used in the present paper are as follows; Prednisolone (11/8, 17, 21-trihydroxypregna-l, 4-diene-3, 20-dione), Dexamethasone (9a-fluoro-ll/3,17,21-trihydroxy16a-methyl-pregna-l,4-diene-3,20-dione), betamethasone (9a-fluoro-11/3,17,21-trihydroxy-16/3methyl-pregna-l,4-diene-3,20-dione) and triamcinolone (9a-fluoro-ll/3,16a,17,21-tetrahydroxypregnal,4-diene-3,20-dione).

renalectomized rats, and found the presence of some specific binding sites for the natural glucocorticoids in the plasma membranes. Materials and Methods Plasma membranes (PM) were prepared from the liver of male Wistar rats adrenalectomized 3 days in advance according to the method of Berman et al. (7) with a slight modification by Yamamoto et al. (8). PM fractions separated from the crude nuclear fraction at the interface between density (d) = 1.13 and 1.16 sucrose layers as well as at the interface between d = 1.16 and 1.18 sucrose layers were referred to as PM0 and PMX respectively. The cytosol and microsome fractions were prepared from 30-50% tissue (liver, kidney and spleen) homogenates in cold Tris-sucrose-salt buffer (TSS buffer) consisting of 0.25M sucrose, 25 mM KC1, 10 mM MgSO4) and 50 mM Tris-HCl buffer, pH 7.55, by a conventional differential centrifugation method. The supernatant obtained by centrifuging the homogenate at 20,000 x g for 20 min was recentrifuged at 105,000 x g for 90 min to separate the cytosol (supernatant) from the microsomes (precipitates), which were washed once with the TSS buffer. [3H]Cortisol (cortisol-l,2-3H; 42.0-51.7 Ci/ mmol) and [ 3 H]dexamethasone (dexamethasone-l,2- 3 H; 25 Ci/mmol) were purchased

1499

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 23:53 For personal use only. No other uses without permission. . All rights reserved.

1500

Kudo • 1975 Vol96 • No 6

SUYEMITSU AND TERAYAM/v

from New England Nuclear Corp., U.S.A. and Radiochemical Centre, G.B., respectively. Nonlabelled steroids such as cortisol (Wako Pure Chem. Ind., Japan), 17/3-estradiol (Sigma Chem. Co., U.S.A.), progesterone (Tokyo Chem. Ind., Japan) and testosterone (Schering AG-Berlin, Germany) were of reagent grade. The other steroids such as corticosterone, cortisone, 11-deoxycorticosterone acetate, prednisolone and dexamethasone were from the Research Laboratory of the Shionogi Pharm. Ind., Japan, and were chromatographically pure. For equilibrium dialysis, a mixture of 0.5 ml each of 3H-cortisol in at least 4 different concentrations in a range of 0.5-20 x 10~9 M in the TSS buffer and a subcellular component dissolved or suspended in the TSS buffer was put into cellophane tubing (Visking; inflated diameter of 6.4 mm) and dialyzed with gentle shaking at 4 C against 5 ml of the TSS buffer containing [3H]cortisol at the same concentration as that in the tubing. The outer liquid was replaced once with fresh buffer containing [3H]cortisol 20-24 h later in order to maintain the free [3H]cortisol concentration nearly constant, and dialysis was continued for another 24 h period under the same conditions. After dialysis, aliquots of the inner and outer liquids were assayed for the radioactivity in a liquid scintillation counter according to the method of Bottoms et al. (9). At the same time another aliquot of the inner liquid was assayed for protein concentration (g/liter) according to the method of Lowry et al. (10) with bovine serum albumin as standard. Dissociation constant, Kd (M), and number of cortisol-binding sites, N (nmol per g protein), were obtained graphically by the least squares method from several experimental poimrs in the double reciprocal plots (Lineweaver-Burk) according to the following equation:

x g for 20 min. The difference between the [3H]cortisol level in the initial mixture and the [3H]cortisol level in the supernatant after centrifugation was assumed to correspond to the level of [3H]cortisol bound to the plasma membranes, while the [3H]cortisol level in the supernatant was assumed to correspond to the free cortisol concentration. Kd and N were obtained in the same way as described in the case of equilibrium dialysis. Results

Results of preliminary experiments have indicated that dialysis equilibrium seems to be achieved 12-24 h after the onset of dialysis under the present conditions, as evidenced in Fig. 1. In the following experiments dialysis was continued for another 24 h period after the renewal of the outer liquid as described in Materials and Methods. The [3H]cortisol binding parameters obtained by equilibrium dialysis analyses carried out with the subcellular components (cytosol, microsome and plasma membrane fractions) of some tissues such as liver, kidney and spleen of the rat are summarized in Table 1. Figure 2 illustrates some of the double reciprocal plots related to the binding of [3H]cortisol to plasma membranes (Fig. 2a), cytosol (Fig. 2b) and microsomes (Fig. 2c) prepared from the liver of rats. Two types of [3H]cortisol binding sites appear to exist in the plasma membranes of rat liver. The higher affinity binding sites with Kd of about 1.4 x 10~9 M seem to be associated mainly with PM0, and the lower [Bound cortisol]"1 x [Protein] affinity ones with Kd of about 13 x 10~9 M appear to be associated mainly with PMj. = Kd x N"1 x [Free cortisol]"1 + N"1 The rat liver cytosol was also found to For plasma membrane preparations isolated contain higher affinity binders with Kd of from the rat liver, the [3H]cortisol binding 1.6 x 10~9 M in accordance with the results parameters were also measured by the cen- obtained by Beato et al. (11). The binding trifugal method which is much quicker than the of [3H]cortisol to rat serum (probably to equilibrium dialysis method. A mixture of 1 ml serum transcortin) was found to show a each of the plasma membrane suspension in slightly larger Kd value as compared with TSS buffer and TSS buffer containing various those detected for binding of [3H]cortisol to concentrations of [3H]cortisol was incubated at PM0 and cytosol. The number of cortisol0 C (in ice water) for 2 h unless otherwise binding sites in PM0 and PMj of rat liver specified, followed by centrifugation at 17,000

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 23:53 For personal use only. No other uses without permission. . All rights reserved.

CORTICOID-BINDING BY PLASMA MEMBRANES tentatively qjbtained by the equilibrium dialysis was almost equal, amounting to 2.5 x 10~9 mol per g membrane protein. The liver microsomes appear to have a considerably larger number of [3H]cortisol binding sites (N = 10 x 10~9 mol per g microsomal protein) but their affinity for cortisol appears to be low (Kd = 31 x 10~9 M). These features in the binding parameters seem to reduce the possibility that the cortisol binding sites in the liver plasma membranes, especially in the PM0, might be due to contamination with microsomal or serum binders in the plasma membrane preparations. Moreover, the possibility that [ 3 H]cortisol binding

150 OE

1501

TABLE 1. Number of [3H]cortisol-binding sites (N) in the subcellular components of rat tissues and dissociation constants of cortisol-ligand complexes (Kd) determined by the equilibrium dialysis method Nb

Subcellular component (tissue)

Number of experiments

(nmolyg protein)

Kdb (nM)

PM0 (liver)

2

2.5 ± 0.0

1.4 ± 0.4

PM, (liver)

2

2.6 ± 0.1

13 ± 3

Microsomes (liver)

2

10 ± 0

31 ± 0

Microsomes (kidney)

1

1.3

21

Cytosol (liver)

4

0.48 ± 0.07

1.6 ± 0.4

Cytosol (liver8)

1

0.20

1.6

Cytosol (kidney)

3

2.9 ± 0.1

1.1 ±0.1

Cytosol (kidney8)

1

1.4

1.3

Cytosol (spleen)

1

1.3

0.77

Serum

3

37 ± 8

4.7 ± 1.3

8 Tissues were first sliced, and then washed with 0.9% NaCl prior to homogenization. b Means of 2-4 separate experiments with average deviations. Subcellular fractions prepared from tissues of adrenalectomized rats were subjected to equilibrium 0 3 6 12 24 dialysis as described in Materials and Methods. The Time of dialysis ( h ) liver was perfused in situ with cold sucrose-K+-Mg2+but the other tissues FlG. 1. Kinetics of [3H]cortisol binding to plasma Tris buffer prior to homogenization were not. The concentration of [3H]cortisol was varied membranes in the equilibrium dialysis system. from 0.5 nM to 20 nM. Concentrations (g protein per Plasma membranes, PM0 and PM,, were prepared liter) of the subcellular fractions and serum were from livers of 10 adrenalectomized Wistar rats. A 0.35-1.19 (PM0), 1.17-4.08 (PM,), 7.55-14.7 (microsomes, liver), 14.2-14.8 (microsomes, kidney), 12.0mixture of 0.5 ml of the plasma membrane suspension liver and liver8), 5.8-8.2 (cytosol, kidney in TSS buffer and 0.5 ml of [3H]cortisol (4 nM) 20.7 (cytosol, 8 and kidney ), 5.6-6.2 (cytosol, spleen) and 6.8-11.1 solution in TSS buffer in a cellophane bag was (serum). dialyzed against 5 ml of TSS buffer containing the same concentration of [3H]cortisol (2 nM) at 4 C (in a cold room) with gentle shaking for periods of time sites in the plasma membrane preparations (3, 6, 12 and 24 h). Aliquots of the inner and outer fluids might be due to the cytosol binders adfrom each dialysis system were assayed for the sorbed on or attached to the membranes radioactivities and protein content. The apparent amount of cortisol bound to the membranes at each. also seems to be small, partly because the time point was tentatively calculated from the differnumber of binding sites per g protein of ence in [3H]cortisol concentrations between the inner the liver cytosol is much lower than that in and the outer fluids and the protein concentration in the plasma membranes and partly because the inner fluid. For a series of dialysis systems the repeated washing of the plasma memconcentrations of plasma membranes in the inner fluid were 0.60-0.76 g protein per liter (PM0) and 0.66-0.84 branes with physiological saline did not 3 g protein per liter (PM,). o and • indicate PM0 and reduce the [ H]cortisol binding sites (unPM, respectively. presented data). The extraordinarily large

50

It

u

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 23:53 For personal use only. No other uses without permission. . All rights reserved.

SUYEMITSU AND TERAYAMA

1502

Endo • 1975 Vol 96 • No 6

In order to examine whether the [3H]cortisol binding sites in the subcellular components of rat liver are specific to the glucocorticoids, competition experiments were carried out by simultaneously adding various nonlabelled steroids in concentrations 3-9-fold greater than that of [3Hlcortisol. As summarized in Table 2, binding of [3H]cortisol to rat liver PM0 was found to be reduced by the addition of corticosterone, cortisol and cortisone in order of decreasing effects. In contrast to PM0, binding of [3H]cortisol to the liver microsomes was affected less markedly and less specifically by the addition of these nonlabelled glucocorticoids. The pattern of competition by these nonlabelled glucocor-

10 -1

-0.5

0

05

1

LFree cortisol J"1 ( vf 1 ) FlG. 2. Double reciprocal plots of binding of [3H]cortisol to rat liver plasma membranes, cytosol and microsomes analyzed by the equilibrium dialysis, (a) Plasma membranes, PM0 and PM! were prepared from livers of 8 adrenalectomized rats in each of experiments I and II. Equilibrium dialysis was carried out as described in Materials and Methods. Concentrations (g protein per liter) of plasma membranes in the inner fluids were: for PM0, 0.99-1.19 (I) and 0.35-0.38 (II), and for PM1; 3.60-4.08 (I) and 1.17-1.72 (II). The concentration of [3H]cortisol was varied in a range of 1.0 nM to 10 nM. • and o indicate PM0 in I and II respectively, and A and A indicate PM, in I and II respectively. Kd (nM) thus calculated for PM0 were: 1.7 (I) and 1.0 (II), and N (nmol per g protein) for PM0 were 2.5 (I) and 2.5 (II), while Kd (nM) for PMi were 16 (I) and 10 (II) and N (nmol per g protein) for PM, were 2.6 (I) and 2.5 (II).

ou

5

T3 C D O CO

-1 C Free cortison" 1 (M~1)

number of cortisol binders found in kidney cytosol might be due partly to contamination by serum binders because in contrast to the liver, which was perfused with saline prior to homogenization, the other tissues such as kidney and spleen were homogenized without prior perfusion.

*10

FIG. 2 (b). See legend for Fig. 2 (a). Cytosol prepared from the livers of 3 adrenalectomized rats was used. [3H]Cortisol was in a range of 0.5 nM-5.0 nM. Duplicate determinations at each cortisol concentration are illustrated. Concentrations of cytosol in the inner fluids were 18.0-20.0 g protein per liter. Kd and N thus calculated were 1.6 nM and 0.41 nmol per g protein respectively.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 23:53 For personal use only. No other uses without permission. . All rights reserved.

1503

CORTICOID-BINDING BY PLASMA MEMBRANES

might have been similarly inactivated during the equilibrium dialysis, the binding parameters for PM 0 and PMj were reexamined by the centrifugation method. The freshly prepared PM0 and PMX were incubated at 0 C for 1, 2, 4, 6 and 12 h in the presence of various concentrations of [3H]cortisol and then centrifuged to separate the plasma membranes from the supernatant. The double reciprocal plots for the [3H]cortisol binding to PM0 and PMj after 2 hr incubation at 0 C are illustrated in Fig. 3. Similar plots in the experiments after 1, 6 and 12 h incubation are illustrated in Fig. 4a (PM0) and Fig. 4b (PM,). The binding parameters obtained with PM0 and PMX by 0

0.5 1 CFree cortisolJ"1 (M" 1 )

1.5 XIO9

FIG. 2 (c). See legend for Fig. 2 (a). Microsomes prepared from the livers of 3 adrenalectomized rats were used in each of experiments I and II. Concentrations (g protein per liter) of microsomes were 7.55-8.05 (I) and 13.7-14.7 (II) and the concentration of [3H]cortisol was varied in a range of 1.0 nM-10 nM. • and o indicate I and II, respectively. Kd (nM) and N (nmol per g protein) were; 32 and 10 in I, and 30 and 10 in II. 3

ticoids in [ H]cortisol binding to PMj appears to be intermediate between the above two cases (PM0 and microsomes). It should be noted that various synthetic glucocorticsids including dexamethasone did not affect the binding of [3H]cortisol to the liver cytosol. Since rat liver cytosol is known to contain G-binders which can bind not only the natural glucocorticoids but also synthetic ones such as dexamethasone, but is more labile than the other cytosol binders (A- and B-binders) that bind only to the natural glucocorticoids (2), the present results suggest that G-binders in the cytosol might have been inactivated during a long term (2 days) of dialysis. Taking into consideration that some of the binding sites in the plasma membranes

TABLE 2. Inhibition of binding of [3H]cortisol to subcellular components of rat liver by nonlabelled steroids Nonlabelled steroid added

Ratio of [3H]cortisol bound" PM0

PM,

Micros.

Cytosol

Serum

Corticosterone

0.16

0.44

0.73

0.25

0.18

Cortisol

0.26

0.40

0.63

0.45

0.49

Cortisone

0.70

0.70

0.69

1.03

0.97

0.72

0.87

11-Deoxycorticosterone acetate Progesterone

—

—

0.85

0.93

—

Testosterone

—

—

0.95

0.95

-

0.94

-

17/3-Estradiol

-

-

1.00

Prednisolone

-

-

0.78

0.70

-

Dexamethasone

—

-

0.99

0.93

1.00

Betamethasone

-

-

0.90

0.96

-

Triamcinolone

-

-

1.06

1.04

-

Concentration of [3H]cortisol (nM) Concentration of nonlabelled steroid (IIM)

1

10

20

4

20

9

90

80

12

60

3

• Ratios of bound [ H]cortisol in the presence and absence of nonlabelled steroid are presented. Numerical values were the results of duplicate experiments with a range of deviation of i 0-10%. The liver subcellular fractions as well as serum were subjected to equilibrium dialysis in duplicate in the presence of [3H]cortisol with or without nonlabelled steroids simultaneously added at the concentrations indicated. Concentrations (g protein per liter) of the liver subcellular fractions and serum were: 0.48-0.54 (PM0), 1.06-1.20 (PM,), 14.0-15.2 (microsomes), 12.8-13.7 (cytosol) and 7.1-8.3 (serum).

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 23:53 For personal use only. No other uses without permission. . All rights reserved.

SUYEMITSU AND TERAYAMA

1504

xiO

c ^

5 °» Q- o

V E

•o

c 3 O CD

0

1

2

3 1

H

[Free cortisoU" (M )

4 10 9

FIG. 3. Double reciprocal plots of binding of [3H]cortisol to rat liver plasma membranes analyzed by the centrifugal method. Freshly prepared plasma membrane fractions, PM0 and PM,, from 10 adrenalectomized rats were incubated with [3H]cortisol at various concentrations (0.3 nM-10 nM) at 0 C for 2 h. Concentrations of plasma membranes wexe 0.21 g protein per liter for PM0 and 0.37 g protein per liter for PM,. o and • indicate PM0 and PM, respectively. Biphasic binding equilibrium is shown only in the case of PM,. The values of Kd (nM) and N (nmol per g protein) thus calculated were as follows: For PM0; Kd = 1.4, N = 5.6, and for PM,; Higher affinity sites Kd = 1.9, N = 3.2 and intermediate affinity sites Kd = 3.9, N = 4.1.

changing the incubation time are summarized in Tables 3 and 4, respectively. These results seem to indicate that the higher affinity binding sites (with Kd = 1.5-1.9 x 10"9 M) in PM0 can be detected by the centrifugal method as by the equilibrium dialysis method. However, the number of the higher affinity binding sites in PM0 appears to change to some extent depending on the length of incubation at 0 C. The greatest number of binding sites, amounting to almost twice that obtained by the equilibrium dialysis method,

Endo i , 1975 V o l 9 6 i No 6

was detected after 2 h of incubation of PM0 with [3H]cortisol at 0 C. Upon increasing the incubation time further the apparent N value tended to decrease. In the case of PMX, the double reciprocal plots in the centrifugal analysis showed a biphasic binding equilibrium suggesting the presence of two types of binding sites having different affinities for cortisol. This was clearly observed when PMX was incubated with [3H]cortisol for less than 6 h. Under these conditions, PMX was shown to contain higher affinity binding sites with Kd of 1.6-1.7 x 10"9 M in addition to the intermediate affinity binding sites with Kd of 3.1-4.7 X 10~9 M. The higher affinity binding sites in PMX, however, were scarcely detected after 12 h of incubation with [3H]cortisol at 0 C, and only the intermediate affinity binding sites were detected, although the number of them was found to be reduced by about 40% compared with the maximal number detected at 2 h of incubation. Instead of the lower affinity binding sites with Kd of 13 x 10~9 M which were detected by equilibrium dialysis, PMX was found to contain the intermediate affinity binding sites by the centrifugal method. It is not clear at the moment whether the intermediate affinity sites in PMX detected by the centrifugal method are identical to the lower affinity ones detected by the equilibrium dialysis method, showing different Kd values because of the slightly different temperature conditions (0 C in the centrifugation analysis but 4 C in the equilibrium dialysis), or whether the former are transformed into the latter after a prolonged incubation (2 days). In order to examine the possibility that the apparent loss of the higher affinity binding sites in PMX during a long period of incubation with [3H]cortisol at 0 C might be due to the solubilization or release of the membrane associated binders into the medium, the supernatant obtained by centrifuging the mixture incubated with [3H]cortisol (2 x 10"9 M) for 24 h at 0 C

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 23:53 For personal use only. No other uses without permission. . All rights reserved.

1505

CORTICOID-BINDING BY PLASMA MEMBRANES xiO

FIG. 4. The change of the double ~ reciprocal plots of [3H]cortisol binding to rat liver plasma membranes, analyzed by the centrifugal method, upon varying the time of incubation with [3H]cortisol at 0 C. Three batches of plasma membranes, PM0 and PMj, were prepared, each from 10 adrenalectomized rats. PM0 and PMi of each batch were incubated with [3H]cortisol at various concentrations (0.3 nM10 nM) at 0 C for 1, 6 or 12 h. • , • and • indicate 1, 6 and 12 h incubations respectively, (a) PM0; Concentrations (g protein 1 2 1 2 per liter) of PM0 in each of the C Free cortisolf1 ( M"1) C Free cortisolf1 (M'1) 3 different incubation time groups were; 0.32 (1 h), 0.54 (6 h) and 0.42 (12 h). (b) PMt; Concentrations (g protein per liter) of PM, in each of the 3 different incubation time groups were; 0.58 (1 h), 0.90 (6 h) and 0.89 (12 h).

was subjected to gel filtration through Sephadex G-25. The results (not presented) showed that all the radioactivity in the supernatant was recovered in a single band eluted at the position corresponding to free cortisol without any appreciable radioactivity in the macromolecular region. Thus no evidence was obtained for this possibility. The stability of cortisol binding sites in PM1} however, seems to be dependent on the presence of cortisol. When PMX previously incubated at 0 C for 2 days in the absence of cortisol was subjected to binding site analysis by the centrifugal method, TABLE 3. Number of [3H]cortisol-binding sites (N) and dissociation constant of [3H]cortisol-ligand complex (Kd) in rat liver PM0 determined by the centrifugal method. Incubation time (h) 1 2 4 6 12

Number of expts.

N (nmol/g protein) 3.7 4.6 4.2 3.4

± 0.2 ± 0.8 ± 0.9 ± 0.1 2.9

the binding sites in PMX were found to be retained without any serious change in both Kd and N. This result together with that described earlier (Fig. 4b) may suggest that the binding sites in PMX are more stable in the absence of cortisol than in the presence of it. The results of similar experiments carried out with PM0 indicated that the higher affinity binding sites in PM0 may be more stable regardless of the presence or the absence of cortisol. The results described above indicate that centrifugation analysis may be more TABLE 4. Number of [3H]cortisol-binding sites (N) and dissociation constant of [3H]cortisol-ligand complex (Kd) in rat liver PM, determined by the centrifugal method Higher affinity binding

Intermediate affinity binding

Incubation time (h)

N (nmol/g protein)

Kd (nM)

N (nmol/g protein)

Kd (nM)

1 2 4 6 12

2.6 3.2 2.5 1.9 0

1.7 1.9 1.6 1.6

4.3 4.1 4.3 2.7 2.4

4.3 3.9 4.5 3.1 4.2

Kd (nM)

1.5 1.8 1.9 1.7

± 0.2 ± 0.4 ± 0.4 ± 0.3 2.3

PM0 of 0.18-0.54 g protein per liter was incubated with various concentrations of [3H]cortisol (0.1-10 nM) in the TSS buffer for various times at 0 C and then centrifuged. Numerical figures for N and Kd are means of 2-4 separate experiments with average deviations.

PMj of 0.42-2.0 g protein per liter was incubated with various concentrations of [3H]cortisol (0.3-10 nM) for various times at 0 C and centrifuged.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 23:53 For personal use only. No other uses without permission. . All rights reserved.

SUYEMITSU AND TERAYAMA

1506

appropriate than the equilibrium dialysis method for obtaining accurate parameters for [3H]cortisol binding to the plasma membranes. Based on the above findings, competition by several nonlabelled steroids added simultaneously with [3H]cortisol in studies on binding of [3H]cortisol to plasma membrane was reexamined by the centrifugation method. The results summarized in Table £> seem to indicate that higher affinity binding sites present in both PM0 and PMX show the highest affinity for corticosterone, which is followed by cortisol. These sites seem to have much lower affinity for the other steroids such as cortisone, dexamethasone or estradiol, since the binding of [3H]cortisol to the membranes was scarcely affected by the simultaneous addition of a steroid at 4 times the concentration of [3H]cortisol. In order to confirm that the rat liver plasma membranes may be devoid of dexamethasone binding sites, PM0 and PMi were incubated with [3H]dexamethasone (10"6-10-9 M) at 0C for 2 h and the binding of [3H]dexamethasone to the membranes was investigated by the centrifugation method. The results (Fig. 5) indicate that the binding of [3H]dexamethaTABLE 5. Inhibition of binding of [3H]cortisol to rat liver plasma membranes, PM0 and PMi, by nonlabelled steroids Nonlabelled steroid added Corticosterone Cortisol Cortisone Dexamethasone 17/3-Estradiol

No. of expts.

Ratio of [3H]cortisol bound8 PMn

PM,

0.30 ± 0.12 0.50 ±0.11 0.90 ± 0.05 0.99 ± 0.04 1.00

0.37 ± 0.02 0.53 ± 0.06 0.95 ± 0.05 0.98 ± 0.03 1.01

8 Ratios of bound [3H]cortisol in the presence and absence of nonlabelled steroid are presented. Numerical figures are means of 2-4 separate experiments with average deviations. Plasma membranes of 0.20-0.35 g proteins per liter for PM0 and of 0.35-0.90 g proteins per liter for PMj were incubated with [3H]cortisol (1 n\i) in the presence or the absence of nonlabelled steroid (4 nM) for 2 h at 0 C and then centrifuged. Amounts of [3H]cortisol bound to plasma membranes were assayed as described in Materials and Methods (the centrifugal method).

Endo • 1975 Vol96 • No 6

xiO 8 en 4 -

"3;

I 1-

3 O CD

0

5 10 [Free dexamethasone ] ( M"1)

15 xJ

?

°

FIG. 5. Double reciprocal plots of binding of [3H]dexamethasone to rat liver plasma membranes analyzed by the centrifugal method. Freshly prepared PM0 and PM, from 10 adrenalectomized rats were incubated with [3H]dexamethasone at various concentrations (10 nM-100 nM) for 2 h at 0 C before centrifugation. Concentrations of the plasma membranes were 0.77 g protein per liter for PM0 and 1.73 g protein per liter for PMj. No saturating tendency in the binding of [3H]dexamethasone was observed and the plots appear to pass through the origin, o and • indicate PM0 and PMi, respectively. In separate experiments the concentrations of [3H]dexamethasone were varied in ranges of 1 nM-10 nM as well as 100 nM-1000 nM, but the results were almost similar to Fig. 5, showing no specific binding of [3H]dexamethasone to either PM0 or PMi.

sone to plasma membranes does not show any concentration-dependent saturation tendency, suggesting that no specific dexamethasone binding sites may exist in both PM0 and PM^ Discussion In the present study we have shown that rat liver plasma membranes may have sites specifically responsible for binding the natural glucocorticoids such as corticosterone and cortisol. The affinity of cortisol

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 23:53 For personal use only. No other uses without permission. . All rights reserved.

CORTICOID-BINDING BY PLASMA MEMBRANES to some of these membrane-associated binding sites seems to be similar to the affinity of cortisol to the known cytosol binders of the rat liver (2). These higher affinity binding sites were shown to be present iri both PM0 and PMj. However, the higher affinity binding sites in PMi seem to be more labile than those in PM0, especially in the presence of cortisol. This fact may account for the failure to detect the higher affinity binding sites in PMi by the equilibrium dialysis method. The finding that the loss of the higher affinity binding sites in PMj during a long-term of incubation at 0 C may not be due to the release of hormone-binders into the medium but is probably due to their inactivation, would be important upon considering the hormone-mediated functions of the binding sites in the membranes. At the moment we do not know whether the higher affinity binding sites in PM0 and PMX are identical to each other, or why the higher affinity binding sites in PMj are more labile than those in PM0. PM0, a lighter plasma membrane fraction, seems to be derived mainly from the free cell surface areas facing the sinusoids, being rich in small vesicles (electron micrograph) probably derived from the microvilli. PMl9 a heavier fraction, seems to be mainly derived from the lateral surface areas that are in cell-to-cell contact, and is rich in tight junctions or desmosomes in accordance with the observation of Evans (12). It would be interesting to examine whether the intermediate affinity binding sites which are found only in the heavier plasma membrane fraction (PMJ by the centrifugation method are associated with the bile canaliculi. Although the substantial data presented in the present study seem to show that the cortisol binding sites found in the isolated plasma membranes may be intrinsic to the membranes, and are not mere contaminations of corticoid binders in the serum, cytosol or microsomes, the biological roles or significance of the plasma membrane-

1507

associated corticoid-binding sites will require further' studies. Preliminary investigations in our laboratory have indicated a marked decrease in the plasma membrane associated cortisol binding sites in some ascites hepatoma cells having as many cytosol binders'as in the normal liver cells, suggesting that the cortisol binding sites in the plasma membranes may be related somehow to the regulation of cell proliferation by glucocorticoids (13,14). The eorticoid binding sites in the plasma membranes seem to be specific for the natural glucocorticoids such as corticosterone and cortisol. The fact that they show the highest affinity for corticosterone is reasonable if we consider that corticosterone is a major glucocorticoid in the rat. Dexamethasone, a synthetic hormone known to be able to induce gluconeogenesis and enzymes related to amino acid catabolism, such as tyrosine aminotransferase, was found not to be bound specifically to the plasma membranes. According to Beato et al. (1,2) rat liver cytosol contains at least 3 different types of corticoid binders; one (G-binder) is labile and binds not only the natural glucocorticoids but also the synthetic ones like dexamethasone, while the others (A- and B-binders) are less labile and bind only the natural glucocorticoids. The results of the present study have shown that rat liver plasma membranes are devoid of the sites specifically binding dexamethasone, although the possibility that the dexamethasone binding sites in the plasma membranes might be extremely labile and be lost during the course of plasma membrane preparation cannot be excluded at the moment. Studies are now in progress in our laboratory on the biological functions of the corticoid-binding sites in the plasma membranes. Acknowledgments The authors are grateful to Dr. H. Nakamura of the Research Institute of the Shionogi Pharmaceutical Company, Osaka, Japan for supplying some of the nonlabelled corticoids of high purity used in the

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 23:53 For personal use only. No other uses without permission. . All rights reserved.

SUYEMITSU AND TERAYAMA

1508

present study. A part of expenditures for the present study was covered by the Grant-in-Aid for Scientific Research from the Ministry of Education, Japan, as well as the Ford Foundation Grant, No. 740-0403.

References 1. Koblinsky, M., M. Beato, M. Kalint, and P. FeigelsonJ Biol Chem 247: 7897, 1972. 2. Beato, M., and P. Feigelson, J Biol chem 247: 7890, 1972. 3. Baxter, J. D., and G. M. Tomkins, Proc Natl Acad Sci USA 68: 932, 1971. 4. Sutherland, E. W., and T. W. Kail, Pharmacol Rev 12: 265, 1960. 5. Goldfine, I. D., J. Roth, and L. Birnbaumer,/ Biol Chem 247: 1211, 1972.

Endo • 1975 Vol 96 • No 6

6. Giogio, N. A., C. B. Johnson, and M. Blecher, / Biol Chem 249: 428, 1974. 7. Berman, H. M., W. Gram, and M. A. Spirtes, Biochim Biophys Ada 183: 10, 1969. 8. Yamamoto, K., S. Omata, T. Ohnishi, and H. Terayama, Cancer Res 33: 567, 1973. 9. Bottoms, G. D., R. D. Stith, and R. O. Burger, Proc Soc Exp Biol Med 132: 1133, 1969. 10. Lowry, O. H., N. J. Rosebrough, A. L. Farr, and R. L. Randall,/ Biol Chem 193: 265, 1951. 11. Beato, M., W. Schmid, and C. E. Sekeris, Biochim Biophys Ada 263: 764, 1972. 12. Evans, W. H., BiochemJ 116: 833, 1970. 13. Henderson, I. C., R. E. Fischel, and J. N. Loeb, Endocrinology 88: 1471, 1971. 14. Loeb, J. N., C. Borek, and L. L. Yeung, Proc Natl Acad Sci USA 70: 3852, 1973.

«•*-