Cholesterol modulation of β-adrenergic receptor characteristics

Biochimica et Biophysica Acta 845 (1985) 520-525 Elsevier

520

BBA 11490

C h o l e s t e r o l m o d u l a t i o n of f l - a d r e n e r g i c r e c e p t o r c h a r a c t e r i s t i c s Philip J. S c a r p a c e *, S t e p h e n W. O ' C o n n o r a n d I t a m a r B. A b r a s s ** Geriatric Research, Education and Clinical Center (GRECC), Sepulveda VA Medical Center, Sepulveda, CA 91343, and Department of Medicine, UCLA School of Medicine, Los Angeles, CA 90024 (U.S.A.) (Received January 7th, 1985)

Key words: Cholesterol modulation; fl-Adrenergic receptor

Cholesterol, a major structural component of plasma membranes, has a profound influence on cell surface receptor characteristics and on adenylate cyclase activity, fl-Adrenergic receptor number, adenylate cyclase activity, and receptor-cyclase coupling were assessed in rat lung membranes following preincubation with cholesteryl hemisuccinate, fl-Adrenergic receptor number increased by 50% without a change in antagonist affinity. However, fl-adrenergic receptor affinity for isoproterenol increased 2-fold as a result of an increase in the affinity of the isoproterenol high-affinity binding site. This increase in agonist affinity did not potentiate hormone-stimulated adenylate cyclase activity, which decreased 3-fold following cholesterol incorporation. However, the ratio of isoproterenol to GTP-stimulated activity was unchanged with cholesterol. Stimulation distal to the receptor by GTP, NaF, GppNHp, Mn 2+ and forskolin also demonstrated 50-80% reduced enzyme activity following cholesterol incorporation. These data suggest that membrane cholesterol incorporation decreases catalytic unit activity without affecting transduction of the hormone signal. Introduction

Cholesterol and phospholipids are major structural components of plasma membranes and have profound influences on cell surface receptors and hormone-stimulated adenylate cyclase activity. /3Adrenergic receptor numbers are decreased following phospholipid incorporation into membranes of Chang liver cells [1], increased following phospholipid methylation in rat reticulocytes [2], increased following addition of phospholipid to solubilized rabbit heart membrane [3], and decreased in rat atria following dietary inclusion of

* To whom correspondence should be addressed at: GRECC (llE), VAMC, Sepulveda, CA 91343 U.S.A. ** Present address: Harborview Medical Center (ZA-87), 3259th Avenue, Seattle, WA 98104, U.S.A. Abbreviation: Hepes, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid.

sunflower oil [4]. In mouse brain, serotonin receptor number is increased and affinity decreased following treatment with cholesteryl hemisuccinate or steric acid, while the number of receptors decreases following treatment with phosphatidylcholine or linolic acid [5]. It has been suggested that alterations in receptor density are secondary to changes in membrane fluidity [5]. Cholesterol and steric acid decrease membrane fluidity and increase serotonin receptor number, while phosphatidylcholine and linolic acid increase membrane fluidity and decrease serotonin receptor number [5]. However, changes in /3adrenergic receptors do not necessarily correlate with membrane fluidity. Increases in fluidity decrease the number of receptors in Chang liver cells [1], increase the number of receptors in rat reticulocyte ghosts [2], and have no effect on receptor number in turkey erythrocytes [6]. We reported earlier that /3-adrenergic receptor density in rat lung membranes increases in a

521 dose-dependent manner following incorporation of cholesteryl hemisuccinate [7]. This increase in receptor density was associated with a corresponding decrease in both hormone- and NaFstimulated adenylate cyclase activity. However, the effects of cholesterol on receptor adenylate cyclase interaction have not been defined. The adenylate cyclase complex represents a transmembrane regulatory system consisting of three components. Located at the outer membrane surface is the receptor component containing a specific site for binding of hormones and neurotransmitters. Distal to the receptor at the inner surface of the membrane are the nucleotide-regulatory protein and the catalytic unit. The lateral movement of components of the adenylate cyclase system is important for the activity of the enzyme, and a modification in the lipid composition of the membrane can result in diminished enzyme activity [1]. A prerequisite for hormone activation of adenylate cyclase is the formation of a hormonereceptor-N-protein (HRN) complex. This HRN complex displays high-affinity binding for agonists and its formation enhances the binding of guanyl nucleotides [9]. The ability or inability to form the H R N complex is functionally assessed by changes in the ability of the receptors to stimulate adenylate cylase. It can also be seen as a change in the agonist-binding properties of the receptor. Uncoupled receptors are less able to form the active ternary HRN complex. As a result, agonist binding to these receptors is of lower affinity and is less responsive to guanine nucleotides. This can be measured as a rightward shift and steepening of an agonist competition curve. Using the agonist competition curves, the affinity of the high- (KH) and low-affinity (KL) states of the receptor and the ratio of the number of receptors in the high compared to the low-affinity state (% Rh) can be calculated with the aid of a computer program [10]. A small, high-affinity to low-affinity ratio is consistent with the uncoupled state, whereas a large ratio is consistent with a tightly coupled state. To further investigate the effects of cholesterol on hormone-stimulated adenylate cyclase and the complex interactions of receptor and enzyme, we asessed fl-adrenergic receptor number, receptoragonist binding, and adenylate cyclase activity in

rat lung membranes following in vitro incorporation of cholesteryl hemisuccinate. Methods

[3H]Dihydroalprenolol, spec. act. 49.1 C i / mmol, was obtained from New England Nuclear (Boston, MA). Biochemicals were obtained from Sigma (St. Louis, MO). Female 170-200 g CDF (F-344) rats were obtained from Charles River Breeding Laboratories (North Wilmington, MA). Lung membranes. Lung membranes were prepared as described previously [11]. Briefly, rats were euthanized by cervical dislocation and the circulatory system perfused with 10 ml cold saline. Lungs were minced, disrupted for 30 s in a Tekmar tissuemizer (Tekmar, Cincinnati, OH), and homogenized 10 strokes with a motor-driven Teflontipped pestle. The homogenate was centrifuged at 48000 x g for 15 min and the pellet harvested. Membranes (1 mg protein/ml) were preincubated for 2 h at 24°C in 45 mM Tris-HC1 (pH 7.4)/9.5% ethanol, either with or without 0.5 mg/ml cholesteryl hemisuccinate. When adenylate cyclase activity was measured, 10 mM NaF was included in the incubation mix to stabilize the enzyme [12]. Membrane preparations were recentrifuged and washed twice at 48000 X g and the pellet suspended in 0.08 mM ascorbic acid/18 mM MgC12/50 mM Hepes (pH 7.4). fl-Adrenergic receptor assay, fl-Adrenergic receptors and antagonist affinity were determined from an eight-point Scatchard analysis of [3H]dihydroalprenolol binding as described previously [13]. Apparent dissociation constants (agonist affinity) for isoproterenol were calculated from the abscissa intercept (IC50) of Hill plots of isoproterenol competition with 2.5 nM [3H]dihydroalprenolol. The number of high- and low-affinity binding sites and the corresponding dissociation constants were calculated using the competition binding data (Hill plots) and the Ligand program (Biomedical Computing Technology Information Center, Vanderbilt Medical Center, Nashville, TN). These calculations determine the percent of receptors in the high-affinity binding state (% R h ) , the high-affinity binding constant for isoproterenol (KH), and the low-affinity binding constant for isoproterenol (KL). Antagonist (dihydroalpre-

522

nolol) affinity was determined from Scatchard plots. Adenylate cyclase assay. Approx. 75/*g of membrane protein were assayed for adenylate cyclase activity as described earlier [14]. In some tubes, either 10 mM NaF, 10 - 4 M GppNHp, 33.3 /~M forskolin, 10 mM Mn 2+, 10 - 4 M GTP, o r 10 - 4 M GTP plus 0-10 -5 M isoproterenol were present as enzyme stimulants. Protein was determined by a modification of the biuret method [14]. Membrane cholesterol and phosphate. Membrane cholesterol was extracted with chloroform,/ methanol (2: 1, v/v) by the method of Frolich et al. [15]. The extract was evaporated to dryness and the cholesterol was assayed enzymatically by the method of Allain et al. [16], (Sigma Chemical Co. Kit No. 350-A, St. Louis, MO). The dried chloroform/methanol extract was assayed for total phosphate by the colormetric method of Chen et al. [17]. Results

fl-Adrenergic receptor characteristics were assessed following preincubation with 0.5 m g / m l cholesteryl hemisuccinate for 2 h. Cholesteryl hemisuccinate incorporated into the membranes readily with maximum incorporation of 0.266 mg

2,0-

TABLE I EFFECT OF CHOLESTERYL HEMISUCCINATE ON /3A D R E N E R G I C RECEPTOR CHARACTERISTICS Membranes were preincubated for 1 h at 24°C with either 9.5% ethanol (control) or 9.5% ethanol plus cholesteryl hemisuccinate, 0.5 mg/ml. Data represent the mean+S.E, of five (receptor number) or ten (isoproterenol affinity) determinations. K H = K d high-affinity state; K L = K d low-affinity state. P values were determined from one-way ANOVA. n.s., not significant. Control

Cholesterol

Receptor number (fmol/mgprotein) " 138 +36 216 _+30 Dissociation constant for dihydroalprenolol ( X I 0 -9 M) " 0.83+ 0.29 0.83+ 0.28 Dissociation constant for isoproterenol ( × 1 0 7M) b 1.7 + 0.1 0.89+ 0.12 Receptors in highaffinity state(%)c 49.9 + 1.8 49.8 + 2.5 KH (×10-SM) c 4.5 _+ 0.6 2.2 _+ 0.3 KL (X10-6M) c 2.1 + 0.3 2.0 + 0.9 KL/K H 48.5 + 3.6 92.0 +19

P value < 0.001

n.s.

<0.001 n.s. <0.02 n.s. <0.022

Receptor number and antagonist (dihydroalprenolol) affinity were determined from Scatchard plots. b Apparent dissociation constants for isoproterenol were calculated from the ICs0 of agonist competition curves with 2.5 nM [3H]dihydroalprenolol. c Hill plots were subjected to two-site computer analysis for determination of % receptors in the high-affinity state and the binding constants of the two sites. a

o

16c~ t2<

-r 0.8"

,."h C~ Z

0.4-

m

40

80

120

160

200

DHA BOUND (f m01/mg)

Fig. 1. Scatchard analysis of [3H]dihydroalprenolol binding to lung membranes preincubated without (O) or with (e) 0.5 m g / m l cholesteryl hemisuccinate for 2 h. Lines were determined by regression analysis with correlation coefficients of 0.99 and 0.99, receptor densities of 133 and 209 f m o l / m g protein (indicated by abscissa intercept) and dissociation constants of 0.94 and 0.82 nM (indicated by the negative reciprocal of the slope of the line), respectively for without and with cholesterol preincubation. DHA, dihydroalprenolol.

cholesterol/mg protein compared to untreated value of 0.022 mg cholesterol/mg protein. Total membrane phosphorus was unchanged with cholesterol incorporation. Scatchard analysis of [3H]dihydroalprenolol binding to membranes incubated with or without cholesteryl hemisuccinate yielded a straight line consistent with a single class of antagonist-binding sites (Fig. 1). Preincubation with cholesteryl hemisuccinate was associated with a 50% increase in the number of fl-adrenergic receptors compared to lung membranes incubated with ethanol (Fig. 1, Table I). The increase in receptor number was not associated with any change in receptor antagonist affinity (Fig. 1, Table I). There were no differences in receptor number between membranes incubated with ethanol

523

30

~.0" 08" 0.60.4.

•

o

3

02.

(:nl~

0. . . . . . . . . . . . . . . . . . . . . . kd

o~ -02

~ t5

-0.4

>-

-0.6 -0.8

-to,

,~

~ f -log ~soproterenol] (M)

@

Fig 2. Hill plots constructed from isoproterenol competition with [3H]dihydroalprenolol in lung membranes preincubated without (O) or with (e) 0.5 mg/ml cholesteryl hemisuccinate for 2 h. Lines were determined by regression analysis. Hill coefficients (indicated by the slope of the line) are 0.52 and 0.53, and concentrations for half-maximal displacement are 7.7 and 1.8.10 7 M, respectively for without and with cholesterol.

and untreated membranes (138 + 36 vs. 138 + 10 fmol/mg protein, respectively). /3-Adrenergic receptor affinity for isoproterenol was assessed by Hill plots of isoproterenol competition with [3H]dihydroalprenolol in lung membranes preincubated with or without cholesteryl hemisuccinate (Fig. 2). There was a significant increase in the affinity (decreased dissociation constant) of the receptor for isoproterenol in mem-

TABLE II EFFECT OF CHOLESTERYL HEMISUCCINATE ON ADENYLATE CYCLASE ACTIVITY Membranes were preincubated for 2 h at 24°C with either 9.5% ethanol (control) or 9.5% ethanol plus cholesteryl hemisuccinate, 0.5 mg/ml. Data represent the meaniS.E, of 12 determinations. The values for isoproterenol represent activities above GTP. P < 0.001 for the differences between control and cholesterol by ANOVA. Stimulant

None GTP Isoproterenol NaF GppNHp Forskolin Mn 2+

Adenylate cyclase activity (pmol cAMP/rain per mg protein) control

cholesterol

36.7 ± 5.7 43.5 ± 2.0 39.1± 8.1 100.0 ± 14.9 74.4 ± 11.5 51.2± 8.7 50.5 ± 10.5

8.8 ± 2.8 16.9± 6.3 14.9± 5.1 42.2 ± 9.0 24.7 ± 5.5 11.5 ± 3.5 26.8 ± 6.3

g 9

8 ? 6 -log [ISOPROTERENOL]

5

Fig. 3. Isoproterenol-stimulated adenylate cyclase activity in lung membranes preincubated without (O) or with (O) 0.5 mg/ml cholesteryl hemisuccinate for 2 h. Data are the ratio of enzyme activity in the presence of isoproterenol plus GTP compared to GTP. The concentration of isoproterenol for half-maximal stimulation is 1.3.10 -7 M for both incubated with and without cholesterol. Data represent mean+S.E, of three experiments.

branes incubated with cholesteryl hemisuccinate compared to controls (Fig. 2, Table I). The competition binding data was subjected to computer analysis for determination of the number of highand low-affinity binding sites and the corresponding dissociation constants. There was no change in the percent of high-affinity isoproterenol binding sites; however, there was an increase in the affinity (decreased dissociation constant) of the high-affinity site in cholesterol-treated compared to control membranes (Table I). There were no differences in affinity of the low-affinity binding site. The ratio of the low-affinity to highaffinity binding constants (Kt/KH) increased 2fold (Table I). In these experiments, the changes in the apparent K d observed in Hill plots reflect changes in the dissociation constant of the high-affinity binding site rather than changes in the relative amount of high-and low-affinity binding sites. Adenylate cyclase activity was assessed in membranes preincubated with and without cholesteryl hemisuccinate. There was a 50-80% decrease in GTP, NaF, GppNHp, forskolin and Mn2+-stimulated adenylate cyclase activity following preincubation with cholesteryl hemisuccinate (Table II). The increase in receptor affinity for isoproterenol

524

following cholesterol incorporation did not potentiate the maximal isoproterenol stimulation of adenylate cyclase activity. Total hormone-stimulated enzyme activity was 63% less following cholesterol incorporation (Table II); however, the ratio of maximal isoproterenol to GTP-stimulated activity and the concentration for half-maximal stimulation were unchanged (Fig. 3). At low concentrations of isoproterenol, this ratio was greater in the cholesterol compared to untreated membranes (Fig. 3). Discussion

In the present study, we have investigated the modulation by cholesterol of fl-adrenergic receptor binding properties and adenylate cyclase activity in rat lung membranes. This study and our previous report indicate that incorporation of cholesterol into membranes increases the number of fl-adrenergic receptor binding sites [7]. It has been suggested that cholesterol modulation of cell surface receptors are secondary to the changes in membrane fluidity [5]. We (unpublished data) and others have shown that temperature, another modulator of membrane fluidity, does not alter fladrenergic receptor density [18,19]. Treatment with cholesterol, which decreases membrane fluidity, increases receptor number, while a decrease in temperature, which also decreases fluidity, has no effect on receptor number. When turkey erythrocyte membranes are delipidated, fl-adrenergic receptor ligand binding is lost [20]. Binding can be restored by the addition of cholesterol and phospholipids. Taken together, these data suggest that cholesterol modulation of ligand binding is a result of an interaction with fl-adrenergic receptors rather than secondary to changes in membrane fluidity. We and others have reported that cholesterol incorporation into membranes decreases adenylate cyclase activity [7,21]; however, another report indicated no change in enzyme activity following cholesterol incorporation in platelet membranes [22]. Incorporation of phosphatidylinositol into turkey erythrocytes decreases hormone-stimulated adenylate cyclase activity, while enzyme activity stimulated distal to the receptors is unchanged. This decrease in receptor-adenylate cyclase cou-

piing was associated with a decrease in the ratio of high- to low-affinity agonist binding sites [23]. In the present study, cholesterol incorporation into membranes increased the fl-adrenergic receptor agonist affinity for isoproterenol. This change in agonist affinity was determined to be a consequence of an increase in the affinity constant of the high-affinity isoproterenol binding site. An increase in fl-adrenergic receptor number and agonist affinity may potentiate isoproterenol stimulation of adenylate cyclase. However, isoproterenol-stimulated enzyme activity was substantially decreased following cholesterol incorporation. Other stimulators, acting distal to the receptor, such as GTP, NaF and GppNHp which act through N-protein or Mn 2+ and forskolin which act through the catalytic unit, also demonstrated decreased ability to stimulate enzyme activity following cholesterol incorporation. These data suggest that cholesterol decreases the activity of the catalytic unit of adenylate cyclase. When isoproterenol-stimulated enzyme activity is expressed as a percent increase above GTP-stimulated activity or when the concentration for half-maximal stimulation is analyzed, there is no difference between control and cholesterol-treated membranes. Agonist-induced changes in receptor-adenylate cyclase coupling are often associated with a change in % R h due to a shift of receptors from one affinity state to the other. In the present study, cholesterol had no effect on % Rh; however, cholesterol lead to a decrease in K H (increased affinity). This decrease in K H may be related to the increased sensitivity to isoproterenol of agonist-stimulated adenylate cyclase activity at low isoproterenol concentrations. In summary, these data support the notion that membrane cholesterol, though decreasing adenylate cyclase activity, does not affect overall transduction of the hormone signal in the rat lung fl-adrenergic system. In this setting, the catalytic unit appears to be rate-limiting; thus the increase in receptor number and agonist affinity is not translated into increased agonist-stimulated enzyme activity. Acknowledgements

This work is supported by the Medical Research Service of the Veterans Administration. The

525

authors appreciate the excellent technical assistance of Charlene Kim, and the assistance of Jo Ann Phillips in the preparation of the manuscript. References 1 Bakardjieva, A., Galla, H.J. and Helmreich, E.J.M. (1979) Biochemistry 18, 3016-3023 2 Strittmatter, W.J., Hirata, F. and Axelrod, J. (1979) Science 204, 1205-1207 3 Thank, N.X., Borsodi, A. and Wollermann, M. (1980) Biochem. Pharmacol. 29, 2791-2797 4 Wince, L.C. and Rutledge, C.O. (1981) J. Pharmacol. Exp. Ther. 219, 625-631 5 Heron, D.S., Shinitzky, M., Hershkowitz, M. and Samuel, D. (1980) Proc. Natl. Acad. Sci. USA 77, 7463-7467 6 Hanski, E., Rimon, G. and Levitzki, A. (1979) Biochemistry 18, 846-853 7 O'Connor, S.W., Scarpace, P.J. and Abrass, I.B. (1984) Biochim. Biophys. Acta 778, 497-502 8 Rodbell, M. (1980) Nature 282, 17-22 9 Stadel, J.M., Shorr, R.G.L., Limbird, L.E. and Lefkowitz, R.J. (1981) J. Biol. Chem. 256, 8718-8723 10 DeLean, A., Stadel, J.M. and Lefkowitz, R.J. (1980) J. Biol. Chem. 255, 7108-7117

11 Scarpace, P.J. and Abrass, I.B., (1982) J. Pharmacol. Exp. Ther. 223, 327-331 12 Scarpace, P.J., O'Connor, S.W. and Abrass, I.B. (1983) Life Sci. 32, 817-824 13 Scarpace, P.J., and Abrass, I.B. (1981) Endocrinology 108, 1007-1011 14 O'Connor, S.W., Scarpace, P.J. and Abrass, I.B. (1983) Mech. Ageing Dev. 21,357-363 15 Frolich, J., Lees, M. and Sloane-Stanley, G.H. (1957) J. Biol. Chem. 226, 479-509 16 Allain, C.C., Poon, L.S. and Chan, C.S.G. (1974) Clin. Chem. 20, 470-475 17 Chen, P.S., Toribara, T.Y. and Warner, H. (1956) Anal. Chem. 28, 1756-1758 18 Weiland, G.A., Minnerman, K.P. and Molinoff, P.B. (1980) Mol. Pharmacol. 18, 341-347 19 Weiland, G.A., Minnerman, K.P. and Molinoff, P.B. (1979) Nature 281, 114-117 20 Kirilovsky, J.K. and Schramm, M. (1983) J. Biol. Chem. 258, 6841-6849 21 Klein, I., Moore, L. and Pastan, I. (1978) Biochim. Biophys. Acta 506, 42-53 22 Insel, P.A., Nirenberg, P., Turnbull, J. and Shattil, S.J. (1978) Biochemistry 17, 5269-5274 23 McOsker, C.C., Weiland, G.A. and Zilversmit, D.B. (1983) J. Biol. Chem. 258, 13017-13028