Fat cell adenylate cyclase and beta-adrenergic receptors in altered thyroid states.

The lipolytic sensitivity of fat cells from hypothyroid rats to epinephrine is severely blunted although 100 PM epinephrine stimulates lipolysis to the same level demonstrated by fat cells from euthyroid rats. Hypothyroidism shifts the lipolytic log-dose response curve to epinephrine to the right by a factor of 10. Epinephrine produced no detectable increase in adenosine 3’:5’-monophosphate (cyclic AMP) accumulation in fat cells from hypothyroid rats. In contrast, fat cells from hyperthyroid rats accumulated more cyclic AMP in response to epinephrine than did those from euthyroid rats. Fat cell ghosts prepared from hypothyroid rats display reduced catecholamine-stimulated adenylate cyclase activity, although the maximal catalytic activity of fat cell ghost adenylate cyclase as measured in the presence of 10 mM sodium fluoride is the same as that of the euthyroid. Fat cell ghosts from hypothyroid rats were incubated with the guanine nucleotide analog, guanyl-5’-yl imidodiphosphate in an attempt to restore the reduced catecholamine-stimulated adenylate cyclase activity. This preincubation with guanyl-5’-yl imidodiphosphate activated the adenylate cyclase and increased catecholamine-stimulated activity of ghosts prepared from fat cells of both euthyroid and hypothyroid rats but did so to a much greater extent in the fat cell ghosts from euthyroid rats. Fat cell ghosts from hypothyroid rats incubated in the presence of a maximal concentration of guanyl-V-y1 imidodiphosphate displayed 37 f 3% less adenylate cyclase activity than the fat cell ghosts obtained from euthyroid animals. Cyclic AMP phosphodiesterase activity of fat cell ghosts preincubated with or without guanyl-5’-yl imidodiphosphate, when measured at 0.125 PM cyclic AMP, was the same in ghosts from hypothyroid rats as in those from normal rats. Similarly, l-methyl-

The lipolytic sensitivity of fat cells from hypothyroid rats to epinephrine is severely blunted although 100 PM epinephrine stimulates lipolysis to the same level demonstrated by fat cells from euthyroid rats. Hypothyroidism shifts the lipolytic log-dose response curve to epinephrine to the right by a factor of 10. Epinephrine produced no detectable increase in adenosine 3':5'-monophosphate (cyclic AMP) accumulation in fat cells from hypothyroid rats. In contrast, fat cells from hyperthyroid rats accumulated more cyclic AMP in response to epinephrine than did those from euthyroid rats.
Fat cell ghosts prepared from hypothyroid rats display reduced catecholamine-stimulated adenylate cyclase activity, although the maximal catalytic activity of fat cell ghost adenylate cyclase as measured in the presence of 10 mM sodium fluoride is the same as that of the euthyroid.
Fat cell ghosts from hypothyroid rats were incubated with the guanine nucleotide analog, guanyl-5'-yl imidodiphosphate in an attempt to restore the reduced catecholamine-stimulated adenylate cyclase activity. This preincubation with guanyl-5'-yl imidodiphosphate activated the adenylate cyclase and increased catecholamine-stimulated activity of ghosts prepared from fat cells of both euthyroid and hypothyroid rats but did so to a much greater extent in the fat cell ghosts from euthyroid rats. Fat cell ghosts from hypothyroid rats incubated in the presence of a maximal concentration of guanyl-V-y1 imidodiphosphate displayed 37 f 3% less adenylate cyclase activity than the fat cell ghosts obtained from euthyroid animals. Cyclic AMP phosphodiesterase activity of fat cell ghosts preincubated with or without guanyl-5'-yl imidodiphosphate, when measured at 0.125 PM cyclic AMP, was the same in ghosts from hypothyroid rats as in those from normal rats. Similarly, l-methyl- 3-isobutyl xanthine (1 mM) did not rectify the reduced response of adenylate cyclase of the fat cell ghosts from hypothyroid rats to catecholamine stimulation. j?-Adrenergic receptors were examined in membranes prepared from isolated fat cells obtained from hyperthyroid, hypothyroid, and euthyroid rats. Putative P-adrenergic receptors were identified with the use of the potent, /3-adrenergic antagonist, (-)-[3Hldihydroalprenolol.
Specific binding of ( -WHldihydroalprenolol to fat cell membranes was rapid, reversible, and saturated at 80 nM. Scatchard plots of specific binding were curvilinear with upward concavity; Hill plots yield coefficients of 0.7. Competition studies display stereospecificity and a potency order of (-)-agonists (isoproterenol >> epinephrine =norepinephrine) indicative of a &-adrenergic receptor. Calculating from membrane binding data, there appear to be approximately lo5 P-adrenergic receptors/fat cell. Maximum specific binding of (-)-[3Hldihydroalprenolol/mg of membrane protein was the same for fat cell membranes prepared from hypothyroid, hyperthyroid, and euthyroid rats. Scatchard plots of binding studies performed with 1 to 100 nM (-)-[3HIdihydroalprenolol were nearly identical for fat cell membranes prepared from the three different groups, each displaying curvilinearity with an upward concavity. P-Adrenergic agonists compete for (-)-[3Hldihydrolalprenolol binding with the same potency when measured in fat cell membranes of euthyroid or hypothyroid rats. The total number of fi-adrenergic receptors per fat cell was the same for fat cells obtained from hyperthyroid, hypothyroid, and control rats. These data suggest that thyroid hormones alter neither the maximum catalytic activity of adenylate cyclase nor the number and affinity of putative /3-adrenergic receptors of the fat cell. Thyroid hormones, thus, may exert their influence on fat cells by regulating the transduction of information between hormone receptors and adenylate cyclase. Debons and Schwartz (1) reported that the stimulation by epinephrine of free fatty acid release in vitro was impaired in adipose tissue from rats rendered hypothyroid by 6-N-propyl-2-thiouracil administration in viuo. The reduction in catecholamine-stimulated lipolysis in fat pads, tissue, and isolated cells of hypothyroid rats is now well documented (2-6). Correze et al. (5) reported a reduction in the maximal responses of adenylate cyclase in fat cell membranes to adrenocorticotropic hormone, glucagon, or epinephrine following thyroidectomy. Armstrong et al. (4), in contrast, reported the activation by hormones and fluoride of adenylate cyclase of fat cell ghost preparations to be the same in hypothyroid and control preparations. These investigators reported, in addition, increased activity of particulate, high affinity forms of adenosine 3'S'monophosphate phosphodiesterase in fat cells from hypothyroid rats (4, 6).
One mechanism by which thyroid hormones could influence hormone-stimulated lipolysis in the fat cell is via modulation of hormone receptor number or affinity (or both). Ciaraldi and Marinetti (7) and Williams et al. (8) recently reported increases in the number of p-adrenergic receptors in the heart by thyroid hormone treatment in ho. The present study was designed to examine the influence of thyroid status on fat cell adenylate cyclase and P-adrenergic receptor number and affinity. Corp.) and water containing 0.00625% 6-N-propyl-2-thiouracil for 21 to 24 days. Hyperthyroid rats were rats given subcutaneous administration of 30 pg of (-)-3,3',5-triiodothyroninellO0 g, body weight, daily for 5 days. An additional group of rats were rendered hypothyroid as above and then given 30 pg of T,'/lOO g, body weight, daily for 5 days while remaining on the iodine-deficient diet and the PTU.
White fat cells were obtained by enzymatic digestion of parametrial adipose tissue according to the procedure of Rodbell (9). Pooled adipose tissue (-100 gf from 15 to 20 rats was minced with scissors and placed in small plastic bottles. Each bottle, containing 8 to 10 g of tissue and 10 ml of Krebs-Ringer phosphate buffer containing 3% albumin and 1 mg/ml of crude collagenase (Clostridium histolyticam, Worthington Biochemical Corp., lot CLS 46E168P1, was incubated for 60 min at 37". The Krebs-Ringer phosphate buffer contained 128 rnM NaCl, 1.4 mM CaCl,, 1.4 mM MgSO,, 5.2 mM KCl, and 10 rnM Na*HPO,. The albumin buffer was made fresh daily and the pH adjusted to 7.4 with NaOH after addition of the bovine serum albumin Fraction V powder (Armour, lots NlOlOl and P56607). At the end of 60-min digestion, cells were filtered through one layer of nylon chiffon and washed twice with the albumin buffer. Cyclic AMP accumulation was measured in cells plus medium after a 0.2-ml aliquot of cells plus medium, in duplicate, was extracted from 1 ml of incubation volume and added to tubes on ice containing 20 ~1 of 2 N HCl. The tubes were then placed in a boiling water bath for 1 min. The tubes were allowed to cool before 10 ~1 of 4 N NaOH was added. The contents of the tubes were mixed and centrifuged prior to removal of 20-~1 aliquots for determination of cyclic AMP. In each experiment, no more than 50 mg of fat cells were incubated/ml of medium which means that the 20-111 aliquots taken for cyclic AMP analysis represented less than 1 mg of fat cells. Cyclic AMP release to the medium was analyzed by taking 20-~1 aliquots of the medium at the end of the incubation just prior to removing the 0.2-ml aliquots of cells plus medium. The cyclic AMP standards were prepared in incubation medium containing albumin which was treated in the same manner as the unknown samples by adding acid and then boiling and neutralizing.
The assay for cyclic AMP was done by a modification of the Gilman (10) protein kinase binding procedure using rabbit muscle protein kinase. The free cyclic AMP was separated from the bound cyclic AMP by charcoal absorption (11). Glycerol release was analyzed in 50-pl aliquots taken from the incubation mixture as described previously (12). In each experiment, the values for lipolysis and cyclic AMP were based on the mean of duplicate determinations. The data are expressed per lo6 fat cells ' The abbreviations used are: T,, 3,3',5(-)-triiodothyronine; PTU, 6-A'-propyl-2-thiouracil; ( Fat cell ghosts were preincubated for 60 min at 4" in 1 rnM KHCO,, pH 7.4, buffer without or with 1, 5, or 50 PM guanyl-5'-yl imidodiphosphate. Following this incubation, the ghosts were washed once and resuspended in 1 mM KHCOI, pH 7.4. Adenylate cyclase activity was measured immediately by incubating the ghosts for 20 min at 37" in the absence or presence of 10 mM NaF or the indicated concentrations of norepinephrine in total volume of 100 ~1 containing 40 mM Tris-HCl (pH 8.0), 5 mM MgCl,, 30 mM KCl, 9 mM creatine phosphate, 0.5 unit of creatine phosphokinase, and 1 rnM ATP. Cyclic AMP was determined on 20-~1 aliquots taken after the reaction mixture was boiled for 3 min and then diluted to a volume of 1 ml. The cyclic AMP binding protein used in this assay was from the 10,000 x g,,, supernatant fraction of homogenized bovine adrenal glands and the assay was conducted as described by Brown et al. (11) to eliminate possible interference by ATP.
Cyclic AMP phosphodiesterase activity was measured by a modification of the method of Thompson and Appleman (14). The final volume of the assay mixture was 0. Laboratories. The (-)-stereoisomers, bitartram form, of norepinephrine, epinephrine, and isoproterenol were obtained from Sigma. The bitartrate had no effect on either total or specific binding of (-I-PHIDHA to these membranes (data not shown). The stock solutions of p-adrenergic agonists and antagonists were made fresh daily and contained 200 ELM metabisulfite to retard oxidation.
Metabiaulfite had no effect on either total or specific binding of (-)-PH]DHA to these membranes. The guanine nucleotide analog, guanyl-5'-yl imidodiphosphate, was purchased from ICN.

Fat cells isolated
from hypothyroid rats (maintained on an iodine-deficient diet and drinking water containing 0.00525% PTU) demonstrated no detectable stimulation of cyclic AMP accumulation over basal in response to epinephrine at ccmcentrations as high as 100 PM (Fig.  1). Epinephrine-stimulated lipolysis was clearly blunted in the fat cells from these animals.
The dose dependence for epinephrine-stimulated lipelysis is shifted to the right by approximately one order of magnitude, although no cyclic AMP accumulation was noted even at 100 PM epinephrine. Fig. 1  ling" between /3-adrenergic hormone receptor occupancy and activation of adenylate cyclase (20,21). We explored the possibility of restoring the reduced catecholamine-stimulated adenylate cyclase activity of the hypothyroid rat fat cell ghosts by prior incubation with Gpp(NH)p. Ghosts prepared from fat cells of euthyroid and hypothyroid rats were incubated for 60 Gpp(NH)p activation of adenylate cyclase (18,19,22,23). As further shown in Fig. 2, both activation of adenylate cyclase and the potentiation of norepinephrine-stimulated activity by Gpp(NH)p was reduced in fat cell ghosts from hypothyroid rats when compared to the response of fat cell ghosts from euthyroid rats. Additional experiments were performed in which fat cell ghosts were incubated with a maximal concentration of Gpp(NH)p (100 FM) under the adenylate cyclase assay conditions for a period of 20 min at 37". The Gpp(NH)pstimulated adenylate cyclase activity was 38 f 3% (n = 4) lower in the fat cell ghosts prepared from hypothyroid rats than in those prepared from the euthyroid rats, although sodium fluoride-stimulated activity was identical in both preparations (data not shown). Low K, cyclic AMP phosphodiesterase activity was measured in the same fat cell ghosts used in the adenylate cyclase assay of Fig. 2 (i.e. incubated at 4" for 60 min with and without Gpp(NH)p). As shown in Table I,2 cyclic AMP phosphodiesterase activity measured at 0.125 PM substrate was the same in fat cell ghosts prepared from hypothyroid and euthyroid rats with and without prior incubation of 50 /.LM Gpp(NH)p. As further shown in Table I, 1 mM I-methyl-Sisobutyl xanthine almost completely inhibited the cyclic AMP phosphodiesterase activity of fat cell ghosts prepared from euthyroid and hypothyroid rats. This same concentration of lmethyl-3-isobutyl xanthine failed to rectify the disparity in both the norepinephrine-stimulated adenyiate cyclase activity and its potentiation by 50 pM Gpp(NH)p between control and hypothyroid preparations (Table IP). Correze et al. (5) reported that the reduced maximal response of fat cell membrane adenylate cyclase from thyroidectomized rats to epinephrine, adrenocorticotrophic hormone, or glucagon could not be normalized to control levels by the presence of theophylline. These data suggest that hypothyroidism reduces lipolytic hormone-stimulation of adenylate cyclase but not its activation by fluoride. In addition, this reduction in hormonestimulated adenylate cyclase activity is not a reflection of increased cyclic AMP phosphodiesterase activity in purified fat cell membranes from thyroidectomized rats (5) or fat cell ghosts prepared from PTU-treated, hypothyroid rats (present study). P-Adrenergic receptors were next examined. An in vitro assay for the identification and characterization of p-adrenergic receptors has been reported utilizing the high specific activity labeled padrenergic antagonist, ['251]iodohydroxybenzylpindolol (24) or (-)-13H]dihydroalprenolol (25). Williams et al. (15) recently identified and characterized P-adrenergic receptors in fat cell membranes utilizing (-)-L3H]DHA.
was used in the present study in an attempt to probe the number and character of putative p-adrenergic receptors in membranes prepared from fat cells obtained from animals with altered thyroid status.
Binding of (-)-[3H]DHA to the fat cell membranes was a rapid process at 37" and reached equilibrium within 2 min (Fig. 3A*). The binding was saturated at approximately 80 nM (-)-["HIDHA (Fig, 3B). The maximum binding capacity in this experiment was 0.5 pmol of (-)-[3H]DHA bound/mg of protein of fat cell membrane. Half-maximal saturation was attained at 19 nM, an approximation of the equilibrium dissociation constant (K,) for this interaction of (-)-VHIDHA with its binding site(s).
Additional saturation studies were performed and analyzed by the method of Scatchard (26) (Fig. 42). Scatchard plots of the binding of (-)-[3H]DHA to fat cell membranes were uniformly curvilinear with an upward concavity in each of the 10 separate experiments analyzed by this procedure. Transformation of this data to a Hill plot is shown in the inset to Fig.  4. The Hill coefficient, nH, was 0.7, less than unity. A simple interpretation of these data would suggest negative cooperative site-to-site interaction among the binding sites or the existence of multiple populations of binding sites of differing affinities. The K, derived from the Hill transformation was 19 nM which is the same value obtained from Fig. 3B. However, the Hill coefficient and curvilinear Scatchard plot indicate these values are, at best, only approximations of the true affinity of this (these) binding site(s) for the radioligand.
As shown in Fig. 5,' (-)-propranolol, a potent P-adrenergic antagonist, produced half-maximal inhibition of (-VHIDHA binding to fat cell membranes at 50 nM. A final concentration of -700 nM (+)-propranolol was required to produce this same inhibition, a demonstration of the stereospecificity of the binding process. The specificity of the binding of ( -)-13HlDHA to fat cell membrane was further examined by competition studies utilizing p-adrenergic agonists (Fig. 6'). The concentration of (+)-isoproterenol required for half-maximal inhibition of ( -)-[3H]DHA binding was almost 3 orders of magnitude higher than that of (-)-isoproterenol.
The potency order of the (-)-p-adrenergic agonists for competition with ( -)$'HlDHA binding to fat cell membranes was (-)-isoproterenol 3 (-)norepinephrine = (-)-epinephrine, suggesting, according to Lands et aE. (271, a &-type adrenergic receptor. These data, confirming the study of Williams et al. (15), suggested that the number and affinity of p-adrenergic receptors in membranes prepared from fat cells of euthyroid, hypothyroid, and hyperthyroid rats could be assessed utilizing this methodol-WY.
As shown in Table III,2 hypothyroid animals displayed a reduction in mean body weight and a concomitant increase in parametrial adipose tissue mass. Treatment with triiodothyronine did not significantly alter mean body weight but did, however, reduce the parametrial adipose tissue mass by almost 50%. Despite these changes in the mass of adipose tissue, the total number of isolated fat cells and the total membrane protein obtained from these cells was the same in hypothyroid, hyperthyroid, and euthyroid rats. Specific binding of (-)-[3H]DHA to freshly prepared fat cell membranes from these animals is also shown in Table III. Specific binding of (-)+?H]DHA measured at 10 and 100 nM expressed in femtomoles bound/mg of protein was statistically the same in fat cell membranes prepared from hyperthyroid, hypothyroid, and euthyroid rats. Calculating specific binding of (-)-13HlDHA on a per cell basis from cell counts and membrane binding data suggests again that neither the T, treatment nor the PTU treatment in uiuo significantly alters the total number of (-)-[3H]DHA binding sites/fat cell over the control value measured at a saturating concentration of (-)-[3H1DHA.
Further studies examining the specific binding of c-l-[3H]DHA to fat cell membranes freshly prepared from euthyroid, hyperthyroid, and hypothyroid animals performed over a range of 1 to 100 nM (-)-13H1DHA concentrations are shown in Table IV. Fat cell membranes prepared from all three groups display essentially equivalent specific binding capacities for (-)-13HlDHA over this range of concentrations. Additional binding studies utilizing fat cell membranes prepared and frozen prior to use were performed with (--I-[3H]DHA at 12 concentrations yielding from 5 to 70% occupancy under equilibrium conditions. The results of these studies analyzed by the method of Scatchard (261, are shown in Fig. 7.' Scatchard plots of specific binding of (-)-13HlDHA to fat cell membranes prepared from hypothyroid (PTUtreated) and hyperthyroid (T,-treated) rats were curvilinear displaying upward concavity. Although the nonlinear nature of these plots prohibit accurate assessment of the total number of (-)-DHA binding sites, it is obvious from these data that the apparent density of binding sites is the same over this spectrum of (-)-[3H]dihydroalprenolo1 concentrations in fat cell membranes obtained from euthyroid (Fig. 4), hyperthyroid, and hypothyroid ( Fig. 7) rats. The marked shift to the right of the dose dependence of epinephrine-stimulated lipolysis in fat cells from hypothyroid animals ( Fig. 1) prompted us to extend our investigations to examine P-agonists competition for (-)-13H1DHA binding in fat cell membranes obtained from hypothyroid (PTU-treated) rats. Lefkowitz et al. (20) reported that purine nucleotides decrease the affinity of (-)-[3H1DHA binding sites of frog erythrocyte membranes for P-adrenergic agonists but not for antagonists. These observations suggest that alterations in padrenergic receptor affinity may not be discerned by the present approach and require competition studies utilizing padrenergic agonists. The blunted response of these fat cells to epinephrine may, in fact, reflect a change in affinity for agonists which was not detected by saturation studies performed with (+)-propranolol. Fig. 8' demonstrates that in competition studies with the potent P-agonist, (-)-isoproterenol, inhibition of (-)-[3H1DHA binding was the same when measured in fat cell membranes obtained from euthyroid (control) and hypothyroid (PTU-treated) rats.

DISCUSSION
This paper confirms the observations that hypothyroidism reduces the lipolytic response of adipose tissue (l-3) and isolated fat cells (4, 6) to submaximal concentrations of catecholamines and that this reduced lipolytic response of fat cells from hypothyroid rats is restored to control levels by administration of T, (6). Cyclic AMP accumulation in isolated fat cells in response to epinephrine could not be detected in the cells from hypothyroid rats and was higher than controls for cells from both the T,-treated groups and the hypothyroid rats subsequently treated with T,. Correze et al. (5) reported a reduction in the maximal response of adenylate cyclase to catecholamines in purified fat cell membranes obtained from rats following thyroidectomy.
This group of investigators further demonstrated that the amount of fluoride-activatable adenylate cyclase activity of fat cell membranes obtained from thyroidectomized rats was equal to that of preparations obtained from euthyroid rats. We confirmed these observations utilizing chemically induced hypothyroid rats and a fat cell ghost preparation rather than purified fat cell membranes.
Incubating fat cell ghosts obtained from euthyroid rats with Gpp(NH)p activated adenylate cyclase and enhanced the stimulation of this enzyme by norepinephrine.
However, this treatment failed to restore the reduced catecholamine-stimulated adenylate cyclase activity of fat cell ghosts from hypothyroid rats to the euthyroid level. In addition, incubating fat cell ghosts obtained from hypothyroid rats with a maximal concentration of this guanine nucleotide also failed to stimulate adenylate cyclase activity to the level demonstrated by the fat cell ghosts from the euthyroid rats. Thus, the reduced response of the hypothyroid rat preparations to Gpp(NH)p may be a reflection of an alteration of the guanine nucleotideactivating component of the adenylate cyclase which may, in turn, influence hormone stimulation.
Alternatively, it may reflect an alteration at some common point through which both catecholamines and guanine nucleotides activate the fat cell adenylate cyclase.
It has been suggested that thyroid hormones modulate hormone-dependent lipolysis in fat cells via regulation of the activity of low K, microsomal (5)