Identification of alpha-adrenergic receptors in uterine smooth muscle membranes by [3H]dihydroergocryptine binding.

[3H]Dihydroergocryptine, a potent alpha-adrenergic antagonist, was used to label smooth muscle membrane binding sites which have the characteristics expected of alpha-adrenergic receptors. Binding of [3H]dihydroergocryptine to rabbit uterine membranes was rapid and reversible with rate constants of 1.26 X 10(7) M-1 min-1 and 0.034 min-1 for the forward and reverse reactions, respectively. [3H]Dihydroergocryptine binding was of high affinity, with an equilibrium dissociation constant (KD) of 8 to 10 nM. Binding was saturable with 0.14 to 0.17 pmol of [3H]dihydroergocryptine bound/mg of protein at maximal occupancy of the sites. No cooperative interactions among the sites were detected. The specificity of the binding sites for a large number of adrenergic agonists and antagonists was identical with the specificity of alpha-adrenergic responses to these agents. The alpha-adrenergic agonist (-)-epinephrine competed for binding with a KD of 0.23 muM. The order of potencies for several adrenergic agonists in competing for the binding sites was (-)-epinephrine greater than (-)-norepinephrine greater than (-)-phenylephrine greater than (-)-isoproterenol in agreement with their alpha-adrenergic potencies. A series of 19 phenylethylamine adrenergic agonists competed for binding in a manner paralleling their potencies as alpha-adrenergic agonists. alpha-Adrenergic antagonists such as phentolamine (KD = 15 nM) and phenoxybenzamine (KD = 18 nM) potently competed for the binding sites. In contrast, beta-adrenergic antagonists such as propranolol (KD = 27,000 nM) and practolol (KD greater than 10(6) nM) did not have high affinity for the binding sites. A series of ergot alkaloids competed for [3H]dihydroergocryptine binding in a manner which paralleled their potencies as alpha-adrenergic agents. Competition for binding sites by alpha-adrenergic agonists and antagonists was a stereospecific process. The (-)-stereoi somers of epinephrine, norepinephrine, and ergotamine were at least 20- to 50-fold more potent than the corresponding (+)-stereoisomers. Compounds devoid of significant alpha-adrenergic activity, such as pyrocatechol, 3,4-dihydroxymandelic acid, normetanephrine, and D-lysergic acid, did not effectively compete for [3H]dihydroergocryptine binding sites. These rabbit uterine binding sites for [3H]dihydroergocryptine appear to have characteristics indistinguishable from those of the physiologically active alpha-adrenergic receptors.

J. LEFKOWITZ  [3H]Dihydroergocryptine, a potent a-adrenergic antagonist, was used to label smooth muscle membrane binding sites which have the characteristics expected of a-adrenergic receptors. Binding of [3H]dihydroergocryptine to rabbit uterine membranes was rapid and reversible with rate constants of 1.26 x lo1 M-I min ~' and 0.034 min' for the forward and reverse reactions, respectively.
[3H]Dihydroergocryptine binding was of high affinity, with an equilibrium dissociation constant (K,) of 8 to 10 nM. Binding was saturable with 0.14 to 0.17 pmol of [3H]dihydroergocryptine bound/mg of protein at maximal occupancy of the sites. No cooperative interactions among the sites were detected.
The specificity of the binding sites for a large number of adrenergic agonists and antagonists was identical with the specificity of cu-adrenergic responses to these agents. The cr-adrenergic agonist (-)-epinephrine competed for binding with a K,, of 0.23 PM. The order of potencies for several adrenergic agonists in competing for the binding sites was (~ )-epinephrine > (-)-norepinephrine > (-)-phenylephrine > (-)-isoproterenol in agreement with their cu-adrenergic potencies. A series of 19 phenylethylamine adrenergic agonists competed for binding in a manner paralleling their potencies as a-adrenergic agonists. a-Adrenergic antagonists such as phentolamine (K a = 15 nM) and phenoxybenzamine (Ka = 18 nM) potently competed for the binding sites. In contrast, P-adrenergic antagonists such as propranolol (K, = 27,000 nM) and practolol (K, > lo6 nM) did not have high affinity for the binding sites. A series of ergot alkaloids competed for [3H]dihydroergocryptine binding in a manner which paralleled their potencies as cu-adrenergic agents. Competition for binding sites by cu-adrenergic agonists and antagonists was a stereospecific process. The (-)-stereoisomers of epinephrine, norepinephrine, and ergotamine were at least 20-to 50.fold more potent than the corresponding (+)-stereoisomers. Compounds devoid of significant a-adrenergic activity, such as pyrocatechol, 3,4-dihydroxymandelic acid, normetanephrine, and n-lysergic acid, did not effectively compete for [3H]dihydroergocryptine binding sites. These rabbit uterine binding sites for [3H]dihydroergocryptine appear to have characteristics indistinguishable from those of the physiologically active oc-adrenergic receptors.
A variety of physiological processes are regulated by the order to elicit the observed (Y or p response. u-Adrenergic endogenous catecholamines, epinephrine and norepinephrine. responses include contraction of smooth muscle in vascular, Ahlquist (1) observed that the adrenergic responses to cate-uterine, and other tissues (2). Such responses are stimulated by cholamines could be divided into two distinct groups, 01 and 0, catecholamines with a typical potency series of epinephrine > based on the order of potency of a series of adrenergic agonists norepinephrine > phenylephrine > isoproterenol (1) and are in eliciting the responses. This classification was further blocked by specific "Lu-adrenergic" antagonists such as phenstrengthened by the observation that one group of competitive tolamine, phenoxybenzamine, and ergot alkaloids (3). Typical adrenergic antagonists specifically blocked cu-adrenergic re-@-adrenergic responses are smooth muscle relaxation and sponses, while another group of compounds specifically stimulation of cardiac contractility (2). Such responses are blocked /3-adrenergic responses. The remarkable specificity of stimulated by catecholamines with a typical potency series of each type of response suggested that catecholamines must isoproterenol > epinephrine >_ norepinephrine > phenylephinteract with distinct 01 or p receptors in the target tissue in rine (1) and are blocked by specific "P-adrenergic" antagonists such as proprl: :olol and dichlorisoproterenol (4,5). *This study was supported by Health, Education and Welfare In recent years the molecular mechanism of P-adrenergic re-Grant HL 16037 and by a grant-in-aid from the American Heart ceptor activation has been studied extensively. Many if not all Association with funds contributed in part by the North Carolina fl-adrenergic responses are mediated by intracellular cyclic Heart Association.
$ Student in the Medical Scientist Training Program supported by AMP, the formation of which is catalyzed by the enzyme National Institutes of Health Grant 5T 05-6 M01678.
adenylate cyclase (6,7). Adenylate cyclase appears to be $ Established Investigator of the American Heart Association. localized in the plasma membrane (8) (14). These techniques have provided a direct approach to studying the fl-adrenergic receptor as a physicochemical entity in a variety of tissues.
In contrast with studies of the P-adrenergic system, little information has been reported on the molecular mechanism of a-receptor-mediated adrenergic stimulation. Some evidence has suggested that cu-adrenergic receptors may be coupled to processes which elevate intracellular cyclic GMP (15) or may be related to regulation of cation permeability of the plasma membrane (16). However, to date most of the reported information about the characteristics of (Y receptors has been inferred from measurement of physiological responses to adrenergic agents, and no previous attempts have been successful in directly identifying the a-adrenergic receptor by binding techniques. Identification of a-adrenergic receptors by radioligand binding would require fulfillment of several criteria which are based on the known characteristics of a-adrenergic responses.
1. The affinity of the radioligand for the binding sites should reflect the known biological potency of the ligand.
2. The time course of the binding of radioligand to receptor sites should be consistent with the rate at which the ligand elicits its physiological response. Similarly, if the physiological effect of the ligand is reversible, the binding of radioligand to receptor sites should be reversible.
3. Radioligand binding should reflect an ol-adrenergic specificity. a-Adrenergic agonists and antagonists should compete for the binding sites in an appropriate order of affinity. Specific P-adrenergic agents should have a much lower affinity for the sites than cu-adrenergic agents. Compounds devoid of CYadrenergic activity should not interact with the binding sites. Binding sites should exhibit stereospecificity toward adrenergic agents since (-)-isomers of a-adrenergic agonists and antagonists are considerably more potent than (+)-isomers in physiological systems (2,17,18).
We have recently described a method for the direct identification of a-adrenergic binding sites in uterine smooth muscle membranes (19). The binding sites are identified by measuring the binding of the potent a-adrenergic antagonist, [3H]dihydroergocryptine, to uterine membranes. We now report a detailed analysis of the kinetics, affinity, and specificity of the interaction of [3H]dihydroergocryptine with its binding sites in smooth muscle membranes from rabbit uterus. This approach provides a method for the direct study of the molecular nature of cu-adrenergic receptors.

MATERIALS AND METHODS
Radioligund-[3H]Dihydroergocryptine ( Fig. 1) was chosen as the cu.adrenergic radioligand for this study for several reasons. First, the compound can be prepared by reduction of the double bond at position 9,lO of ergocryptine by established procedures (20 Inc. and maintained at -70" were used as a source of smooth muscle. These uteri gave binding results indistinguishable from those obtained with membranes prepared from fresh tissue. Uteri were cleaned of fat, opened longitudinally, and stripped of endometrium using a scalpel. Uteri were then minced and homogenized in ice cold buffer (0.25 M sucrose, 1 mM MgCl,, 5 mM Tris, pH 7.4) for four 10-s periods using a Tekmar tissue grinder at high speed. After filtration through a single layer of gauze, the homogenate was centrifuged at 400 x g for 10 min at 4" and the pellet was discarded. The supernatant was centrifuged at 28,000 x g for 10 min at 4". The resulting pellet was washed twice in ice cold incubation buffer (10 rnM MgCl,, 50 mM Tris, pH 7.5) by resuspension and centrifugation at 28,000 x g for 10 min. The final pellet was resuspended in incubation buffer for use in the binding assay.
Binding Assay-[3H jDihydroergocryptine (8 nM unless otherwise specified) and uterine membranes (-4 mg/ml unless otherwise specified) were incubated for 15 min (unless otherwise specified) at 25" with shaking in a total volume of 150 ~1 of incubation buffer.
is the concentraion of (3H]dihydroergocryptine in the assay (8 nM), KDHE is the K, for dihydroergocryptine (8 nM) computed from equilibrium binding studies (Fig. 2), and EC,, is the concentration of the competing compound which inhibits 50% of the [3H]dihydroergocryptine binding. Hence under these conditions K, = EC,,/2.
An alternative calculation of these data as previously described (11) using the method of Rodbard and Lewald (23)  were prepared at 6 x lo-' M in absolute ethanol and diluted in water. Ergotaminine was dissolved at 6 x 10e3 M in glacial acetic and diluted to 6 x lOmE M in water for use in the assay. All other stock solutions were made in water. All stock solutions were prepared less than 2 h before use in the binding assay.

Number
and Affinity of Binding Sites-The binding of [3H]dihydroergocryptine to rabbit uterine membranes was a saturable process ( Fig. 2) (24) demonstrates a single order of sites with a KD of 10 nM (Fig. 2, inset). The intercept of this plot (Fig. 2, inset) with the abscissa provides an estimate of the number of binding sites in rabbit uterine membranes (n = 0.17 pmol/mg of protein). No evidence of cooperative interactions is apparent from Fig. 2  Kinetics of Binding-The binding of [Wldihydroergocryptine to uterine membranes was rapid (tH = 4.5 min) (Fig. 3) and reversible (Fig. 4) at 25". Since the concentration of radioligand (8 nM) was much greater than the binding site concentration in the assay (0.76 nM), the forward reaction could be considered to be a pseudo-first order reaction which depends on the binding site concentration. Taking into account the reversible nature of the binding reaction, the reaction can be described by In [X,,/(X,, ~ X)] = k,, t where X is the amount of [3H]dihydroergocryptine bound at each time (t), and X,, is the amount bound at equilibrium (0.4 nM). Thus the line in Fig. 3  where 12, is the rate constant for the reverse (dissociation) reaction (Fig. 4) and [DHE] is the concentration of [3H]dihydroergocryptine in the reaction mixture. The reverse reaction rate constant, k,, is 0.034 min-' (determined by linear regression analysis (Fig. 4)) (F = 0.97). The ratio k,@, = 3 nM is a kinetically derived estimate of the K, for the reaction of [3H]dihydroergocryptine with its binding site. This value is comparable to the K, (8 to 10 nM) determined by equilibrium studies (Fig. 2) Specificity of Binding-Adrenergic agonists competed for the [3H]diLydroergocryptine binding sites (Fig. 5) in an order of potency: (-)-epinephrine > (-0-norepinephrine > (-)phenylephrine > > ( P)-isoproterenol. This potency order is identical with the well known order of potency for these agents in eliciting physiological ol-adrenergic responses (1,2,17). This order is in contrast with the potency order previously demonstrated for competition of the /3-adrenergic receptor: (-)-isoproterenol, (-)-epinephrine 2 (-)-norepinephrine > > (-)phenylephrine (9-14). The a-adrenergic agonists had high af-  *In addition to the indicated substituents, methoxamine has methoxy substituents at positions 2 and 5 of the phenyl group. ** Two methyl groups are substituted on the a-carbon of mephentermine. as an a-antagonist and the noncompetitive irreversible nature of phenoxybenzamine as an a-adrenergic antagonist (4). The specific competitive Lu-adrenergic antagonists dibozane and yohimbine also competed for the binding sites with KD values of 0.13 fiM and 0.22 pM, respectively.
In contrast to the marked potency of these oc-adreneric antagonists in competing for the [3H]dihydroergocryptine binding sites, P-adrenergic antagonists such as propranolol, dichlorisoproterenol, and practolol competed for the sites only at very high concentrations (Fig. 6, Table II). These results agree well with the previously reported finding that in physiological studies P-adrenergic antagonists such as propranolol and dichlorisoproterenol weakly block a-adrenergic receptors with K, values of 7 pM and 5 pM, respectively (27).
A variety of ergot alkaloids were tested for their abilities to compete for the [3H]dihydroergocryptine binding sites. Affinities as assessed by binding studies (Table II) (Table  II). Ergotamine, a potent a-adrenergic agent, competed for binding with a K, of 21 nM (Fig. 7A, Table II). Dihydroergotamine, the oc-adrenergic antagonist prepared by reduction of ergotamine, had slightly higher affinity than ergotamine for the binding sites. The (+)-stereoisomer of ergotamine, which is called ergotaminine, did not have high affinity for the binding site (Fig. 7A) thus demonstrating the stereospecificity of the binding sites toward the ergot alkaloids as well as toward catecholamines (Fig. 7A). Ergonovine, an amine ergot alkaloid with a-adrenergic agonist activity (28) also competed for the binding sites. In the ergotoxin group of alkaloids (Fig. 7B) the dihydroergot alkaloids which are reduced at position 9, 10 were more potent in competing for the binding sites than the corresponding native ergot alkaloids. This is consistent with physiological studies (3, 21) which have shown that the reduced ergot alkaloids are more potent as ol-adrenergic antagonists than the native alkaloids. The ergot alkaloid demonstrating the highest affinity for the receptor was dihydroergocryptine (K,-10 nM). The KD for dihydroergocryptine determined by competition binding (Fig. 7B) agrees well with that determined by equilibrium binding studies with [3H]dihydroergocryptine (Fig. 2). This further confirms the biological equivalence of [3H]dihydroergocryptine and dihydroergocryptine at the a-adrenergic receptors. Dihydroergocryptine was lo-fold more potent than ergocryptine (Fig. 7B) (Table II). Thus chlorpromazine, trifluoperazine, and azapetine inhibited binding with K, values in the range of 0.25 to 1.25 pM. In addition, the vasoactive amine serotonin competed for binding with a K, of 20 PM. Histamine had a low affinity for the binding sites (K, = 300 PM).
Clonidine, an a-adrenergic agonist, inhibited binding with a K, of 0.26 WM.

DISCUSSION
These data demonstrate that the a-adrenergic antagonist, can be used to identify uterine smooth muscle binding sites which have the characteristics expected of physiological Lu-adrenergic receptors. Although the existence of a-adrenergic receptors has been postulated for over two decades (l), information about the receptor could only be inferred from physiological studies of the abilities of various agents to stimulate or block smooth muscle contraction. By contrast, the radioligand binding methods used here permit for the first time the direct study of the interaction of a-adrenergic agonists and antagonists with the receptor binding sites.
The binding sites identified with [3H]dihydroergocryptine appear to satisfy the criteria which must be met for identification of physiological ol-adrenergic receptors. The kinetics and affinity of the receptor ligand interaction are appropriate. There is remarkable agreement in the specificity of binding and the specificity of the a-adrenergic effects of the 50 compounds tested.
The group of structurally related phenylethylamine compounds demonstrated a distinct specificity in competing for [SH]dihydroergocryptine binding (Tables IA and IB). Several generalizations can be made about the structure-activity relationships which determine the affinities of these compounds for the binding sites.
Aromatic Ring Substituents-Compounds such as (-)-epinephrine and (-)-norepinephrine with hydroxyl substituents at positions 3 and 4 on the aromatic ring had higher affinity for the receptor than analogous compounds lacking one or both hydroxyl substituents (Table IA). The next most potent configuration with respect to phenyl substituents was the group with only a 3-hydroxyl substituent (phenylephrine, metaraminol, ethylnorphenylephrine). Methoxamine, which has methoxy groups at positions 3 and 5 was slightly less potent than the 3-hydroxy compounds. Methoxamine is of particular interest since physiologically it is a relatively specific (Yadrenergic agonist and is not susceptible to catecholamine uptake processes (27,29). Compounds with only a 4-hydroxy substituent (octopamine, tyramine, deoxyisoproterenol) had somewhat lower affinities as did the compounds with no ring substituents (ephedrine and mephentermine).
Alkyl Substituents on Amino Nitrogen-( -)-Epinephrine, which has a methyl substituent on the amino nitrogen had a slightly higher affinity than the corresponding compound with no substituent on the amino nitrogen (( -)-norepinephrine).
A further increase in size of the amino nitrogen substituent to an isopropyl group (( -)-isoproterenol) markedly reduced the affinity for the binding site (Table IA). A comparable reduction in the physiological potency of these compounds when the amino substituent is increased from a methyl to an isopropyl group has been reported (17, 18). Among the 3-hydroxy compounds a similar relationship is apparent with respect to increasing the size of the amino substituent, i.e. the compound Identification of a-Adrenergic Receptors with a methyl substituent on the amino nitrogen (phenylephrine) had a 7-fold higher affinity than the analogous compound with an ethyl group on the nitrogen (ethylnorphenylephrine). Similarly for the 4..hydroxylphenyl compounds, increasing the size of the amino substituent from hydrogen (octopamine) to isopropropyl (deoxyisoproterenol) resulted in a significant reduction in affinity. Hence the order of affinity of compounds with respect to alkyl substituents on the amino nitrogen appears to be CH, > H > CH, CH, > CH(CH,),. This is in agreement with the physiological potency order of these compounds as ol-adrenergic agonists (30).

Aralkyl
Substituents on Amino Nitrogen-In physiological studies it has been demonstrated (17, 18) that substitution of large aralkyl groups on the amino nitrogen eliminates the activity of the compound as an a-adrenergic agonist, while enhancing its affinity as an a-adrenergic antagonist. The K, values derived from inhibition of [3H]dihydroergocryptine binding (Table IB) by these compounds are in close agreement with their reported K, values as a-adrenergic antagonists. &Carbon Hydroxyl'Group-The highest affinity phenylethylamine agonists contain a P-carbon hydroxyl group which is in the levo (-)-stereoconfiguration (e.g. (-)-norepinephrine)(Table IA). The presence of a P-hydroxyl in the dextro (+)-configuration ((+)-norepinephrine) or the absence of a P-hydroxyl group (dopamine) is accompanied by a 30-fold loss of affinity. This stereospecificity is also manifest in the physiological response to these agents (2,17).
Several previous attempts by others to directly label aadrenergic receptors using radioligands did not succeed in labeling sites which have the appropriate characteristics (31-34). Possible reasons for the success of the present study in contrast with previous attempts at a-receptor identification are as follows.
1. In general, previous attempts have utilized radioactively labeled ligands of relatively low specific radioactivity (-25 to 50 mCi/mmol). Thus in order to identify the true receptors, of which there are a very small number, high concentrations of ligand were required. At these high concentrations, nonspecific binding of ligands to high capacity low affinity nonreceptor sites presumably obscured binding to the small number of physiological high affinity receptors. The ligand used in the present study ( [3H]dihydroergocryptine) not only possesses high specific radioactivity (23 Ci/mmol) but has very high affinity for the ol-adrenergic receptor. Hence at the low concentrations of ligand used in the assays (-8 nM) specific o( receptor binding could be detected in the absence of significant nonspecific binding. 2. In several previous attempts to identify a-adrenergic receptors by radioligand binding, intact strips of tissue were used as a source of receptors (31,32). The use of membrane preparations in the present study might allow the attainment of higher receptor site concentrations and may also eliminate some of the drug and hormone uptake processes which occur in intact tissue (29).
The techniques described in this report for the identification of cu-adrenergic receptors by [3H]dihydroergocryptine binding may also be of value in the study of cu-receptors in tissues other than uterine smooth muscle. In each tissue and species studied, however, a careful documentation of the specificity of binding should be undertaken. It has been reported that under appropriate conditions in certain tissues ergot alkaloids interact with pharmacological responses other than cu-adrenergic responses (4). Of particular interest in this regard, is that Davis et al. ' have demonstrated that [3H]dihydroergocryptine binds to sites in rat brain membranes which have a specificity pattern somewhat different from that of the rabbit uterine binding sites reported here. The specificity of the binding in the brain membranes is such as to suggest that either several receptors are being labeled (including cy receptors, dopamine receptors, and serotonin receptors) or that a receptor of hybrid specificity is involved. Hence, a careful study of binding specificity should precede application of these techniques to the study of a-adrenergic receptors in tissues other than rabbit uterine smooth muscle.