Evidence That the Gh Protein Is a Signal Mediator from al-Adrenoceptor to a Phospholipase C I. IDENTIFICATION OF a,-ADRENOCEPTOR-COUPLED Gh FAMILY AND PURIFICATION OF Gh7 FROM BOVINE HEART*

Our previous studies on al-adrenoceptor-mediated signaling suggested that Gh is a signal mediator. Gh consists of a 74-kDa GTP-binding a-subunit and a 50- kDa ,%subunit. Studies using the a,-agonist-receptor-G-protein ternary complexes from various tissues and species revealed that the intensity (GTP-binding) of the [a-s2P]GTP-labeled proteins resulting from acti- vating the al-receptor was significantly attenuated by phentolamine. The molecular masses of GTP-binding proteins were 74 kDa in rat heart and liver, 77 kDa in dog heart, 78 kDa (Gh7d in bovine heart and liver, and 80 kDa in human heart. Supporting these observations, a specific antibody to Gh7. not only recognized these GTP-binding proteins in the ternary complex preparations, but also co-immunoprecipitated al-adrenocep- tors, indicating a tight association of these GTP-bind-ing proteins with the al-adrenoceptor. These results also demonstrate that functional and structural simi-larities exist among these GTP-binding proteins. Ad- ditionally, one of the identified G-proteins (termed Gt.7) was purified from bovine heart. Gh7 consisted of the 78-kDa GTP-binding protein and

binding regulatory proteins (G-proteins) and various effectors (3)(4)(5)(6)(7)(8). Pharmacological studies of the al-adrenoceptors have indicated that at least two subtypes of the a,-receptors exist; these subtypes are designated aIS and (Ylb. Based on biochemical studies with various tissues and cell types, two signal pathways of the al-receptors can clearly be observed. Stimulation of alb-receptor leads to the formation of inositol 1,4,5triphosphate and diacylglycerol via activation of a phospholipase C through a toxin-insensitive G-protein (5, 9). The formation of arachidonic acid via activation of phospholipase A*, is stimulated by the ala-receptor through a pertussis toxinsensitive G-protein (5,lO). Recently, the al-receptor subtypes have been cloned including ala-, alb-, and al,-receptors (6, 7, 11, 12), and a new member designated as type aid, cloned by Perez et al. (13). Transfection of cells with cDNAs encoding the ale-, (Ylb-, and tylc-receptors stimulated the hydrolysis of phosphatidylinositides via phospholipase C-Dl through the activation of the Gq family (14) (see also Refs. 15 and 16).
We previously reported that in rat liver al-receptor (possibly the (Ylb type) coupled to a 74-kDa GTP-binding protein. This GTP-binding protein was identified by inducing the alagonist-receptor-G-protein ternary complex and by direct photoaffinity labeling of the G-protein in the ternary complex with radiolabeled GTP (17). The isolated Gh consisted of the 74-kDa GTP-binding protein (a-subunit) and a -50-kDa protein (@-subunit) (18). We have also shown that Gh coupled to an al-receptor (18) and stimulated a membrane-bound phospholipase C in an in vitro reconstitution system (19).
In this report, we have extended our studies in order to further evaluate the coupling ability of the al-receptor with Gh. The following studies have been described. 1) A new high molecular mass GTP-binding regulatory protein (Gh7)' was purified from bovine heart; 2) (-)-epinephrine-al-receptor-Gh ternary complexes from various species were induced and isolated; 3) Gh?=-specific antibody was used to assess the direct interaction of the Gh family with the al-receptor. Herein, we have demonstrated that the Gh family which couples to the al-adrenoceptor exhibits different molecular masses in various species.

EXPERIMENTAL PROCEDURES
Materials-Sucrose monolaurate (SM-1200) was a gift from the Mitsubishi-Kasei Company (Tokyo, Japan). Lubrol PX and Nonidet Gh7: The term Gh7 is used in anticipation of an increase in the number of Gh family proteins from different species, because this Gprotein is functionally and structurally similar to GI,. We presented this G-protein as G, at the 1992 AHA meeting in New Orleans, because earlier data from Gh7= antibody experiments and cross-coupling studies were not available. P-40 were from Sigma. Protein A-agarose, guanine nucleotides, and other nucleotides were obtained from Boehringer Mannheim. The column chromatographic resins were obtained from Pharmacia LKB Biotechnology Inc. [cY-~'P]GTP (3000 Ci/mmol) and 1261-labeled protein A (30 mCi/mg of protein A) were obtained from Amersham Corp., and [35S]GTPrS (-1300 Ci/mmol) and [3H]prazosin (76 Ci/ mmol) were from DuPont NEN. Other chemical and biochemical materials were used as described previously (17)(18)(19).
Membrane Preparation from Various Sources-The rat and bovine liver membranes were prepared by Percoll gradient centrifugation using the method of Prpic et al. (20). Bovine hearts (4 kg), obtained from Pel-Freez Biologicals (Rogers, AR), were minced, homogenized briefly with a Waring blender, and then with a mechanical homogenizer (Ultra-Turrax, Janke & Kunkel) in 30 liters of 10 mM Hepes buffer (pH 7.5) containing 250 mM sucrose, 5 mM EGTA, and protease inhibitors (bacitracin, 2 pg/ml; benzamidine, 100 pg/ml; leupeptin, 2 pg/ml; pepstatin A, 2 pg/ml; trypsin inhibitor, 2 pg/ml; phenylmethylsulfonyl fluoride, 2 pg/ml; and antipain, 20 pg/ml). The homogenate was filtered through four layers of cheesecloth and centrifuged at 500 X g for 5 min. The supernatant was collected and centrifuged at 40,000 X g for 1 h. Pellets were harvested and washed three times with 50 mM Hepes buffer (pH 7.5) containing 10 mM MgCl,, 5 mM EGTA, and the protease inhibitors listed above. Crude membranes from various species and various tissues of rat were prepared using essentially the same protocol. The membranes were suspended at 10 mg/ml protein concentration in the same buffer containing 10% glycerol and 1 mM dithiothreitol and stored at -80 "C until use. The heart tissues were obtained from various sources in different lengths of storage periods. Human heart tissue was obtained from the heart transplantation program at the Cleveland Clinic Foundation. Rat tissues were prepared freshly, whereas dog and bovine tissues were obtained from Pel-Freez.
Purification of 78-kDa GTP-binding Protein (GhJ from Bovine Heart-The 78-kDa GTP-binding protein (a-subunit of Gh7) was isolated by previously described Gh purification method with some modifications (18) (see also Ref. 19). The following modifications were made. 1) Protein was solubilized without detergent, but the purification of the Gh7 protein was carried out in the presence of 0.1% sucrose monolaurate, 2) Q-Sepharose ion-exchange resin was used for the first step instead of heparin-agarose, and 3) the isolation of Gh7 was achieved without using hydrophobic resin. Throughout the purification, the 78-kDa protein was monitored by photoaffinity labeling with [a-3ZP]GTP as well as by [36S]GTPrS binding. The purifications were carried out at 4 "C, and protease inhibitors were included in the buffers listed above. Glycerol (10%) and 0.1% sucrose monolaurate were included in HED buffer (20 mM Hepes, 1 mM EGTA, and 0.5 mM dithiothreitol, pH 7.5) to stabilize the proteins throughout the purification. To assess the subunit association of the 78-kDa GTP-binding protein, the protein was purified under nonactivated conditions. The protocol and results described here are representative of several independent purifications.
The crude membranes (10 g of protein) were washed with ice-cold HED buffer containing protease inhibitors, as described above. After centrifugation at 40,000 X g for 40 min, the pellets were collected, resuspended at 5 mg/ml protein in the same buffer (2 liters) supplemented with 250 mM NaCl, then solubilized with gentle agitation for 1 h at 4 'C. After centrifugation at 40,000 X g for 1 h, the supernatant (1,120 ml) was collected and diluted 4-fold with HED buffer containing 0.1% sucrose monolaurate and 10% glycerol. The sample was then applied to a Q-Sepharose column (3.6 X 20 cm) which had been equilibrated with HED buffer containing 70 mM NaC1. The column was washed with 600 ml of the equilibration buffer. The retained materials were eluted using 1,000 ml of a linear salt gradient (50-700 mM) in the same buffer, and 9-ml fractions were collected at a flow rate of 30-40 ml/h. The GhT-COntaining fractions (135 ml, fractions 49-63 of peak 11) were pooled and concentrated to 5-7 ml using Amicon PM-30 membranes. The sample was applied to an Ultrogel AcA 34 column (450 ml) that had been equilibrated with HED buffer containing 100 mM NaCl. The column was eluted overnight with the same buffer, and 4.5-ml fractions were collected at a flow rate of 25 ml/h. The pooled fractions (76 ml, fractions 49-65) which contained Gh, protein were diluted with 80 ml of HED buffer and loaded onto a Q-Sepharose column (1.4 X 10 cm). The column was washed with 70 ml of HED buffer containing 250 mM NaCl. Elution was achieved using 80 ml of a linear sodium chloride gradient (200-700 mM). Fractions (1.5 ml) were collected at a flow rate of 30 ml/h. The Gh7containing fractions (27 ml) were pooled and diluted 7-fold with HED buffer. The sample was applied to a hydroxylapatite column (3-5 ml) which had been equilibrated with HED buffer containing 50 mM NaC1. After washing the column with the equilibration buffer (15 ml), the bound materials were eluted using 40 ml of a linear phosphate gradient (0-100 mM), and 1 ml fractions were collected.
Purification of Gh"Ch was purified from rat liver membranes isolated by Percoll gradient centrifugation, as described previously (18,19).

al-Agonist-Receptor-G-protein Ternary Complex
Formation and Purification-The ternary complex comprised of (-)-epinephrine, a1receptor, and G-protein was induced by incubating the membranes with (-)-epinephrine. Thus, the rat liver and heart, bovine liver and heart, and dog and human heart membranes (1 g of protein) were preincubated for 3 h with 5 X loe6 M (-)-epinephrine, M (+)propranolol, and lo-' M rauwolscine. The complexes then were solubilized using 0.2% sucrose monolaurate in HED containing 100 mM NaCl and isolated using heparin-agarose and wheat germ agglutinin-agarose by the protocol previously reported (17). For experiments, the ternary complex preparations (100 pl) were used without adjusting the amount of the al-receptors and G-proteins. GTPase Assay-GTPase activity was measured essentially as described previously (18). Briefly, Gh or Gh7 (20 pmol) was incorporated into phospholipid vesicles using a Sephadex G-50 column (0.6 X 15 cm). The incorporation of Gh and Gh7 into vesicles was 65-70%. Vesicles (300-400 pl) containing G-proteins were mixed with 0.25 p M GTP and 5 pCi of [-p3'P]GTP, and release of phosphate from [Y-~'P] GTP was measured using Norit A charcoal (5% suspension, w/v). rH]Prazosin and f5S]GTPrS Binding Assay and [CX-~~PIGTP Photoaffinity Labeling-These assays were performed essentially as described previously (17, 18). Briefly, the amount of al-adrenoceptor was quantitated by specific [3H]prazosin binding after incubation at room temperature for 1 h. The total amount of Gh or Gh7 in 0.05% sucrose monolaurate solution was determined by means of the binding of [35S]GTPyS in the presence of 1-2 mM MgC12. The photoaffinity labeling of G-proteins was performed in the presence of 5-20 pCi of [w~'P]GTP and 0.5-2 mM MgClz in an ice bath under 254-nm UV irradiation for 6-10 min.
Antibody Experiments-A polyclonal antibody to Gh7-was generated in New Zealand White rabbits. Gh7* was separated using a Q-Sepharose column as described (19). Seventy micrograms of (200 pl) were emulsified with an equal volume of complete Freund's adjuvant and injected subcutaneously into the rabbits. At 3-week intervals, three booster injections were given with 70-100 pg of Gh7* and incomplete Freund's adjuvant. Rabbit antisera were characterized by immunoblots using the methods of Harris et al. (21) or by immunoprecipitation. Briefly, for Western blots, proteins were separated on 7% gel by SDS-PAGE and then transferred to Immobilon-P (Millipore, Bedford, MA). Immunoblots were incubated in LDB solution (low detergent blotto, 80 mM NaC1, 2 mM CaCI', 0.02% NaN3, 0.2% NP-40, 50 mM Tris/HCl, pH 8.0, containing 5% nonfat dry milk) for 2 h at room temperature to block nonspecific binding, then transferred to LDB containing antibody (1:500 dilution), and incubated for 1 h at room temperature. After washing three times with LDB, the immunoblots were incubated with lZ5I-labeled protein A (0.2 pCi/ml) in HDB (high detergent blotto, 2% Nonidet P-40 in LDB) for 1 h at room temperature. After washing intensively with HDB and non-detergent blotting buffer, the dried blots were subjected to autoradiography on Kodak XAR-5 x-ray film with intensifying screens for 1-2 days. For immunoprecipitation, Gh7, Gh, the a,agonist-receptor-G-protein ternary complex preparations and solubilized tissue membranes were photoaffinity-labeled with 10 GCi of [w3'P]GTP in the presence of 0.1 mM App(NH)p and 2 mM MgC1, (17,18) and then incubated for 2 h with 5 pl of Gh7= antibody at room temperature. The antigen-antibody complexes were precipitated using 10 p1 of protein A-agarose (binding capacity, 22 mg of rabbit IgG/ml of agarose gel). The pellets were collected by centrifugation at 3000 rpm and washed three times with 20 mM Hepes buffer (pH 7.4) containing 500 mM NaCl and 0.01% sucrose monolaurate. The samples were denatured by boiling in the presence of Laemmli buffer (22) and subjected to SDS-PAGE (7% gels) and autoradiography for 1-2 days. mined by the method of Bradford (23) using a Bio-Rad protein Protein Determinution-The protein concentration was deterdetermination kit and bovine serum albumin as a standard.  we tested various conditions; salt concentrations were critical for increasing the specific solubilization. The optimal concentration of salt was 200-400 mM in HED buffer. An increase in the salt concentration (400-2000 mM) did increase the G T P r S binding, but the specific binding activity was decreased because the amount of total protein extracted from the membranes also increased. At 250 mM NaCl the extraction of G-proteins and other proteins from the membranes reached approximately 13-14% and 10-11%, respectively. When the pellets were resuspended and photolabeled with 10 pCi of [a-32P]GTP in the presence of 2 mM MgC12, the 78-kDa GTPbinding protein was not detectable in most experiments, suggesting that the protein was substantially solubilized (data not shown). It should be noted that the 78-kDa GTP-binding protein could be solubilized with salt alone, but in the absence of the detergent, sucrose monolaurate, the protein became easily aggregated and lost the ligand binding activity within less than 1 week. The elution profile of the 78-kDa GTPbinding protein was similar to that of Gh purified from rat liver membranes with Q-Sepharose (Fig. lA). However, the 78-kDa protein from gel filtration column was eluted in somewhat later fractions (at least more than 10 fractions) than Gh (Fig. lB,see Ref. 18). The reasons for this later elution are not clear yet. Washing of the second Q-Sepharose column with high salt (250 mM) resulted in separation of most protein from the 78-kDa GTP-binding protein which was obtained with >25% purity (Fig. 1C). Further purification and concentration of the 78-kDa GTP-binding protein were achieved using a hydroxylapatite column, as shown in Fig. 1D.

Purification
The overall results of the Gh7 protein purification scheme are summarized in Table I. Based on the specific [35S]GTPySbinding activity, the Ghla protein obtained after the last purification step was of 80% purity. The yield of was -20% compared with that of salt extraction. Based on the final recovery, the amount of Ghlo was 0.26% of total Gproteins and 0.0065% of the total protein in the membranes. The 78-kDa GTP-binding protein (Gh7-) was copurified with a BO-kDa protein (Gh7& which has the same molecular mass as GhS and does not bind GTP (see Fig. 3, A and B ) . Differences in the molecular masses between Gb and Gh7 were seen only with the GTP-binding proteins. The purified protein was stable in this buffer for less than 3 weeks without the addition of a G-protein stabilizer. However, when the aluminum fluoride was included, the protein was stable for a month.
Biochemical Properties of Gh7-The specificity of the nucleotide binding was determined by the photolabeling method tography. See "Experimental Procedures" and "Results" for details. These results demonstrate that Gh7 is a specific GTP-binding protein and has an affinity for the ligands similar to Gh (18). To further evaluate the properties of Gh7, we compared the magnesium ion requirement for GTPyS-binding of Gh and Gh7. As shown in Fig. 2, when the GTPyS-binding assays were carried out in 0.05% Lubrol PX solution, the maximal ligand binding was observed at 1-3 mM M&12 for Gh7 and 10-20 mM for Gh which was similar to that previously reported (18). The ligand binding by Gh7 was subsequently inhibited -55% of the maximal when the MgZ+ concentration was further increased. On the other hand, when the GTPyS-binding experiments were performed in 0.05% sucrose monolaurate solution, the maximal ligand binding by Gh7 was obtained at <0.2 mM MgC12 and did not change with any further increases of the metal ion concentration (0.2-20 mM), whereas Mg2+ requirement for the maximal GTPySbinding by Gh was dramatically decreased to 0.8-3 mM in this detergent solution. By further increasing M F concentration (>4 mM), the ligand binding activity was subsequently inhibited -60% of the maximal ligand binding. However, GTPySbinding by Gh and Gh7 was not completely inhibited by further increases of MgC12 up to 50 mM. In the absence of MgC12, the GTPyS-binding by Gh and Gh7 could not be detected in either Lubrol PX or sucrose monolaurate solution. These data indicate that the different amounts of magnesium ion requirement for GTPyS binding by these G-proteins is due rather to the detergent effect than to the specific character of these Gproteins. This property, however, could be an indicator to distinguish between these G-proteins. When intrinsic GTPase activities of Gh and Gh7 were measured, the turnover of GTPase activity of Gh7 was slower than that of Gh. Thus, Gh and Gh7 hydrolyzed GTP; the turnover numbers were 2-3 and -0.25-0.6 min", respectively. It is not yet clear whether the different turnover of GTPase activities is specific for these G-proteins, since the isolated Gh7 protein was more unstable than Gh, although these proteins are structurally and functionally similar (see below). In addition, Gh7 was not a toxin substrate to be ADP-ribosylated either by cholera or pertussis toxin, even with the addition of 87-subunits of the heterotrimeric G-proteins (data not shown).
Immunological Cross-reactivity of Antibody-As mentioned above, overall biophysical and biochemical properties of Gh7 are similar to Gh. The distinct difference is in the molecular masses of the GTP-binding a-subunits (Fig. 3, A   and B ) , whereas the 50-kDa 8-subunits of Gh7 and Gh have the same molecular mass (Fig. 3A). To assess whether Gh7 is distinct from Gh, an antibody raised against the native protein was used to test the immunological cross-reactivity. As demonstrated in Fig. 3c, when Gh and Gh7 were subjected to immunoprecipitation, the G b protein was also effectively immunoprecipitated, indicating that Gh and Gh7 are homologs. The specificity of antibody was demonstrated in Fig. 4. The [a-32P]GTP-labeled Gh7. was incubated with Gh7,, antibody, nonimmune sera, or the antibody preincubated with unlabeled Gh7a. The results revealed that the Gh7a protein was specifically recognized by the antibody (Fig. 4A, lane 1 ). Thus, [a-"PIGTP-labeled Gh7. was not precipitated by nonimmune sera ( l o n e 2 ) or by the antibody pretreated with the unlabeled ( l a n e 3). To further evaluate the homology of the G b family and other G-proteins, the purified Gh, Gh7, and the membrane extracts from rat, dog, bovine, and human heart were tested by immunoblota. As demonstrated in Fig.   4B, the antibody recognized Gh7. . L, and a 74-kDa protein in rat, a 77-kDa in dog, a 78-kDa in bovine, and an 80-kDa in human heart membranes. The 78-kDa protein in bovine heart is most likely the and the 74-kDa protein in rat heart is Gho (see below). The 77-and 80-kDa proteins in dog

Ro. 4. Determination of
antibody specificity by immunoprecipitation and immunoblots. Panel A, immunoprecipitation of the purified Gh7-. Purified Gh7. (50 ng/30 pl), labeled with 10 pCi of [cx-~'P]GTP in the presence of 2 mM MgC12, was incubated at room temperature for 2 h with 5 pl of Gh7. antibody (lone 1 ) or nonimmune sera (lone 2), or with the antibody which was preincubated with unlabeled Gh7 (lane 3). Preincubation with unlabeled Gb7 and antibody was carried out at mom temperature for 2 h prior to incubation with labeled Gh7. Panel B, immunoblots of the purified Gh,, G h , and heart membranes from various species. Purified Gh7 (20 ng) and Gh (20 ng) and heart membranes from rat (200 pg, R H ) , dog (200 pg, DH), bovine (200 pg, BH), and human (300 pg, HH) were Subjected to SDS-PAGE (7% gel) and transferred to Immobilon-P. Immunoblotting was accomplished using 1 5 0 0 dilution of antibody as described under "Experimental Procedures." To evaluate the specificity of antibody, the protein-transferred Immobilon-P was pretreated with nonimmune sera (1500 dilution) at room temperature for 1 h followed by incubation with unlabeled protein A (5 pg/ml) for 1 h. Each step of washing was carried out as described under "Experimental Procedures." The pretreated blot was subjected to immunoblotting with the antibody and '2bI-labeled protein A. and human, respectively, might also be homologs of G b (G~T,,).
These results substantially demonstrate the specificity of Gh7. antibody for Gh7. and its family and its recognition of the native and denatured Gh,, family proteins. Moreover, the antibody did not recognize any other proteins, indicating that the Ghrr family is distinct from other G-proteins. When similar experiments were carried out with nonimmune sera or antibody preincubated with Gh or Gh7, the results were negative (data not shown).
T o assure whether these proteins recognized by the G7,, antibody are GTP-binding proteins, the membrane extracts from hearts of various species and various tissues of rat were examined for cross-reactivity with Gh7. antibody. Prior to immunoprecipitation, the extracts were photolabeled with [a-"PIGTP in the presence of 0.1 mM App(NH)p and 2 mM MgC12. As shown in Fig. 5A, after immunoprecipitation with the antibody usingprotein A-agarose, only the range of labeled 74-80-kDa proteins were precipitated with Gh7,, antibody. When extracts of the rat tissues were subjected to immunoprecipitation after labeling with [cY-~'P]GTP, a 74-kDa molecular mass protein similar to crhp was detected in all tissues tested (Fig. 5 8 ) . These results clearly indicate that the GTPbinding proteins are the same proteins recognized in the membranes from various species by immunoblots (see Fig.   4B) and demonstrate that these GTP-binding proteins are homologous to G h (Gh,,,) and are species-specific in molecular mass.

G-Proteins Which Couple to alddrenuceptor in the Ternary
Complez Preparations-Utilizing the specific properties of the ternary complex comprised of hormone-receptor-G-protein, the coupling ability of Gh with the al-adrenoceptors was examined in various tissues and species. The ternary complex is a result of a sequential process in which the hormone,

A B
I-43 1 1 2 3 4 E+, H Lu B St SpK Liv hc. 5. Immunoprecipitation of membrane est-from various species' heart and rat t i m e . The heart membranes (10 mg/ml) from various species were preincubated with 5 X lo-' M (-)epinephrine for 30 min at 30 "C to increase the labeling of >74-kDa proteins and then extracted with 0.5% sucrose monolaurate in HED buffer containing 250 mM NaCl at 4 'C for 1 h. After centrifugation at 40,000 X g at 4 "C, the extracts (400 pg of protein) were incubated with 20 pCi of [cz-~PICTP in the presence of 0.2 rnM App(NH)p and 2 mM MgC12 for 10 min at 30 'C and then photolabeled in an ice bath for 8 min. Panel A, immunoprecipitation of membrane extracts from various species. The labeled samples described above were incubated with 5 pl of antibody at room temperature for 2 h; then precipitation was performed with 10 pl of protein A-agarose (see "Experimental Procedures"). The -48-kDa bands in lams 2,3, and 4 are probably proteolytic fragments of the corresponding G-proteins, since these bands are not observed in Figs. 4, 6, and 7. The lanes indicated are: 1, rat; 2, dog; 3, bovine; 4, human. Panel B, immunoprecipitation of various tissues of rat. Membranes (10 mg/ml) of various tissues of rat were preincubated with 5 X loA6 M (-)-epinephrine for 40 rnin at 30 'C and then were solubilized. The extracts (200 pg) were incubated with 10 pCi of [cz-~*P]GTP and labeled in an ice bath, prior to performing immunoprecipitation experiments, as described above. As a control, the purified G h (20 ng) was also tested. The -35-kDa bands are probably photoleolytic fragments, since the intensity of the bands match with of GL. Autoradiography was for 2 days. The lanes are: Gh; H, heart; L u , lung; B. brain; St, stomach; Sp, spleen; K , kidney; Liv. liver. receptor, and G-protein become associated, forming a heterotrimeric intermediate. As a result of this process, the Gprotein in the ternary complex is primed for G T P binding that is fast and occurs at 0-4 'C (17, 24). Therefore, the vesicles containing the complexes were incubated with (-)epinephrine or (-)-epinephrine plus phentolamine a t 30 'C for 30 min and then chilled in an ice bath for 10 min. To observe the specific reaction of the ternary complexes, the samples were immediately subjected to UV irradiation after the addition of 10 pCi of [CY-~~PJGTP. The results of these studies are demonstrated in Fig. 6. Fig. 6A showed the hormone-mediated G T P binding with the vesicles containing the ternary complexes from rat liver and heart and bovine liver and heart. Thus, the apparent labeling of the 74-kDa proteins of either rat liver (lane I ) or heart (lane 3) obtained in the presence of (-)-epinephrine, was significantly inhibited in the presence of phentolamine (lanes 2 and 4 ) . On the other hand, with the ternary complex preparations from bovine liver (lanes 5 and 6 ) and heart membranes (lanes 7 and 8), the labeling of 78-kDa proteins in the presence of (-)-epinephrine (lanes 5 and 7) was substantially attenuated by phentolamine (lanes 6 and 8). These data show that these different molecular mass G-proteins couple to the al-adrenoceptor. The residual labeling in the presence of phentolamine is probably due to incomplete blocking of the receptor by the antagonist since the al-receptor in the ternary complex is in the state which has high affinity for agonist (17,18,24). It is also possible, because of the intrinsic guanine nucleotide exchange of Gh, that it is receptor-independent. The slight differences in molecular masses of 74-and 78-kDa proteins between liver Propranolol ( lo6 M ) and rauwolscine ( M) were also included in these reaction mixtures. The original ternary complex preparations isolated under the exact same condition were used for the following experiments without adjusting the protein concentration. Panel A, GTP-binding activity of the al-agonist-receptor-Gb ternary complexes prepared from rat liver and heart and bovine liver and heart. Vesicles containing the ternary complexes (100 p l ) were photolabeled with 10 pCi of [c~-~*P]GTP for 6 min in an ice bath under 254-nm UV irradiation. The reactions were carried out in the presence of 0.5 mM MgC12 and 0.1 mM App(NH)p in a 20 mM Hepes buffer (pH 7.4) containing 100 mM NaCl and 0.2 mM ascorbic acid. The reactions were terminated by the addition of Laemmli solution, followed by SDS-PAGE (7% gel) and autoradiography overnight. Each sample (100 pl) contained 150 fmol of al-receptors and 153 fmol of G-proteins from rat liver, 120 fmol of al-receptor and 145 fmol of G-proteins from rat heart, 137 fmol of al-receptor and 145 fmol of G-protein from bovine liver, and 116 fmol of a,-receptor and 124 fmol of Gprotein from bovine heart. RL, rat liver; RH, rat heart, B L , bovine liver; BH, bovine heart, 1, (-)-epinephrine; 2, (-)-epinephrine + phentolamine. Panel B , GTP-binding activity of the al-agonist-recep-tOr-Gh ternary complexes from various species. These studies were performed exactly the same way as detailed above. The samples (100 p l ) contained 132 fmol of al-receptors and 156 fmol of G-proteins from rat, 77 fmol al-receptors and 85 fmol C-protein from dog, 110 fmol of al-receptors and 148 fmol of G-proteins from bovine, and 45 fmol of al-receptor and 47 fmol of G-protein from human. The data shown are representative of five independent experiments. RH, rat heart; DH, dog heart; B H , bovine heart; HH, human heart, 1, (-)epinephrine; 2, (-)-epinephrine + phentolamine. and heart tissues are probably due to the purity of the complex preparations since it is evident that in the same species the same molecular mass of Gha exists, as shown in Figs. 5B and

7A.
To further assess the al-receptor-coupled Gh family in various species preparations, the al-agonist-receptor-G-protein ternary complexes from rat, dog, bovine, and human heart membranes were also examined. As presented in Fig.  6B, (-)-epinephrine-stimulated [c~-~*P]GTP labeling of the different molecular mass GTP-binding proteins was significantly decreased by the al-receptor antagonist, phentolamine. Thus, the molecular masses of GTP-binding proteins which coupled to the al-receptors were 74 kDa in rat heart, 77 kDa in dog heart, 78 kDa in bovine heart, and 80 kDa in human heart. To ensure that these GTP-binding proteins were the same proteins recognized by antibody (see Figs. 4 and 5), the ternary complex preparations were photoaffinity-labeled with [cc-~*P]GTP, incubated with antibody, and precipitated using protein A-agarose. As shown in Fig. 7A, the GTPbinding proteins which coupled to the al-receptors were the same proteins observed above by immunoblots and immuno- an ice bath. The samples (100 p l ) were immunoprecipitated with 5 pl of Gh7. antibody, followed by SDS-PAGE (7% gel) and autoradiography overnight. Difference in the intensity of the G-proteins as compared to Fig. 6, is probably due to the experimental conditions, since photolabeling in this experiment was performed in the detergent solution to facilitate immunoprecipitation. The data shown are representative of three independent experiments. RL. rat liver; RH, rat heart; DH, dog heart; R H , bovine heart; HH, human heart. Panel R.
co-immunoprecipitation of the al-adrenoceptors in the ternary complex preparations with Gh7a antibody. The density of the al-receptors in the ternary complex preparations was measured in the nupernatant using 3 nM [3H]prazosin before and after immunoprecipitation with antibody. For these studies Ga, . antibody which was purified using protein A-agarose was further purified by hydroxylapatite column. The loaded antibody in the column was eluted using a phosphate gradient (0-300 mM). The inorganic phosphate in the eluate wan removed through a dry Sephadex G-25 column (17-19). For the measurement of the amount of the receptors in the samples. (-1epinephrine was removed by a dry Sephadex C-25 column (3 ml) equilibrated with a 20 mM Hepea buffer (pH 7.4) containing 100 mM NaCI. 1 mM EDTA, and 0.05% nucrose monolaurate. Samples were also incubated with protein A-agarose (+) or nonimmune sera (0) to determine the specificity of the Gh,. antibody. Treatment of the ternary complex preparations with these probes did not significantly change the amount of the receptors in the supernatants. The data shown are representative of three independent experiments. RL, rat liver; BL, bovine liver; RH, rat heart; DH. dog heart; RH, bovine heart; HH, human heart. precipitation. Supporting this notion, when the amounta of the receptors in the ternary complex preparations were measured in the supernatants after immunoprecipitation, the receptors remaining were 13% in rat liver and 5% in bovine liver. The preparations from rat, dog, bovine, and human heart showed a range of 28-4596 of the al-receptors remaining (Fig. 7B). The samples treated with nonimmune sera or protein-A agarose did not significantly decrease the receptor density, remaining -82-93% and -94-100% in the supernatants, respectively (Fig. 7 B ) . These data clearly demonstrate the tight association of the al-receptor with the Gh family proteins and the specificity of the a,-receptor coupling to the Gh family. The reason is not clear why less precipitation of the al-receptor by Gh7P antibody occurs in t h e ternary complex preparations from the heart tissues. However, it is probably due to the instability of the ternary complex from heart tissues or because of the proteolytic fragments of the related proteins as seen in rat and dog hearts (Fig. 6 B ) . It is also possible that other al-receptor subtypes present in this tissue are not Coupling to Gh (5,25).

DISCUSSION
To further understand the al-receptor-Gh coupling mechanism, we have made three different determinations. 1) A new