THE JOURNAL OF BIOLOGICAL CHEMISTRK 0 1991 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 266, No. 30, Issue of October 25. pp. 20329-20336.1991 Printed in U. S.A.

Atypical &Adrenergic Receptor in 3T3-F442A Adipocytes PHARMACOLOGICAL AND MOLECULAR RELATIONSHIP WITH THE HUMAN &ADRENERGIC RECEPTOR*

(Received for publication, August 13, 1990)

Bruno FeveSB, Laurent J. Emorinell, Franpoise LasnierS, Nathalie Blinll, Beatrice Baudell, Clara Nahmiasll, A. Donny Strosbergll, and Jacques PairaultS From the $Unit6 282 Institut National de la Santh et de la Recherche Medicale-Centre National de la Recherche Scientifique, H6pital Henri Mondor, 94010 Creteil, France and the 9Uniti Propre de Recherche 0415 Centre National de la Recherche Scientifique, and Universiti Paris VU, Institut Cochin de Genitique Mokculaire, 22, rue Mechain, 75014 Paris, France

Expression of ligand binding properties for an atyp- ical &adrenergic receptor W-AR) subtype was studied during the adipose differentiation of murine 3T3- F442A cells and compared with that of the human &- AR expressed in Chinese hamster ovary cells stably transfected with the human 83-AR gene (CHO-B, cells) Emorine, L. J., Marullo, S., Briend-Sutren, M. M., Patey, G., Tate, K., Delavier-Klutchko, C., and Stros- berg, A. D. (1989) Science 245, 1118-1121). 3T3- F442A adipocytes exhibited high and low affinity binding sites for (-)-4-(3-t-butylamino-2-hydroxy- propo~y)[5,7-~H]benzimidazole-2-one ((-)-['H]CGP- 12177) (KO = 1.2 and 38.3 nM) and (-)-['2sI]iodo- cyanopindolol (['261]CYP) (KO = 47 and 1,610 PM). The high affinity sites corresponded to the classical B1- and Bz-AR subtypes whereas the KO values of the low affin- ity sites for the radioligands were similar to those measured in CHO-Bs cells (KO = 28 nM and 1,890 PM for (-)-[3H]CGP12177 and ['2sI]CYP, respectively). These low affinity sites were undetectable in preadi- pocytes but represented about 90% of total &ARs in adipocytes. The atypical &AR and the human D3-AR had similarly low affinities (Ki = 3-5 PM) for (2)-(2- (3-carbamoyl-4-hydroxyphenoxy)ethylamino-3)-(4- (l-methyl-4-trifluormethyl-2-imidazolyl)-phenoxy)- 2-propanol methane sulfonate (CGP20712A) or ery- thro-(f)-l-(7-methylindan-4-yloxy)-3-isopropylami- nobutan-2-01 (ICI118561), highly selective &- and f12-

AR antagonists, respectively, in agreement with the poor inhibitory effect of the compounds on (-)-isopro- terenol (1PR)-stimulated adenylate cyclase activity. Atypical @-AR and 83-AR had an affinity about 10-50 times higher for sodium-4-(2-[2-hydroxy-2-(3-chloro- phenyl)ethylamino]propyl)phenoxyacetate sesquihy- drate (BRL37344) than the B1-AR subtype. This cor- relates with the potent lipolytic effect of BRL37344 in adipocytes. The rank order of potency of agonists in functional and binding studies was BRL37344 > IPR < (-)-norepinephrine > (-)-epinephrine both in 3T3 adipocytes and CHO-B, cells. As in CHO-& cells, the classical B1- and &-antagonists CGP12177, oxpreno-

* This work was supported by the Centre National de la Recherche Scientifique and the Institut National de la Recherche Scientifique, the Ministry of Research, the University Paris VI1 and the Bristol Meyer's-Squibb Company. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1 Recipient of a Poste d'Accueil INSERM. To whom correspond- ence should be sent: U 282 INSERM, H6pital Henri Mondor, 51, av. Markcha1 De Lattre de Tassigny, 94010 Crkteil, France. Fax: 1.48- 98-04-69.

101, and pindolol were partial agonists in adipocytes. Although undetectable in preadipocytes, a major mRNA species of 2.3 kilobases (kb) and a minor one of 2.8 kb were observed in adipocytes by hybridization to a human D3-AR specific probe. The 2.3-kb transcript was also found in mouse white adipose tissue. Polym- erase chain reaction analysis of cDNA reverse-tran- scribed from 3T3-F442A RNA revealed the presence in adipocytes but not in preadipocytes of transcripts homologous to the human &-AR gene.

*The close parallelism between the pharmacological and mRNA analyses supports the existence in 3T3 adipocytes of a predominant &AR population in- volved in the lipolytic effect of catecholamines.

P-ARs' are integral plasma membrane proteins that mediate a wide variety of tissue-specific responses. Originally, the multiple effects of catecholamines were thought to be me- diated by two receptor subtypes, called ,&- and &ARs (Lands et al., 1967). The molecular cloning of two genes coding for PI- and &ARs (Dixon et al., 1986; Chung et al., 1987; Emorine et al., 1987; Frielle et al., 1987; Kobilka et ,al., 1987) has definitively established this distinction.

In adipose tissue, the PI-AR has generally been considered as the main effector of lipolysis (Lands et al., 1967; Buko- wiecki et al., 1978; Rothwell et al., 1985; Levin and Sullivan, 1986; Bahouth and Malbon, 1988). The availability of an increasing number of selective 0-AR ligands has however led to the hypothesis that an atypical /3-AR subtype, that is non- &, non-& could mediate lipolysis in rodent adipocytes (re- viewed by Zaagsma and Nahorski, 1990). So far, direct char- acterization of this additional 8-AR subtype has failed. The cloning from a human genomic library of a gene coding for a third P-AR (termed &AR) has brought new insights into the diversity of the P-AR family (Emorine et al., 1989). This p3- AR exhibits peculiar pharmacological properties, clearly dis- tinct from those of the &- and &-AR subtypes, which suggest

The abbreviations used are: @-AR(s), @-adrenergic receptor(s); bp, base pair; BRL37344, sodium-4-~2-[2-hydroxy-2-(3-chloro- phenyl)ethylamino]propyl)phenoxyacetate sesquihydrate (RR.SS distereoisomer); (-)-[3H]CGP12177, (-)-4-(3-t-butylamino-2-hy- droxypropoxy)[5,7-3H]benzimidazole-2-one; CGP20712A, (f)-(2-(3- carbamoyl-4-hydroxyphenoxy)-ethylamino)-3-(4-(l-methyl-4-tri- fluormethyl-2-imidazolyl)-phenoxy)-2-propanol methane sulfonate; E, (-)-epinephrine; Hepes, N-2-hydroxyethylpiperazine-W-2-eth- anesulfonic acid; ['251]CYP, (-)-['251]iodocyanopindolol; ICI118551, erythro-(+.)-l-(7-methylindan-4-yloxy)-3-isopropylaminobutan-2-o~ IPR, (-)-isoproterenol; kb, kilobase(s); PCR, polymerase chain re- action; SDS, sodium dodecyl sulfate; CHO, Chinese hamster ovary.

20329

20330 Atypical @-Adrenergic Receptor in 3T3 Adipocytes

its relationship with the postulated atypical 0-AR. Murine 3T3 preadipose cells which undergo adipose con-

version in vitro (Green, 1979) offer an excellent opportunity to elucidate this paradigm. We (Five et al., 1990) and others (Lai et al., 1982; Nakada et al., 1987; Guest et al., 1990) have already studied some aspects of &-AR and &AR expression during 3T3 adipose development.

In the present work we have further investigated, on phar- macological and molecular bases, @-AR expression in differ- entiating 3T3-F442A cells. The differentiation of the cells into adipocytes reveals a major and coordinated expression of an atypical @-AR whose pharmacological properties have been compared in parallel experiments with those of the human &-AR expressed in Chinese hamster ovary cells (CHO-p3 cells). The adipocyte atypical P-AR and the human &TAR have similarly low affinities for several 6-AR antagonists including the radioligands ["51]CYP and (-)-[3H]CGP12177. Among other properties shared by the two receptors are the partial agonistic activities of the &- and &antagonists CGP12177, oxprenolol, and pindolol and a high affinity for the selective lipolytic agonist BRL37344.

In 3T3-F442A adipocytes the emergence of the atypical p- AR protein occurs together with the appearance of a major mRNA species hybridizing to probes derived from the human P3-AR gene and selective for this P-AR subtype. This devel- opmentally regulated mRNA species, also found in mouse white adipose tissue, probably represents the relevant tran- script for the atypical P-AR characterized in this study.

Our results support the conclusion that the atypical /3-AR of 3T3 adipocytes is the murine homolog of the human 8 3 -

AR. The functional coupling of this &AR to catecholamine- stimulated lipolysis in 3T3-F442A cells suggests its partici- pation in the energy balance of rodent adipocytes.

MATERIALS AND METHODS

Cell Culture-3T3-F442A cells (Green and Kehinde, 1976) were grown and differentiated in tissue culture dishes (Falcon) at 37 "C in an atmosphere of air/COz (90:10, v/v) in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum (GIBCO). Upon cell confluence (day 0), culture media were continuously supplemented with insulin (1 pglml). Cells progressively transformed their fibro- blast-like phenotype into an adipocyte-like phenotype. At day 8, more than 90% of the cells had accumulated lipid droplets and were thus considered as adipocytes. Subconfluent CHO-83 cells, grown in Ham's nutrient mixture F-12 supplemented with 10% fetal calf serum (Tate et al., 1991), were used for experiments.

Lipolysis Experiments-Lipolysis was assessed as glycerol release from adherent cells in 24-well plates. Adipocyte monolayers were washed with Krebs-Ringer phosphate buffer (pH 7.4), supplemented with 2% bovine serum albumin, 4.5 g/liter glucose, and 50 pg/ml Na&05 as antioxidant. Cells were then incubated with or without 8- adrenergic effectors for 90 min at 37 "C in 1 ml of this medium in a humidified atmosphere of 5% COz in air. Aliquots of the incubation medium (100 pl) were removed to determine glycerol (Lin and Ma- gasanik, 1960). Reduction of NAD to NADH was recorded at 340 nm in the presence of glycerol dehydrogenase (EC 1.1.1.6) from Entero- bacter aerogenes (Boehringer Mannheim). Reaction buffer consisted of 130 mM (NH&S04, 10 p M MnClZ, 3.3 mM NAD, and 125 mM K2C03/NaHC03 (pH 10).

Cell Extracts and Adenylate Cyclase Assays-After washing of cell

homogenized at 4 "C in 1 mM EDTA, 25 mM Tris-HC1 (pH 7.5) (Fbve monolayers with phosphate-buffered saline, cells were harvested and

et al., 1989). Homogenates were centrifuged at 10,000 X g for 5 min (adenylate cyclase assay) or at 100,000 X g for 90 min (binding experiments). Membrane pellets were resuspended in the homogeni- zation buffer and stored at -80 "C. Protein content was assayed (Lowry et al., 1951) using bovine serum albumin as a standard.

Adenylate cyclase (EC 4.6.1.1) activity was measured in triplicate for 10 min at 35 "C in a 50-pl standard assay consisting of 0.2 mM [~u-~'P]ATP (Amersham Corp., PB 171), 1 mM CAMP, 10 mM phos- phocreatine, 0.5 unit of creatine phosphokinase, 100 p M GTP, 5 mM MgClZ, 0.2 mM EDTA, and 50 mM Tris-HC1 (pH 7.5), with or without

@-adrenergic effector. The reaction was initiated by the addition of crude membranes (15-25 pg of protein) and terminated as described (Pairault et al., 1982).

(-)-C3H]CGP12177 Binding to Intact Cel.k"-)-[3H]CGP12177 binding studies were performed on adherent cells (Fbve et al., 1990). Briefly, cells were incubated at room temperature for 15 min in Dulbecco's modified Eagle's medium adjusted to 10 mM NaHC03 and supplemented with 20 mM Hepes, 1 mM ascorbic acid (final pH 7.35), and in the presence of varying concentrations (0.1-200 nM) of (-)- [3H]CGP12177 (Amersham Corp., TRK 751). Cells were then quickly washed three times with ice-cold phosphate-buffered saline and col- lected in 0.5% SDS before counting. Nonspecific binding, determined in the presence of 100 p~ (-)-IPR, was usually 15-20% and 35-40% of the total binding at 1 nM and 40 nM (-)-[3H]CGP12177, respec- tively. Internalization of the radioligand was assessed by dissociation of bound radioactivity for 10 min at 4 "C with a mixture of 0.2 M acetic acid, 0.5 M NaCl (pH 2.5); it was negligible ( 4 % ) at 4 or 40 nM radioligand.

[lZ5Z]CYP Binding to Subcellular Fractions-Membrane aliquots (20-50 pg of protein) were incubated for 20 min at 37 "C with [1261] CYP (Amersham Corp., IM 142), with or without competing ligand, in a final volume of 200 p1 of 10 mM MgC12, 1 mM ascorbic acid, 100 p~ GTP, and 50 mM Tris-HC1 (pH 7.4). After dilution with ice-cold 10 mM MgClZ, 50 mM Tris-HC1 buffer (pH 7.41, samples were im- mediately filtered through glass fiber discs (Whatman GF/C) pre- soaked in 0.3% polyethyleneimine. Saturation experiments were per- formed with varying concentrations (5-3,000 pM) of radioligand. Competition experiments were carried out a t 30 or 300 pM [12sI]cYP. At these concentrations nonspecific binding, defined as that occurring in the presence of 100 p~ (-)-IPR, represented 10-15% and 25-30% of total binding, respectively. CGP12177 and CGP20712A were gift from Ciba-Geigy. IC1118551 was provided by IC1 Pharma (France Division, Cergy-Pontoise) and BRL37344 by Smith Kline Beecham Pharmaceuticals (Epsom, UK). (-)-IPR, (-)-epinephrine, (-)-nor- epinephrine, (&)-oxprenolol and (+)-pindolo1 were purchased from Sigma.

Data Analysis-Saturation and competition binding experiments were analyzed with the EBDA and LIGAND programs (Biosoft- Elsevier, Cambridge, England) (Munson and Rodbard, 1980; Mc- Pherson, 1985). Unless otherwise specified, pooled data are presented in figures and tables as mean & S.E. of at least three independent experiments performed in duplicate.

RNA Analysis-Total RNA was extracted from cells by the method of Cathala et al. (1983), and from mouse adipose tissue as described (Chomczynski and Sacchi, 1987). Poly(A)+ RNA was isolated by oligo(dT)-cellulose chromatography, electrophoresed through a 1.5% agarose, 2.2 M formaldehyde gel (Ausubel et al., 1987), and transferred to nylon Hybond membranes N+ (Amersham Corp.). Probes were labeled (0.5-1 X 10' cpm/pg of DNA) by nick translation with ["PI dCTP (Amersham, PB 10205). The &AR probe (il&) was a 206-bp AccI-ApaLI fragment encompassing the third intracytoplasmic loop of the human B3-AR gene. The pl- and the &TAR probes have already been described (Fkve et al., 1990). After prehybridization and hybrid- ization (Fbve et al., 1989), membranes were washed in 2 X SSC, 0.1% SDS for 10 min at room temperature and twice for 30 min at 45 "C. Final washing was in 0.2 X SSC, 0.1% SDS at 45 "C for 30 min (1 X SSC = 150 mM NaCl, 15 mM sodium citrate).

For polymerase chain reaction analyses, oligonucleotides were syn- thetized according to the sequence of the murine &TAR (Nahmias et al., 1991). The sense (5'-GCTCCGTGGCCTCACAG-3') and the antisense (5'-CTCGGCATCTGCCCCTA-3') primers corresponded to amino acid residues 2-7 and 177-182 (180-185 of the human 8 3 -

AR) of the receptor, respectively. For standardization purposes, two oligonucleotides were derived from the sequence of the murine hom- olog of the HepG2/erythrocyte glucose transporter whose expression does not vary during the adipose conversion of 3T3 cells (Kaestner et al., 1989). The sense (5'-GTGGAGCAACTGTGCGGC-3') and the antisense (5'-TCACACTTGGGAGTCCGC-3') primers correspond to codons 425-430 and 488-493 (stop codon) of the molecule, respec- tively. First strand DNA synthesis was performed on 5 pg of total RNA from 3T3-F442A cells using an oligo(dT) primer and avian myeloblastosis virus reverse transcriptase. After phenol and chloro- form extractions and ethanol precipitation, the sample was subjected to 30 cycles of amplification (93 "C for 1.5 min, 55 "C for 2 min, and 72 "C for 2 min) using 2.5 units of Thermophylus aquaticus polymerase (Cetus), in 100 p1 of buffer comprised of 67 mM Tris-HC1 (pH 8.41, 6.7 mM MgCl2, 100 pg/ml gelatin, 6.7 p~ EDTA, 10 mM Pz-mercap- toethanol, 16 mM (NH4)zS04, 10% (v/v) dimethyl sulfoxide, 5%

Atypical @-Adrenergic Receptor in 3T3 Adipocytes 20331

formamide (v/v), 125 p~ each of the deoxynucleotide triphosphates, and 125 nM each of the primer. When amplification was performed on genomic DNA, 1 pg of template was used. One-tenth of the reaction volume was then electrophoresed through a 2% agarose gel, trans- ferred to nylon membranes, and hybridized to a 314-bp probe obtained by amplification of the human 03-AR gene between a sense (5'- GCTCCGTGGCCTCACGAGAA-3') and an antisense (5"CCCA- ACGGCCAGTGGCCAGTCAGCG-3') primer corresponding to amino acid residues 2-8 and 98-106, respectively, of the human p3-

AR. This probe thus encompasses the N terminus of the receptor, a region of low homology with the &- and &-AR subtypes. Blot was treated as above except for a final washing of 60 min in 0.1 X SSC, 0.05% SDS at 50 "C.

RESULTS

Relationship between Atypical &Adrenergic Sensitivity in 3T3-F442A Adipocytes and &AR Coupling in cH0-p~ Cells- During 3T3 adipose differentiation, cells acquire a new cate- cholamine sensitivity (Rubin et al., 1977; Pairault et al., 1982). To address the issue of whether an atypical /3-AR could be involved in the remodeling of the transmembrane signaling system of these cells, functional experiments have been car- ried out to compare the potency of BRL37344 and CGP12177 with that of the classical P-AR agonist IPR. BRL37344 be- longs to a novel group of 8-AR agonists which selectively stimulate lipolysis and thermogenesis in rat adipocytes (Arch et al., 1984; Wilson et al., 1984; Muzzin et al., 1988; Hollenga et al., 1990). CGP12177 is a strong &- and &-AR antagonist (Staehelin et al., 1983) that unexpectedly displays potent agonistic effects in adipose tissue (Mohell and Dicker, 1989; Langin et al., 1991).

We studied the development of adenylate cyclase activity induced by a saturating concentration (100 @M) of IPR, BRL37344, and CGP12177 in differentiating 3T3-F442A cells (Fig. 1). Along with conversion from preadipocytes to adipo- cytes, the response to IPR increased steadily from 1.8 k 0.2- to 8.0 f 2.2-fold the basal value. By contrast, responsiveness t o BRL37344 and CGP12177, which was negligible in pread- ipocytes (day 0 ) only increased to 33 and lo%, respectively, of that to IPR at day 4 and attained its maximal level (78 and 34%, respectively) at day 8 after confluence. The delayed response to BRL37344 and CGP12177 suggests that they act through a target distinct from that of IPR.

The relative potencies of these /3-AR effects in eliciting functional 8-AR coupling in 3T3-F442A adipocytes were then studied. With respect to Kact values, BRL37344 was found to be the best activator of both the adenylate cyclase and lipol- ysis pathways. For activation of adenylate cyclase, this com-

0 2 4 6 8 1 0 1 2 DAYSOFCULTURE AFTERCONFLUENCE

FIG. 1. Development of IPR-, BRL37344- and CGP12177- stimulated adenylate cyclase activity during the adipose con- version of 3T3-F442A cells. At the indicated times, crude mem- branes were prepared and assayed for basal (0) and IPR- (O), BRL37344- (O), and CGP12177- (W) stimulated (100 b ~ ) adenylate cyclase activity. Results represent the mean of triplicates for a typical experiment.

pound was followed by CGP12177, then by IPR (Fig. 2A) whereas in lipolysis experiments IPR had a higher efficiency than CGP12177 (Fig. 2 B ) . The reversal in the order of these two compounds suggests that limited levels of CAMP, better attained with IPR than with CGP12177, may be sufficient for a full lipolytic response. This is illustrated by the fact that BRL37344 and CGP12177, which are partial agonists for CAMP production, almost became full agonists for lipolysis (93 and 70% of the IPR response, respectively).

The potency of a panel of P-AR effectors in stimulating adenylate cyclase activity was studied in parallel in 3T3- F442A and CHO-P3 cells (Table I). As observed in adipocytes, CGP12177 exhibited a partial agonistic activity in CHO-O,

Log lElfecbrl (M)

FIG. 2. Dose-response curve of agonist-stimulated adenyl- ate cyclase activity and lipolysis in adipocytes. Adenylate cy- clase activity ( A ) and lipolysis ( B ) of adipocytes were determined in the presence of various concentrations of IPR (O), BRL37344 (0) and CGP12177 (W). For adenylate cyclase activation, Kact values for IPR, BRL37344, and CGP12177 were 1,860 f 190,180 f 20, and 480

90 min/well, and K,,, values for IPR, BRL37344, and CGP12177 were k 30 nM, respectively. For glycerol release basal value was 51.7 nmol/

8 f 1, 1.5 f 0.5, and 20 f 4 nM, respectively. The figure represents mean of triplicates for a typical experiment.

TABLE I Comparative analysis of the effector-stimulated adenylate cyclase

activities in 3T3-F442A adipocytes and cHo-B3 cells Adenylate cyclase activity was determined in membranes from

3T3-F442A adipocytes and CHO-& cells, in the presence of various concentrations of the indicated ligands. VmaX values represent the percentage of the maximal effect of IPR. K,,, values correspond to the concentration of ligand giving 50% of the maximal stimulation. Basal adenylate cyclase activity was 10.3 k 1.3 and 3.7 k 1.2 pmol cAMP/min/mg of protein in adipocytes and CHO-P, cells, respec- tively. Maximal IPR-stimulated adenylate cyclase activity was 107.6 k 24.5 and 10.3 f 1.3 pmol cAMP/min/mg of protein in adipocytes and CHO-8, cells, respectively.

% of IPR nM % o f I P R rZM

IPR 100 1,860 f 190 100 495 f 220 Norepinephrine 95 k 4 4,000 f 600 88 f 5 1,250 f 350 Epinephrine 86 k 7 6,800 f 550 93 f 6 1,350 f 70 BRL37344 CGP12177

7 8 2 3 180k 20 57 f 13 7 8 k 31

Pindolol 3 4 k 1 480 f 30 60 k 5 330 f 110 10 k 2 10,000 f 3,500 45 f 10 1,100 f 200

Oxprenolol 11 k 2 7,500 f 2,000 40 k 8 1,200 & 100

20332 Atypical @-Adrenergic Receptor in 3T3 Adipocytes

B

0 20 4 0 60 ao Bound (fmol/ IO6 cells)

FIG. 3. Characterization of (-)-[3H]CGP12177 binding sites on 3T3-F442A cells. Panel A , saturation curves of a typical binding experiment performed either on preadipocytes (0) or on adipocytes (0) (inset, details of binding data a t low concentrations of radioligand). Panel B , Scatchard plot of simulated one-ligand, two- binding site data, corrected for nonspecific binding. The dashed curue is the curvilinear best fit for a two-site model in adipocytes (0). The two binding components are indicated by the solid straight lines. In preadipocytes (0) the data are best fitted by a one-site model.

cells (roughly 60% of the IPR response). Similarly, the &- and &AR antagonists oxprenolol and pindolol behaved as agonists in adipocytes as well as in CHO-& cells. In both cell lines, this agonistic activity corresponded roughly to a dou- bling of the basal enzyme activity. Finally, the preferential activation of adenylate cyclase activity by BRL37344 was also observed in cH0-83 cells with a rank order of potency for agonists (BRL37344 > IPR > norepinephrine > epinephrine) being identical in adipocytes and CH0-B3 cells (Table I).

Taken together, these findings show the similar behavior of &- and adipocyte @-AR toward adenylate cyclase activation.

Relationship between the Atypical 8-AR Binding Sites in 3T3-F442A Adipocytes and &AR in cH0-p~ Cells-To char- acterize the atypical P-AR of 3T3-F442A cells and its rela- tionship with the human P3-AR, saturation experiments, using the hydrophilic radioligand (-)-[3H]CGP12177 over a wide range of concentrations (0.1-200 nM), were carried out on intact preadipocytes and adipocytes and also on CHO-p3 cells. In preadipocytes, (-)-[3H]CGP12177 binding was saturable with a plateau at about 10 nM of the radioligand (Fig. 3A). Scatchard representation (Fig. 3B) indicated the presence of a single class of sites (2,750 f 1,030 sites/cell) with a high affinity (& = 0.7 & 0.1 nM) for the radioligand. In contrast, the binding of (-)-[3H]CGP12177 to adipocytes showed an initial plateau of saturation between 7 and 10 nM of the radioligand (Fig. 3A 1, but revealed a second stereospecific and saturable binding component at higher concentrations. The curvilinear Scatchard plot was fitted with a two-binding site model, corresponding to a high (KO = 1.2 f 0.5 nM) and a low (KO = 38.3 k 8.1 nM) affinity component with respective B,,, values of 7,800 f 1,1002 and 50,000 f 6,400 sites/cells. In

* In a previous study we identified this component as fll-AR subtype (Five et al., 1990). Since the concentration of radioligand varied from 0.1 to only 20 nM, the low affinity component could not be detected, leading to a slight overestimation of high affinity sites (-15,000 sites/ cell).

150

0 20 40 [-)-[JHICGPl2177 [nMl

60

FIG. 4. Characterization of (-)-13H1CGP12177 binding . . - ~~

sites in CHO-83 cells. Experiments were carried out on adherent CHO-fl3 cells, and a typical saturation curve is shown. Inset, Scat- chard plot of a simulated one-site model.

~~

CHO-P3 cells, Scatchard representation of (-)-13H]CGP12177 binding (Fig. 4) indicated the presence of a single class of sites with an affinity (KD = 28 f 8.5 nM) comparable to that of the low affinity sites of 3T3 adip~cytes.~

To ensure that in 3T3-F442A adipocytes the two affinity components for (-)-[3H]CGP12177 did not represent a single class of sites in a low and a high affinity state related to the partial agonism of this ligand, saturation experiments were performed with another radioligand. Thus, ['251]CYP binding experiments were done on crude membranes in the presence of 100 p~ GTP to shift all receptor subtypes into the low affinity state for agonists (Kent et al., 1980). Here again, high and low affinity sites were detected (Table 11), both exhibiting saturation and stereoselectivity (not shown). The low affinity component (-90% of total B-ARs) had a K, value for [lZ5I] CYP similar to that of the human P3-AR (Table 11). The proportion and estimated number of the two classes of binding sites were in close agreement in (-)-[3H]CGP12177 or [1251] CYP saturation experiments. For both ligands, the low and the high affinity components exhibited KD values differing by more than 1 order of magnitude.

The involvement of the two classes of binding sites in mediating biological effects was assessed by performing com- petition experiments against ['251]CYP binding with the li- gands used in functional experiments. To address specifically the ligand binding profiles of the low or of the high affinity sites, experiments were performed at both low (30 pM) and high (300 PM) radioligand concentrations. In preadipocytes, whatever the [1z51]CYP concentration used, we always ob- served the following rank order of IC5, (50% inhibitory con- centration) values: CGP12177 < IPR < BRL37344 (Fig. 5, A and B). By contrast, in adipocytes an inversion of the ligand binding profile occurred, depending on the radioligand con- centration. Thus, at 30 PM ['251]CYP, the rank order of potency in adipocytes was the same as that observed in preadipocytes, whereas at 300 pM [1251]CYP there was a reversal in the ICs0 values with BRL37344 < IPR (Fig. 5, C and D).

These findings were analyzed by assuming that the two classes of binding sites were independent, and no heterotropic interactions did exist within a given class of binding sites. Based on the binding constants and numbers of high and low affinity sites measured, we estimated that the high affinity sites accounted for 60% of specific radioligand binding at 30

In our initial characterization of the human &AR (Emorine et al., 1989), we had reported that CGP12177 did not bind to this receptor. This was based on the absence of significant binding at concentration of (-)-[3H]CGP12177 up to 3 nM. As illustrated in the present study much higher concentrations of radioligand are neces- sary to label the &AR.

Atypical &Adrenergic Receptor in 3T3 Adipocytes 20333 TABLE I1

Characteristics of ['251]CYP binding sites in membranes from 3T3-F442A and CHO-& cells Saturation experiments were performed on crude membranes from 3T3-F442A preadipocytes and adipocytes or

from CHO-& cells within a wide range of concentration of ['251]CYP (5-3,000 PM). Data were analyzed according to the method of Scatchard (1949) using the EBDA/LIGAND computer program. Membrane protein content was 0.19, 0.62, and 0.12 mg of protein/106 cells in preadipocytes, adipocytes, and CHO-03 cells, respectively, and was used to normalize the number of sites/cell.

High affinity sites Cell type

Low affinity sites

KIJ Sites/cell KO Sites/cell

P M P M

3T3-F442A preadipocyte 62 t- 19 2,640 f 670 None

3T3-F442A adipocyte 47 f 29 6,670 f 4,450 1,510 f 10 90,700 f 16,000

CHO-83 None 1,890 k 20 121,000 f 10,000

-10 -9 -8 -7 -6 -5 - 4

6o 1 40 4 \ ,,\ \

-10 -9 -6 -7 -6 -5 - 4 -10 -9 -8 -7 -6 -5 -4

log ICornpetitorl (M) Log lcornpetitorl (M)

['aaI]CYP binding to 3T3-F442A preadipocyte and adipocyte FIG. 5. Competition curves of &adrenergic agonists for

membranes. Membranes from preadipocytes (A, B ) and adipocytes (C, D) were incubated with 30 PM (A, C ) or 300 pM (B, D) ['251]CYP and varying concentrations of IPR (O), BRL37344 (O), or CGP12177 (m). The figure represents displacement curves from a typical exper- iment. The corresponding inhibition constants obtained by LIGAND computer analysis according to a model for two 8-AR subpopulations are given in Table 111.

PM ['261]CYP. By contrast, at 300 pM ['251]CYP, because of their large predominance, the low affinity sites represented 75% of the labeled receptors. Thus, our findings suggested that binding sites of low affinity for ['251]CYP had a higher affinity for BRL37344 than for IPR. Reciprocally, binding sites of high affinity for the radioligand exhibited a better affinity for IPR than for BRL37344. This was verified by computer analysis of the data, in which we assumed that the high affinity sites for ['251]CYP represented only classical 8'- ARs. Indeed, the a-AR population could be considered as negligible at 300 PM ['251]CYP, it being by far the minor @- AR species in adipocytes (Fbve et al., 1990). Under these conditions we identified two classes of binding sites with different affinities for each competitor (Table 111). The sites of high affinity for ['251]CYP were better inhibited by CGP12177, followed by IPR and BRL37344 with inhibition constants (Ki) corresponding to binding of the compounds to the PI-AR. In contrast, the atypical sites, of low affinity for ['261]CYP, had a similar affinity for BRL37344 and CGP12177, which was higher than that for IPR.

The same experimental approach was used to analyze epi- nephrine and norepinephrine binding to the two classes of 8- AR of 3T3-F442A cells. The potency of epinephrine, norepi-

nephrine, and also IPR, BRL37344, and CGP12177 to inhibit ['261]CYP binding to membranes from CHO-& cells was de- termined in parallel. The Ki values for the human &AR were in the same range as those for the atypical B-AR of 3T3- F442A cells but clearly different from those for the 81 subtype (Table 111). It is noteworthy that the human &-AR and the atypical P-AR of 3T3 adipocytes both had a 10-50-fold higher affinity for the lipolytic compound BRL37344 than for IPR.

To assess further the respective participation in 3T3-F442A adipocytes of classical /3-ARs and of the atypical 8-AR in functional coupling, we carried out in parallel ['251]CYP bind- ing and adenylate cyclase experiments in the presence of CGP20712A and ICI118551, 8'- and &AR selective antago- nists, respectively (Dooley and Bittiger, 1987; O'Donnell and Wanstall, 1980). If the biological effects were being mediated by either of these receptors, one would expect to find a close correspondence between the Ki of these compounds for inhi- bition of ['251]CYP and their IC, for inhibition of adenylate cyclase. As shown in Table IV, however, the IC, values are some 2-3 logs higher than the Ki values for binding to B1 but are in close agreement with the Ki for interaction with the atypical P-AR. Again, the human P3-AR and the atypical 8- AR of adipocytes shared similar binding and functional pa- rameters of CGP20712A and ICI118551.

Taken together, these findings suggest that the atypical p- AR, which resembles the human &AR, is a major effector of catecholamine responsiveness in 3T3-F442A adipocytes.

Analysis of &AR mRNA in 3T3-F442A Cells-The phar- macological similarities between the atypical 3T3-F442A /3- AR and the human &AR prompted us to analyze 83-AR mRNA expression during 3T3-F442A adipose differentiation. Northern blot analysis of poly(A)+ RNA was performed using a &AR subtype-specific DNA probe derived from the third intracytoplasmic loop (i3p3) of the human &AR, a region that is less than 50% homologous to the corresponding parts of the pl- and &-AR genes (Emorine et al., 1989). With this probe no hybridization signal could be detected in preadipo- cytes whereas we observed the emergence in adipocytes of a strong &-AR signal corresponding to a transcript of 2.3 kb in size (Fig. 6). The appearance of a 2.8-kb minor species that could result from the use of alternative promoters or polyad- enylation sites was also noted in mature adipocytes. The major 2.3-kb transcript was also present in poly(A)+ RNA from mouse white adipose tissue: These signals cannot be related

* In a previous report we found different sizes for murine O3-AR mRNA transcripts (Emorine et al., 1989). This was because of the use of a cRNA probe that retrospectively displayed low homology (less than 65%) with the mouse 83-AR gene (Nahmias et al., 1991). In the present study a specific DNA probe derived from the third intracytoplasmic loop of the human P3-AR gene has been used; this probe exhibits 85% homology with the corresponding part of the mouse &-AR gene.

20334 Atypical @-Adrenergic Receptor in 3T3 Adipocytes TABLE I11

Binding parameters of the P-AR binding sites of 3T3-F442A adipocytes and CHO-P:t cells for various ligands Competition experiments were performed on membranes from adipocytes or CHO-Ps cells a t 300 pM ['2sI]cYP

in the presence of various ligands. In adipocytes Ki values were calculated by computer analysis according to a model of two 8-AR classes of high (PI) and low (atypical p) affinity for ["'IJCYP. In CHO-P, cells Ki values were analyzed according to a one-site model.

Ligand 3T3-F442A adipocyte K, CHO-P3 cell K,

P I Atypical P a f l M

IPR 75 f 27 6,500 k 1,500 4,210 f 130 Norepinephrine 593 f 323 121,000 f 50,000 38,300 f 4,900 Epinephrine 2,020 f 1,480 237,500 f 81,000 90,000 f 7,500 BRL37344 7,300 f 2,800 131 f 71 649 f 197 CGP12177 0.65 f 0.12 152 f 48 207 t 23

TABLE IV Relationship between binding parameters and biological effects of CGP20712A and IC1118553

in 3T3-F442A adipocytes and CHO-P, cells Competition experiments against 300 p~ ['2sIJCYP were performed in 3T3-F442A adipocytes or CHO-8s cells

in the presence of a PI- (CGP20712A) or &-selective (ICI118551) antagonist. Ki were obtained from the EBDA/ LIGAND computer analysis of displacement curves resolved in a two-site (adipocyte) or a one-site (CHO-PZ cell) model. Adenylate cyclase was stimulated by a submaximal dose of IPR (1 p ~ ) in the absence or presence of varying concentrations of CGP20712A or ICI118551. The ICso values was defined as the concentration of antagonist required to induce a half-maximal inhibition of IPR-stimulated adenylate cyclase activity.

Adipocyte CHO-P3 cell

Selective antagonist Ki

PI

ICs" K, ICs0 Atypical

f l M nM

CGP20712A 7.6 f 1.3 4,600 f 700 15,050 f 8,370 5,770 k 300 17,000 f 1,200

IC1118551 78.5 f 21 5,400 f 1,500 8,400 f 2,800 2,350 f 140 4,400 f 500

2.4- 7 I 1 . 4 9 , , , , I I

01 02 ' 0 3 '

c

FIG. 6. Northern analysis of &AR mRNA in 3T3-F442A cells. Blots of poly(A)+ RNA (20 pg) from adipocytes (lanes 1,2, and 4 ) , preadipocytes (lane 3 ) , and mouse inguinal adipose tissue (lane 5) were hybridized to the Dl-, &, and &AR subtype-specific probes. Membranes were exposed for 65 h to Kodak XAR-5 films a t -80 "C, with intensifying screens. Sizes of RNA standards (Bethesda Re- search Laboratories) are indicated (in kb) on the left side. Arrows show the position of 28 and 18 S ribosomal RNAs.

to the &-AR mRNA, which is 3.2 kb in size (Fig. 6), nor to the P2-AR mRNA, which was not detected with the P2-AR probe. Indeed, the P2-AR message has been only occasionally observed with this probe in mature adipocytes unless they are treated with glucocorticoids (F6ve et al., 1990).

Polymerase chain reaction experiments were performed (Fig. 7) using oligonucleotides derived from the sequence of the murine PZ-AR (Nahmias et al., 1991). A 543-bp fragment was amplified from the cDNA generated from 3T3-F442A adipocyte RNA. Since this fragment was not observed with preadipocytes, increased expression of the corresponding

mRNA was linked to the differentiation process. As an inter- nal control, the levels of the 206-bp fragment corresponding to HepGZ/erythrocyte glucose transporter mRNA were com- pared in preadipocytes and adipocytes. The &-AR primers were derived from regions of low homology with PI- and P2-

AR genes (Fig. 7), thus precluding the amplification of these genes. The fragment generated from 3T3-F442A mRNA had the size expected from the structure of the Pa-AR gene. More- over, it was a single fragment amplified from mouse DNA, and it specifically hybridized (Fig. 7) to the &-selective probe derived from the N-terminal region of the human Ps-AR.

DISCUSSION

The experiments reported here utilized the differentiating 3T3-F442A adipose cells to document several aspects of the differentiation-linked emergence of an atypical P-AR, differ- ent from the classical P1- and &ARs as judged by ligand- binding profile, functional efficiency, and expression at the mRNA level. The study establishes that this atypical P-AR is the prominent subtype expressed in 3T3-F442A adipocytes and suggests that it is the murine equivalent of the human Ps-AR.

An Atypical P-AR Is the Prominent P-AR Subtype Expressed in 3T3 Adipocytes. Relationship with the P-AR of Rodent Adipocytes-The increase of P-AR number is one of the major mechanisms for the regulation of the adrenergic transmem- brane signaling system during the adipose differentiation of 3T3 cells (Lai et al., 1982; Pairault et al., 1982; Nakada et al., 1987; Fkve et al., 1990; Guest et al., 1990). In 3T3-F442A cells, the induction of P1-AR is driven by the differentiation process per se whereas the up-regulation of P2-AR is essentially under

Atypical P-Adrenergic Receptor in 3T3 Adipocytes 20335

b3-AR-

GT1-

Sense primer Anti-seas. p r d r

MoE3 GCTCCGTGGCCTCACAG CTCCGCATCTCCCCCTA H"E3 "_"_""""~ Mob2 -GG--ACACGGGA-CGA

GA """G"c-""" AC-TTTC-TCTCCCTCC

RaEl -GCG--G--G-G-T-GC -C-TT-C--GET-T-CC

FIG. 7. Gene amplification analysis of &-AR mRNA from 3T3-F442A cells. Left panel, ethidium bromide staining of polym- erase chain reaction products obtained from mouse DNA (Mo DNA and from cDNA from 3T3 preadipocytes (Pread) and adipocytes (Ad). p&, p&, and p&AR genes, respectively. These plasmids were digested with restriction enzymes to yield control hybridization frag- ments. Positions of the fragments amplified with 83 and HepG2/ erythrocyte glucose transporter (GTI) primers are indicated on the left side. The sizes of the molecular weight markers (123-bp ladder from Bethesda Research Laboratories) are indicated in bp in the middle of the figure. Right panel, hybridization of a blot obtained from the left panel to a human &-AR N-terminal probe. The bottom of the figure shows the sequences of the primers derived from the mouse D3-AR gene sequence (Mo&) compared with the corresponding regions of the genes for the human b3-AR (HuP3; Genbank, accession code HUMARBBA), the mouse &AR (Moo2; Nakada et al., 1989), and the rat P,-AR (RapI; Machida et al., 1990). Nucleotide identity is indicated with hyphens.

glucocorticoid control (Fhe et al., 1990). Beside these classical @-ARs, which exhibit high affinity for (-)-[3H]CGP12177 and ['251]CYP, the present work demonstrates the existence of a major population of atypical P-AR representing about 90% of total P-ARs and exhibiting a much lower affinity for these radioligands. The KD of these low affinity sites for (-)-[3H] CGP12177 measured in saturation binding experiments is roughly comparable to the K j obtained for CGP12177 used as an inhibitor of ['251]CYP binding. Moreover, receptor densi- ties measured by both radioligands are in good agreement. These findings indicate that both ['251]CYP and (-)-[3H] CGP12177 label the same atypical receptor population. Be- cause such low affinity binding sites are undetectable in preadipocytes, they represent a 8-AR population whose expression in 3T3-F442A cells is linked to the differentiation process. This conclusion seems to be contradictory to that reached by other investigators (Guest et al., 1990), who only found an increased &AR expression during 3T3-Ll adipose conversion. However, differentiation of this cell line necessi- tates exposure to dexamethasone, a treatment that has been shown to switch P-AR subtype expression toward the &-AR (Lai et al., 1982; F6ve et al., 1990). Moreover, detection of the atypical P-AR in 3T3-Ll cells would require an appropriate pharmacological approach.

An important issue is whether the atypical B-AR of 3T3- F442A cells reflects ectopic P-AR expression in this cell line or truly represents the prototype of the adipocyte P-AR. It is noteworthy that in our binding and functional experiments, BRL37344 has a marked selectivity for the atypical B-AR as compared with IPR. These observations firmly corroborate the preferential lipolytic and/or thermogenic effect of BRL37344 in white and brown adipose tissue (Arch et al., 1984; Wilson et al., 1984, Muzzin et al., 1988; Hollenga et al.,

1990). Moreover, adenylate cyclase activity and glycerol re- lease in 3T3-F442A adipocytes are stimulated by the potent Dl- and &AR antagonist CGP12177. The Kact of CGP12177 for eliciting lipolysis roughly agrees with its low affinity for the atypical B-AR measured in intact adipocytes but can be contrasted with its antagonistic effect and high affinity to- ward classical o-ARs. Again, this suggests that the pro- nounced thermogenic and lipolytic effects of CGP12177 in rodent adipocytes (Mohell and Dicker, 1989; Langin et aL, 1991) are caused by the presence in these cells of an atypical /3-AR, equivalent to that of 3T3 cells.

Another feature of rodent adipocyte P-AR which has char- acterized its atypical nature is the low efficiency of potent Dl- and &selective antagonists, such as CGP20712A and ICI118551, to block catecholamine responsiveness (Bojanic et al., 1985; Zaagsma et al., 1985; Bahouth and Malbon, 1988; Arch, 1989; Hollenga and Zaagsma, 1989). Here we show that the atypical P-AR of 3T3 adipocytes displays a low and similar affinity for CGP20712A and ICI118551, consistent with the poor unselective inhibition of agonist-stimulated adenylate cyclase by these two compounds.

Taken together, our findings provided evidence for the prominent involvement in catecholamine responsiveness of 3T3 adipocytes of an atypical 0-AR, the properties of which mimic the picture encountered in rodent adipocytes. The numerous inconsistencies observed in previous works between binding and functional studies can now be revised on the basis of the pharmacological properties of the 3T3 atypical

Pharmacological and Molecular Relationship between the Atypical B-AR of 3T3 Adipocytes and the Human p3-AR-The human &AR, whose gene has recently been cloned (Emorine et al., 1989), shares numerous pharmacological features with the 3T3-F442A atypical 8-AR. Using similar experimental approaches for the two receptors, our study illustrates their similar and low affinity for classical 0-AR antagonists, such as CGP12177, CYP, CGP20712A, and ICI118551, but their high affinity for the atypical P-AR selective agonist BRL37344. In addition we show the close correlation existing between the rank order of potency of several agonists in activating adenylate cyclase of CHO-83 cells and 3T3-F442A adipocytes. KO and Ki values for several ligands are similar and range in the same rank order for both receptors. Hence the adenylate cyclase systems of 3T3 adipocytes and CHO-A cells are both activated by the potent unselective 8'- and 02-

AR antagonists CGP12177, oxprenolol, and pindolol. Additional evidence for the close relationship between the

atypical 0-AR of adipocytes and the human &AR is provided by molecular analysis of RNA. In Northern blot experiments, a major mRNA species hybridizing to the humad i3P3 probe is expressed in differentiated adipocytes whereas it is absent from preadipocytes. This mRNA probable corresponds to that for the atypical P-AR of rodent adipose tissue since it is also found in the mouse inguinal fat pad. This transcript is neither that for the &-AR, which has a higher size, nor that for the &AR, which is not present in adipocytes unless they are exposed to dexamethasone (F6ve et al., 1990). Moreover, the i3O3 probe presents less than 50% homology with the corre- sponding domains of 0'- and B2-AR genes, thus precluding the possibility that it cross-hybridizes with these receptor sub- types. Nor can this mRNA species represent that for another yet unidentified P-AR since the i3P3 probe only hybridizes to the P3-AR gene in the human genome and only detects a single sequence in the mouse (and rat) genome (Emorine et al., 1991). The corresponding gene has now been cloned from a mouse genomic library (Nahmias et al., 1991). Its overall

8-AR.

20336 Atypical p-Adrenergic Receptor in 3T3 Adipocytes

sequence is highly homologous to that of the human P3-AR (i.e. 83% identity within the i3P3 domain) whereas homology to h- and &AR does not exceed 50%. Full characterization of the murine &-AR is presented elsewhere (Nahmias et al., 1991).

By performing polymerase chain reaction using oligonucle- otides derived from the sequence of this murine P3-AR gene we have been able to amplify specifically a fragment of the expected size for the &AR. This fragment hybridizes to a subtype selective probe, derived from the N terminus of the human &AR, which reacts neither with the PI- nor with the @TAR genes. Thus, we have demonstrated that the transcript induced during 3T3 adipose conversion can hybridize to two nonoverlapping probes encompassing more than 40% of the entire &-AR sequence and derived from regions of low ho- mology with PI- and &AR genes. Furthermore, the apparition of this mRNA correlates with the emergence in 3T3-F442A adipocytes of an atypical P-AR pharmacological profile resem- bling that of the human ,f3,-AR. Together, these observations support the notion that the atypical P-AR expressed in 3T3- F442A, and tissular adipocytes, is the murine counterpart of the recently described human Pa-AR.

The low affinity of the atypical P-AR, compared with that of classical P-ARs, for the endogenous catecholamines raises the question of its physiological significance. Besides the possible existence of an unknown endogenous effector for the atypical (3-AR, this low affinity suggests that this receptor might be primarily involved under circumstances in which high levels of catecholamines are achieved, as is the case in sympathetic synapses. In this regard, because of the major noradrenergic innervation of the brown adipose tissue, the atypical 8-AR might be involved in thermogenesis rather than in lipolysis. Other metabolic pathways requiring higher recep- tor occupancy and cAMP levels could be controlled through this atypical P-AR. Alternatively, since the &AR does not display canonical sites for phosphorylation by a CAMP-de- pendent protein kinase or by the P-AR kinase (Emorine et al., 1991), it may maintain a minimal adrenergic sensitivity after desensitization of PI- and &ARs induced by prolonged exposure to high doses of catecholamines.

In conclusion, we demonstrate for the first time the exist- ence in 3T3-F442A adipocytes of a predominant &-AR pop- ulation, besides the classical PI- and PZ-ARs. The expression of these three pharmacologically distinct receptors is coordi- nated with the differentiation process and is hormonally regulated (Lai et al., 1982; Nakada et al., 1987; FBve et al., 1990). The preponderance of the f13-AR population suggests that it may play a central role in mediating the catecholamine- induced cAMP production in adipocytes. The hormonal and/ or environmental regulation of &AR expression should be a promising field of investigation for a better understanding of the energetic homeostasis.

Acknowledgments-We thank N. Scharapan for her secretarial services and Dr. A. Szabo for critical reading of the manuscript. W e gratefully acknowledge Dr. H. Green for the gift of the 3T3-F442A cell subclone.

REFERENCES

Arch, J. R. S. (1989) Proc. Nutr. Soc. 48,215-223 Arch, J. R. S., Ainsworth, A. T., Cawthorne, M. A,, Piercy, V., Sennitt, M. V.,

Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G., Smith Thody, V. E., Wilson, C., and Wilson, S. (1984) Nature 309 , 163-165

J. A., and Struhl, K. (1987) in Current Protocols in Molecular Siob& (Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G., Smith, J. A., and Struhl, K., eds) p. 495, John Wiley & Sons, New York

P h o r m o l . 8 4 , 131-137

Bahouth, S. W., and Malbon, C. C. (1988) Mol. Pharmacol. 34,318-326 Bojanic, D., Jansen, J. D., Nahorski, S. R., and Zaagsma, J. (1985) Br. J .

Bukowiecki, L.. Follea. N.. Vallieres. J.. and Leblanc. J. (1978) Eur. J. P h r - m o l . 92,189-196

. . , . .

Cathala, C., Savouret, J. F., Mendez, B., Karin, M., Martial, J. A., and Baxter,

Chomczynskl, P., and Sacchl, N. (1987) Anal. Btoehem. 162,156-159 Chung, F. Z. Lentes, K. U., Gocayne, J. Fitzgerald, M., Robinson, D., Kerlav-

age, A. R.,'Fraser, C. M., and Venter, b. C. (1987) FEBS Lett. 211,200-206 Dixon, R. A. F., Kobilka, B. K., Strader, D. J., Benovic, J. L., Dohlman, H. G.,

Frielle, T., Bolanowski, M. A., Bennett, C. D., Rands, E., Diehl, R. E., Mumford, R. A., Slater, E. E., Sigal, I. S., Caron, M. G., Lefkowitz, R. J., and Strader, C. D. (1986) Nature 321,75-79

J. D. (1983) DNA ( N Y ) 2: 329-335

Dooley, D. J., and Bittiger, H. (1987) J. P h a r m o l . Methods 18,131-136 Emorine, L. J., Marullo, S., Delavier-Klutchko, C., Karevi, S. V., Durieu-

Trautmann. 0.. and Strosbera A. D. (1987) Proc. Natl. Acad. Sci. U. S. A. 84,6995-6999 '

Emorine, L. J., Marullo, S., Briend-Sutren, M. M., Patey, G., Tate, K., Delavier-

Emorine, L. J.. and FQve, B., Pairault, J., Briend-Sutren, M. M., Marullo, S., Klutchko, C., and Strosberg, A. D. (1989) Science 246,1118-1121

Delavier-Klutchko, C., and Strosberg, A. D. (1991) Biochem. Phormacol. 4 1 ,

FQve. B.. Antras. J.. Lasnier. F.. Hilliou. F.. and Pairault. J. (1989) Mol. Cell.

-~

853-859 , .

Endocrinol. 6 7 , i7-27 , . . . .

F k e , B., Emorine, L. J., Briend-Sutren, M.-M., Lasnier, F., Strosberg, A. D.,

Frielle, T., Collins, S., Daniel, K. W., Caron, M. G., Lefkowitz, R. J., and

Green, H. (1979) in Obesity: Cellular and Molecular Aspects (Ailhaud, G., ed)

Green, H., and Kehinde, 0. (1976) Cell 7 , 105-113 Guest, S. J., Hadcock, J. R., Watkins, D. C., and Malbon, C . C. (1990) J. Biol.

Hollenga, C., and Zaa sma, J. (1989) Br. J. Pharmol . 9 8 , 1420-1424 Hollenga, C., Haas, d., Deinum, J. T., and Zaagsma, J. (1990) Horm. Metab.

and Pairault, J. (1990) J. Biol. Chem. 265,16343-16349

Kobilka, B. K. (1987) Proc. Natl. Acad. Sci. U. S. A. 8 4 , 7920-7924

Vol. 87, p. 15-24, Editions INSERM, Paris

Chem. 266,5370-5375

Res. 22.17-21 Kaestner, K. H., Christy, R. J., McLenithan, J. C., B r a h m a n , L. T., Cornelius,

P., Pekala, P. H., and Lane, M. D. (1989) Proc. Natl. Acad. Sci. U. S. A. 8 6 , 3150-3154

Kent, R. S., De Lean, A., and Lefiowitz, R. J. (1980) Mol. Pharmmol. 17, 14-

Kobllka, B. K., Frielle, T., Dohlman, H. G., Bolanowski, M. A., Dixon, R. A. F., Keller, P., Caron, M. G., and Lefltowltz, R. J. (1987) J. Bwl. Chem. 262,

Lai, E., Rosen, 0. M., and Rubin, C. S. (1982) J. Biol. Chem. 267,6691-6696 7321-7327

Lands, A. M., Arnold, A., McAuliff, J. P., Luduena, F. P., and Brown, T. G.,

Langin, D., Portillo, M. P., Saulnier-Blache, J. S., and Lafontan, M. (1991)

Levin, B. E., and Sullivan, A. C. (1986) J. Pharmacol. Exp. Therap. 2 3 6 , 681-

Lin, E. C. C., and Magasanik, B. (1960) J. Biol. Chem. 2 3 6 , 1820-1823 Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol.

Machida, C. A., Bunzow, J. R., Searles, R. P., Van Tol, H., Tester, B., Neve, Chem. 193,265-275

K. A,, Teal, P., Nipper, V., and Civelli, 0. (1990) J. Biol. Chem. 265,12960- 12965

--. - . ~~

23.

Jr. (1967) Nature 214,597-598

Eur. J. Pharmacol., 199,291-302

688

McPherson, G. A. (1985) J. Pharnacol. Methods 14,213-228 Mohell, N., and Dicker, A. (1989) Biochem. J. 261,401-405 Munson, P. J., and Rodbard, D. (1980) Aml. Bwchenz. 107,220-239 Muzzin, P., Seydoux, J., Giacobino, J.-P., Venter, J.-C., and Fraser, C. (1988)

Nahmias, C., &in, N., Elalouf, J. M., Mattei, M. G., Strosberg, A. D., and

Nakada. M. T.. Stadel. J. M.. Poksav. K. S.. and Crooke. S. T. (1987) Mol.

Biochem. Bio hys Res Commun. 156,375-382

Emorine, L. J. (1991) EMBOJ., in press

~ Phar&ol. 3'1, 377-384 '

Nakada. M. T.. Haskell. K. M.. Ecker. D. J.. Stadel. J. M.. and Crooke. S. T. I,

. ~~ ~ .. ~

(1989) Biochkm. J.~260,53-59 ~ ' '

O'Donnell, S. R., and Wanstall, J. C. (1980) Life Sci. 2 7 , 671-677 Pairault, J., Lasnier, F., and Laudat, M.-H. (1982) Eur. J . Bwchem. 127,351-

9 r o

Rothwell, N. J., Stock, M. J., and Sudera, D. K. (1985) Am. J. Physiol. 248 ,

Rubin C. S., Lai, E., and Rosen, 0. M. (1977) J. Biol. Chem. 252,3554-3557 Scatciard, G. (1949) Ann. N. Y. Acad. Sci. 51,660-672 Staehelin, M., Simons, P., Jaeggi, K., and Wigger, N. (1983) J. Biol. Chem.

Tate, K. M., Briend-Sutren, M. M., Emorine, L. J., Delavier-Klutchko, C.,

Wilson, C., Wilson, S., Piercy, V., Sennitt, M. V., and Arch, J. R. S. (1984)

Zaagsma, J., and Nahorski, S. R. (1990) Trends Phorrnacol. Sei. 11 , 3-7 Zaagsma, J., de Vente, J., Harms, H. H., and Jansen, J. D. (1985) in Phorma-

cobgy of Adrenoreceptors (Szabadi, E., Bradshaw, C. M., and Nahorski, S. R., eds) pp. 247-256, Macmillan Press, London

*do

E397-E402

268,3496-3502

Marullo, S., and Strosberg, A. D. (1991) Eur. J . Biochem. 196,357-361

Eur. J. Phormacol. 100,309-319