The kappa-opioid receptor from human placenta: Hydrodynamic characteristics and evidence for its association with A G protein

Life Sciences, Vol. 43, pp. 559-567 Printed in the U.S.A.

Pergamon Press

THE KAPPA-OPIOID RECEPTOR FROM HUMAN PLACENTA : HYDRODYNAMIC CHARACTERISTICS AND EVIDENCE FOR ITS ASSOCIATION WITH A G PROTEIN

*

PorthC G., Frances B., Verrier B. , Cros .I. and Meunier J.-C.

Laboratoire de Pharmacologic et de Toxicologic Fondamentales, Centre National de la Recherche Scientifique 205, route de Narbonne, 31077 TOULOlJSECCdex, F rance. Laboratoire de Biologie Cellulaire des Systemes Organises, INSERM U270 Facult6 de Mddecine Nord, Boulevard Pierre Dramard, 13326 MARSEILLE CBdex 15, France. (Received in final form June 16, 1988) SUMMARY

The K nature of opioid binding sites in a brush border membrane (BBM) fraction from human placenta has been confirmed : these sites display considerably higher apparent affinity (KI - 1.2 nM) for the K selective ligaqd U-504g8 than5they do for the u and 6 selective ligands CD-Ala ,2MePh3 , Glyol ! enkeCD-Thr ,Leu ! enkephalyl-Thr phalin (KI = 1.5-2 vM) and = lo-15 JIM), respectively. (KI The BBM fraction from human placenta was incubated either with the agonist 'Ii-etorphine or with the antagonist 'H-diprenorphine and subsequently solubilixed with digitonin. The solubilized macromolecular radioactivity was found to behave as a homogenous entity both in molecular exclusion chromatography (app. rs - 6.1 nm) and in linear sucrose gradients (app. S20 . w = 12 S). Two lines of evidence indicated that the placental K opioid receptor is capable of interacting with a guanine nucleotide regulatory (G) protein : (i) equilibrium binding of the agonist 'II-etorphine in the BBM fraction was clearly inhibited by 5'-guanylyl$nidodiphosphate (Gpp(NB)p), especially in the presence of Na ions while binding of the antagonist 'H-diprenorphine was significantly less so and (ii) the sedimentation velocity of the K opioid receptor was decreased down to about 10 S when the BBM fraction was prelabeled with radioligand in the presence of Gpp(NR)p prior to its solubilization with digitonin. The G protein that mediates the effect of Gpp(NH)p might be neither G nor G since no adenylate cyclase activity could be demonstrated in t#le BBM fraction from human placenta.

The presence of opiate binding sites (1,2) and of opioid peptides (8 endorphin (3), enkephalins (4,5) and dynorphin (6)) has been demonstrated in human placenta. This tissue contains a homogenous population of kappa opioid binding sites of glycoproteic nature which were located on syncitial brush border membranes (7,8,9). 0024-3205/88 $3.00 + .OO Copyright (c) 1988 Pergamon Press plc

560

Human Placental K-Gpioid Receptor Complex

Vol. 43, No. 6, 1988

In nerve tissue preparations, _in vitro equilibrium binding of agonists at u, 6 and K opioid sites is selectively inhibited in the presence of guanine nucleotides (10,ll) suggesting that a regulatory site for these effecters is involved in recognition of opioids by the three types of opioid receptor. In addition, JJ, 6 and K opioid receptors appear to be capable of interacting with JJ and K opioid receptors a guanine nucleotide regulatory CG) protein : (i) have recently been shown to exist under two molecular forms, the form of higher apparent molecular size being convertible into the other in the presence of 5'-guanylylimidodiphosphate, !12,13) (ii) the 6 opioid receptor interacts with a G protein since it is negatively coupled with adenylate cyclase in the hybr?d cell line NG 108 x 15 (14) and in rat striatum (15). In the present study, we have completed characterization of the human placental opjoid sites as a K type by use of u, 6 and K selective ligands in competition experiments against 'H-etorphine. We also provide indirect evidence for the interaction between the placental K opioid receptor and a G equilihrium binding of 'H-etorphine and of sH-diprenorphine in protein : (i) brush border membranes is differentially regulated by Gpp(NH)p and (ii) there is a decrease of apparent molecular size (sedimentation velocity) of the placental kappa opioid receptor when the BBM fraction has been prelabeled with 'H-diprenorphine in the presence of Gpp(NH)p prior to its solubilization with digitonin. The G protein that might underly these effects is likely to be different from Gi and G since we were unable to demonstrate any adenylate cyclase activity in the bSNsh border membrane fraction from human placenta.

METHODS Membrane preparations Human term placenta were collected on ice immediately after Cesarian section. A microvillus brush border memhrane (BBM) fraction was prepared as previously described (9). A basal membrane (BM) fraction was prepared as control according to Whitsett et al. (16). Equilibrium binding studies The BBM fraction (0.35 mg protein) was incubated, in triplicate, with 'H-etorphine or 'H-diprenorphine in the presence or in the absence of NaCl (see legend of the figures). Non specific binding was and/or of Gpp(NH$ measured with 10 M levorphanol. After 30 minutes at 37'C, the samples were rapidly filtered under reduced pressure through Whatman glass fiber disks (GF/B) and washed with IO ml of ice cold buffer (Tris-HCl, 0.05 M, pH 7.4). The filters were dried and counted in 3 ml of MP Ready-solv cocktail (Beckman) in a Kontron model MR 300 automatic liquid scintillation system. Protein concentrations were estimated by the method of Lowry et al. (17). Labeling and solubilixation of the BBM fraction The BBM suspension (5 mg protein/ml) was incubated for 30 minutes at 37'C in buffer either with 3 nM 'H-etorphine or with 3 nM 'H-diprenorphine in ite absence (total labeling) or in the presence (non specific labeling) of 10 M levorphanol. All subsequent steps were carried out at 0-4°C. Hundred ul aliquots of the labeled suspension were assayed, in triplicate, for specifically bound radioactivity by filtration (see above). The bulk of the suspension was supplemented with enough sol_Lddigitonin to make 1 X (w:v), gently stirred for 30 min and centrifuged in a Beckman rotor type 40.3 for 45 min at 38000 rpm. The resulting supernatant was taken as the soluble extract, 0.1 ml of which was immediatly assayed, in triplicate, for specifically bound radioactivity by rapid filtration on polyethyleneimine (0.33 X) treated glass fiber djsks (18). The yield of solubilization, i.e. the ratio of the amount of soluble to membrane bound radioactivity retained on filters was in the range ._ 30-40 X. Molecular exclusion chromatography An aliquot (2.5 ml) of the radioactive soluble extract supplemented with bovine liver catalase (Cl00 Sigma Chem. Co) as reference was applied to

Vol. 43, No. 6, 1988

Human Placental r-Gpioid Receptor Complex

561

columns (1.5 x 40 cm) of packed Sepharose 6B (Pharmacia).The columns were fully equilibrated and eluted with buffer containing 0.1 XI (w:v) digitonin. Elution was carried out by a peristalticpump (Technicon)at a constant flow rate of - 0.15 mllmin. Sixty fractions of 1.65 ml were collected. The radioactivity in 0.2 ml aliquots of each fraction was counted as described above. The fractions were assayed for catalase activity by the method of Chance and Maehly (19) with H202 as substrate. Sedimentationin sucrose gradient The stock solutions consisted of 5 and 20 X (w:v) sucrose (Merck) in 50 mM Tris HCl pH = 7.4 containing 0.1 X (w:v) digitonin. Linear 5 to 20 XI sucrose gradients of 2 x 5,2 = 10.4 ml were formed in "Ultra clear" (Beckman) centrifuge tubes using a classical lucite gradient maker. The radioactive soluble extract (0.5 ml) with catalase (0.02 ml) was layered on top of the gradients and centrifugedat 40.000 rpm (Beckmanrotor SW 41 Ti) for 16 hrs at 2'C. Fractions of about 0.16 ml (7 drops) were collected directly into scintillationvials and counted for radioactivityas described above. Adenylate cyclase assay Assays were performed (for 15 min) at 30°C in a final volume of 50 1.11. The incubation medium contained 120 ug protein oJ2 BBM or basal membrane fraction with 4 mM MgCl ; 0.1 mM ATP ; 0.75 JJC~ ( P) ATP ; 0.5 mM EGTA, a creatine phosphokinaseJrP regenerating system (15 mM creatine phosphate, 8 U/ml creatine kinase) and 0.1 X bovine serum albumine in 20 mM Tris HCl pH = 7.8. Cyclic AMP produced was measured as in (20).

RESULTS 1) Equilibriumbinding studies a- Pharmacologicalprofile : In the absence of effecters,eH-etorphineor 'Ii-diprenorphine labeled the same maximal number of binding sites (- 0.25 pmol/mg of protein) with KD values of 0.36 nM and 0.31 nM, respectively(fig. 1) in the BBM fraction from human placenta. We had previously shown (8) that the pharmacologicalprofile of 'H-etorphineand 'H-ethylketocyclazocine binding in placental membranes was typically K as the latter type had been defined in guinea-pigbrain homogenates (21). This notion was confirmed here by use of ligands more selective of N, 6 and K opioid sites, i.e. DAGO, DTLET and U-50488, respectively.Table I shows that U-50488 was much more potent than DAGO and DTLET in competing with binding of 'H-etorphinein the BBM fraction from human placenta. TABLE I

Drug U-50488 DAGO DTLET

KI (nM)

nHil1

1.3 1,800 13,000

0.94 1.03 1.12

Apparent affinities of N (DAGG), 6 (DTLET) and K (U50488) selective ligande for placental opioId binding sites. K values were calculated from the concentrations of unlabeled ligand causing f0 X inhibitionof binding of "H-etorphine used at the fixed concentrationof 1 nM (22). U50488 = trans 3,4 dichloro N methyl N (2(1-pyrolinyl-cyclohexyl)-benzene-acetamide was kindly supplied by Dr. R. Lahti (Upjohn, Kalamazoo, MI, USA). DAGG - Tyr-D Ala-Gly-MePhe-Glyol and DTLET - Tyr-D Thr-Gly-Phe-Leu-Thrwere from Cambridge Research (Harston, England)

562

Human Placental K-Opioid Receptor Complex

Vol. 43, No. 6, 1988

In addition, the apparent Hill coefficient of the inhibition curves of the three unlabeled ligands were close to unity (Table 1) indicating a homogenous population of opioid binding sites in this preparation. The values of the apparent affinity of the JJ, 6 and K selective ligands for the placental opioid sites were very similar to those reported for the K opioid sites in guinea-pig brain (21) and cerebellum (23,lO) membrane preparations. b- Regulation of 'II-etorphine and 'H-diprenorphine binding by Gpp(NH)p and NaCl. Equilibrium binding of the opiate agoniqt, 'H-etorphine, to the placental ions and/or Gpp(NH)p (fig. 1A). K receptor was fqund to be sensitive to Na Inhibition by Na ions was mainly characterized by a decrease in the maximal number of sites, whereas Gpp(NH)p caused a 2-3 fold reduction of 'I-I-etorphine affinity. The simultaneous presence of the two effecters resulted in a maximally reduced binding of 'H-etorphine : Gpp(NH)p appeared to decrease the affinity of the sites remaining in the presence of sodium. The effect of Gpp(NH)p on 'II-etorphine (1 nM) binding was concentration-dependent with 50 X of inhibition at about 250 JJM of Gpp(NH)p (fig. 2). Equilibrium binding of the opiate an+agonist, 'H-diprenorphine was less sensitive to the inhibitory effects of Na ions and@r of Gpp(NH)p than binding of 'II-etorphine (fig. 1B). In particular, Na ions had no effect on 'H-diprenorphine's maximal binding capacity while Gpp(NH)p only slightly decreased the affinity of the tritiated opiate antagonist. In fact, binding of 'H-diprenorphine (1 nM) in human placenta BBM was found to be quite insensitive to inhibition bv the GTP analonue L. at concentration of the latter up to 500 JJM (fig. 2).

B/F

I

100

FIG. 1 Equilibrium saturation binding of 'Hetorphine (A) and 'H-diprenorphine (B) in placental brush border membrane fraction in 50 mM Tris-HCl pH 7.4 (0) + 100 JJM Gpp(NH)p (0) ; + 120 mM NaCl (A) ; + 100 uM Gpp(NH)p + 120 mM NaCl (A). Scatchard transformation of the data. B (pmol/mg of protein) is the amount of specifically bound radioligand and F (nmol/l) is the concentration of free radioligand.

20

5

GPP(NH)P.

PM

FIG. 2 Effect of different concentrations of Gpp(NH)p on equilibrium binding of 'H-etorphine, 1 nM(O) and 'H-diprenorphine, 1 nM(O) in a BBM fraction from human placenta. Specific binding is expressed as percent of control measured in Tris-HCl buffer.

Human Placental

Vol. 43, No. 6, 1988

2) Characterization cular complexes

of

solubilized

K-Opioid

‘II-etorphine

Receptor

Complex

563

or %I-diprenorphine

macromole-

a- Molecular exclusion chromatography Digitonin extracts form the ‘H-etorphine prelabeled BBM fraction were submitted to molecular exclusion Sepharose 6B. No chromatography on radioactivity was recovered in the void volume (Vo) of the column, indicating excellent solubility in 0.1 X (w/v) digitonin (fig. 3). Instead, part of the radioactivity was eluted along with ferritin (rs - 6.1 nm) as a rather sharp, highly symmetrical peak totally separated from free ‘H-etorphine. The “Hetorphine macromolecular complex in question appeared to be specific since it was absent in soluble extracts from BBM that had been labeled in the presence of an excess of levorphanol (fig. 3).

fraction

number

FIG. 3 Molecular exclusion chromatography in Sepharose 6B of the radioactivity solubilized from ‘H-etorphine labeled brush border membrane ._,$iBMwere labeled in the absence (0) and in the presence (0) of 10 M levorphanol. The column was calibrated with standard globular proteins of known Stokes radius (rs) : bovine serum albumin (BSA, 3.55 nm), catalase (5.22 nm), ferritin (6.1 nm) and thyroglobulin (8.5 nm). The calibration curve f.insert)was according to Porath (24) with Kav = (Ve - Vo>/(Vt - Vo), Ve being the elution volume, Vo the void volume and Vt the total volume of the column. b- Sedimentation in linear sucrose gradient Fig. 4 shows sedimentation profiles of the radioactivity solubilized from 'H-etorphine (A) or 'H-diprenorphine (B) labeled placental BBM. In the absence of effecters, a very sharp peak of macromolecular radioactivity was observed which was sedimented slighly ahead (- 12 S> of catalase (S20 w = 11.4 S), the

,

564

Human Placental K-Gpioid Receptor Complex

Vol. 43, No. 6, 1988

protein used as velocity marker. At first sight, the macromolecular radioactivity peak appeared to be symmetrical and therefore homogenous with respect to its constituents. When the BBM had been incubated in the presence of 120 mM NaCl, both "Hetorphine and 'H-diprenorphine labeled K placental receptors were recovered at the same position as in controls, i.e. 12 S, with a small shoulder at position 10 S. Recovery of bound 'H-dtorphine was substantially reduced, whereas recovery of macromolecular 3H-diprenorphine was unchanged. This reduction could be related to the specific decrease of 'H-etorphine capacity caused by NaCl in binding studies (Fig. 1). Next, we examined the effect of preincubating RBM and radioligand in the presence of 200 JJM Gpp(NH)p. This effector strongly inhibited labeling of the 12 S component with either tritiatecl ligand agonist ('H-etorphine) or antagonist (sH-diprenorphine). Under these conditions, a radioactivity peak appeared at position 10 S for the two ligands, indicating that Gpp(NH)p had shifted a part of the 'H-diprenorphine or 3H-etorphine labeled K receptor from its position 12 S to the position 10 S. This effect seems to be specific of guanine nucleotide since App(NH)p did not induce this shift (data not shown).

fractbn

number

fraction

number

FIG. 4 Sedimentation ?n linear (5-20X) sucrose gradients of the radioactivity solubilized from the 'H-etorphine (A) and 3H-diprenorphine (B) prelabeled BBM fraction from human placenta. Prelabeling had been carried out in 50 mM TrfsI+Kl, pH 7.4 CO), + 120 mM NaCl (A), + 200 pM Gpp(NH)p (A), + 10 M levorphanol (non specific) (0) with 3 r&l radioligand.

Vol.

43, No. 6, 1988

Human Placental

K-Gpioid

Receptor

Complex

565

However, the 10 S peak of bound ‘II-etorphine was much smaller than the one of bound sH- diprenorphine and total recovery of the bound agonist was considerably lower than total recovery of bound antagonist. Such a difference was not expected from saturation binding data since the amounts of membrane-bound ‘Ii-etorphine (used at the saturating concentration of 3 nM) were nearly the same in the absence and in the presence of Gpp(NH)p. It could be, hovever, that under these sedimentation conditions (non equilidissociated faster than did ‘II-diprenorphine from the 10 brium) , “H-etorphine S component. A similar discrepancy was actually observed, under comparable in the case of the u-opioid receptor from rabbit experimental conditions, cerebellum (12). 3) Adenylate cyclase activity Adenvlate cvclase (AC) activity had been vreviouslv demonstrated in a basal membrane (BM) fraction from human placenta-(25). We-therefore used this particular fraction as a positive control to assess for AC activity in the BBM fraction. Indeed, as Table II shovs, the BM fraction exhibited an AC activity which could be activated either indirectly (in the presence of & uM Gp@H)p or of 100 mMNaF) or directly by forskolin in the presence of Mn orMg . In contrast, no adenylate cyclase activity could be demonstrated under identical conditions in the BBM fraction from human placenta, i.e. in the fraction where the K opioid receptors are concentrated.

Adenylate

cyclase

assay

TABLE II on placental

membrane fractions

Enzyme specific_fctivityI (pm01 AMPc x mg x min BBM basal Gpp(NH)p (10-5M) NaF (100 mM) For&#in + (IO M)

MgCl2 (4 mM) MnCl, (4 mM)

< < < < <

1.0 1.0 1.0 1.0 1.0

)

Basal. membrane 2.6 6.0 34 9.3 26

DISCUSSION The present study has confirmed previous data from this laboratory indicating that the human placenta contains a homogenous population of K opioid : in this preparation there vas no evidence for a high affinity composites nent for the p and 6 selective ligands DAGC and DTLET, respectively. The brush horder membrane fraction from human placenta therefore designates itself clearly as a convenient human preparation to biochemically characterize a K opioid receptor. The placental K opioid site appears to be carried by a unique macromolecular entity which, in digitonin solution, displays an apparent sedimentation coefficient - 12 S and an apparent stokes radius - 6.1 nm. The latter value is somewhat lower than the one (- 7 nm) reported by others (26) for the CpAPS solubilized placental opioid receptor. Such difference may actually reflect the use of different detergents in the two studies. In fact, the placental K-opioid receptor exhibits hydrodynamic characteristics which under identical experimental conditions, are indjstinpuishable from those of the K opioid

566

Human Placental K-Opioid Receptor Complex

Vol. 43, No. 6, 1988

receptor from guinea-pig cerebellum (submitted for publication). In placental brush border membranes, equilibrium bind,$g of the agonist 'H-etorphine was found to be sensitive to inhibition by Na ions yet significantly less so than previously found in guinea-pig cerebellum membranes (IO). Most importantly however, binding of the tritiated agonist to the membrane bound placental K receptor was sensitive to the inhibitory effect of Gpp(NH)p and this, as in the case of+the membrane bound cerebellar K receptor, especially in the presence of Na ions. Guanine nucleotide regulation of hormonereceptor interactions is generally considered symptomatic of the association between receptor and a guanine nucleotide regulatory (G) protein. In the present study, the latter hypothesis was strongly supported by the observation that the apparent molecular size (apparent sedimentation coefficient) of the placental K opioid receptor was decreased in the presence of Gpp(NH)p. Although one cannot exclude the possibility that the Gpp(NH)p-induced decrease of K-opioid receptor size follows from changes in receptor conformation, a similar phenomenom has been reported for several other types of hormone receptor (27,28,29,30) and is generally taken to reflect a guanine nucleotide induced dissociation of a receptor-G protein complex. Interestingly, in all cases the difference between the apparent sedimentation of the "free" (uncoupled) and of the G coupled receptors is about 2 Svedberg units, a value which corresponds well to the apparent Mr of the a subunit of a G protein in solution (31). The identity of the G protein that is reported here to interact with K opioid receptors in brush border membranes from human placenta remains to be determined. Several such regulatory proteins have been characterized to date (32,33) including one, G , in human placenta (34). In general, the functionnal role of the G protein i% not clearly established, except for the three best characterized ones, G and Gi which are known to positively and negatively couple hormone receppors to adenylate cyclase, respectively and G which allows rhodopsin to stimulate a high affinity cyclic GMP phosphodietsterase upon photoactivation (33). Although adenylate cyclase activity was readily detected in basal membranes, we failed to demonstrate it in brush border membranes from human placenta and this, not only in the presence of the "indirect" activators Gpp(NH)p and NaF, in agreement with Whitsett et al . (35) but also and above all in the presence of forskolin which is known to directly activate the eneyme. Placental BBMs would then appear as a unique preparation which lacks adenylate cyclase but which does contain Kc-opioid receptors. This situation implies that K-opioid receptors cannot be functionally coupled with this enzyme in the human placenta and it is therefore likely that the G-protein which interacts with K-opioid receptor in placenta3 BBMs (the present study) is neither Gs nor Gi.

REFERENCES 1. A. VALETTE, G. PONTONNIER and J. CROS, FEBS Lett. 103, 362-365 (1979). 2. A. VALETTE, J.M. REME, G. PONTONNIER and J. CROS, Biochem. Pharmacol. -29, 2657-2661 (1980). 3. Y. NAKAI, K. NAKAO, S. OK1 and H. IMURA, Life Sc1. 2, 2013-2018 (1978). 4. B.V.R. SASTRY, S.L. RARNWELL, O.S. TAYEB, V.E. JANSON and L.K. OWENS, Biochem. Pharmacol. 2, 475-478 (1980). 5. L. TAN and P. YU, Biochem. Rjophys. Res. Can. 98, 752-760 (1981). 6. S. LEMAIRE, A. VALETTE, L. CHOUINARD, M. DUPUIS, R. DAY, G. PORTHE and J. CROS, Neuropeptides, 2, 181-191 (1983). 7. G. PORTHE, A. VALETTE and J. CROS, Biochem. Biophys. Res. Corn.,3, 1-6 (1981). 8. A. VALETTE, P. ROUGE, E. COULAIS, G. PONTONNIER, J. CROS and E.J. SIMON tife Sci. 40, 535-540 (1987)

Vol. 43, No. 6, 1988

Human Placental K-Opioid Receptor Complex

567

9. G. PORTHE, A. VALETTE, A. MOISAN'D,M. TAFFANI and J. CROS, Life Sci. 2, 2647-2654 (1982). 10. B. FRANCES, C. MOISAND and J.C. MEUNIER, Eur. J. Pharmacol. -117, 223-231 (1985). 11. A.J. BLUME, Life Sci. 22, 1843-1852 (1978) 12. P. JAUZAC, B. FRANCES,T. PUGET, C. MOISAND and J.C. MEUNIER, J. Rec. Res. 6, l-25 (1986) 13. B. FRANCES, G. PORTHE, C. MOISAND and J.C. MEUNIER, I.N.R.C. Abstracts 166 (1987) 14. A.J. BLUME, D. LICHTSHTEIN and G. BOONE, Proc. Natl. Acad. Sci. USA 76, 5626-5630 (1979) 15. P. Y. LAW, J. WU, J.E. KOEHLER and H.H. LOH, J. Neurochem. 2, 1834-1846 (1981) 16. J.A. WHITSETT, J. Pediatr. 96, 600-604 (1980) 17. O.H. LOWRY, M. J. ROSEBROUGH,A.L. FARR and R.J. RANDALL, J. Biol. Chem. 193,

265-275

(1951)

18.

R.F. BRUNS, LAWSON-WENDLINGand T.A. PUGSLEY, Anal. Biochem. 132, 74-81 (1983) 19. B. CHANCE and A.C. MAEHLY, Academic Press. New York 2, 764-768 (1955) 20. Y. SALOMON, C. LONDOS and M. RODBELL, Anal. Biochem. 2, 541-548 (1974) 21. H.W. KOSTERLITZ and S.J. PATERSON, Proc. R. Sot. Lond. B. -210, 113-122 (1980) 22. Y.C. CHENG and W.H. PRUSSOF, Biochem. Pharmacol.2, 3099-3102 (1973) 23. L.E. ROBSON, R.W. FOOTE, R. MAURER and H.W. KOSTERLITZ, Neurosci. 2, 621-627

(1984)

24. 25.

J. PORATH, Pure Applied Chem. 2, 233-244 (1963) J.A. WHITSETT, C.L. JOHNSON, A. NOGUCHI, C. DARVEC-BECKEMAN and M. COSTELLO, J. of Clin. Endocrinol.Metab. 0, 27-32 (1980) 26. M.S. AHMED, Membrane Biochem.2, 35-47 (1983) 27. L.L. LIMBIRD and R.J. LEFKOWITZ, Proc. Natl. Acad. Sci. USA 11, 228-232 (1978) 28. S.K. SMITH and L.L. LIMBIRD, Proc. Natl. Acad. Sci. USA 78, 4026-4030 (1981) 29. A. COUVINEAU,B. AMIRANOFF and M. LABURTHE, J. Biol. Chem. 261, 14482-14489 (1986) 30.

T.J. FITZGERALD,R.J. UHING and J.H. EXTON, J. Biol. Chem. 26J,

16871-16878

(1986) 31.

32. 33. 34. 35.

L. BIRNBAUMER, J. CODINA, R. MATTERA, R.A. CERIONE, J.D. HILDEBRANDT,T. SUNYER, F.J. ROYAS, M.G. CARON, R.J. LEFKOWITZ and R. IYENGAR, Recent Progress in Hormone Research 41, 41-99 (1985) S. COKCROFT,Trends Biochem. Scl. 12, 75-78 (1987) J. BOCKAERT,V. HOMBURGER and B. ROUOT, Biochimie 69, 329-338 (1987) T. EVANS, M.L. BROWN, E.D. FRASER and J.K. NORm, J. Biol. Chem. 261, 7052-7059 (1986) J.A. WHITSETT, C.L. JOHNSON and K. HAMKINS, Am. J. Gbstet. Glynecol.133, 204-207 (1979)