Eur. J . Biochem. 74, 243-252 (1977)
LABORATOIRE ez
mcd :o.'ogio SAZT TllMAN LIEGE 1
/r?>lilut c's i'r
4000
Thyrotropin Binding to and Adenylate Cyclase Activity of Porcine Thyroid Plasma Membranes Bernard VERRIER, Richard PLANELLS, and Serge LISSITZKY Laboratoire de Biochimie Medicale et Unite 38 de I'Institut National de la SantC et de la Recherche Medicale, Faculte de Medecine, Marseille (Received November 24, 1976)
Plasma membranes have been purified from porcine thyroid gland homogenate by discontinuous sucrose gradient centrifugation. The preparations contained specific binding sites for thyrotropin but not for luteinizing hormone or the p subunits of thyrotropin and luteinizing hormone. Optimum conditions of '2sI-labeled thyrotropin binding were pH 6.0-6.5 and 37 "C. Thyrotropin binding was reduced by divalent (Ca", Mg Zf)and monovalent cations (Na', K', Li'), 50% inhibition being obtained at 10 mM and 50 mM respectively. Displacement curves of '2sI-labeled bovine or porcine thyrotropin by the unlabeled hormone from three species was in the order of increasing concentrations (bovine > porcine > human) which is the order of decreasing biological activity of these hormone preparations in the assay in vivo in the mouse. The validity of the results was established by controlling that porcine membranes bound the native and the '251-labeled hormones with equal affinity. A single type of high-affinity (& = 0.28 nM) binding sites was detected for bovine and porcine thyrotropins. In contrast, porcine plasma membranes bound human thyrotropin with a lower affinity (& = 70 nM). A good correlation was found at equilibrium and in the conditions o f the cyclase assay, between receptor occupancy and adenylate cyclase activation for the three hormones. The consensus of investigators agrees that the first event in the action of thyrotropin on target tissues is the binding to receptors in the plasma membrane and the consecutive stimulation of adenylate cyclase (see [l] for a review). In previous studies [2,3], we have demonstrated in both i n t h porcine thyroid cells and their derived plasma membranes the presence of a single class of specific sites that bind '251-labeled porcine thyrotropin with high affinity. In addition a quantitative correlation between binding and adenylate cyclase activation was established. Other investigators [4,5]using plasma membranes from bovine glands suggested the possible existence of two types of binding sites as shown by non-linear Scatchard plots of saturation binding experiments and were unable to correlate binding with adenylate cyclase activation. The interpretation of non-linear Scatchard plots is not clear and many models have been proposed to account for non-linearity such as cooperative binding interactions, polymerization of hormone, heteroAbbreviation. Cyclic AMP, adenosine 3' : 5'-monophosphate. Enzymes. Adenylate cyclase (EC 4.6.1.1) ; p-nitrophenylphosphatase (EC 3.1.3.41); glucose-6-phosphatase (EC 3.1.3.9); SUCcinate dehydrogenase (EC 1.3.99.1).
geneity of receptors or of labeled hormones (see [6] for references) and recently difference of affinity of the receptor for the labeled and unlabeled hormones [61. Since differences in binding characteristics may be related to the nature of the starting material (gland or intact cultured cells) used to prepare plasma membranes, we have investigated the specific binding of thyrotropin to plasma membranes purified from porcine thyroid glands. In this paper, we describe the preparation of these membranes, the analysis of the binding of thyrotropins from different species and the correlation between binding and adenylate cyclase activation for these hormones. MATERIALS AND METHODS Purfication o j Plasma Membranes
Porcine thyroid glands were collected at the local slaughter house, immediately immersed in chilled Hank's saline buffer and processed as soon as possible. After careful dissection to remove fat and connective tissue, slices were obtained using a Stadie-Riggs slicer and homogenized in 0.25 M sucrose, 1 mM MgC12,
Porcine Thyroid Plasma Membranes and Thyrotropin
244
10 mM Tris-CI pH 7.4 (1 g slices in 2 ml buffer), first with an Ultra-Turrax at minimum speed for a few seconds and then, by hand, in a glass teflon Potter homogenizer (two strokes). The homogenate obtained from 100 g glands was diluted to obtain a final volume of 800 ml and centrifuged at 50 x g for 10 min. The supernatant was filtered on a stainless steel sieve. The residual tissue fragments remaining on the sieve and the 50 x g pellets were rehomogenized in 200 ml buffer and the homogenate was centrifuged in the same conditions. Both supernatants were then spun at 500 x g for 10 min, the pellets discarded and the supernatant centrifuged at 10000 x g for 10 min. The 10000 x g pellets were gently homogenized in 1 mM MgC12, 10 mM Tris-CI pH 7.4, using a glass/teflon homogenizer and were enriched with sucrose in buffer to obtain a final concentration of 43.3% (w/w) for a total volume of 90 ml. 15 ml of the sucrose-enriched suspension were layered in 38.5-nil tubes over 8 ml 45 (w/w) sucrose solution and overlaid by sucrose solutions of 41% (10 ml) and 8 % ( 5 ml). The six discontinuous sucrose gradients thus obtained were spun in a Spinco model L2-50 centrifuge in rotor SW27 at 24000 rev./min for 2 h at 1 "C. Plasma membranes were harvested by aspiration with a pasteur pipette at the interface 41 %,/8%, diluted (1 : 6) with 10 mM Tris-C1 pH 7.4 and centrifuged at 10000 x g for 10 min. The pellet was resuspended in 50 mM Tris-CI pH 7.4 to obtain a final concentration of 1.2- 2 mg protein/ml. Aliquots of 0.2-0.4 ml were stored at -196 "C. Hormone Preparations Highly purified porcine and human thyrotropins were obtained from Dr G. Hennen (Liege, Belgium). The thyrotropic activity of porcine thyrotropin was initially 39 U/mg as determined by the McKenzie bioassay in the mouse [7]; after storage for two years in the dry form the activity decreased with time to reach values below 10 U/mg (lot 2 = 9.5 U/mg, lot 1 = 7.4 Ujmg) when the experiments related in this paper were performed. The specific bioactivity of the human hormone was 3.3 U/mg. The high purity of this preparation was checked by amino acid composition closely resembling that of the bovine and porcine hormones, N-terminal amino acid analysis and polyacrylamide gel electrophoresis (G. Hennen and J. Closset, personal communication). Bovine thyrotropin (42 U/mg) was the generous gift of Dr J. G. Pierce (Los Angeles, Calif., U.S.A.). Labeling of thyrotropin with "'I was carried out as previously described [8] using the lactoperoxidase method of iodide oxidation except that the iodination mixture was made in 0.5 M aceto-acetate buffer pH 5.6, and separation of the labeled hormone was performed in a single step on a Sephadex G-100 column (1 x 70 cm) equilibrated in
50 mM Tris-C1 pH 7 containing 0.1 % bovine serum albumin [l]. Specific radioactivities of 60 - 90 Ci/g were obtained. As compared to the native hormone, no loss of biological activity due to iodination was observed using the bioassay in vitro previously described [9]. I-Labeled Tlzyrotropin to Plasma Membranes
Binding
Measurement of labeled thyrotropin binding to plasma membranes was performed in 50 mM Tris-C1 pH 7.4containing 1 %bovine serum albumin (standard binding buffer), 0.4- 0.8 mg/ml of membrane protein, and labeled thyrotropin in a final volume of 0.0500.2 ml. Incubations were routinely performed at 27 'C and separation of membrane-bound from free thyrotropin was carried out by centrifugation as previously described [2]. Radioactivity of the pellets was estimated in a Packard Autogamma spectrometer with a counting efficiency of 25 Specific binding was obtained by substracting from the total radioactivity bound the amount which was not displaced by an excess of native thyrotropin. All assays including control were performed in duplicate or triplicate. Membranes which have bound porcine '"I-labeled thyrotropin were analyzed by gradient centrifugation as described below. Membranes were incubated at 27 "C in 50 mM Tris-C1 pH 7.4, 1 bovine serum albumin for 25 min, and then washed from free 1251-labeledthyrotropin by centrifugation at 10000 x g for 10 min. The pellet was suspended in 50 mM Tris-C1 pH 7.4 containing 1 '%; bovine serum albumin and 200 pg of native membrane proteins as carrier and was loaded on a 20-45% (w/w) sucrose gradient in 10 mM Tris-C1 pH 7.4. The gradient was spun at 24000 rev.jmin in a 17-ml SW27 bucket (Beckman) at 1 L'C.Gradients were fractionated and absorbance at 280 nm and "'I radioactivity measured using standard conditions.
'x.
Assay of Adenylate Cjduse Activity Assays were performed in a final volume of 50 p1 in 50 mM Tris-C1 pH 7.5, containing 0.1 04 bovine serum albumin, 4 mM MgC12, 4 mM phosphoenolpyruvate, 400 mU pyruvate kinase, 1 mM cyclic AMP and the indicated concentrations of [a-"PIATP (10- 50 counts min-' pmol-I). Membrane protein at a concentration of 500 - 760 pg/ml and thyrotropin at different concentrations were added. When used, NaF concentration was 10 mM. Total incubation medium without ATP was preincubated for 7 min at 30 "C. In the case of kinetic experiments 50-pl aliquots were taken off at given intervals from a conveniently sized pool. The reaction was started by addition of substrate after preincubation for 7 min at 30 "C. At various times of incubation, the reaction was stopped
B. Verrier, R. Planells, and S. Lissitzky
by the addition of 150 p1 50 mM Tris-C1 pH 7.5 containing 1.33% sodium dodecylsulfate, 40 mM ATP and 1.4 mM cyclic AMP. 50 pl of cyclic [3H]AMP (2- 3 x lo4 counts/min per assay) and 750 pl of water were added. Cyclic AMP was determined as described by Saloinon et al. [lo] with slight modifications. Columns of Dowex AG-50-W x 4 containing 4.5 ml of resin were washed with 5 ml 1 M NaOH, 10 ml bidistilled water, 5 ml 3 M HC1 and then 10 ml of bidistilled water. After use, the column was regenerated by an identical cycle. Between assays resin columns were kept in water. For further purification of cyclic AMP, alumina columns were utilized as indicated by the authors. In all the assays performed (more than 500) cyclic AMP recovery was 60 - 80 %. Assay blanks containing no membrane were processed in the same way. Radioactivity in the blanks were less than 0.0005% of the total added. Assays were performed in triplicate or quadruplicate. We have verified by an electrophoretic separation [ l l ] that in our conditions, no more than 10 % cyclic AMP was hydrolysed during the course of 20-min incubations. Degradation of ATP which could occur in our conditions of incubation was determined by thin-layer chromatography on poly(ethy1eneimine)-cellulose in 1.2 M LiCl [12]. N o more than 20% of the nucleotide was destroyed, even at 0.1 mM ATP. To verify that the radioactive material isolated by double-column chromatography was effectively cyclic AMP we processed as described. At the end of the 20-min incubation period, membranes were removed by centrifugation at 10000 x g for 10 min and the supernatants were incubated for 10 min at 30 'C with 50mU beef heart phosphodiesterase. These samples did not show radioactivity in the cyclic AMP fraction while control assays processed in the same way retained radioactivity. Orlzer. Assays
Ouabain-sensitive K+-stimulated p-nitrophenylphosphatase activity was determined at 25 "C and 410 nm in 50 mM Tris-C1 buffer at pH 7.5 containing 20inM KCI, 20mM MgClz and 20mM p-nitrophenyl phosphate in the presence or in the absence of 0.2 mM ouabain [13]. Glucose-6-phosphatase and antimycin-sensitive succinate dehydrogenase activities were assayed using standard methods [14,15]. Membrane protein was determined according to Lowry et al. [16] using serum albumin as standard, after solubilization in 1 M NaOH. A z x I / A ~ ratios ~o were estimated in 2 '%,sodium dodecylsulfate. Materials [a-32P]ATP(0.5- 1 Ci/mol) and cyclic [3H]AMP (30 Ci/mmol) were purchased from the Radiochemical Center (Amersham, England). ATP, cyclic AMP, p-
245
nitrophen ylphosphate and crystallized pyruvate kinase were obtained from Boehringer, Mannheim (West Germany). Phosphoenolpyruvate and lyophilized pyruvate kinase were obtained from Sigma (St Louis, U.S.A.). All other chemicals were of the highest purity available from commercial sources. Other.Details
All curves were traced using linear regression calculation. Typical experiments are presented. A molecular weight of 30000 for thyrotropin was used for all calculations. RESULTS
Plasma Membrane Preparations The plasma membrane fraction was obtained from porcine thyroid gland homogenates after discontinuous sucrose density gradient centrifugation of the 10000 x g pellet at the interface 8%/41% (d = 1.18 g/ ml). In 15 preparations the yield was about 0.2 mg membrane protein per g fresh slices. Marker enzyme data are summarized in Table 1. As compared to the particulate fraction obtained by centrifugation at 105000 x g of the 50 x g supernatant fraction, the purified membranes showed a decrease in glucose-6phosphatase and succinate dehydrogenase activities of about 2-fold and an increase in K 'pnitrophenylphosphatase activity (a plasma membrane enzyme marker) of about 5-fold. As compared to the 50 x g fraction, an 18-fold increase in the latter enzyme activity was observed. Although less purified than the membranes previously obtained from isolated porcine thyroid cells [2], these membrane preparations contained an active thyrotropin-stimulated and F--stimulated adenylate cyclase and showed a good capacity to bind labeled thyrotropin (see below). Basal, thyrotropinTable 1. Comparison of enzymic activities of total particles and purified plu,smamembranes of porcine thyroid gland homogenate.v Each number is the mean of duplicate determinations in three different preparations; the individual values agreed within 5- 10 Fraction
Ouabainsensitive p-nitrophenyl phosphatase
Glucose-6phosphatase
Antimycinsensitive succinate dehydrogenase
A28~iA~h~
nmol rn1n-l (mg protein)-' _ ~ _ _ _ ~ - - 105000 x g pellet 33 Plasma membranes 15 1
29 0
0 65
0 49
14.0
0 30
0 98
Porcine Thyroid Plasma Membranes and Thyrotropin
246
Fraction number
Fig. 1. Sucrose density gradient centrifugation ofporcine thyroid plusma niembranes. 200 pg membrane protein were preincubated with Ilabeled porcine thyrotropin as described in Methods and analyzed on a 20-45 7"sucrose gradient in 10 mM Tris-C1 pH 7.4. (0)Absorbance at 280 nm: (A) radioactivity; (0)sucrose concentration "
' ?* \
I
I
B
'*
A\4A\
'\
A*
0
10
20
Time (min)
I
100
0.1
1
10
100
[Cation] (mM)
Fig. 2. Time course of '251-labeledporcine th-vrotropin binding to thyroid plasma membrunes and effect of temperuture. (0) 4,'C; (0) 10 T; (0)17 "C; (m) 27 "C; (A) 37 'C. Each assay contained 30 pg membrane protein in 120 p1 of standard binding buffer in the presence of 8 ng/ml labeled thyrotropin
Fig. 3. Inhibition oJ porcine '251-kuheled thyrotropin binding ( A ) monovalent and ( B J divalent cations. Membranes (40 pg protein) were incubated at 27 'C for 20 min in standard binding buffer containing increasing amounts of the following cations as chloride: in (A) (0) Na'; ( 0 ) Li'; (*) K t ; in (B) (A) Ca"; ( x ) M g 2 + . B = binding in presence of cation, BO = binding in its absence
stimulated and F--stimulated adenylate cyclase activities were 70,150 and 480 pmol min-' mg protein-', respectively. Plasma membrane homogeneity toward hormone binding was controlled by linear sucrose density gradient centrifugation. After incubation with '251-labeled thyrotropin, the thyrotropin-binding material sedimented as a sharp boundary at 37 % sucrose concentration when analyzed in a 20-45 "/, sucrose gradient (Fig. 1). The plasma membrane fraction thus obtained will be referred to further as membranes.
Thyrotropin Binding to Membranes
As previously shown for intact porcine thyroid cells and their derived plasma membranes [ 2 ] , the specific binding of porcine '251-labeled thyrotropin to porcine thyroid-gland-derived membranes was saturable, time-dependent and temperature-dependent. As shown in Fig.2 the rate of binding and the amount of hormone bound increased with increasing temperature up to 37 "(2, binding equilibrium was obtained after 10 min of incubation. Optimum pH
B. Verrier, R. Planells, and S. Lissitzky
247
Fig.4. Scatchard plot ofthe displacement of ( A ) bovine and ( B )porcine 1251-laheledthyrotropin by the corresponding unlabeled hormones; ( C ) theoretical Scatchardplot ofb/f versus b for K = k/2 und 3 k / 4 . is defined by Eqn (2), K and k are the dissociation constants of the unlabeled and labeled hormones, respectively; values of Kd = 8.42 ng/ml and S = 7.9 ng/mg protein were determined as in (A)
of binding was found between 6.0 and 6.5 (not illustrated). A striking effect of increasing concentrations of monovalent and divalent cations on thyrotropin binding was observed (Fig. 3). Using membranes in 50 mM Tris-C1 pH 7.4, 1 % bovine serum albumin and the cations as chloride, 50% inhibition of 12’1labeled thyrotropin binding was obtained at 5060 mM Na’, K + or Lit, and 10 mM Ca” or Mg2+ concentrations. In most studies of the interaction of hormone and receptor, inhibition of labeled hormone binding to receptor by increasing amounts of unlabeled hormone is analyzed. This analysis of the inhibition binding data relies on the assumption that both labeled and unlabeled hormones are bound with equal affinity. Recently, Taylor [6] showed that when this assumption is made incorrectly, non-linear Scatchard plots result. Taking into accou’nt the concentrations of labeled and unlabeled ligands, a modified Scatchard equation was derived : b l f = Slk - P/k
(1)
where b and f represent bound and free forms of the labeled hormone, S the total concentration of all forms of receptor and k the dissociation constant of the labeled hormone. The variable p is defined as:
P
=
[(Sf- k b ) ( f K + b k ) + f 2 b ( b - K)]/fiok (2)
K being the dissociation constant of the unlabeled hormone. When K = k , the formula simplifies to /3 = b (1
+ Iolio)
(3)
with I0 = B + F, B and F representing bound and free forms of the unlabeled hormone and io = b .f: The plot of blf’ versus gives a straight line.
+
The equal affinity of thyroid membrane receptors to bind labeled and unlabeled thyrotropin was verified in the case of the displacement of bovine and porcine 12’I-labeled thyrotropins by their respective unlabeled hormones (Fig.4A and 4B). For both hormones, the plot of b/f vs p was a straight line from which a Kd of 0.28 nM was derived for bovine thyrotropin and a capacity, S, of 0.26 pmol/mg protein. Fig. 4 C shows the theoretical plot of b/f vs P obtained in the case of a lower affinity of the receptor for the labeled hormone. With K = 3k/4, i.e. if the labeled hormone binds to the receptor with 75 of the affinity shown by the unlabeled hormone, a typical curvilinear plot is obtained. These results are thus compatible with the conclusions that (a) the membrane receptor binds native and 12’I-modified hormones with equal affinities, (b) porcine thyroid membranes contain a single type of high-affinityllow-capacity binding sites for thyrotropin. To gain insight on the possible specificity of the porcine receptor toward thyrotropins from different species, the displacement curves of porcine 125 Ilabeled thyrotropin by unlabeled porcine, bovine and human thyrotropins was studied. As shown in Fig. 5 , 50 % of the labeled hormone was displaced in this order by increasing concentrations of unlabeled hormones: bovine > porcine-2 2 porcine-1 > human. This is the order of decreasing biological activity as shown by the bioassay in vivo in the mouse (42, 9.5, 7.4 and 3.3 U/mg for bovine, porcine-2, porcine-1 and human thyrotropins, respectively). Similar results were obtained using bovine ‘2’I-labeled thyrotropin (Fig. 6). Fig. 5 also demonstrates that porcine luteinizing hormone and its p subunit at a concentration about
Porcine Thyroid Plasma Membranes and Thyrotropin
248 1 .0
'6
0.5
0
0.01
.o
0.1 1 Hormone (Lgirnl)
100
10
Fig. 5. Displacement of porcine 'Z51-iubbeledrhyroiropin by uniubeied thyroiropin f r o m d#&smi species und hj. gijoiprotrin siihuniis. (A) Bovine thyrotropin; (0) porcine thyrotropin lot 1; (0) porcine thyrotropin lot 2 ; (A) human thyrotropin: (m) porcine luteinizing hormone; ( x ) porcine thyrotropin p subunit; (0)porcine luteinizing hormone /Isubunit. 40 pg membrane protein were incubated at 27 ' C for 20 min in 120 p1 standard binding buffer with 8 ng/ml porcine '251-labeled thyrotropin and increasing amounts of unlabeled hormone or subunit dissolved in the same buffer. B = binding in the presence of the hormone, Bo = binding in its absence
" \A \ 0.0 0.1 \D
0
1
I 5 .
1 .0
10
Hormone ( L g l m i )
Fig. 6 . Displucernent ojbocinr'251-luheledth~rotro~~in b y uniaheiedbovine rhyrorropin ( ~ j , p ~ r c i thyrotropin ne i 0) andhurnarr rh,wotropitt ( A ] . Details as in Fig. 5
400-times higher than bovine thyrotropin, and porcine thyrotropin-P at a concentration 4000-times higher could not displace porcine 1251-labeled thyrotropin thus confirming previous results obtained with intact cells that the thyrotropin receptor is not able to recognize to a physiological significant extent luteinizing hormone and the b-subunits of luteinizing hormone and thyrotropin [17]. The preparations of porcine thyrotropin used in this work were chemically of very high purity as shown by analysis of their primary structure [18]. Their specific biological activity was 39 U/mg at the time of purification. However they lose activity with time of storage in the dry form and at the time when the experiments were performed two batches of the initial preparation disclosed activities of 9.5 (lot 1) and 7.4 U/mg (lot 2). Repurification by gel filtration did not modify their activity, nor their homogeneity in dodecylsulfate-polyacrylamide gel electrophoresis and their aptitude to react with anti-porcine-tliyrotropin antibodies.
As compared to the highly purified preparation of bovine thyrotropin with a biological activity of 42 U/mg, the decrease of activity of porcine thyrotropin might be related to a decrease in activity of all the molecules resulting in a loss of binding affinity to the receptor or to a loss of activity of a certain percentage of the molecules which would coexist in the preparation with native fully active molecules. The validity of the latter possibility was verified by measuring the relative affinities of the porcine membranes for both hormones. As shown in Fig. 7, the displacement of porcine 1251-labeledthyrotropin by unlabeled bovine thyrotropin (Fig. 7A) and of bovine '2sI-labeled thyrotropin by unlabeled porcine thyrotropin (Fig. 7 B) gave linear Scatchard plots in the representation of Taylor. In the case of the displacement of bovine 1251-labeledthyrotropin by unlabeled porcine thyrotropin and assuming an identical capacity of the membrane for binding both hormones, the theoretical plot of blf' versus P was calculated using different
B. Verrier, R . Planells, and S. Lissitzky I
I
A
I
0.1
249
0.3
I
0.
3.2
, . .
3
0.1
C
3
5
1
1
Fig. I. Plot of bjf versus p ,for the displacement of’ porcine ‘251laheled thjrotropin by unlabeled bovine thyrotropin ( A ) and of hovine ‘2s1-labeledthyrotropin by unlabeledporcinr tl7yrotropin ( B ) . (0) in (A) and (A)in (B) represent the b/fvalues of labeled hormone binding without added unlabeled hormone
Fig. 9. Scatchard plot qf b,/f versus3! , for displaremmt of poi-cinc I-labeled th,yrotropin by unlabeled humun tlijrotropin
125
In contrast, the plot of bifversus p for the displacement of porcine 1251-labeledthyrotropin by unlabeled human thyrotropin, is curvilinear with an upward concavity which indicates that the porcine receptor discloses a lower affinity for the human hormone (Fig. 9).
Adenylute Cycluse Activity of Membranes
0
10
20
P (nglml)
30
Fig. 8. Throretical Scatchard plot of b/f versus ,for values of’ K from 0 10 10 k, where K and k are the dissociation constants of porcine and bovine lhyrotropin, respectivelv See Fig.4C for details
values of K (the dissociation constant of porcine thyrotropin) as a function of k , the dissociation constant of bovine thyrotropin, from K = 2 k to K = 10k. As shown in Fig. 8 the value of K ( K = 8 k ) which is necessary to account for the experimental results (Fig. 7B), gives a non-linear plot. In addition if we recall the analysis made above (Fig.4C), the displacement of 1251-labeledporcine thyrotropin by unlabeled bovine thyrotropin should have given a curvature toward the left which is not observed (see Fig. 7A). It is thus concluded that the observed differences between the preparations of bovine and porcine thyrotropin are not related to differences in affinity but to the decrease of the amount of native molecules in the porcine preparations.
Membrane adenylate cyclase activity measurements were performed by incubating membranes and various amounts of thyrotropin at 30 ‘ C for 7 min prior to [U-~’P]ATPaddition; incubation was continued at the same temperature for various periods of time after which cyclic [32P]AMP was estimated. In these experiments, the enzyme of the ATP-regenerating system was crystalline pyruvate kinase in ammonium sulfate which was used without prior elimination of this salt. In these conditions, adenylate cyclase activity was linearly related to time up to 20 min and increased with increasing thyrotropin concentration (Fig. 10). A linear relation was also found between enzyme activity and membrane concentration up to 1.2 mg membrane protein/ml (not illustrated). However, replacement of pyruvate kinase in ammonium sulfate by a lyophilized preparation solubilized in 50 mM Tris-C1 pH 7.4 resulted in a decrease of stimulation by thyrotropin without modification of the fluoride-stimulated activity. In fact, the presence of ammonium sulfate in the medium of adenylate cyclase assay decreased the basal activity and increased the stimulation by thyrotropin (Table 2). For a concentration of 64 mM which corresponds to the final
hyroid Plasma Membranes and Thyrotropin
250
I
I 7
I
1
5 ll[ATP](mM-')
21
14 Time (min)
Fig. 10. Time course oj udenylute cyclasc acrivation of' porcine thyroid plasma membranes by increasing concentrations of bovinethyrotropin. (m) No thyrotropin; (0) 10 ng/ml; (0) 100 ng/ml; (0) 1 pg/ml; (A) 10 vg/ml
I
10
Fig. 11. Double reciprocal plot of' adenylate ryclase uctivit!: as a function of A T P concenlration. (W) N o thyrotropin; (A)100 ng/ml bovine thyrotropin.
Table 2. Basal and thyrotropin-stimulated adenylate cyclase activity of porcine thyroid plasma membranes us a function of ammonium sulfute concentrution -~
~~
(NH4)zS04
Adenylate cyclase activity with thyrotropin (100 ng ml-')
basal
mM
pmol min-
0 6.4 32 64 128 256
78 k 3 76 4 72 i 2 58 i 1 37 3 16 It: 2
0 Thyrotropin (Kgirni)
*
Fig. 12. Effect of' niedium on fhe binding of hoi'ine '251-I~beled
Table 3. Adenylatc cyclase uctivirj of'porcinc thyroid plasma membrancs versus thyrutropin concentration in rhe presence and absence sf 64 m M ammonium sulfute
fliyroiropin to thyroid plasma nzembrane.~.(A) Standard binding buffer; (A) standard binding buffer, 64 niM ammonium sulfate; ( 0 ) complete adenylate cyclase assay medium containing 64 mM ammonium sulfate. B = binding in presence of thyrotropin, Bo in its absence
-
~
Thyrolropin
Adenylate cyclase activity
pg ml-'
p o l min-' (mg protein)-'
0 0.00 I 0 01 01 10 10.0
33 _+ 1 31 1 33 i- 3 46 i: 2 61 k 4 69 k 1
_ _ _ _ _
~~
_ _ _ _
~~
_ _ _ _
41 1 44 3 44 2 50 3 52 & 2 53 _+ 2
**
concentration of ammonium sulfate in the assays, the stimulation by 100 ng/ml thyrotropin was I .65-fold as compared to 1.17-fold in its absence. Table 3 illustrates the stimulation of adenylate cyclase activity
by increasing amounts of thyrotropin in the presence or in the absence of 64 mM ammonium sulfate. In the complete system, a Michaelian relationship was found between velocity and ATP concentration, both in the absence and in the presence of thyrotropin (Fig.11). The hormone increased the value of V whereas an identical K , (520 pM) were observed in the presence or in the absence of thyrotropin. To specify the relationship between thyrotropin binding and adenylate cyclase activation, binding was reexamined in the medium conditions of adenylate cyclase assay containing 64 mM ammonium sulfate. Fig. 12 shows the displacement curves of bovine '''Ilabeled thyrotropin by unlabeled bovine thyrotropin in the standard binding medium (50 mM Tris-C1 pH 7.4) with and without ammonium sulfate and in
B. Verrier, R. Planells, and S. Lissitzky
25 1
Table 4. Equilibrium dissociation constant (Kd) and capacity ( S ) of bovine 251-iubeiedthyrotropin bincling to porcinr thyroid nzembrane.~ as a,function of the cmr~positionof ihe incubation medium In medium A 5 experiments were carried out and in medium C, 3 experiments; the range of values is shown in parentheses
Ki
S
nM
pmol (mg protein)-'
A. 50 m M Tris-CI, 0.1 bovine serum albumin
0.28 (0.20- 0.33)
0.29 (0.22 - 0.30)
B. Adenylate cyclase assay medium without (N&)ZS04
0.25
0.22
C. As (A) + 64 mM (NH&S04
3.77 (1.67- 2.00)
0.21 (0.19-0.23)
2.00
0.28
Medium
D. As (B) + 64 m M (NH4)2S04
T
I
I
1
I
I
3
0
u
00
0
A
I
0
I
0
I
I
0.001 0.01 0.1 1.0 Thyro ttopin (pgirnl)
I
I
10
100
Fig. 13. Adenylate cyclase activity of thyroid plasnza mrmhrancs as a function o/ rhyrotropin concentration. (A) Bovine; (0) porcine lot 2; (0)human
the complete adenylate cyclase assay medium. The Kd values are almost identical in media without ammonium sulfate and about 6-times lower than in the complete medium of adenylate cyclase assays (Table 4). In all conditions, the capacity of membranes for thyrotropin binding was almost the same. Adenylate cyclase stimulation by thyrotropin was measured with the bovine, porcine and human hormones. As shown in Fig. 13 a sigmoidal relationship between enzyme activity and hormone concentration was observed. Half-maximum stimulation was obtained for 80 ng/ml(2.7 nM) and 1040 ng/ml(34.7 nM) of the bovine and porcine hormones respectively. With a concentration of 100 vg/ml human thyrotropin, maximum stimulation was not obtained. Assuming the same maximum value for the human as for the bovine and porcine hormones, a half-maximum value of 20 pg/ml (0.7 pM) was computed for the human hormone.
DISCUSSION Plasma membranes purified from porcine thyroid glands contain a single type of high-affinity (& = 0.28 nM) binding site for porcine and bovine thyrotropins. As for plasma membranes purified from bovine thyroid glands [4,5],binding is optimum at pH 6.0-6.5 and is inhibited by monovalent (Na', K', Li') and divalent (Mg2+,Ca2+)cations. In contrast with bovine membranes [4,5] binding increases with temperature up to 37 "C. These properties are very similar to those previously described for intact cultured porcine thyroid cells and their derived plasma membranes [2,3]. In this paper, we have verified that the porcine membrane receptor for thyrotropin binds the native hormone and its '251-labeled derivative with identical affinity. An identical affinity to the receptor for bovine and porcine thyrotropin was demonstrated whereas the affinity for the human hormone was about 200times lower. Thyrotropin and the other pituitary glycoprotein hormones (luteinizing and follicle-stimulating hormones) are formed of two dissimilar non-covalently associated n and /3 subunits. In the three hormones the CI subunit primary structure appears to be identical, whereas the hormone-specific p subunits are different [19]. The sequences of the ci and p subunit of bovine and porcine thyrotropin have been elucidated [20,18]. The porcine thyrotropin n subunit is identical to the CI subunit of porcine luteinizing hormone [21] but is shorter by six amino acid residues at the amino terminus than the bovine thyrotropin ci subunit [18]. In addition three amino acid replacements occur in chain, the the chain. When compared to the bovine /l porcine p subunit differs by six amino acid replacements and by the absence of a methionyl residue at the carboxy terminus. These mutations in the gene are conservative in terms of chemical function in the peptide chain. If other factors such as the nature of the carbohydrate moiety are not involved in the activity of the hormones, the slight differences observed in the primary structure of the ci and p subunits of the bovine and porcine hormones do not affect the affinity of the porcine receptor for both hormones. As compared to bovine thyrotropin the human n subunit is one amino acid residue longer at its amino terminus and seems to differ by 22 amino acid replacements [22]. The differences between the p subunit of bovine and human thyrotropins concern 2 1 amino acid residue positions [23]. However, nine of these changes can be considered to be highly acceptable replacements. On the whole and as compared to the bovine hormone, the differences in the primary structure of the two chains of the human hormone are larger than
252
B. Verrier, R. Planells, and S. Lissitzky: Porcine Thyroid Plasma Membranes and Thyrotropin
that showed by porcine thyrotropin. How these differences can explain the lower affinity of the human hormone for the porcine receptor remain to be determined. Stimulation of adenylate cyclase activity of the porcine thyroid plasma membranes by thyrotropin is a saturatable phenomenon. Correlation between occupancy of receptor sites by the hormone and stimulation of adenylate cyclase was demonstrated. At equilibrium and in the conditions of adenylate cyclase assay, the Kdof binding for the bovine hormone (2.0 nM) was close to the concentration of hormone which provoked half-stimulation of the enzyme (2.7 nM). The small difference observed lies within the limits of experimental errors. As for thyrotropin binding, the associated stimulation of adenylate cyclase correlates well with the biological activity of the bovine, porcine and human hormones in the bioassay in vivo in the mouse. These results are comparable to those obtained with porcine plasma membranes derived from intact cells [2] and in bovine membranes derived from glands, to the concentration of hormone giving half-stimulation of adenylate cyclase [24]. Basal and thyrotropin-stimulated adenylate cyclase activities of the porcine membranes are comparable to those observed for bovine [25- 271 or human [28] membranes. However, the factor of stimulation of the enzyme by thyrotropin is 2 - 2.5-fold and lower than that found by other investigators for bovine membranes. This low value seems more related to an increase of the basal activity of the porcine membranes than to a decrease of the thyrotropin-stimulated activity. In addition, the stimulation factor was constant for all the values of the basal adenylate cyclase activity (40- 70 pmol min-' mg protein-'). In the presence of a concentration of Mg2+ ions (4 mM) in excess of that necessary to complex all the ATP present in the assay, the K,,, of adenylate cyclase for ATP is not modified by the binding of thyrotropin to the receptor. This agrees with similar results obtained with the systems epinephrineisarcolemma [29] and glucagon/liver membranes [30]. The K , value found (520pM) is comparable to that found for the adenylate cyclase in other systems [29,31]. We are greatly indebted to D r G. Hennen (LiBge, Belgium) and t o D r J. G . Pierce (Los Angeles, Calif. U.S.A.) for the kind supply of thyrotropin. The technical assistance of J.-C. Bugeia and J. Garcia is gratefully acknowledged. We are also indebted to Mr B. Rousset (Service rl' Entlocrinologir, Hopiral de I'Antiyuuille, Lyon) who tested the biological activities of the thyrotropin preparations used
in our investigations. This work was supported in part by the Centre National de lu Recherche Scientifique (contract 1919 ATP Micanisme d'action des hormones).
REFERENCES 1. Lissitzky, S., Fayet, G. & Verrier, B. (1976) Methods in Receptor Research (Blecher, M., ed.) vol. 2. pp. 641-665, Marcel Dekker, New York. 2. Verrier, B., Fayet, G . & Lissitzky, S. (1974) Eur. J . Bioci7on. 42,355 - 365. 3. Lissitzky, S., Fayet, G . & Verrier, B. (1975) Adi,. C:vclic Nucleotide Res. 5 , 133-152. 4. Amir, S. M., Carraway, T. F.. Kohn, L. D. & Winand, R. J . (1973) J . Biol. C'hem. 248, 4092-4100. 5. Moore, W. V . & WolH, J.(1973)J. Biul. Clii.m.248,5705-5711. 6. Taylor, S. I. (1975) Biochen?istry, 14, 2357-2361. 7. McKenzie, J. M. (1958) Endocrinology, 63, 372-378. 8. Jaquet, P., Hennen, G. & Lissitzky, S. (1974) Biochiinie (Parly) 56,169 - 774. 9. Planells, R., Fayet, G., Lissitzky. S., Hennen, G. & Closset, J . (1975) FEBS Lett. 53, 87-92 10. Solomon, Y., Londos, C. & Rodbell, M. (1974) Anal. Biocl7rrn. 58,541 - 548. 11. Delaage, M. A., Bellon, N. & Cailla, H . L. (1974) Anal. Biochem. 62,417-425. 12. Randerath, K . & Randerath. E. (1967) Mi,rizoifs E ~ i : y / n d , 12, 323 - 347. 13. Cache, C., Rossi, B. & Lazdunski, M . (1976) Eur. J . B i u d 7 ~ ~ 7 . 65,293 - 306. 14. Aronson, N. N . Jr & Touster, 0. (1974) Methods Enzj.mol. 31. 90- 102. 15. Weaver, R. A. & Boyle, W. (1969) Biochim. Biophj.s. Acta. 173. 377-388. 16. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall. R. J. (1951) J . B i d . Chem. 193, 265-275. 17. Lissitzky, S.. Fayet, G., Verrier, B., Closset, J. B Hennen, G. (1974) FEBS Lett. 48, 275-278. 18. Maghuin-Rogister, G., Hennen, G., Closset. J . & Kopeyitn. C. (1976) Eur. J . Biochem. 61, 157- 163. 19. Pierce, J. G. (1971) Endocrinolug~.,8Y, 1331- 1344. 20. Liao, T. H. &Pierce, J . G. (1971)J. Biol. Chcw;. 246, 850-865. 21. Maghuin-Rogister, G. & Hennen, G . (1973) Eur. J . Bioc~lron. 39,255- 263. 22. Sairain, M. R . & Li, C. H. (1973) Biorkeni. Bioplij~.~.Rrs. Commun. 51, 336-342. 23. Sairam, M. R. & Li, C. H. (1973) Biochem. Bioplij's. Res. C'ommun. 54.426-431. 24. Yamashita. K. & Field, J . B. (1972) J . C'lin. Inwst. 51, 463472. 25. Wolff, J. & Cook, G. H. (1973) J. Biol. Chenz. 248, 350-355. 26. Mashiter, K., Mashiter, G. D. & Field, J. B. (1954) Endxrino/O,~V, 94, 370- 376. 27. Marshall, N . J., Von Borcke, S. & Malan, P. G . (1975) Endocrinologj', 96, 1513- 1519. 28. Sato, S., Yamada, T., Furihata, R. & Makinchi, M . (1974) Biochim. Biop1zJ.s.Actu, 332, 166- 174. 29. Severson, D. L., Drummond, G. I. & Sulakhe, P. V. (1972) J. Biol. Chem. 247, 2949- 2958. 30 Lin, M. C., Salomon, Y., Rendell, M. & Rodbell, M . (1975) J . Biol. Chcm. 25U, 4246- 4252. 31 Ho, R. J . B Sutherland, E. W . (1975) Proc. Nail Acud. &i. U . S . A . 72, 3773--1777.
B. Verrier, R. Planells, and S. Lissitzky, Laboratoire de Biochimie MCdicale, Faculte de Medecine, Universiti d'Aix-Marseille, 27 Boulevard Jean-Moulin, F-I 3385 Marseille-Cedex-4, France
LABORATOIRE ez
mcd :o.'ogio SAZT TllMAN LIEGE 1
/r?>lilut c's i'r
4000
Thyrotropin Binding to and Adenylate Cyclase Activity of Porcine Thyroid Plasma Membranes Bernard VERRIER, Richard PLANELLS, and Serge LISSITZKY Laboratoire de Biochimie Medicale et Unite 38 de I'Institut National de la SantC et de la Recherche Medicale, Faculte de Medecine, Marseille (Received November 24, 1976)
Plasma membranes have been purified from porcine thyroid gland homogenate by discontinuous sucrose gradient centrifugation. The preparations contained specific binding sites for thyrotropin but not for luteinizing hormone or the p subunits of thyrotropin and luteinizing hormone. Optimum conditions of '2sI-labeled thyrotropin binding were pH 6.0-6.5 and 37 "C. Thyrotropin binding was reduced by divalent (Ca", Mg Zf)and monovalent cations (Na', K', Li'), 50% inhibition being obtained at 10 mM and 50 mM respectively. Displacement curves of '2sI-labeled bovine or porcine thyrotropin by the unlabeled hormone from three species was in the order of increasing concentrations (bovine > porcine > human) which is the order of decreasing biological activity of these hormone preparations in the assay in vivo in the mouse. The validity of the results was established by controlling that porcine membranes bound the native and the '251-labeled hormones with equal affinity. A single type of high-affinity (& = 0.28 nM) binding sites was detected for bovine and porcine thyrotropins. In contrast, porcine plasma membranes bound human thyrotropin with a lower affinity (& = 70 nM). A good correlation was found at equilibrium and in the conditions o f the cyclase assay, between receptor occupancy and adenylate cyclase activation for the three hormones. The consensus of investigators agrees that the first event in the action of thyrotropin on target tissues is the binding to receptors in the plasma membrane and the consecutive stimulation of adenylate cyclase (see [l] for a review). In previous studies [2,3], we have demonstrated in both i n t h porcine thyroid cells and their derived plasma membranes the presence of a single class of specific sites that bind '251-labeled porcine thyrotropin with high affinity. In addition a quantitative correlation between binding and adenylate cyclase activation was established. Other investigators [4,5]using plasma membranes from bovine glands suggested the possible existence of two types of binding sites as shown by non-linear Scatchard plots of saturation binding experiments and were unable to correlate binding with adenylate cyclase activation. The interpretation of non-linear Scatchard plots is not clear and many models have been proposed to account for non-linearity such as cooperative binding interactions, polymerization of hormone, heteroAbbreviation. Cyclic AMP, adenosine 3' : 5'-monophosphate. Enzymes. Adenylate cyclase (EC 4.6.1.1) ; p-nitrophenylphosphatase (EC 3.1.3.41); glucose-6-phosphatase (EC 3.1.3.9); SUCcinate dehydrogenase (EC 1.3.99.1).
geneity of receptors or of labeled hormones (see [6] for references) and recently difference of affinity of the receptor for the labeled and unlabeled hormones [61. Since differences in binding characteristics may be related to the nature of the starting material (gland or intact cultured cells) used to prepare plasma membranes, we have investigated the specific binding of thyrotropin to plasma membranes purified from porcine thyroid glands. In this paper, we describe the preparation of these membranes, the analysis of the binding of thyrotropins from different species and the correlation between binding and adenylate cyclase activation for these hormones. MATERIALS AND METHODS Purfication o j Plasma Membranes
Porcine thyroid glands were collected at the local slaughter house, immediately immersed in chilled Hank's saline buffer and processed as soon as possible. After careful dissection to remove fat and connective tissue, slices were obtained using a Stadie-Riggs slicer and homogenized in 0.25 M sucrose, 1 mM MgC12,
Porcine Thyroid Plasma Membranes and Thyrotropin
244
10 mM Tris-CI pH 7.4 (1 g slices in 2 ml buffer), first with an Ultra-Turrax at minimum speed for a few seconds and then, by hand, in a glass teflon Potter homogenizer (two strokes). The homogenate obtained from 100 g glands was diluted to obtain a final volume of 800 ml and centrifuged at 50 x g for 10 min. The supernatant was filtered on a stainless steel sieve. The residual tissue fragments remaining on the sieve and the 50 x g pellets were rehomogenized in 200 ml buffer and the homogenate was centrifuged in the same conditions. Both supernatants were then spun at 500 x g for 10 min, the pellets discarded and the supernatant centrifuged at 10000 x g for 10 min. The 10000 x g pellets were gently homogenized in 1 mM MgC12, 10 mM Tris-CI pH 7.4, using a glass/teflon homogenizer and were enriched with sucrose in buffer to obtain a final concentration of 43.3% (w/w) for a total volume of 90 ml. 15 ml of the sucrose-enriched suspension were layered in 38.5-nil tubes over 8 ml 45 (w/w) sucrose solution and overlaid by sucrose solutions of 41% (10 ml) and 8 % ( 5 ml). The six discontinuous sucrose gradients thus obtained were spun in a Spinco model L2-50 centrifuge in rotor SW27 at 24000 rev./min for 2 h at 1 "C. Plasma membranes were harvested by aspiration with a pasteur pipette at the interface 41 %,/8%, diluted (1 : 6) with 10 mM Tris-C1 pH 7.4 and centrifuged at 10000 x g for 10 min. The pellet was resuspended in 50 mM Tris-CI pH 7.4 to obtain a final concentration of 1.2- 2 mg protein/ml. Aliquots of 0.2-0.4 ml were stored at -196 "C. Hormone Preparations Highly purified porcine and human thyrotropins were obtained from Dr G. Hennen (Liege, Belgium). The thyrotropic activity of porcine thyrotropin was initially 39 U/mg as determined by the McKenzie bioassay in the mouse [7]; after storage for two years in the dry form the activity decreased with time to reach values below 10 U/mg (lot 2 = 9.5 U/mg, lot 1 = 7.4 Ujmg) when the experiments related in this paper were performed. The specific bioactivity of the human hormone was 3.3 U/mg. The high purity of this preparation was checked by amino acid composition closely resembling that of the bovine and porcine hormones, N-terminal amino acid analysis and polyacrylamide gel electrophoresis (G. Hennen and J. Closset, personal communication). Bovine thyrotropin (42 U/mg) was the generous gift of Dr J. G. Pierce (Los Angeles, Calif., U.S.A.). Labeling of thyrotropin with "'I was carried out as previously described [8] using the lactoperoxidase method of iodide oxidation except that the iodination mixture was made in 0.5 M aceto-acetate buffer pH 5.6, and separation of the labeled hormone was performed in a single step on a Sephadex G-100 column (1 x 70 cm) equilibrated in
50 mM Tris-C1 pH 7 containing 0.1 % bovine serum albumin [l]. Specific radioactivities of 60 - 90 Ci/g were obtained. As compared to the native hormone, no loss of biological activity due to iodination was observed using the bioassay in vitro previously described [9]. I-Labeled Tlzyrotropin to Plasma Membranes
Binding
Measurement of labeled thyrotropin binding to plasma membranes was performed in 50 mM Tris-C1 pH 7.4containing 1 %bovine serum albumin (standard binding buffer), 0.4- 0.8 mg/ml of membrane protein, and labeled thyrotropin in a final volume of 0.0500.2 ml. Incubations were routinely performed at 27 'C and separation of membrane-bound from free thyrotropin was carried out by centrifugation as previously described [2]. Radioactivity of the pellets was estimated in a Packard Autogamma spectrometer with a counting efficiency of 25 Specific binding was obtained by substracting from the total radioactivity bound the amount which was not displaced by an excess of native thyrotropin. All assays including control were performed in duplicate or triplicate. Membranes which have bound porcine '"I-labeled thyrotropin were analyzed by gradient centrifugation as described below. Membranes were incubated at 27 "C in 50 mM Tris-C1 pH 7.4, 1 bovine serum albumin for 25 min, and then washed from free 1251-labeledthyrotropin by centrifugation at 10000 x g for 10 min. The pellet was suspended in 50 mM Tris-C1 pH 7.4 containing 1 '%; bovine serum albumin and 200 pg of native membrane proteins as carrier and was loaded on a 20-45% (w/w) sucrose gradient in 10 mM Tris-C1 pH 7.4. The gradient was spun at 24000 rev.jmin in a 17-ml SW27 bucket (Beckman) at 1 L'C.Gradients were fractionated and absorbance at 280 nm and "'I radioactivity measured using standard conditions.
'x.
Assay of Adenylate Cjduse Activity Assays were performed in a final volume of 50 p1 in 50 mM Tris-C1 pH 7.5, containing 0.1 04 bovine serum albumin, 4 mM MgC12, 4 mM phosphoenolpyruvate, 400 mU pyruvate kinase, 1 mM cyclic AMP and the indicated concentrations of [a-"PIATP (10- 50 counts min-' pmol-I). Membrane protein at a concentration of 500 - 760 pg/ml and thyrotropin at different concentrations were added. When used, NaF concentration was 10 mM. Total incubation medium without ATP was preincubated for 7 min at 30 "C. In the case of kinetic experiments 50-pl aliquots were taken off at given intervals from a conveniently sized pool. The reaction was started by addition of substrate after preincubation for 7 min at 30 "C. At various times of incubation, the reaction was stopped
B. Verrier, R. Planells, and S. Lissitzky
by the addition of 150 p1 50 mM Tris-C1 pH 7.5 containing 1.33% sodium dodecylsulfate, 40 mM ATP and 1.4 mM cyclic AMP. 50 pl of cyclic [3H]AMP (2- 3 x lo4 counts/min per assay) and 750 pl of water were added. Cyclic AMP was determined as described by Saloinon et al. [lo] with slight modifications. Columns of Dowex AG-50-W x 4 containing 4.5 ml of resin were washed with 5 ml 1 M NaOH, 10 ml bidistilled water, 5 ml 3 M HC1 and then 10 ml of bidistilled water. After use, the column was regenerated by an identical cycle. Between assays resin columns were kept in water. For further purification of cyclic AMP, alumina columns were utilized as indicated by the authors. In all the assays performed (more than 500) cyclic AMP recovery was 60 - 80 %. Assay blanks containing no membrane were processed in the same way. Radioactivity in the blanks were less than 0.0005% of the total added. Assays were performed in triplicate or quadruplicate. We have verified by an electrophoretic separation [ l l ] that in our conditions, no more than 10 % cyclic AMP was hydrolysed during the course of 20-min incubations. Degradation of ATP which could occur in our conditions of incubation was determined by thin-layer chromatography on poly(ethy1eneimine)-cellulose in 1.2 M LiCl [12]. N o more than 20% of the nucleotide was destroyed, even at 0.1 mM ATP. To verify that the radioactive material isolated by double-column chromatography was effectively cyclic AMP we processed as described. At the end of the 20-min incubation period, membranes were removed by centrifugation at 10000 x g for 10 min and the supernatants were incubated for 10 min at 30 'C with 50mU beef heart phosphodiesterase. These samples did not show radioactivity in the cyclic AMP fraction while control assays processed in the same way retained radioactivity. Orlzer. Assays
Ouabain-sensitive K+-stimulated p-nitrophenylphosphatase activity was determined at 25 "C and 410 nm in 50 mM Tris-C1 buffer at pH 7.5 containing 20inM KCI, 20mM MgClz and 20mM p-nitrophenyl phosphate in the presence or in the absence of 0.2 mM ouabain [13]. Glucose-6-phosphatase and antimycin-sensitive succinate dehydrogenase activities were assayed using standard methods [14,15]. Membrane protein was determined according to Lowry et al. [16] using serum albumin as standard, after solubilization in 1 M NaOH. A z x I / A ~ ratios ~o were estimated in 2 '%,sodium dodecylsulfate. Materials [a-32P]ATP(0.5- 1 Ci/mol) and cyclic [3H]AMP (30 Ci/mmol) were purchased from the Radiochemical Center (Amersham, England). ATP, cyclic AMP, p-
245
nitrophen ylphosphate and crystallized pyruvate kinase were obtained from Boehringer, Mannheim (West Germany). Phosphoenolpyruvate and lyophilized pyruvate kinase were obtained from Sigma (St Louis, U.S.A.). All other chemicals were of the highest purity available from commercial sources. Other.Details
All curves were traced using linear regression calculation. Typical experiments are presented. A molecular weight of 30000 for thyrotropin was used for all calculations. RESULTS
Plasma Membrane Preparations The plasma membrane fraction was obtained from porcine thyroid gland homogenates after discontinuous sucrose density gradient centrifugation of the 10000 x g pellet at the interface 8%/41% (d = 1.18 g/ ml). In 15 preparations the yield was about 0.2 mg membrane protein per g fresh slices. Marker enzyme data are summarized in Table 1. As compared to the particulate fraction obtained by centrifugation at 105000 x g of the 50 x g supernatant fraction, the purified membranes showed a decrease in glucose-6phosphatase and succinate dehydrogenase activities of about 2-fold and an increase in K 'pnitrophenylphosphatase activity (a plasma membrane enzyme marker) of about 5-fold. As compared to the 50 x g fraction, an 18-fold increase in the latter enzyme activity was observed. Although less purified than the membranes previously obtained from isolated porcine thyroid cells [2], these membrane preparations contained an active thyrotropin-stimulated and F--stimulated adenylate cyclase and showed a good capacity to bind labeled thyrotropin (see below). Basal, thyrotropinTable 1. Comparison of enzymic activities of total particles and purified plu,smamembranes of porcine thyroid gland homogenate.v Each number is the mean of duplicate determinations in three different preparations; the individual values agreed within 5- 10 Fraction
Ouabainsensitive p-nitrophenyl phosphatase
Glucose-6phosphatase
Antimycinsensitive succinate dehydrogenase
A28~iA~h~
nmol rn1n-l (mg protein)-' _ ~ _ _ _ ~ - - 105000 x g pellet 33 Plasma membranes 15 1
29 0
0 65
0 49
14.0
0 30
0 98
Porcine Thyroid Plasma Membranes and Thyrotropin
246
Fraction number
Fig. 1. Sucrose density gradient centrifugation ofporcine thyroid plusma niembranes. 200 pg membrane protein were preincubated with Ilabeled porcine thyrotropin as described in Methods and analyzed on a 20-45 7"sucrose gradient in 10 mM Tris-C1 pH 7.4. (0)Absorbance at 280 nm: (A) radioactivity; (0)sucrose concentration "
' ?* \
I
I
B
'*
A\4A\
'\
A*
0
10
20
Time (min)
I
100
0.1
1
10
100
[Cation] (mM)
Fig. 2. Time course of '251-labeledporcine th-vrotropin binding to thyroid plasma membrunes and effect of temperuture. (0) 4,'C; (0) 10 T; (0)17 "C; (m) 27 "C; (A) 37 'C. Each assay contained 30 pg membrane protein in 120 p1 of standard binding buffer in the presence of 8 ng/ml labeled thyrotropin
Fig. 3. Inhibition oJ porcine '251-kuheled thyrotropin binding ( A ) monovalent and ( B J divalent cations. Membranes (40 pg protein) were incubated at 27 'C for 20 min in standard binding buffer containing increasing amounts of the following cations as chloride: in (A) (0) Na'; ( 0 ) Li'; (*) K t ; in (B) (A) Ca"; ( x ) M g 2 + . B = binding in presence of cation, BO = binding in its absence
stimulated and F--stimulated adenylate cyclase activities were 70,150 and 480 pmol min-' mg protein-', respectively. Plasma membrane homogeneity toward hormone binding was controlled by linear sucrose density gradient centrifugation. After incubation with '251-labeled thyrotropin, the thyrotropin-binding material sedimented as a sharp boundary at 37 % sucrose concentration when analyzed in a 20-45 "/, sucrose gradient (Fig. 1). The plasma membrane fraction thus obtained will be referred to further as membranes.
Thyrotropin Binding to Membranes
As previously shown for intact porcine thyroid cells and their derived plasma membranes [ 2 ] , the specific binding of porcine '251-labeled thyrotropin to porcine thyroid-gland-derived membranes was saturable, time-dependent and temperature-dependent. As shown in Fig.2 the rate of binding and the amount of hormone bound increased with increasing temperature up to 37 "(2, binding equilibrium was obtained after 10 min of incubation. Optimum pH
B. Verrier, R. Planells, and S. Lissitzky
247
Fig.4. Scatchard plot ofthe displacement of ( A ) bovine and ( B )porcine 1251-laheledthyrotropin by the corresponding unlabeled hormones; ( C ) theoretical Scatchardplot ofb/f versus b for K = k/2 und 3 k / 4 . is defined by Eqn (2), K and k are the dissociation constants of the unlabeled and labeled hormones, respectively; values of Kd = 8.42 ng/ml and S = 7.9 ng/mg protein were determined as in (A)
of binding was found between 6.0 and 6.5 (not illustrated). A striking effect of increasing concentrations of monovalent and divalent cations on thyrotropin binding was observed (Fig. 3). Using membranes in 50 mM Tris-C1 pH 7.4, 1 % bovine serum albumin and the cations as chloride, 50% inhibition of 12’1labeled thyrotropin binding was obtained at 5060 mM Na’, K + or Lit, and 10 mM Ca” or Mg2+ concentrations. In most studies of the interaction of hormone and receptor, inhibition of labeled hormone binding to receptor by increasing amounts of unlabeled hormone is analyzed. This analysis of the inhibition binding data relies on the assumption that both labeled and unlabeled hormones are bound with equal affinity. Recently, Taylor [6] showed that when this assumption is made incorrectly, non-linear Scatchard plots result. Taking into accou’nt the concentrations of labeled and unlabeled ligands, a modified Scatchard equation was derived : b l f = Slk - P/k
(1)
where b and f represent bound and free forms of the labeled hormone, S the total concentration of all forms of receptor and k the dissociation constant of the labeled hormone. The variable p is defined as:
P
=
[(Sf- k b ) ( f K + b k ) + f 2 b ( b - K)]/fiok (2)
K being the dissociation constant of the unlabeled hormone. When K = k , the formula simplifies to /3 = b (1
+ Iolio)
(3)
with I0 = B + F, B and F representing bound and free forms of the unlabeled hormone and io = b .f: The plot of blf’ versus gives a straight line.
+
The equal affinity of thyroid membrane receptors to bind labeled and unlabeled thyrotropin was verified in the case of the displacement of bovine and porcine 12’I-labeled thyrotropins by their respective unlabeled hormones (Fig.4A and 4B). For both hormones, the plot of b/f vs p was a straight line from which a Kd of 0.28 nM was derived for bovine thyrotropin and a capacity, S, of 0.26 pmol/mg protein. Fig. 4 C shows the theoretical plot of b/f vs P obtained in the case of a lower affinity of the receptor for the labeled hormone. With K = 3k/4, i.e. if the labeled hormone binds to the receptor with 75 of the affinity shown by the unlabeled hormone, a typical curvilinear plot is obtained. These results are thus compatible with the conclusions that (a) the membrane receptor binds native and 12’I-modified hormones with equal affinities, (b) porcine thyroid membranes contain a single type of high-affinityllow-capacity binding sites for thyrotropin. To gain insight on the possible specificity of the porcine receptor toward thyrotropins from different species, the displacement curves of porcine 125 Ilabeled thyrotropin by unlabeled porcine, bovine and human thyrotropins was studied. As shown in Fig. 5 , 50 % of the labeled hormone was displaced in this order by increasing concentrations of unlabeled hormones: bovine > porcine-2 2 porcine-1 > human. This is the order of decreasing biological activity as shown by the bioassay in vivo in the mouse (42, 9.5, 7.4 and 3.3 U/mg for bovine, porcine-2, porcine-1 and human thyrotropins, respectively). Similar results were obtained using bovine ‘2’I-labeled thyrotropin (Fig. 6). Fig. 5 also demonstrates that porcine luteinizing hormone and its p subunit at a concentration about
Porcine Thyroid Plasma Membranes and Thyrotropin
248 1 .0
'6
0.5
0
0.01
.o
0.1 1 Hormone (Lgirnl)
100
10
Fig. 5. Displacement of porcine 'Z51-iubbeledrhyroiropin by uniubeied thyroiropin f r o m d#&smi species und hj. gijoiprotrin siihuniis. (A) Bovine thyrotropin; (0) porcine thyrotropin lot 1; (0) porcine thyrotropin lot 2 ; (A) human thyrotropin: (m) porcine luteinizing hormone; ( x ) porcine thyrotropin p subunit; (0)porcine luteinizing hormone /Isubunit. 40 pg membrane protein were incubated at 27 ' C for 20 min in 120 p1 standard binding buffer with 8 ng/ml porcine '251-labeled thyrotropin and increasing amounts of unlabeled hormone or subunit dissolved in the same buffer. B = binding in the presence of the hormone, Bo = binding in its absence
" \A \ 0.0 0.1 \D
0
1
I 5 .
1 .0
10
Hormone ( L g l m i )
Fig. 6 . Displucernent ojbocinr'251-luheledth~rotro~~in b y uniaheiedbovine rhyrorropin ( ~ j , p ~ r c i thyrotropin ne i 0) andhurnarr rh,wotropitt ( A ] . Details as in Fig. 5
400-times higher than bovine thyrotropin, and porcine thyrotropin-P at a concentration 4000-times higher could not displace porcine 1251-labeled thyrotropin thus confirming previous results obtained with intact cells that the thyrotropin receptor is not able to recognize to a physiological significant extent luteinizing hormone and the b-subunits of luteinizing hormone and thyrotropin [17]. The preparations of porcine thyrotropin used in this work were chemically of very high purity as shown by analysis of their primary structure [18]. Their specific biological activity was 39 U/mg at the time of purification. However they lose activity with time of storage in the dry form and at the time when the experiments were performed two batches of the initial preparation disclosed activities of 9.5 (lot 1) and 7.4 U/mg (lot 2). Repurification by gel filtration did not modify their activity, nor their homogeneity in dodecylsulfate-polyacrylamide gel electrophoresis and their aptitude to react with anti-porcine-tliyrotropin antibodies.
As compared to the highly purified preparation of bovine thyrotropin with a biological activity of 42 U/mg, the decrease of activity of porcine thyrotropin might be related to a decrease in activity of all the molecules resulting in a loss of binding affinity to the receptor or to a loss of activity of a certain percentage of the molecules which would coexist in the preparation with native fully active molecules. The validity of the latter possibility was verified by measuring the relative affinities of the porcine membranes for both hormones. As shown in Fig. 7, the displacement of porcine 1251-labeledthyrotropin by unlabeled bovine thyrotropin (Fig. 7A) and of bovine '2sI-labeled thyrotropin by unlabeled porcine thyrotropin (Fig. 7 B) gave linear Scatchard plots in the representation of Taylor. In the case of the displacement of bovine 1251-labeledthyrotropin by unlabeled porcine thyrotropin and assuming an identical capacity of the membrane for binding both hormones, the theoretical plot of blf' versus P was calculated using different
B. Verrier, R . Planells, and S. Lissitzky I
I
A
I
0.1
249
0.3
I
0.
3.2
, . .
3
0.1
C
3
5
1
1
Fig. I. Plot of bjf versus p ,for the displacement of’ porcine ‘251laheled thjrotropin by unlabeled bovine thyrotropin ( A ) and of hovine ‘2s1-labeledthyrotropin by unlabeledporcinr tl7yrotropin ( B ) . (0) in (A) and (A)in (B) represent the b/fvalues of labeled hormone binding without added unlabeled hormone
Fig. 9. Scatchard plot qf b,/f versus3! , for displaremmt of poi-cinc I-labeled th,yrotropin by unlabeled humun tlijrotropin
125
In contrast, the plot of bifversus p for the displacement of porcine 1251-labeledthyrotropin by unlabeled human thyrotropin, is curvilinear with an upward concavity which indicates that the porcine receptor discloses a lower affinity for the human hormone (Fig. 9).
Adenylute Cycluse Activity of Membranes
0
10
20
P (nglml)
30
Fig. 8. Throretical Scatchard plot of b/f versus ,for values of’ K from 0 10 10 k, where K and k are the dissociation constants of porcine and bovine lhyrotropin, respectivelv See Fig.4C for details
values of K (the dissociation constant of porcine thyrotropin) as a function of k , the dissociation constant of bovine thyrotropin, from K = 2 k to K = 10k. As shown in Fig. 8 the value of K ( K = 8 k ) which is necessary to account for the experimental results (Fig. 7B), gives a non-linear plot. In addition if we recall the analysis made above (Fig.4C), the displacement of 1251-labeledporcine thyrotropin by unlabeled bovine thyrotropin should have given a curvature toward the left which is not observed (see Fig. 7A). It is thus concluded that the observed differences between the preparations of bovine and porcine thyrotropin are not related to differences in affinity but to the decrease of the amount of native molecules in the porcine preparations.
Membrane adenylate cyclase activity measurements were performed by incubating membranes and various amounts of thyrotropin at 30 ‘ C for 7 min prior to [U-~’P]ATPaddition; incubation was continued at the same temperature for various periods of time after which cyclic [32P]AMP was estimated. In these experiments, the enzyme of the ATP-regenerating system was crystalline pyruvate kinase in ammonium sulfate which was used without prior elimination of this salt. In these conditions, adenylate cyclase activity was linearly related to time up to 20 min and increased with increasing thyrotropin concentration (Fig. 10). A linear relation was also found between enzyme activity and membrane concentration up to 1.2 mg membrane protein/ml (not illustrated). However, replacement of pyruvate kinase in ammonium sulfate by a lyophilized preparation solubilized in 50 mM Tris-C1 pH 7.4 resulted in a decrease of stimulation by thyrotropin without modification of the fluoride-stimulated activity. In fact, the presence of ammonium sulfate in the medium of adenylate cyclase assay decreased the basal activity and increased the stimulation by thyrotropin (Table 2). For a concentration of 64 mM which corresponds to the final
hyroid Plasma Membranes and Thyrotropin
250
I
I 7
I
1
5 ll[ATP](mM-')
21
14 Time (min)
Fig. 10. Time course oj udenylute cyclasc acrivation of' porcine thyroid plasma membranes by increasing concentrations of bovinethyrotropin. (m) No thyrotropin; (0) 10 ng/ml; (0) 100 ng/ml; (0) 1 pg/ml; (A) 10 vg/ml
I
10
Fig. 11. Double reciprocal plot of' adenylate ryclase uctivit!: as a function of A T P concenlration. (W) N o thyrotropin; (A)100 ng/ml bovine thyrotropin.
Table 2. Basal and thyrotropin-stimulated adenylate cyclase activity of porcine thyroid plasma membranes us a function of ammonium sulfute concentrution -~
~~
(NH4)zS04
Adenylate cyclase activity with thyrotropin (100 ng ml-')
basal
mM
pmol min-
0 6.4 32 64 128 256
78 k 3 76 4 72 i 2 58 i 1 37 3 16 It: 2
0 Thyrotropin (Kgirni)
*
Fig. 12. Effect of' niedium on fhe binding of hoi'ine '251-I~beled
Table 3. Adenylatc cyclase uctivirj of'porcinc thyroid plasma membrancs versus thyrutropin concentration in rhe presence and absence sf 64 m M ammonium sulfute
fliyroiropin to thyroid plasma nzembrane.~.(A) Standard binding buffer; (A) standard binding buffer, 64 niM ammonium sulfate; ( 0 ) complete adenylate cyclase assay medium containing 64 mM ammonium sulfate. B = binding in presence of thyrotropin, Bo in its absence
-
~
Thyrolropin
Adenylate cyclase activity
pg ml-'
p o l min-' (mg protein)-'
0 0.00 I 0 01 01 10 10.0
33 _+ 1 31 1 33 i- 3 46 i: 2 61 k 4 69 k 1
_ _ _ _ _
~~
_ _ _ _
~~
_ _ _ _
41 1 44 3 44 2 50 3 52 & 2 53 _+ 2
**
concentration of ammonium sulfate in the assays, the stimulation by 100 ng/ml thyrotropin was I .65-fold as compared to 1.17-fold in its absence. Table 3 illustrates the stimulation of adenylate cyclase activity
by increasing amounts of thyrotropin in the presence or in the absence of 64 mM ammonium sulfate. In the complete system, a Michaelian relationship was found between velocity and ATP concentration, both in the absence and in the presence of thyrotropin (Fig.11). The hormone increased the value of V whereas an identical K , (520 pM) were observed in the presence or in the absence of thyrotropin. To specify the relationship between thyrotropin binding and adenylate cyclase activation, binding was reexamined in the medium conditions of adenylate cyclase assay containing 64 mM ammonium sulfate. Fig. 12 shows the displacement curves of bovine '''Ilabeled thyrotropin by unlabeled bovine thyrotropin in the standard binding medium (50 mM Tris-C1 pH 7.4) with and without ammonium sulfate and in
B. Verrier, R. Planells, and S. Lissitzky
25 1
Table 4. Equilibrium dissociation constant (Kd) and capacity ( S ) of bovine 251-iubeiedthyrotropin bincling to porcinr thyroid nzembrane.~ as a,function of the cmr~positionof ihe incubation medium In medium A 5 experiments were carried out and in medium C, 3 experiments; the range of values is shown in parentheses
Ki
S
nM
pmol (mg protein)-'
A. 50 m M Tris-CI, 0.1 bovine serum albumin
0.28 (0.20- 0.33)
0.29 (0.22 - 0.30)
B. Adenylate cyclase assay medium without (N&)ZS04
0.25
0.22
C. As (A) + 64 mM (NH&S04
3.77 (1.67- 2.00)
0.21 (0.19-0.23)
2.00
0.28
Medium
D. As (B) + 64 m M (NH4)2S04
T
I
I
1
I
I
3
0
u
00
0
A
I
0
I
0
I
I
0.001 0.01 0.1 1.0 Thyro ttopin (pgirnl)
I
I
10
100
Fig. 13. Adenylate cyclase activity of thyroid plasnza mrmhrancs as a function o/ rhyrotropin concentration. (A) Bovine; (0) porcine lot 2; (0)human
the complete adenylate cyclase assay medium. The Kd values are almost identical in media without ammonium sulfate and about 6-times lower than in the complete medium of adenylate cyclase assays (Table 4). In all conditions, the capacity of membranes for thyrotropin binding was almost the same. Adenylate cyclase stimulation by thyrotropin was measured with the bovine, porcine and human hormones. As shown in Fig. 13 a sigmoidal relationship between enzyme activity and hormone concentration was observed. Half-maximum stimulation was obtained for 80 ng/ml(2.7 nM) and 1040 ng/ml(34.7 nM) of the bovine and porcine hormones respectively. With a concentration of 100 vg/ml human thyrotropin, maximum stimulation was not obtained. Assuming the same maximum value for the human as for the bovine and porcine hormones, a half-maximum value of 20 pg/ml (0.7 pM) was computed for the human hormone.
DISCUSSION Plasma membranes purified from porcine thyroid glands contain a single type of high-affinity (& = 0.28 nM) binding site for porcine and bovine thyrotropins. As for plasma membranes purified from bovine thyroid glands [4,5],binding is optimum at pH 6.0-6.5 and is inhibited by monovalent (Na', K', Li') and divalent (Mg2+,Ca2+)cations. In contrast with bovine membranes [4,5] binding increases with temperature up to 37 "C. These properties are very similar to those previously described for intact cultured porcine thyroid cells and their derived plasma membranes [2,3]. In this paper, we have verified that the porcine membrane receptor for thyrotropin binds the native hormone and its '251-labeled derivative with identical affinity. An identical affinity to the receptor for bovine and porcine thyrotropin was demonstrated whereas the affinity for the human hormone was about 200times lower. Thyrotropin and the other pituitary glycoprotein hormones (luteinizing and follicle-stimulating hormones) are formed of two dissimilar non-covalently associated n and /3 subunits. In the three hormones the CI subunit primary structure appears to be identical, whereas the hormone-specific p subunits are different [19]. The sequences of the ci and p subunit of bovine and porcine thyrotropin have been elucidated [20,18]. The porcine thyrotropin n subunit is identical to the CI subunit of porcine luteinizing hormone [21] but is shorter by six amino acid residues at the amino terminus than the bovine thyrotropin ci subunit [18]. In addition three amino acid replacements occur in chain, the the chain. When compared to the bovine /l porcine p subunit differs by six amino acid replacements and by the absence of a methionyl residue at the carboxy terminus. These mutations in the gene are conservative in terms of chemical function in the peptide chain. If other factors such as the nature of the carbohydrate moiety are not involved in the activity of the hormones, the slight differences observed in the primary structure of the ci and p subunits of the bovine and porcine hormones do not affect the affinity of the porcine receptor for both hormones. As compared to bovine thyrotropin the human n subunit is one amino acid residue longer at its amino terminus and seems to differ by 22 amino acid replacements [22]. The differences between the p subunit of bovine and human thyrotropins concern 2 1 amino acid residue positions [23]. However, nine of these changes can be considered to be highly acceptable replacements. On the whole and as compared to the bovine hormone, the differences in the primary structure of the two chains of the human hormone are larger than
252
B. Verrier, R. Planells, and S. Lissitzky: Porcine Thyroid Plasma Membranes and Thyrotropin
that showed by porcine thyrotropin. How these differences can explain the lower affinity of the human hormone for the porcine receptor remain to be determined. Stimulation of adenylate cyclase activity of the porcine thyroid plasma membranes by thyrotropin is a saturatable phenomenon. Correlation between occupancy of receptor sites by the hormone and stimulation of adenylate cyclase was demonstrated. At equilibrium and in the conditions of adenylate cyclase assay, the Kdof binding for the bovine hormone (2.0 nM) was close to the concentration of hormone which provoked half-stimulation of the enzyme (2.7 nM). The small difference observed lies within the limits of experimental errors. As for thyrotropin binding, the associated stimulation of adenylate cyclase correlates well with the biological activity of the bovine, porcine and human hormones in the bioassay in vivo in the mouse. These results are comparable to those obtained with porcine plasma membranes derived from intact cells [2] and in bovine membranes derived from glands, to the concentration of hormone giving half-stimulation of adenylate cyclase [24]. Basal and thyrotropin-stimulated adenylate cyclase activities of the porcine membranes are comparable to those observed for bovine [25- 271 or human [28] membranes. However, the factor of stimulation of the enzyme by thyrotropin is 2 - 2.5-fold and lower than that found by other investigators for bovine membranes. This low value seems more related to an increase of the basal activity of the porcine membranes than to a decrease of the thyrotropin-stimulated activity. In addition, the stimulation factor was constant for all the values of the basal adenylate cyclase activity (40- 70 pmol min-' mg protein-'). In the presence of a concentration of Mg2+ ions (4 mM) in excess of that necessary to complex all the ATP present in the assay, the K,,, of adenylate cyclase for ATP is not modified by the binding of thyrotropin to the receptor. This agrees with similar results obtained with the systems epinephrineisarcolemma [29] and glucagon/liver membranes [30]. The K , value found (520pM) is comparable to that found for the adenylate cyclase in other systems [29,31]. We are greatly indebted to D r G. Hennen (LiBge, Belgium) and t o D r J. G . Pierce (Los Angeles, Calif. U.S.A.) for the kind supply of thyrotropin. The technical assistance of J.-C. Bugeia and J. Garcia is gratefully acknowledged. We are also indebted to Mr B. Rousset (Service rl' Entlocrinologir, Hopiral de I'Antiyuuille, Lyon) who tested the biological activities of the thyrotropin preparations used
in our investigations. This work was supported in part by the Centre National de lu Recherche Scientifique (contract 1919 ATP Micanisme d'action des hormones).
REFERENCES 1. Lissitzky, S., Fayet, G. & Verrier, B. (1976) Methods in Receptor Research (Blecher, M., ed.) vol. 2. pp. 641-665, Marcel Dekker, New York. 2. Verrier, B., Fayet, G . & Lissitzky, S. (1974) Eur. J . Bioci7on. 42,355 - 365. 3. Lissitzky, S., Fayet, G . & Verrier, B. (1975) Adi,. C:vclic Nucleotide Res. 5 , 133-152. 4. Amir, S. M., Carraway, T. F.. Kohn, L. D. & Winand, R. J . (1973) J . Biol. C'hem. 248, 4092-4100. 5. Moore, W. V . & WolH, J.(1973)J. Biul. Clii.m.248,5705-5711. 6. Taylor, S. I. (1975) Biochen?istry, 14, 2357-2361. 7. McKenzie, J. M. (1958) Endocrinology, 63, 372-378. 8. Jaquet, P., Hennen, G. & Lissitzky, S. (1974) Biochiinie (Parly) 56,169 - 774. 9. Planells, R., Fayet, G., Lissitzky. S., Hennen, G. & Closset, J . (1975) FEBS Lett. 53, 87-92 10. Solomon, Y., Londos, C. & Rodbell, M. (1974) Anal. Biocl7rrn. 58,541 - 548. 11. Delaage, M. A., Bellon, N. & Cailla, H . L. (1974) Anal. Biochem. 62,417-425. 12. Randerath, K . & Randerath. E. (1967) Mi,rizoifs E ~ i : y / n d , 12, 323 - 347. 13. Cache, C., Rossi, B. & Lazdunski, M . (1976) Eur. J . B i u d 7 ~ ~ 7 . 65,293 - 306. 14. Aronson, N. N . Jr & Touster, 0. (1974) Methods Enzj.mol. 31. 90- 102. 15. Weaver, R. A. & Boyle, W. (1969) Biochim. Biophj.s. Acta. 173. 377-388. 16. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall. R. J. (1951) J . B i d . Chem. 193, 265-275. 17. Lissitzky, S.. Fayet, G., Verrier, B., Closset, J. B Hennen, G. (1974) FEBS Lett. 48, 275-278. 18. Maghuin-Rogister, G., Hennen, G., Closset. J . & Kopeyitn. C. (1976) Eur. J . Biochem. 61, 157- 163. 19. Pierce, J. G. (1971) Endocrinolug~.,8Y, 1331- 1344. 20. Liao, T. H. &Pierce, J . G. (1971)J. Biol. Chcw;. 246, 850-865. 21. Maghuin-Rogister, G. & Hennen, G . (1973) Eur. J . Bioc~lron. 39,255- 263. 22. Sairain, M. R . & Li, C. H. (1973) Biorkeni. Bioplij~.~.Rrs. Commun. 51, 336-342. 23. Sairam, M. R. & Li, C. H. (1973) Biochem. Bioplij's. Res. C'ommun. 54.426-431. 24. Yamashita. K. & Field, J . B. (1972) J . C'lin. Inwst. 51, 463472. 25. Wolff, J. & Cook, G. H. (1973) J. Biol. Chenz. 248, 350-355. 26. Mashiter, K., Mashiter, G. D. & Field, J. B. (1954) Endxrino/O,~V, 94, 370- 376. 27. Marshall, N . J., Von Borcke, S. & Malan, P. G . (1975) Endocrinologj', 96, 1513- 1519. 28. Sato, S., Yamada, T., Furihata, R. & Makinchi, M . (1974) Biochim. Biop1zJ.s.Actu, 332, 166- 174. 29. Severson, D. L., Drummond, G. I. & Sulakhe, P. V. (1972) J. Biol. Chem. 247, 2949- 2958. 30 Lin, M. C., Salomon, Y., Rendell, M. & Rodbell, M . (1975) J . Biol. Chcm. 25U, 4246- 4252. 31 Ho, R. J . B Sutherland, E. W . (1975) Proc. Nail Acud. &i. U . S . A . 72, 3773--1777.
B. Verrier, R. Planells, and S. Lissitzky, Laboratoire de Biochimie MCdicale, Faculte de Medecine, Universiti d'Aix-Marseille, 27 Boulevard Jean-Moulin, F-I 3385 Marseille-Cedex-4, France
