Brain Research, 528 (1990) 291-299
291
Elsevier BRES 15881
Characterization of [125I-Tyr1°]human growth hormone-releasing factor(i-44) amide binding to rat pituitary: evidence for high and low affinity classes of sites T. Abribat, L. Boulanger and R Gaudreau Neuroendocrinology Laboratory, Notre-DameHospital Research Center, University of Montreal, Montreal, Que. (Canada) (Accepted 27 March 1990)
Key words: Growth hormone-releasing factor; Binding site; Pituitary; Rat
A sensitive binding assay was developed to determine binding characteristics of the commercially available [125I-Tyr1°]humangrowth hormone-releasing factor (hGRF) (1-44)NH z in rat pituitary using 0.1 gland homogenate (70-75 gg protein) per incubation tube. Under standard assay conditions, addition of 5 mM EDTA efficiently prevented the degradation of both human and rat GRF for at least 3 h. Association of the ligand was time-dependent: equilibrium was reached within 30 min of incubation at 23 °C and remained stable for an additional 150 min (K1 = 5.01 _+ 0.86 nM-l.min-a). Specific binding increased linearly with the amount of protein present in the assay, from 15 to 170 pg per incubation tube. This binding was reversible, dissociation occurring almost completely after a 120-rain period (/(1 = 8.13 _+ 0.29 × 10-3 rain-l). A concentration of 5-10 mM Mg2+ was required for optimal specific binding whereas 50 mM Mg2+ or monovalent cations such as Na +, K÷, Li÷ decreased it. Scatchard analysis of cold saturation studies by the Ligand program statistically revealed the presence of two distinct classes of binding sites: the first was of high affinity (0.68 _+ 0.11 nM) and low capacity (140 + 22 fmoi/pituitary), the second was of lower affinity (590 + 347 nM) and higher capacity (38.7 + 18.7 pmol/pituitary). Similar values were obtained with various bovine serum albumin (BSA) concentrations and when using crude or washed pituitary homogenates, suggesting that the second low affinity site was not BSA or a soluble protein from the homogenate. Moreover, since vasoactive intestinal polypeptide, peptide histidine isoleucine (PHI), secretin or somatostatin (SRIF) could not displace [125I-Tyrt°]hGRF(1-44)NH2binding at a concentration of 2.4/~M, it might be supposed that the second site is rather specific for GRF. Finally, a partial inhibitory effect of hydrolysable and non-hydrolysable guanyl nucleotides was observed, and the affinity of various GRF analogs and related peptides, as determined by competition studies, correlated with their documented biological activity on GH release in vitro. The high affinity and specificity of [125I-Tyr~°]hGRF(1-44)NH2,together with the miniaturization of the assay, allowed us to achieve further characterization of GRF binding to rat pituitaries, in the course of which we have pinpointed the presence of two classes of binding sites for this new radioligand. Whether they represent two different receptor entities and whether they both mediate some physiological actions of GRF remains to be elucidated.
INTRODUCTION Since its discovery in 19828 , growth hormone-releasing factor ( G R F ) has been the subject of n u m e r o u s studies regarding its ability to induce growth h o r m o n e (GH) secretion from the somatotroph cells. Numerous clinical and zootechnical applications in man, pig and cattle, using h u m a n ( h ) G R F ( 1 - 4 4 ) N H 2 or its analogs are currently u n d e r investigation18'24. However, binding characteristics of this e n d o g e n o u s peptide to anterior pituitary m e m b r a n e s receptors have never been documented. Specific binding sites for [HisX,125I-Tyr1°,Nle27]hGRF(132)NH 2 have been described in rat pituitary cells2° and homogenates 21 and in h u m a n pituitary adenomas 9, while specific binding sites for [125I-Tyr1°]hGRF(1-40)OH have been d o c u m e n t e d in bovine pituitary m e m b r a n e s 25. So far all binding studies in rat have been performed with
[His1,125I-Tyra°,NIe27]hGRF(1-32)NHz,
a
radioligand
that possesses a 75% sequence homology with rat ( r ) G R F ( 1 - 3 2 ) amide and exhibits a high biological potency in this species. These initial reports have docum e n t e d the presence of a single class of G R F binding sites. Native h u m a n G R F ( 1 - 4 4 ) N H 2 possesses a high degree of sequence homology not only with rat G R F ( 1 - 4 3 ) O H (67%), but also with porcine and bovine G R F ( 1 - 4 4 ) N H 2 ( > 8 9 % ) , and shows a high biological activity in all these species. For these reasons, we wanted to examine the binding characteristics of [125I-Tyrl°]hGRF(1-44)NH2 to rat pituitaries. We describe here a new sensitive binding assay of the commercially available m o n o i o d i n a t e d analog of natural h u m a n G R F ( 1 - 4 4 ) N H 2 , using a pituitary protein concentration as low as 70-75 p g (0.1 pituitary) per incubation tube. This microbinding assay allowed us
Correspondence: P. Gaudreau, Neuroendocrinology Laboratory, Notre-Dame Hospital Research Center, 1560 East Sherbrooke, Montreal, H2L 4M1, Que., Canada. 0006-8993/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
292 to c h a r a c t e r i z e
[125I-Tyrl°]hGRF(1-44)NH2
binding in
rat p i t u i t a r y h o m o g e n a t e s , and to d e m o n s t r a t e that this r a d i o l i g a n d binds to two d i f f e r e n t classes of sites in this tissue.
MATERIALS AND METHODS
Chemicals [125I-Tyr1°]hGRF(1-44)NH2 (125I-GRF; 2000 Ci/mmol) was purchased from Amersham (Oakville, Ont.). The lyophylized radioligand was reconstituted in 0.1% aqueous acetic acid, aliquoted in polypropylene microcentrifuge tubes (1/~Ci/10 pl/tube) and stored at -20 °C. Under these conditions, no major radiochemical degradation of 125I-GRF affecting binding parameters occurred within at least 3 weeks. All peptides were synthesized by solid-phase methodology TM in our laboratory, except for secretin, vasoactive intestinal polypeptide (VIP) and peptide histidine isoleucine (PHI), which were generously given by Dr. Alain Fournier (INRS-Santr, Pointe Claire, Qu6.). Rat GRF and its analogs were solubilized at a concentration of 0.15-0.3 mM in 51% 0.1 N HCI and 49% 0.1 N NaOH. Dilutions were freshly prepared before use in the assay buffer containing 0.5% bovine serum albumin (BSA). Other peptides and guanyl nucleotides (Sigma, St. Louis, MO) were solubilized in the assay buffer just before the experiments. Tissue preparations Male Sprague-Dawley rats (200-225 g) were purchased from Charles River Canada Inc. (St Constant, Qur.) and had free access to food and water. Anterior pituitaries were dissected out immediately after decapitation, rinsed and homogenized in ice-cold buffer (one pituitary in 500 pl of 50 mM Tris-HCl buffer, pH 7.4, containing 5 mM MgCI2 and 5 mM EDTA) using a Potter-Elvehjem tissue grinder. This homogenate was used for all binding experiments (50 pl/tube), except when otherwise mentioned. For some comparative cold saturation studies, this homogenate was centrifuged at 2000 g for 30 min at 4 °C, the pellet was rehomogenized in a same volume of buffer and used in the assay at a concentration of 50 pl/tube. Protein concentrations were determined by the Lowry method TM using BSA as standard. Binding assay ~2SI-GRF binding assay was carried out at 23 °C for 60 rain, in 50 mM Tris-HCl buffer, pH 7.4, containing 5 mM MgC12, 5 mM EDTA, 0.42% BSA, 35-50 pM [lzSI-Tyrl°]hGRF(1-44)NH2 (around 50,000 cpm), and 70-75 #g protein homogenate, corresponding to 0.1 pituitary, in a final volume of 300 pl. Incubation was terminated by immersing the 12 x 75 mm borosilicate tubes in ice-water, transferring 280 pl of the preparation in cold polyethylene microfuge tubes and centrifuging it at 12,000 g in a Beckman microfuge for 5 min at 4 °C. Supernatants were then aspirated, tube tips were cut off and counted for radioactivity in a y-counter. Specific binding was defined as the difference in radioactivity bound in the absence and presence of 2.4 /~M rGRF(1-29)NH 2 and represented 55-70% of total binding. Preliminary experiments using rapid filtration (Wattman GF/C or GF/B filters) to terminate the reaction resulted in an apparent specific binding in the absence of pituitary homogenate, even when the filters were presoaked with 1% BSA, 0.01 or 0.1% polyethyleneimine, 0.1% poly-L-lysine or 0.01% GRF (data not shown). Some experiments were performed with tissue protein concentrations ranging from 15 to 170/~g in order to determine the minimal amount of protein required for a sufficient specific binding. Others were conducted with various concentrations of cations and EDTA. In kinetic experiments, incubation time varied from 5 to 180 min. In hot saturation studies, increasing concentrations of radioligand (9-350 pM) were tested with and without 2.4 pM rGRF(1-29)NH 2 in order to determine non-specific binding. In cold saturation studies, 35-50 pM radioligand was used in all tubes, with increasing
concentrations (0-100 nM) of non-radioactive [1z51-Tyr1°]hGRF(144)NH2, or with 2.4 pM of rGRF(1-29)NH 2 for non-specific binding evaluation. Finally, competition studies were performed with increasing concentrations of GRFs and other peptides (10 pM to 2.4 gM), or with guanyl nucleotides (10-100 gM).
Degradation assay of hGRF(1-44)NH2 and rGRF(1-29)NH2 in pituitary homogenates Enzymatic breakdown of hGRF(1-44)NH 2 and rGRF(1-29)NH 2 during incubation was investigated in the standard binding assay conditions by high performance-liquid chromatography (HPLC). The reaction was performed in the incubation buffer with or without 5 mM EDTA. Samples of pituitary homogenates (150/tg protein/100 /~l) were incubated with 10 ktM hGRF(1-44)NH 2 or rGRF(129)NH 2 in a final volume of 500 ktl at 23 °C for 0-180 min. The degradation reaction was stopped by the addition of 300 pl cold 50 mM phosphate buffer, pH 0.8, and immediate centrifugation at 12,000 g for 5 min at 4 °C. Parallel experiments were conducted without pituitary homogenates to control GRF's adsorption to borosilicate tubes. The residual GRF concentrations were analyzed by reverse-phase HPLC, on a ~-Bondapak C~g column (39 x 150 mm, 10-pm particles; Waters, Mississauga, Canada) using a Waters instrument equipped with an automated injector, a gradient controller and a two-wavelength (214,280 nm) UV detector coupled to a two-pen recorder. The mobile phase was composed of 0.01% trifluoroacetic acid (TFA) in acetonitrile (CH3CN) (eluent A) and of 0.01% aqueous TFA, pH 2.9, (eluent B). A linear gradient of 1.0% increase of eluent B/min, for 30 rain, was used with 20% of B as initial condition and a flow rate of 1.5 ml/min. Quantification was achieved by peak height measurement which gave identical results to surface integration in preliminary experiments. Protein dilution followed by immediate centrifugation efficiently stopped the peptide degradation. This was shown by equivalent peptide recoveries for time zero of incubation with or without membranes. Data analysis The Ligand computerized, non-linear regression curve fitting program of Munson and Rodbard 16 was used to analyze association, dissociation, competition and saturation studies. The Scatchard transforms 19 of hot and cold saturation curves were used to derive estimates of the apparent affinity constants (Ka) and binding sites capacities (Bmax). First, each curve was individually analyzed; then, all data related to a particular study were simultaneously coanalyzed. One important feature of the Ligand program is its ability to pool and fit data from multiple experiments: this allows to correct data for variable receptor capacities and improves the quality of statistical analysis when comparing different models and hypotheses TM. For each analysis, estimates of binding parameters were calculated for 3 models, whenever the fit was possible: one class of sites, two classes of sites, and one class of sites, corrected for non-specific binding. In this last fit, the non-specific binding is allowed to 'float', the model considering the second site as another form of non-specific binding which is displaced by rGRF(1-29)NHz, used to determine non-specific binding. Finally, the Ligand program determined the best fit for each analysis, by comparing residual sums of squares of experimental points to fitted curves with the F-test. The existence of a second site was statistically judged in all analyses.
RESULTS U s i n g t h e c o n d i t i o n s d e s c r i b e d in M a t e r i a l s and M e t h ods, specific [125I-Tyr1°]hGRF(1-44)NH2 b i n d i n g (ratio o f r a d i o l i g a n d specifically b o u n d o v e r t o t a l r a d i o l i g a n d a d d e d ) was l i n e a r as a f u n c t i o n o f p r o t e i n c o n c e n t r a t i o n s that r a n g e d b e t w e e n 15 and 1 7 0 / ~ g (Fig. 1). It r e p r e s e n t e d 7 5 % of the total b i n d i n g at 170 ~tg p r o t e i n , this
293 percentage slightly decreasing with decreasing amounts of protein. The best compromise between specific binding and amount of protein was obtained in our view with 70-75 /tg protein per assay, corresponding to 0.1 of pituitary equivalent, with 55-70% of total binding. In the standard conditions of our assay (35-50 pM radioligand, 70-75 /tg protein, 0.42% BSA), there was a slight adsorption of the radioligand on the incubation tubes (typically 8-10% of 125I-GRF added). This adsorption did not change with addition of unlabeled GRFs (10 pM to 2.4/~M). The kinetic of association at 23 °C showed that 125I-GRF specific binding increased with time, reached a steady-state within 30 min and remained stable for at least an additional 150 min (Fig. 2, top panel). Subsequent experiments were carried out for 60 min. The high levels of total and non-specific binding in the initial phase of the reaction led us to investigate a possible interference of BSA, through a non-specific adsorption of GRFs to this protein. Therefore, experiments were conducted with various concentrations of BSA. At 0.17, 0.42 and 0.75% BSA (n = 3, for each concentration), there was no statistical difference at equilibrium (60 min incubation) between total binding (9.6 + 0.9, 9.5 _+ 1.1, 8.7 _+ 0.9%, respectively, P > 0.05), non-specific binding (3.9 + 0.4, 3.7 + 0.5, 3.6 __+0.4% respectively, P > 0.05) and specific binding (5.7 + 0.6, 5.8 + 0.7, 5.1 + 0.5% respectively, P > 0.05) (all values being expressed as B/T, i.e. the ratio of total, non-specific or specific binding over total radioligand added). These results suggest that BSA does not interfere with GRF binding at equilibrium in our assay and enabled the use of 0.42% BSA in the incubation buffer. Stability of GRFs to proteolysis was evaluated for hGRF(1-44)NH2, a close analog to the radioligand, and for rGRF(1-29)NH 2, a GRF N-terminal fragment used to determine non-specific binding. As shown in Fig. 2 (bottom panel), hGRF(1-44)NH 2 concentration did not significantly diminish after 60 or 180
min incubation in the assay conditions, in the presence of 5 mM EDTA. The remaining concentrations were, respectively, 84 + 8% and 79 _+ 9% of those at time zero. Similar results were obtained with rGRF(1-29)NH2, with a 107 _+ 13% and 90 _+ 13% recovery, under the same experimental conditions. Omission of EDTA, a metalloprotease inhibitor, revealed the presence of GRF-sensitive protease activities, since 76 _+ 7% and 29 +_ 3% of hGRF(1-44)NH2 initial concentration was recovered after 60 and 180 min, respectively, while 53 _+ 10% and 12 + 4% of rGRF(1-29)NH 2 initial concentration remained after the same periods of time. Glass adsorption of rGRF under the same experimental conditions was 3 _+ 1% after 60 min and did not increase further, while it represented 3 _+ 9% and 11 _+ 5% for hGRF after 60 and 180 min, respectively. The reversibility of 125I-GRF binding was evaluated by adding 2.4/~M of rGRF(129)NH 2 at equilibrium (60 min) (Fig. 3). Dissociation occurred quickly at first, then slowed down and was almost complete after 120 min of additional incubation time. Association and dissociation rate constants, as estimated by the Ligand program were, respectively, 5.01 _+ 0.86 nM-l.min -1 and 8.13 + 0.29 × 10 -3 min -1. To optimize the ratio of specific to total binding, incubations were performed with various cations and
~..
0.10
lIB
0.05
0.00 z
0.10
= - - e Specific binding o ~ o Nonspecific binding
0.15
lOOJ
~--
80
Z
60
Z
0.08
8§
0.06
_~40
0.04
u_
\
"r--, • • ----
m
20
0
hGRF(1-4A,)NH2 rGRF(1-29)NH 2 5mM EDTA OmM EDTA
0
0.02
0 i
0.00
40
i
i
,
80 120 160 200 TOTALPROTEIN(b~J/oesay) Fig. 1. Effect of protein concentration on specific 125I-GRF binding. Each point represents the mean of triplicate determinations. Data shown are gathered from two independent experiments. 0
60
t
I
120
180
TIME (rain) Fig. 2. Time course of 125I-GRF association to pituitary homogehates at 23 °C (top panel, mean _+ S.E.M. of 4 independent experiments) and stability of hGRF(1-44)NH 2 and r G R F ( 1 29)NH2 in pituitary homogenate in the absence and presence of 5 mM EDTA (bottom panel, mean _+ S.E.M. of 3 experiments).
294 0.08
0o8
"~
0.04
0.02
0.00
-
~ -
-
-@
~
'
'
'
0
60
120
180
TIME (rain)
Fig. 3. Dissociation of specific 125I-GRF binding to rat pituitary homogenates. The reaction was initiated by adding 2.4 /~M unlabeled rGRF(1-29)NH 2 at equilibrium (60 rain). Each point represents the mean + S.E.M. of 4 experiments, each done in triplicate.
E D T A concentrations (Fig. 4). Removing EDTA from the incubation buffer resulted in a decrease of specific binding, likely due to GRF degradation. There was no difference between 5 and 10 mM, confirming that 5 mM is enough to prevent GRF breakdown by pituitary proteolytic enzymes. As shown in Fig. 4, optimal specific binding required the presence of Mg z+ in the incubation buffer; 5 and 10 mM were equally potent and sufficient, while 50 mM Mg 2÷ resulted in an 87% decrease in specific binding. Addition of 50 mM Na ÷ also reduced the specific binding of 125I-GRF by 60%. This effect was Na+-concentration-dependent and also occurred to a lesser extent with K ÷ and Li ÷ (data not shown). Therefore, the incubation buffer for all subsequent experiments contained 5 mM EDTA and 5 mM MgCI 2. In hot saturation experiments, 125I-GRF specifically bound increased with radioligand concentrations ranging from 9 to 350 pM. Larger concentrations could not be tested in these studies, due to an increased adsorption on incubation tubes at high radioligand concentrations, and
120 I00
~o
8o
,.o_ o
60
~
40
......
i
2O 0 0 5 0
10 5 0
5 0 0
5 10 0
5 50 0
5 5 50
EDTA (mM) MgCl2 (raM) NaCl (mM)
Fig. 4. Effect of EDTA and cations on specific 125I-GRFbinding to rat pituitary homogenates. Specific binding with 5 mM EDTA and 5 mM MgCI2 was defined as 100%. Each value is the mean + S.E.M. of 3 experiments, each done in triplicate.
to possible interferences of scavengers present in the commercial preparation of 125I-GRF, such as lactose (5%), L-cysteine hydrochloride (0.1%) and aprotinin (800 KIU/ml). Scatchard transformation of 3 separate experiments gave linear plots and fitting these experimental data with the Ligand program revealed an apparent single class of binding sites (Table I), fitting with a two-sites model being impossible. We then synthesized [127I-Tyr1°]hGRF(1-44)NH2 to perform cold saturation studies. Therefore, we were able to test a much larger range of ligand concentrations (10 pM to 100 nM). The displacement curves of 125I-GRF by its non-radioactive analog suggested that this peptide bound to two classes of sites (Fig. 5, top panel). Using computerized Scatchard analysis, the data of these competition curves were described as curvilinear plots (data not shown). When resolved in two components, the first class of sites was of high affinity and low capacity, while the second had a 500-fold lower affinity and was present in a 200-fold excess relative to the high affinity site (Table I). According to the Ligand interpretation, the two-sites model was statistically preferred to the one-site model (P < 0.001 for all 6 experiments and for coanalysis of all data) and to the one-site model corrected for non-specific binding (P < 0.05 for 4 of 6 experiments and P < 0.02 for coanalysis of all data). To further investigate the nature of this second binding site, we performed cold saturation experiments with 2000 g pituitary rehomogenized pellets instead of crude homogenates and 0.1% BSA instead of 0.42% BSA. This was meant to determine whether the adsorption of BSA or of a soluble protein from the homogenate could be responsible for the appearance of the second low affinity site. As with homogenates, 0.1 pituitary equivalent per assay tube was used, corresponding to about 40 pg protein. Binding experiments of 125I-GRF to washed membranes resulted in a decreased percentage of specific binding (defined as the ratio of specific to total binding, 55-70% with homogenate, 35-40% after washing), due to an increase of non-specific binding, the ratio of specific binding to total ligand added (B/T) being unchanged. The displacement curves of 125I-GRF by its non-radioactive congener suggested again two classes of sites (Fig. 5, bottom panel), as well as the curvilinear Scatchard transformations of the data. When fitting these data with Ligand, the two-site model was statistically preferred to the one-site model (P < 0.001 or 0.01 for all 4 experiments and P < 0.001 for coanalysis of all data) and to the one-site model corrected for non-specific binding (P < 0.05 for 2 of 4 experiments, P < 0.005 for coanalysis of all data). As shown in Table I, all saturation studies gave nearly identical parameters (K d and Bmax) for the high affinity
295 TABLE I Estimates of binding parameters (Kd = affinity and Bmax = capacity) obtained by analysis with Ligand of hot and cold saturation studies (S.S.) (a) Mean + S.E.M. of n experiments, individually analyzed by Ligand. (b) Mean ___S.E.M. as calculated by Ligand when all n experiments were coanalyzed. S.S.
Kdl (nM)
Hot (n = 3) homogenates 0.42% BSA
(a) (b) (a) (b) (a) (b)
Cold (n = 6) homogenates 0.42% BSA Cold (n = 4) 2000 g pellets 0.10% BSA
Ka2 (nM)
0.28 + 0.07 0.24 + 0.07 0.68-+0.11 0.86 + 0.15 0.35 + 0.09 0.36 _+0.12
site. H o w e v e r , the estimated precision of the parameters for the low affinity site was less good, due to the increased mathematical complexity of the two-site model 16. Despite that, the existence of the second site was judged statistically significant; moreover, gd2 and Bmax2, as determined with pituitary homogenates and 0.42% B S A or with 2000 g pellets and 0.10% B S A were not different ( P > 0.05). Some attempts to perform cold
120
,ooi ~R
v
(.9
Z
2O
Z 0 E
IM fit. (/1
,oo
:, -.:\
Bmaxl
-
590-+347 400 +_210 80 _+35 67 _+31
B,n~x2
(fmol/ pituitary)
(fmol/ mg prot)
57 + 12 78 + 21 140+22 202 + 35 96 + 24 111 _+33
76 __+16 104 + 28 187+29 269 -+ 47 240 + 60 277 + 83
(pmol/ pituitary) -
38.7-+18.7 31.5 -+ 14.4 5.4 _ 2.3 4.8 _+1.6
(pmol/ mg prot) -
51.6-+25 42.0 -+ 19.2 13.5 -+ 5.8 12.0 + 4.0
saturation experiments without B S A resulted in inconsistent values for affinities, due to a dramatic increase in G R F ' s adsorption to preparation and incubation tubes; however, these data were again best fitted with a two-site model (data not shown). Various guanyl nucleotides inhibited the specific 125IG R F binding to rat pituitary homogenates (Fig. 6). Hydrolysable guanosine 5"-triphosphate (GTP), guanosine 5"-diphosphate ( G D P ) and non-hydrolysable 5"guanylylimidodiphosphate ( G p p N H p ) , guanosine 5"-o(3-thiotriphosphate) (GTP-7-S) nucleotides were equally potent at 10/~M, whereas cyclic guanosine 5"-monophosphate (cGMP) had no effect. A concentration-effect was observed for G p p N H p and GTP-7-S, with an improved inhibitory effect on specific binding at 100/~M. Since concentrations greater than 100 a M of nucleotides have not been evaluated, a complete inhibition was not obtained. Selectivity of 125I-GRF binding was examined by the ability of several G R F analogs and unrelated peptides to compete with the radioligand. Their ICsos and relative
100
80
20 0 --1 1
I I --10
I --9 LOG
I --8 TOTAL
•
I --7
--6
(M)
Fig. 5. Competition studies of 125I-GRFspecifically bound with [127I-Tyrl°]hGRF(1-44)NH2 as competitor, with crude homogehates and 0.42% BSA (top panel) or with washed homogenates and 0.10% BSA (bottom panel). Data are plotted as percentage of radioligand specifically bound (B/T) vs log Total [I-Tyrl°]hGRF(1 44)NH 2 (radioactive and non-radioactive). Each point is the average of duplicate or triplicate tubes. The mean curves were obtained from the Ligand program, when coanalysing all curves (top panel, n = 6, and bottom panel, n = 4).
"
~
40
20 o
G'rP
lo~u
GDP GMP¢ GppNHp GppNl.,Ip GTP.yS GTPTS IO#M IO/=M lO~la tOOj~M ~O0#M 100~
Fig. 6. Inhibition of specific 125I-GRFbinding by guanyl nucleotides at various concentrations. Specific binding in the absence of nucleotides was defined as 100%. Each value represents the mean _+ S.E.M. of 3 independent experiments, each done in triplicate.
296 TABLE II
Relativeaffinities of different GRF analogs Affinities were estimated by /(?5o, concentration of peptide displacing 50% of specifically bound 125I-GRF, and was determined with Ligand program. Relative affinities (R.A.) to rGRF(1-29)NH 2 were calculated for each experiment as the ratio of 1C5oof the tested GRF peptides to the ICso of rGRF(1-29)NH 2. Values are expressed as the mean + S.E.M. of 3 experiments, each performed in triplicate.
Peptide* rGRF(1-29)NH 2 rGRF(1-43)OH hGRF(1-29)NH 2 hGRF(1-44)NH 2 [127I-TyrX°lhGRF(1-44)NH2 [desaminoTyr1,D-Ala2, AlalS]hGRF(1-29)NH2 [His~,Nle27]hGRF(1-29)NHz [AIaXS]hGRF(2-29)NH2 [Ala~5]hGRF(3-29)NH2 [desLys21]hGRF(1-29)NH2
1C5o(nM) 0.72 + 0.10 0.36 + 0.04 6.30 + 1.70 0.85 + 0.47 1.16 _+0.23
R.A. (%) 100 170 + 38 12.8 + 4.0 104 _+25 90 _+22
0.73 _+0.03 67.4 _+10.3 0.82 _+0.18 80.3 _+10.3 376.7 _+153.2 0.4 + 0.1 311.0 _+97.6 0.4 + 0.1 >2400
291
Elsevier BRES 15881
Characterization of [125I-Tyr1°]human growth hormone-releasing factor(i-44) amide binding to rat pituitary: evidence for high and low affinity classes of sites T. Abribat, L. Boulanger and R Gaudreau Neuroendocrinology Laboratory, Notre-DameHospital Research Center, University of Montreal, Montreal, Que. (Canada) (Accepted 27 March 1990)
Key words: Growth hormone-releasing factor; Binding site; Pituitary; Rat
A sensitive binding assay was developed to determine binding characteristics of the commercially available [125I-Tyr1°]humangrowth hormone-releasing factor (hGRF) (1-44)NH z in rat pituitary using 0.1 gland homogenate (70-75 gg protein) per incubation tube. Under standard assay conditions, addition of 5 mM EDTA efficiently prevented the degradation of both human and rat GRF for at least 3 h. Association of the ligand was time-dependent: equilibrium was reached within 30 min of incubation at 23 °C and remained stable for an additional 150 min (K1 = 5.01 _+ 0.86 nM-l.min-a). Specific binding increased linearly with the amount of protein present in the assay, from 15 to 170 pg per incubation tube. This binding was reversible, dissociation occurring almost completely after a 120-rain period (/(1 = 8.13 _+ 0.29 × 10-3 rain-l). A concentration of 5-10 mM Mg2+ was required for optimal specific binding whereas 50 mM Mg2+ or monovalent cations such as Na +, K÷, Li÷ decreased it. Scatchard analysis of cold saturation studies by the Ligand program statistically revealed the presence of two distinct classes of binding sites: the first was of high affinity (0.68 _+ 0.11 nM) and low capacity (140 + 22 fmoi/pituitary), the second was of lower affinity (590 + 347 nM) and higher capacity (38.7 + 18.7 pmol/pituitary). Similar values were obtained with various bovine serum albumin (BSA) concentrations and when using crude or washed pituitary homogenates, suggesting that the second low affinity site was not BSA or a soluble protein from the homogenate. Moreover, since vasoactive intestinal polypeptide, peptide histidine isoleucine (PHI), secretin or somatostatin (SRIF) could not displace [125I-Tyrt°]hGRF(1-44)NH2binding at a concentration of 2.4/~M, it might be supposed that the second site is rather specific for GRF. Finally, a partial inhibitory effect of hydrolysable and non-hydrolysable guanyl nucleotides was observed, and the affinity of various GRF analogs and related peptides, as determined by competition studies, correlated with their documented biological activity on GH release in vitro. The high affinity and specificity of [125I-Tyr~°]hGRF(1-44)NH2,together with the miniaturization of the assay, allowed us to achieve further characterization of GRF binding to rat pituitaries, in the course of which we have pinpointed the presence of two classes of binding sites for this new radioligand. Whether they represent two different receptor entities and whether they both mediate some physiological actions of GRF remains to be elucidated.
INTRODUCTION Since its discovery in 19828 , growth hormone-releasing factor ( G R F ) has been the subject of n u m e r o u s studies regarding its ability to induce growth h o r m o n e (GH) secretion from the somatotroph cells. Numerous clinical and zootechnical applications in man, pig and cattle, using h u m a n ( h ) G R F ( 1 - 4 4 ) N H 2 or its analogs are currently u n d e r investigation18'24. However, binding characteristics of this e n d o g e n o u s peptide to anterior pituitary m e m b r a n e s receptors have never been documented. Specific binding sites for [HisX,125I-Tyr1°,Nle27]hGRF(132)NH 2 have been described in rat pituitary cells2° and homogenates 21 and in h u m a n pituitary adenomas 9, while specific binding sites for [125I-Tyr1°]hGRF(1-40)OH have been d o c u m e n t e d in bovine pituitary m e m b r a n e s 25. So far all binding studies in rat have been performed with
[His1,125I-Tyra°,NIe27]hGRF(1-32)NHz,
a
radioligand
that possesses a 75% sequence homology with rat ( r ) G R F ( 1 - 3 2 ) amide and exhibits a high biological potency in this species. These initial reports have docum e n t e d the presence of a single class of G R F binding sites. Native h u m a n G R F ( 1 - 4 4 ) N H 2 possesses a high degree of sequence homology not only with rat G R F ( 1 - 4 3 ) O H (67%), but also with porcine and bovine G R F ( 1 - 4 4 ) N H 2 ( > 8 9 % ) , and shows a high biological activity in all these species. For these reasons, we wanted to examine the binding characteristics of [125I-Tyrl°]hGRF(1-44)NH2 to rat pituitaries. We describe here a new sensitive binding assay of the commercially available m o n o i o d i n a t e d analog of natural h u m a n G R F ( 1 - 4 4 ) N H 2 , using a pituitary protein concentration as low as 70-75 p g (0.1 pituitary) per incubation tube. This microbinding assay allowed us
Correspondence: P. Gaudreau, Neuroendocrinology Laboratory, Notre-Dame Hospital Research Center, 1560 East Sherbrooke, Montreal, H2L 4M1, Que., Canada. 0006-8993/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
292 to c h a r a c t e r i z e
[125I-Tyrl°]hGRF(1-44)NH2
binding in
rat p i t u i t a r y h o m o g e n a t e s , and to d e m o n s t r a t e that this r a d i o l i g a n d binds to two d i f f e r e n t classes of sites in this tissue.
MATERIALS AND METHODS
Chemicals [125I-Tyr1°]hGRF(1-44)NH2 (125I-GRF; 2000 Ci/mmol) was purchased from Amersham (Oakville, Ont.). The lyophylized radioligand was reconstituted in 0.1% aqueous acetic acid, aliquoted in polypropylene microcentrifuge tubes (1/~Ci/10 pl/tube) and stored at -20 °C. Under these conditions, no major radiochemical degradation of 125I-GRF affecting binding parameters occurred within at least 3 weeks. All peptides were synthesized by solid-phase methodology TM in our laboratory, except for secretin, vasoactive intestinal polypeptide (VIP) and peptide histidine isoleucine (PHI), which were generously given by Dr. Alain Fournier (INRS-Santr, Pointe Claire, Qu6.). Rat GRF and its analogs were solubilized at a concentration of 0.15-0.3 mM in 51% 0.1 N HCI and 49% 0.1 N NaOH. Dilutions were freshly prepared before use in the assay buffer containing 0.5% bovine serum albumin (BSA). Other peptides and guanyl nucleotides (Sigma, St. Louis, MO) were solubilized in the assay buffer just before the experiments. Tissue preparations Male Sprague-Dawley rats (200-225 g) were purchased from Charles River Canada Inc. (St Constant, Qur.) and had free access to food and water. Anterior pituitaries were dissected out immediately after decapitation, rinsed and homogenized in ice-cold buffer (one pituitary in 500 pl of 50 mM Tris-HCl buffer, pH 7.4, containing 5 mM MgCI2 and 5 mM EDTA) using a Potter-Elvehjem tissue grinder. This homogenate was used for all binding experiments (50 pl/tube), except when otherwise mentioned. For some comparative cold saturation studies, this homogenate was centrifuged at 2000 g for 30 min at 4 °C, the pellet was rehomogenized in a same volume of buffer and used in the assay at a concentration of 50 pl/tube. Protein concentrations were determined by the Lowry method TM using BSA as standard. Binding assay ~2SI-GRF binding assay was carried out at 23 °C for 60 rain, in 50 mM Tris-HCl buffer, pH 7.4, containing 5 mM MgC12, 5 mM EDTA, 0.42% BSA, 35-50 pM [lzSI-Tyrl°]hGRF(1-44)NH2 (around 50,000 cpm), and 70-75 #g protein homogenate, corresponding to 0.1 pituitary, in a final volume of 300 pl. Incubation was terminated by immersing the 12 x 75 mm borosilicate tubes in ice-water, transferring 280 pl of the preparation in cold polyethylene microfuge tubes and centrifuging it at 12,000 g in a Beckman microfuge for 5 min at 4 °C. Supernatants were then aspirated, tube tips were cut off and counted for radioactivity in a y-counter. Specific binding was defined as the difference in radioactivity bound in the absence and presence of 2.4 /~M rGRF(1-29)NH 2 and represented 55-70% of total binding. Preliminary experiments using rapid filtration (Wattman GF/C or GF/B filters) to terminate the reaction resulted in an apparent specific binding in the absence of pituitary homogenate, even when the filters were presoaked with 1% BSA, 0.01 or 0.1% polyethyleneimine, 0.1% poly-L-lysine or 0.01% GRF (data not shown). Some experiments were performed with tissue protein concentrations ranging from 15 to 170/~g in order to determine the minimal amount of protein required for a sufficient specific binding. Others were conducted with various concentrations of cations and EDTA. In kinetic experiments, incubation time varied from 5 to 180 min. In hot saturation studies, increasing concentrations of radioligand (9-350 pM) were tested with and without 2.4 pM rGRF(1-29)NH 2 in order to determine non-specific binding. In cold saturation studies, 35-50 pM radioligand was used in all tubes, with increasing
concentrations (0-100 nM) of non-radioactive [1z51-Tyr1°]hGRF(144)NH2, or with 2.4 pM of rGRF(1-29)NH 2 for non-specific binding evaluation. Finally, competition studies were performed with increasing concentrations of GRFs and other peptides (10 pM to 2.4 gM), or with guanyl nucleotides (10-100 gM).
Degradation assay of hGRF(1-44)NH2 and rGRF(1-29)NH2 in pituitary homogenates Enzymatic breakdown of hGRF(1-44)NH 2 and rGRF(1-29)NH 2 during incubation was investigated in the standard binding assay conditions by high performance-liquid chromatography (HPLC). The reaction was performed in the incubation buffer with or without 5 mM EDTA. Samples of pituitary homogenates (150/tg protein/100 /~l) were incubated with 10 ktM hGRF(1-44)NH 2 or rGRF(129)NH 2 in a final volume of 500 ktl at 23 °C for 0-180 min. The degradation reaction was stopped by the addition of 300 pl cold 50 mM phosphate buffer, pH 0.8, and immediate centrifugation at 12,000 g for 5 min at 4 °C. Parallel experiments were conducted without pituitary homogenates to control GRF's adsorption to borosilicate tubes. The residual GRF concentrations were analyzed by reverse-phase HPLC, on a ~-Bondapak C~g column (39 x 150 mm, 10-pm particles; Waters, Mississauga, Canada) using a Waters instrument equipped with an automated injector, a gradient controller and a two-wavelength (214,280 nm) UV detector coupled to a two-pen recorder. The mobile phase was composed of 0.01% trifluoroacetic acid (TFA) in acetonitrile (CH3CN) (eluent A) and of 0.01% aqueous TFA, pH 2.9, (eluent B). A linear gradient of 1.0% increase of eluent B/min, for 30 rain, was used with 20% of B as initial condition and a flow rate of 1.5 ml/min. Quantification was achieved by peak height measurement which gave identical results to surface integration in preliminary experiments. Protein dilution followed by immediate centrifugation efficiently stopped the peptide degradation. This was shown by equivalent peptide recoveries for time zero of incubation with or without membranes. Data analysis The Ligand computerized, non-linear regression curve fitting program of Munson and Rodbard 16 was used to analyze association, dissociation, competition and saturation studies. The Scatchard transforms 19 of hot and cold saturation curves were used to derive estimates of the apparent affinity constants (Ka) and binding sites capacities (Bmax). First, each curve was individually analyzed; then, all data related to a particular study were simultaneously coanalyzed. One important feature of the Ligand program is its ability to pool and fit data from multiple experiments: this allows to correct data for variable receptor capacities and improves the quality of statistical analysis when comparing different models and hypotheses TM. For each analysis, estimates of binding parameters were calculated for 3 models, whenever the fit was possible: one class of sites, two classes of sites, and one class of sites, corrected for non-specific binding. In this last fit, the non-specific binding is allowed to 'float', the model considering the second site as another form of non-specific binding which is displaced by rGRF(1-29)NHz, used to determine non-specific binding. Finally, the Ligand program determined the best fit for each analysis, by comparing residual sums of squares of experimental points to fitted curves with the F-test. The existence of a second site was statistically judged in all analyses.
RESULTS U s i n g t h e c o n d i t i o n s d e s c r i b e d in M a t e r i a l s and M e t h ods, specific [125I-Tyr1°]hGRF(1-44)NH2 b i n d i n g (ratio o f r a d i o l i g a n d specifically b o u n d o v e r t o t a l r a d i o l i g a n d a d d e d ) was l i n e a r as a f u n c t i o n o f p r o t e i n c o n c e n t r a t i o n s that r a n g e d b e t w e e n 15 and 1 7 0 / ~ g (Fig. 1). It r e p r e s e n t e d 7 5 % of the total b i n d i n g at 170 ~tg p r o t e i n , this
293 percentage slightly decreasing with decreasing amounts of protein. The best compromise between specific binding and amount of protein was obtained in our view with 70-75 /tg protein per assay, corresponding to 0.1 of pituitary equivalent, with 55-70% of total binding. In the standard conditions of our assay (35-50 pM radioligand, 70-75 /tg protein, 0.42% BSA), there was a slight adsorption of the radioligand on the incubation tubes (typically 8-10% of 125I-GRF added). This adsorption did not change with addition of unlabeled GRFs (10 pM to 2.4/~M). The kinetic of association at 23 °C showed that 125I-GRF specific binding increased with time, reached a steady-state within 30 min and remained stable for at least an additional 150 min (Fig. 2, top panel). Subsequent experiments were carried out for 60 min. The high levels of total and non-specific binding in the initial phase of the reaction led us to investigate a possible interference of BSA, through a non-specific adsorption of GRFs to this protein. Therefore, experiments were conducted with various concentrations of BSA. At 0.17, 0.42 and 0.75% BSA (n = 3, for each concentration), there was no statistical difference at equilibrium (60 min incubation) between total binding (9.6 + 0.9, 9.5 _+ 1.1, 8.7 _+ 0.9%, respectively, P > 0.05), non-specific binding (3.9 + 0.4, 3.7 + 0.5, 3.6 __+0.4% respectively, P > 0.05) and specific binding (5.7 + 0.6, 5.8 + 0.7, 5.1 + 0.5% respectively, P > 0.05) (all values being expressed as B/T, i.e. the ratio of total, non-specific or specific binding over total radioligand added). These results suggest that BSA does not interfere with GRF binding at equilibrium in our assay and enabled the use of 0.42% BSA in the incubation buffer. Stability of GRFs to proteolysis was evaluated for hGRF(1-44)NH2, a close analog to the radioligand, and for rGRF(1-29)NH 2, a GRF N-terminal fragment used to determine non-specific binding. As shown in Fig. 2 (bottom panel), hGRF(1-44)NH 2 concentration did not significantly diminish after 60 or 180
min incubation in the assay conditions, in the presence of 5 mM EDTA. The remaining concentrations were, respectively, 84 + 8% and 79 _+ 9% of those at time zero. Similar results were obtained with rGRF(1-29)NH2, with a 107 _+ 13% and 90 _+ 13% recovery, under the same experimental conditions. Omission of EDTA, a metalloprotease inhibitor, revealed the presence of GRF-sensitive protease activities, since 76 _+ 7% and 29 +_ 3% of hGRF(1-44)NH2 initial concentration was recovered after 60 and 180 min, respectively, while 53 _+ 10% and 12 + 4% of rGRF(1-29)NH 2 initial concentration remained after the same periods of time. Glass adsorption of rGRF under the same experimental conditions was 3 _+ 1% after 60 min and did not increase further, while it represented 3 _+ 9% and 11 _+ 5% for hGRF after 60 and 180 min, respectively. The reversibility of 125I-GRF binding was evaluated by adding 2.4/~M of rGRF(129)NH 2 at equilibrium (60 min) (Fig. 3). Dissociation occurred quickly at first, then slowed down and was almost complete after 120 min of additional incubation time. Association and dissociation rate constants, as estimated by the Ligand program were, respectively, 5.01 _+ 0.86 nM-l.min -1 and 8.13 + 0.29 × 10 -3 min -1. To optimize the ratio of specific to total binding, incubations were performed with various cations and
~..
0.10
lIB
0.05
0.00 z
0.10
= - - e Specific binding o ~ o Nonspecific binding
0.15
lOOJ
~--
80
Z
60
Z
0.08
8§
0.06
_~40
0.04
u_
\
"r--, • • ----
m
20
0
hGRF(1-4A,)NH2 rGRF(1-29)NH 2 5mM EDTA OmM EDTA
0
0.02
0 i
0.00
40
i
i
,
80 120 160 200 TOTALPROTEIN(b~J/oesay) Fig. 1. Effect of protein concentration on specific 125I-GRF binding. Each point represents the mean of triplicate determinations. Data shown are gathered from two independent experiments. 0
60
t
I
120
180
TIME (rain) Fig. 2. Time course of 125I-GRF association to pituitary homogehates at 23 °C (top panel, mean _+ S.E.M. of 4 independent experiments) and stability of hGRF(1-44)NH 2 and r G R F ( 1 29)NH2 in pituitary homogenate in the absence and presence of 5 mM EDTA (bottom panel, mean _+ S.E.M. of 3 experiments).
294 0.08
0o8
"~
0.04
0.02
0.00
-
~ -
-
-@
~
'
'
'
0
60
120
180
TIME (rain)
Fig. 3. Dissociation of specific 125I-GRF binding to rat pituitary homogenates. The reaction was initiated by adding 2.4 /~M unlabeled rGRF(1-29)NH 2 at equilibrium (60 rain). Each point represents the mean + S.E.M. of 4 experiments, each done in triplicate.
E D T A concentrations (Fig. 4). Removing EDTA from the incubation buffer resulted in a decrease of specific binding, likely due to GRF degradation. There was no difference between 5 and 10 mM, confirming that 5 mM is enough to prevent GRF breakdown by pituitary proteolytic enzymes. As shown in Fig. 4, optimal specific binding required the presence of Mg z+ in the incubation buffer; 5 and 10 mM were equally potent and sufficient, while 50 mM Mg 2÷ resulted in an 87% decrease in specific binding. Addition of 50 mM Na ÷ also reduced the specific binding of 125I-GRF by 60%. This effect was Na+-concentration-dependent and also occurred to a lesser extent with K ÷ and Li ÷ (data not shown). Therefore, the incubation buffer for all subsequent experiments contained 5 mM EDTA and 5 mM MgCI 2. In hot saturation experiments, 125I-GRF specifically bound increased with radioligand concentrations ranging from 9 to 350 pM. Larger concentrations could not be tested in these studies, due to an increased adsorption on incubation tubes at high radioligand concentrations, and
120 I00
~o
8o
,.o_ o
60
~
40
......
i
2O 0 0 5 0
10 5 0
5 0 0
5 10 0
5 50 0
5 5 50
EDTA (mM) MgCl2 (raM) NaCl (mM)
Fig. 4. Effect of EDTA and cations on specific 125I-GRFbinding to rat pituitary homogenates. Specific binding with 5 mM EDTA and 5 mM MgCI2 was defined as 100%. Each value is the mean + S.E.M. of 3 experiments, each done in triplicate.
to possible interferences of scavengers present in the commercial preparation of 125I-GRF, such as lactose (5%), L-cysteine hydrochloride (0.1%) and aprotinin (800 KIU/ml). Scatchard transformation of 3 separate experiments gave linear plots and fitting these experimental data with the Ligand program revealed an apparent single class of binding sites (Table I), fitting with a two-sites model being impossible. We then synthesized [127I-Tyr1°]hGRF(1-44)NH2 to perform cold saturation studies. Therefore, we were able to test a much larger range of ligand concentrations (10 pM to 100 nM). The displacement curves of 125I-GRF by its non-radioactive analog suggested that this peptide bound to two classes of sites (Fig. 5, top panel). Using computerized Scatchard analysis, the data of these competition curves were described as curvilinear plots (data not shown). When resolved in two components, the first class of sites was of high affinity and low capacity, while the second had a 500-fold lower affinity and was present in a 200-fold excess relative to the high affinity site (Table I). According to the Ligand interpretation, the two-sites model was statistically preferred to the one-site model (P < 0.001 for all 6 experiments and for coanalysis of all data) and to the one-site model corrected for non-specific binding (P < 0.05 for 4 of 6 experiments and P < 0.02 for coanalysis of all data). To further investigate the nature of this second binding site, we performed cold saturation experiments with 2000 g pituitary rehomogenized pellets instead of crude homogenates and 0.1% BSA instead of 0.42% BSA. This was meant to determine whether the adsorption of BSA or of a soluble protein from the homogenate could be responsible for the appearance of the second low affinity site. As with homogenates, 0.1 pituitary equivalent per assay tube was used, corresponding to about 40 pg protein. Binding experiments of 125I-GRF to washed membranes resulted in a decreased percentage of specific binding (defined as the ratio of specific to total binding, 55-70% with homogenate, 35-40% after washing), due to an increase of non-specific binding, the ratio of specific binding to total ligand added (B/T) being unchanged. The displacement curves of 125I-GRF by its non-radioactive congener suggested again two classes of sites (Fig. 5, bottom panel), as well as the curvilinear Scatchard transformations of the data. When fitting these data with Ligand, the two-site model was statistically preferred to the one-site model (P < 0.001 or 0.01 for all 4 experiments and P < 0.001 for coanalysis of all data) and to the one-site model corrected for non-specific binding (P < 0.05 for 2 of 4 experiments, P < 0.005 for coanalysis of all data). As shown in Table I, all saturation studies gave nearly identical parameters (K d and Bmax) for the high affinity
295 TABLE I Estimates of binding parameters (Kd = affinity and Bmax = capacity) obtained by analysis with Ligand of hot and cold saturation studies (S.S.) (a) Mean + S.E.M. of n experiments, individually analyzed by Ligand. (b) Mean ___S.E.M. as calculated by Ligand when all n experiments were coanalyzed. S.S.
Kdl (nM)
Hot (n = 3) homogenates 0.42% BSA
(a) (b) (a) (b) (a) (b)
Cold (n = 6) homogenates 0.42% BSA Cold (n = 4) 2000 g pellets 0.10% BSA
Ka2 (nM)
0.28 + 0.07 0.24 + 0.07 0.68-+0.11 0.86 + 0.15 0.35 + 0.09 0.36 _+0.12
site. H o w e v e r , the estimated precision of the parameters for the low affinity site was less good, due to the increased mathematical complexity of the two-site model 16. Despite that, the existence of the second site was judged statistically significant; moreover, gd2 and Bmax2, as determined with pituitary homogenates and 0.42% B S A or with 2000 g pellets and 0.10% B S A were not different ( P > 0.05). Some attempts to perform cold
120
,ooi ~R
v
(.9
Z
2O
Z 0 E
IM fit. (/1
,oo
:, -.:\
Bmaxl
-
590-+347 400 +_210 80 _+35 67 _+31
B,n~x2
(fmol/ pituitary)
(fmol/ mg prot)
57 + 12 78 + 21 140+22 202 + 35 96 + 24 111 _+33
76 __+16 104 + 28 187+29 269 -+ 47 240 + 60 277 + 83
(pmol/ pituitary) -
38.7-+18.7 31.5 -+ 14.4 5.4 _ 2.3 4.8 _+1.6
(pmol/ mg prot) -
51.6-+25 42.0 -+ 19.2 13.5 -+ 5.8 12.0 + 4.0
saturation experiments without B S A resulted in inconsistent values for affinities, due to a dramatic increase in G R F ' s adsorption to preparation and incubation tubes; however, these data were again best fitted with a two-site model (data not shown). Various guanyl nucleotides inhibited the specific 125IG R F binding to rat pituitary homogenates (Fig. 6). Hydrolysable guanosine 5"-triphosphate (GTP), guanosine 5"-diphosphate ( G D P ) and non-hydrolysable 5"guanylylimidodiphosphate ( G p p N H p ) , guanosine 5"-o(3-thiotriphosphate) (GTP-7-S) nucleotides were equally potent at 10/~M, whereas cyclic guanosine 5"-monophosphate (cGMP) had no effect. A concentration-effect was observed for G p p N H p and GTP-7-S, with an improved inhibitory effect on specific binding at 100/~M. Since concentrations greater than 100 a M of nucleotides have not been evaluated, a complete inhibition was not obtained. Selectivity of 125I-GRF binding was examined by the ability of several G R F analogs and unrelated peptides to compete with the radioligand. Their ICsos and relative
100
80
20 0 --1 1
I I --10
I --9 LOG
I --8 TOTAL
•
I --7
--6
(M)
Fig. 5. Competition studies of 125I-GRFspecifically bound with [127I-Tyrl°]hGRF(1-44)NH2 as competitor, with crude homogehates and 0.42% BSA (top panel) or with washed homogenates and 0.10% BSA (bottom panel). Data are plotted as percentage of radioligand specifically bound (B/T) vs log Total [I-Tyrl°]hGRF(1 44)NH 2 (radioactive and non-radioactive). Each point is the average of duplicate or triplicate tubes. The mean curves were obtained from the Ligand program, when coanalysing all curves (top panel, n = 6, and bottom panel, n = 4).
"
~
40
20 o
G'rP
lo~u
GDP GMP¢ GppNHp GppNl.,Ip GTP.yS GTPTS IO#M IO/=M lO~la tOOj~M ~O0#M 100~
Fig. 6. Inhibition of specific 125I-GRFbinding by guanyl nucleotides at various concentrations. Specific binding in the absence of nucleotides was defined as 100%. Each value represents the mean _+ S.E.M. of 3 independent experiments, each done in triplicate.
296 TABLE II
Relativeaffinities of different GRF analogs Affinities were estimated by /(?5o, concentration of peptide displacing 50% of specifically bound 125I-GRF, and was determined with Ligand program. Relative affinities (R.A.) to rGRF(1-29)NH 2 were calculated for each experiment as the ratio of 1C5oof the tested GRF peptides to the ICso of rGRF(1-29)NH 2. Values are expressed as the mean + S.E.M. of 3 experiments, each performed in triplicate.
Peptide* rGRF(1-29)NH 2 rGRF(1-43)OH hGRF(1-29)NH 2 hGRF(1-44)NH 2 [127I-TyrX°lhGRF(1-44)NH2 [desaminoTyr1,D-Ala2, AlalS]hGRF(1-29)NH2 [His~,Nle27]hGRF(1-29)NHz [AIaXS]hGRF(2-29)NH2 [Ala~5]hGRF(3-29)NH2 [desLys21]hGRF(1-29)NH2
1C5o(nM) 0.72 + 0.10 0.36 + 0.04 6.30 + 1.70 0.85 + 0.47 1.16 _+0.23
R.A. (%) 100 170 + 38 12.8 + 4.0 104 _+25 90 _+22
0.73 _+0.03 67.4 _+10.3 0.82 _+0.18 80.3 _+10.3 376.7 _+153.2 0.4 + 0.1 311.0 _+97.6 0.4 + 0.1 >2400
