- Email: [email protected]
Guanine nucleotide modulation of high affinity gonadotropin-releasing hormone receptors in rat mammary tumors
Molecular and Cellular Endocrinology, 0 1992 Elsevier Scientific Publishers
MOLCEL
109
85 (1992) 109-l 16 Ireland, Ltd. 0303-7207/92/$05.00
02749
Guanine nucleotide modulation of high affinity gonadotropin-releasing hormone receptors in rat mammary tumors Tzvia Segal-Abramson,
Judith Giat, Joseph Levy and Yoav Sharoni
Clinical Biochemistry Unit, Ben-Gurion University of the Negev and Soroka University Hospital, 84105 Beer-Sheva, Israel (Received
Key words: Mammary
cancer;
16 September
Gonadotropin-releasing
hormone
1991; accepted
receptor;
15 January
G-binding
1992)
protein
Summary We previously suggested that gonadotropin-releasing-hormone (GnRH) analogues activate the phosphoinositide pathway in rat mammary tumor membranes. In the present study we analyzed the binding of GnRH analogues to these membranes and assessed its modulation by guanine nucleotides. [‘251]Buserelin (a GnRH superagonist) binding is specific because it is displaced only by GnRH analogues. Scatchard plot analysis reveals high affinity binding sites (K, = 2.5 k 0.8 nM, B,,, = 250 _t 120 fmol/mg membrane protein) and low affinity binding sites (K, 1.1 k 0.3 PM, B,,, = 200 k 105 pmol/mg membrane protein). Guanine nucleotides increased the ED,, of [‘2511buserelin displacement, and almost completely eliminated the high affinity binding. Similar results were obtained with [‘251]D-Trp6-GnRH another GnRH superagonist. The inhibition of buserelin binding by guanine nucleotides was specific for nucleotides that interact with G-binding proteins and was dose-dependent with a maximal effect at 10 PM GTPyS. Kinetic analysis of buserelin binding revealed that the dissociation rate increased at least 4-fold in the presence of 10 PM GTPyS. These results support the hypothesis that GnRH analogues interact directly with mammary tumors and activate a G-protein-dependent transducing mechanism.
Introduction Correspondence to: Yoav Sharoni, Clinical Biochemistry Unit, Ben-Gurion University of the Negev and Soroka University Hospital, 84105 Beer-Sheva, Israel. Tel. 972-57-660349; Fax 972-57-37342. Abbreviations: GnRH, gonadotropin-releasing hormone; SB-75, [Ac-o-NaI(2)‘,D-Phe(4C1)2,D-Pa1(3)3,D-Cit6,o-Ala’o]GnRH; buserelin, [D-Ser(t-Bu)6,des-Gly’o-etylamide]-GnRH; TRH, tyrotropin-releasing hormone; ADH, antidiuretic hormone; DMBA, 7,12-dimethylbenz[a]anthracene; GTPyS, guanosine 5’-[y-thioltriphosphate; GDPPS, guanosine 5’-[pthioldiphosphate; App[NH]p, adenosine 5’-[&imido]triphosphate.
Several recent studies have suggested that the hypothalamic hormone, GnRH, has a therapeutic effect on hormone-dependent tumors such as carcinema of the prostate (Labrie et al., 1986) and breast (Klijn, 1984; Manni et al., 1986). Continuous administration of GnRH inhibits the release of gonadotropins from the pituitary, causing a decline in blood estrogen levels. The growth arrest or regression of estrogen-dependent mam-
110
mary tumors has been attributed to this mechanism. However, GnRH may also exert a direct effect through a mechanism independent of gonadal suppression. This is supported by the response to GnRH analogues of postmenopausal breast cancer patients with an already low level of plasma estrogens (Plowman et al., 1986). Furthermore, GnRH analogues inhibit the growth of mammary tumor cells in culture (Eidne et al., 1987; Sharoni et al., 1989). The mechanism of transduction of the GnRH signal in the pituitary, its major target tissue, involves activation of phospholipase C which is mediated by G-binding proteins (Huckle and Conn, 1988; Naor, 1990). Supporting our working hypothesis on the direct effect of GnRH in mammary tumors, we recently demonstrated that GnRH analogues directly activate phospholipase C in DMBA-induced rat mammary tumor membranes in a GTP-dependent manner (Segal et al., 1987). In analogy with the adenylate cyclase systems (Hoffman and Lefkowitz, 1980), it is expected that guanyl nucleotides would reduce the binding of hormones that activate phospholipase C as well. Findings which show that the binding of tyrotropin-releasing hormone (TRHI to the pituitary (Sullivan et al., 1987) and antidiuretic hormone (ADH) to liver plasma membranes (Bojanic and Fain, 1986) is inhibited by guanine nucleotides lend support for the coupling of these receptors to G-binding proteins. Guanyl nucleotides also reduced GnRH binding to the pituitary receptor (Perrin et al., 1989). The presence of GnRH binding sites in human mammary cancer cells (Eidne et al., 1985) and in breast cancer specimens has been reported (Fekete et al., 1989). We were interested in determining whether guanine nucleotides also modulate the interaction of the GnRH agonist with its receptor in mammary tumors in an analogous manner to the one previously described for the pituitary receptor. For these studies we employed the DMBA-induced rat mammary tumor since large quantities of relatively homogeneous tumor tissue can be easily obtained. The rat mammary tumor is an excellent model for hormonal dependent human breast cancer, as the tumor growth rate is easily manipulated by estrogens (Levy et
al., 1981b) and other hormones (Levy et al., 1981a; Johnson et al., 1983; Sharoni et al., 1985). Here we show that in rat mammary tumors, GnRH agonists are bound to high affinity binding sites which are modulated by guanine nucleotides. Materials
and methods
Materials Buserelin was kindly provided by J. Sandow (Hoechst, Frankfurt, Germany); D-Trph-GnRH was kindly provided by Profs. Ghraf and Schroder (Ferring, Kiel, Germany); SB-75 was kindly provided by Dr. A.V. Schally (Tulane University, New Orleans, LA, USA). DMBA was purchased from Sigma (St. Louis, MO, USA), [lz51]Na from Amersham (UK), GTPyS, GDPpS and App[NH]p from Boehringer-Mannheim (Indianapolis, IN, USA). Preparation of rat mammary tumor membranes Mammary tumors induced in rats, as described previously (Sharoni et al., 19841, were excised and rapidly frozen in liquid nitrogen and stored at -70°C. The frozen tissue was diced and then homogenized at 0-4°C in a buffer containing 10 mM Tris-HCl pH 7.4, 0.1% bovine serum albumin (BSA), 100 IU/ml Trasylol, 1 mM phenylmethylsulfonyl fluoride (PMSF) and 2 pg/ml leupeptine. Nuclei were removed by centrifugation at 800 x g for 10 min and the supernatant was centrifuged at 20,000 X g for 20 min. The pellet was resuspended in the same buffer and stored at -70°C until used. Protein concentration was determined according to Bradford (1976) with bovine albumin (crystallized, Sigma, No. A-4378) as a standard. Iodination of GnRH agonists Buserelin or D-Trp6-GnRH was iodinated by the chloramine-T method, with modification according to Fekete et al. (1989). An aliquot of the peptide (5 pg in 10 ~1 0.01 N acetic acid) was mixed with 40 ~1 0.5 M phosphate buffer pH 7.4 and 1 mCi [‘*‘I]Na (10 ~1). 10 ~1 of chloramine-T (1 mg/ml) were added. After 20 s the reaction was terminated by adding 10 ~1 sodium meta-bisulphate (1 mg/ml) and 120 /*l 0.5 M phosphate buffer pH 7.4 plus 1 ~1 2.5 M KI were added.
111
The purification of the labelled peptide was carried out by high-performance liquid chromatography (HPLC) on a C-18 reverse-phase column (LKB). The specific activity of the labelled peptides was about 600 cpm/fmol, determined by displacement curves performed with increasing concentrations of the labelled peptide alone.
the absence or presence of guanine nucleotides and the incubation continued until terminated as described above. Association essays were performed by incubating membranes with [1z51]buserelin in the presence or absence of 1 PM unlabelfed busereIin and the incubation was terminated as above at various time points.
Radioligand binding assays 15 pl unlabelled peptides (various concentra-
Results
tions) solubilized in 0.005 M acetic acid, and labelled peptide (100,000 cpm) were incubated with 100 pg mammary tumor membrane suspension in a total volume of 0.5 ml homogenization buffer for 90 min at 4°C. These experimental conditions were established as the most suitable in prelimina~ studies (not shown). Incubation was terminated by the addition of 3 ml phosphate-buffered saline (PBS - 8 mM sodium phosphate, 1.5 mM potassium phosphate, 2.7 mM KCI, 137 mM NaCl, pH 7.4), containing 0.1% bovine albumin and filtrated on presoaked Whatman GF/C filters. The filters were washed 3 times with 3 ml of the same buffer and counted in an LKB crystal multidetector y-system. Total binding was determined in the absence of unlabelled peptide and nonspecific binding in the presence of LO-” M peptjde, if not otherwise indicated. In regular experiments the maximal binding for [‘2”Ilbuserelin was between 7000 and 10,000 cpm which represent on the average 6-8% of the total counts present in the tube. For [“‘l]D-Trp6-GnRH the maximal binding was between 20,000 and 30,000 cpm which represent on the average 20-30% of the total counts present in the tube. The specific binding was 50-60% of the maximal binding. The K, values and the number of binding sites were analyzed as described by Munson and Rodbard (1980) using the EBDA program for initial estimates of binding parameters and the LIGAND program for the final Scatchard analyses (Elsevier-Biosoft, Cambridge, UK).
binding of GnRH agon~ts to rat mammu9 membranes. Effect of GTPyS
Kinetic assays
Dissociation assays were performed by incubating membrane aliquots (100 pg) with [‘2511buserelin for 90 min as described above. Then 1 PM unlabelled buserelin was added in
tamo~
[‘251]Buserelin interacted with specific binding sites in mammary tumor membranes. Bound [ 125Ilbuserelin was equally displaced by the GnRH agonists buserelin and D-Trp6-GnRH and the antagonist SB-75 (Fig 1). Unrelated peptides such as TRH and ADH did not affect buserelin binding. Fig. 2A shows that 10 yM GTPyS (a nonhydrolyzable analogue of GTP) increased the ED,, of [12sIlbuserelin displacement by 20- to 40-fold.
100 80
* Buserelin
60 40 20 0 0
11
9 -
7
5
log (peptide
Fig. 1. Competitive displacement of [~*51]buserelin binding by uniabefled peptides. Mammary tumor membranes were incubated with a constant concentration of [‘251]buserelin (approx. 0.1 nM) and with increasing concentrations of the unlabelled peptides as indicated. The binding was performed as described under Material and methods. The results are presented as ‘% of maximal binding’ which refers to the specific [‘~I]buserelin binding after subtraction of the nonspecific binding as determined in the presence of 10 FM unlabelled buserelin. Each point is the average of duplicate determinations. The graph is a representative of three experiments with comparable results.
9
8
7
6
5
7
6
5
-log [unlabeled buserelin]
B. f25~j~-TrpG-G~~
9
8
-log [unlabeled D-Tr$-GnRHl Fig. 2. GTPyS causes a shift in the displacement of labelled GnRH agonist binding. A: Mammary tumor membranes were incubated with [‘251]buserelin (approx. 0.1 nM) and with increasing concentrations of the unlabelled buserelin in the absence (0) or the presence (0) of 10 FM GTPyS. The binding assays were performed as described under Material and methods and in the legend to Fig. 1. 100% binding in the absence and presence of GTPyS was 5128 cpm and 2600 cpm, respectively. B: Binding of [1251]p-Trp6-GnRH was performed as in A. 100% binding in the absence and presence of GTPyS was 9200 cpm and 4500 cpm, respectively. Each point is the average of duplicate determinations. The graph is a representative of five experiments with comparable results.
Similar results were obtained with [12s1]~-Trp6GnRH (Fig. 2%). Scatchard plot of [‘EI]buserelin binding was not linear (open circles Fig. 3). Analysis by the LIGAND program using a two-site model showed high affinity binding sites with a K, of 2.5 f 0.8 nM and a B,,, of 250 f 120 fmol/mg membrane protein, and low affini~ binding sites with a K, of 1.1 + 0.3 FM and a B,,, of 200 + 10.5 pmol/mg membrane protein. These results are the mean k SEM from five experiments. GTPyS almost completely ehminated the high affinity binding sites. In three out of five experiments, the hvo-site model of LIGAND was not valid in the presence of GTPrS, showing that there were no high affinity binding sites in this condition. The results in Fig. 3 (filled circles) are from one of the two experiments where residual high affinity binding sites were still present. The K, of these sites (4 nM) was not different from the control value without GTPyS. However, the number of
0
5
10
15
20
bound buserelin (&I) Fig. 3. Scatchard analysis of [‘251]buserelin binding to mammary tumor membranes: effect of GTPyS. Mammary tumor membranes were incubated with ~‘~l]busereljn (approx. 0.1 nM) and with increasing concentrations of the unlabelled buserelin in the absence (0) and the presence (0) of 10 PM GTP$$. The binding was performed as described under Material and methods and in the legend to Fig. 1. Results were analyzed by LIGAND, using a two-site model. For clarity, only the line of the control values for the high affinity binding sites was drawn. The graph depicts one of five experiments, see text for details.
113
60
4d-77 0
I
I
I
. I
.1
1
10
100
I .Ol
Nucleotide
1
I 1000
@Ml
Fig. 4. Dose-dependent inhibition of [‘*‘Ilo-Trp6-GnRH binding by GTPyS and ATP. Mammary tumor membranes were incubated with [1251]o-Trp6-GnRH (approx. 0.1 nM) and increasing concentrations of GTPyS (*) or ATP CO).The assays were performed as described under Material and methods. The results are expressed as the percentage of the maximal specific binding in the absence of nucleotides. 100% binding = 12,400 cpm. Each point is the average of duplicate determinations. The graph is a representative of three experiments with comparable results.
binding sites (70 fmol/mg protein) was only 22% of the control value in the same experiment. There was no significant effect on the low affinity
sites which in the presence of 10 PM GTPyS showed a K, of 1.3 rt 0.4 PM and B,,, of 300 k 200 pmol/mg membrane protein. The effect of GTPyS on [1251]buserelin binding was dose-dependent with maximal inhibition at 10 PM (Fig. 4). ATP had only a slight inhibitory effect on buserelin binding at a much higher concentration (1 mM). The high affinity binding of buserelin was inhibited equally by 10 PM GTPyS and GDPPS, whereas GMP, ATP and the nonhydrolyzable ATP analogne AppNHp had no effect on the binding at these concentrations (Table 1). Kinetic analysis of [‘251]buserelin mammary tumor membranes
binding to rat
In order to investigate the mechanism underlying the inhibitory effect of guanine nucleotides in more detail, we analyzed the association and dissociation of buserelin binding. The association and dissociation rates were measured in the presence of 1 PM unlabelled buserelin. Under these experimental conditions the high and part of the low affinity binding sites (Kd w 1 PM) are measured concomitantly. We attempted to study the
3000k
Control GTPyS
0
30
60
90
Time (min)
120
I 150
0.1-j. , . , . , . , . , . , 0
30
60
Time (min)
effect of GTPyS. Membranes were incubated Fig. 5. Association kinetics of [ ‘251]buserelin binding to mammary tumor membranes: for the indicated time with [1251jbuserelin (approx. 0.1 nM) in the absence (0) and presence (0) of 10 FM GTPyS as described under Material and methods. A: Specific binding was determined after the subtraction of the nonspecific binding in the presence vs. time (Leeb-Lundberg and Mathis, 1990): B,, is the of 1 FM unlabelled buserelin. B: The data are presented as (B,,-B)/B,, bound radioligand at equilibrium (90 min) and B is the bound radioligand at the indicated time. Each point is the average of duplicate determinations. The graph is a representative of three experiments with similar results.
114 TABLE
1
EFFECTS OF GUANINE AND ADENINE NUCLEOTIDES ON HIGH AFFFINITY [‘*‘I]BUSERELIN BINDING TO RAT TUMOR MEMBRANES Buserelin binding was determined in the presence of the indicated nucleotides. The high affinity binding was calculated by LIGAND using a two-site model. The B,,, of the high affinity binding sites in the absence of nucleotides was 280 5 85 fmol/mg membrane protein. Results are the mean + SEM of four experiments, each done in duplicate. Nucleotide (10 PM)
B,,, (% of control)
None GTPyS GDPpS GMP ATP
100 11+2 I*2 81&7 93k8 92k9
APP[NHIP
high affinity binding independently of the low affinity binding, using 10 nM buserelin, but the minor capacity of the high affinity binding sites as
Control
1 li 0
TiZe
&nB”
Fig. 6. Dissociation kinetics of [‘*‘I]buserelin from mammary tumor membranes: effect of GTPyS. Membranes were incubated for 90 min with [ ‘251]buserelin (approx. 0.1 nM). Dissociation was then initiated by adding 1 *M buserelin with (0) or without (0) 10 PM GTPyS. The results are presented as percentage of the specific binding present at the beginning of dissociation in the absence of added nucleotides (4300 cpm). The nonspecific binding at this time was determined with 1 +M unlabelled buserelin (5000 cpm). Each point is the average of duplicate determinations. The graph is a representative of three experiments with similar results.
compared to the low affinity binding at the short incubation periods did not attain definitive results (not shown). The binding of buserelin reached equilibrium in 60-90 min of incubation in the presence and in the absence of GTPyS. This nucleotide decreased the specific binding of [‘251]buserelin at all time points (Fig. 5A). Analysis of these results according to Leeb-Lundberg and Mathis (1990) (Fig. 5B) reveals no apparent effect of GTPyS on the association rate of [‘251]buserelin to the binding sites. However, the results did not permit conclusive estimation of the association constants by LIGAND. Fig. 6 shows the dissociation of [‘251]buserelin. As can be seen from the dissociation slopes, in the presence of 10 PM GTPyS, [ 12SI]buserelin dissociates from its binding sites more rapidly (at least 4-fold). Discussion Within the framework of our working hypothesis of a direct effect of GnRH analogues on mammary tumors we examined the binding of GnRH-superagonists buserelin and D-Trp6GnRH to the tumor membranes and assessed their modulation by guanine nucleotides. [‘*‘I]Buserelin binds to specific binding sites on mammary tumor membranes as it is displaced by GnRH analogues but not by unrelated peptides. Both high and low affinity binding sites were found. Extrapituitary tissues were found to contain low affinity binding sites for GnRH analogues. In the placenta their affinity was reported to be 1.6 X 10d6 M. Therefore it was suggested that these sites may be activated by high local concentrations of a GnRH-like peptide (Iwashita et al., 1986). Breast cancer specimens and cell lines (Eidne et al., 1985; Manni et al., 1986) also revealed low affinity binding sites for GnRH. However, in this study we report the presence of high affinity binding sites for GnRH in rat mammary tumors with an affinity similar to that of the pituitary receptor. High affinity binding sites for GnRH were recently reported in human breast tumor biopsies (Fekete et al., 1989). The interaction of [ ‘251]buserelin with the high affinity binding sites involves a guanine nucleotide-sensitive binding state. Guanine nu-
11.5
cleotides eliminated the high affinity binding sites almost completely without affecting the low affinity binding. The small capacity of the high affinity binding sites does not allow us to determine whether guanine nucleotides convert the high affinity binding sites into low affinity ones or just inactivate them. A similar effect of GTPyS was obtained with [ *251]o-Trp6-GnRH binding. Because clinical and in vitro studies were performed with either buserelin or o-Trp’-GnRH, we examined the effect of GTPyS on the binding of both agonists. Results with native GnRH are not shown because this peptide is very unstable in the experiments with cancer specimens probably due to high protease activity. Both GTPyS and GDPpS inhibited the agonist binding. However, GMP and various adenine nucleotides had no effect. An inhibitory effect of GDPpS on hormone binding was previously described in other systems such as the chemoattractant receptor in human polymorphonuclear leukocytes (Koo et al., 19831, TRH binding to rat pituitary cells (Sullivan et al., 1987) and GnRH binding to the pituitary receptor (Perrin et al., 1989). As Perrin et al. (1989) have suggested, it is possible that crude membrane preparations contain a nucleoside triphosphate regenerating system. Furthermore, it is possible that either GTP or GDP modulate the interaction between the receptor and the G-binding protein (Perrin et al., 1989). Decrease of receptor affinity by GTPyS was reported previously for receptors coupled to adenylate cyclase, such as the glucagon receptor (Lad et al., 19771, and other receptors (Burgisser et al., 1982; Stadel et al., 1982). Recently, a similar effect of GTPyS was described for hormones that activate phospholipase C, such as the chemoattractant receptor in human polymorphonuclear leukocytes (Koo et al., 19831, the bombesin receptor in different cell types (Fischer and Schonbrunn, 19881, TRH (Aub et al., 19871, ADH (Bojanic and Fain, 1986) and GnRH in the pituitary (Perrin et al., 1989). It has been proposed that these findings are a consequence of the coupling of the receptors to G-protein. Agonist-promoted inositol phospholipid metabolism has been shown to involve G-proteins (Cockroft,
1987; Sharoni et al., 1990). There is no consensus about the type of G-protein involved in the modulation of phospholipase C activity. Both hetrotrimeric (Gilman, 1987; Shenker et al., 1991) and low molecular weight G-proteins (Lapetina and Reep, 1987) were implicated in this process. Low molecular weight G-binding proteins which are translocated by GTP from the cytosol to the membranes were identified in mammary cancer cells (Levy and King, 1990). We previously showed that GnRH agonists activate phospholipase C in rat mammary tumor membranes in a GTP-dependent manner (Segal et al., 1987). The EC,, of o-Trp6-GnRH in the presence of GTPyS was about lop9 M. Given the large binding capacity of the low affinity binding sites, it can be calculated that at 10e9 M peptide concentration, the number of occupied low affinity binding sites may be higher than the number of occupied high affinity sites. Thus, it is not possible to conclude unequivocally which binding sites are involved in the activation of phospholipase C. We here show in the same experimental model that the high affinity binding of GnRH is modulated by GTP analogues, indicating that these GnRH binding sites are linked to a G-protein and thus may be involved in the activation of phospholipase C. The kinetic analysis performed in this study revealed that GTPyS’ affects the dissociation of the agonists from their binding sites. The rate of dissociation was at least 4-fold faster in the presence of GTPyS. This result may mean that with GTPrS a larger proportion of the binding is to a low affinity binding site from which the ligand appears to dissociate more rapidly. However, the large difference between the number of the high and low affinity sites (- 10”) did not permit measurement of any change in the number of low affinity sites. Even though conclusive evidence from the kinetic results would require an assay of the high affinity sites without the interference of the low affinity sites, the results suggest that GTPyS did not affect considerably the association rate. Assuming that GTPyS affects only the high affinity binding, we can propose that the formation of a complex between the G-protein and the high affinity receptor converts the latter to a low affin-
116
ity conformation followed by a rapid dissociation of the ligand from the receptor. Therefore, it is possible that this regulatory process is involved in the termination of the GnRH analogues’ signal in rat mammary tumors. The dissociation rate of GnRH from its receptor in the pituitary was found to be affected by GTP analogues in a similar manner (Perrin et al., 1989). We have demonstrated the presence of guanine nucleotide-regulated high affinity binding sites for GnRH analogues in DMBA-induced rat mammary tumor membranes. The effects of guanine nucleotides on GnRH high affinity binding and on the stimulation of phospholipase C by GnRH agonists are both mediated by G-proteins. These results strongly support our working hypothesis of a direct interaction of GnRH analogues with mammary tumor cells. Acknowledgements
We are grateful to Drs. Sandow and Rechenberg (Hoechst, Frankfurt, Germany) for supplying buserelin, to Profs. Ghraf and Schroder (Ferring, Kiel, Germany) for supplying o-Trp6-GnRH and to Dr. A.V. Schally (Tulane University, New Orleans, LA, USA) for supplying SB-75. This work was supported by the Israeli Cancer Association and by the Israel Cancer Research Fund (New York). References Aub, D.L., Gosse, M.E. and Cote, T.E. (1987) J. Biol. Chem. 262, 9521-9528. Bojanic, D. and Fain, J.N. (1986) Biochem. J. 240, 361-365. Bradford, U. (1976) Anal. Biochem. 72, 248-255. Burgisser, E., DeLean, A. and Lefkowitz, R.J. (1982) Proc. Natl. Acad. Sci. USA 79, 1732-1736. Cockroft, S. (1987) Trends Biochem. Sci. 12, 75-78. Eidne, K.A., Flanagan, C.A. and Millar, R.P. (1985) Science 229, 989-991. Eidne, K.A., Flanagan, C.A., Harris, N.S. and Millar, R.P. (1987) J. Clin. Endocrinol. Metab. 64, 425-432. Fekete, M., Wittliff, J.L. and Schally, A.V. (1989) J. Clin. Lab. Anal. 3, 137-147.
Fischer, J.B. and Schonbrunn, A. (1988) J. Biol. Chem. 263, 2808-2816. Gilman, A.G. (1987) Annu. Rev. Biochem. 56, 615-649. Hoffman, B.B. and Lefkowitz, R.J. (1980) Annu. Rev. Pharmacol. Toxicol. 20, 581-608. Huckle, W.R. and Conn, P.M. (1988) Endocr. Rev. 9,387-395. Iwashita, M., Evans, M.I. and Catt, K.J. (1986) J. Clin. Endocrinol. Metab. 62, 127-133. Johnson, M.L., Levy, J. and Rosen, J.M. (1983) Cancer Res. 43, 2199-2209. Klijn, J.G.M. (1984) Med. Oncol. Tumor Pharmacol. 1, 123128. Koo, C., Lefkowitz, R.J. and Snyderman, R. (1983) J. Clin. Invest. 72, 748-753. Labrie, P., DuPont, A., Belanger, A., St.-Amaud, R., Giguere, M., Lacourciere, Y., Emond, J. and Monfette, G. (1986) Endocr. Rev. 7, 67-73. Lad, P.M., Welton, A.F. and Rodbell, M. (1977) J. Biol. Chem. 252, 5942-5946. Lapetina, E.G. and Reep, B.R. (1987) Proc. Natl. Acad. Sci. USA 84, 2261-2265. Leeb-Lundberg, L.M.F. and Mathis, S.A. (1990) J. Biol. Chem. 265, 9621-9627. Levy, J. and King, R.J.B. (1990) Biochem. J. 271, 223-229. Levy, J., Liel, Y., Feldman, B., Aflallo, L. and Glick, SM. (1981a) Eur. J. Cancer Clin. Oncol. 17, 1023-1026. Levy, J., Liel, Y. and Glick, S.M. (1981b) Isr. J. Med. Sci. 17, 970-975. Manni, A., Santen, R., Harvey, H., Lipton, A. and Max, D. (1986) Endocr. Rev. 7, 89-94. Munson, P.J. and Rodbard, D. (1980) Anal. Biochem. 107, 220-239. Naor, 2. (1990) Endocr. Rev. 11, 326-353. Perrin, M.H., Haas, Y., Porter, J., Rivier, J. and Vale, W. (1989) Endocrinology 124, 798-804. Plowman, P.N., Nicholson, R.I. and Walker, K.J. (1986) Br. J. Cancer 54, 903-909. Segal, T., Levy, J. and Sharoni, J. (1987) Mol. Cell. Endocrinol. 53, 239-243. Sharoni, Y., Graziani, Y., Karny, N., Feldman, B. and Levy, J. (1984) Eur. J. Cancer Clin. Oncol. 20, 277-281. Sharoni, Y., Radian, D. and Levy, J. (1985) FEBS Lett. 189, 133-136. Sharoni, Y., Bosin, E., Miinster, A., Levy, J. and Schally, A.V. (1989) Proc. Natl. Acad. Sci. USA 86, 1648-1651. Sharoni, Y., Viallet, J., Trepel, J.B. and Sausville, E.A. (1990) Cancer Res. 50, 5257-5262. Shenker, A., Goldsmith, P., Unson, C.G. and Spiegel, A.M. (1991) J. Biol. Chem. 266, 9309-9313. Stadel, J.M., DeLean, A. and Lefkowitz, J. (1982) Adv. Enzymol. 53, l-43. Sullivan, N.J., Lautens, L.L. and Tashjian, J.A.H. (1987) Mol. Endocrinol. 1, 889.
MOLCEL
109
85 (1992) 109-l 16 Ireland, Ltd. 0303-7207/92/$05.00
02749
Guanine nucleotide modulation of high affinity gonadotropin-releasing hormone receptors in rat mammary tumors Tzvia Segal-Abramson,
Judith Giat, Joseph Levy and Yoav Sharoni
Clinical Biochemistry Unit, Ben-Gurion University of the Negev and Soroka University Hospital, 84105 Beer-Sheva, Israel (Received
Key words: Mammary
cancer;
16 September
Gonadotropin-releasing
hormone
1991; accepted
receptor;
15 January
G-binding
1992)
protein
Summary We previously suggested that gonadotropin-releasing-hormone (GnRH) analogues activate the phosphoinositide pathway in rat mammary tumor membranes. In the present study we analyzed the binding of GnRH analogues to these membranes and assessed its modulation by guanine nucleotides. [‘251]Buserelin (a GnRH superagonist) binding is specific because it is displaced only by GnRH analogues. Scatchard plot analysis reveals high affinity binding sites (K, = 2.5 k 0.8 nM, B,,, = 250 _t 120 fmol/mg membrane protein) and low affinity binding sites (K, 1.1 k 0.3 PM, B,,, = 200 k 105 pmol/mg membrane protein). Guanine nucleotides increased the ED,, of [‘2511buserelin displacement, and almost completely eliminated the high affinity binding. Similar results were obtained with [‘251]D-Trp6-GnRH another GnRH superagonist. The inhibition of buserelin binding by guanine nucleotides was specific for nucleotides that interact with G-binding proteins and was dose-dependent with a maximal effect at 10 PM GTPyS. Kinetic analysis of buserelin binding revealed that the dissociation rate increased at least 4-fold in the presence of 10 PM GTPyS. These results support the hypothesis that GnRH analogues interact directly with mammary tumors and activate a G-protein-dependent transducing mechanism.
Introduction Correspondence to: Yoav Sharoni, Clinical Biochemistry Unit, Ben-Gurion University of the Negev and Soroka University Hospital, 84105 Beer-Sheva, Israel. Tel. 972-57-660349; Fax 972-57-37342. Abbreviations: GnRH, gonadotropin-releasing hormone; SB-75, [Ac-o-NaI(2)‘,D-Phe(4C1)2,D-Pa1(3)3,D-Cit6,o-Ala’o]GnRH; buserelin, [D-Ser(t-Bu)6,des-Gly’o-etylamide]-GnRH; TRH, tyrotropin-releasing hormone; ADH, antidiuretic hormone; DMBA, 7,12-dimethylbenz[a]anthracene; GTPyS, guanosine 5’-[y-thioltriphosphate; GDPPS, guanosine 5’-[pthioldiphosphate; App[NH]p, adenosine 5’-[&imido]triphosphate.
Several recent studies have suggested that the hypothalamic hormone, GnRH, has a therapeutic effect on hormone-dependent tumors such as carcinema of the prostate (Labrie et al., 1986) and breast (Klijn, 1984; Manni et al., 1986). Continuous administration of GnRH inhibits the release of gonadotropins from the pituitary, causing a decline in blood estrogen levels. The growth arrest or regression of estrogen-dependent mam-
110
mary tumors has been attributed to this mechanism. However, GnRH may also exert a direct effect through a mechanism independent of gonadal suppression. This is supported by the response to GnRH analogues of postmenopausal breast cancer patients with an already low level of plasma estrogens (Plowman et al., 1986). Furthermore, GnRH analogues inhibit the growth of mammary tumor cells in culture (Eidne et al., 1987; Sharoni et al., 1989). The mechanism of transduction of the GnRH signal in the pituitary, its major target tissue, involves activation of phospholipase C which is mediated by G-binding proteins (Huckle and Conn, 1988; Naor, 1990). Supporting our working hypothesis on the direct effect of GnRH in mammary tumors, we recently demonstrated that GnRH analogues directly activate phospholipase C in DMBA-induced rat mammary tumor membranes in a GTP-dependent manner (Segal et al., 1987). In analogy with the adenylate cyclase systems (Hoffman and Lefkowitz, 1980), it is expected that guanyl nucleotides would reduce the binding of hormones that activate phospholipase C as well. Findings which show that the binding of tyrotropin-releasing hormone (TRHI to the pituitary (Sullivan et al., 1987) and antidiuretic hormone (ADH) to liver plasma membranes (Bojanic and Fain, 1986) is inhibited by guanine nucleotides lend support for the coupling of these receptors to G-binding proteins. Guanyl nucleotides also reduced GnRH binding to the pituitary receptor (Perrin et al., 1989). The presence of GnRH binding sites in human mammary cancer cells (Eidne et al., 1985) and in breast cancer specimens has been reported (Fekete et al., 1989). We were interested in determining whether guanine nucleotides also modulate the interaction of the GnRH agonist with its receptor in mammary tumors in an analogous manner to the one previously described for the pituitary receptor. For these studies we employed the DMBA-induced rat mammary tumor since large quantities of relatively homogeneous tumor tissue can be easily obtained. The rat mammary tumor is an excellent model for hormonal dependent human breast cancer, as the tumor growth rate is easily manipulated by estrogens (Levy et
al., 1981b) and other hormones (Levy et al., 1981a; Johnson et al., 1983; Sharoni et al., 1985). Here we show that in rat mammary tumors, GnRH agonists are bound to high affinity binding sites which are modulated by guanine nucleotides. Materials
and methods
Materials Buserelin was kindly provided by J. Sandow (Hoechst, Frankfurt, Germany); D-Trph-GnRH was kindly provided by Profs. Ghraf and Schroder (Ferring, Kiel, Germany); SB-75 was kindly provided by Dr. A.V. Schally (Tulane University, New Orleans, LA, USA). DMBA was purchased from Sigma (St. Louis, MO, USA), [lz51]Na from Amersham (UK), GTPyS, GDPpS and App[NH]p from Boehringer-Mannheim (Indianapolis, IN, USA). Preparation of rat mammary tumor membranes Mammary tumors induced in rats, as described previously (Sharoni et al., 19841, were excised and rapidly frozen in liquid nitrogen and stored at -70°C. The frozen tissue was diced and then homogenized at 0-4°C in a buffer containing 10 mM Tris-HCl pH 7.4, 0.1% bovine serum albumin (BSA), 100 IU/ml Trasylol, 1 mM phenylmethylsulfonyl fluoride (PMSF) and 2 pg/ml leupeptine. Nuclei were removed by centrifugation at 800 x g for 10 min and the supernatant was centrifuged at 20,000 X g for 20 min. The pellet was resuspended in the same buffer and stored at -70°C until used. Protein concentration was determined according to Bradford (1976) with bovine albumin (crystallized, Sigma, No. A-4378) as a standard. Iodination of GnRH agonists Buserelin or D-Trp6-GnRH was iodinated by the chloramine-T method, with modification according to Fekete et al. (1989). An aliquot of the peptide (5 pg in 10 ~1 0.01 N acetic acid) was mixed with 40 ~1 0.5 M phosphate buffer pH 7.4 and 1 mCi [‘*‘I]Na (10 ~1). 10 ~1 of chloramine-T (1 mg/ml) were added. After 20 s the reaction was terminated by adding 10 ~1 sodium meta-bisulphate (1 mg/ml) and 120 /*l 0.5 M phosphate buffer pH 7.4 plus 1 ~1 2.5 M KI were added.
111
The purification of the labelled peptide was carried out by high-performance liquid chromatography (HPLC) on a C-18 reverse-phase column (LKB). The specific activity of the labelled peptides was about 600 cpm/fmol, determined by displacement curves performed with increasing concentrations of the labelled peptide alone.
the absence or presence of guanine nucleotides and the incubation continued until terminated as described above. Association essays were performed by incubating membranes with [1z51]buserelin in the presence or absence of 1 PM unlabelfed busereIin and the incubation was terminated as above at various time points.
Radioligand binding assays 15 pl unlabelled peptides (various concentra-
Results
tions) solubilized in 0.005 M acetic acid, and labelled peptide (100,000 cpm) were incubated with 100 pg mammary tumor membrane suspension in a total volume of 0.5 ml homogenization buffer for 90 min at 4°C. These experimental conditions were established as the most suitable in prelimina~ studies (not shown). Incubation was terminated by the addition of 3 ml phosphate-buffered saline (PBS - 8 mM sodium phosphate, 1.5 mM potassium phosphate, 2.7 mM KCI, 137 mM NaCl, pH 7.4), containing 0.1% bovine albumin and filtrated on presoaked Whatman GF/C filters. The filters were washed 3 times with 3 ml of the same buffer and counted in an LKB crystal multidetector y-system. Total binding was determined in the absence of unlabelled peptide and nonspecific binding in the presence of LO-” M peptjde, if not otherwise indicated. In regular experiments the maximal binding for [‘2”Ilbuserelin was between 7000 and 10,000 cpm which represent on the average 6-8% of the total counts present in the tube. For [“‘l]D-Trp6-GnRH the maximal binding was between 20,000 and 30,000 cpm which represent on the average 20-30% of the total counts present in the tube. The specific binding was 50-60% of the maximal binding. The K, values and the number of binding sites were analyzed as described by Munson and Rodbard (1980) using the EBDA program for initial estimates of binding parameters and the LIGAND program for the final Scatchard analyses (Elsevier-Biosoft, Cambridge, UK).
binding of GnRH agon~ts to rat mammu9 membranes. Effect of GTPyS
Kinetic assays
Dissociation assays were performed by incubating membrane aliquots (100 pg) with [‘2511buserelin for 90 min as described above. Then 1 PM unlabelled buserelin was added in
tamo~
[‘251]Buserelin interacted with specific binding sites in mammary tumor membranes. Bound [ 125Ilbuserelin was equally displaced by the GnRH agonists buserelin and D-Trp6-GnRH and the antagonist SB-75 (Fig 1). Unrelated peptides such as TRH and ADH did not affect buserelin binding. Fig. 2A shows that 10 yM GTPyS (a nonhydrolyzable analogue of GTP) increased the ED,, of [12sIlbuserelin displacement by 20- to 40-fold.
100 80
* Buserelin
60 40 20 0 0
11
9 -
7
5
log (peptide
Fig. 1. Competitive displacement of [~*51]buserelin binding by uniabefled peptides. Mammary tumor membranes were incubated with a constant concentration of [‘251]buserelin (approx. 0.1 nM) and with increasing concentrations of the unlabelled peptides as indicated. The binding was performed as described under Material and methods. The results are presented as ‘% of maximal binding’ which refers to the specific [‘~I]buserelin binding after subtraction of the nonspecific binding as determined in the presence of 10 FM unlabelled buserelin. Each point is the average of duplicate determinations. The graph is a representative of three experiments with comparable results.
9
8
7
6
5
7
6
5
-log [unlabeled buserelin]
B. f25~j~-TrpG-G~~
9
8
-log [unlabeled D-Tr$-GnRHl Fig. 2. GTPyS causes a shift in the displacement of labelled GnRH agonist binding. A: Mammary tumor membranes were incubated with [‘251]buserelin (approx. 0.1 nM) and with increasing concentrations of the unlabelled buserelin in the absence (0) or the presence (0) of 10 FM GTPyS. The binding assays were performed as described under Material and methods and in the legend to Fig. 1. 100% binding in the absence and presence of GTPyS was 5128 cpm and 2600 cpm, respectively. B: Binding of [1251]p-Trp6-GnRH was performed as in A. 100% binding in the absence and presence of GTPyS was 9200 cpm and 4500 cpm, respectively. Each point is the average of duplicate determinations. The graph is a representative of five experiments with comparable results.
Similar results were obtained with [12s1]~-Trp6GnRH (Fig. 2%). Scatchard plot of [‘EI]buserelin binding was not linear (open circles Fig. 3). Analysis by the LIGAND program using a two-site model showed high affinity binding sites with a K, of 2.5 f 0.8 nM and a B,,, of 250 f 120 fmol/mg membrane protein, and low affini~ binding sites with a K, of 1.1 + 0.3 FM and a B,,, of 200 + 10.5 pmol/mg membrane protein. These results are the mean k SEM from five experiments. GTPyS almost completely ehminated the high affinity binding sites. In three out of five experiments, the hvo-site model of LIGAND was not valid in the presence of GTPrS, showing that there were no high affinity binding sites in this condition. The results in Fig. 3 (filled circles) are from one of the two experiments where residual high affinity binding sites were still present. The K, of these sites (4 nM) was not different from the control value without GTPyS. However, the number of
0
5
10
15
20
bound buserelin (&I) Fig. 3. Scatchard analysis of [‘251]buserelin binding to mammary tumor membranes: effect of GTPyS. Mammary tumor membranes were incubated with ~‘~l]busereljn (approx. 0.1 nM) and with increasing concentrations of the unlabelled buserelin in the absence (0) and the presence (0) of 10 PM GTP$$. The binding was performed as described under Material and methods and in the legend to Fig. 1. Results were analyzed by LIGAND, using a two-site model. For clarity, only the line of the control values for the high affinity binding sites was drawn. The graph depicts one of five experiments, see text for details.
113
60
4d-77 0
I
I
I
. I
.1
1
10
100
I .Ol
Nucleotide
1
I 1000
@Ml
Fig. 4. Dose-dependent inhibition of [‘*‘Ilo-Trp6-GnRH binding by GTPyS and ATP. Mammary tumor membranes were incubated with [1251]o-Trp6-GnRH (approx. 0.1 nM) and increasing concentrations of GTPyS (*) or ATP CO).The assays were performed as described under Material and methods. The results are expressed as the percentage of the maximal specific binding in the absence of nucleotides. 100% binding = 12,400 cpm. Each point is the average of duplicate determinations. The graph is a representative of three experiments with comparable results.
binding sites (70 fmol/mg protein) was only 22% of the control value in the same experiment. There was no significant effect on the low affinity
sites which in the presence of 10 PM GTPyS showed a K, of 1.3 rt 0.4 PM and B,,, of 300 k 200 pmol/mg membrane protein. The effect of GTPyS on [1251]buserelin binding was dose-dependent with maximal inhibition at 10 PM (Fig. 4). ATP had only a slight inhibitory effect on buserelin binding at a much higher concentration (1 mM). The high affinity binding of buserelin was inhibited equally by 10 PM GTPyS and GDPPS, whereas GMP, ATP and the nonhydrolyzable ATP analogne AppNHp had no effect on the binding at these concentrations (Table 1). Kinetic analysis of [‘251]buserelin mammary tumor membranes
binding to rat
In order to investigate the mechanism underlying the inhibitory effect of guanine nucleotides in more detail, we analyzed the association and dissociation of buserelin binding. The association and dissociation rates were measured in the presence of 1 PM unlabelled buserelin. Under these experimental conditions the high and part of the low affinity binding sites (Kd w 1 PM) are measured concomitantly. We attempted to study the
3000k
Control GTPyS
0
30
60
90
Time (min)
120
I 150
0.1-j. , . , . , . , . , . , 0
30
60
Time (min)
effect of GTPyS. Membranes were incubated Fig. 5. Association kinetics of [ ‘251]buserelin binding to mammary tumor membranes: for the indicated time with [1251jbuserelin (approx. 0.1 nM) in the absence (0) and presence (0) of 10 FM GTPyS as described under Material and methods. A: Specific binding was determined after the subtraction of the nonspecific binding in the presence vs. time (Leeb-Lundberg and Mathis, 1990): B,, is the of 1 FM unlabelled buserelin. B: The data are presented as (B,,-B)/B,, bound radioligand at equilibrium (90 min) and B is the bound radioligand at the indicated time. Each point is the average of duplicate determinations. The graph is a representative of three experiments with similar results.
114 TABLE
1
EFFECTS OF GUANINE AND ADENINE NUCLEOTIDES ON HIGH AFFFINITY [‘*‘I]BUSERELIN BINDING TO RAT TUMOR MEMBRANES Buserelin binding was determined in the presence of the indicated nucleotides. The high affinity binding was calculated by LIGAND using a two-site model. The B,,, of the high affinity binding sites in the absence of nucleotides was 280 5 85 fmol/mg membrane protein. Results are the mean + SEM of four experiments, each done in duplicate. Nucleotide (10 PM)
B,,, (% of control)
None GTPyS GDPpS GMP ATP
100 11+2 I*2 81&7 93k8 92k9
APP[NHIP
high affinity binding independently of the low affinity binding, using 10 nM buserelin, but the minor capacity of the high affinity binding sites as
Control
1 li 0
TiZe
&nB”
Fig. 6. Dissociation kinetics of [‘*‘I]buserelin from mammary tumor membranes: effect of GTPyS. Membranes were incubated for 90 min with [ ‘251]buserelin (approx. 0.1 nM). Dissociation was then initiated by adding 1 *M buserelin with (0) or without (0) 10 PM GTPyS. The results are presented as percentage of the specific binding present at the beginning of dissociation in the absence of added nucleotides (4300 cpm). The nonspecific binding at this time was determined with 1 +M unlabelled buserelin (5000 cpm). Each point is the average of duplicate determinations. The graph is a representative of three experiments with similar results.
compared to the low affinity binding at the short incubation periods did not attain definitive results (not shown). The binding of buserelin reached equilibrium in 60-90 min of incubation in the presence and in the absence of GTPyS. This nucleotide decreased the specific binding of [‘251]buserelin at all time points (Fig. 5A). Analysis of these results according to Leeb-Lundberg and Mathis (1990) (Fig. 5B) reveals no apparent effect of GTPyS on the association rate of [‘251]buserelin to the binding sites. However, the results did not permit conclusive estimation of the association constants by LIGAND. Fig. 6 shows the dissociation of [‘251]buserelin. As can be seen from the dissociation slopes, in the presence of 10 PM GTPyS, [ 12SI]buserelin dissociates from its binding sites more rapidly (at least 4-fold). Discussion Within the framework of our working hypothesis of a direct effect of GnRH analogues on mammary tumors we examined the binding of GnRH-superagonists buserelin and D-Trp6GnRH to the tumor membranes and assessed their modulation by guanine nucleotides. [‘*‘I]Buserelin binds to specific binding sites on mammary tumor membranes as it is displaced by GnRH analogues but not by unrelated peptides. Both high and low affinity binding sites were found. Extrapituitary tissues were found to contain low affinity binding sites for GnRH analogues. In the placenta their affinity was reported to be 1.6 X 10d6 M. Therefore it was suggested that these sites may be activated by high local concentrations of a GnRH-like peptide (Iwashita et al., 1986). Breast cancer specimens and cell lines (Eidne et al., 1985; Manni et al., 1986) also revealed low affinity binding sites for GnRH. However, in this study we report the presence of high affinity binding sites for GnRH in rat mammary tumors with an affinity similar to that of the pituitary receptor. High affinity binding sites for GnRH were recently reported in human breast tumor biopsies (Fekete et al., 1989). The interaction of [ ‘251]buserelin with the high affinity binding sites involves a guanine nucleotide-sensitive binding state. Guanine nu-
11.5
cleotides eliminated the high affinity binding sites almost completely without affecting the low affinity binding. The small capacity of the high affinity binding sites does not allow us to determine whether guanine nucleotides convert the high affinity binding sites into low affinity ones or just inactivate them. A similar effect of GTPyS was obtained with [ *251]o-Trp6-GnRH binding. Because clinical and in vitro studies were performed with either buserelin or o-Trp’-GnRH, we examined the effect of GTPyS on the binding of both agonists. Results with native GnRH are not shown because this peptide is very unstable in the experiments with cancer specimens probably due to high protease activity. Both GTPyS and GDPpS inhibited the agonist binding. However, GMP and various adenine nucleotides had no effect. An inhibitory effect of GDPpS on hormone binding was previously described in other systems such as the chemoattractant receptor in human polymorphonuclear leukocytes (Koo et al., 19831, TRH binding to rat pituitary cells (Sullivan et al., 1987) and GnRH binding to the pituitary receptor (Perrin et al., 1989). As Perrin et al. (1989) have suggested, it is possible that crude membrane preparations contain a nucleoside triphosphate regenerating system. Furthermore, it is possible that either GTP or GDP modulate the interaction between the receptor and the G-binding protein (Perrin et al., 1989). Decrease of receptor affinity by GTPyS was reported previously for receptors coupled to adenylate cyclase, such as the glucagon receptor (Lad et al., 19771, and other receptors (Burgisser et al., 1982; Stadel et al., 1982). Recently, a similar effect of GTPyS was described for hormones that activate phospholipase C, such as the chemoattractant receptor in human polymorphonuclear leukocytes (Koo et al., 19831, the bombesin receptor in different cell types (Fischer and Schonbrunn, 19881, TRH (Aub et al., 19871, ADH (Bojanic and Fain, 1986) and GnRH in the pituitary (Perrin et al., 1989). It has been proposed that these findings are a consequence of the coupling of the receptors to G-protein. Agonist-promoted inositol phospholipid metabolism has been shown to involve G-proteins (Cockroft,
1987; Sharoni et al., 1990). There is no consensus about the type of G-protein involved in the modulation of phospholipase C activity. Both hetrotrimeric (Gilman, 1987; Shenker et al., 1991) and low molecular weight G-proteins (Lapetina and Reep, 1987) were implicated in this process. Low molecular weight G-binding proteins which are translocated by GTP from the cytosol to the membranes were identified in mammary cancer cells (Levy and King, 1990). We previously showed that GnRH agonists activate phospholipase C in rat mammary tumor membranes in a GTP-dependent manner (Segal et al., 1987). The EC,, of o-Trp6-GnRH in the presence of GTPyS was about lop9 M. Given the large binding capacity of the low affinity binding sites, it can be calculated that at 10e9 M peptide concentration, the number of occupied low affinity binding sites may be higher than the number of occupied high affinity sites. Thus, it is not possible to conclude unequivocally which binding sites are involved in the activation of phospholipase C. We here show in the same experimental model that the high affinity binding of GnRH is modulated by GTP analogues, indicating that these GnRH binding sites are linked to a G-protein and thus may be involved in the activation of phospholipase C. The kinetic analysis performed in this study revealed that GTPyS’ affects the dissociation of the agonists from their binding sites. The rate of dissociation was at least 4-fold faster in the presence of GTPyS. This result may mean that with GTPrS a larger proportion of the binding is to a low affinity binding site from which the ligand appears to dissociate more rapidly. However, the large difference between the number of the high and low affinity sites (- 10”) did not permit measurement of any change in the number of low affinity sites. Even though conclusive evidence from the kinetic results would require an assay of the high affinity sites without the interference of the low affinity sites, the results suggest that GTPyS did not affect considerably the association rate. Assuming that GTPyS affects only the high affinity binding, we can propose that the formation of a complex between the G-protein and the high affinity receptor converts the latter to a low affin-
116
ity conformation followed by a rapid dissociation of the ligand from the receptor. Therefore, it is possible that this regulatory process is involved in the termination of the GnRH analogues’ signal in rat mammary tumors. The dissociation rate of GnRH from its receptor in the pituitary was found to be affected by GTP analogues in a similar manner (Perrin et al., 1989). We have demonstrated the presence of guanine nucleotide-regulated high affinity binding sites for GnRH analogues in DMBA-induced rat mammary tumor membranes. The effects of guanine nucleotides on GnRH high affinity binding and on the stimulation of phospholipase C by GnRH agonists are both mediated by G-proteins. These results strongly support our working hypothesis of a direct interaction of GnRH analogues with mammary tumor cells. Acknowledgements
We are grateful to Drs. Sandow and Rechenberg (Hoechst, Frankfurt, Germany) for supplying buserelin, to Profs. Ghraf and Schroder (Ferring, Kiel, Germany) for supplying o-Trp6-GnRH and to Dr. A.V. Schally (Tulane University, New Orleans, LA, USA) for supplying SB-75. This work was supported by the Israeli Cancer Association and by the Israel Cancer Research Fund (New York). References Aub, D.L., Gosse, M.E. and Cote, T.E. (1987) J. Biol. Chem. 262, 9521-9528. Bojanic, D. and Fain, J.N. (1986) Biochem. J. 240, 361-365. Bradford, U. (1976) Anal. Biochem. 72, 248-255. Burgisser, E., DeLean, A. and Lefkowitz, R.J. (1982) Proc. Natl. Acad. Sci. USA 79, 1732-1736. Cockroft, S. (1987) Trends Biochem. Sci. 12, 75-78. Eidne, K.A., Flanagan, C.A. and Millar, R.P. (1985) Science 229, 989-991. Eidne, K.A., Flanagan, C.A., Harris, N.S. and Millar, R.P. (1987) J. Clin. Endocrinol. Metab. 64, 425-432. Fekete, M., Wittliff, J.L. and Schally, A.V. (1989) J. Clin. Lab. Anal. 3, 137-147.
Fischer, J.B. and Schonbrunn, A. (1988) J. Biol. Chem. 263, 2808-2816. Gilman, A.G. (1987) Annu. Rev. Biochem. 56, 615-649. Hoffman, B.B. and Lefkowitz, R.J. (1980) Annu. Rev. Pharmacol. Toxicol. 20, 581-608. Huckle, W.R. and Conn, P.M. (1988) Endocr. Rev. 9,387-395. Iwashita, M., Evans, M.I. and Catt, K.J. (1986) J. Clin. Endocrinol. Metab. 62, 127-133. Johnson, M.L., Levy, J. and Rosen, J.M. (1983) Cancer Res. 43, 2199-2209. Klijn, J.G.M. (1984) Med. Oncol. Tumor Pharmacol. 1, 123128. Koo, C., Lefkowitz, R.J. and Snyderman, R. (1983) J. Clin. Invest. 72, 748-753. Labrie, P., DuPont, A., Belanger, A., St.-Amaud, R., Giguere, M., Lacourciere, Y., Emond, J. and Monfette, G. (1986) Endocr. Rev. 7, 67-73. Lad, P.M., Welton, A.F. and Rodbell, M. (1977) J. Biol. Chem. 252, 5942-5946. Lapetina, E.G. and Reep, B.R. (1987) Proc. Natl. Acad. Sci. USA 84, 2261-2265. Leeb-Lundberg, L.M.F. and Mathis, S.A. (1990) J. Biol. Chem. 265, 9621-9627. Levy, J. and King, R.J.B. (1990) Biochem. J. 271, 223-229. Levy, J., Liel, Y., Feldman, B., Aflallo, L. and Glick, SM. (1981a) Eur. J. Cancer Clin. Oncol. 17, 1023-1026. Levy, J., Liel, Y. and Glick, S.M. (1981b) Isr. J. Med. Sci. 17, 970-975. Manni, A., Santen, R., Harvey, H., Lipton, A. and Max, D. (1986) Endocr. Rev. 7, 89-94. Munson, P.J. and Rodbard, D. (1980) Anal. Biochem. 107, 220-239. Naor, 2. (1990) Endocr. Rev. 11, 326-353. Perrin, M.H., Haas, Y., Porter, J., Rivier, J. and Vale, W. (1989) Endocrinology 124, 798-804. Plowman, P.N., Nicholson, R.I. and Walker, K.J. (1986) Br. J. Cancer 54, 903-909. Segal, T., Levy, J. and Sharoni, J. (1987) Mol. Cell. Endocrinol. 53, 239-243. Sharoni, Y., Graziani, Y., Karny, N., Feldman, B. and Levy, J. (1984) Eur. J. Cancer Clin. Oncol. 20, 277-281. Sharoni, Y., Radian, D. and Levy, J. (1985) FEBS Lett. 189, 133-136. Sharoni, Y., Bosin, E., Miinster, A., Levy, J. and Schally, A.V. (1989) Proc. Natl. Acad. Sci. USA 86, 1648-1651. Sharoni, Y., Viallet, J., Trepel, J.B. and Sausville, E.A. (1990) Cancer Res. 50, 5257-5262. Shenker, A., Goldsmith, P., Unson, C.G. and Spiegel, A.M. (1991) J. Biol. Chem. 266, 9309-9313. Stadel, J.M., DeLean, A. and Lefkowitz, J. (1982) Adv. Enzymol. 53, l-43. Sullivan, N.J., Lautens, L.L. and Tashjian, J.A.H. (1987) Mol. Endocrinol. 1, 889.