Opposite effects of ethanol on the activation of adenylyl cyclase in human corpus luteum membranes

~aig~I~

and Cellular En~rjn~o~,

129

43(1985) 129-136

Elsevier Scientific publishers Ireland, Ltd. MCE 01294

Opposite effects of ethanol on the activation of adenylyl cyclase in human corpus luteum membranes * F.J. Rojas and R.H. Asch ~e~f~enf

of Obstetrics and Gynecology, ‘13e Un~versi~of Texas Health Science Center at Son Antonio, San Antonio, TX 78284 (U.S.A.) (Received 17 October 1984; accepted 17 January 1985)

Kerr&:

adenylyl cyclase; human corpus luteum; ethanol; gonadotropin binding; forskoiin.

The influence of an acute exposure to ethanol on adenylyl cyclase activity in membrane fractions prepared from human corpus luteum was investigated. Ethanol up to a concentration of 5% (v/v) was without effect on basal luteal adenylyl cyclase activity, but markedly potentiated sedation of NaF and hCG in a dose-dependent marmer. In contrast, ethanol progressively inhibited forskolin stimulation at the same range of ethanol concentrations. Maximal NaF and hCG responsiveness of adenylyl cyclase activity was observed at 5% ethanol and reached values 80% and 100% higher than controls without ethanol, respectively. However, at the same ethanol concentration, forskolin-stimulated enzymatic activity was reduced by 40% relative to controls. Equilibrium binding studies involving [i2?JhCG interaction with luteal membr~es in the presence of the ~n~ntration of ethanol showing maximal hCG responsiveness indicated that ethanol slightly affected (15% increase) the hCG binding compared to controls, without any appreciable change on the Kd for the hormone. This minor effect of ethanol on gonadotropin binding sites contrasted greatly with the extent at which ethanol maximally potentiated the gonadotropin-stimulated adenylyl cyclase. GTP was found to be less effective than GMP-P(NH)P in sustaining ethanol potentiation, suggesting that ethanol is unlikely to act by i~biting GTPase activity. These data indicate that the acute effects of ethanol inhibit forsko~-s~ulat~ adenylyl cyclase at concentrations potentiating stimulatory effects of NaF and of hCG, and that the synergistic interaction of ethanol and gonadotropin stimulation of adenylyl cyclase is, at least in part, due to an increase in the functional coupling of the occupied hCG-receptor complex with the components of the enzyme system.

Our current understanding of the functioning of hormone-stimulated adenylyl cyclase involves the interaction of at least three components of the * Supportedby an Institutional Research Grant (!SO7-RRO5654) from The University of Texas Health Science Center at San Antonio. This work was presented in part at the 7th Intemational Congress of Endocrinology, l-7 July 1984, Quebec (Canada). 0303-7207/85/$03.~

system: the hormone receptor (R), the regulatory subunit (N) and the catalytic subunit (C) (Rodbell, 1980). Functions of N subunit include interaction with R, binding to guanyl nucleotide, interaction with C and possibly a GTPase activity. The binding of R to the hormone induces its coupling with N, which, in turn, promotes the association of the activated N with C, allowing the stimulation of CAMP production (Rodbell, 1980; Bimbaumer and

8 1985 Elsevier Scientific Publishers Ireland, Ltd.

130

Iyengar, 1982). It is also known that adenylyl cyclase activity is modulated not only by hormones, neurotransmitters and guanyl nucleotide, but also by a variety of different types of agents such as NaF, cholera toxin and the diterpene forskolin (Rodbell, 1980; Seamon and Daly, 1981; Bimbaumer and Iyengar, 1982). Ethanol has been suggested to induce changes in membrane properties, such as lipid composition and membrane fluidity, which can influence the function of critical membrane proteins (Chin and Goldstein, 1977; Chin et al., 1978; Houslay and Gordon, 1983; Illiano et al., 1983). In various tissues, ethanol was reported to modify the action of several peptide hormones and neurotransmitters on the membrane-bound enzyme adenylyl cyclase (Abramowitz and Bimbaumer, 1979; Uhlemann et al., 1979; Harper and Brooker, 1980; Rabin and Molinoff, 1981), as well as other membrane functions, including sodium channel function (Mullin and Hunt, 1984), calcium uptake (Harris and Hood, 1980) and the density of certain membrane-bound receptors (Gnanaprakasam et al., 1979; Ticku and Burch, 1980; Cameron and Stouffer, 1982; Chamess et al., 1983). Reports about in vitro examinations on the acute effects of ethanol on gonadotropin receptors show conflicting data. A dose-dependent increase in gonadotropin binding sites by in vitro exposure to ethanol (0.8-8%, v/v) has been reported in monkey luteal membranes (Cameron and Stouffer, 1982). However, at the same concentrations, ethanol was found not to affect gonadotropin binding sites in porcine luteal membranes (Cameron and Stouffer, 1982). In contrast, Bhalla (1978) has shown that there is a marked decrease in LH/hCG receptor sites after in vitro treatment of rat testicular homogenate with 2.55% (v/v) alcohol. Unfortunately, in all these studies it was not clear whether the ethanol-induced effects on gonadotropin receptor sites did indeed represent functional hormone receptors. On the other hand, Abramowitz and Bimbaumer (1979) have reported that incubation of membranes from rat and rabbit corpora lutea with 5% (v/v) alcohol increases gonadotropin-stimulated adenylyl cyclase activity about 2- and 7-fold, respectively. However, it was not determined whether this treatment affected gonadotropin binding capacity in these mem-

branes. Although the mechanism and physiological relevance of the activation of adenylyl cyclase by ethanol remain unresolved, it is apparent that the understanding of the involvement of ethanol in this process may enhance our knowledge of cellular regulation by hormone. The present studies were undertaken to define the influence of an acute exposure to ethanol on adenylyl cyclase activity and to investigate the correlation of these effects with the number of gonadotropin binding sites in membrane fractions prepared from human corpus luteum. Materials and methods Materials [a- 32P]ATP (20-25 Ci/mmole)

was purchased from ICN (Irvine, CA); [3H]cAMP (20-30 Ci/mmole) from Amersham (Arlington Heights, IL); “‘1 (carrier-free) from Iso-Tex Diagnostics (Friendswood, TX); creatine phosphate, myokinase, creatine phosphokinase, ATP (Na salt; catalog no. A-2383), GTP, CAMP, EDTA and Tris from Sigma Chemical Co. (St. Louis, MO); GMPP(NH)P (guanyl 5’-yl imidodiphosphate) from Boehringer-Mannheim (Mannheim, F.R.G.); and NaF from Fisher Scientific Co. (Waltham, MA). Forskolin was purchased from Calbiochem (La Jolla, CA). A highly purified hCG (14000 IU/mg) was purchased from Radioassay Systems Laboratories, Inc. (Carson, CA). All other chemicals and reagents were of the highest commercially available purity and were used without further purification. Collection of corpora lutea

Corpora lutea were obtained from the ovaries of three women, aged 23 to 36 years, undergoing exploratory laparotomies at the Medical Center Hospital, San Antonio, TX, for a variety of benign gynecological conditions. All the subjects were informed on the nature of the study and signed an informed consent approved by the Institutional Review Board. None of the women was pregnant or hormonally medicated, and the ages of the corpora lutea were between 6 and 12 days, as assessed by histology and cycle dates. Immediately after removal, the luteal tissue was placed in iced Krebs-Ringer bicarbonate buffer prepared with

131

one-half the recommended amount of CaCl, (Dawson et al., 1969) and transported to the laboratory. Preparation

of membrane particles

Washed membrane particles were prepared as described elsewhere (Rojas and Asch, 1984). Briefly, the corpora lutea were bivalved, teased from the surrounding stroma with forceps and weighed. A small fraction was excised and placed in 10% formaldehyde solution for histological examination. The remaining tissue was then minced finely and homogenized in 20 ~01s. (w/v) of 27% (w/w) sucrose, 10 mM Tris-HCl, pH 7.5, and 1 mM EDTA in a 40 ml Dounce homogenizer (Wheaton Industries, Millville, NJ), using 10 up/down strokes with the loose pestle and 10 up/down strokes with the tight pestle. The homogenate was filtered through silk screen, size 12 (G.F. Muth Co., Washington, DC) and centrifuged for 5 min at 800 x g,,. The pellet was discarded and the supernatant suspension transferred to a second centrifuge tube and recentrifuged for 45 min at 10000 X g,, in a JA-20 rotor (Beckman Instruments, Palo Alto, CA). The resulting pellet was resuspended in 20 ~01s. (with respect to original tissue weight) of homogenization medium and subjected to a second centrifugation at 10 000 X g,, for 45 min in the JA-20 rotor. The supematant was discarded and the pellet - ‘washed membrane particles’ - was resuspended in 5 ~01s. (with respect to original tissue weight) of homogenization medium with the aid of a small Dounce homogenizer, using 10 up/down strokes with the loose pestle. After aliquoting, membranes were stored at - 70°C. Assay of adenylyl cyclase

Adenylyl cyclase activity was determined by the method of Bimbaumer et al. (1976). The optimal conditions for the assay of the enzyme from human luteal membranes have been described elsewhere (Rojas and Asch, 1984). Aliquots (10 ~1) of membrane particle preparation (5-20 pg protein) were assayed in triplicate for adenylyl cyclase activity in a final volume of 50 ccl containing 2.0 mM [ (Y-~‘P]ATP ( - 50 cpm/pmole), 5.0 mM MgCl,, 1.0 mM EDTA, 1.0 mM [ ‘HIcAMP ( - 10 000 cpm), a nucleoside triphosphate-regener-

ating system consisting of 20 mM creatine phosphate, 0.2 mg/ml creatine kinase (111 U/ml), 0.1 mg/ml myokinase (1950 U/mg) and 25 mM Tris-HCl buffer, pH 7.5. Incubations were performed at 32°C for 10 min and were terminated by the addition of 0.1 ml of a solution containing 40 mM ATP, 10 mM CAMP and 1% sodium dodecyl sulfate. The [32P]cAMP formed and [ ‘HIcAMP present as recovery marker were isolated by double chromatography over Dowex 50 and alumina according to the procedure of Salomon et al. (1974) as modified by Bockaert et al. (1976), and quantified by liquid scintillation counting. The enzyme activities measured were linear with respect to time for at least 40 min and with respect to the amount of membrane protein used in the assay. In each experiment, a batch of membrane particles from a single corpus luteum was used. 3-6 determinations of the same tissue were assayed. In order to test tissue variability, the experiments were repeated with at least one batch of membranes from a different corpus luteum. All experiments showed a similar pattern of results. Because of variations in absolute values among experiments, results from different experiments were not averaged. Thus, the figures and table present data obtained in single representative experiments. Inter-experimental variability in absolute values among separate batches of tissues was 40-60%. [‘2SI]hCG

binding assays

Assays of [‘251]hCG binding were performed using approximately 10 pg membrane protein in the presence of 25 mM Tris-HCl, pH 7.5, 1 mM EDTA and 0.1% bovine serum albumin. Final incubation volume was 100 ~1. The incubations were carried out for 90 min at 32’C. Bound labeled hormone was separated from free by polyethylene glycol (PEG; Carbowax 8000) as described by Iyengar et al. (1980a). Briefly, 1.5 ml ice-cold bovine y-globulin (2 mg/ml) in 100 mM NaCl, followed by 0.5 ml ice-cold 20% PEG, were added to the reaction mixtures. The tubes were kept on ice for 10 min, centrifuged, and the supematant discarded by aspiration. The precipitates were resuspended by the addition of 1.5 ml 100 mM NaCl and reprecipitated by the further addition of 0.5 ml 20% PEG. After centrifugation and aspiration

132

of the supematants, the tubes were counted to determine the amount of [‘251]hCG bound to the precipitates. Nonspecific binding was measured in the presence of an excess of hCG (100 IU; Ayerst Laboratories, New York, NY). Iodination of hCG was carried out by the lactoperoxidase procedure as described by Abramowitz et al. (1982). Unreacted “‘1 was separated from [‘251]hCG by gel filtration on Sephadex G-25. The specific activity of the [‘251]hCG ranged from 45 to 49 &i/pg. Other procedures To minimize contamination with ‘guanyl nucleotide-like’ material, the components of the nucleoside triphosphate-regenerating system were subjected to purification. To accomplish this, creatine phosphokinase and myokinase were passed over Sephadex G-25 and creatine phosphate was mixed with activated charcoal (Norit A) at 0-4’C according to the procedures described by Iyengar et al. (1980b). Forskolin was added as a dilution in distilled water of a stock solution of 20 mM dissolved in absolute ethanol. Final concentration of ethanol in the assay was 0.5%. Proteins were measured by the method of Lowry et al. (1951), using bovine serum albumin as standard. Comparison between means to test statistical differences was performed by analysis of variance, followed by Duncan’s multiple range test (Snedecor and Cochrane, 1967). The results are expressed as the mean &-SEM. RCdtS

Activation of human luteal adenylyl cyclase by hCG, NaF and forskolin in the absence or presence of a saturating concentration of GMPP(NH)P is shown in Table 1. The enzyme activity was stimulated by the addition of guanyl nucleotide in the absence of hormonal stimuli. In the presence of hCG, the inclusion of guanyl nucleotide markedly enhanced the hormonal stimulation of the absolute enzyme activity. Forskolin was a potent activator of the luteal adenylyl cyclase activity without addition of guanyl nucleotide; this stimulation by forskolin, however, was markedly potentiated by a maximally effective concentration of GMP-P(NH)P. The activity measured when forskolin and NaF were added simultaneously was

TABLE 1 EFFECTS OF hCG, NaF AND FORSKOLIN ON ADENYLYL CYCLASE ACTIVITY FROM HUMAN CORPUS LUTEUM MEMBRANES IN THE ABSENCE OR PRESENCE OF GMP-P(NH)P Additions

None hCG NaF Forskolin Forskolin+

’

Adenylyl cyclase activity b (pmoles/min/mg)

NaF

No GMP-P(NH)P

100 CM GMP-P(NH)P

8.2 26.2 130.5 214.0 556.9

63.6f 3.2 ’ (6) 174.7* 4.5 ’ (6) 122.5* 4.9 d (3) 460.3 f 17.2 = (3) 455.7* 11.8 = (3)

f 0.6 f 2.8 f 1.9 St 14.3 f 21.8

(6) (6) (6) (6) (3)

When present, hCG was 10 &ml, GMP-P(NH)P and forskolin were 100 pM, and NaF was 10 mM. Adenylyl cyclase activity was assayed for 10 min at 32’C as described in Materials and Methods in the absence or presence of 100 PM GMP-P(NH)P. Values are the mean f SEM; values in parentheses represent number of determinations. Significantly different by at least P-C 0.05 from the value observed with the respective group in the absence of added GMP-P(NH)P. Not significantly different from the value observed with the respective group in the absence of added GMP-P(NH)P.

higher than the sum of the individual effects (i.e. synergistic). Addition of GMP-P(NH)P, however, produced a significant decrease (about 20%; P < 0.05) in the activity observed with forskolin plus NaF. This negative effect by guanyl nucleotide might imply the presence of an inhibitory guanyl nucleotide regulatory protein in the human luteal membranes, similar to that described in other membrane systems (Rodbell, 1980; Cooper, 1982; Hildebrandt et al., 1982). This possibility would agree with previous data from our laboratory indicating that the human luteal adenylyl cyclase is under positive and negative regulation by guanyl nucleotide (Rojas and Asch, 1984). The effects of increasing concentrations of ethanol on the human luteal adenylyl cyclase activity stimulated by NaF, forskolin and hCG are shown in Fig. 1. To achieve maximal gonadotropin responsiveness, assays were carried out under optimal conditions, i.e. in the presence of maximally effective concentrations of both hCG and GMPP(NH)P (Rojas and Asch, 1984; see also Table 1). Ethanol at O-5% (v/v) was without significant effect on basal enzymatic activity, but markedly

133

hCG+~P-P(N~)P

A

02.55

1020

0 2.5 5 IO20

Percent Ethanol (vol/vol)

Fig. 1. Effects of increasing concentrations of ethanol on basal, hCG-, NaF- or forskolin-stimulated adenylyl cyclase activity from human luteal membranes. Gonadotropin responsiveness was determined under optimal conditions, i.e. in the presence of maximally effective ~n~nt~tions of both KG and GMPP(NH)P. Wben present, hCG was 10 &ml, GMP-P(NH)P and forskolin were 100 pM and NaF was 10 mM. Adenylyl cyclase activity was assayed for 10 min at 32V as described in Materials and Methods. Data shown are the means* SEM of triplicate determinations. Means bearing the same letters are not significantly different (P > 0.05). Comparable results were obtained in 3 separate experiments.

potentiated NaF and hCG stimulation in a dose-dependent manner. In contrast, ethanol progressively inhibited forskolin stimulation at the same range of ethanol concentration. Maximal NaF and hCG responsiveness of adenylyl cyclase activity was observed at 5% ethanol (approx. 0.85 M) and reached values 80% and 100% higher than controls without ethanol, respectively. However, at the same ethanol concentration, forskolin-stimulated enzymatic activity was reduced by 40% relative to controls. These results demonstrate that ethanol may affect the activation of the adenylyl cyclase system from the human corpus luteum in opposite ways. This concept was also supported by the observation that, at 10% ethanol, stim~ation by gonadotropin decreased as compared to the

values observed at 5% ethanol, but it was still higher than control without alcohol; in contrast, forsko~-stimulated activity was further inhibited, while the NaF stimulation was not significantly affected as compared to the values seen at 5% ethanol. Concentrations higher than 10% ethanol depressed the activity of all activators (Fig. 1). To investigate whether the potentiation of hCG-stimulable adenylyl cyclase by ethanol was related to an increase in the number of hCG binding sites on human luteal membranes, equilibrium binding studies involving [‘251JhCG interaction with the membranes in the presence of a concentration of ethanol sustaining the highest gonadotropin responsiveness were carried out. Results indicated that 5% ethanol causes a small (about 15%), although significant, increase in the hCG binding to luteal membranes compared to controls without ethanol when membranes were incubated with 1.6 nM [‘251JhCG (data not shown). This increase in binding sites was not accompanied by an appreciable change in the affinity ( Kd) of the gonadotropin binding sites for hCG (Fig. 2). Thus, Isr, values for controls as well as for the group incubated with 5% ethanol were about 0.5 nM. These data demonstrate that, although ethanol has some ability to increase gonadotropin binding sites, this increase was small and, therefore, not proportional with the extent at which ethanol maximally potentiated the gonadotropin-stimulated adenylyl cyclase. In agreement with the observations in other gonadal tissues (Bhalla and Reichert, 1974; Cameron and Stouffer, 1982), higher doses of alcohol significantly reduced hCG binding in human luteal membranes. At 20% ethanol, binding was about 60% less than controls without ethanol when using 1.6 nM [‘251JhCG (data not shown). In an attempt to explore the possibility that ethanol may increase adenylyl cyclase by inhibiting GTPase activity, the effects of ethanol on hCG-stimulated adenylyl cyclase were studied in the presence of a maximally effective concentration of GTP and the results were compared with those observed in the presence of the hydrolysisresistant GTP analog, GMP-PfNH)P (Fig. 3). With GTP, a less pronounced eth~ol-educe potentiation of gonadotropin responsiveness was detectable, being bighest at 2.5% ethanol and declining

134

-10 0

12

‘:

Q

10 20

c

30

t

IO

as the concentration of ethanol increased further. No potentiation by GTP was observed at 10% ethanol. These results contrasted greatly with those seen in the presence of GMP-P(NH)P in which gonadotropin-stimulated activity was markedly potentiated by ethanol and was still 40-50% over control at 10% ethanol (Fig. 3; see also Fig. 1B). Since an inactivation of GTPase would be expected to allow GTP to be as effective as GMPP(NH)P in sustaining ethanol potentiation, these observations suggest that ethanol is unlikely to act by inhibiting GTPase activity.

5% Ethanol

Discussion

OF

2 0.25

0.5

[‘*‘I]

0.75

1.0

1.25

I

hCG(nM)

Fig. 2. Effect of 5% ethanol on [“’ I]hCG binding to membrane fraction preparations of human corpus luteum. Aliquots of human Iuteal membranes were incubated with increasing concentrations of f’asI]hCG in the absence or presence of 5% ethanol for 90 min at 32°C as described in Materials and Methods. inset: Data analyzed by double reciprocal plots of bound [i2’I]hCG vs. free [ ‘251JhCG. The regression for curve control was y =15.4x +32.3; for curve in the presence of ethanol, y =12.7x + 25.0. The correlation coefficient for the lines was 0.99 in both cases. Values in the graph represent the mean of 3 replicates. Comparable results were obtained in 2 separate experiments.

Percent Ethaml bol/vol 1 Fig. 3. Effects of increasing concentrations of ethanol on hCG-stimulated adenylyl cyclase from human luteal membranes in the presence of GMP-P(NH)P or GTP. Aliquots of human luteal membranes were incubated with 10 &ml of hCG in the presence of 100 CM of either GMP-P(NH)P or GTP at 32°C for 10 min. The data are expressed as a percentage of the values, observed for the corresponding control without ethanol and are the mean of 3 replicates. Control values in the absence of ethanol were 240.3 f 11.1 for hCG+ GMP-P(NH)P and 77.6*4.7 for hCG +GTP. Means bearing the same letters are not significantly different (P > 0.05). Cornparable results were observed in 3 separate experiments.

In the present work we have demonstrated that an acute exposure of alcohol exerts opposite effects on stimulation of adenylyl cyclase in membranes from human corpus luteum. Thus, ethanol inhibited forskolin-stimulated adenylyl cyclase at concentrations potentiating stimulatory effects of NaF and of hCG. These opposite changes on the adenylyl cyclase system were unexpected, since alcohols have been proposed to act directly on the C unit of the adenylyl cyclase system (Uhlemann et al., 1979) and, therefore, one might expect that ethanol potentiates the activation of any stimulator of the enzyme. Inhibition of forskolin-stimulated adenylyl cyclase by ethanol is not a property inherent only to human luteal membranes, since the same effect has also been reported in membranes from rat heart (Robber~ht et al., 1983) and in homogenates from bovine corpus luteum (Huang et al., 1982). In homogenates from bovine corpus luteum, however, ethanol was found to produce opposite effects on NaF-stimulated activity. Thus, while NaF-stimulated activity was potentiated by ethanol in rat cardiac membranes (Robber~ht et al., 1983), in agreement with our data using human luteal membranes, this activity was markedly inhibited by ethanol in homogenates from bovine corpus luteum (Huang et al., 1982). These observations would indicate that ethanol does not cause the same effects on adenylyl cyclase activity in all membrane systems. Interestingly, Wells et al. (1984) have recently reported that ethanol has disparate effects upon gonadotropin binding sites in particulate membrane preparations of the

135

monkey and rat corpus luteum and that these effects are related to species-dependent alterations in membrane fluidity. Collectively, these findings suggest that important species differences may exist in the receptor milieu and composition of luteal membranes. The data presented here constitute the first attempt to examine the in vitro acute effects of ethanol on gonadotropin receptors and the correlation of the effects with hormonal responsiveness in a cell-free preparation. We have found that the enhancement in gonadotropin responsiveness observed in the presence of ethanol could not be totally accounted for by an unmasking of gonadotropin receptor sites on luteal membranes, since ethanol at concentrations producing maximal hormone stimulation showed only a minor increase in hCG binding sites. Considering that gonadotropin binding, unlike other hormone receptors coupled to adenylyl cyclase, is not modified by guanyl nucleotide at the concentrations used in the present studies (LaBarbera et al., 1980; Abramowitz et al., 1982), these data strongly suggest that the synergistic interaction of ethanol and gonadotropin stimulation on adenylyl cyclase activity is, at least in part, due to an increase in the functional coupling of the occupied hCG-receptor complex with the components of the adenylyl cyclase system. In contrast to the conflicting data concerning in vitro effects of ethanol at concentrations lower than 10% there is good agreement among investigators that higher doses of alcohol inhibit either gonadotropin binding sites or adenylyl cyclase activity in most gonadal tissues (Bhalla and Reichert, 1974; Abramowitz and Birnbaumer, 1979; Cameron and Stouffer, 1982; and this report). Also, in vitro treatment with 30% (v/v) alcohol has been shown to solubilize gonadotropin binding sites from rat testicular homogenates (Bhalla and Reichert, 1974; Bhalla et al., 1976). The suggestion that ethanol does not inhibit GTPase activity in the luteal membranes agrees with the recent observations reported by Rabin and Molinoff (1983) studying adenylyl cyclase activity in mouse striatal membranes. These investigators found that pretreatment of the membranes with cholera toxin, an agent able to inhibit GTPase fully, does not significantly change the activation

of either basal or dopamine-stimulated adenylyl cyclase activities by ethanol. Taken together, these data support the concept that GTPase turnoff reaction might not be involved in the ethanolmediated effects on adenylyl cyclase activity. Although the precise site(s) of ethanol action is not yet defined, the present data have at least two implications. On the one hand, the observation that ethanol inhibits forskolin-stimulated adenylyl cyclase activity, but potentiates NaF and hCG responsiveness, suggests that forskolin might act via a mechanism that differs from that by which NaF or guanyl nucleotide affect the luteal adenylyl cyclase enzyme. This suggestion would agree with data showing that the mode of enzyme activation by forskolin appears to be complex. We found, for example, that even though the absence of a requirement for guanyl nucleotide suggests the possibility that the forskolin effect may be exerted on the C unit of the adenylyl cyclase system (Seamon and Daly, 1981), in the presence of GMP-P(NH)P the activation by forskolin was increased synergistically, suggesting that an activated N protein is able to potentiate the forskolin response (Table 1). Similar potentiation of forskolin-stimulated adenylyl cyclase activity by guanyl nucleotide has been reported in other membrane systems (Stengel et al., 1982; Ho and Shi, 1983; Wong and Martin, 1983). This evidence has led to the hypothesis that another regulatory component of adenylyl cyclase, different from the GTP-binding regulatory complex, may exist in the plasma membrane (Stengel et al., 1982; Bimbaumer et al., 1983; Brooker et al., 1983; Wong and Martin, 1983). In such a hypothesis, the question is raised whether the opposite effects of ethanol on forskolin response, compared with those on NaF and gonadotropin responses, involve a well-differentiated alteration by ethanol upon the activity of the different components of the adenylyl cyclase system. On the other hand, in agreement with observations reported in other adenylyl cyclase systems (Huang et al., 1982; Robberecht et al., 1983), a practical implication derived from the present data is the solvent effect in studies involving forskolin activation. In most studies, ethanol is the solvent of choice to prepare forskolin solutions and to ensure solubility in the assay. Therefore, a proper

136

design in such experiments, in order to minimize the inhibitory effects upon forskolin stimulation, should consider (1) the use of low final concentrations of ethanol and (2), in the case of studies involving varied forskolin ~ncentrations, the use of a constant amount of solvent. In conclusion, the data presented here clearly show that the acute effects of ethanol cause opposite changes on the adenylyl cyclase of human luteal membranes. Since experimentai perturbation of adenylyl cyclase is one of the approaches for exploring the mechanisms involved in the regulation of enzyme activity, the disparate effects exerted by ethanol may prove to be a useful tool in differentiating the roles of the various components of the adenylyl cyclase system. Acknowledgements We wish to thank Ms. Rowena Bray for excellent technical assistance and Ms. Gretta Small and Ms. Rim Francis for expert assistance in preparing this manuscript. References Abramowitz, J. and Bimbaumer, L. (1979) Biol. Reprod. 21, 213-217. Abramowita, J., Iyengar, R. and Birnbaumer, L. (1982) Endocrinology 110,336-346. Bhalla, V.K. (1978) Int. J. Androl. (Suppl.) 2, 354-373. Bhalla, V.K. and Reichert, L.E. (1974) J. Biol. Chem. 249, 7996-8004. Bhalla, V.K., Haskell, J., Grier, H. and Mahesh, V.B. (1976) J. Biol. Chem. 251,4947-4957. Bimbaumer, L. and Iyengar, R. (1982) In: Cyclic Nucleotides. 1. Handbook of Experimental Pharmacology, Eds.: J.A. Nathanson and J.W. Kebabian (Springer-Verlag, Berlin) pp. 153-183. Bimbaumer, L., Yang, P.-C., Hunzicker-Dunn, hf., Bockaert, J. and Duran, J.M. (1976) Endocrinology 99,163-184. Bimbaumer, L., Stengle, D., Desmier, M. and Hanoune, T. (1983) Eur. J, B&hem. 136,107-112. Bockaert, J., Hunaicker-Dunn, M. and Bimbaumer, L. (1976) J. Biol. Chem. 251,2653-2663. Brooker, G., Pedone, C. and Barovsky, K. (1983) Science 220, 1169-1170. Cameron, J.L. and Stouffer, RL. (1982) Endocrinology 110, 1451-1453. Chamess, ME., Gordon, AS. and Diamond, I. (1983) Science 222.1246-1248. Chin, J.H. and Goldstein, D.B. (1977) Science 196, 684-685.

Chin, J.H., Parsons, L.M. and Goldstein, D.B. (1978) Biochim. Biophys. Acta 513, 358-363. Cooper, D.M.F. (1982) FEBS Len. 138, 157-163. Dawson, R.M.C., Elliot, D.C., Elliot, W.H. and Jones, K.M. (1969) Data for Bi~he~c~ Research (oxford University Press, New York) p. 507. Gnanaprakasam, M.S., Chen, C.J.N., Sutherland, J.G. and Bhalla, V.K. (1979) Biol. Reprod. 20,991-1000. Harper, J.F. and Brooker, G. (1980) J. Cyclic Nucleotide Res. 6, 51-52. Harris, RA. and Hood, W.F. (1980) J. Pharmacol. Exp. Ther. 213,.562-568. Hildebrandt, J.D., Hanoune, J. and Bimbaumer, L. (1982) J. Bid. Chem. 257, 14723-14725. Ho, R.-J. and Shi, Q.-H. (1983) Mol. Cell. Endocrinol. 30, 279-290. Houslay, M.D. and Gordon, L.M. (1983) Curr. Topics Membrane Transport 18,179-231. Huang, R.-D., Smith, M.F. and Zahler, W. (1982) J. Cyclic Nucleotide Res. 8,385-394. Illiano, G., Cbiosi, E., Draetta, G.F. and Laurenza, A. (1983) Gen. Pharmacol. 14, 669-672. Iyengar, R., Abramowitz, J., Bordelon-Riser, M., Blume, A.J. and Birnbaumer, L. (1980a) J. Biol. Chem. 255, 10312-10321. Iyengar, R., Mintz, P.W., Swartz, T.L. and Bimbaumer, L. (1980b) J. Biol. Chem. 255,11875-11882. LaBarbera, A.R., Richert, N.D. and Ryan, R.J. (1980) Arch. B&hem. Biophys. 200, 177-185. Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275. Mullin, M.J. and Hunt, W.A. (1984) Life Sci. 34,287-292. Rabin, R.A. and Molinoff, P.B. (1981) J. Pharmacol. Exp. Ther. 216, 129-134. Rabin, R.A. and Molinoff, P.B. (1983) J. Pharmacol. Exp. Ther. 227, 551-556. Robberecht, P., Waelbroeck, M., Chatelain, P., Camus, J.-C. and Christophe, J. (1983) FEBS Lett. 154, 205-208. Rodbell, M. (1980) Nature (London) 284, 17-22. Rojas, F.J. and Asch, R.H. (1984) J. Clin. Endocrinol. Metab. 59,219-227. Salomon, Y., Londos, C. and Rodbell, M. (1974) Anal. Biothem. 58, 541-548. Seamon, K. and Daly, J.W. (1981) J. Biol. Chem. 256, 9799-9801. Snedecor, W.G. and Co&ran, W.G. (1967) Statistical Methods, 6th Edn. (Iowa State College Press, Ames, IA). Stengel, D., Guenet, L., Desmier, M., Insel, P. and Hanoune, T. (1982) Mol. Cell. Endocrinol. 28, 681-690. Ticku, M.K. and Burch, T. (1980) J. Neurochem. 34,417-423. Uhlemann, E.R., Robberecht, P. and Gardner, J.D. (1979) Gastroenterology 76, 917-925. Well, M.A. and Stouffer, R.L. (1984) Program for the 17th Annual Meeting of the Society for the Study of Reproduction, Laramie, WY, 23-26 July, Abstract no. 176, p. 118. Wong, S.K.-F. and Martin, B.R. (1983) B&hem. J. 216, 153-759.