Molecular Mechanisms of Adenylyl Cyclase Desensitization in Pregnant Rat Myometrium following In Vivo Administration of the ß-Adrenergic Agonist, Isoproterenol

^a Laboratoire de Physiologie de la Reproduction, CNRS URA 1449, Université P.M. Curie, 75252 Paris Cedex 05, France

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
ß-Adrenergic agonists are widely used for preterm labor treatment, but their effectiveness may be limited by desensitization. We thus investigated the effects of a ß-agonist, isoproterenol, on the myometrial ß-adrenergic receptor (ß-AR)/adenylyl cyclase pathway after administration in vivo to late-pregnant rats (8 mg/kg, twice-daily injections). One hour after the first injection, isoproterenol-stimulated adenylyl cyclase activity was reduced by 37%. This was associated with a rapid and transient uncoupling of the ß₂-ARs (53% reduction of high-affinity receptors). After prolonged isoproterenol treatment (76 h), adenylyl cyclase activity was desensitized not only to isoproterenol but also to guanosine triphosphate and forskolin. Such treatment induced 1) a selective decrease of ß₂-ARs as assessed by ¹²⁵I-cyanopindolol binding, which was reversed by 5'-guanylylimidodiphosphate and thus probably did not involve irreversible loss of receptors, and 2) a rapid alteration of their transcript levels. Prolonged isoproterenol treatment also led to myometrial Gi2 and Gi3 increase (44% and 70%) as assessed by Western blotting. Furthermore, pertussis toxin pretreatment of membranes abolished the decrease in isoproterenol-stimulated adenylyl cyclase activity. Thus, we demonstrated that myometrial adenylyl cyclase desensitization after ß-agonist treatment results mainly from ß₂-AR uncoupling and increase in Gi activity.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Regulation of uterine muscle contractility is critical for fetal development and ultimately for parturition. The ß-adrenergic system is one of the main factors maintaining uterine quiescence during pregnancy. Myometrial ß-adrenergic receptors (ß-ARs) are mainly of the ß₂ subtype (85%) and are coupled via a guanosine triphosphate (GTP)-regulatory protein (Gs) to adenylyl cyclase. The cAMP produced in response to ß-adrenergic stimulation induces myometrial relaxation, mainly via inhibition of the key regulatory enzyme (the myosin light-chain kinase) and via reduction of free intracellular calcium concentration (for review, see [1]). During pregnancy, progesterone enhances ß-adrenergic transduction by increasing ß₂-AR gene transcription [2] and thus ß₂-AR expression and by up-regulating Gs, resulting in a higher coupling of ß₂-ARs to adenylyl cyclase [3, 4].

Clinical investigations [5, 6] have shown that ß-adrenergic agonists are potent inhibitors of contractile activity. They are thus commonly used as tocolytic agents for the treatment of preterm labor. However, such treatment may lead to loss of myometrial responsiveness to further ß-adrenergic stimulation, a phenomenon referred to as homologous desensitization. ß-Adrenergic desensitization involves the contribution of a combination of events, well characterized in cultured cell lines (see [7–9] for reviews). One of these mechanisms is rapid uncoupling of the receptor from G-protein interaction, which is shown by the disappearance of the high-affinity state of the receptor. Functional uncoupling of ß-ARs is triggered by receptor phosphorylation via G-protein-coupled receptor kinases and second messenger kinases. Another slower mechanism is the agonist-induced reduction of the receptor number, often termed receptor down-regulation. The attenuation of signal transduction may also result from regulations at the postreceptor level, i.e., G-protein levels and/or function. Desensitization is then characterized by a decreased responsiveness of adenylyl cyclase, not only to the desensitizing agent but also to multiple classes of activators, and is referred to as heterologous desensitization. It has been reported in several studies that down-regulation of stimulatory G-proteins [10] as well as up-regulation of inhibitory G-proteins [11–13] underlies such desensitization.

Only a few studies have analyzed the molecular basis of ß-AR desensitization in rat myometrium [14-16], and most of these were conducted in vitro. Moreover, no in vivo studies have been performed in pregnant females. Pregnant rat myometrium may, however, constitute an interesting animal model for investigation of the tocolytic treatment-induced alterations of the human myometrial ß-AR system.

The present study was thus undertaken to analyze the effects of the ß-agonist isoproterenol on the ß-AR transduction pathway when administered in vivo to late-pregnant rats. We first evaluated the extent of agonist-induced adenylyl cyclase desensitization. Then, to determine molecular mechanisms underlying adenylyl cyclase desensitization, we evaluated agonist-induced effects on 1) selective regulation of ß₁- and ß₂-ARs and coupling to Gs-proteins and 2) alterations of Gs- and Gi-protein expression.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Chemicals and Reagents

The [³H]cAMP assay system, ¹²⁵I-cyanopindolol (¹²⁵I-CYP, spec. act. 2200 Ci/mmol), [-³²P]dCTP (spec. act. 3000 Ci/mmol), RM/1, AS/7, and EC/2 antibodies, and Gs synthetic oligodeoxynucleotide probe were obtained from DuPont NEN Life Science Products (Les Ulis, France); cesium-trifluoroacetate and prepacked oligo(dT)-cellulose columns were from Pharmacia LKB Biotechnology Inc. (Pharmacia Biotech, Orsay, France). ICI-118,551 was from Imperial Chemical Industries (ICI Pharmaceuticals, Cheshire, UK); (-)-isoproterenol-HCl, (±)-propranolol-HCl, atenolol, forskolin, 5'-guanylylimidodiphosphate (Gpp(NH)p), pertussis toxin, N-ethylmaleimide, guanosine 5'-O-(2-thiodiphosphate) (GDPßS), and all other reagents of the highest grade commercially available were from Sigma (L'Isle d'Abeau, France).

Animals and Treatments

Sprague-Dawley rats (250–300 g) were obtained from Iffa-Credo (L'Arbresle, France). They were maintained in accordance with the guidelines for care and use of laboratory animals (NIH Guide). The females were caged with males overnight, and successful mating was determined by the presence of spermatozoa in the vaginal smear (Day 1 of pregnancy). In our breeding colony, parturition occurs between 1200 and 1900 h on Day 22 for 80% of rats.

The rats were divided into three experimental groups, which were treated from Day 18 to Day 21 of pregnancy with either NaCl 0.9% as control or with (-)-isoproterenol-HCl (8 mg/kg; Iso) or (-)-isoproterenol-HCl in combination with (±)-propranolol (40 mg/kg; P+Iso). Rats were injected i.p. with saline or drugs twice daily. Propranolol was administered 30 min before isoproterenol according to the method of Kimura et al. [17]. Animals were killed by cervical dislocation at the indicated times after the first injection or at various gestational ages: Day 18, Day 19, Day 20, and Day 21 for control animals (D18, D19, D20, and D21). Uterine horns were immediately excised; myometrial tissues were then rapidly trimmed of fat and connective tissues and scraped from adherent endometrium.

Adenylyl Cyclase Assay

Membranes were prepared from fresh myometrium, and adenylyl cyclase activity was measured as described previously [18] by incubating, for 10 min at 30°C, membrane fraction (100 µg proteins) in a medium (final volume 120 µl) containing 50 mM Tris-HCl (pH 7.4), 5 mM MgSO₄, 5 mM creatine phosphate, 12 units of creatine phosphokinase, 0.5 mM ATP, 1 mM isobutylmethylxanthine, 0.1% BSA, and the agent to be tested. Adenylyl cyclase activity was evaluated in response to 0.1 mM GTP, 0.1 mM Gpp(NH)p, or 0.1 mM forskolin in presence of 0.1 mM GTP. ß-Adrenergic stimulation was determined using 0.1 mM isoproterenol in the presence of 0.1 mM GTP [18]. To directly evaluate the activity of the catalyst, we substituted Mn²⁺ for Mg²⁺ and added GDPßS (300 µM) to further reduce the possible influence of Gs on adenylyl cyclase activity [19]. In the latter case, membranes were incubated in the buffer described above without Mg²⁺. The enzymatic reaction was stopped by a 100°C water bath (5 min) followed by centrifugation (12 000 x g, 30 min), and the cAMP concentration in the supernatant solution was radioimmunoassayed. In some experiments, purified membranes were preincubated for 30 min at 30°C with 10 µg/ml of preactivated pertussis toxin, a concentration causing full inactivation of myometrial Gi-proteins [18]. Results are expressed as pmol of cAMP/mg protein per 10 min. Protein concentration was determined by the method of Schacterle and Pollack [20], using BSA as a standard.

ß-AR Binding Assay

Radioligand binding assays were performed according to Strasser et al. [21]. Total binding was determined by incubating myometrial membranes (150 µg proteins) with 25-600 pM ¹²⁵I-CYP. Nonspecific binding was determined in the presence of 10 µM (±)-propranolol. Binding site density and affinity for ligand were determined by Scatchard plot analysis. The proportion of ß-AR subtypes was determined as previously described [2]. Density of ß₁-ARs was measured by addition to the incubation medium of 1 µM unlabeled ICI-118,551, a ß₂-selective antagonist, allowing ¹²⁵I-CYP to occupy only ß₁-AR sites. ß₂-ARs were measured by addition of 5 µM atenolol, a ß₁-selective antagonist.

Low-affinity state (R_L) and high-affinity state (R_H) ß-ARs were quantified using the method of Nerme et al. [22]. Low-affinity ß-ARs were quantified after preincubation of membranes (20 min at 30°C) in 50 mM Tris-HCl, 25 mM MgCl₂, pH 7.4 (buffer A), containing 10 µM isoproterenol and 0.5 mM N-ethylmaleimide, which stabilizes the ternary agonist-receptor-Gs complex via alkylation of Gs. The membranes were then washed twice in buffer A, and the remaining receptors, i.e., those not involved in tight agonist binding (R_L), were quantified by ¹²⁵I-CYP binding. Total ß-AR density was determined by incubating myometrial membranes with ¹²⁵I-CYP in the presence of 0.1 mM Gpp(NH)p after preincubation with isoproterenol. The density of high-affinity ß-AR was estimated as total minus low-affinity ß-AR. Both quantifications were performed on membrane aliquots prepared from the same uterine horn.

Immunoblotting of G-Protein Subunits

Myometrial membranes (100-µg proteins) were resolved by electrophoresis in 10% SDS-polyacrylamide gels, and nitrocellulose filters were then probed with RM/1, AS/7, and EC/2 antibodies (diluted 1:1000) as previously described [23]. Antibodies were raised against the carboxy-terminal decapeptide of Gs (RM/1), Gi1 and Gi2 (AS/7), and Gi3 and Go (EC/2). Antibody complexes were detected using a chemiluminescent method (ECL; Amersham Corp., Arlington Heights, IL) and quantified by scanning densitometry on a densitometer GS300 (Hoefer Scientific Instruments, San Francisco, CA). Data were obtained under conditions in which a linear relationship existed between ECL signal intensity and protein amount as illustrated for Gi2 (see Fig. 6A).

View larger version (19K): [in this window] [in a new window]   FIG. 6. Effects of isoproterenol on Gi2- and Gi3-protein levels in pregnant rat myometrium. A) Linear relationship between Gi2 immunoreactivity, as detected by densitometry, and protein concentration. Quantification of myometrial Gi2 (B) and Gi3 (C) was performed on membranes of control Day 21 (D21) and 76-h isoproterenol-treated rats (Iso 76h). Myometrial membranes were probed with AS/7 (B) and EC/2 (C) antibodies. Data represent mean ± SEM of three to four independent determinations. * p < 0.01 vs. control D21 (unpaired Student's t-test). Representative autoradiograms are shown in the upper portion of each figure.

RNA Preparation and Northern Blotting

Total RNA was extracted from myometrial tissues by the cesium-trifluoroacetate gradient method, and poly(A)⁺ RNAs were purified using oligo(dT) columns as previously described [23]. Then 15 µg total RNA (Gs blotting) or 10 µg poly(A)⁺ RNA (ß₂-AR blotting) was denatured by formaldehyde, fractionated by 1% agarose gel electrophoresis, and transferred to GeneScreen-Plus membranes (DuPont NEN) by overnight capillary blotting. The blots were prehybridized in 45% formamide, 4-strength SSC (single-strength SSC is 0.15 M sodium chloride and 0.015 M sodium citrate), 5-strength Denhardt's solution, 0.5% SDS, and 75 µg/ml denatured salmon sperm DNA at 42°C for 4 h. Hybridization was performed overnight at 42°C in the same buffer containing 10% dextran sulfate and 10⁶ cpm/ml ³²P-radiolabeled probes. The probes used were a rat heart ß₂-AR cDNA probe corresponding to the entire coding sequence of the gene and a synthetic oligodeoxynucleotide Gs probe complementary to bases encoding amino acids 194-206 as described elsewhere [2, 3]. A rat cyclophilin ³²P-radiolabeled probe was used as internal control for RNA hybridization. Size estimates of the RNA species were established by comparison with an RNA ladder. Autoradiographic bands were quantified by densitometric analysis.

Data Analysis

Results are expressed as mean ± SEM. Statistical analyses were performed using unpaired Student's t-test. Values were considered statistically different when p < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Adenylyl Cyclase Desensitization

As a first step in analyzing which elements of the ß-AR/adenylyl cyclase pathway are desensitized after repeated administration of isoproterenol, we measured adenylyl cyclase activity in response to several activators: isoproterenol acting through ß₁- and ß₂-ARs, GTP, and forskolin, bypassing the receptors. Results were compared to those obtained with Mn²⁺-GDPßS, directly activating the catalytic unit of adenylyl cyclase (Fig. 1).

View larger version (48K): [in this window] [in a new window]   FIG. 1. Effect of isoproterenol treatment on adenylyl cyclase activity in pregnant rat myometrial membranes. Pregnant rats from Day 18 to Day 21 of pregnancy were injected twice daily with isoproterenol and killed 1 h (Iso 1h) or 76 h (Iso 76h) after the beginning of the treatment. Control rats at Day 18 (D18) and Day 21 (D21) were injected with vehicle (0.9% NaCl). Adenylyl cyclase activity was assayed in response to 0.1 mM isoproterenol, 0.1 mM GTP, 0.1 mM forskolin, and 10 mM Mn2+ in presence of 300 µM GDPßS as described in Materials and Methods. Results are mean ± SEM of 4 to 20 independent determinations. * p < 0.05; ** p < 0.01 vs. respective controls (unpaired Student's t-test).

Figure 1 shows that the response elicited by the ß-agonist isoproterenol was markedly diminished by 37% at 1 h after the first injection as compared to the control value at Day 18 of pregnancy (121 ± 15 vs. 192 ± 18 pmol cAMP/mg protein per 10 min, respectively, p < 0.01). A 31% desensitization was still observed after 76-h treatment as compared to the control value at Day 21 of pregnancy (111 ± 6 vs. 160 ± 11 pmol cAMP/mg protein per 10 min, respectively, p < 0.01). There was no statistical difference in adenylyl cyclase response to isoproterenol between control D18 and D21. Enzyme basal activity was not modified by isoproterenol treatment (77 ± 7 pmol cAMP/mg protein per 10 min in 76 h isoproterenol-treated rats vs. 81 ± 7 pmol cAMP/mg protein per 10 min in control D21, data not shown). Adenylyl cyclase activity in response to GTP, which activates G-proteins, was not altered at 1 h after the first isoproterenol injection but was significantly reduced at 76 h (-22%) as compared to that in control animals (isoproterenol-treated rats, 108 ± 5 vs. control D21, 138 ± 10 pmol cAMP/mg protein per 10 min, p < 0.01). The same results were obtained with the nonhydrolyzable analogue Gpp(NH)p (data not shown). Forskolin-stimulated adenylyl cyclase activity was also decreased in myometrium of rats treated for 76 h as compared to D21 controls (289 ± 27 vs. 422 ± 55 pmol cAMP/mg protein per 10 min, p < 0.05), while not being significantly altered at 1 h after the beginning of treatment. The activity of the catalytic unit of the adenylyl cyclase, as evaluated with Mn²⁺-GDPßS stimulation, was not affected at either 1 h or 76 h after treatment beginning.

To examine the possible involvement of Gi-proteins in adenylyl cyclase desensitization, we evaluated the effect of pertussis toxin pretreatment on isoproterenol-stimulated adenylyl cyclase activity in myometrial membranes (Fig. 2). Without pertussis toxin pretreatment, adenylyl cyclase responsiveness to isoproterenol was decreased after 76 h of agonist treatment as observed in experiments illustrated in Figure 1. This decrease was completely abolished by pertussis toxin preincubation of myometrial membranes. Thus, isoproterenol-stimulated activities were similar in D21 and isoproterenol-treated membranes (232 ± 12 vs. 216 ± 16 pmol cAMP/mg protein per 10 min, respectively). Pertussis toxin pretreatment increased adenylyl cyclase responsiveness to ß-agonist in isoproterenol-treated membranes (+76%) and also in control D21 membranes, albeit to a lesser extent (+31%).

View larger version (23K): [in this window] [in a new window]   FIG. 2. Influence of pertussis toxin (PTX) treatment on isoproterenol-stimulated adenylyl cyclase activity in pregnant rat myometrial membranes. Pregnant rats from Day 18 to Day 21 of pregnancy were injected twice daily with isoproterenol and killed 76 h after the beginning of the treatment (Iso 76h, filled columns). Control rats at Day 21 (D21) were injected with saline (D21, open columns). Myometrial membranes were pretreated (+PTX) or not (-PTX) with 10 µg/ml of preactivated pertussis toxin. Adenylyl cyclase activity in response to 0.1 mM isoproterenol was assayed as described in Materials and Methods. The data represent mean ± SEM of three independent determinations.

Quantification of ß₂-ARs

To elucidate the mechanisms of adenylyl cyclase desensitization to isoproterenol stimulation, we quantified ß-AR number using the high-affinity radioligand ¹²⁵I-CYP. The number of ß-ARs was not statistically different in control D18 and D21 myometria (118 ± 18 vs. 107 ± 10 fmol/mg protein, respectively). At 10 h after treatment beginning, i.e., 6 h after the second injection, ß-AR density was significantly reduced, by 60%, in comparison with that in control D18 animals (Fig. 3). This maximal reduction was still observed at 24 h and 76 h (mean value in isoproterenol-treated rats: 47 ± 3 fmol/mg proteins vs. 107 ± 10 fmol/mg proteins in control D21, p < 0.01). The simultaneous administration of the ß-antagonist (±)-propranolol prevented the ß-AR decrease (Fig. 3), thus demonstrating a ß-AR-mediated process.

View larger version (27K): [in this window] [in a new window]   FIG. 3. Effect of isoproterenol treatment on ß-AR density in pregnant rat myometrial membranes. Pregnant rats from Day 18 to Day 21 of pregnancy were injected twice daily with isoproterenol (Iso) and killed at the indicated times after the first injection. Control rats at Day 18 (D18) and Day 21 (D21) of pregnancy were treated with vehicle (0.9% NaCl). A group of animals were injected with isoproterenol plus propranolol (P+Iso 76h). Data represent mean ± SEM of three to five independent determinations. ** p < 0.01 vs. D18 and D21 controls (unpaired Student's t-test).

Selective quantification of ß₁ and ß₂ subtypes agrees with a 85% proportion of ß₂-AR binding sites, as demonstrated in previous data from our laboratory [2]. Isoproterenol treatment selectively decreased the number of ß₂-ARs. At 76 h of isoproterenol treatment, i.e., 4 h after the last injection, a 72% decrease in ß₂-ARs was observed, while the number of ß₁-ARs remained stable over the treatment period with an average B_max of 15 ± 1 fmol/mg protein (Fig. 4). Isoproterenol administration did not change ß₂-AR affinity for ¹²⁵I-CYP (mean K_d value in isoproterenol-treated rats, 215 ± 27 vs. 177 ± 37 pM in control rats).

View larger version (22K): [in this window] [in a new window]   FIG. 4. Differential effect of isoproterenol treatment on ß1- and ß2-ARs in pregnant rat myometrial membranes. Pregnant rats from Day 18 to Day 21 of pregnancy were injected twice daily with isoproterenol (Iso) and killed at the indicated times after the first injection. Results represent mean ± SEM of three to six independent determinations. ** p < 0.01 vs. D18 and D21 controls (unpaired Student's t-test).

Quantification of ß₂-AR mRNA Levels

Analysis by Northern blotting using the rat ß₂-AR cDNA probe revealed a single product of 2.2-kilobase (kb) size, in agreement with the expected size of ß₂-AR mRNA in pregnant rat myometrium [2]. The steady-state level of ß₂-AR mRNA was then quantified by densitometric scanning and normalized using cyclophilin hybridization (Fig. 5). ß₂-AR mRNA levels increased 2.3-fold from Day 18 to Day 21 of pregnancy. Treatment with isoproterenol significantly reduced ß₂-AR mRNA density. At 4 h after the first isoproterenol injection, the level of ß₂-AR mRNA was significantly reduced, by 33%, in comparison to the level in control D18 animals. Then, irrespective of the treatment time, the ß₂-AR mRNA levels remained lower than those in the respective controls.

View larger version (28K): [in this window] [in a new window]   FIG. 5. Time course of changes in ß2-AR mRNA levels after isoproterenol treatment. Pregnant rats were injected twice daily with isoproterenol from Day 18 to Day 21 of pregnancy. Poly(A)+ RNAs were prepared from control rats (D18, 19, 20, 21 of pregnancy) and isoproterenol-treated rats as indicated in Materials and Methods. ß2-AR mRNA levels are expressed relative to control D18, taken as 100. Data represent mean ± SEM of three to six independent experiments. ß2-AR mRNA levels are lower in isoproterenol-treated rats when compared with respective controls (Iso 4h: 33 ± 10% reduction vs. D18; Iso 24h: 53 ± 1% reduction vs. D19; Iso 48h: 34 ± 1% reduction vs. D20; Iso 76h: 40 ± 7% reduction vs. D21). * p < 0.05; ** p < 0.01 vs. respective control (unpaired Student's t-test).

ß-AR Uncoupling

To determine whether uncoupling of ß-ARs may also contribute to reduced adenylyl cyclase activity in response to ß-adrenergic stimulation, we evaluated the number of receptors in the high-affinity state (R_H). As shown in Table 1, the percentage of high-affinity ß-ARs in myometrial membranes was markedly reduced at 1 h after isoproterenol treatment in comparison with the percentage for control D18 (28 ± 3% vs. 51 ± 9%, respectively, p < 0.05). Propranolol significantly antagonized the isoproterenol effect (58% of R_H). By 76 h, the receptors appeared to be recoupled, since the absolute number of coupled ß-ARs was not significantly different in myometrial membranes of control D21 and 76-h isoproterenol-treated rats (50 ± 1 fmol/mg protein vs. 57 ± 1 fmol/mg protein, respectively, Table 1). Interestingly, when ß-AR assay was carried out in the presence of Gpp(NH)p in order to evaluate total ß-ARs (Table 1), the receptor density for 76-h isoproterenol-treated rats was restored to control levels (51 ± 4 fmol/mg protein without Gpp(NH)p vs. 104 ± 12 fmol/mg protein in presence of Gpp(NH)p, p < 0.01). There was no significant difference in total ß-AR density whether binding was performed with or without Gpp(NH)p (Day 18 of pregnancy: 118 ± 8 fmol/mg proteins without Gpp(NH)p and 85 ± 11 fmol/mg proteins with Gpp(NH)p; Day 21 of pregnancy: 107 ± 10 fmol/mg proteins without Gpp(NH)p and 89 ± 4 fmol/mg proteins with Gpp(NH)p; cf Table 1 and Fig. 3).

G-proteins and mRNA Studies

To examine a possible involvement of G-protein alterations in adenylyl cyclase desensitization, we quantified myometrial Gs- and Gi-protein levels after isoproterenol treatment.

Gi antibodies each revealed one band corresponding to Gi2 (40 kDa, AS/7) and Gi3 (41 kDa, EC/2), the predominant isoforms of Gi in pregnant rat myometrium (Fig. 6, B and C). Isoproterenol treatment for 76 h significantly increased Gi2 and Gi3 levels (+44% and +70%, respectively) as compared with those for control D21.

Gs subunits were assayed by quantitative Western blotting using the RM/1 antiserum. In the pregnant rat myometrium, the antibody revealed two main bands of 42 and 47 kDa, corresponding to the small (Gs-S) and the large (Gs-L) isoforms of Gs, respectively. As illustrated in Figure 7A, Gs-S and Gs-L expression levels were similar in control D18 and D21 myometrial membranes. Levels of both subunits were not affected 1 h after isoproterenol treatment but were significantly increased at 76 h of treatment. Consequently, at this time, the total myometrial Gs (Gs-S plus Gs-L) amount revealed a 35% increase (Fig. 7A). This increase was antagonized by propranolol, demonstrating the implication of the ß-adrenergic pathway (data not shown). Using Gs oligodeoxynucleotide probe, we demonstrated that this change was not related to modifications in the Gs mRNA level, since the transcript amount (1.9 kb) was not significantly modified by isoproterenol treatment when compared with the amount for control D21 (Fig. 7B). The Gs mRNA level increased as pregnancy progressed (+43% between Day 18 and Day 21) in agreement with our previous data [3].

View larger version (21K): [in this window] [in a new window]   FIG. 7. Effect of isoproterenol on Gs-protein levels and mRNA expression in pregnant rat myometrium. A) Effect of isoproterenol treatment on immunoreactive Gs levels. Quantification of Gs was performed on myometrial membranes from controls, Day 18 and Day 21 (D18 and D21), and 76 h isoproterenol-treated rats (Iso 76h) by immunoblotting using RM/1 antibodies as described in Materials and Methods. Data represent mean ± SEM of three to six independent determinations. * p < 0.05 vs. D18 and D21 controls (unpaired Student's t-test). A representative autoradiogram is shown in the upper portion of the figure. B) Effect of isoproterenol on Gs mRNA levels. Total RNA were prepared from control rats of Day 18 (D18) and Day 21 (D21) and from isoproterenol-treated rats as indicated in Materials and Methods. Data represent mean ± SEM of three to six independent determinations.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Our results show a rapid desensitization of myometrial adenylyl cyclase to ß-agonist stimulation after in vivo isoproterenol treatment. The desensitization extent (37%) is similar to that described in other in vivo models (rat lung [21], neonatal rat liver [24]). The adenylyl cyclase activity in response to agents acting at postreceptor levels (GTP, Gpp(NH)p, and forskolin) was not significantly decreased, suggesting therefore a homologous desensitization. This is further substantiated by the fact that we demonstrated a rapid and concomitant uncoupling of ß-ARs from Gs. In the few in vivo studies in which ß-AR uncoupling was investigated, it was associated with an increased ß-AR kinase (ß-ARK) activity [24, 25]. A similar mechanism may also be involved in our model, since we identified ß-ARK protein and transcript in pregnant rat myometrium (unpublished results). Moreover, recent work has shown increases in rat myometrial ß-ARK expression at the end of pregnancy [26]. No alterations in ß-AR or Gs densities, which might have contributed to the observed decrease in adenylyl cyclase stimulability, are detected in such a short time, i.e., 1 h after the beginning of treatment.

After repeated exposure of myometrium to isoproterenol (76 h), adenylyl cyclase activity in response to ß-agonist is still desensitized. Quantification of ß-ARs reveals a progressive decrease in the apparent number of myometrial ß-ARs as assessed by ¹²⁵I-CYP binding. This phenomenon occurs later than uncoupling (10 h) and is maintained for up to 76 h of isoproterenol treatment. Furthermore, this decrease selectively involves the ß₂-AR subtype, which predominates in myometrium [2] and which displays the highest coupling efficiency to Gs/AC [27]. A decrease in ß₂-AR binding sites following ß-agonist treatment has already been reported in vitro and in vivo in other models. This decrease is usually interpreted as an actual reduction in cellular ß₂-AR content due to enhanced degradation and/or decreased synthesis of receptors. In our experimental model, however, it is worth noting that ¹²⁵I-CYP binding sites in 76-h isoproterenol-treated rats are significantly increased and are restored to control levels when receptor quantification is done in the presence of 0.1 mM Gpp(NH)p. A similar result was obtained on cyclic rat myometrium [14]. These authors indeed reported that when ß-AR assays of membranes from muscle strips, desensitized by incubation with isoproterenol, were carried out in the presence of Gpp(NH)p, receptor density was restored to near control levels. Mukherjee et al. [28] further reported that injection of frogs with isoproterenol produced a fall in the apparent number of ß-ARs in the erythrocyte membranes and that its reversal was not blocked by cycloheximide. Together, these results led the authors to suggest that changes in ß₂-AR number as assessed by ligand binding studies may not reflect actual changes in the rate of receptor turnover. Such a hypothesis is supported by the work of Wang et al. [29] using immunochemical and immunocytochemical methods. Indeed, the authors demonstrated that the decline in specific ¹²⁵I-CYP binding in isoproterenol-stimulated A431 cells was not correlated with any change in immunoreactive ß-AR levels in plasma membranes. Since Gpp(NH)p reverses the decrease in ß₂-ARs in our experiment, observed changes in the myometrial ß₂-AR number may not reflect an actual loss of ß₂-AR content but rather isoproterenol-induced conformational changes leading to inaccessibility of ß-ARs to the radioligand. An alternative hypothesis is that the apparent decrease in ß-AR number may reflect occupation of the receptor with isoproterenol, as suggested by Nishikawa et al. [30] with formoterol treatments.

While ß₂-AR concentration in myometrial membranes appears not to be altered by isoproterenol treatment, we demonstrate a reduction in ß₂-AR mRNA expression taking place rapidly, at 4 h, after the beginning of treatment. This rapid decrease in ß₂-AR transcript levels after agonist exposure is in agreement with results reported in cultured cell lines (DDT₁ MF-2 cells [31], rat heart cell line H9c2 [32]). Such experiments have led to the conclusion that the decline in mRNA levels involves a decrease in transcript stability and may result in part from the elevated cAMP level elicited by the agonist [7, 33]. A similar mechanism may be involved in the pregnant rat uterus. In the myometrial model, agonist exposure thus leads to a reduction in ß₂-AR transcripts that is not followed by a reduction in the corresponding receptors. This may be explained by the fact that the ß₂-AR half-life is quite long in animal tissues, reaching several days [34].

Our results illustrate that myometrial ß₂-AR uncoupling after isoproterenol treatment is only transitory. Indeed, at 76 h after treatment beginning, the number of coupled receptors is identical to that measured in control myometrial membranes, suggesting that uncoupling may not be a factor limiting adenylyl cyclase responsiveness to ß-agonist at this time of treatment.

Long-term isoproterenol treatment leads not only to reduced adenylyl cyclase stimulability in response to ß-agonists but also to agents acting at postreceptor levels such as GTP and forskolin, illustrating a heterologous desensitization. Heterologous desensitization may affect either the catalytic unit of adenylyl cyclase or G-protein coupling. Since the catalytic unit appears not to be altered, as assessed by Mn²⁺ stimulation, the defect leading to the observed desensitization must be located at the level of G-proteins, Gs or Gi. Arguing for this hypothesis is the observed decreased adenylyl cyclase activity in response to GTP and Gpp(NH)p, as well as to forskolin, which enhances Gs/C complexes formation [35]. In rat myocardium, desensitization was reported to be associated with decreases in Gs amounts [17]. Our model suggests a different mechanism, since Gs expression is not reduced and is even slightly increased 76 h after the beginning of isoproterenol treatment. Up-regulation of Gs-proteins is not paralleled by increases in amounts of their mRNAs. We previously discussed such discrepancy between Gs-proteins and mRNA levels during pregnancy [3]. Similar observations have been described in other mammalian tissues or cell lines [36, 37]. Translational or posttranslational modifications may obscure the relationship existing between absolute levels of Gs mRNAs and associated proteins.

Although we cannot exclude the possibility of a decrease in Gs activity that would not be correlated with Gs levels, two significant findings of the present study strongly suggest a contribution of Gi-proteins in adenylyl cyclase desensitization. Using immunoblotting studies, we first established that isoproterenol treatment increases Gi2 and Gi3 levels in myometrial membranes. Previous results from our laboratory demonstrated that this increase results from transcription rate activation of both genes [38]. Up-regulation of Gi expression after activation of the ß-adrenergic pathway was also reported in myocardium [12, 39] as well as in in vitro models [11, 13]. We also established that pretreatment of isoproterenol-treated myometrial membranes with pertussis toxin, functionally inactivating Gi-proteins, abolishes the decrease in adenylyl cyclase responsiveness to isoproterenol. It is worth noting that pertussis toxin also increased the adenylyl cyclase response to isoproterenol in D21 control animals, albeit to a lesser extent. This result confirms our previous findings of a tonic inhibition of adenylyl cyclase mediated by Gi-proteins in late-pregnant rat myometrium [18], which may rely on high levels of Gi-proteins at this stage of pregnancy [23].

Our data are not in agreement with recent work on human myometrium undergoing tocolytic therapy with fenoterol [40]. Surprisingly, the authors detected only a reduction in the ß₂-AR level; they did not report any changes in Gs- and Gi-protein levels, adenylyl cyclase activity in response to guanine nucleotides and forskolin, or ß₂-AR mRNA levels. Our results, showing that administration of isoproterenol to late-pregnant rats leads to several pronounced alterations of the ß₂-adrenergic pathway and up-regulation of Gi-proteins, are in better agreement with results from numerous other experimental models (myocardium [12, 17], lung [21, 22], liver [24]) as well as a chronic heart failure model [41]. The discrepancy between the work of Engelhardt et al. and ours is surprising, since both studies were conducted in myometrium. This may be explained by the fenoterol regimen, which may not have led to any actual desensitization. The discrepancy may also relate to differences in steroid environment in human and rat myometrium at the time experiments were performed.

In conclusion, our study illustrates that prolonged ß-agonist treatment induces marked myometrial adenylyl cyclase desensitization that results, at least partly, from two main alterations of the ß₂-AR pathway: transient ß₂-AR uncoupling and increases in functional activity of Gi-proteins.

	ACKNOWLEDGMENTS

 
The authors wish to thank M.T. Robin for manuscript illustrations.

	FOOTNOTES

 
¹ Correspondence: Joëlle Cohen-Tannoudji, Laboratoire de Physiologie de la Reproduction, CNRS URA 1449, Université P.M. Curie, 4 Place Jussieu, 75252 Paris Cedex 05, France. FAX: 33-1-44-27-26-50; jtannoud{at}snv.jussieu.fr

Accepted: February 19, 1998.

Received: October 6, 1997.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Wray S. Uterine contraction and physiological mechanisms of modulation. Am J Physiol 1993; 264:C1–C18.
2. Vivat V, Cohen-Tannoudji J, Revelli JP, Muzzin P, Giacobino JP, Maltier JP, Legrand C. Progesterone transcriptionally regulates ß₂-adrenergic receptor gene in pregnant rat myometrium. J Biol Chem 1992; 267:7975–7978.[Abstract/Free Full Text]
3. Elwardy-Mérézak J, Maltier JP, Cohen-Tannoudji J, Lécrivain JL, Vivat V, Legrand C. Pregnancy-related modifications of rat myometrial Gs proteins: ADP-ribosylation, immunoreactivity and gene expression studies. J Mol Endocrinol 1994; 13:23– 27.[Abstract/Free Full Text]
4. Cohen-Tannoudji J, Vivat V, Heilmann J, Legrand C, Maltier JP. Regulation by progesterone of the high-affinity state of myometrial ß-adrenergic receptor and of adenylate cyclase activity in the pregnant rat. J Mol Endocrinol 1991; 6:137–145.[Abstract/Free Full Text]
5. Andersson KE, Ingermarsson I, Persson CGA. Effects of terbutaline on human uterine mobility at term. Acta Obstet Gynecol Scand 1975; 54:165–172.[Medline]
6. Caritis SN, Lin LS, Taig G, Wong LK. Pharmaco-dynamics of ritodrine in pregnant women during preterm labour. Am J Obstet Gynecol 1983; 147:752–759.[Medline]
7. Collins S, Caron MG, Lefkowitz RJ. From ligand binding to gene expression: new insights into the regulation of G-protein-coupled receptors. Trends Pharmacol Sci 1992; 17:37-39.
8. Lohse MJ. Molecular mechanisms of membrane receptor desensitization. Biochim Biophys Acta 1993; 1179:171–188.[Medline]
9. Ferguson SSG, Barak LS, Zhang J, Caron MG. G-Protein-coupled receptor regulation: role of G-protein receptor kinases and arrestins. Can J Physiol Pharmacol 1996; 74:1095– 1110.[CrossRef][Medline]
10. Jena BP, Abramowitz J. Catecholamine-induced heterologous desensitization of rabbit luteal adenylyl cyclase: loss of luteinizing hormone responsiveness is associated with impaired G-protein function. Endocrinology 1989; 124:1942–1948.[Abstract/Free Full Text]
11. Reithman C, Gierschik P, Sidiropoulos D, Werdan K, Jacobs KH. Mechanism of noradrenaline-induced heterologous desensitization of adenylate cyclase stimulation in rat heart muscle cells: increase in the level of inhibitory G-protein -subunits. Eur J Pharmacol 1989; 172:211–221.[CrossRef][Medline]
12. Eschenhagen T, Mende U, Diederich M, Nose M, Schmitz W, Scholz H, Schulte am Esch J, Warnholtz A, Schäfer H. Long term ß-adrenoceptor-mediated up-regulation of G_i and G_o mRNA levels and pertussis toxin-sensitive guanine nucleotide-binding proteins in rat heart. Mol Pharmacol 1992; 42:773–783.[Abstract]
13. Hadcock JR, Ros M, Watkins DC, Malbon CC. Cross-regulation between G-protein-mediated pathways: stimulation of adenylyl cyclase increases expression of the inhibitory G-protein G_i2. J Biol Chem 1990; 265:14784–14790.[Abstract/Free Full Text]
14. Levin LC, Korenman SG, Krall JF. Agonist-dependent desensitization of myometrial ß-adrenergic catecholamine-sensitive adenylate cyclase. Biol Reprod 1980; 22:493–499.[Abstract]
15. Tanfin-Tougui Z, Do-Khac L, Harbon S. Agonist-induced desensitization of ß-adrenergic receptors in rat myometrium. FEBS Lett 1981; 135:31–37.[CrossRef][Medline]
16. Herman-Gnjidic Z, MacLusky NJ, Lye SJ. Dexamethasone partially protects the myometrium against ß-adrenergic agonist-induced desensitization in vivo in the rat. Am J Obstet Gynecol 1994; 171:1651–1659.[Medline]
17. Kimura H, Miyamoto A, Ohshika H. Down-regulation of ß-adrenoceptors and loss of Gs subunit levels in ventricular myocardium of rats treated with isoproterenol. Life Sci 1993; 53:171–176.[CrossRef]
18. Mhaouty S, Cohen-Tannoudji J, Bouet-Alard R, Limon-Boulez I, Maltier JP, Legrand C. Characteristics of the ₂/ß₂-adrenergic receptor-coupled adenylyl cyclase system in rat myometrium during pregnancy. J Biol Chem 1995; 270:11012–11016.[Abstract/Free Full Text]
19. Chang FH, Bourne HR. Dexamethasone increases adenylyl cyclase activity and expression of the -subunit of G_s in GH₃ cells. Endocrinology 1987; 121:1711–1715.[Abstract/Free Full Text]
20. Schacterle GR, Pollack RL. Simplified method for the quantitative assay of small amounts of proteins in biological materials. Anal Biochem 1973; 51:654–655.[CrossRef][Medline]
21. Strasser RH, Stiles GL, Lefkowitz RJ. Translocation and uncoupling of the ß-adrenergic receptor in rat lung after catecholamine promoted desensitization in vivo. Endocrinology 1984; 115:1392–1400.[Abstract/Free Full Text]
22. Nerme V, Abrahamsson T, Vauquelin G. Chronic isoproterenol administration causes altered beta adrenoceptor-G_s-coupling in guinea pig lung. J Pharmacol Exp Ther 1990; 252:1341–1346.[Abstract/Free Full Text]
23. Cohen-Tannoudji J, Mhaouty S, Elwardy-Mérézak J, Lécrivain JL, Robin MT, Legrand C, Maltier JP. Regulation of myometrial Gi2, Gi3 and Gq expression during pregnancy. Effects of progesterone and estradiol. Biol Reprod 1995; 53:55– 64.[Abstract]
24. Garcia-Higuera I, Mayor F. Rapid desensitization of neonatal rat liver ß-adrenergic receptor kinase. A role for ß-adrenergic receptor kinase. J Clin Invest 1994; 93:937– 943.
25. Ungerer M, Parruti G, Böhm M, Puzicha M, DeBlasi A, Erdmannn E, Lohse MJ. Expression of ß-arrestins and ß-adrenergic receptor kinases in the failing human heart. Circ Res 1994; 74:206–213.[Abstract/Free Full Text]
26. Ruzycky AL, Deloia JA. Expression of beta-adrenergic receptor kinase subtypes in the pregnant rat myometrium. Am J Obstet Gynecol 1997; 176:1077–1083.[CrossRef][Medline]
27. Levy FO, Zhu X, Kaumann AJ, Birnbaumer L. Efficacy of ß₁-adrenergic receptors is lower than that of ß₂-adrenergic receptors. Proc Natl Acad Sci USA 1993; 90:10798–10802.[Abstract/Free Full Text]
28. Mukherjee C, Caron MG, Lefkowitz RJ. Regulation of adenylate cyclase coupled ß-adrenergic receptors by ß-adrenergic catecholamines. Endocrinology 1976; 99:347–357.[Abstract/Free Full Text]
29. Wang H, Berrios M, Malbon CC. Localization of ß-adrenergic receptors in A431 cells in situ. Effects of chronic exposure to agonist. Biochem J 1989; 263:533–538.[Medline]
30. Nishikawa M, Mak JCW, Barnes PJ. Effect of short- and long-acting ß₂-adrenoceptor agonists on pulmonary ß₂-adrenoceptor expression in human lung. Eur J Pharmacol 1996; 318:123–129.[CrossRef][Medline]
31. Hadcock JR, Malbon CC. Down regulation of ß-adrenergic receptors: agonist-induced reduction in receptor mRNA levels. Proc Natl Acad Sci USA 1988; 85:5021–5025.[Abstract/Free Full Text]
32. Dangel V, Giray J, Ratge D, Wisser H. Regulation of ß-adrenoceptor density and mRNA levels in the rat heart cell line H9c2. Biochem J 1996; 317:925–931.
33. Hadcock JR, Wang H, Malbon CC. Agonist-induced destabilisation of ß-adrenergic receptor mRNA. Attenuation of glucocorticoid-induced up-regulation of ß-adrenergic receptors. J Biol Chem 1989; 264:19928–19933.[Abstract/Free Full Text]
34. Mahan LC, McKernan RM, Insel PA. Metabolism of alpha- and beta-adrenergic receptors in vitro and in vivo. Annu Rev Pharmacol Toxicol 1987; 27:215–235.[Medline]
35. Darfler FJ, Mahan LC, Koachman AM, Insel PA. Stimulation by forskolin of intact S49 lymphoma cells involves the nucleotide regulatory protein of adenylate cyclase. J Biol Chem 1982; 257:11901–11907.[Free Full Text]
36. Longabaugh JP, Didsbury J, Spiegel A, Stiles GL. Modification of the rat adipocyte A₁ adenosine receptor-adenylate cyclase system during exposure to an A₁ adenosine receptor agonist: alterations in the quantity of G_s and Gi are not associated with changes in their mRNAs. Mol Pharmacol 1989; 36:681–688.[Abstract]
37. Granneman JG, Lahners KN. Differential adrenergic regulation of ß1- and ß₃-adrenoreceptor messenger ribonucleic acids in adipose tissues. Endocrinology 1992; 130:109–114.[Abstract/Free Full Text]
38. L;aaecrivain JL, Mhaouty-Kodja S, Cohen-Tannoudji J, Maltier JP, Legrand C. In vivo stimulation of the ß₂-adrenergic pathway increases expression of the G_i proteins and the _2A-adrenergic receptor genes in the pregnant rat myometrium. J Endrocrinol 1998; 156:in press.
39. Mende U, Eschenhagen T, Geertz B, Schmitz W, Scholz H, Schulte am Esch J, Sempell R, Steinath M. Isoprenaline-induced increase in the 40/41 kDa pertussis toxin substrates and functional consequences on contractile response in rat heart. Naunyn-Schmiedebergs Arch Pharmacol 1992; 345:44–50.[Medline]
40. Engelhardt S, Zieger W, Kanubek J, Michel MC, Lohse MJ, Brodde OE. Tocolytic therapy with fenoterol induces selective down-regulation of ß-adrenergic receptors in human myometrium. J Clin Endocrinol & Metab 1997; 82:1235–1242.[Abstract/Free Full Text]
41. Brodde OE, Michel MC, Zerkowski HR. Signal transduction mechanisms controlling cardiac contractility and their alterations in chronic heart failure. Cardiovasc Res 1995; 30:570–584.[CrossRef][Medline]