Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Endothelins (ETs) are a family of 21-amino acid peptides that include ET-1, ET-2, and ET-3. The actions of ET are mediated by binding to distinct cell surface receptors, designated ET_A, ET_B, and ET_C. ET receptors belong to the G protein-coupled receptor superfamily (Rubanyi and Polokoff, 1994). In addition to its vasoconstrictive activity, a large body of evidence emphasizes the important role of ET-1 in the regulation of different functions of nonvascular smooth muscles, including the myometrium (Rae et al., 1995). Rat uterus smooth muscle cells express binding sites selective for ET-1 (Bousso-Mittler et al., 1989). ET-1 synthesized in intrauterine tissues (Rae et al., 1995) can modulate uterine activities in a paracrine manner. We previously reported that ET-1 activates ET_A receptor subtypes in rat myometrium, resulting in both the activation of the phosphatidylinositol-4,5-bisphosphate/phospholipase C (PtdInsP2/PLC) pathway via a pertussis toxin-insensitive G protein and the inhibition of adenylyl cyclase via a G_i-mediated process (Dokhac et al., 1994). The resulting increase in Ca²⁺ concentration and decrease in cAMP level are major determinants of ET-1-induced uterine contraction. Recently, a contribution of PKC to ET-1-induced contraction and proliferation of human myometrial cells was reported (Tertrin-Clary et al., 1999). Protein kinase C (PKC) is an enzyme activated by diacylglycerol (DAG), a second messenger produced by the PLC-catalyzed hydrolysis of PtdInsP2 (Nishizuka, 1992). Another source of DAG has been proposed with the hydrolysis of phosphatidylcholine (PC) through the PLD pathway (Liscovitch et al., 1993). Recently, we demonstrated that activated ETA receptors are also associated with the phospholipase D (PLD) pathway in rat myometrium (Naze et al., 1997).

Activation of PLD after the interaction of agonists with G protein-coupled receptors and receptors with tyrosine kinase activity leads to the hydrolysis of PC to produce phosphatidic acid (PA) along with choline. PA may serve as a potential lipid second messenger. It can also be converted into two messenger molecules, DAG and lysophosphatidic acid (LPA; Exton, 1997). The mechanisms regulating PLD and the means by which its lipid products function in diverse physiological processes are beginning to be elucidated. PLD can be activated by multiple pathways, involving PKC, heterotrimeric and small G proteins, protein tyrosine kinases, Ca²⁺, and unsaturated fatty acids (Exton, 1997; Frohman and Morris, 1999). Negative regulation of PLD has also been described. The PLD-inhibitory protein factors characterized include fodrin, the clathrin assembly protein 3, synaptojanin, and synucleins (Exton, 1997; Frohman and Morris, 1999). It has also been reported that ceramide, a lipid-derived metabolite of sphingomyelin, the generation of which is consistently associated with antiproliferative action (Obeid and Hannun, 1995), may also be an important negative regulator of the PLD pathway (Riboni et al., 1997).

We have previously reported that in rat myometrium, the activation of PKC that occurs after the breakdown of PtdInsP2 by PLC is important in the regulation of PLD stimulated by ET-1 (Naze et al., 1997). However, PKC activation cannot entirely account for the stimulatory effect of ET-1. This led us to examine the additional mechanisms involved in ET-1-mediated PLD activation. We further examined whether specific signaling pathways could inhibit the activation of PLD by ET-1. The current data revealed that a PTK-mediated process, independent of PKC activation, made an important contribution to the stimulatory effect of ET-1. We also found that the peptide-induced activation of PLD was negatively regulated by two different signaling pathways: the sphingomyelinase/ceramide and cAMP/PKA pathways. Ceramide specifically affected the PKC-dependent process, whereas cAMP selectively inhibited the PTK-dependent pathway of PLD activation triggered by ET-1.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals. [³H]Myristic acid (30-40 Ci/mmol) was purchased from New England Nuclear (Les Ulis, France). myo-[2-³H]Inositol (10-20 Ci/mmol) was obtained from Amersham International (Les Ulis, France). ET-1 was provided by Neosystem (Strasbourg, France). 3-Isobutyl-1-methylxanthine (IBMX) was obtained from Aldrich Chemical Co. (Milwaukee, WI). -Estradiol 3-benzoate, BSA, ceramides, D-erythro-N-acetyl-sphingosine (C2-ceramide), D-erythro-N-hexanoyl-sphingosine (C6-ceramide), forskolin, N-oleoylethanolamine, PDBu, dibutyryl cAMP (dbcAMP), sphingosine, sphingomyelin, sphingomyelinase (Staphylococcus aureus), tyrphostin-47, and tyrphostin-63 were obtained from Sigma Chemical Co. (St. Louis, MO). D-Erythro-N-acetyl-dihydrosphingosine (C2-dihydro ceramide) was obtained from Calbiochem (Meudon, France). Genistein was purchased from ICN Biomedicals France (Orsay, France). H-85 and H-89 were purchased from Seikagaku-Coger (Paris, France). Silica gel 60 plates were obtained from Whatman (Kent, England). Ro-31-8220 was a generous gift from Dr. Bradshaw (Roche Ltd., Hertfordshire, England). All other chemicals were of the highest grade available.

Preparation of Sphingolipids. For the stock solution in fatty acid-free BSA, 100 mM sphingolipid (C2-ceramide, C6-ceramide, C2-dihydroceramide, or sphingosine), in absolute ethanol, was diluted slowly into a solution of 2.5 mM BSA to a final concentration of 2 mM. The solution was then shaken at 37°C for 30 to 60 min before use. The solution could be frozen and was redispersed by an additional 30-min incubation as above, before use (Lambeth et al., 1988).

Preparation of Pervanadate. Pervanadate was prepared by mixing 5 parts of 5 mM sodium orthovanadate with 1 part of 250 mM H₂O₂ in distilled water and incubating at room temperature for 30 min, as described (Palmier et al., 1996) with minor modifications. Then, 1 part of catalase (100 µg/ml) was added to rapidly remove excess H₂O₂. Under these conditions, pervanadate was effective for at least 2 h after the removal of H₂O₂.

Animals and Tissue Processing. Immature Wistar female rats (4-5 weeks of age) were treated with 30 µg of estradiol for 2 days and sacrificed the next day with 1 min of carbon dioxide inhalation. Uteri were removed, and the myometrium was prepared free of endometrium (Dokhac et al., 1994; Naze et al., 1997). Histological section of myometrial preparations showed the samples to consist almost exclusively of longitudinal muscle.

PLD Assay. PLD activity was determined in [³H]myristic acid-labeled myometrium by measuring the accumulation of [³H]PBut, which is the product of its transphosphatidylation reaction and is considered to be a definitive assay for PLD (Liscovitch et al., 1993). Incubations were carried out as described (Naze et al., 1997) with minor modifications. Briefly, myometrial strips (25 mg) were equilibrated by incubation for 20 min in 5 ml of Krebs-Ringer-bicarbonate buffer (pH 7.4) containing 117 mM NaCl, 4.7 mM KCl, 1.1 mM MgSO₄, 1.2 mM KHPO₄, 2.4 mM NaHCO₃, 0.8 mM CaCl₂, and 1 mM glucose (gas phase 95% O₂/5% CO₂) under constant agitation. Tissues were then incubated with 8 µCi/ml [³H]myristic acid in 800 µl of fresh buffer for 5 h. After three successive washings with nonradioactive buffer, myometrial strips were incubated in 1 ml of fresh buffer containing 0.3% butanol for 10 min before exposure to the agents. Reactions were stopped by the rapid immersion of myometrial strips in liquid nitrogen. Lipids were extracted according to a modification of the method of Bligh and Dyer (1959). Frozen tissues were extracted in 1.8 ml of chloroform/methanol/HCl (50:100/1, v/v/v) with an Ultra-Turrax homogenizer and left at 4°C overnight. Then, 0.5 ml of H₂O was added, and after a brief homogenization, the monophase was split by the addition of 0.6 ml of 2 M KCl and 0.6 ml of chloroform. After a vigorous mixing, phases were separated by centrifugation for 10 min at 1000g, and the aqueous phase was removed. The chloroform extract was dried, using a speed vacuum concentrator, and the residue was suspended in 50 µl of methanol/chloroform (95:5, v/v) before thin-layer chromatography on heat-activated precoated 20 × 20-cm silica gel plates. To analyze the production of [³H]PBut, plates were developed in the organic phase of a mixture of ethyl acetate/2,2,4-trimethylpentane/acetic acid/water (13:2:3:10; v/v/v/v; Van Blitterswijk et al., 1991) and analyzed with a computerized Berthold radiochromatoscanner. The production of [³H]PBut was expressed as a percentage of the total radioactivity in phospholipid obtained from the same sample.

Measurement of [³H]Ceramides. Tissues were incubated with 10 µCi/ml [³H]myristic acid for 5 h. After three successive washings with nonradioactive buffer, myometrial strips were incubated in 1 ml of fresh buffer in the absence or presence of sphingomyelinase at the concentration and for the time indicated. Lipids were extracted according to the method of Bligh and Dyer (1959) as described earlier. Lipids were then alkaline hydrolyzed for 1 h with 0.1 N methanolic KOH (Dressler and Kolesnick, 1990) and reextracted. The chloroform extract was dried, and the residue was dissolved in 50 µl of methanol/chloroform (95:5, v/v). [³H]Ceramides were resolved by thin-layer chromatography on silica gel 60 plates using a solvent system of chloroform/methanol/acetic acid/H₂O (85:4:5.5:0.5, v/v/v/v) and were detected through comigration of ceramide standards (Quintans et al., 1994). The plates were then analyzed with a computerized Berthold radiochromatoscanner. The production of [³H]ceramides was expressed as a percentage of total radioactivity recovered on the plate.

Measurement of [³H]Inositol Phosphates. Myometrial strips (25 mg) were labeled with 5 µCi of myo-[2-³H]inositol (0.4 µM) in 800 µl of fresh buffer for 4 h, essentially as described previously (Dokhac et al., 1994). After three successive washings with nonradioactive buffer, tissues were incubated for 10 min in 1 ml of fresh buffer containing 10 mM LiCl before exposure to the agents indicated. Reactions were stopped by immersion of the tissue strips in liquid nitrogen. Tissues were then homogenized in 1.5 ml of cold 7% (w/v) trichloroacetic acid and centrifuged at 3000g for 20 min at 4°C. The trichloroacetic acid-soluble supernatants were extracted with diethyl ether, neutralized with Tris base, and applied to a column of anion exchange resin (AG1-X8, formate form, 200-400 mesh). Free inositol, glycerophosphoinositol, and total inositol phosphates were eluted successively with: 1) 12 ml of water, 2) 10 ml of 60 mM ammonium formate/5 mM sodium tetraborate, and 3) 12 ml of 1 M ammonium formate/0.1 M formic acid. The ³H content of the various fractions was determined by scintillation counting in Quicksafe (Zinsser Analytic). Results were expressed as cpm/100 mg of tissue.

Measurement of [³H]Phosphoinositides. Phosphoinositides present in the pellets obtained after centrifugation of the trichloroacetic acid homogenates were extracted according to the method of Bligh and Dyer (1959), essentially as described previously (Dokhac et al., 1994). The different phosphoinositides were separated by developing the plates in chloroform/methanol/4 M NH₄OH (90:70:20, v/v/v; Garret and Garret, 1976). The phospholipids were located according to their migration, compared with authentic standards (detected with iodine vapor). The plates were then analyzed with a computerized Berthold radiochromatoscanner. The radioactivity associated with PtdInsP2 was expressed as the percentage of the radioactivity in total phosphoinositides.

cAMP Assay. Myometrials strips (25 mg) were incubated in the presence of 250 µM IBMX, with 10 µM forskolin or 50 µM C6-ceramide for the time indicated. Reactions were stopped by immersion of the tissue strips in liquid nitrogen, followed by homogenization in 1 ml of cold 7% (w/v) trichloroacetic acid and centrifugation at 10,000g for 15 min at 4°C. The trichloroacetic acid supernatants were extracted with diethyl ether, and cAMP was estimated according to the method of Gilman (1970), as described previously (Dokhac et al., 1994). The pellets were used for protein determination (Lowry et al., 1951). cAMP level was expressed as picomoles per milligram of protein.

Data Analysis. The results are expressed as the mean ± S.E. and were analyzed statistically using Student's t test. P < .05 was considered significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effect of PKC and PTK Inhibitors on Production of [³H]PBut Triggered by ET-1 in Myometrium. We have previously detailed (Naze et al., 1997) the involvement of PKC in the stimulatory effects of both ET-1 and PDBu on the production of [³H]PBut in rat myometrium. It was found that in the presence of different PKC inhibitors, particularly Ro-31-8220, as well as under conditions of PKC depletion, there was a marked attenuation (<80%) of the generation of [3H]PBut promoted by PDBu. Under these conditions, the stimulatory effect of ET-1 on the production of [³H]PBut was only partially decreased. Data in Fig. 1 confirm the partial inhibition (53%) displayed by Ro-31-8220, used at 5 µM, on the production of [³H]PBut induced by ET-1, consistent with our previous interpretation that the peptide-mediated activation of PLD is dependent, albeit partially, on PKC activation. We have demonstrated that treatment of rat myometrium for 20 min with 100 µM pervanadate, a protein tyrosine phosphatase inhibitor, increases the phosphotyrosine content of a number of cellular proteins, including phosphorylation and activation of PLC-1 (Palmier et al., 1996). In similar conditions, pervanadate also enhanced the production of [³H]PBut (Fig. 1). Pervanadate stimulation was markedly attenuated (80%) when the myometrium was pretreated with a PTK inhibitor, genistein, used at 100 µM (Fig. 1). Inhibition was similarly observed with another PTK inhibitor, the active tyrphostin-47 but not with the inactive tyrphostin-63, consistent with the involvement of PTK activities. Pervanadate-mediated PLD activation was almost unaffected (<14% inhibition) by Ro-31-8220 (Fig. 1). Data in Fig. 1 show that the accumulation of [³H]PBut in response to ET-1 was partially reduced by both genistein and tyrphostin-47 (53 and 35% inhibition, respectively) but not by tyrphostin-63. It was previously reported that under these conditions of genistein and tyrphostin-47 treatment, the stimulation of PTK activities by ET-1 was totally abolished (Palmier et al., 1999). The simultaneous use of genistein and Ro-31-8220, at their maximally effective concentrations, resulted in the additive and complete inhibition of the ET-1 response (96% inhibition). Thus, two independent pathways, one mediated by PKC and the other by PTK, are involved in the activation of PLD by ET-1.

View larger version (25K): [in this window] [in a new window]   Fig. 1.   Effects of PTK and PKC inhibitors on [3H]PBut accumulation induced by ET-1 and pervanadate. [3H]Myristic acid-labeled myometrial strips were exposed to 0.3% butanol for 10 min without or with genistein (100 µM), tyrphostin-47 (100 µM), tyrphostin-63 (100 µM), and Ro-31-8220 (5 µM) added alone or combined. Incubations were further continued in the absence or presence of 0.2 µM ET-1 for 10 min or 100 µM pervanadate (PV) for 20 min. [3H]PBut accumulation was expressed as percent of total label in phospholipid, after substraction of basal values (0.18 ± 0.03, 0.19 ± 0.04, 0.20 ± 0.04, 0.22 ± 0.02, and 0.21 ± 0.02% of label in total phospholipid for untreated and Ro-31-8220-, tyrphostin-47-, tyrphostin-63-, and genistein-treated tissues, respectively). Data represent the mean ± S.E. of four independent experiments, each performed in duplicate. *P < .05 versus respective control. #P < .05 for ET-1 + genistein + Ro-31-8220 versus ET-1 + genistein. NS, not significantly different from respective control.

Effect of Cell-Permeable Ceramides and Exogenous Sphingomyelinase on Production of [³H]PBut Triggered by ET-1. Recent reports have shown that sphingomyelin hydrolysis products are able to regulate specific signal transduction pathways (Obeid and Hannun, 1995; Spiegel et al., 1996; Riboni et al., 1997). To evaluate the effects of ceramide, the primary metabolite of sphingomyelin hydrolysis, on PLD activation by ET-1, myometrial strips were treated with two cell-permeable ceramides: C2-ceramide and C6-ceramide. Results in Fig. 2 show that both C2-ceramide and C6-ceramide inhibited, in a concentration- and time-dependent manner, the production of [³H]PBut induced by ET-1. PLD inhibition occurred at C2- and C6-ceramide concentrations as low as 15 µM with maximum inhibition (53 and 55%, respectively), reached at a concentration of 50 µM. This maximum inhibition was recorded after 60 min of treatment with C2- and C6-ceramide and was maintained for at least 120 min.

View larger version (23K): [in this window] [in a new window]   Fig. 2.   Time- and concentration-dependent inhibitory effect of ceramide analogs on PLD activation triggered by ET-1. Myometrial strips were incubated with [3H]myristic acid (8 µCi/ml) for 5 h. After three successive washings with nonradioactive buffer, [3H]myristic acid-labeled myometrial strips were pretreated (left) without or with 50 µM C2-ceramide (, ) or C6-ceramide (, ) for varying lengths of time and (right) for 120 min with varying concentrations of C2- and C6-ceramide. Left and right, 0.3% butanol was added during the last 10 min of treatment, and tissues were incubated for an additional 10 min in the absence (, ) or presence (, ) of 0.2 µM ET-1. PLD activity was determined through the production of [3H]PBut and was expressed as a percent of total label in phospholipid. Values represent the mean ± S.E. of four independent experiments, each performed in duplicate.

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Effects of exogenous sphingomyelinase on the production of [3H]ceramides. Myometrial strips were incubated with [3H]myristic acid (10 µCi/ml) for 5 h. After three successive washings with nonradioactive buffer, [3H]myristic acid-labeled myometrial strips were treated with 0.1 U/ml exogenous sphingomyelinase for varying lengths of time. [3H]Ceramides were extracted as described in Materials and Methods. Production of [3H]ceramides was estimated as a percentage of total radioactivity recovered on the thin-layer chromatographic plate. Values are expressed as percent of those measured at time 0 in the absence of sphingomyelinase. Values represent the mean ± S.E. of three independent experiments, each performed in duplicate.

View this table: [in this window] [in a new window]   TABLE 1 Effects of ceramides, sphingosine, and sphingomyelinase on PLD activation mediated by ET-1 and AlF4 [3H]Myristic acid-labeled myometrial strips were exposed for 120 min to 50 µM C6-ceramide, C2-ceramide, C2-dihydroceramide, and sphingosine or to exogenous sphingomyelinase (SMase: 0.1 U/ml) in the absence or the presence of 500 µM N-oleoylethanolamine (NOE), added 60 min before SMase. Butanol (0.3%) was added during the final 10 min of treatment, and tissues were incubated in the absence or presence of 0.2 µM ET-1 for 10 min or with 20 mM NaF + 10 µM AlCl3 (AlF4) for 20 min. PLD activity was determined through the production of [3H]PBut, which was expressed as percentage of total label in phospholipid. Values represent the mean ± S.E. of four independent experiments, each performed in duplicate.

Effect of Cell-Permeable Ceramides and Exogenous Sphingomyelinase on ET-1-Mediated Activation of the PLC Pathway and on Level of PtInsP2. Data in Table 1 show that inhibition by C6-ceramide was not restricted to PLD stimulation triggered by receptor activation but that the cell-permeable ceramide was similarly able to reduce (48%) the production of [³H]PBut elicited by AlF₄, a direct activator of heterotrimeric G proteins. Because PKC, a signal downstream from receptor/G protein activation, is involved in ET-1-mediated [³H]PBut production, we examined whether the inhibition of PLD activation by ceramide was due to alteration of the PtdInsP₂/PLC cascade. Under conditions that attenuated ET-1-mediated PLD activation, neither C6-ceramide nor exogenous sphingomyelinase had a significant effect on the increased production of [³H]inositol phosphates caused by the peptide (Fig. 4). Similarly, neither the cell-permeable ceramide nor sphingomyelinase caused any appreciable change in the level of PtdInsP₂, as estimated by the amount of [³H]inositol incorporated into the phosphoinositide (Fig. 4).

View larger version (24K): [in this window] [in a new window]   Fig. 4.   Effects of C6-ceramide and exogenous sphingomyelinase on the relative distribution of [3H]inositol in PtdInsP2 and on the production of [3H]inositol phosphates triggered by ET-1. [3H]Inositol-labeled myometrial strips were exposed for 120 min to C6-ceramide (C6-cer; 50 µM) or to exogenous sphingomyelinase (Smase; 0.1 U/ml). Left, 10 mM LiCl was added during the final 10 min of treatment, and incubations were continued for an additional 10 min in the presence of 0.2 µM ET-1. [3H]Inositol phosphates were extracted as described in Materials and Methods. Production of [3H]inositol phosphates was expressed as cpm/100 mg of tissue, after substraction of basal values (18,385 ± 1,655 cpm/100 mg of tissue). Right, 3H-labeled phosphoinositides were extracted and subjected to thin-layer chromatography as described in Materials and Methods. Radioactivity associated with PtdInsP2 was expressed as a percentage of total [3H]phosphoinositides (92,090 ± 7,360, 95,250 ± 8,570, and 73,050 ± 10,227 cpm/100 mg of tissue for untreated, C6-ceramide-, and sphingomyelinase-treated tissues, respectively). Values represent the mean ± S.E. of three independent experiments, each performed in duplicate.

Differential Effects of Cell-Permeable Ceramides on PDBu and Pervanadate-Mediated Production of [³H]PBut. Results reported in Fig. 5 show that the PTK-dependent process that is involved in the stimulation by pervanadate of [³H]PBut production was totally insensitive to C6-ceramide. However, both C2- and C6-ceramide caused an attenuation (52 and 53% inhibition, respectively) of the PDBu-mediated response, with the inactive C2-dihydroceramide having no effect.

View larger version (21K): [in this window] [in a new window]   Fig. 5.   Differential effects of ceramide analogs on [3H]PBut accumulation induced by PDBu and pervanadate. [3H]Myristic acid-labeled myometrial strips were exposed for 120 min to 50 µM C6-ceramide (C6-cer), C2-ceramide (C2-cer), or C2-dihydroceramide (C2i). Butanol (0.3%) was added during the final 10 min of treatment Tissues were then stimulated with 1 µM PDBu for 10 min or with 100 µM pervanadate (PV) for 20 min. The production of [3H]PBut was expressed as a percentage of total label in phospholipid, after the substraction of basal values (0.18 ± 0.03, 0.21 ± 0.04, 0.22 ± 0.03, and 20 ± 0.02% of label in total phospholipid for untreated and C6-ceramide-, C2-ceramide-, and C2-dihydroceramide-treated tissues, respectively). Data represent the mean ± S.E. of three independent experiments, each performed in duplicate. *P < .05 versus respective control. NS, not significantly different from respective control.

Combined Effect of C6-Ceramide, Ro-31-8220, and Genistein on Activation of PLD by ET-1. Treatment of the myometrium with genistein and C6-ceramide together gave greater inhibition (81%) of the [³H]PBut accumulation induced by ET-1 than did genistein (48% inhibition) or C6-ceramide (48% inhibition) alone (Fig. 6). In contrast, treatment of the myometrium with Ro-31-8220 and C6-ceramide together did not result in an inhibition greater than that obtained with each agent alone. This implies that the PKC-mediated process, and not the PTK-mediated process involved in PLD activation by ET-1, is the target for inhibition by C6-ceramide.

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Effects of different combinations of C6-ceramide, genistein, and Ro-31-8220 on the production of [3H]PBut induced by ET-1. [3H]Myristic acid-labeled myometrial strips were incubated for 120 min without or with 50 µM C6-ceramide (C6-cer). Butanol (0.3%) was added, in the absence or presence of genistein (100 µM) and Ro-31-8220 (5 µM) during the final 10 min of treatment. Tissues were then stimulated for an additional 10 min with 0.2 µM ET-1. The production of [3H]PBut was expressed as a percentage of total label in phospholipid, after substraction of basal values (0.23 ± 0.05, 0.21 ± 0.03, 0.18 ± 0.06, and 0.20 ± 0.03% of label in total phospholipid for untreated and C6-ceramide-, genistein-, and Ro-31-8220-treated tissues, respectively). Data represent the mean ± S.E. of four independent experiments, each performed in duplicate. *P < .05 versus ET-1 alone. #P < .05 for ET-1 + C6-ceramide + genistein versus ET-1 + C6-ceramide. NS, not significantly different for ET-1 + C6-ceramide + Ro-31-8220 versus ET-1 + C6-ceramide.

Effects of cAMP-Elevating Agents on ET-1-Mediated Activation of PLD and PLC Pathways. It has been reported that depending on the cell type, agents that elevate cAMP may either stimulate PLD activity (Tyagi et al., 1991) or inhibit agonist-mediated PLD activation (Ginsberg et al., 1997). Thus, we analyzed the action of forskolin, a direct activator of the catalytic unit of adenylyl cyclase, and iloprost, a stable analog of prostacyclin, both of which elevate cAMP levels in rat myometrium (Mokhtari et al., 1985; Goureau et al., 1992). Under conditions in which both agents increased the production of cAMP, basal levels of [³H]PBut production were not affected (Table 2). In contrast, both forskolin (10 µM) and iloprost (5 µM) consistently reduced (51 and 52%, respectively) the amount of [³H]PBut generated by ET-1. Similarly, dbcAMP, a cell-permeable cAMP analog, also attenuated (47%) ET-1-mediated [³H]PBut formation. Treatment of the myometrial strips with compound H-89 (30 µM), a known PKA inhibitor (Chijiwa et al., 1990), abolished the ability of forskolin to attenuate PLD activation by ET-1. Under similar conditions, the structurally related, but inactive, H-85 (30 µM) had no effect. Thus, ET-1-mediated PLD activation in the myometrium appears to be negatively modulated by the cAMP/PKA cascade. This inhibition of the production of [³H]PBut caused by ET-1 could not be attributed to an alteration of the PLC cascade. Neither forskolin nor dbcAMP and, similarly, iloprost, had any effect on the increased production of [³H]inositol phosphates, with the attendant rise in Ca²⁺ and activation of PKC, triggered by the peptide.

Differential Effects of cAMP-Elevating Agents on PdBu and Pervanadate-Mediated Production of [³H]PBut. Results in Fig. 7 show that forskolin decreased (51%) the [³H]PBut response to pervanadate. This decrease also appears to involve the cAMP/PKA pathway in that inhibition by forskolin was markedly reduced (8% inhibition) by H-89, whereas it was unaffected by H-85 (45% inhibition). In marked contrast, the increase in [³H]PBut production mediated by PDBu was totally insensitive to both forskolin and dbcAMP.

View larger version (28K): [in this window] [in a new window]   Fig. 7.   Inhibitory effects of cAMP-elevating agents on PLD activation triggered by PDBu and pervanadate. [3H]Myristic acid-labeled myometrial strips were incubated for 20 min without or with dbcAMP (1 mM), H-89 (30 µM), or H-85 (30 µM). Butanol was added, in the absence or presence of 10 µM forskolin (FK), during the final 10 min of treatment. Incubations were further continued in the presence of 1 µM PDBu for 10 min and 100 µM pervanadate (PV) for 20 min. The production of [3H]PBut was expressed as a percentage of total label in phospholipid, after substraction of basal values (0.21 ± 0.03, 0.19 ± 0.05, and 0.22 ± 0.02% of label in total phospholipid for untreated and forskolin- and dbcAMP-treated tissues, respectively). Data represent the mean ± S.E. of three independent experiments, each performed in duplicate. *P < .05 versus pervanadate alone. NS, no significant difference with their respective control.

Combined Effects of Forskolin, Ro-31-8220, and Genistein on Activation of PLD by ET-1. Treatment of the myometrium with forskolin and Ro-31-8220 together increased the inhibition of [³H]PBut production triggered by ET-1 (Fig. 8). Under these conditions, 85% inhibition was achieved versus the 46 and 48% inhibition obtained with Ro-31-8220 and forskolin, respectively, added alone. In marked contrast, treatment of myometrium with both forskolin and genistein gave no greater inhibition of [³H]PBut production than that obtained with each agent alone (Fig. 8). Thus, cAMP appears to inhibit the PTK-mediated process rather than the PKC-mediated process involved in the activation of PLD by ET-1.

View larger version (21K): [in this window] [in a new window]   Fig. 8.   Inhibitory effects of different combinations of forskolin, C6-ceramide, genistein, and Ro-31-8220 on PLD activation induced by ET-1. [3H]Myristic acid-labeled myometrial strips were incubated for 120 min without or with 50 µM C6-ceramide (C6-cer). Butanol (0.3%) was added, in the absence or the presence of forskolin (10 µM), Ro-31-8220 (5 µM), and genistein (100 µM), during the final 10 min of treatment. Tissues were stimulated for an additional 10 min with 0.2 µM ET-1. The production of [3H]PBut was expressed as a percentage of label in total phospholipid, after substraction of basal values (0.19 ± 0.03, 0.22 ± 0.01, 0.18 ± 0.04, 0.21 ± 0.02, and 0.19 ± 0.02% of total label in phospholipid for untreated and C6-ceramide-, forskolin-, Ro-31-8220-, and genistein-treated tissues, respectively). Data represent the mean ± S.E. of three independent experiments, each performed in duplicate. *P < .05 versus ET-1 alone. #P < .05 versus ET-1 + forskolin. NS, not significantly different for ET-1 + forskolin + genistein versus ET-1 + forskolin.

Independent Mechanisms for Inhibition of ET-1-Mediated PLD Activation by cAMP and Ceramides. C6-ceramide, under conditions that decreased the production of [³H]PBut evoked by ET-1, caused no increment in intracellular cAMP levels (Table 2). Kinetic experiments (not reported) demonstrated that the level of cAMP was not affected by incubation for 15 to 120 min with C6-ceramide. This indicates that the mechanism via which ceramides inhibit PLD activation by ET-1 is distinct from that involved in inhibition by cAMP. This interpretation was supported by further experiments (Fig. 8) in which treatment with both forskolin and C6-ceramide resulted in additive inhibition of the production of [³H]PBut mediated by ET-1. The extent of inhibition was 48 and 50% for forskolin and C6-ceramide, respectively, and reached 87% if both were present.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

In a preceding report (Naze et al., 1997), we showed that among other signaling systems (Dokhac et al., 1994; Palmier et al., 1999), the PLD/PC cascade is associated with activated ET receptors in the rat myometrium. Our present study extended these observations and revealed that PLD in the myometrium is activated by pathways involving PKC and PTK and that both pathways make a major contribution to ET-1-mediated PLD activation. The data further show that stimulation of PLD by ET-1 may be negatively regulated by two different pathways. One pathway involves the sphingomyelinase/ceramide cascade that specifically affected the PKC-dependent process involved in PLD activation triggered by ET-1. The second pathway is mediated by the cAMP/PKA cascade, which, in contrast, specifically affected the PTK-dependent arm of the ET-1 response.

There is abundant evidence that PLD is regulated by PKC in most mammalian cells (Exton, 1997). We have previously demonstrated (Naze et al., 1997) and herein confirmed that PKC is involved in both PDBu- and ET-1-stimulated PLD activity in rat myometrium. However, PKC activation accounted for no more than 50% of the ET-1 response. Work from our laboratory (Palmier et al., 1996) has shown that treatment of the myometrium with pervanadate, a protein tyrosine phosphatase inhibitor, increases the tyrosine phosphorylation of a number of proteins, including PLC-1. Recently, ET-1 was also reported to promote tyrosine phosphorylation of proteins in rat myometrium (Palmier et al., 1999). In the present study, we found that pervanadate activated PLD and that this activation was totally PTK-dependent but PKC-independent. Our observations that two PTK inhibitors, genistein and tyrphostin-47, but not the inactive tyrphostin-63, reduced the [³H]PBut accumulation induced by ET-1 support the contention that a PTK-regulated mechanism is involved in ET-1-stimulated PLD activity. It is worth mentioning that even in the presence of Ro-31-8220, genistein still displayed its maximal inhibitory effect, supporting the contention that a PTK-regulated mechanism, independent of PKC activation, contributed to ET-1-stimulated PLD activity. This is consistent with a previous report implicating different roles for PKC and PTK in the regulation of PLD by bombesin (Briscoe et al., 1995). Two mammalian PLD isoforms, PLD1 and PLD2, have been cloned and characterized to date (Frohman and Morris, 1999). Lopez et al. (1998) reported the coexpression of PLD1 and PLD2 in the intact human uterus. We also found through the use of specific isoform antibodies that PLD1 and PLD2 are coexpressed in the rat myometrial preparations (H. Le S., L.D., and S.H., manuscript in preparation). It is difficult to establish from the present observations which of the PLD isoforms is responsible for ET-1-mediated PLD activation. The possibility still remains that each of the PKC and PTK regulatory processes may activate a specific PLD isoform and that the sum of the response to ET-1 reflects the activation of both isoforms. It is now well established that PLD1 is activated by PKC (Frohman and Morris, 1999). Both PLD1 and PLD2 isoforms have been shown to undergo tyrosine phosphorylation in certain stimulatory conditions (Marcil et al., 1997; Slaaby et al., 1998), but it is unclear whether phosphorylation per se is responsible for the increase in PLD activity.

There is evidence that ceramide reduces agonist-mediated PLD activation in intact fibroblasts and HL-60 cells (Gomez-Munoz et al., 1994; Jones and Murray, 1995). This study showed that ceramide partly inhibits the stimulation of PLD by ET-1 in the myometrium. Ceramide also decreased PLD activation elicited by AlF₄, indicating that inhibition occurred at a step downstream from receptor/G protein activation. The effects of cell-permeable ceramides were concentration- and time-dependent and specific in that the closely related dihydroceramide had no effect. The inhibition of ET-1-stimulated PLD activity by exogenous sphingomyelinase, which increases endogenous ceramide levels, suggests that the effect of ceramide may have physiological relevance in rat myometrium. Inhibition could not be ascribed to ceramide being converted to sphingosine because: 1) sphingosine, as previously observed in other cells (Spiegel et al., 1996), stimulated PLD activity in the myometrium and 2) inhibition by sphingomyelinase was not affected by the presence of an inhibitor of ceramidase, the enzyme involved in the catabolism of ceramide.

Our observation for an additive inhibition by C6-ceramide and genistein on the activation of PLD by ET-1 indicates that ceramides did not affect the PTK-dependent component of the ET-1 response. This interpretation is supported by the failure of C6-ceramide to significantly attenuate the stimulation of [³H]PBut production by pervanadate. In contrast, the inhibition of ET-1-mediated effects by C6 ceramide was clearly not additive to inhibition by the PKC inhibitor Ro-31-8220, suggesting that both compounds target a common regulatory pathway, the PKC-dependent pathway. It is worth considering that ceramide did not affect the ability of ET-1 to stimulate the PtdInsP2/PLC pathway, a prerequisite step for PKC activation via the generation of DAG. Similarly, ceramide did not affect the level of PtdInsP2, described as an essential cofactor for PLD activation (Exton, 1997). The interpretation that ceramide negatively regulates a PKC-mediated process that contributes to PLD activation by ET-1 is supported by the inhibition displayed by C6-ceramide versus the [³H]PBut production stimulated by PDBu. The data are in accordance with similar observations demonstrating that ceramides interfere with PKC-mediated PLD activation in both intact and cell-free systems (Gomez-Munoz et al., 1994; Jones and Murray, 1995; Abousalham et al., 1997). However, the exact mechanism of such an inhibition is not clearly delineated. It has been reported by two independent groups (Jones and Murray, 1995; Abousalham et al., 1997) that ceramides, which inhibit agonist- or phorbol ester-stimulated PLD activity in intact cells, block the membrane translocation of Ca²⁺-dependent PKC isozymes, whereas Venable et al. (1996) described no effect of ceramides on PKC translocation. Ceramides were also reported (Abousalham et al., 1997) to prevent the translocation to the membrane of ADP-ribosylating factor and RhoA, which are required for PLD activation. The molecular mechanisms via which ceramides inhibit the PKC component of ET-1-mediated PLD activation in myometrium remain to be clarified.

Another example of cross-talk between signal transduction pathways in the myometrium was provided by our demonstration of inhibition of ET-1-mediated PLD activation by the adenylyl cyclase/PKA cascade. It has been reported that cAMP displays different effects on PLD activation. Tyagi et al. (1991) reported inhibition of PLD activation by fMet-Leu-Phe in neutrophils by agents that increased cAMP levels, whereas the report of Ginsberg et al. (1997) demonstrated that cAMP-elevating agents activate PLD in FTRL-5 thyroid cells. In this study, forskolin, a direct activator of the catalytic unit of adenylyl cyclase, attenuated by about 50% the activation of PLD by ET-1 without affecting ET-1-mediated PLC activation. The inhibitory effect of forskolin was correlated with its ability to raise intracellular cAMP. It was reproduced by the cell-permeable cAMP analog dbcAMP and implied a PKA-mediated process. Also of interest is the inhibition of ET-1-mediated PLD activation by iloprost, a stable analog of prostacyclin that is the major prostaglandin compound that modulates cAMP levels in rat myometrium (Vesin et al., 1979). These observations suggest that the effect of cAMP on PLD activity may have functional significance in rat myometrium.

Although ceramides were shown to affect the levels of cAMP in some cells (Riboni et al., 1997), inhibition of PLD activation by ceramides in the myometrium could not be ascribed to a cAMP-mediated process. Indeed, C6-ceramide caused no increase in tissue cAMP content under conditions in which it attenuated ET-1 responses in terms of PLD activation. Instead, our data provide evidence for the involvement of two independent mechanisms in inhibition by ceramides and cAMP. The PKC-dependent process, involved in the activation of PLD, the target for ceramide inhibition, was insensitive to cAMP. This is illustrated: 1) by the inability of cAMP to affect the activation of PLD by PDBu and 2) by the additive inhibition of ET-1-induced PLD activation by the PKC inhibitor and cAMP, or ceramide and cAMP. It is interesting that cAMP inhibited the pervanadate-mediated PLD activation and that the combination of genistein and cAMP gave no additive inhibition of ET-1 responses. Our reported findings are consistent with the interpretation that the PTK-dependent pathway operating in the activation of PLD by ET-1 is the target for cAMP/PKA-mediated inhibition. These data provide the first insight into how cAMP may interfere with PLD activation. The site of PKA-evoked phosphorylation responsible for attenuating the PTK-dependent process in the activation of PLD is unknown. It may be the PLD itself or an associated protein that, once phosphorylated, renders the enzyme insensitive to PTK-dependent regulation. The possibility that cAMP interferes with the activation of PTK downstream from G protein activation cannot be ruled out.

In summary, we have shown that PLD activation by ET-1 is regulated by both stimulatory and inhibitory mechanisms that result from the complex interaction of several signal-generating systems. It is recognized that the PtdInsP2/PLC pathway (Dokhac et al., 1994) and PTK (Palmier et al., 1996) are important regulators of myometrial contractility, whereas cAMP makes a major contribution to myometrial relaxation (Dokhac et al., 1986). There also is some evidence that LPA, which can be generated from PA, the metabolite of the PLD/PC pathway, stimulates contraction of different smooth muscle preparations (Nietgen and Durieux, 1998). Both PA and LPA have also been described as mitogens in some cell lines (Exton, 1997; Nietgen and Durieux, 1998), in contrast to ceramides and cAMP, which inhibit proliferation of smooth muscle cells (Tomlinson et al., 1995; Johns et al., 1998). It is quite reasonable to consider that the diverse regulatory processes, described in the present study, that would ultimately tend to modulate the level of PA and/or its metabolites are of physiological relevance, particularly in the control of myometrial cell motility and proliferation by ET-1. These concerns are the subjects of our current work.