Luminal and basolateral endothelin inhibit chloride reabsorption in the mouse thick ascending limb via a Ca2+-Independent Pathway

Authors

  • Marie Celeste de Jesus Ferreira,

    1. URA CNRS 1859, Département de Biologie Cellulaire et Moléculaire, CEA Saclay, 91191 Gif sur Yvette, France
    Search for more papers by this author
  • Claire Bailly

    Corresponding author
    1. URA CNRS 1859, Département de Biologie Cellulaire et Moléculaire, CEA Saclay, 91191 Gif sur Yvette, France
    • To whom correspondence should be addressed at URA 1859, SBCe/DBCM, Bât.520, Commissariat à l'Energie Atomique, Saclay, 91191 Gif sur Yvette Cedex, France.

    Search for more papers by this author

Abstract

  • 1The recent localization of endothelin synthesis and receptors in the thick ascending limb (TAL) prompted us to investigate a possible autocrine and/or paracrine effect of this agent. The net chloride flux (JCl) has been determined in isolated cortical and medullary TALs by the in vitro microperfusion technique.
  • 2In both segments, endothelin 1 (ET-1) at 10−8m in the bath significantly decreased JCl, an effect which was partially reversible and observed at concentrations equal to or greater than 10−13m.
  • 3This (JCl inhibition (by 33.9 ± 3.2%) was blocked by BQ788 and was also observed with sarafotoxin 6C and ET-3, indicating that endothelin receptor B (ETB) are present in TAL.
  • 4ET-1 did not affect cAMP content under basal or hormone-stimulated conditions. The presence of a prostaglandin synthesis inhibitor also did not prevent the ET-1 action on JCl.
  • 5The ET-1-induced inhibition of JCl was prevented by protein kinase C inhibitors (staurosporine or GF 109203) and was reproduced by diacylglycerol analogues (OAG and DiC8). However, ET-1 failed to increase intracellular Ca2+ concentration.
  • 6Addition of ET-1 or ET-3 to the apical surface induced a decrease of JCl through ETB receptors, an effect which was not additive with that induced by basolateral ET-1, and was not concomitant with an increase in intracellular Ca2+ concentration.
  • 7It is concluded that the basolateral and luminal inhibitions of JCl by ET-1 in TAL, through ETB receptors, is mediated by a protein kinase C activation which is independent of intracellular Ca2+ increase.

Since the discovery of endothelin as a potent endothelium-derived vasoactive peptide, three distinct isoforms have been identified to date, namely endothelin 1, 2, and 3 (ET-1, ET-2, and ET-3). Identification of the sites of production of these isoforms and evidence of their action outside the vascular system have furnished new insights on possible differences in their respective physiological roles (for review see Sokolovsky, 1995). In the kidney, recent studies (Ujiie, Terada, Nonogushi, Shinohara, Tomita & Marumo, 1992; Terada, Tomito, Nonogushi, Yang & Marumo, 1993) have localized ET-1 and ET-3 synthesis along the rat nephron: ET-1 is exclusively produced by glomeruli and inner medullary collecting tubules, whereas ET-3 is mainly synthetized in proximal tubules and outer medullary collecting tubules. A significant production of ET-3, but not of ET-1, has also been found in the medullary part (M) of the thick ascending limb (TAL) of the loop of Henle. It is of interest that early evidence suggested that ET-1 and ET-3 productions were controlled in different ways related to physiopathological states (Terada et al. 1993; Firth, Schricker, Ratcliffe & Kurtz, 1995).

Three subtypes of endothelin receptor have been cloned (ETA, ETB, and ETC) (for review, see Sokolovsky, 1995). They all belong to the protein-G-coupled receptor family and differ from one another in their respective affinities for ligands: ETA binds ET-1 and ET-2 with a 100-fold higher affinity than ET-3, ETB displays similar affinities for the three peptides, whereas ETC seems highly selective for ET-3. In the rat kidney, ETA and ETB are not equally distributed: mRNA coding for ETA is present in glomeruli, vascular bundles, and interstitial cells, whereas mKNA coding for ETB is found in glomeruli and tubular epithelium (Terada, Tomita, Nonogushi & Marumo, 1992; Chow, Subramanian, Nuovo, Miller & Nord, 1995). Furthermore, these studies pointed out that whereas ETB mRNA is most abundant in inner medullary collecting tubules, small but detectable amounts have also been found in MTAL (Terada et al. 1992). No information about the cortical part (C) of TAL is available. In most target cells, ETA and ETB receptors are coupled to phospholipase C, triggering intracellular Ca2+ release and protein kinase C activation which are associated in some cases with the subsequent induction of phospholipases D and A2. Inhibition of adenylyl cyclase has also been described (for review, see Sokolovsky, 1995).

In agreement with these biochemical studies, functional studies have shown that ET-1 plays a physiological role in renal function: it was reported to decrease the glomerular filtration rate in several ways including contraction of the glomerular mesangial cells and vasoconstriction of the afferent and efferent arteries (Badr, Murray, Breyer, Takahashi, Inagami & Harris, 1989). In contrast to these haemodynamic effects, there are also indications that ET-1 increases water and sodium excretion by a direct action on renal tubules. In the collecting duct, this effect has been ascribed, in part, to ET-1-induced inhibition of the anti-diuretic hormone (ADH)-stimulated cAMP generation and, consequently, of the ADH-induced increase in water permeability (Tomita, Nonogushi & Marumo, 1991; Nadler, Zimpelmann & Hébert, 1992; Tomita, Nonogushi, Terada & Marumo, 1993).

Since NaCl reabsorption in TAL contributes to the urinary dilution–concentration process, the question arises as to whether the natriuretic and diuretic effects of endothelin could be accounted for, in part, by an inhibitory action on the TAL reabsorptive capacity. The purpose of this work was, therefore, to evaluate the possible effect of endothelin on net Cl flux (JCl) in both MTAL and CTAL. Moreover, since endothelin synthesis probably occurs in TAL, we investigated whether this agent is active at both the luminal and basolateral surfaces of the tubular cells.

METHODS

Microperfusion experiments

CTALs and MTALs were microperfused in vitro following the technique first described by Burg (Burg, Grantham, Abramow & Orloff, 1996) and routinely used in our laboratory. Animals were bred and killed according to the guidelines from the Ministère de l'Agriculture et de la Pêche (authorization no. 5443). Briefly, male Swiss mice weighing 18–20 g were killed by cervical dislocation and exsanguinated. Coronal slices were cut from both kidneys and immediately immersed in a cold perfusion solution (for composition see below) with 0.1% bovine serum albumin added. CTAL and MTAL were dissected from the medullary rays of the cortex and from the inner strip of the outer medulla, respectively. Each tubule was then transferred to a Lucite chamber thermostatically maintained at 36.0 ± 0.1 °C, with a flow rate of ∼5 ml min−1.

Each perfused tubule was allowed to equilibrate for 1 h for CTAL and 0.5 h for MTAL. After a 30 min control period the agent under study was added to the bath or to the lumen, according to the experimental protocol. After 10 min, a 30 min experimental period followed. For recovery experiments, an additional 30 min experimental period was performed after removal of the agent. The luminal fluid was collected every 10 min. When agents were present on the luminal side, the luminal fluid was manually changed with a syringe, under microscopic observation, as we have already reported (Néant & Bailly, 1993). This technique was found not to modify the JCl in control tubules.

The composition of the perfusion solution was as follows (mm): NaCl, 140; K2HPO4, 2.4; NaH2PO4, 0.6; MgCl2, 1; CaCl2, 1; Hepes, 10; urea, 10. Glucose (5 mm) was added to the bathing solution. All solutions were adjusted to pH 7.39–7.42.

Chloride concentrations in collected fluid (Cc and perfusate (Cp) were determined by microelectrometric titration. The tubular flow rate (V) was calculated from the volume of the collected sample, assuming that water reabsorption was negligible. The length (L) of the perfused tubule was measured with an eyepiece micrometer at × 400 magnification. The net chloride flux was calculated as JCl= (CcCp)V/L, expressed in pmol min−1 (mm tubular length)−1.

Data from the three collections of each 30 min period were pooled and considered as one datum. Values are expressed as means ±s.e.m. Statistical significance was evaluated within each series by Student's paired t test. For comparing the different series, results were expressed as the percentage of JCl inhibition for each tubule since a linear relationship exists between JCl in the control and the experimental period (see Results). Statistical treatment was the one-way analysis of variance followed by Fisher's least significant difference test, using as the parameter the difference of log(JCl) between the control and experimental periods.

Determination of intracellular calcium concentration ([Ca2+]i)

Experiments were carried out on isolated tubules, according to the technique used in our laboratory (Champigneulle, Siga, Vassent & Imbert-Teboul, 1993). Briefly, tubules were dissected from eight male Swiss mice (18–20 g body weight), following the protocol described above for microperfusion experiments. Each tubule was then transferred onto a slide where it was loaded for 1 h at room temperature in the dark, with a solution containing fura-2 AM (acetomethoxy ester; 10−5 M) in DMSO (final dilution 500). Each fura-2-loaded tubule was then transferred to a Lucite chamber, held tightly by pipettes at both sides, and superfused at 36 °C at a rate of 10–12 ml min−1, so that total equilibrium in the chamber was achieved in about 15–20 s. The bathing solution contained (mm): NaCl, 140; KCl, 5; NaHCO3, 4; NaH2PO4, 0.44; Na2HPO4, 0.33; MgCl2, 1; MgSO4, 0.8; CaCl2, 1; glucose, 2; Hepes, 10; pH 7.4. When ET-1 was added on the apical side, the tubules were microperfused as described for JCl determinations (see above) and the luminal fluid was manually changed with a syringe. This manoeuvre could be made rapidly since with no sampling required there was no oil in the collecting pipette and rapid luminal fluid change did not provoke tubule swelling. For each tubule, measurements of fura-2 fluorescence were carried out after a 5–10 min equilibration period.

The portion of the tubule selected for fluorescence measurements was centered across a circular field diaphragm (60 mm diameter) and included about twenty to thirty cells. Fluorescence intensities were recorded at two excitation wavelengths, 340 and 380 nm. Tubule autofluorescence at the two wavelengths was determined on homologous segments not loaded with fura-2 and superfused as described above. For each tubule, the net fluorescence intensities of fura-2 at 340 and 380 nm were obtained by subtracting the corresponding autofluorescence from all measurements. [Ca2+]i was calculated from the ratio R of the fluorescence intensities at 340 and 380 nm according to the following equation:

display math

where Kd (224 nm) is the dissociation constant of fura-2 for calcium; Rmin and Rmax are the values of R at 0 and the saturating calcium concentration, respectively; (Sf2/Sb2) are the fluorescence intensities at 380 nm at 0 and the saturating calcium concentration, respectively.

cAMP content

cAMP content was determined by radioimmunoassay in tubules microdissected after collagenase dissociation of kidney tissue, following a procedure currently used in our laboratory (Firstov, Aarab, Mandon, Siaume-Perez, De Rouffignac & Chabardes, 1995). Twenty-one male Swiss mice were anaesthetized with sodium pentobarbitone (0.1 mg (10 g body wt)−1). The left kidney was perfused in situ via the abdominal aorta with 5 ml of incubation solution containing (nm): NaCl, 120; KCl, 5; CaCl2, 1; MgSO4, 1; Na2HPO4, 4; NaHCO3, 4; glucose, 5; CH3COONa, 10; and Hepes, 20. Dextran 40000 (0.3% w/v), collagenase (151 U mg−1) 0.3% w/v, and BSA 0.1% w/v were added. Thin pyramids were excised along the cortico-papillary axis of the left kidney and were then incubated for 10–15min at 35 °C in the incubation solution containing 0.1% collagenase.

For cAMP determinations, one or two tubule segments were transferred to a slide in 2 ml of incubation solution with 5 × 10−6m Ro 201724, an inhibitor of the high affinity cAMP phosphodiesterase (type IV). Individual segments were photographed for subsequent determination of their lengths. Each sample was pre-incubated for 10 min and then incubated for 4 min at 35 °C in the presence of 2 μl of either the solution alone (basis) or the relevant hormone. The reaction was stopped by transferring the sample into a tube containing 25 μl of a mixture of formic acid and absolute ethanol (5% v/v). After evaporation, acetate buffer was added and cAMP was determined by radioimmunoassay after acetylation. The limit of detection was one femtomole of cAMP per tube.

Five to eight replicates were determined for each experimental condition. The mean of these replicate values was considered as one datum and was used to calculate the mean and standard error of the response in several animals. Statistical significance was evaluated by the one-way analysis of variance followed by Fisher's least significant difference test.

Materials

ET-1 and ET-3 were purchased from Alexis Co. (Laüfalfingen, Switzerland), GF 109203 and BQ788 from RBI (Natick, MA, USA), OAG (1-oleoyl, 2-acetyl-sn-glycerol) and DiC8 (1,2-dioctanoyl glycerol) from Calbiochem (La Jolla, CA, USA). cAMP antibodies were obtained from INSERM 64 (Hôpital Tenon, Paris, France). All other compounds were purchased from Sigma.

RESULTS

Effects of basolateral and luminal endothelin on Cl reabsorption

Addition of 10−8m ET-1 to the bath, significantly decreased JCl in MTAL (Fig. 1). This inhibitory effect (33.1 ± 4.2%) was significantly higher than that observed in time control tubules (4.7 ± 2.5%; P < 0.001). A similar effect was observed in CTAL, in which ET-1 significantly decreased JCl by 35.9 ± 4.0% (Fig. 1). This effect of ET-1 on JCl resulted from a significant reduction of the chloride concentration gradient between the perfused and collected fluids (10.7 ± 1.4 vs. 14.7 ± 1.6 mm, P < 0.001, and 15.7 ± 5.3 vs. 24.4 ± 7.8 mm, P < 0.05, in MTAL and CTAL, respectively) while the tubule flow rates did not change significantly (3.2 ± 0.3 vs. 3.4 ± 0.3 nl min−1, and 3.2 ± 0.6 vs. 3.2 ± 0.6 nl min−1, n.s., for MTAL and CTAL, respectively). The variability of JCl between individual tubules did not invalidate the overall effect since: (1) each tubule was used as its own control, and (2) a significant linear relationship existed between the JCl in the experimental period (y) and that in the control period (x). This is shown in either the presence of endothelin (y= 0.70x–0.68, r= 0.87, P < 0.001; pooled data from ET-1 and ET-3, see below) or in time-control tubules (y= 0.94x+ 2.51, r= 0.98, P < 0.001). Moreover, the slopes of these regression lines were significantly different (P < 0.02), indicating that the relative effect of ET-1 was independent of the basal JCl value over the range measured in these experiments. Since the responses of MTAL and CTAL to ET-1 were similar with regard to JCl, cAMP and [Ca2+]i (see below), the data have been obtained indiscriminately from one or the other segment, and classified as TAL, in the following experiments.

Figure 1.

Inhibitory effects of endothelin on net chloride reabsorption in MTAL and CTAL

After a control period (stippled bars), the bath either remained unchanged, for time control tubules (Control), or 10−8m endothelin 1 (ET-1, hatched bars) was added, (n), number of tubular samples. Significantly different from the preceding period: *P < 0.05; †P < 0.001.

When concentrations of ET-1 from 10−13m to 10−8m were tested on separate groups of tubules, a dose-dependent effect was observed (Fig. 2). Endothelin 1-induced inhibition of JCl was reversible as JCl increased significantly by 21.4 ± 3.4% after removal of the agent (from 76.5 ± 10.6 to 91.0 ± 11.1 pmol min−1 mm−1, n= 8, P < 0.001, Fig. 3). However, when the control and recovery periods were performed during individual experiments (Fig. 3), the reversibility was found to be only partial, since the JCl remained significantly (P < 0.01) lower after ET-1 removal compared with initial values (105.0 ± 10.6, 68.9 ± 11.9, and 82.1 ± 12.0 pmol min−1 mm−1, before, during, and after the presence of ET-1, respectively).

Figure 2.

Inhibitory effects of different concentrations of endothelin 1 on net chloride reabsorption

The ordinate represents the decrease of JCl in the experimental period as a percentage of the control value, determined in 4 separate groups of TALs. C, control group, (n), number of tubules. Significance of difference from the preceding value: *P < 0.05, **P < 0.01.

Figure 3.

Reversibility of the endothelin effect on net chloride reabsorption, in individual TALs

After a control period (C), endothelin 1 (ET) was added to the bath at a concentration of 10−8 M, and then removed from the bath (R).

The diminution of JCl induced by 10−8M ET-1 (Fig. 4, column 2) was prevented by BQ788, a selective antagonist of the ETB receptor (Fig. 4, column 3, from 103.7 ± 19.5 to 97.4 ± 16.8 pmol min−1 mm−1; n.s.). This inhibitory effect of ET-1 (10−8m) was also obtained by the presence of 10−11m ET-3 (Fig. 4, column 4, from 133.6 ± 24.0 to 98.9 ± 18.0 pmol min−1 mm−1, P < 0.02), and 10−8m sarafotoxin 6C, a specific agonist of the ETB receptor, (Fig. 4, column 5, from 85.5 ± 18.8 to 61.1 ± 17.7 pmol min−1 mm−1).

Figure 4.

Effects of endothelin agonists and antagonists on net chloride reabsorption, in TAL

The ordinate represents the decrease of JCl in the experimental period as a percentage of the control value. All agents were added to the bath, either during the experimental period in the case of endothelin 1 (ET-1, 10−8m), endothelin 3 (ET-3, 10−11 M), and sarafotoxin 6C (S6C, 10−8 M), or from the control period onwards in the case of BQ788 (10−7 M). (n), number of tubules. Significantly different from the control plus ET-1 + BQ788 values: *P < 0.05, **P < 0.001.

Apical addition of ET-1 (10−8m) induced a 35.1 ± 7.6% inhibition of JCl (Fig. 5, column 1). This effect was also seen with the same concentration of ET-3 (33.2 ± 4.2% inhibition, Fig. 5, column 2). The inhibitory effect of endothelin was eliminated by a simultaneous luminal administration of BQ788 (2.6 ± 3.3% inhibition, Fig. 5, column 3) whilst addition of the receptor antagonist to the bath was without effect (34.5 ± 6.1% inhibition, Fig. 5, column 4). Simultaneous additions of endothelin to the bath and the perfusate (Fig. 5, column 5) failed to induce a greater effect (28.9 ± 3.2%) than that elicited by the agent present on only one or the other of the two sides.

Figure 5.

Effect of luminal endothelin 1 or 3 on net chloride reabsorption, in TAL

Endothelin was used at a concentration of 10−8m for both ET-1 and ET-3 in the lumen (ET-1, 1 and ET-3, l). BQ788 (10−7 M) was added either to the lumen (BQ, l) simultaneously with ET (ET-1 or ET-3) during the experimental period, or to the bath (BQ, b), from the control period onwards. ET, l + ET, b: endothelin was added simultaneously to the bath and perfusate. *Significantly different (P < 0.05) from the other groups.

In the presence of 3 × 10−6m Ibuprofen, an inhibitor of constitutive and inducible cyclo-oxygenase activities, PT-1 significantly decreased JCl by 29.3 ± 4.0% (n= 5), a value similar to that obtained in the absence of prostaglandin synthesis inhibitor.

Absence of an endothelin effect on the cAMP pathway

Accumulation of cAMP was unaffected by ET-1 under either basal or ADH-stimulated conditions (Table 1). This lack of an ET-1 effect was shown not to be due to damaged tubules or to altered hormone solutions. As expected, the ADH-stimulated cAMP accumulation was inhibited by PGE2 in the TAL, and by ET-1 in the cortical collecting duct. In addition, the effect of ET-1 on JCl was not potentiated by the presence of glucagon, which like ADH stimulates adenylyl cyclase activity in TAL (Morel, Chabardès, Imbert-Teboul, Le Bouffant, Hus-Citharel & Montégut, 1982) since there was a 28.5 ± 5.3% inhibition (n= 7) in the presence of glucagon and 33.9 ± 3.2% (n= 17, n.s.) in its absence.

Table 1. cAMP contents of isolated thick ascending limbs and cortical collecting ducts
 TAL n CCD n
  1. cAMP content measured in fmol (4 min)−1 mm−1. Values are means ± S.E.M. (n), number of experiments. For each experiment, 5 to 8 tubule samples per condition were tested. Cyclic AMP accumulation was determined in the presence of Ro 201724, an inhibitor of the phosphodiesterases IV. ET-1, endothelin 1, 10−7 M; ADH, 8 arginine vasopressin, 10−8 M; PGE2, prostaglandin E2, 3 × 10−7 M. n.t., not tested. *P < 0.001 relative to the basal value, †P < 0.01 relative to ADH.

Basis3.7 ± 0.5(15)5.8 ± 2.0(5)
ET-17.6 ± 2.4(8)5.9 ± 3.3(4)
ADH23.7 ± 3.8*(13)66.3 ± 9.6(6)
ADH + ET-126.1 ± 5.8*(10)21.4 ± 6.4†(6)
ADH + PGE26.6 ± 0.7(5)n.t. 

Transduction pathway involved in the effect of endothelin on JCl

There was no effect of the addition of basolateral ET-1 when protein kinase C inhibitors (staurosporine or GF 109203) were also present at a concentration of 10−7m, from the control period onwards (Table 2). Indeed, the slight decrease of JCl observed in the presence of protein kinase C inhibitors was not significantly different from that observed in time-control experiments (3.0 ± 2.4%). However, the inhibitory effect of basolateral ET-1 on JCl was also seen with 10−6m of the diacylglycerol analogues OAG or DiC8, since addition to the bath of either of these agents reduced JCl by 31.7 ± 4.4%.

Table 2. Role of protein kinase C in the endothelin effect on net chloride reabsorption, in TAL
  J Cl (pmol min−1 mm−1) 
 ControlExperimentalPercentage inhibition
  1. Values are means ± S.E.M. (n), number of tubule samples. JCl, net chloride reabsorption. ET, endothelin 1 or added at a concentration of 10−8m to either the bath (rows 1 and 2) or the lumen (row 4), during the experimental period. Protein kinase C inhibitors, GF109203 or staurosporine (STP), were present at a concentration of 100−7m in either the bath (rows 1 and 2) or the lumen (row 4) from the control period onwards. OAG and DiC8 were added to the bath at a concentration of 10−8m during the experimental period. Within each group (one row), differences in JCl between control and experimental periods were compared by Student's paired t test (*P < 0.05, **P < 0.01 and ***P < 0.001); between the groups columns), the percentages of inhibition were compared by ANOVA (†P < 0.001 when compared with the three other groups).

Bath   
ET + GF (5)133.9 ± 18.2124.0 ± 16.9*7.4 ± 2.2
ET + STP (5)179.3 ± 41.0166.1 ± 37.1**6.8 ± 1.2
OAG/DiC8 (11)156.3 ± 18.2109.7 ± 15.6***31.7 ± 4.4†
Lumen   
ET + GF (5)111.1 ± 15.4103.5 ± 13.5*6.1 ± 1.6

Inhibition of JCl by luminal ET-3 was abolished by previous luminal addition of GF 109203 (6.1 ± 1.6% inhibition, Table 2). That this experimental procedure did not impair tubular function was verified: in two tubules, the presence of the protein kinase C inhibitor failed to stop a further inhibitory effect of luminal cyclic GMP (35.9 and 23.3% inhibition; Néant & Bailly, 1993). It is noteworthy that the presence of GF 109203 in the bath failed to suppress the luminal effect of ET-3 (41.0 ± 7.5% inhibition, n= 4), although it reduced the basolateral effect of the hormone in the same tubule.

To determine whether ET-1 stimulates protein kinase C through activation of phospholipase C, the effect of ET-1 on cytosolic calcium was evaluated. In TAL, [Ca2+]i remained unchanged after addition of 10−8m ET-1 to the bath (n= 16, Fig. 6A). As a test of tubule viability, [Ca2]i responses were measured following activation of calcium sensors by peritubular addition of 5 mm calcium (Paulais, Baudoin-Legros & Teulon, 1996) or of purinoreceptors by 10−4m ATP (Paulais, Baudoin-Legros & Teulon, 1995) both of which increased [Ca2+]i by 273 ± 41 nm (n= 11) and 254 ± 66 nm (n= 5), respectively. It should be noted that in the rat also, endothelin failed to increase [Ca2+]i although a peak could be provoked by addition of 10−8m angiotensin II (data not shown). Moreover, as a further control, it was verified that ET-1 elicited a [Ca2+]i increment of 172 ± 54 nm (n= 5) in the mouse cortical collecting duct.

Figure 6.

Recordings from two representative experiments of determination of intracellular calcium concentration ([Ca2+]i), in TAL

A, endothelin 1 (ET-1, 10−8 M) and adenosine triphosphate (ATP, 10−4 M) were present in the bath. B, 5 mm calcium (5 Ca) was present in the bath; endothelin 3 (ET-3,10−8 M) was present in the lumen (1 μm).

Similarly, 10−8m ET-3 did not alter [Ca2+]i when added to the luminal side of TAL (n= 5, Fig. 6B). With the same tubule, however, increasing the bath calcium concentration from 1 to 5 mm evoked a [Ca2+]i peak of 204 ± 32 nm above the basal value.

DISCUSSION

The results of this study demonstrate that endothelin, whether added luminally or basolaterally, inhibits the reabsorptive function of MTAL and CTAL. In contrast to mechanisms characterized in other target cells, the transduction pathway in these kidney cells does not involve inhibition of cAMP production but is related to activation of a protein kinase C which, however, is independent of increases in intracellular calcium.

Effect of endothelin on JCl

The decrease of JCl observed in the presence of ET-1 is unlikely to result from a toxic action of the peptide as this effect is inhibited by the use of a specific antagonist. Relevant to the partial reversibility of the effect of endothelin are the very slow rates of dissociation of receptor–ligand complexes reported in many structures (for review see Sokolovsky, 1995), including renal papillary membranes (Martin, Marsden, Brenner & Ballermann, 1989). Our study presents pharmacological evidence to indicate that the ET-1-induced inhibition of JCl occurs through the ETB receptor. Moreover, the fact that JCl inhibition can be induced by ET-1 precludes the involvement of ETC receptors. Terada et al. (1992) observed a small signal for the ETB receptor whereas other workers did not detect ETA or ETB receptor mRNA (Chow et al. 1995). Our study confirms the presence of functional ETB receptors in the mouse CTAL and MTAL and also emphasizes either species differences or a translation efficiency sufficient to allow protein expression from very low levels of mRNAs in this segment. It should be noted that mRNAs encoding the ETB receptor have not been detected in the proximal tubule (Terada et al. 1992), yet ET-1 action in this segment is well-documented (Garcia & Garvin, 1994; Guntupalli & DuBose, 1994; Walter, Helmle-Kolb, Forgo, Binswanger & Murer, 1995) and is shown to be partly linked to activation of the ETB receptor (Knotek, Jaksic, Selmani, Skoric & Banfic, 1996). It must be kept in mind, however, that although mRNA encoding ETC has not been detected in murine species this is because this receptor has not yet been cloned in mammals.

A significant inhibitory effect of ET-1 in TAL was observed at a concentration of 10−13 M, which suggests that this action is of physiological relevance since: (i) the threshold for proximal tubule effects is 10−13m (Garcia & Garvin, 1994) which is three orders of magnitude lower than the threshold found in the collecting duct (Tomita et al. 1991), both segments being well-recognized target sites of endothelin; and (ii) detectable amounts of ET-3 and ET-1 have been found in rat renal medulla, under basal conditions for the former (Terada et al. 1993) and following stimulation by inflammatory agents for the latter (Ujiie et al. 1992). Thus these results, taken together, strongly suggest that ET-3 and ET-1 act as autocrine and/or paracrine agents in TAL.

Transduction pathway involved in the endothelin effect in TAL

The present data indicate that the ET-1 inhibitory effect on JCl in TAL is not coupled to an inhibition of cAMP production. It may be assumed that both basolateral and apical ET-1 fail to impair generation of this nucleotide, because the luminal surface is presumably accessible to ET-1 in cAMP experiments as firstly, microscopic observation revealed that most microdissected tubules displayed a fairly open lumen, and secondly, numerous studies report that electrolyte transport through the apical Na+–K+–2Cl cotransporter in MTAL suspensions is inhibited by incubation with selective blockers such as furosemide and bumetanide, suggesting that they can freely reach the luminal side of the tubules. The lack of effect of ET-1 on cAMP production in mouse TAL is in agreement with a previous report by Tomita (Tomita, Nonogushi & Marumo, 1990) concerning the rat nephron. In this latter study, however, cAMP content was determined while phosphodiesterase activity was fully inhibited by IBMX, indicating that endothelin fails to modify adenylate cyclase activity. The present experiments, in which phosphodiesterase activity was partially preserved, show that endothelin is unlikely to act on cAMP degradation. The observation that in TAL ET-1 did not affect cAMP production shows that the process here differs from that of the collecting duct in which ET-1 prevented the ADH-stimulated increase in water permeability at the cAMP generation step (Tomita et al. 1991; Nadlere et al. 1992).

In many target cells, including renal medullary interstitial cells (Wilkes et al. 1991) and inner medullary collecting duct cells (Kohan, Padilla & Hughes, 1993), endothelin has been shown to generate arachidonic acid and metabolites including PGE2 which may explain ET-1 actions on TAL. However, this is unlikely because (i) the ET-1 effects were unchanged after cyclo-oxygenase inhibition, and (ii) both arachidonic acid and PGEj are known to attenuate ADH-stimulated cAMP accumulation in this segment (Torikai & Kurokawa, 1983; Firstov et al. 1995).

The results presented here provide evidence for the activation of protein kinase C by endothelin. Surprisingly, administration of the phorbol ester PMA at a concentration of 10−7m does not alter JCl in mouse CTAL, as already reported by us (Néant, Imbert-Teboul & Bailly, 1994). A similar anomaly has been observed for proximal tubule, in that PMA failed to produce the ET-1-induced increase in the Na+–H+ exchanger although this effect was consistently suppressed by protein kinase C inhibitors (Guntupalli & DuBose, 1994; Walter et al. 1995). The presence of other effects of phorbol esters in addition to protein kinase C activation may explain these discrepancies.

There is clear evidence from this work that basolateral or luminal ET-1 does not stimulate [Ca2+]i increases in TAL, in contrast to its observed effect made in collecting tubules (Naruse, Uchida, Ogata & Kurokawa, 1991, and our results). This lack of an ET-1 effect on [Ca2+]i is in agreement with the report concerning basolateral ET-1 in mouse TAL (Naruse et al. 1991). To our knowledge, the present results show for the first time a functional effect of endothelin at a target site where the hormone does not evoke intracellular calcium mobilization. This absence of endothelin-induced [Ca2+]i increase probably excludes a possible activation of the so-called classical protein kinase C isoforms, as a consequence of phospholipase C stimulation. Rather, the results afford evidence that endothelin action is mediated by the so-called novel protein kinase C isoforms, the activation of which requires diacyl glycerol but not [Ca2+]i increases. Moreover, the recent demonstration in the renal outer medulla of several protein kinase C isoforms, including an atypical one independent of calcium and phorbol ester (Östlund et al. 1995), raises the possibility that endothelin activates such a protein kinase C, in addition to the phorbol ester-activated isoforms. The mechanism underlying protein kinase C activation requires further investigation of the direct stimulation of phospholipase D by endothelin. Some data show that ET-1 induces phospholipase D stimulation in renal structures (Wilkes et al. 1991; Baldi, Musial & Kester, 1994). In these studies, however, phospholipase D was activated by protein kinase C and was, thereby, primarily linked to phospholipase C induction.

Polarization of the endothelin effect on JCl

It is interesting that endothelin inhibits JCl when it is added on either the basolateral or the luminal side of the tubule. The fact that the TAL epithelium is tight and that the apical effect of endothelin occurs in the presence of a receptor antagonist in the bath allows us to exclude a possible diffusion of endothelin from the lumen to the bath. A luminal action of ET-3, implying ETB receptors, has recently been demonstrated in proximal tubules while the agonist action on the basolateral side was associated with ETC receptors (Knotek et al. 1996). On the other hand, the present study strongly suggests that the same receptor subtype, namely ETB, is present on both sides of the epithelium. Moreover, since luminal and basolateral endothelin effects are not additive, it is probable that they are both mediated via similar transduction pathways. However, differences in signal transduction located at a step proximal to kinase activation remain to be elucidated.

In conclusion, the present work indicates that ET-1 and ET-3 exert an inhibitory action on MTAL and CTAL function after either basolateral or luminal addition. This effect is linked to ETB receptors, unrelated to cAMP pathways, and mediated by a protein kinase C activation independent of increases in intracellular calcium. The effect might, in part, contribute to the potentiation of the natriuretic and diuretic action of the hormone on the collecting duct. However, the respective roles of ET-1 and ET-3 in physiological and pathological conditions remain to be elaborated.

Acknowledgements

This work was supported by grants from the Centre National de la Recherche Scientifique and from the Commissariat à l'Energie Atomique. The technical assistance of Huguette Moysan is gratefully acknowledged.

Ancillary