Endothelin-1 Inhibits Thick Ascending Limb Transport via Akt-stimulated Nitric Oxide Production

Abstract

Endothelin-1 inhibits sodium reabsorption in the thick ascending limb (THAL) via stimulation of nitric oxide (NO) production. The mechanism whereby endothelin-1 stimulates THAL NO is unknown. We hypothesized that endothelin-1 stimulates THAL NO production by activating phosphatidylinositol 3-kinase (PI3K), stimulating Akt activity, and phosphorylating NOS3 at Ser¹¹⁷⁷. This enhances NO production and inhibits sodium transport. We measured 1) NO production by fluorescence microscopy using DAF2-DA, 2) Akt activity using a fluorescence resonance energy transfer-based Akt reporter, 3) phosphorylated NOS3 and Akt by Western blotting, and 4) NKCC2 activity by fluorescence microscopy. In isolated THAL, endothelin-1 (1 nmol/liter) increased NO production from 0.23 ± 0.24 to 2.81 ± 0.32 fluorescence units/min (p < 0.001; n = 5) but failed to stimulate NO production in THALs isolated from NOS3^–/– mice. Wortmannin (150 nmol/liter), a PI3K inhibitor, reduced endothelin-1-stimulated NO by 83% (0.49 ± 0.13 versus 3.31 ± 0.49 fluorescence units/min for endothelin-1 alone; p < 0.006; n = 5). Endothelin-1 stimulated Akt activity by 0.16 ± 0.02 arbitrary units as measured by fluorescence resonance energy transfer (p < 0.001; n = 5) and increased phosphorylation of Akt at Ser⁴⁷³ by 56 ± 11% (p < 0.002; n = 7). Dominant-negative Akt blocked endothelin-1-induced NO by 60 ± 8% (p < 0.001 versus control; n = 6), and an Akt inhibitor had a similar effect. Endothelin-1 increased phosphorylation of NOS3 at Ser¹¹⁷⁷ by 89 ± 24% (p < 0.01; n = 7) but had no effect on Ser⁶³³. Endothelin-1 inhibited NKCC2 activity, an effect that was blocked by dominant-negative Akt and NOS inhibition. We conclude that endothelin-1 stimulates THAL NO production by activating PI3K, stimulating Akt activity, and phosphorylating NOS3 at Ser¹¹⁷⁷. This enhances NO production and inhibits sodium transport.

Nitric oxide (NO) augments salt and water excretion by the kidney (1–6). NO produced by both NOS1 and NOS3 (neuronal and endothelial NOS2) contributes to this effect (7–9). Endothelin-1 appears to be one factor that stimulates NO production by both enzymes in the kidney (7–10). Inhibition of endothelin-induced NOS activation can cause salt-sensitive hypertension (6). The thick ascending limb reabsorbs ∼30% of the filtered NaCl, and improper regulation of sodium reabsorption by this segment has been implicated in salt-sensitive hypertension (11, 12). Thus, studying the effects of endothelin-1 on the thick ascending limb is physiologically significant.

Endothelin-1 inhibits thick ascending limb NaCl reabsorption via stimulation of NO (9). NO has been shown to inhibit apical Na⁺-K⁺-2Cl^– co-transport (NKCC2) (13), the main route for sodium entry in this segment and the first step in NaCl absorption (14, 15). The thick ascending limb expresses all three NOS isoforms. The actions of endothelin-1 are likely due to NOS3 activation because 1) this isoform is responsible for regulating thick ascending limb NaCl reabsorption (8), and 2) endothelin-1 stimulates NOS3 expression in the thick ascending limb (16). However, whether endothelin-1 acutely stimulates NO production via NOS3 activation in the thick ascending limb is uncertain.

NOS3 can be activated by several signaling pathways, including those that involve Ca²⁺/calmodulin and phosphatidylinositol 3-kinase (PI3K). In endothelial cells, both pathways are important. However, in the thick ascending limb, only the latter has been shown to activate NOS3 (17, 18). Thus, the signaling cascades that activate NOS3 in the thick ascending limb and endothelial cells likely differ (19). The mechanisms by which endothelin-1 stimulates NOS3 and inhibits sodium transport in this segment are unknown. We hypothesized that endothelin-1 stimulates thick ascending limb NO production by activating PI3K, stimulating Akt activity, and phosphorylating NOS3 at Ser¹¹⁷⁷. This enhances NO production and inhibits sodium transport.

EXPERIMENTAL PROCEDURES

Animals—Male Sprague-Dawley rats (Charles River, Kalamazoo, MI) and C57BL/6J and NOS3 knock-out (–/–) mice (The Jackson Laboratory, Bar Harbor, ME) were fed a diet containing 0.22% sodium and 1.1% potassium (Purina, Richmond, IN) for at least 7 days. On the day of the experiment, animals were anesthetized with ketamine (100 mg/kg of body weight intraperitoneally) and xylazine (20 mg/kg of body weight intraperitoneally).

Medullary Thick Ascending Limb Suspensions—Suspensions of thick ascending limbs were obtained from rats weighing 150–220 g as reported previously (20). A solution containing 130 mmol/liter NaCl, 2.5 mmol/liter NaH₂PO₄, 4 mmol/liter KCl, 1.2 mmol/liter MgSO₄, 6 mmol/liter alanine, 1 mmol/liter disodium citrate, 5.5 mmol/liter glucose, 2 mmol/liter calcium lactate, and 10 mmol/liter HEPES (pH 7.4) was used (Solution A). This procedure yields a 92% pure suspension of thick ascending limbs (16), and therefore, the contributions of other cell types in our preparation (if any) were minimal.

Measurement of NO Production by Fluorescence Microscopy—One-tenth of the thick ascending limb suspension was seeded at 4 °C in a chamber designed for live cell imaging on the stage of an inverted microscope (Nikon Eclipse TE-2000-U). After 5 min, the bath was started at 0.3 ml/min, and the chamber was warmed to 37.0 ± 0.5 °C. The bath was Solution A plus 50 μmol/liter l-arginine, the substrate for NOS. At this concentration, l-arginine supports NO production but does not alone stimulate NO production (21). Tubules were loaded with the NO-selective fluorescent dye DAF2-DA (5 μmol/liter; EMD Biosciences, Gibbstown, NJ) for 15 min, followed by a 30-min wash. Tubules were imaged using a 100× oil immersion objective (numerical aperture = 1.3), and the dye was excited with an argon laser at 488 nm. The fluorescence emitted by NO-bound dye (>500 nm) was measured using a laser scanning confocal microscope equipped with data acquisition and analysis software (VisiTech International). Measurements were recorded once every 30 s for a 5-min control period. Either endothelin-1 (1 nmol/liter; Bachem, Torrance, CA) or vehicle (0.005% acetic acid) was then added to the bath. Fluorescence was measured once every 30 s during a 15-min experimental period. NO production was calculated from the slope of the initial rate of increase in DAF2 fluorescence. In some experiments, wortmannin (a PI3K inhibitor; 150 nmol/liter; Sigma) or Akt inhibitor IV (5 μmol/liter; EMD Biosciences) was added to the bath during the last 5 min of the washing period as indicated below. We did not use isolated and perfused thick ascending limbs because we have reported that flow stimulates NO production by this segment (18).

Measurement of NOS3 and Akt Phosphorylation by Western Blotting—Thick ascending limb suspensions obtained from the same rat were divided into four 2-ml centrifuge tubes and incubated for 15 min in 250 μl of Solution A at 37 °C. Then, 250 μl of Solution A were added to tubes 1 and 2, whereas 250 μl of Solution A plus wortmannin were added to tubes 3 and 4 (final concentration of 150 nmol/liter). All tubes were incubated for 5 min at 37 °C. Then, 500 μl of Solution A containing (final concentration) 1) 0.005% acetic acid (vehicle), 2) endothelin-1 (1 nmol/liter), 3) wortmannin (150 nmol/liter), and 4) endothelin-1 plus wortmannin were added to the tubes, respectively. After a 15-min incubation at 37 °C, suspensions were cooled by adding 1 ml of ice-cold Solution A to each tube. Tubules were centrifuged and lysed by vortexing in 100 μl of buffer containing 20 mmol/liter HEPES (pH 7.4), 2 mmol/liter EDTA, 300 mmol/liter sucrose, 1.0% Nonidet P-40, 0.1% SDS, 5 μg/ml anti-pain, 10 μg/ml aprotinin, 5 μg/ml leupeptin, 4 mmol/liter benzamidine, 5 μg/ml chymostatin, 5 μg/ml pepstatin A, and 0.105 mol/liter 4-(2-aminoethyl)benzenesulfonyl fluoride (Sigma). A 1:100 dilution of phosphatase inhibitor mixture II (EMD Biosciences) was added to the buffer prior to use. Samples were centrifuged at 6000 × g for 5 min at 4 °C, and protein content in the supernatant measured. For total and phosphorylated NOS3, 5 and 50 μg of total protein, respectively, were loaded into each lane of an 8% SDS-polyacrylamide gel, separated by electrophoresis, and transferred to a polyvinylidene difluoride membrane (Millipore, Bedford, MA). The membrane was incubated in blocking buffer containing 20 mmol/liter Tris, 137 mmol/liter NaCl, 5% nonfat dried milk, and 0.1% Tween 20 for 60 min and then with a 1:1000 dilution of a NOS3-specific monoclonal antibody (BD Transduction Laboratories), a 1:1000 dilution of a monoclonal antibody against NOS3 phosphorylated at Ser¹¹⁷⁷ (BD Transduction Laboratories), or a 1:500 dilution of a polyclonal antibody against NOS3 phosphorylated at Ser⁶³³ (Upstate, Lake Placid, NY) in blocking buffer for 2 h at room temperature. The membrane was washed with buffer containing 20 mmol/liter Tris, 137 mmol/liter NaCl, and 0.1% Tween 20 and incubated with a 1:1000 dilution of a secondary antibody against the appropriate IgG conjugated to horseradish peroxidase (Amersham Biosciences). The reaction products were detected with a chemiluminescence kit (Amersham Biosciences). Chemiluminescence was detected by exposure to Fuji RX film and quantified by densitometry. For phosphorylated Akt, the process was similar except that 1) 5 μg of protein and 2) a 1:2000 dilution of a polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA) were used to detect Akt phosphorylated at Ser⁴⁷³. Total NOS3 was used as control for loading. The data are expressed as percent change in the protein of interest/NOS3 at each time point.

Akt Reporter—Akt activity was measured using a fluorescence resonance energy transfer (FRET)-based Akt reporter. The Akt reporter BKAR was generously provided by Drs. Alexandra C. Newton and Maya T. Kunkel (Howard Hughes Medical Institute, University of California at San Diego) (22). An increase in cyan fluorescent protein (CFP)/yellow fluorescent protein (YFP) ratio indicates increased Akt activity. The probe was subcloned into the adenoviral shuttle vector pVQ Ad5CMV K-NpA, which contains the cytomegalovirus promoter, and sent to ViraQuest (North Liberty, IA) for viral production.

Dominant-negative Akt—Dominant-negative Akt plasmid was kindly provided by Dr. Kenneth Walsh (Boston University School of Medicine). Dominant-negative Akt is a mutant form (T308A/S473A) that cannot be activated by phosphorylation (23). Dominant-negative Akt and a scrambled DNA (control) were subcloned into the adenoviral shuttle vector pVQ Ad5CMV K-NpA and sent to ViraQuest for viral production.

In Vivo Gene Delivery of the FRET-based Akt Reporter and Dominant-negative Akt—Thick ascending limbs were transduced in vivo with adenoviruses expressing 1) an Akt reporter, 2) dominant-negative Akt, or 3) a scrambled DNA sequence (all driven by a cytomegalovirus promoter) as we reported previously (24, 25). Briefly, the left kidney of a 95–105-g rat was exposed via a flank incision, and the renal artery and vein were clamped. Four 20-μl virus injections (1 × 10¹² particles/ml) were performed along the longitudinal axis of the kidney at a flow rate of 20 μl/min. The clamp on the renal vessels was then released, the kidney was returned to the abdominal cavity, the muscle incision was sutured, and the skin was clipped. Expression of the reporter and dominant-negative Akt in the thick ascending limb was investigated by Western blotting at 3, 5, and 7 days after gene transfer. We found that maximum expression occurred 3–5 days after injection. Thus, all experiments were performed within these time points.

Measurement of Akt Activity by FRET—The day of the experiment, thick ascending limb suspensions were obtained from the transduced kidney as indicated above, and one-fifth of the suspension was seeded in a temperature-controlled chamber and warmed to 37 °C. The flow rate of the bath (Solution A plus 50 μmol/liter arginine) was 0.3 ml/min. During the 30-min equilibration period, images were acquired (100× oil objective, numerical aperture = 1.3) by alternately exciting CFP (442 nm) and YFP (514 nm) and monitoring YFP emission at 540 nm to determine expression of the FRET sensor and to draw regions of interest. During the control period, CFP/YFP emission ratios were measured by exciting CFP at 442 nm once/min for 5 min and simultaneously monitoring CFP and YFP emissions at 440–480 nm (CFP) and 540–545 nm (YFP). At the end of the control period, endothelin-1 was added to the bath, and the CFP/YFP ratio was monitored once/min for 15 min. The averages corresponding to the 5-min control period and the last 5 min of the experimental period were compared. When wortmannin (150 nmol/liter) was used, it was added to the bath during the equilibration period. To confirm that the YFP signal was due to FRET, control experiments were performed by photobleaching CFP and measuring the decrease in YFP emission. Images were acquired using the same settings (laser intensity, detector gain and offset, resolution, and exposure time).

Measurement of Na⁺-K⁺-2Cl^– Co-transporter (NKCC2) Activity—Tubules were isolated and perfused as routinely done in our laboratory at 37 °C (26). Thick ascending limbs were bathed for 15 min in Solution A containing 1 μmol/liter sodium green (Invitrogen) and washed for 20 min with Solution A containing 50 μmol/liter l-arginine. Tubules were perfused with a solution designed to prevent Na⁺-K⁺-2Cl^– co-transport activity that contained 2.5 mmol/liter NaH₂PO₄, 1.2 mmol/liter MgSO₄, 6 mmol/liter l-alanine, 1 mmol/liter disodium citrate, 5.5 mmol/liter glucose, 2 mmol/liter calcium lactate, 10 mmol/liter HEPES, 260 mmol/liter mannitol, and 100 μmol/liter dimethlyamiloride (to inhibit sodium/hydrogen exchange). During the control period, sodium green was excited at 488 nm every 5 s, and fluorescence greater than 515 nm was digitally imaged and recorded utilizing Metafluor software (Universal Imaging, West Chester, PA). Once a stable base line was obtained, the luminal perfusate was switched to one containing 140 mmol/liter sodium, 134 mmol/liter chloride, and 4 mmol/liter potassium. The initial increase in intracellular sodium caused by the switch is due to activation of Na⁺-K⁺-2Cl^– co-transport and was used to calculate NKCC2 activity. The luminal solution was then switched back to one containing 4.5 mmol/liter sodium, 0 mmol/liter chloride, and 0 mmol/liter potassium. After 15 min, endothelin-1 (1 nmol/liter) was added to the bath. Ten min later, the luminal solution was switched again. In some experiments, the protocol was reversed. When 5 mmol/liter l-N^G-nitroarginine methyl ester (l-NAME) was used, it was present throughout the experiment.

Statistics—Results are expressed as the mean ± S.E. All statistical analyses were performed by the Biostatistics Department at Henry Ford Hospital. Student's t test was used to analyze data in Figs. 1, 2, 3, 5, and 7. In experiments in which the variance differed, Welch's version of Student's t test, which adjusts the degrees of freedom to account for the difference in variances, was used. A paired t test with Hochberg's method for multiple comparisons was used to analyze data in Figs. 4 and 6. p < 0.05 was considered significant.

FIGURE 1.

A, effect of 1 nmol/liter endothelin-1 (ET-1) on NO production by isolated thick ascending limbs from wild-type (WT) and NOS3^–/– mice (n = 4). B, effect of 1 nmol/liter endothelin-1 on NO production by isolated thick ascending limbs from rats (n = 5).

FIGURE 2.

Effect of 150 nmol/liter wortmannin, a PI3K inhibitor, on endothelin-1-stimulated NO production by isolated thick ascending limbs. ET-1, endothelin-1 (n = 5).

FIGURE 3.

Effect of 1 nmol/liter endothelin-1 and 150 nmol/liter wortmannin, a PI3K inhibitor, on Akt activity measured by FRET in isolated thick ascending limbs. A, representative traces of CFP and YFP fluorescence in response to endothelin-1 (ET-1). B, representative trace of the CFP/YFP ratio calculated from the experiment in A. C, mean data (n = 5).

FIGURE 5.

A, effect of 1 nmol/liter endothelin-1 on NO production by thick ascending limbs that have been transduced in vivo with adenoviruses expressing a scrambled DNA sequence (scr) or dominant-negative Akt (dn-Akt) (n = 6). B, effect of 5 μmol/liter Akt inhibitor IV on endothelin-1 (ET-1)-stimulated NO production by isolated thick ascending limbs from non-transduced rats (n = 4).

FIGURE 7.

Effect of 1 nmol/liter endothelin-1 on NKCC2 activity in 1) thick ascending limbs in the absence (control) and presence of the NOS inhibitor l-NAME (n = 4) and 2) tubules isolated from kidneys that have been transduced in vivo with adenoviruses expressing dominant-negative Akt (dn Akt) or a scrambled DNA sequence (control) (n = 5).

FIGURE 4.

Effect of 1 nmol/liter endothelin-1 and 150 nmol/liter wortmannin on Akt phosphorylation at Ser⁴⁷³ in isolated thick ascending limbs. Changes in phosphorylation of Akt at Ser⁴⁷³ (pAKT-Ser⁴⁷³) were normalized by total NOS3. Upper panel, representative Western blot; lower panel, mean data (n = 7). C, control (vehicle); ET-1, endothelin-1; W, wortmannin.

FIGURE 6.

Effect of 150 nmol/liter wortmannin, a PI3K inhibitor, on endothelin-1-induced NOS3 phosphorylation at Ser¹¹⁷⁷ in thick ascending limbs suspensions. Changes in phospho-NOS3 were normalized by total NOS3. Upper panel, representative Western blot; lower panel, mean data (n = 7). C, control (vehicle); ET-1, endothelin-1; W, wortmannin; pNOS3-Ser¹¹⁷⁷, NOS3 phosphorylated at Ser¹¹⁷⁷.

RESULTS

To test our hypothesis, we first investigated whether endothelin-1 increases NO via NOS3 activation by measuring changes in DAF2 fluorescence in thick ascending limbs isolated from wild-type and NOS3^–/– mice. In isolated thick ascending limbs from wild-type mice, 1 nmol/liter endothelin-1 increased fluorescence by 3.81 ± 0.46 fluorescence units/min. In contrast, endothelin-1 failed to stimulate NO production by tubules from NOS3^–/– mice (Δ= 0.38 ± 0.22 fluorescence units/min; p < 0.001 versus wild-type mice; n = 4) (Fig. 1A). These data indicate that endothelin-1 enhances NO production by thick ascending limbs via activation of NOS3 and that DAF2 is measuring NO.

Because the remainder of our experiments were performed using rats, we repeated the above experiments in rat isolated thick ascending limbs. We found that 1 nmol/liter endothelin-1 increased NO production by 2.81 ± 0.32 fluorescence units/min, whereas vehicle had no effect (0.23 ± 0.24 fluorescence units/min; p < 0.001; n = 5) (Fig. 1B). In the absence of endothelin-1, there was no change in NO, indicating that there was no photobleaching of the dye.

We next studied the mechanism by which endothelin-1 stimulates NO production. Previously, we have shown that other factors that stimulate NO production in the thick ascending limb act via PI3K (17, 18). Consequently, we first studied the effect of wortmannin, a PI3K inhibitor. In isolated thick ascending limbs, 1 nmol/liter endothelin-1 increased NO production by 3.31 ± 0.49 fluorescence units/min. In the presence of wortmannin (150 nmol/liter), endothelin-1 increased NO production by only 0.49 ± 0.13 fluorescence units/min (p < 0.004 versus endothelin-1 alone; n = 5) (Fig. 2), an inhibition of 83%. Wortmannin alone did not change basal fluorescence. These data suggest that endothelin-1 stimulates thick ascending limb NO production via PI3K activation.

Akt mediates most of the actions of PI3K (27–30). To study how PI3K stimulates NO production, we first measured Akt activity and phosphorylation. In isolated transduced thick ascending limbs, 1 nmol/liter endothelin-1 increased Akt activity by 0.16 ± 0.02 from the base line (p < 0.001; n = 5). The PI3K inhibitor wortmannin (150 nmol/liter) completely blunted this effect (0.02 ± 0.02 from the base line; n = 5) (Fig. 3). Control experiments showed no effect of time or wortmannin alone. Control experiments also showed that fluorescence emission from YFP was due to FRET from CFP rather than direct excitation of the dye. We also measured Akt phosphorylation and found that endothelin-1 increased phosphorylation at Ser⁴⁷³ by 53 ± 10% (p < 0.002; n = 7). In the presence of wortmannin (150 nmol/liter), endothelin-1 did not increase Akt phosphorylation (Δ= 6 ± 5% versus wortmannin alone; n = 7) (Fig. 4). Wortmannin alone significantly reduced Akt phosphorylation by 76 ± 3% (p < 0.001; n = 7). These data suggest that endothelin-1 enhances Akt activity via phosphorylation. Moreover, PI3K appears to regulate the basal phosphorylation state of Akt.

To study whether endothelin-1 stimulates NO production via Akt activation, we next investigated the effect of inhibiting Akt on endothelin-induced NO production using dominant-negative Akt and Akt inhibitor IV. In tubules expressing scrambled DNA, endothelin-1 stimulated NO production by 5.14 ± 0.58 fluorescence units/min. Endothelin-1-induced NO production was significantly reduced in tubules expressing dominant-negative Akt (1.87 ± 0.26 fluorescence units/min; p < 0.001 versus scrambled; n = 6) (Fig. 5A). The Akt inhibitor also blocked stimulation of NO by endothelin-1 (Fig. 5B). These data suggest that endothelin-1 stimulates thick ascending limb NO production via activation of Akt.

Phosphorylation of NOS3 at Ser¹¹⁷⁷ and Ser⁶³³ increases NO production (31–33). To investigate the mechanism by which the PI3K/Akt pathway enhances NO production, we measured the effect of endothelin-1 on NOS3 phosphorylation in the presence and absence of wortmannin. Endothelin-1 (1 nmol/liter) increased NOS3 phosphorylation at Ser¹¹⁷⁷ by 85 ± 24% (p < 0.01; n = 6) in thick ascending limb suspensions. In the presence of wortmannin (150 nmol/liter), the ability of endothelin-1 to phosphorylate NOS3 at Ser¹¹⁷⁷ was completely blocked (6 ± 7% versus wortmannin alone; n = 7) (Fig. 6). However, wortmannin alone significantly reduced NOS3 phosphorylation at Ser¹¹⁷⁷ by 38 ± 10% (p < 0.001; n = 7). Endothelin-1 did not affect phosphorylation of Ser⁶³³. These data suggest that endothelin-1 stimulates NOS3 by phosphorylating Ser¹¹⁷⁷ via activation of the PI3K pathway. Moreover, PI3K appears to regulate the basal phosphorylation state of NOS3.

To investigate the physiological significance of the endothelin-1/PI3K/Akt/NO pathway, we measured the effect of l-NAME and dominant-negative Akt on the ability of endothelin-1 to inhibit thick ascending limb transport (Fig. 7). In isolated perfused tubules, 1 nmol/liter endothelin-1 inhibited NKCC2 activity by 41 ± 3% (p < 0.01; n = 4). l-NAME (5 mmol/liter) blocked the inhibitory effect of endothelin-1 on NKCC2 activity (Δ= 9 ± 9%; n = 4). Next, we investigated the ability of endothelin-1 to inhibit NKCC2 activity in tubules expressing dominant-negative Akt. In tubules expressing scrambled DNA, endothelin-1 inhibited NKCC2 activity by 57 ± 6% (p < 0.01; n = 5). In contrast, in tubules expressing dominant-negative Akt, endothelin-1 failed to inhibit NKCC2 activity (17 ± 11%; n = 5). These data indicate that endothelin-1 inhibits NKCC2 activity via activation of Akt and NO in thick ascending limbs.

DISCUSSION

Our hypothesis was that endothelin-1 inhibits thick ascending limb sodium reabsorption by stimulating PI3K, activating Akt, and phosphorylating NOS3 at Ser¹¹⁷⁷, thus enhancing NO production. To test this hypothesis, we first directly measured the effect of endothelin-1 on NO production by the thick ascending limb. We found that endothelin-1 stimulated thick ascending limb NO production.

To investigate whether NOS3 is the NOS isoform responsible for the stimulatory effect of endothelin-1 on NO production, we measured the effect of endothelin-1 on NO production in thick ascending limbs isolated from wild-type and NOS3^–/– mice. We found that endothelin-1 did not stimulate NO production in tubules isolated from NOS3^–/– mice. Thus, endothelin-1 directly stimulates thick ascending limb NOS3 to increase NO production.

After showing that endothelin-1 stimulated NO production, we next addressed the mechanism by which this occurs. We found that the PI3K inhibitor blocked endothelin-1-induced NO production. We then directly measured the effects of endothelin-1 on Akt activity using a FRET-based Akt sensor. We found that endothelin-1 activated Akt in thick ascending limbs and that wortmannin inhibited this effect. Activation was correlated with an increase in Akt phosphorylation, which was also blocked by wortmannin.

The above data indicate that endothelin-1 stimulates Akt via PI3K, but they do not demonstrate that this signaling pathway is responsible for NOS3 activation. To test whether endothelin-1 stimulates NO production via Akt activation, we studied 1) the ability of endothelin-1 to increase thick ascending limb NO production in tubules expressing dominant-negative Akt and 2) the ability of an Akt inhibitor to block the stimulatory effect of endothelin-1. We found that expression of dominant-negative Akt reduced the ability of endothelin-1 to stimulate NO production, whereas expression of a scrambled DNA sequence did not. The Akt inhibitor also blocked the endothelin-1-induced increase in NO production. Taken together, these data show that endothelin-1 stimulates NO production by NOS3 via the PI3K/Akt pathway.

To address the mechanism by which Akt stimulates NOS3 activity, we measured NOS3 phosphorylation at Ser¹¹⁷⁷ and Ser⁶³³, both positive regulatory sites. Treatment with endothelin-1 increased NOS3 phosphorylation at Ser¹¹⁷⁷ but had no effect on NOS3 phosphorylation at Ser⁶³³. Furthermore, wortmannin prevented endothelin-1-induced phosphorylation of NOS3 at Ser¹¹⁷⁷. These data suggest that in the thick ascending limb, endothelin-1 activates the PI3K/Akt pathway, leading to phosphorylation of NOS3 at Ser¹¹⁷⁷ and thus increased NO generation.

To study the physiological relevance of the endothelin-1-stimulated Akt pathway, we next investigated whether NO and Akt mediate the inhibitory effect of endothelin-1 on NKCC2 activity in the thick ascending limb. This is the first and rate-limiting step in NaCl reabsorption in this segment. We found that endothelin-1 inhibited NKCC2 activity by 41%. The inhibitory effect of endothelin-1 was completely blocked by the NOS inhibitor l-NAME. Thus, NO production is required for endothelin-1 to inhibit NKCC2 activity. Dominant-negative Akt also blocked the effect of endothelin-1 on NKCC2 activity, indicating that Akt activation is necessary for endothelin-1 to inhibit NKCC2. These data indicate that endothelin-stimulated NO production via the PI3K/Akt pathway is physiologically important.

Our finding that endothelin stimulates NO production by NOS3 in the thick ascending limb is novel. To our knowledge, there have been no direct studies on the effects of endothelin-1 on this NOS isoform prior to our study. However, these data are consistent with previous findings. We showed that endothelin-1 stimulates NOS3 expression and NO production in primary cultures of thick ascending limbs (16). Nakano et al. (7) reported that endothelin-1 enhances NO production in the renal inner medulla, and Notenboom et al. (34) have shown that it stimulates NO production by isolated proximal tubules. More recently, Stricklett et al. (35) reported that endothelin-1 increases NO production in rat inner medullary collecting duct suspensions.

The finding that endothelin-1 stimulates NO production via the PI3K/Akt pathway appears to be different from endothelial cells. In the latter cell type, NOS3 activation occurs primarily via an increase in intracellular calcium (36). Our conclusion that the PI3K/Akt pathway is the primary mechanism by which endothelin-1 stimulates NO production in the thick ascending limb is supported by several studies. Jesus Ferreira and Bailly (37) reported no change in intracellular calcium in thick ascending limbs challenged with 10 nmol/liter endothelin-1. We have reported that luminal flow activates NOS3 via a PI3K-dependent pathway that involves phosphorylation of NOS3 at Ser¹¹⁷⁷ (18, 38). We also found that the α₂-adrenergic agonist clonidine stimulates thick ascending limb NO production without measurable changes in intracellular calcium (17), but PI3K activation is required.

The finding that endothelin-1 stimulated NOS3 activity by enhancing Ser¹¹⁷⁷ phosphorylation while having no effect on Ser⁶³³ is consistent with other data. Other humoral and mechanical factors have been reported to acutely activate NOS3 via phosphorylation at this residue (30, 38–41), whereas other phosphorylation sites appear to be involved in the chronic regulation of enzyme activity (42).

Our results show that wortmannin alone decreased both Akt and NOS3 phosphorylation. These data suggest that PI3K is an important regulator of Akt activity and possibly sodium transport in the thick ascending limb under basal conditions.

The fact that dominant-negative Akt decreased endothelin-1-stimulated NO production by 60%, whereas pharmacological inhibition of Akt resulted in complete blockade, is not surprising. The most likely explanation of these results is incomplete inhibition of endogenous Akt activity by the dominant-negative mutant. Complete blockade of endothelin-1-stimulated NO by the Akt inhibitor is not likely due to cytotoxicity. We tested for such a possibility and found that at the concentration used in our study, Akt inhibitor IV does not affect viability.

Although dominant-negative Akt reduced NO production by 60%, it completely blocked the inhibitory effect of endothelin-1 on NKCC2 activity. These data may indicate that the NO concentration required to inhibit transport is above the remaining levels after dominant-negative Akt treatment. However, it may simply be due to differences in expression of dominant-negative Akt in the two groups of rats.

In summary, we found that 1) endothelin-1 acutely stimulates NO production by isolated thick ascending limbs from rats and wild-type mice but not NOS3^–/– mice; 2) wortmannin, a PI3K inhibitor, blocks endothelin-1-stimulated NO production and NOS3 phosphorylation; 3) endothelin-1 increases Akt activity and phosphorylates Akt at Ser⁴⁷³, and wortmannin blocks these effects; 4) dominant-negative Akt reduces the effect of endothelin-1 on NO production; 5) Akt inhibition prevents the endothelin-1-induced increase in thick ascending limb NO production; 6) endothelin-1 acutely phosphorylates NOS3 at Ser¹¹⁷⁷, a positive regulatory site, and wortmannin blocks this effect; 7) l-NAME blocks the inhibitory effect of endothelin-1 on NKCC2 activity; and 8) dominant-negative Akt blocks the inhibitory effect of endothelin-1 on NKCC2 activity. Therefore, we conclude that endothelin-1 activates PI3K, inducing phosphorylation and activation of Akt, which in turn phosphorylates NOS3, resulting in increased NO production and inhibition of sodium reabsorption in the thick ascending limb. Given the importance of NO in the regulation of thick ascending limb NaCl absorption (25), this mechanism likely plays an important role in the regulation of thick ascending limb NO production and sodium homeostasis.