Mechanism of the Nitric Oxide-Induced Blockade of Collecting Duct Water Permeability
Abstract Nitric oxide has a diuretic effect in vivo. We have shown that nitric oxide inhibits antidiuretic hormone–stimulated osmotic water permeability in the collecting duct; however, the mechanism by which this occurs is unknown. We hypothesized that inhibition of antidiuretic hormone–stimulated water permeability by nitric oxide in the collecting duct is the result of activation of cGMP-dependent protein kinase, which in turn decreases intracellular cAMP. To test this hypothesis, we microperfused cortical collecting ducts. Antidiuretic hormone–stimulated water permeability was 317±47 μm/s (P<.001). Addition of spermine NONOate, a nitric oxide donor, to the bath decreased water permeability to 74±38 μm/s (P<.002). In the presence of LY 83583, an inhibitor of soluble guanylate cyclase, spermine NONOate did not change water permeability. Addition of spermine NONOate increased cGMP production (P<.01). In the presence of the cGMP-dependent protein kinase inhibitor, spermine NONOate did not change water permeability. Since antidiuretic hormone increases water permeability by increasing cAMP, we hypothesized that nitric oxide inhibits water permeability by decreasing cAMP. In tubules pretreated with antidiuretic hormone, intracellular cAMP was 18.9±3.9 fmol/mm. In tubules treated with antidiuretic hormone and spermine NONOate, cAMP was 9.3±1.7 fmol/mm (P<.03). We also examined the effect of spermine NONOate on dibutyryl-cAMP–stimulated water permeability. In the presence of dibutyryl-cAMP, water permeability was 388±30 μm/s. Addition of spermine NONOate had no significant effect on water permeability. Time controls and inhibitors by themselves did not change antidiuretic hormone–stimulated water permeability. We concluded that nitric oxide decreases antidiuretic hormone–stimulated water permeability by increasing cGMP via soluble guanylate cyclase, activating cGMP-dependent protein kinase and decreasing cAMP.
Several studies have shown the diuretic and natriuretic effects of NO in vivo.1 2 3 4 5 Using isolated perfused cortical collecting ducts in vitro, we have shown that NO inhibits basal sodium absorption6 as well as ADH-stimulated sodium absorption and Pf.7 These three effects may contribute to the diuresis induced by NO observed in vivo; however, the intracellular mechanism whereby NO inhibits ADH-stimulated transport is unknown.
In vivo studies suggest that NO-induced diuresis and natriuresis are mediated by cGMP. Infusion of agents that release NO or NO donors increases cGMP content of the kidney8 and has been correlated with diuresis.9 Additionally, exogenous cGMP reverses the antidiuretic effect of NO synthesis inhibition by NG-nitro-l-arginine methyl ester.5 In vitro, using cultured cortical collecting duct cells, Stoos et al10 showed that endothelium-derived relaxing factor decreased sodium flux while increasing cGMP content. Using medullary collecting duct cells, Zeidel et al11 found that cGMP decreases oxygen consumption, suggesting that cGMP decreases transport in this segment of the nephron.
Although these data suggest that NO in the distal nephron decreases transport via cGMP, the steps involved beyond production of cGMP have not been investigated. We hypothesized that NO inhibits ADH-stimulated Pf in the rat cortical collecting duct by decreasing ADH-stimulated cAMP by a mechanism dependent on cGMP generated by NO. We can now report that NO activates guanylate cyclase, increases cGMP, and subsequently activates PKG. Stimulation of this cascade results in a decrease in intracellular cAMP and consequently ADH-stimulated Pf in the cortical collecting duct.
Male Sprague-Dawley rats weighing 120 to 150 g (Charles River Breeding Laboratories, Wilmington, Mass) were maintained on a diet containing 0.22% sodium and 1.1% potassium (Ralston Purina) for at least 5 days. Rats were pretreated with deoxycorticosterone pivalate (5 mg IM) (Ciba-GEIGY) 5 to 9 days before the experiment to enhance sodium transport at the level of the cortical collecting duct.12 On the day of the experiment, the rat was anesthetized with ketamine (100 mg/kg, Parke-Davis), and the left kidney was excised and placed in cold dissection medium equilibrated with 95% O2/5% CO2. Coronal slices were transferred to a dissection dish containing physiological saline at 10°C. Medullary rays were dissected from the slices, and cortical collecting ducts were dissected from the rays. Cortical collecting ducts were transferred to a perfusion chamber, mounted on concentric pipettes, and perfused at 37°C.13
The composition of the bath (mmol/L) was NaCl2 114.0, KCl 4.0, NaH2PO4 2.5, MgSO4 1.2, Na3 citrate 1.0, NaHCO3 25.0, alanine 6.0, calcium lactate2 2.0, glucose 5.5, and raffinose 5.0. The osmolarity was 295±3 mOsm/kg as measured by freezing-point depression. The composition of the perfusion solution was the same as that of the bath except that it had an osmolarity of 190±2 mOsm/kg, which was achieved by decreasing NaCl content. The pH of the bath was 7.4. Solutions were gassed with 95% O2/5% CO2.
The dose of ADH (10 pmol/L) used to increase Pf was determined from our previous work.7 ADH and dibutyryl-cAMP were purchased from Sigma Chemical Co; KT 5823, a PKG inhibitor, was obtained from Kamyia Biomedical Co; and LY 83583, a guanylate cyclase inhibitor, was from Biomol Research Laboratories. SPM was purchased from Cayman Chemical Co. ADH, SPM, dibutyryl-cAMP, and inhibitors were added to the bath.
Osmotic Water Permeability
Pf was measured using raffinose as a volume marker. Raffinose in the perfusate and collected fluid was measured with a continuous-flow ultramicrofluorometer using a previously described enzymatic assay.7 Pf was measured as described by García et al.7 A transepithelial osmotic gradient of approximately 100 mOsm/kg was imposed, resulting in volume flow in the presence of ADH. At least two samples were obtained during each period.
In experiments designed to investigate the effect of NO on ADH-stimulated cAMP, tubules were microdissected and incubated in 30 μL perfusion solution containing no additives, 10 pmol/L ADH, or 10 pmol/L ADH plus 1 μmol/L SPM at 37°C for 30 minutes. The reaction was stopped by addition of 250 μL methanol. Samples were dried in a Savant drier (Forma Scientific) and then reconstituted in 50 μL sodium acetate/sodium nitrate buffer. cAMP was measured with a radioimmunoassay purchased from Biomedical Technologies. cAMP standards and blanks were treated similarly to the samples.
cGMP production was measured by the conversion of [3H]guanine to [3H]cGMP. Isolated tubules were incubated in Hanks’ balanced salt solution containing 20 mmol/L HEPES and [3H]guanine for 2 hours at 37°C. Cortical collecting ducts were then perfused as described above. They were bathed in guanine-free Hanks’ solution for 30 minutes, after which the bath was replaced with Hanks’ solution containing 0.5 mmol/L isobutylmethylxanthine. Bath and perfusate were collected for 10 minutes for basal measurement of cGMP accumulation, and then SPM was added to the bath. Perfusate and basolateral bath were collected over the next 120 minutes. Samples, blanks, and recoveries were run through Al2O3 columns to separate cGMP, guanosine, and guanine from GTP, GDP, and 5′GMP, which are retained on the column.14 An aliquot of this eluate was passed through a Dowex anion-exchange column, which retains cGMP and passes guanine and guanosine.15 cGMP accumulation was determined by subtracting eluate activity from the Dowex columns from eluate activity from the Al2O3 columns. cGMP and guanine recoveries were 88±2.3% and 98±2.1%, respectively.
Values are reported as mean±SE. Student’s paired t test was used to test for significant differences.
Previously we reported that NO decreases ADH-stimulated Pf in the cortical collecting duct and Pf returns to stimulated values after removal of NO.7 The purpose of the present work was to determine how NO decreases ADH-stimulated Pf. We first demonstrated that NO decreases ADH-stimulated Pf. ADH (10 pmol/L) stimulated basal Pf from −94±31 to 317±47 μm/s (P<.001). SPM, an NO donor, decreased Pf to 74±38 μm/s (P<.002, n=6, Fig 1⇓). ADH-stimulated Pf did not change during the course of the experiment unless SPM was added (P>.30).
In several tissues, the effects of NO have been found to depend on cGMP production. In view of this, we investigated the effect of SPM on ADH-stimulated Pf in the presence of 1 μmol/L LY 83583, an inhibitor of soluble guanylate cyclase. Basal Pf was −38±28 μm/s. When ADH was added to the bath, Pf increased to 208±25 μm/s (P<.001). Subsequent addition of LY 83583 to the bath did not change Pf (P>.50). With further addition of SPM to the bath, Pf also did not change (from 195±31 to 171±29 μm/s, P>.30, n=5; Fig 2⇓). Next we measured the effect of SPM on cGMP production. Basal production was 0.08±0.07 fmol/mm per minute. When SPM was added, production increased to 0.89±0.20 fmol/mm per minute (P<.01, n=6; Fig 3⇓). These data indicate that NO increases cGMP via soluble guanylate cyclase.
cGMP may activate several enzymes, one of which is PKG. To test whether PKG activation is a necessary step in the NO second messenger cascade, we tested the effect of SPM on ADH-stimulated Pf in the presence of KT 5823, a PKG inhibitor. Basal Pf was −2±6 μm/s and ADH-stimulated Pf was 127±11 μm/s (P<.001). SPM at 1 μmol/L did not decrease ADH-stimulated Pf in the presence of 2 μmol/L KT 5823 (from 127±11 to 173±31 μm/s, P>.15, n=6; Fig 4⇓). The inhibitor alone did not affect ADH-stimulated Pf (from 134.3±42.0 to 133.9±44.5 μm/s, P>.90).
ADH increases Pf by activation of adenylate cyclase and subsequent generation of cAMP.16 To investigate whether NO decreases ADH-stimulated Pf by decreasing intracellular cAMP content, we studied the effect of SPM on the cAMP content of ADH-treated tubules. ADH-stimulated cAMP content was 18.9±3.9 fmol/mm. In the presence of ADH and SPM, it was 9.3±1.7 fmol/mm (P<.03, n=12; Fig 5⇓). To determine whether a decrease in cAMP is required for the inhibitory action of SPM on Pf, we examined whether SPM could decrease dibutyryl-cAMP–stimulated Pf. Basal Pf was 10±11 μm/s; after addition of 0.5 mmol/L dibutyryl-cAMP, Pf increased to 388±30 μm/s (P<.0001). Addition of SPM did not affect Pf (from 388±30 to 393±29 μm/s, P>.70, n=4; Fig 6⇓).
NO decreases ADH-stimulated Pf in the rat cortical collecting duct by stimulating soluble guanylate cyclase, increasing cGMP, and activating PKG. Activation of this kinase decreases cAMP, which in turn inhibits Pf.
The diuretic effect of NO is well documented.1 2 3 4 5 We have shown that SPM and nitroglycerin, both NO donors, decrease basal6 and ADH-stimulated7 transport by the cortical collecting duct. The inhibitory effect of NO on ADH-stimulated fluid absorption was the result of inhibition of Pf and active transport of sodium. To investigate how NO decreases ADH-stimulated Pf, we first investigated the second messenger of NO. Initially, the role of soluble guanylate cyclase was examined using a specific inhibitor. When soluble guanylate cyclase was blocked, NO was unable to reduce ADH-stimulated Pf. Next we measured the production of cGMP induced by NO, since activation of soluble guanylate cyclase was required for NO to decrease Pf. When SPM was added to the bath, basal production of cGMP increased, indicating that NO increases cGMP as previously reported.8 9 Taking these results together, we conclude that NO stimulates soluble guanylate cyclase and increases cGMP production in the cortical collecting duct.
Increased intracellular cGMP may activate PKG or affect other cGMP-dependent enzymes.17 18 19 20 Additionally, NO inhibits Na,K-ATPase activity in the proximal tubule through a mechanism involving PKG.20 Thus we hypothesized that NO increases cGMP and activates PKG. To test this hypothesis, we inhibited PKG. When this enzyme was inhibited, NO did not change Pf. This is the first demonstration that activation of PKG is involved in decreasing Pf by agents that stimulate cGMP levels.
In agreement with our results, it has been previously shown with cultured cortical collecting duct cells that NO decreases net sodium transport via a mechanism mediated by cGMP.10 In humans, in vivo infusion of l-arginine, a precursor of NO, increased urinary sodium excretion. This increase was correlated with an increase in urinary cGMP and nitrite, suggesting that NO-induced diuresis is mediated by cGMP.1 Regulation of NO by cGMP in vivo was also investigated by Lahera et al.5 They inhibited NO synthesis with NG-nitro-l-arginine methyl ester, and the antidiuretic response induced by the inhibitor was blocked by 8-bromo-cGMP, indicating that the natriuretic effect maintained by NO is due to cGMP production.
ADH is a major regulator of Pf of the cortical collecting duct. This effect is mediated by an increase in intracellular cAMP, leading to activation of cAMP-dependent protein kinase. Activation of this protein kinase increases the number of water channels in the apical membrane.16 Although there are a number of mechanisms whereby PKG could affect ADH-stimulated Pf, the simplest explanation is that it alters cAMP levels. Consequently, we examined the effect of NO on cAMP levels. To examine whether PKG modulates ADH-stimulated Pf at the level of cAMP rather than at the level of some regulatory protein, we measured intracellular levels of cAMP before and after NO treatment. Treatment of cortical collecting ducts with ADH increased cAMP content as expected. Interestingly, NO decreased cAMP content. To demonstrate that NO lowers Pf by decreasing cAMP, we studied the effect of NO on Pf stimulated by an exogenous cAMP analogue. In the presence of dibutyryl-cAMP, NO did not affect Pf, indicating that NO decreases Pf by decreasing intracellular cAMP levels. These are the first studies to show that NO lowers Pf by decreasing cAMP.
PKG could alter cAMP levels by a number of mechanisms. PKG could directly alter cAMP production by affecting the ADH receptor, the stimulatory G protein, or adenylate cyclase. Alternatively, it may alter cAMP degradation by affecting phosphodiesterase activity. We are now attempting to resolve this issue.
An inhibitory effect of cGMP on ADH-stimulated Pf has been reported by investigators studying ANF. Dillingham and Anderson21 reported that ANF decreased ADH-stimulated Pf in rabbit cortical collecting ducts. Since ANF did not affect chlorophenyl thio cAMP–stimulated Pf, these authors concluded that ANF lowered Pf by decreasing cAMP formation. Nonoguchi et al22 reported that ANF decreased ADH-stimulated Pf in the inner medullary collecting duct, likewise suggesting that ANF alters cAMP levels by influencing its production or degradation. Interestingly, this same group found little or no effect of ANF on ADH-stimulated Pf in the cortical collecting duct.23 Explanations for the differences between our results and theirs are not obvious; however, compartmentalization of cGMP may be one explanation, since NO increases soluble guanylate cyclase and ANF stimulates particulate guanylate cyclase.
In summary, we propose the following cascade for the inhibition of ADH-stimulated Pf by NO in the cortical collecting duct: Activation of soluble guanylate cyclase by NO increases cGMP. Increases in cGMP activate PKG. PKG activation may decrease cAMP content, leading to a decline in Pf.
Selected Abbreviations and Acronyms
|ANF||=||atrial natriuretic factor|
|Pf||=||osmotic water permeability|
|PKG||=||cGMP-dependent protein kinase|
This work was supported by a grant from the National Institutes of Health (HL-28982). Jeffrey L. Garvin is the recipient of a Research Career Development Award (HL-02891).
Majid DSA, Williams A, Kadowitz PJ, Navar G. Renal responses to intra-arterial administration of nitric oxide donor in dogs. Hypertension. 1993;22:535-541.
Majid DSA, Godfrey M, Grisham MB, Navar LG. Relationship between pressure natriuresis and urinary excretion of nitrate/nitrite in anesthetized dogs. Hypertension. 1995;25:860-865.
Lahera V, Navarro J, Biondi ML, Ruilope LM, Romero JC. Exogenous cGMP prevents decrease in diuresis and natriuresis induced by inhibition of NO synthesis. Am J Physiol. 1993;264:F344-F347.
Stoos BA, García NH, Garvin JL. Nitric oxide inhibits sodium reabsorption in the isolated perfused cortical collecting duct. J Am Soc Nephrol. 1995;6:89-92.
García NH, Pomposiello SI, Garvin JL. Nitric oxide inhibits ADH-stimulated osmotic water permeability in cortical collecting ducts. Am J Physiol. 1996;270:F206-F210.
Grandes S, Gallego MJ, Riesco A, Farre AL, Millas I, Casado S, Hernando L, Caramelo C. Mechanisms of renal effects of different agents stimulating production of cGMP. Am J Physiol. 1991;261:H1109-H1114.
Stoos BA, Carretero OA, Farhy RD, Scicli G, Garvin JL. Endothelium-derived relaxing factor inhibits transport and increases cGMP content in cultured mouse cortical collecting duct cells. J Clin Invest. 1992;89:761-765.
Zeidel ML, Seifter JL, Lear S, Brenner BM, Silva P. Atrial peptides inhibit oxygen consumption in kidney medullary collecting duct cells. Am J Physiol. 1986;251:F379-F383.
Chen L, Williams SK, Schafer JA. Differences in synergistic actions of vasopressin and deoxycorticosterone in rat and rabbit CCD. Am J Physiol. 1990;259:F147-F156.
Burg MB, Grantham JJ, Abramov M, Orloff J. Preparation and study of fragments of single rabbit nephrons. Am J Physiol. 1966;210:1293-1298.
Lincoln TM, Cornwell TL. Intracellular cyclic GMP receptor proteins. FASEB J. 1993;7:328-338.
McKee M, Scavone C, Nathanson JA. Nitric oxide, cGMP, and hormonal regulation of active sodium transport. Proc Natl Acad Sci U S A. 1994;91:12056-12060.
Dillingham MA, Anderson RJ. Inhibition of vasopressin action by atrial natriuretic factor. Science. 1986;231:1572-1573.
Nonoguchi H, Sands JM, Knepper MA. Atrial natriuretic factor inhibits vasopressin-stimulated osmotic water permeability in rat inner medullary collecting duct. J Clin Invest. 1988;82:1383-1390.
Nonoguchi H, Sands JM, Knepper MA. ANF inhibits NaCl and fluid absorption in cortical collecting duct of the rat kidney. Am J Physiol. 1989;256:F179-F186.