ENaC is regulated by natriuretic peptide receptor-dependent cGMP signaling

Abstract

Epithelial sodium channels (ENaCs) located at the apical membrane of polarized epithelial cells are regulated by the second messenger guanosine 3′,5′-cyclic monophosphate (cGMP). The mechanism for this regulation has not been completely characterized. Guanylyl cyclases synthesize cGMP in response to various intracellular and extracellular signals. We investigated the regulation of ENaC activity by natriuretic peptide-dependent activation of guanylyl cyclases in Xenopus 2F3 cells. Confocal microscopy studies show natriuretic peptide receptors (NPRs), including those coupled to guanylyl cyclases, are expressed at the apical membrane of 2F3 cells. Single-channel patch-clamp studies using 2F3 cells revealed that atrial natriuretic peptide (ANP) or 8-(4-chlorophenylthio)-cGMP, but not C-type natriuretic peptide or cANP, decreased the open probability of ENaC. This suggests that NPR-A, but not NPR-B or NPR-C, is involved in the natriuretic peptide-mediated regulation of ENaC activity. Also, it is likely that a signaling pathway involving cGMP and nitric oxide (NO) are involved in this mechanism, since inhibitors of soluble guanylyl cyclase, protein kinase G, inducible NO synthase, or an NO scavenger blocked or reduced the effect of ANP on ENaC activity.

natriuretic peptides (NPs) meditate their biological effects by binding to specific cell surface receptors. There are at least three members of the NP family: atrial NP (ANP), brain NP (BNP), and C-type NP (CNP). NPs and their cognate receptors are coexpressed in various extracardiac tissues, including the kidneys (39), lungs (13, 35), thyroid (14), pancreas (10, 36), and colon (16). ANP and BNP preferentially binds to NP receptor (NPR)-A, while CNP mainly binds to NPR-B (23, 34). NPs indiscriminately bind NPR-C (8) and are then internalized and degraded. In addition to clearing NPs from circulation, NPR-C is involved in signaling transduction (3). Its activation is coupled to adenylyl cyclase inhibition and phospholipase C activation through the inhibitory guanine nucleotide regulatory protein (G_i) (7, 28). The binding of NPs to their cognate cell-surface guanylyl cyclase-coupled receptors (NPR-A and NPR-B) results in activation of intrinsic guanylyl cyclase activity associated with the intracellular domain of the receptor and, consequently, leads to a rapid increase in guanosine 3′,5′-cyclic monophosphate (cGMP). Diffusible cGMP acts as a second messenger primarily by stimulating protein kinase G (PKG). The intracellular level of cGMP is regulated by balancing synthesis by guanylyl cyclases with its degradation by phosphodiesterases. Both cytosolic and membrane-bound guanylyl cyclases contribute to the production of cGMP.

The epithelial sodium channel (ENaC) functions as a heterotrimeric complex composed of α-, β-, and γ-subunits at the apical membrane of polarized epithelial cells. ENaC is essential for sodium transport and fluid reabsorption across many epithelial tissues, including the kidneys, lungs, and colon. ENaC can be regulated by second messengers, including cAMP (11), Ca²⁺ (26), and cGMP (29). Nie et al. (30) reported the activation of ENaC with a membrane-permeable cGMP analog using a Xenopus oocyte expression system. Zhao et al. (43) reported low doses of ANP increases distal nephron sodium delivery, but does not change the fractional reabsorption of distal sodium delivery. Yamada et al. (37, 38) showed ANP and cGMP-activated ENaC-dependent sodium transport in frog urinary bladder epithelial cells. However, Poschet et al. (31) reported elevating levels of intracellular cGMP inhibited ENaC activity in primary human cystic fibrosis bronchial epithelial cells.

The aim of this study was to investigate the regulation of ENaC activity by cGMP/PKG-dependent and/or -independent mechanisms. Here we show the polarized distribution of endogenously expressed NPR subtypes in sodium-transporting 2F3 renal cells. We also show that ENaC activity decreases in a cGMP-dependent manner, and that the mechanism involves activation of NPR-A.

METHODS

Cell culture.

2F3 cells derived from the Xenopus laevis distal nephron epithelial cell line (A6) and were maintained in DMEM/F-12 (Invitrogen, Carlsbad, CA) medium containing NaHCO₃ and supplemented with 90 mM NaCl, 25 mM NaHCO₃, 3.1 mM KCl, 0.8 mM CaCl₂, 0.4 mM Na₂HPO₄, 0.3 mM NaH₂PO₄, 0.2 mM MgCl₂, 0.3 mM MgSO₄, 5% fetal bovine serum, 1.5 μM aldosterone, 1% penicillin-streptomycin. For single-channel patch-clamp studies, 2F3 cells were subcultured on gluteraldehyde-fixed, collagen-coated Millipore-CM filters (Millipore, Billerica, MA) attached to the bottom of Lucite rings. For all other experiments, 2F3 cells were subcultured on Transwell-permeable supports (Corning, Acton, MA). Cells were cultured for 10 days to form tight junctions before being used for experiments.

Recombinant protein production.

Full-length α, α-NH₂-terminus (M2-V68), α-extracellular loop (S86-G529), α-COOH-terminus (H554-N643), β-NH₂-terminus (M1-K51), β-COOH-terminus (D566-N647), γ-NH₂-terminus (M1-R49) Xenopus ENaC coding sequences were subcloned into the pGEX expression vector. The constructs were transformed into competent bacterial cells, induced with isopropyl-β-d-thiogalactoside for expression, and batch purified from inclusion bodies using glutathione sepharose 4B, as previously described by Alli and Gower (3, 5).

Antibody production.

Polyclonal antibodies against the carboxy terminal domain of Xenopus ENaC-α (ENaC 59) and ENaC-β (ENaC 60) subunits were generated after recombinant Xenopus glutathione-S-transferase ENaC-α and glutathione-S-transferase ENaC-β fusion proteins were purified, as described above, and then White New Zealand rabbits were immunized with 1–2 mg of the purified recombinant fusion protein per animal (Bio-Synthesis, Lewisville, TX). Animals were boosted/bleed at 6, 8, and 10 wk thereafter before being exsanguinated. Each batch of serum was supplemented with sodium azide and evaluated for specificity and cross-reactivity using protein from the wheat germ in vitro translation system (Promega), Xenopus tissue lysates, and cellular lysates of various origins.

Immunofluorescence microscopy.

Confocal microscopy experiments were performed using confluent 2F3 cells, as previously described (1). Briefly, the cells were fixed with 4% paraformaldehyde for 15 min and then permeabilized with 0.1% Triton X-100 for 15 min. To detect the tight junction protein, zonula occludens-1, and to detect NPRs, the cells, were first incubated with mouse antibody to zonula occludens-1 and rabbit antibodies to NPR-A, -B, or -C for 1 h after which the cells were incubated with Alexa Fluor 594 anti-mouse IgG for 1 h, shown in red, and with Alexa Fluor 488 anti-rabbit IgG for 1 h, shown in green.

Adult SV126 mice were maintained on a regular chow diet. The protocol for all animal procedures was approved by the Institutional Animal Care and Use Committee at Emory University. Mice were anesthetized with pentobarbital sodium. Kidneys were fixed with 2.5% paraformaldehyde in PBS, removed, and postfixed in 4% paraformaldehyde at 4°C for 4 h. The kidneys were maintained in 15% sucrose at 4°C overnight before the tissues were then frozen in optimal cutting temperature compound and cut in 7- to 10-μm sections. Frozen kidney sections were washed with PBS and treated with 0.1% Triton X-100 for 5–10 min. Sections were incubated with blocking solution (PBS, 3% BSA, 10% horse serum) for 40 min and then incubated with rabbit anti-NPR antibody (1:1,000) and goat anti-aquaporin-2 (AQP2) (1:200, Santa Cruz Biotechnology) antibodies at 4°C overnight. After washing with PBS, sections were incubated with Alexa Fluor 546-conjugated donkey anti-rabbit IgG (1:800, Invitrogen) and Alexa Fluor 633-conjugated donkey anti-goat IgG (1:800, Invitrogen). Sections were washed with PBS, mounted, and then imaged with an Olympus FV-1000 confocal microscope.

Single-channel patch-clamp studies.

Experiments were performed at room temperature using the cell-attached patch configuration. Patch pipette and extracellular bath solutions consisted of a physiological amphibian saline containing the following (in mM): 95 NaCl, 3.4 KCl, 0.8 CaCl₂, 0.8 MgCl₂, and 10 HEPES or 10 Tris, titrated with 0.1 N NaOH or HCl to a pH of 7.3–7.4. Pharmacological agents were added to the apical or basolateral side of 2F3 cells cultured on gluteraldehyde-fixed, collagen-coated Millipore-CM filters (Millipore, Billerica, MA) attached to the bottoms of small Lucite rings. Open probability (P_o) of a single channel was calculated by dividing NP_o by the number of channels in a patch. For our experiments, we determined mean P_o for 5 min before addition of any agent and for 5 min after addition.

Transepithelial electrical measurements.

Cell confluence was assessed by measuring transepithelial resistance. Cells were returned to the incubator until the resistance was stable. Confluent cells were used for measurements. Reconstituted peptides dissolved in ethanol, or ethanol alone (vehicle), were applied to the apical side of 2F3 cells grown on permeable inserts. Transepithelial voltage (mV) and transepithelial resistance (KΩ) were measured before and after application of the peptide or vehicle with an ohm-volt meter (EVOM; World Precision Instruments, Sarasota, FL). Transepithelial current was calculated by Ohm's law and expressed as microamperes.

Harvesting protein from cells and tissue.

Confluent 2F3 cells cultured on permeable supports were washed once with sterile 1 × PBS. Cells were scraped in ice-cold MPER (Thermo Scientific Pierce), and tissue was homogenized in TPER (Thermo Scientific Pierce). Both MPER and TPER were supplemented with Halt Protease and Phosphatase Inhibitor Cocktail (Thermo Scientific Pierce). Protein concentration was determined by using the BCA Protein Assay Kit (Thermo Scientific Pierce), while following the manufacturer's instructions.

SDS-PAGE and Western blotting.

Fifty micrograms of protein were resolved on 4–20% Tris·HCl polyacrylamide precast gels for 1 h using Tris glycine/SDS buffer (Bio-Rad). The separated proteins were electrophoretically transferred onto Hybond C-extra nitrocellulose membranes (GE Healthcare, Piscataway, NJ) for 1 h using Tris glycine buffer. The membranes were then blocked in 5% (wt/vol) dry milk in 1 × Tris-buffered saline (Bio-Rad) for 1 h at room temperature. Western blotting was performed by first incubating the blocked membranes overnight in a 1:4,000 dilution of NPR-A polyclonal antibody, 1:500 dilution of NPR-B polyclonal antibody (generously provided by Dr. David L. Garbers, University of Texas Southwestern, Dallas, TX), or 1:1,000 dilution of NPR-C antibody (2, 3, 5), each prepared in 5% BSA-1 × Tris-buffered saline. After a series of washes, the membranes were incubated with horseradish peroxidase-conjugated goat anti-rabbit secondary antibody (1:3,000) prepared in blocking solution. The membranes were washed and incubated with SuperSignal Dura Chemiluminescent Substrate (Pierce), according to the manufacturer's instructions, and then exposed using a Kodak Gel Logic 2200 Imager and Molecular Imaging software system (Carestream Health, Rochester, NY).

Statistical analysis.

A paired or unpaired t-test was used to determine statistical significance between two groups. One-way ANOVA, followed by the Holm-Sidak multiple-comparisons test, was used to determine statistical significance for more than one group. Sigma Plot software (Systat Software, Chicago, IL) was used, and values are expressed as means ± SE (or when explicitly stated, SD). P < 0.05 was considered statistically significant.

RESULTS

NPRs are expressed at the apical membrane of Xenopus 2F3 cells.

The expression of NPR subtypes in 2F3 renal cells has not been reported. Figure 1 shows that both classes of NPRs, guanylyl cyclase coupled and nonguanylyl cyclase coupled, are expressed in 2F3 cells (and in mouse kidney). To test the hypothesis that the regulation of ENaC by cGMP likely occurs mainly through activation of particulate guanylyl cyclases, and not soluble guanylyl cyclases (sGC), we show that particulate guanylyl cyclase receptors and ENaC are both expressed at the apical membrane of polarized renal epithelial cells. Immunofluoresence studies show the presence of all three NPR subtypes at the apical membrane of 2F3 cells (Fig. 2, A–C) and at the surface of renal epithelial cells of mouse kidney (Fig. 2, D and E).

Fig. 1.

Western blot (WB) analysis of natriuretic peptide receptor (NPR) subtypes in 2F3 renal cells. A: NPR-A antibody detected a distinct immunoreactive band at ∼130 kDa. B: NPR-B antibody detected multiple immunoreactive bands, including one at ∼130 kDa. C: NPR-C antibody detected a distinct immunoreactive band of ∼60 kDa.

Fig. 2.

Immunofluorescence analysis of NPR expression in 2F3 and mouse kidney cells. Expression of NPR-A (A), NPR-B (B), and NPR-C (C) in 2F3 cells is shown. Immunoreactive staining for NPR-A (D), NPR-B (E), and NPR-C (F) in mouse kidney slices is shown. Zonula occludens-1 is red and a marker for tight junctions. The z-axis shows all three natriuretic peptide subtypes are mainly expressed at the apical membrane. Red scale bar is 10 μm.

NPRs are expressed in AQP2-positive cells of mouse kidney.

NPR subtypes are expressed in 2F3 renal cells. We wished to show NPRs in principal cells of mouse collecting ducts. Figure 3 shows that all three types of NPRs are present in the same cells that express AQP2. In fact, all of the receptors colocalize at the apical membrane with AQP2, but the distribution of the three receptors, A, B, and C, is not the same. NPR-A almost exclusively colocalizes with AQP2 in the apical membranes of principal cells (Fig. 3A). NPR-B also colocalizes with AQP2, but is also present in a few other locations in the sections (Fig. 3B). NPR-C is present in principal cells, but is also more generally found throughout the sections (Fig. 3C). The distributions across the renal slices can best be appreciated in the lower magnification image (Fig. 3D).

Fig. 3.

Distribution of NPRs in mouse kidney. NPRs are expressed in aquaporin-2 (AQP2)-positive cells of mouse kidney. We wished to show that NPRs were in principal cells of mouse collecting ducts where they could influence epithelial sodium channel (ENaC) activity. A–C: distribution of NPR-A, -B, and -C, respectively. For reference, differential interference contrast images are included on the bottom right of each panel (A–C). The images show that all three types of NPRs are present in the cells that express AQP2. A: all of the receptors colocalize at the apical membrane with AQP2, but NPR-A almost exclusively colocalizes with AQP2 in the apical membranes of principal cells. B: NPR-B also colocalizes with AQP2, but is also present at locations different from AQP2. C: NPR-C is present in all cell types. D: the distributions of the three receptor types in the renal slices can best be appreciated in the bottom magnification image. Red scale bar in all images is 10 μm.

Exogenous ANP decreases amiloride-sensitive transepithelial current in Xenopus 2F3 cells.

2F3 cells contain NPRs. To show that at least one class of receptor is functional, we applied ANP to the apical surface of cells grown to confluence on permeable supports and then measured transepithelial voltage, resistance, and current. Application of ANP resulted in a statistically significant reduction in transepithelial current (Fig. 4). Application of ANP to the basolateral surface of the 10 monolayers of cells produced no change in the transepithelial voltage, resistance, or current.

Fig. 4.

Amiloride-sensitive transepithelial current measurements in 2F3 cells after application of exogenous ligands. Cells were cultured on permeable supports until confluent (10 days). When confluent, transepithelial voltage and resistance was measured after vehicle (ethanol), 1 μM atrial natriuretic peptide (ANP), 1 μM β-phenyl-1,N²-etheno-8-bromo-cGMP (8-Br-PET-cGMP), or 1 μM 8-(4-para-chlorophenylthio)-cGMP (8-pCPT-cGMP) were applied to the apical side of the cell monolayers. Total transepithelial current was calculated as the ratio of voltage to resistance. A statistically significant decrease in transepithelial current was observed in response to ANP and 8-pCPT-cGMP compared with vehicle alone, but no significant difference for 8-Br-PET-cGMP-treated cells. Values are the mean current of three plates and the standard deviation. Statistical differences are marked with asterisks and were calculated using a one-way ANOVA with a Holm-Sidak posttest for comparison to the vehicle-treated cells.

The effect of ANP is often mediated by cGMP. To confirm that cGMP affects the activity of ENaC in 2F3 cells and that the effect involves activation of guanylyl cyclase, we measured transepithelial current after applying the cell-permeable cGMP analog 8-(4-para-chlorophenylthio)-cGMP (8-pCPT-cGMP) to the apical side of 2F3 cells. 8-pCPT-cGMP resulted in a statistically significant decrease in amiloride-sensitive transepithelial current (Fig. 4).

ANP, but not CNP or cANP, decreases the P_o of ENaC in 2F3 cells.

The ability of ANP to cause a decrease in ENaC activity was corroborated by single-channel patch-clamp studies. ANP applied to the apical surface of 2F3 cells resulted in a statistically significant decrease in ENaC activity (Fig. 5). ANP binds to all three NPR subtypes with high affinity. Therefore, we examined the ability of the NPR-B preferred ligand, CNP, to alter the activity of ENaC in single-channel patches. CNP applied to the apical side of 2F3 cells did not result in a statistically significant decrease in ENaC activity (Fig. 6, A and B). We also evaluated the ability of the NPR-C-specific agonist, a ring-deleted analog of atrial natriuretic factor, des[Gln18, Ser19, Gly20, Leu21, Gly22]ANF4–23-NH2 (C-ANF4–23), or cANP, to alter the activity of ENaC in single-channel patches. Like CNP, cANP applied to the apical side of 2F3 cells did not result in a statistically significant decrease in ENaC activity (Fig. 6, C and D). We did not examine the basolateral effect of CNP or cANP.

Fig. 5.

Effect of ANP on the open probability (P_o) of ENaC in 2F3 cells. A: representative single-channel current traces obtained from cell-attached patches after application of 1 μM ANP to the apical side of 2F3 cells. The closed level of the channel is marked with c and an arrow. B: summary plot showing a statistically significant decrease (marked with an asterisk) in the P_o of ENaC in response to the ANP treatment. For this figure, we determined mean P_o for 5 min before addition of any agent and for 5 min after addition.

Fig. 6.

Effect of C-type natriuretic peptide (CNP) and NPR-C-specific agonist cANP on the P_o of ENaC in 2F3 cells. A: representative single-channel current traces obtained from cell-attached patches after application of 1 μM CNP to the apical side of 2F3 cells. B: summary plot showing little if any effect on the P_o of ENaC in response to the CNP treatment. C: representative single-channel current traces obtained from cell-attached patches after application of 1 μM cANP (an NPR-C agonist) to the apical side of 2F3 cells. D: summary plot showing little if any effect on the P_o of ENaC in response to the cANP treatment. For this figure, we determined mean P_o for 5 min before addition of any agent and for 5 min after addition.

The decrease in ENaC P_o involves activation of PKG II.

One action of cGMP is to activate PKG, but application of the PKG Iβ activator β-phenyl-1,N²-etheno-8-bromo-cGMP (8-Br-PET-cGMP) resulted in no significant effect on amiloride-sensitive transepithelial current in these cells (Fig. 4). On the other hand, a selective membrane-permeable cGMP analog and activator of PKG II, 8-pCPT-cGMP produced a statistically significant decrease in the P_o of ENaC (Fig. 7, A and B). For comparison, the activator of PKG Iβ, 8-Br-PET-cGMP, did not result in a statistically significant decrease in ENaC activity (Fig. 7, C and D). To further corroborate a role for PKG in cGMP-dependent regulation of ENaC activity, we pretreated 2F3 cells with the selective inhibitor of cGMP-dependent PKG, KT5823, before applying ANP to the apical side and measuring ENaC activity. The inhibitor blocked the ANP-induced decrease in ENaC activity (Fig. 8).

Fig. 7.

Protein kinase G (PKG) II mediates the effect of ANP. A: effect of a PKG II activator on the P_o of ENaC in 2F3 cells. Representative single-channel current (C) record was obtained from cell-attached patches after application of 1 μM 8-pCPT-cGMP to the apical side of 2F3 cells. B: summary plot showing a statistically significant decrease (marked with an asterisk) in the P_o of ENaC in response to the 8-pCPT-cGMP treatment. C: effect of a PKG-1α and PKG-1β activator on the P_o of ENaC in 2F3 cells. Representative single-channel current traces were obtained from cell-attached patches after application of 10 μM 8-Br-PET-cGMP to the apical side of 2F3 cells. D: summary plot showing no change in the P_o of ENaC in response to the 8-Br-PET-cGMP treatment. For this figure, we determined mean P_o for 5 min before addition of any agent and for 5 min after addition.

Fig. 8.

Effect of ANP on the P_o of ENaC in 2F3 cells after inhibiting PKG. A: representative single-channel current record obtained from cell-attached patches after application of 1 μM KT5823 to the apical side of 2F3 cells. B: summary plot showing KT5823 treatment did not have a significant effect on the P_o of ENaC. For this figure, we determined mean P_o for 5 min before addition of any agent and for 5 min after addition.

The source of cGMP that regulates ENaC is at least partially from sGC.

In addition to cGMP produced from cell surface-linked guanylyl cyclase, it is possible that sGC could also contribute to the increase in cGMP levels. Since nitric oxide (NO) activates sGC, we treated the apical and basolateral membranes of 2F3 cells with an NO scavenger, (4-carboxyphenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide (C-PTIO), before applying ANP. Pretreatment with the NO scavenger blocked the ANP-mediated decrease in ENaC P_o (Fig. 9, A and B).

Fig. 9.

Effect of ANP on the P_o of ENaC in 2F3 cells after inhibiting nitric oxide (NO). A: representative single-channel current record obtained from cell-attached patches after application of 200 μM (4-carboxyphenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide (C-PTIO) to the apical side of 2F3 cells. B: summary plot showing C-PTIO treatment blocked the effect of ANP on the P_o of ENaC. C: effect of ANP on the P_o of ENaC in 2F3 cells after inhibiting inducible NO synthase (iNOS). (The large upward deflections in the current record are spurious electrical noise from ventilation motors in the vicinity of the patch setup.) Representative single-channel current record was obtained from cell-attached patches after application of 50 μM N^G-nitro-l-arginine methyl ester (l-NAME) to the apical side of 2F3 cells. D: summary plot showing l-NAME treatment blocked the effect of ANP on the P_o of ENaC. For this figure, we determined mean P_o for 5 min before addition of any agent and for 5 min after addition.

The inducible isoform [inducible NO synthase (iNOS)] is a major source of NO (17). We applied the iNOS inhibitor, N^G-nitro-l-arginine methyl ester, to the apical and basolateral surfaces of 2F3 cells, to block NO production and, thereby, reduce cGMP production from sGC. After inhibiting iNOS, ANP was no longer able to inhibit ENaC P_o (Fig. 9, C and D).

DISCUSSION

We show here that ANP acting through apical receptors can inhibit ENaC activity. Activation of NPRs is often associated with the production of cGMP, which is known to inhibit ENaC. We investigated the mechanism for this cGMP-dependent effect mediated by NPs. The second-messenger cGMP can be synthesized by either soluble or particulate forms of guanylyl cyclases. There are multiple isoforms of soluble and particulate phosphodiesterases that are capable of hydrolyzing cGMP. The production and degradation of cGMP by multiple sources that are present in the cytosol and at the membrane suggest diffusion of the second messenger to its target is not necessary and the effects occur locally. Therefore, in order for the cGMP-dependent regulation of ENaC to be mediated by NPs, particulate guanylyl cyclases should at least share the same polarized membrane distribution. Although ENaC has only been found at the apical membrane, particulate guanylyl cyclases have been reported at the apical and basolateral membranes.

The discrepancy in the literature about the polarized expression of NPR can be attributed to the use of different polyclonal antibodies for the various receptor subtypes. Also, the lack of consistency in the polarized distribution of NPRs may be due to responses to specific pathophysiological conditions. For example, renal ischemia can alter the polarization of the Na⁺-K⁺-ATPase in renal proximal tubular cells. Spiegel et al. (33) showed that this enzyme, which normally functions at the basolateral membrane, redistributed to the apical membrane following ischemic injury. Consistent with this observation, Ritter et al. (32) showed that the polarization of NPR-A is lost in the ischemic kidney. The polarized expression of ion channels and receptors is important because it determines their accessibility to specific ligands. The kidney-specific form of ANP, urodilatin, is produced by renal tubular cells and is then secreted into the lumen of the distal tubules. Therefore, the NPRs would need to be present at the luminal side to mediate the effects of the peptide. Our results suggest that ANP activation of NPR-A is the source of the cGMP responsible for the decrease in the P_o of ENaC that we observe. This finding is consistent with NPR-A being present on the apical side of the cells.

While urodilatin is an NPR ligand that is produced by the tubule, ANP is systemic. To activate apical receptors in the collecting duct, ANP would have to be filtered and not degraded before it reached the collecting duct. There is evidence that ANP does reach cortical collecting duct (CCD) receptors (for a review, see Ref. 19). Urinary concentrations of ANP and urodilatin are generally low; therefore, under normal physiological conditions, ANP receptor activation is probably limited (9). However, under conditions of volume loading comparable to the loading associated with congestive heart failure, concentrations of ANP in plasma and urine rise dramatically, suggesting that CCD receptors are likely active (9, 22). Our work suggests that there would be sodium and water loss when the receptors were activated. Our work also implies the PKG II is the mediator of the ANP effect. PKG exists as two major forms, cGMP-dependent protein kinase (cGK) I and cGK II (21). Contrary to what our results imply, in one study, cGK II was localized in juxtaglomerular cells, the thin ascending limbs, and, to a lesser extent, the brush border of proximal tubules in rats but not in collecting ducts (15). However, another report has shown the presence of cGK II in freshly isolated connecting tubules and CCD in rabbit (18).

Our results examine the effect of ANP and cGMP analogs on single ENaC channels in 2F3 cells. Because we are examining single-channel events, there can be no ambiguity about which type of channel is affected by ANP. The channel described in this paper has all the characteristics that are associated with the “classical” ENaC: 5pS conductance, long mean open and closed times, and a high selectivity for sodium over potassium. Our laboratory has previously shown in 2F3 cells and alveolar type 2 cells that both cell types can produce NO, and that application of either NO or cGMP inhibit this channel (17, 20, 40). There is a long history of other investigators examining the effects of cGMP on Na⁺ transport in sodium-transporting epithelia or ENaC-expressing systems with contradictory results. cGMP increased transport in toad urinary bladder (12), in H441 lung cells in culture (29), and ENaC expressed in Xenopus oocytes (30). On the other hand, ANP or cAMP analogs decrease sodium transport activity in a renal epithelial cell line, LLC-PK1, and medullary collecting duct cells (24, 25, 41, 42). The differences in results could be due to a difference in response of channels other than low-conductance, highly sodium-selective ENaC [indeed, our laboratory has previously described a nonselective cation channel in lung cells that is regulated quite differently than ENaC and whose activity may be increased by NO (17, 20)]. Some of the differences, even when the channels are ENaC, might be due to the off-target effects of 8-Br-cAMP which, besides activating PKG, also alters the activity of phosphodiesterases, thereby altering levels of all cyclic nucleotides in cells (27). Alternatively, the responses might be attributed to the differential expression of different PKG isoforms in different epithelial cells.

The fact that the cell-permeable cGMP analog, 8-pCPT-cGMP and ANP, which binds the membrane-bound guanylyl cyclase receptor, each resulted in a statistically significant decrease in ENaC activity suggests that there could be two different mechanisms for the regulation of ENaC by cGMP. Therefore, we investigated whether or not sGC could induce a similar response on ENaC activity. sGC is the cellular receptor for the intercellular messenger NO. The binding of NO to the heme prosthetic group in the β-subunit of sGC leads to a conformational change of the enzyme and increased cGMP synthesis. Since blocking sGC activation by reducing NO by applying the NO scavenger, C-PTIO, or by reducing NO production by blocking iNOS blocked the effect of ANP, we conclude the sGC play some role in ANP-mediated inhibition of ENaC.

The coexpression of the CNP receptor with NPR-A at the apical membrane will influence the rate and amount of cGMP produced in response to NP binding. It is expected that an upregulation of NPR-C will attenuate the cGMP-mediated inhibition of ENaC as a result of less NP available to bind to NPR-A.

The expression of the C-type NPR may also contribute to the regulation of ENaC independently of cGMP. The surface expression of ENaC increases in response to cAMP, and its activity has been found to decrease after inhibiting cAMP production. NPR-C is linked to the inhibition of adenylyl cyclase, and, therefore, the decreased levels of cAMP may result in inhibition of ENaC activity. This would suggest that there could be cross talk between the two pathways.

We observed a statistically significant decrease in the P_o of ENaC in 2F3 cells in response to exogenous application of ANP. ANP is a potent ligand of both the guanylyl cyclase-linked NPRs, NPR-A and -B, as well as the nonguanylyl cyclase-coupled NPR-C. The fact that application of CNP, the preferential ligand for NPR-B, resulted in no significant effect on the P_o of ENaC suggests that the source of cGMP is from activation of NPR-A. Since NPR-C is more than just a clearance receptor for NPs and has been shown to be involved in the inhibition of adenylyl cyclase and activation of phospholipase C, it was necessary to exclude cross talk between alternate signaling pathways for being responsible for the decrease in ENaC activity, but application of the specific NPR-C agonist, cANP, produced no significant change in the P_o of ENaC.

Together, our results show that ANP is an inhibitor of ENaC, and that NPR-A, but not NPR-B or NPR-C, is involved in the natriuretic-peptide-mediated regulation of ENaC activity. Also, it is likely that a signaling pathway involving cGMP and NO is involved in this mechanism, since inhibitors of sGC, PKG, iNOS, or an NO scavenger blocked or reduced the effect of ANP on ENaC activity. These observations are consistent with the natriuretic effect of ANP on the distal nephron of mammals.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: L.-J.G., A.A.A., and H.-F.B. performed experiments; L.-J.G., A.A.A., D.C.E., and H.-F.B. analyzed data; L.-J.G., A.A.A., D.C.E., and H.-F.B. approved final version of manuscript; A.A.A., D.C.E., and H.-F.B. conception and design of research; A.A.A., D.C.E., and H.-F.B. interpreted results of experiments; A.A.A., D.C.E., and H.-F.B. prepared figures; A.A.A. drafted manuscript; A.A.A., D.C.E., and H.-F.B. edited and revised manuscript.