The effect of nitric oxide (NO) on Na+/H+ exchange (NHE) activity was investigated utilizing Caco-2 cells as an experimental model. Incubation of Caco-2 cells with 10−3 MS-nitroso-N-acetylpenicillamine (SNAP), a conventional donor of NO, for 20 min resulted in a ∼45% dose-dependent decrease in NHE activity, as determined by assay of ethylisopropylamiloride-sensitive 22Na uptake. A similar decrease in NHE activity was observed utilizing another NO-specific donor, sodium nitroprusside. SNAP-mediated inhibition of NHE activity was not secondary to a loss of cell viability. NHE3 activity was significantly reduced by SNAP (P < 0.05), whereas NHE2 activity was essentially unaltered. The effects of SNAP were mediated by the cGMP-dependent signal transduction pathway as follows:1) LY-83583 and 1H-(1,2,4)oxadiazolo(4,3-a)quinoxalin-1-one (ODQ), specific inhibitors of soluble guanylate cyclase, blocked the inhibitory effect of SNAP on NHE; 2) 8-bromo-cGMP mimicked the effects of SNAP on NHE activity; 3) the SNAP-induced decrease in NHE activity was counteracted by a specific protein kinase G inhibitor, KT-5823 (1 μM); 4) chelerythrine chloride (2 μM) or calphostin C (200 nM), specific protein kinase C inhibitors, did not affect inhibition of NHE activity by SNAP;5) there was no cross activation by the protein kinase A-dependent pathway, as the inhibitory effects of SNAP were not blocked by Rp-cAMPS (25 μM), a specific protein kinase A inhibitor. These data provide novel evidence that NO inhibits NHE3 activity via activation of soluble guanylate cyclase, resulting in an increase in intracellular cGMP levels and activation of protein kinase G.
- sodium/hydrogen exchanger
nitric oxide (NO) is a free radical diatomic molecule synthesized by mammalian cells from the guanidino nitrogen of l-arginine by a five-electron oxidation reaction catalyzed by a family of NO synthase (NOS) enzymes (34). A significant role for NO has been demonstrated in the physiology and pathophysiology of the gastrointestinal tract by a series of in vivo and in vitro studies (55). NO has been shown to be involved in regulation of intestinal motility (8), maintenance of blood flow within the layers of the wall of the gut (41), and protection of the mucosal barrier (34). However, besides its important biological functions under normal conditions and in acute inflammation, there is strong evidence that overproduction of NO plays an important role in the pathogenesis of experimental and clinical inflammatory bowel disease (IBD) (14). Potential sources of NO production in the intestinal epithelium include vascular endothelium (15), mesenteric neurons (37), enterocytes (51), macrophages and mast cells of the lamina propria (36). The effects of NO on intestinal epithelial functions have been suggested to occur through different mechanisms. Evidence suggests that NO can exert its effects through stimulation of soluble guanylate cyclase, resulting in the production of second messenger cGMP (53). Also, NO can act through activation of protein kinase C (PKC) in a number of cell types (53) or through ADP-associated ribosylation (53).
Previous functional studies indicate that NO directly affects Na+ and Cl− absorption and osmotic water permeability (17). NO has also been found to inhibit Na+ transport in the rabbit proximal tubule (45) and in the thick ascending limb (18). Plato et al. (42) demonstrated that NO inhibits Cl− transport in the thick ascending limb. Electrophysiological studies utilizing NO donors indicated a decrease in NaCl absorption in rat colon (58) and guinea pig intestine (29). In contrast, the NO precursorl-arginine at low concentrations increased Na+absorption in rat intestine (56) and mouse cecum (21). In this regard, however, the effects of NO on Na+ absorption in the human intestine have not been investigated.
The effects of NO on electrolyte transport are influenced by whether the conditions under study are physiological or pathophysiological. Under physiological conditions, NO is considered to be a proabsorptive molecule, whereas it causes net secretion or inhibits absorption under pathophysiological conditions (53). The decreased NaCl absorption can be a result of a decrease in Na+ transport mediated by membrane proteins known as Na+/H+exchangers (NHEs), which mediate the electroneutral exchange of extracellular Na+ with intracellular H+. Molecular cloning studies have demonstrated the existence of the NHE gene family, of which at least seven members, designated NHE1–NHE7, have been described with different tissue distributions and functional properties. Previous studies from our laboratory have shown that NHE1–NHE3 are expressed in the human intestine (13). NHE1 is localized to the basolateral membrane (53) and is suggested to be involved in housekeeping functions, whereas NHE2 and NHE3 are localized to the apical membrane and are suggested to be involved in vectorial Na+ transport (22).
The present study was undertaken to study the direct effects of NO on possible regulation of NHE activity and mechanism of regulation utilizing Caco-2 cells as an experimental model. The Caco-2 cell line is a human colonic carcinoma cell line that, on differentiation, manifests many anatomic and functional similarities to absorptive small intestinal enterocytes and has been used to study regulation of electrolyte uptake by various hormones and growth factors (3, 35,44). Our present data demonstrate for the first time that NO decreased Caco-2 cell NHE activity by differential inhibition of the activity of NHE3 but not NHE2. This occurs through the activation of soluble guanylate cyclase, resulting in increased production of intracellular cGMP and activation of protein kinase G (PKG). There was no involvement of PKC- or protein kinase A (PKA)-mediated pathways in this process.
MATERIALS AND METHODS
S-nitroso-N-acetylpenicillamine (SNAP) and Rp-cAMPS were obtained from Sigma Chemical (St. Louis, MO),22Na from NEN Life Science Products (Boston, MA), Caco-2 cells from the American Type Culture Collection (Manassas, VA), and chelerythrine chloride, calphostin C, 1H-(1,2,4)oxadiazolo(4,3-a)quinoxalin-1-one (ODQ), 6-anilinoquinoline-5,8-quinonine (LY-83583), 8-bromo-cGMP (8-BrcGMP), and KT-5823 from Biomol (Plymouth Meeting, PA). HOE-694 was a generous gift from Dr. Hans J. Lang (Aventis, Pharma Deutschland Chemical Research, Frankfurt/Main, Germany). All other chemicals were of at least reagent grade and were obtained from Sigma Chemical or Fisher Scientific.
Caco-2 cells were grown routinely in 75-cm2 plastic flasks at 37°C in a 5% CO2-95% air environment. The culture medium consisted of high-glucose DMEM, 20% FBS, 100 IU/ml penicillin, and 100 μg/ml streptomycin. Cells reached confluence after 5–7 days in culture. They were used for these studies betweenpassages 25 and 55 and were plated in 24-well plates at a density of 2 × 104 cells/ml. Cells were used for experiments 10 days after plating and were fed fresh incubation medium on alternate days.
Determination of nitrite levels.
The levels of NO in the culture medium can be measured by estimating nitrite (NO ), a stable metabolite of NO, utilizing the Greiss reaction (20). For these studies, the medium from confluent cells grown in 24-well plates was aspirated, and cells were incubated for 20 min in the presence or absence of the exogenous donor of NO (SNAP). NO levels were measured in aliquots of the incubation medium by using a commercially available NO colorimetric assay kit (Biomol).
Assay of cGMP.
For measuring cGMP levels in Caco-2 cells, confluent monolayers were preincubated for 15 min at room temperature in fresh medium supplemented with 0.5 mM IBMX. SNAP was then added for a further 20 min in the acid load buffer, and the reaction was stopped by addition of 10% (vol/vol, final concentration) trichloroacetic acid. Cell-associated cGMP content was measured in aliquots of the buffer extracted four times with four volumes of water-saturated ether, brought to pH 7.0 with Tris, and assayed for cGMP by radioimmunoassay using a commercially available kit (cGMP kit, Amersham). Results are expressed as picomoles of cGMP per milligram of protein.
Measurement of intracellular pH.
Intracellular pH (pHi) was measured in Caco-2 cells grown on coverslips using the pH-sensitive fluorescent dye 2′,7′-bis(2-carboxyethyl)-5(6-carboxyfluorescein) (BCECF; Sigma Chemical), as previously described (28, 40). Briefly, the cells were washed with buffer containing (in mM) 140 NaCl, 5 KCl, 1 CaCl2, 1 MgCl2, 5 glucose, and 6 HEPES-KOH, pH 7.4. Cells were then incubated at 37°C for 30 min with the same buffer containing 10 μM BCECF-AM. The Caco-2 cells were washed twice and loaded with acid load medium, in the absence or presence of 1 mM SNAP, for 30 min. The ratio of fluorescence of the intracellularly trapped BCECF dye was determined with excitation at wavelengths of 490 and 440 nm and emission at 530 nm utilizing a luminescence spectrometer (model LS50, Perkin-Elmer, Beaconsfield, UK). To estimate pHi, the BCECF excitation fluorescence ratios were calibrated utilizing the K+/nigericin methods, as previously described (28). The calibration curve demonstrated that the fluorescence ratios were linear as a function of pHi 6.0–8.0, as previously reported (28).
Assay of NHE activity.
The NHE activity was determined in acid-loaded Caco-2 cells as ethylisopropylamiloride (EIPA)-sensitive 22Na uptake, as described (3). The activity of NHE isoforms (NHE2 and NHE3) was measured in the presence of HOE-694 (NHE2-specific inhibitor at 50 μM). NHE2 activity was calculated as NHE activity sensitive to 50 μM HOE-694. NHE3 activity was calculated by subtracting 50 μM HOE-694-sensitive NHE activity from total NHE activity (50 μM EIPA-sensitive NHE activity).
For these studies, the cells were placed at room temperature for 15–20 min to allow for equilibration before determination of22Na uptake. Briefly, confluent cell monolayers were preincubated for 20 min at room temperature in acid load solution containing (in mM) 50 NH4Cl, 70 choline chloride, 5 KCl, 1 MgCl2, 2 CaCl2, 5 glucose, and 15 MOPS (pH 7.0). Cells were then incubated further for 20 min with or without SNAP, a conventional donor of NO, in the acid load solution. In a separate set of experiments, cells were also coincubated with 6-anilinoquinoline-5,8-quinonine (LY-83583) and ODQ (inhibitors of soluble guanylate cyclase), calphostin and chelerythrine chloride (inhibitors of PKC), and KT-5823 (an inhibitor of PKG). The cells were then washed with a solution containing (in mM) 120 choline chloride and 15 Tris-HEPES, pH 7.5. The solution was then aspirated, and the cells were incubated in uptake buffer containing (in mM) 10 NaCl, 110 choline chloride, 1 MgCl2, 2 CaCl2, and 20 HEPES (pH 7.4) and 1 mCi/ml of 22Na, with or without 50 μM EIPA or 50 μM HOE-694. After 5 min, the 22Na-containing uptake solution was aspirated and the cells were washed twice with ice-cold PBS. The cells were then solubilized by incubation with 0.5 N NaOH, and incorporated radioactivity was determined. The protein content of cell uptakes was estimated by the method of Bradford (4a). 22Na uptake was normally measured at 5 min (inasmuch as this was in the linear range of the time course) and expressed as nanomoles per milligram of protein per 5 min.
Apical membrane unidirectional 22Na influx utilizing Transwell inserts.
For the experiments using Transwell inserts, Caco-2 cells were plated at a density of 4 × 104 cells/ml for 14 days before experimentation. Unidirectional apical membrane 22Na uptake was measured from the apical side, as described previously (44). Briefly, Caco-2 cells were exposed to SNAP from the apical and basolateral sides for 20 min in the acidifying solution, in the absence or presence of 0.5 mM ouabain applied basolaterally. Data are presented as EIPA (50 μM)-inhibitable component per Transwell insert and are expressed as picomoles per Transwell insert per 5 min.
Values are means ± SE. Student's t-test was utilized in statistical analysis. P < 0.05 was considered statistically significant.
Effects of NO on NHE activity.
To determine the effects of NO on NHE activity, Caco-2 cells were incubated with 0–1,000 μM SNAP, a conventional donor of NO, for 20 min, and NHE activity was measured as EIPA-sensitive22Na uptake after acidification of the cells by NH4Cl prepulse. As shown in Fig.1, incubation with 500 μM and 10−3 M SNAP resulted in a dose-dependent decrease in NHE activity. There was up to ∼45% inhibition with 10−3 M SNAP (P < 0.05). Therefore, 10−3 M SNAP was utilized for all subsequent experiments. To exclude the possibility that the effects of SNAP on NHE activity were due to cytotoxic effects of NO, trypan blue exclusion studies were carried out with different concentrations of SNAP. No alteration in cell viability was observed in Caco-2 cells incubated with SNAP compared with controls (15 ± 3% nonviable cells in control vs. 16 ± 1% at 10−3 M SNAP).
To evaluate whether the observed decrease in NHE activity by SNAP is due to the release of NO, and not a nonspecific effect of the NO donor, we determined the response to another NO donor, sodium nitroprusside (SNP), which is structurally unrelated to SNAP. The results showed that, similar to SNAP, SNP (0–500 μM) also decreased NHE activity in a dose-dependent manner (Fig.2).
To further rule out the possibility that the effects of SNAP on NHE activity are mediated by changes in the buffering capacity of the cells after SNAP treatment, pHi was measured in Caco-2 cells grown on coverslips and incubated in the absence or presence of 1 mM SNAP in the acidifying solution for 20 min using the pH-sensitive dye BCECF-AM. The results demonstrated that, after incubation of Caco-2 cells in the acid load, there was a significant decrease in pHi (7.20 ± 0.02 and 6.50 ± 0.01 for control and acid load, respectively, n = 3). This drop in pHi was essentially unaltered in the presence of SNAP (6.48 ± 0.03 and 6.50 ± 0.01 in SNAP-treated and control cells, respectively). These data indicate almost no change in the buffering capacity of the cells in response to SNAP, and the observed effects of SNAP on NHE activity were indeed due to the effects of NO generated.
To estimate the concentration of NO in solution, the levels of NO , a stable metabolite of NO (30), were measured after exposure of cells to SNAP for 20 min. As shown in Table 1, incubation of cells with SNAP resulted in a concentration-dependent increase in NO production.
Effect of NO on NHE3 vs. NHE2.
The amiloride analogs EIPA and HOE-694 were used to measure the effect of SNAP on apical NHE isoforms NHE2 and NHE3. As described inmaterials and methods, NHE2 activity was calculated as NHE activity sensitive to 50 μM HOE-694 and determined as total NHE activity minus NHE activity in the presence of 50 μM HOE-694. NHE3 activity was calculated as NHE activity (50 μM EIPA sensitive) remaining in the presence of 50 μM HOE-694. Under control conditions, 42 ± 6% of total NHE activity was contributed by NHE2, and 58 ± 8% was contributed by NHE3 (Fig.3 A). Figure 3 Bshows the differential effect of SNAP at two different concentrations on NHE3 activity. NHE3 activity was reduced to 58% and 41% in the presence of 500 μM and 10−3 M SNAP, respectively, compared with the control values. In contrast, however, NHE2 activity essentially remained unaltered in the presence of SNAP (47.5 ± 2.85 and 45.3 ± 6.8% at 500 μM and 10−3 M SNAP, respectively). These observations indicate specific effects of SNAP on NHE3 activity.
Role of cGMP in NHE inhibition by NO.
We next examined the signal transduction pathways mediating the effects of NO on NHE activity in Caco-2 cells. NO stimulates cGMP production in many cell types via activation of soluble guanylate cyclase (45). Experiments were thus performed to measure cGMP levels stimulated by NO. As shown in Table2, SNAP caused a concentration-dependent increase in cGMP production. There was ∼4.0-fold increase in cGMP levels at 1 mM SNAP compared with control. However, in the presence of the soluble guanylate cyclase inhibitor LY-83583 (10 μM), the cGMP levels were equivalent to that of control. To investigate further the role of cGMP in regulation of NHE activity, cells were incubated with the inhibitor of guanylate cyclase, ODQ (1 μM), in the presence of SNAP (10−3 M SNAP), and NHE activity was assayed. As shown in Fig. 4, 10 −3 M SNAP inhibited NHE activity by ∼50%, as expected, but, in the presence of ODQ and LY-83583, specific inhibitors of soluble guanylate cyclase, which act as NO scavengers by specifically inhibiting soluble guanylate cyclase, also blocked the inhibitory effect of SNAP. These results suggested that production of cGMP through activation of guanylate cyclase mediates NO-induced inhibition of NHE activity. cGMP has been shown to inhibit NHE activity in different cell types (9, 47). We also studied the effect of the membrane-permeable analog of cGMP, 8-BrcGMP, on apical NHE activity. As shown in Fig. 5, preincubation of cells for 20 min with 8-BrcGMP resulted in a dose-dependent decrease in NHE activity (P < 0.05). These observations suggest that 8-BrcGMP mimicked the effects of NO in inhibiting NHE activity.
Role of PKG.
NO stimulates soluble guanylate cyclase, resulting in production of cGMP, and increases in cGMP lead to PKG activation (54). To assess the role of PKG in regulation of NHE activity, cells were treated with 10−3 M SNAP in the presence of the PKG inhibitor KT-5823. In the presence of 1 μM KT-5823, the SNAP-mediated inhibition of NHE activity was completely reversed, indicating that PKG activation was involved in inhibition of NHE activity (Fig.6).
Effect of PKC inhibitors on SNAP-mediated inhibition of NHE activity.
To exclude the possibility that PKC might be involved in SNAP-mediated inhibition of NHE activity, experiments were also performed utilizing the specific PKC inhibitors chelerythrine chloride and calphostin C. Caco-2 cells were preincubated with the PKC inhibitors for 60 min before addition of SNAP, and NHE activity was assayed. Results indicated that, in the presence of 2 μM chelerythrine chloride (Fig.7), there was essentially no effect on SNAP-mediated inhibition of NHE activity. Similar results were obtained utilizing another inhibitor of PKC, calphostin C (2.32 ± 0.48, 1.44 ± 0.14, and 1.19 ± 0.07 nmol · mg protein−1 · 5 min−1 for control, SNAP, and SNAP + 200 nM calphostin C, respectively).
Role of possible cross activation of PKA by cGMP.
Previous evidence suggests that cAMP inhibits apical Na+/H+ exchange (7), and high concentrations of cGMP can cross activate PKA (16). To determine whether a PKA-dependent mechanism was involved in NO-mediated inhibition of NHE activity, NHE activity in response to SNAP was measured in the presence of the specific PKA inhibitor Rp-cAMPS (25 μM). As shown in Fig. 8, SNAP (10−3 M)-mediated NHE activity remained unaltered in the presence of Rp-cAMPS, thereby indicating that the effects of NO are specifically mediated by cGMP and activation of PKG, rather than PKA.
Possible role of Na+-K+-ATPase in SNAP-mediated effects on NHE activity.
Previous studies on the effects of NO showed a decrease in the activity of the Na+-K+-ATPase in kidney (31) and T84 cells (49). Therefore, additional experiments were performed to determine whether inhibition of this enzyme by NO would account for the observed decrease in NHE activity by SNAP. For this purpose, Caco-2 cells were grown in Transwell inserts treated with SNAP (from apical and basolateral sides) in the absence or presence of basolaterally applied ouabain (0.5 mM). As shown in Fig. 9, the observed decrease in apical 22Na uptake in response to SNAP was essentially unaltered in the presence of ouabain. These results rule out the possible involvement of the basolaterally expressed Na+-K+-ATPase in mediating the effects of acute SNAP treatment on NHE activity.
The present study demonstrated that high concentrations of NO inhibited Na+ uptake mediated by NHE3 in the Caco-2 cells. High levels of NO have been implicated in the pathogenesis of IBD. Increased NO production has been associated with high inducible NOS immunoreactivity and increased nitrate/NO levels in various animal models of colitis (33, 43, 59). Similar results have also been obtained from clinical findings in patients with ulcerative colitis (2). The fact that NO levels are elevated in diarrhea, associated with IBD, suggests an important role of NO in modulating electrolyte and fluid transport in the intestine. In this regard, previous studies in the literature regarding the effect of NO on electrolyte transport appear contradictory and depend on whether the conditions under study are physiological or pathophysiological. For example, in physiological conditions, endogenous NO is a proabsorptive molecule and may decrease fluid secretion stimulated by various agonists (24). In contrast, under pathophysiological conditions, high levels of NO can provoke net secretion and decrease absorption (24).
Previous electrophysiological studies showed that NO stimulates Cl− and bicarbonate secretion and inhibits Na+and Cl− absorption in rat colon (58). Recent studies by Liang and Knox (27) and Roczniak and Burns (45) suggested a decrease in NHE activity in the proximal tubule cells in response to exogenous donors of NO. In contrast, Marletta et al. (30) reported an increase in Na+ absorption in response to l-arginine (precursor of NO) in mouse cecum (21).
Our studies in Caco-2 cells using SNAP as an exogenous donor of NO demonstrated a dose-dependent decrease in apical NHE activity, with ∼50% inhibition at 1 mM. Estimation of NO concentration by measuring NO levels in the culture medium demonstrated a dose-dependent increase in NO levels with increasing concentrations of SNAP. The concentrations of SNAP used in this study produced NO levels within the range of values considered to be obtainable in vivo (19, 48). Furthermore, patients with colorectal carcinoma and IBD show elevated plasma nitrate/NO levels (60–70 μM) (26,50), which tend to be similar to the values achieved in our studies.
Generation of high levels of NO has been reported to lead to induction of cell death in a number of cell types (38, 39). However, measurement of cell viability by trypan blue exclusion in present studies ruled out the possibility that effects of NO on apical NHE activity are a mere manifestation of cytotoxicity. Consistent with our observations, Salzman et al. (47) found no evidence of cytotoxicity in Caco-2/bb2 cells incubated with 1.25 mM sodium nitroprusside (SNP) for 24 h as seen by confocal and ultrastructural microscopy as well as lactate dehydrogenase release.
Studies from our laboratory (3) and others (25, 35,44) have provided strong evidence for the suitability of Caco-2 cells for investigating the functions of NHE isoforms. NHE3 and NHE2 are suggested to be involved in vectorial Na+ transport across the apical membranes in human small intestine and colon. NHE1 is localized basolaterally (53) and is considered to be important for housekeeping functions such as regulation of cell volume and pH. Several previous studies have demonstrated the differential regulation of the apical NHE isoforms NHE2 and NHE3, and even in some cases these isoforms are regulated reciprocally in intestinal epithelial cells. In C2 intestinal epithelial cells, for example, both isoforms are downregulated by cAMP; however, NHE2 is upregulated by serum and unaffected by cGMP or PKC, whereas NHE3 is downregulated by cGMP, PKC, and serum (4, 32). Our present studies indicate the regulation of NHE3 by NO in Caco-2 cells. There was no effect on NHE2 transport activity as estimated by 22Na uptake studies in the presence of 50 μM HOE-694 to block NHE2 but not NHE3. These observations suggest that NO regulates Na+ absorption in Caco-2 cells by exerting specific effects on NHE3.
It could be argued that elevated levels of NO secondary to 1 mM SNAP treatment may also affect permeability in Caco-2 cell monolayers. In this regard, Salzman et al. (47) showed that NO depletes ATP levels and reversibly dilates the tight junctions in cultured Caco-2/bbe (C2) cells. However, these effects were observed only with long-term (6–24 h) treatment in the presence of high concentrations of NO donor, e.g., 5 mM SNAP. In the present studies, a lower concentration of SNAP (1 mM) for a shorter duration (20 min) was used. Therefore, it is very unlikely that the observed changes in NHE3 activity could be due to alterations in the permeability of the Caco-2 monolayers. On the other hand, considering the fact that even if these shorter incubations, i.e., treatment with 1 mM SNAP for 20 min, possibly caused minor permeability changes, our results showing a decrease in NHE3 activity would remain valid, as NHE3 activity only represents the HOE-694-insensitive component of the total22Na uptake.
NO has been shown to act via a variety of second messenger cascades (6, 45, 52), although most of its effects are mediated by activation of soluble guanylate cyclase, leading to an increase in intracellular levels of cGMP. The guanylate cyclase enzyme occurs in soluble (cytosolic) and membrane-bound (particulate) forms (24). The soluble guanylate cyclase contains heme moiety with bound iron, which can serve as an intracellular receptor for NO, leading to its activation. Although the intestinal epithelium contains 95% of the membrane-associated form of guanylate cyclase, NO-donating compounds have been shown to activate the soluble guanylate cyclase in colonic mucosa (57) and production of high levels of cGMP in ileal mucosal scrapings (29). Roczniak and Burns (45) found that cGMP partially mediated the observed effects of NO on NHE activity in rabbit proximal tubule cells. Our findings from the present study suggest that apical NHE activity in Caco-2 cells is inhibited by NO through a cGMP/PKG-dependent pathway. The data with measurements of cGMP levels in Caco-2 cells confirmed that SNAP stimulation of Caco-2 cells resulted in a rapid (∼4.0-fold) increase in cGMP. This increase was blocked by the specific inhibitor of soluble guanylate cyclase LY-83583. To provide more direct evidence for the role of the cGMP pathway, additional studies were performed; for example, the direct effect of cGMP on NHE activity was examined. Incubation of Caco-2 cells with 8-BrcGMP significantly inhibited NHE activity (Fig. 5). Thus we could mimic the effect of SNAP by a biologically active analog of cGMP. Second, LY-83583 and ODQ (Fig. 4), specific inhibitors of soluble guanylate cyclase, completely abrogated the effect of SNAP on NHE activity. Third, KT-5823, a specific inhibitor of cGMP-dependent PKG, abolished the effect of SNAP. Therefore, the above findings convincingly indicate that inhibition of NHE activity by NO occurs via stimulation of soluble guanylate cyclase with a resultant increase in cGMP and activation of cGMP-dependent PKG in Caco-2 cells. Our studies are also supported by the previous findings of McSwine et al. (32) showing inhibition of NHE3, but not NHE2, by cGMP in the Caco-2/bbe cell line. More recently, Cha et al. (10) demonstrated the role of cGMP kinase II in the regulation of NHE3 by cGMP in the PS120 cell line.
The interactions of PKG with NHE3, whether direct or involving some accessory proteins, however, remain a point of further investigation. In this regard, studies have shown that although NHE3 is a phosphoprotein under basal conditions, changes in phosphorylation of NHE3 are not involved in the regulation by growth factors and protein kinases (60). Further studies are therefore needed to elucidate the downstream effectors of SNAP-mediated inhibition of NHE activity via a cGMP/PKG-mediated pathway.
NO-mediated inhibition of NHE activity in Caco-2 cells could also occur via 1) a PKC-mediated pathway or 2) cross activation of a PKA-mediated pathway by cGMP. PKC isozymes have been involved in mediating the effects of NO in some cell types (52). Our data with specific PKC inhibitors (chelerythrine chloride or calphostin C) exclude a role of this signal transduction cascade as a mediator of NO-induced inhibition of apical NHE activity, as these inhibitors failed to show an effect on decrease in NHE activity by SNAP.
High concentrations of cGMP have been shown to cross activate cAMP-dependent protein kinase in smooth muscle cells (11) and in intestine (16). It was thus considered of interest to determine whether the SNAP-induced decrease in NHE activity was secondary to cross talk and/or activation of cAMP. Our results with the specific PKA inhibitor Rp-cAMPS showing no effect on SNAP-mediated inhibition of NHE activity rule out the possible involvement of PKA in the cGMP/PKG-dependent pathway for inhibition of NHE activity in Caco-2 cells.
NO can interact with superoxide anions in cells, leading to the formation of oxidants such as peroxynitrite, with subsequent liberation of hydroxyl radicals (12). Peroxynitrite has been shown in previous studies (5, 23) to mediate its effect on vascular relaxation through metabolic generation of NO. The present studies suggest that NO directly modulates the activity of NHE3 and that this activity is not due to the possible generation of other reactive intermediates for the following reasons: 1) preliminary studies in our laboratory demonstrated that oxidants and free radicals had no significant effects on NHE activity in Caco-2 cells (unpublished observations); 2) use of LY-83583, which acts as an NO inhibitor by blocking the NO-induced activation of soluble guanylate cyclase, strongly indicated that the observed effects are indeed due to NO and not to reactive intermediates; and 3) we could mimic the effects of SNAP, with a structurally dissimilar NO donor, SNP, suggesting that the effects of SNAP were related to generation of NO and were not due to any direct nonspecific effect of the NO donor. These results also emphasize that the observed effects obtained with 1 mM SNAP were not pharmacological in nature.
In the present studies, the observed effects of NO on NHE3 activity appear to be primary in nature and not secondary to changes in the activity of other transporters. NO has previously been shown to inhibit the activity of basolateral Na+-K+-ATPase. Sugi et al. (49) observed a marked decrease in Na+-K+-ATPase activity in T84 cells grown on Transwell inserts in response to long-term treatment with NO. Similarly, a decrease was observed in response to chronic NO induction in the kidney (31). Our data utilizing Transwell inserts showed that SNAP treatment (20 min) in the presence or absence of ouabain, a well-known inhibitor of Na+-K+-ATPase activity, exhibits no significant effects on the observed inhibition of apical NHE activity by SNAP. These results, although indirect, rule out the possible involvement of Na+-K+-ATPase in SNAP-mediated effects on NHE3 activity in Caco-2 cells. The contribution of NHE1 also seems unlikely, since, in the present studies, basal pHi remained unchanged after acute SNAP treatment. Moreover, NO decreased the activity of only NHE3, and not NHE2, so it appears that NO-induced changes are not secondary to alterations in the Na+ concentration or pHi changes caused by NHE1. Interestingly, previous studies by Aizman et al. (1) showing the effects of NO on K+ transport in the rat distal colon suggested that although NO, being a lipophilic molecule, can traverse the plasma membrane, its effects were observed near its site of production because of its rapid inactivation under aerobic conditions. For example, these studies indicated that NO produced in the lumen affected only apical K+ transporters, whereas NO produced at the serosal side affected only basolateral K+ transporters. Additionally, studies in our laboratory have shown that NO decreased the activity of the apical Cl−/OH− exchanger via a mechanism involving PKG and PKC, different from the mechanism suggested in the present studies (46). Under exactly the same conditions, there was no alteration in the activity of the SO /OH− exchanger in the Caco-2 cells (46). These results further emphasize that the effects of NO on NHE3 are specific and are not secondary to changes in other transporter activities, pHi, or Na+concentration.
In summary, the present study, for the first time, demonstrates that NO regulates the apical membrane NHE activity in Caco-2 cells by inhibiting the activity of NHE3 but not NHE2. These studies also present the novel findings of NO-induced inhibition of NHE3 via a PKG-mediated pathway without any involvement of PKC or PKA.
On the basis of our results, we propose a model (Fig.10) demonstrating the mechanism of inhibition of NHE3 by NO. The observed inhibition appears to be mediated by cGMP/PKG-dependent mechanism(s), whereas there is no involvement of the PKC- or PKA-mediated pathway in this process. Because increased levels of NO and inducible NOS are associated with ulcerative colitis, gastroenteritis, and hypoxia-induced dysfunction, our present data would suggest a crucial role of NO in the regulation of human intestinal electrolyte transport in normal and pathological conditions.
These studies were supported by the Department of Veterans Affairs and National Institute of Diabetes and Digestive and Kidney Diseases Grants DK-54016 (P. K. Dudeja), DK-33349 (K. Ramaswamy), and DK-09930 (W. A. Alrefai).
Address for reprint requests and other correspondence: P. K. Dudeja, University of Illinois at Chicago, Medical Research Service (600/151), Chicago VA: West Side Division, 820 South Damen Ave., Chicago, IL 60612 (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2002 the American Physiological Society