Abstract
In our study we have examined the importance of cyclic guanylate monophosphate (cGMP) in NO-mediated intestinal cellular damage. Epithelial cells were harvested from a 20-cm segment of rat proximal small intestine by dispersion using citrate and ethylenediaminetetraacetic acid. Cell viability was assessed by trypan blue dye exclusion. Incubation of cells with the nitric oxide donors, S-nitroso-N-acetyl penicillamine (SNAP) or sodium nitroprusside (SNP) (10–1000 μM) produced a concentration-dependent increase in cell injury and an increase in cellular cGMP formation as determined by immunoassay. In addition, cell injury was also increased by treatment of cells with the cell permeable analogue, dibutryryl cGMP (db cGMP; 0.1–2.0 mM). Suppression of cellular cGMP production by incubating cells with the guanylate cyclase inhibitor LY83583 (5–20 μM) attenuated the damaging actions of SNAP or SNP. However, LY83583 treatment did not reduce ethanol-mediated (10% v/v) cell injury. Furthermore the cytotoxic actions of SNAP or SNP were enhanced by preincubation of cells with the selective cGMP phosphodiesterase inhibitor, zaprinast (10 mM). The damaging actions of SNAP, SNP and db cGMP were reduced by treating cells with superoxide dismutase (100 U/ml). Similarly SNAP, SNP and db cGMP treatments resulted in an increase in the in vitro production of reactive oxygen metabolites as assessed by the fluorescent probe 2′7′ dichlorofluoresein diacetate. These findings indicate that cGMP mediates intestinal cell injury in response to high levels of nitric oxide as produced by the nitric oxide donors, SNAP and SNP. Furthermore these data suggest that the cGMP-induced damage to intestinal epithelial cells involves the generation of reactive oxidants.
Small amounts of NO formed from l-arginine by a constitutive NOS (Moncada et al., 1991) play a role in maintaining microvascular and epithelial integrity in the intestine (Kubes and Granger, 1992; Whittle, 1993). In contrast, excessive production of NO in response to induction of the calcium-independent NOS as a result of endotoxin treatment has been associated with a reduction in the viability of epithelial cells harvested from the gastrointestinal mucosa (Tepperman et al., 1993, 1994). Furthermore large amounts of NO derived from exogenous sources can also damage the intestine epithelial cells. Incubation of intestinal epithelial cells with the NO liberators, SNAP or SNP, resulted in significant increases in cellular injury (Tepperman et al., 1994).
In most tissues including the intestine, NO stimulates the production of cGMP through a direct action on the soluble guanylate cyclase (Younget al., 1993; Moncada et al., 1991; Teppermanet al., 1994). The role of cGMP in NO-mediated intestinal cellular injury is unknown. However, high levels of cGMP have been shown to damage some cell types including cultured neurons (Frandsenet al., 1992, 1993). Furthermore NO-mediated elevations of cGMP have been shown to enhance TNF-mediated cytotoxicity in a number of tumor cell lines (Higuchi et al., 1991) and in neuronal cells (Sherman et al., 1992). Recently Loweth and colleagues (1997) have shown that high concentrations of NO liberated from the NO donor s-nitrosoglutathione could induce injury to a line of pancreatic β cells and this damage was attenuated by an inhibitor of guanylate cyclase. Furthermore, in pheocromocytoma PC12 cells, SNP-mediated cytotoxicity could be enhanced by addition of a non-metabolizable cGMP analogue, 8-Br-cGMP and the injury was reduced by an inhibitor of guanylate cyclase, methylene blue (Nakamura et al., 1997). These data suggest that cGMP, at least in part, is responsible for the cytotoxic actions of large amounts of NO.
It is unknown if NO-mediated injury to intestinal epithelial cells is similarly dependant, to some extent, on cGMP formation. Therefore, in our study, we have examined the effect of cGMP on cell integrity and have investigated cell injury in response to the NO liberators, SNAP and SNP and have determined the role of cGMP in the epithelial cellular responses to NO.
Materials and Methods
Isolation of intestinal epithelial cells.
Nonfasted male Wistar rats (250–300 g) were killed by cervical dislocation, and intestinal epithelial cells were isolated from segments of the intestine as described by Weiser (1973) and modified by Lentze et al., (1985). Briefly, starting from the gastroduodenal junction, a 20-cm segment of proximal small intestine was excised and slowly flushed with 50 ml of a solution containing 0.15 M NaCl and 0.1 mM DTT. The segment was then filled with 5 ml of a solution containing (in mM) 1.5 KCl, 96 NaCl, 27 sodium citrate, 8 KH2PO4and 5.6 Na2HPO4 (pH 7.3), and the proximal and distal ends were ligated. The segment was immersed in PBS at 37°C and bubbled with 95% 02–56% CO2. After 15 min, the instilled solution was removed and discarded.
Cells were collected after intraluminal instillation of Ca++ and Mg++-free PBS containing 1.5 mM EDTA and 0.5 mM DTT. The cells were incubated for 45 min with 5 ml of the EDTA-DTT instillate described above. Cells were washed twice with PBS (pH 7.4) and centrifuged for 5 min at 800 × g. The cells were resuspended in a buffer containing (in mM) 10 N-2-hydroxyethylpiperazine-N1-2-ethanesulfonic acid, 320 sucrose, 1 DTT, as well as (in mg/ml) 0.01 soybean trypsin inhibitor, 0.01 leupeptin and 0.002 aprotinin (pH 7.4). We have previously used these techniques in the isolation of epithelial cells from rat small intestine (Tepperman et al., 1993)
Effect of NO donors.
In some experiments, cells were incubated for 1 hr at 37°C with either SNAP or SNP at concentrations of 10 to 1000 μM. In some of these studies the cells were also incubated in the presence of the guanylate cyclase inhibitor LY83583 (Mulsch et al., 1989). LY83583 (Biomol Research, Plymouth Meeting, PA) was added to the incubation mixture in the concentration range 5 to 20 μM. To examine the specificity of the response to LY83583, cells were incubated for 15 min with 10% w/v ethanol either in the presence (20 μM) or absence of LY83583. Finally, in some studies, the cGMP phosphodiesterase inhibitor M and B 22948 (Zaprinast; 10 mM; gift of Rhone-Poulene Roher, Montreal, Quebec, Canada) was added to the cell suspension 10 min before addition of SNAP or SNP (1000 μM). At the end of the incubation period, cells were assessed for viability, cGMP content and intracellular oxidant production as described below.
Effect of db cGMP.
db cGMP (N2, 21, 0-dibutyryl guanosine -31 51 cyclic monophosphate; Sigma Chemical Co., St. Louis, MO) was added to the cell suspension in the concentration range 0.1 to 2 mM. Incubations proceeded for 1 hr at 37°C. All incubations were performed in the presence of the selective cGMP phosphodiesterase inhibitor, Zap (10 mM). To assess the specificity of the guanosine group in the damaging actions of db cGMP, in a separate study, cells were incubated with db cCMP; (2 mM; Sigma). Cells were incubated for 1 hr at 37°C.
Assessment of cell viability.
More than 90% of the cells harvested were epithelial cells as determined by light microscopy. The remaining cells were identified as macrophages, endothelial cells and red cells. In all experiments, an aliquot of cells was examined for viability as determined by trypan blue dye exclusion. This method has previously been shown to be a reliable index of gastrointestinal epithelial cell injury (Tepperman et al., 1991, 1993). At the end of the incubation period in each experiment, trypan blue (100 μl of a 0.4% w/v solution; Sigma) was added directly to the incubate and mixed. Within 5 min, the number of stained and nonstained cells in a 10-μL aliquot of the suspension was counted using a hemocytometer chamber at an original magnification of 400×. Cells (at least 100) from each fraction were counted in a randomized manner by a naive observer and the number of nonviable cells was determined by light microscopy by counting those cells that failed to exclude the dye. The number of stained cells is expressed as a percentage of the total.
Determination of cGMP content.
The cGMP content of colonic epithelial cells was measured in those cells harvested from control and NO-treated rats. The cells were incubated (30 min) in the presence of Zap (10 mM). At the end of the incubation period, cGMP was extracted by adding 4 mM EDTA, followed by deproteninization in acidic ethanol (1 ml 1 N HCl in 100 ml absolute alcohol). cGMP was assayed as described for the cGMP RIA kit from Amersham Corp., Arlington Heights, IL, using a specific antiserum and [8-3]-guanosine 3′5′ cyclic phosphate (2.6 Ci). The assay sensitivity was 2 fmol/well. The cGMP content was expressed as fmol/106cells/30-min incubation period.
Intracellular oxidant production.
Cells were suspended in a medium containing 100 μM 2′,7′-dichlorofluorescein diacetate (Molecular Probes, Eugene, OR) for 30 min at 37°C. Cells were washed twice with Hanks’ balanced salt solution (Gibco, Burlington, Canada) and sonicated in a buffer containing 50 mM K2HPO4, 0.1 mM EDTA and 0.1% 3-[(3-cholamidopropyl)-dimethylammonio]-1-propane-sulfonate (pH 7.0). The mixture was centrifuged at 2000 × g for 10 min at 4°C. The supernatant was used to determine intracellular oxidant production by monitoring its fluorescence on a Hitachi F-4010 fluorescence spectrophotometer at 502 nm excitation and 523 nm emission. Results are expressed as relative fluorescent intensity per 5 × 106 cells.
Statistical calculations.
Statistical significance was estimated using analysis of variance and Student-Newman Keuls multiple comparisons test or the t test for paired data. P < .05 was the minimum accepted level of significance for all groups. Data are expressed as means (± S.E.) with n equaling the number of cell preparations, each from a different rat.
Results
Effects of NO donors.
Addition of either SNAP or SNP in the concentration range 1 to 1000 μM to the incubation medium resulted in an increase in the percentage of non-viable epithelial cells (fig.1). Significant increases in cell injury in response to SNAP or SNP were observed in response to concentrations of each NO donor as low as 1 μM. Similarly, incubation of intestinal epithelial cells with SNAP or SNP resulted in an increase in cellular cGMP content (fig. 2). However, only concentrations of 100 and 1000 μM of either SNAP or SNP resulted in significant (P < .05; n = 6–8) increases in cGMP content when compared to control levels.
The effect of increasing concentrations (1–1000 μM) of (A) S-nitrososo-N-acetyl-penicillamine (SNAP) or (B) sodium nitroprusside (SNP) in the incubation medium on the viability of rat intestinal isolated cells as assessed by trypan blue dye uptake. Cell viability is expressed as the mean (+S.E.) percentage (%) of nonviable cells (n = 6–8 cell preparations) Asterisks indicate significant (P < .05) differences from control as determined by analysis of variance and Student-Newman-Keuls test.
The effect of increasing concentrations (1–1000 μM) of (A) S-nitrososo-N-acetyl-penicillamine (SNAP) or (B) sodium nitroprusside (SNP) in the incubation medium on rat intestinal cellular cyclic GMP (cGMP) formation. cGMP formation is expressed as fmol/106 cells and is displayed as the mean (+S.E.) of six to eight cell preparations. Asterisks indicate significant (P < .05) differences from control as determined by analysis of variance and Student-Newman-Keuls test.
SNAP and SNP in the concentration of 1000 μM resulted in a significant (P < .05) increase in the percentage of nonviable intestinal epithelial cells (fig. 3). Preincubation of SNAP- or SNP-treated cells with the cGMP phosphodiesterase inhibitor, M and B 22948 (Zap) at a concentration of 2 mM significantly (P < .05)) increased the degree of cell injury in response to either SNAP or SNP alone (fig. 3). Zap treatment resulted in increases of 31 + 1% (n = 6–8) and 27 + 1% (n = 6–8) over the levels observed in response to SNAP and SNP alone, respectively. By itself, Zap did not have a significant effect in cellular viability.
The effect of SNAP or SNP (1000 μM) in the presence or absence of the cGMP phosphodiesterase inhibitor Zaprinast (Zap; 10 μM) on the mean (+S.E.) percentage (%) of nonviable cells in preparations of rat intestinal isolated epithelial cells. Asterisks indicate significant (P < .05) increases in the presence of Zap over SNAP or SNP alone (n = 6–7 cell preparations) as assessed by the t test for paired data.
Effect of guanylate cyclase inhibitor, LY83583.
Addition of the guanylate cyclase inhibitor, LY83583 in the concentration range 5 to 20 μM, to intestinal epithelial cells challenged with either SNAP or SNP (1000 μM; n = 6–7) resulted in a dose-dependent reduction in the extent of cell injury (fig.4). Significant (P < .05) decreases in the extent of trypan blue uptake were evident in response to treatments with 10 and 20 μM LY83583. These concentrations of LY 83583 did not affect cellular viability. Furthermore, we also determined that 10 and 20 μM LY83583 also reduced cellular cGMP content from 17 ± 4 fmol/106 cells to 9 ± 2 fmol/106 cells and 7 ± 3 fmol/106 cells, respectively, in cells challenged with SNAP (1000 μM;n = 5–6).
The effect of increasing concentrations (5–20 μM) of the guanylate cyclase inhibitor LY83583 on intestinal cellular viability in response to addition of SNAP (A; 1000 μM) or SNP (B; 1000 μM) to the incubation medium. Cell viability was estimated by trypan blue dye uptake and is expressed as the mean (+S.E.) of nonviable cells (n = 6–7 cell preparations). Asterisks indicate significant (P < .05) differences from SNAP or SNP alone as determined by analysis of variance and Student-Newman-Kuels test. By itself, LY83583 in concentrations of 10 and 20 μM did not significantly affect cell viability.
In contrast to its effect on NO-mediated cell injury, LY83583 (5 and 20 μM) did not significantly reduce the cytotoxic effect of addition of ethanol (10% w/v) to the cellular incubation medium.
Detection of oxidative product formation using dichlorofluorescein fluorescence is displayed in figure 5. Both SNAP and SNP resulted in a significant (P < .05) increase in fluorescence of the dichlorofluorescein marker. Addition of LY83583 (20 μM) significantly reduced the fluorescence intensity in response to treatment of cells with either SNAP or SNP.
The effect of addition of SNAP ( ;1000 μM) or SNP (▨; 1000 μM) to the incubation medium in the presence or absence of LY83583 (20 μM) on the in vitro production of intracellular reactive oxygen metabolites in rat intestinal isolated cells. Cells are also incubated in the presence of dibutyryl cGMP (db cGMP ⊞; 0.2 and 2.0 mM). Results are expressed as fluorescence intensity per 5 × 106 cells and are means (+S.E.;n = 7–9 cell preparations) Daggers indicate significant (P < .05) increases over control while asterisks indicate significant (P < .05) decreases from SNAP or SNP alone as determined by the t test for paired data.
Effect of db cGMP.
Incubating intestinal epithelial cells with the cell permeant analogue db cGMP, in the concentration range of 0.1 to 2.0 mM, resulted in a dose-dependent increase in cell damage as assessed by trypan blue dye uptake (fig.6). Significant increases in cell injury were evident in response to db cGMP in concentrations of 0.2 to 2.0 mM. Addition of the derivative dibutyryl cytidine cyclic monophosphate (db cCMP; 2.0 mM) did not significantly increase the extent of cell injury. In addition, db cGMP in the doses of 0.2 and 2.0 mM significantly increased dicholorfluorescien fluorescence in suspensions of intestinal cells (fig. 5).
The effect of increasing concentrations (0.1–2.0 mM) of dibutyryl cyclic GMP (│; db cGMP) in the incubation medium on viability of rat intestinal isolated cells. Cells were also incubated with the derivative, dibutyryl cytidine cyclic monophosphate (▨; db cCMP; 2mM). Cell viability was estimated by trypan blue dye uptake. Data are expressed as mean (+S.E.) percentage (%) of nonviable cells (n = 7 cell preparations). Asterisks indicate significant (P < .05) differences from control as determined by analysis of variance and Student-Newman-Kuels test.
Effect of superoxide dismutase.
In cells incubated in the presence of SNAP (1000 μM), SNP (1000 μM) and db cGMP (2 mM), cell damage was increased over control levels (fig.7). Addition of SOD (1000 and 2000 U/ml) resulted in a significant reduction in the extent of cell injury in response to the NO donors and db cGMP. SOD (1000 or 2000 U/ml) by itself did not significantly alter the extent of cell injury observed in control cells.
The effect of superoxide dismutase (SOD; 1000 and 2000 U/ml) on rat intestinal cellular viability in response to db cGMP (2 mM), SNAP (1000 μM) or SNP (1000 μM). Cell viability was estimated by trypan blue dye uptake and assessed as mean (+S.E.) percentage (%) of nonviable cells (n = 6 cell preparations). Asterisks indicate significant differences from respective controls by t test for paired data.
Discussion
The results of our study indicate that the integrity of epithelial cells isolated from the mucosa of the rat proximal small intestine is reduced upon exposure to high concentration of the NO donors, SNP and SNAP. This confirms the early findings of Lopez-Belmonte et al., (1993) demonstrating that release of high levels of NO,in vivo, in response to local infusions of NO donors induced gastric mucosal injury. Furthermore, Whittle (1993) and Teppermanet al., (1994) have shown that incubation of intestinal epithelial cells with high concentration of NO donors was also accompanied by an increase in the extent of cell injury.
NO stimulates the production of cGMP through a direct action on the soluble guanylate cyclase (see Moncada et al., 1991) and, in our study, both SNP and SNAP increased the levels of cGMP in intestinal epithelial cells. Similar increases in colonic cellular cGMP have been previously demonstrated in response to challenge by high concentrations of the NO donor, SNAP (Tepperman et al., 1994). Additional evidence from our study that NO-mediated damage involves cGMP formation is demonstrated by the finding that the injurious actions of SNAP and SNP are attenuated by the guanylate cyclase inhibitor LY83583. This confirms the previous finding that the cytotoxicity induced by SNP in PC12 cells was reduced by the inhibition of guanylate cyclase, using methylene blue (Nakamura et al., 1997)
However, in our study, increases in the degree of cell injury were also observed to occur at concentrations of SNP and SNAP as low as 1 μM. However, those concentrations were not associated with an increased cellular content of cGMP. Although this finding may only reflect the relatively low sensitivity of the commercially available immunoassay to detect small increases in cGMP produced by low concentrations of NO donors, these results may also suggest that the damaging actions of NO, at least at the lowest concentrations of the NO liberators used here, are not associated with cGMP and are suggestive of a cGMP independent mechanism(s) of intestinal cellular injury. Indeed, although the precise mechanism of NO-induced cell damage is unknown, it could result at least in part, from direct cytotoxicity of the NO radical (seeMoncada et al., 1991). NO can directly interact with a variety of intracellular targets that contain either haem groups on nonsulfur centers including respiratory carriers and metabolic enzymes (McDaniel et al., 1996). The functional consequences of these interactions may mediate some of the cytotoxic effects of NO without the necessity of cGMP generation. A similar dissociation of the cellular effects of NO and cGMP have previously been demonstrated byCampbell et al., (1996) in which the effects of nitrosothiols affected biological responses in cardiac myocytes in the absence of alterations in cGMP levels.
The results of our study suggest that cGMP can function as a direct mediator of cell injury in the small intestine. This is based on findings that the extent of NO-mediated cell injury is augmented by the cGMP phosphodiesterase inhibitor, Zap. The concentration of Zap used in this study has been shown previously to enhance cGMP levels in intestinal cells (Tepperman et al., 1994). These data are analogous to the findings of Nakamura et al., (1997) in which it was shown that the damaging actions of sodium nitroprusside in a pheochromocytoma cell line was enhanced by coincubation with a cGMP analogue. Furthermore, Pollman et al., (1996) have demonstrated that Zap potentiated the effect of NO on vascular smooth muscle cell injury in vitro. Our data also demonstrate that db cGMP exerts a direct damaging action on cells. Although db cGMP affected cell integrity, the dibutyryl moiety was not responsible for this action because dibutyryl cyclic cytidine had no damaging actions. There is a considerable literature detailing the role of cGMP as a mediator of cell injury in nonintestinal tissues and cells. Direct effects of cGMP have been demonstrated in retinal degeneration (Lolleyet al., 1977) and in peptide-induced degeneration of insect muscle (Schwartz and Truman, 1984). Similarly, cGMP analogues have been shown to promote cell death in preparations of pancreatic β cells (Loweth et al., 1997) and vascular smooth muscle cells (Pollman et al., 1996) High levels of cGMP has been shown to damage neuronal cells (Frandsen et al., 1993) and increase the sensitivity to established injurious agents (Higuchi et al., 1991; Frandsen et al., 1992).
The mechanisms through which an increase in cGMP increases intestinal cell injury is unknown. In many tissues, including the gastrointestinal mucosa, NO analogues and cGMP have been shown to enhance calcium influx (Magliola and Jones, 1990; Tripp and Tepperman, 1996). Similarly, increases in cellular cGMP, or addition of exogenous cGMP have also been shown to result in increases in cytosolic calcium in a number of cell types (Geiger et al., 1992; Desole et al., 1994) A sustained increase in cytosolic calcium is associated with damage in many cells including gastric mucosal cells (Farber, 1990;Tepperman et al., 1991). Increases in cGMP content are also associated with increases in levels of cytotoxic oxygen radicals. In our study, the damaging effects of SNAP, SNP and db cGMP were each reduced by incubation of cells with the oxidant scavenger, superoxide dismutase. Although it is unlikely that SOD crosses the plasma membrane, it presumably functions extracellularly to scavenge oxidants released into the incubation medium. Garthwaite and Garthwaite (1988)have shown that in rat cerebellar slices, oxygen radical generation mimicked the pattern of cGMP-mediated cytotoxicity. The effects of hydrogen peroxide in platelets is accompanied by an increase in cGMP formation (Ambrosio et al., 1994). Similarly, inhibition of guanylate cyclase have been associated with protection against oxidative stress in neurosecretory cells (Klyszcz-Nasko et al., 1993). In our study, NO donors as well as db cGMP increased cellular oxidant production and the effects of SNAP and SNP were reduced by LY83583. This confirms previous findings in which LY83583 has been shown to inhibit oxidant production in neutrophils (Sundqvist and Axelsson, 1993) and inhibit antioxidant enzyme metabolism in bovine intestinal mucosa (Luond et al., 1993). The inhibition of SNAP or SNP oxidant mechanism by LY83583 is not complete. Similarly, high concentrations of SOD did not completely inhibit SNAP- and SNP-mediated cell damage. These data suggest that oxidants may be produced by non-cGMP requiring mechanisms and that NO/cGMP-mediated cell injury is not exclusively dependant on oxidant formation.
In summary, the data from our group of studies suggest that NO-induced intestinal cell injury is mediated, at least in part, by an increase in cellular cGMP formation. Furthermore these data also demonstrate that cGMP has a direct cytoxic action on intestinal mucosal cells in vitro. The cytotoxic effects of cGMP appear to be as a result of an increase in cellular reactive oxidant levels.
Footnotes
-
Send reprint requests to: Dr. B. L. Tepperman, Department of Physiology, The University of Western Ontario, London, Ontario Canada N6A 5C1.
-
↵1 This work was supported by Grant MT 6426 from the Medical Research Council of Canada.
- Abbreviations:
- NO
- nitric oxide
- cGMP
- cyclic guanylate monophosphate
- db cCMP
- dibutyryl cyclic cytidine monophosphate
- SNAP
- s-nitrosos-N-acetyl penicillamine
- SNP
- sodium nitroprusside
- Zap
- Zaprinast
- SOD
- superoxide dismutase
- PBS
- phosphate buffered saline
- DTT
- dithiothreitol
- EDTA
- ethylenediamine-tetraacetic acid
- Received July 2, 1997.
- Accepted November 3, 1997.
- The American Society for Pharmacology and Experimental Therapeutics