Recent findings on the mode of action of laxatives: the role of platelet activating factor and nitric oxide.

Platelet activating factor (PAF) is a phospholipid mediator of inflammation and stimulates anion secretion in animals and in isolated preparations of human colon. Nitric oxide (NO), synthesized from the amino acid L-arginine, is an important enteric inhibitory neurotransmitter. In addition, NO-donating compounds stimulate anion secretion in rat and guinea-pig colon. In this article, Angelo A. Izzo and colleagues review the key pharmacological features of the involvement of NO and PAF in the action of laxatives and propose that the classification of laxatives should take into account the important implications of these endogenous mediators.

The osmotic and intrinsic mechanisms of the pharmacological laxative action of oral high doses of magnesium sulphate. Importance of the release of digestive polypeptides and nitric oxide.

A common use for high doses of oral magnesium salts is to produce a laxative effect to treat constipation. In the intestinal lumen the poorly absorbable magnesium ions (and other ions such as sulphate) exert an osmotic effect and cause water to be retained in the intestinal lumen. This increases the fluidity of the intraluminal contents and results in a laxative action. Although the laxative action of magnesium is thought to be due to a local effect in the intestinal tract, it is also possible that released hormones such as cholecystokinin or activation of constitutive nitric oxide synthase might contribute to this pharmacological effect. Under normal circumstances the pharmacological administration of high doses of oral magnesium salts is safe and some salts--such as magnesium hydroxide--also have an antacid effect to neutralize stomach acid. However, high doses of magnesium or prolonged use may allow sufficient absorption into the systemic circulation to cause renal or other organ toxicity.

Relationship between nitric oxide and platelet-activating factor in castor-oil induced mucosal injury in the rat duodenum.

The modulation of platelet activating factor (PAF) formation in duodenal tissue by nitric oxide (NO) released in response to castor oil was studied in rats pretreated with NG-nitro-L-arginine methyl ester (L-NAME, 6.25-25 mg/kg, i.p.), an inhibitor of NO synthase, NG-nitro-D-arginine methyl ester (D-NAME, 25 mg/kg, i.p.), the inactive enantiomer of L-NAME or isosorbide-5-mononitrate (IMN, 30-90 mg/kg, p.o.), a NO donating compound. Castor oil (2 ml/rat orally) increased PAF production in the rat duodenum 3 h after challenge. L-NAME, but not D-NAME, enhanced the amount of PAF formed by duodenal tissue, while IMN (30-90 mg/kg) counteracted the effects of L-NAME (12.5 mg/kg) and also reduced PAF release in the tissue of rats treated with castor oil. L-NAME 12.5 mg/kg, but not D-NAME, enhanced both macroscopic damage and acid phosphatase release induced by castor oil. These effects were reduced by a PAF antagonist BN 52021 (3-t-Butyl-hexahydro-4, 7b, 11-trihydroxy-8-methyl-9H-1, 7a-epoxymethano-1H, 6aH-cyclopenta [c] furo [2, 3b] furo [3'2':3,4] cyclopenta [1.2-d]furan-5,9,12(4H)trione) 10 and 20 mg/kg i.p. Such findings suggest that endogenous nitric oxide could reduce PAF biosynthesis in castor oil-treated rats.

NG-nitro-L-arginine methyl ester reduces senna- and cascara-induced diarrhoea and fluid secretion in the rat.

Senna (60 mg/kg orally) and cascara (800 mg/kg orally)-induced diarrhoea and net fluid secretion were studied in rats for a time period of 1-8 h. NG-Nitro-L-arginine methyl ester (L-NAME) (2.5-25 mg/kg i.p. twice, 15 min before and 4 h after laxative administration), an inhibitor of nitric oxide synthase, reduced the diarrhoeal response. This effect was counteracted by L-arginine (600 and 1500 mg/kg i.p. 15 min before laxative administration), the precursor of nitric oxide (NO). The senna- and cascara-stimulated fluid secretion was reduced by NG-nitro-L-arginine methyl ester 25 mg/kg i.p. (twice, 15 min before and 4 h after laxative administration), while the stereoisomer NG-nitro-D-arginine methyl ester (D-NAME) 25 mg/kg i.p. was without effect. These results suggest a possible involvement of NO in senna- and cascara-induced diarrhoea and fluid secretion.

Eicosanoids and histamine mediate C5a-induced electrolyte secretion in guinea pig ileal mucosa.

C5a is a biologically active polypeptide formed during the course of complement activation and is known to possess histamine-releasing and neutrophil chemotactic properties. In the present study, we have demonstrated that C5a can regulate electrolyte transport across guinea pig ileum, and we have investigated its mechanism of action. Segments of ileum stripped of longitudinal muscle were mounted in Ussing chambers (Krebs' buffer, 37 degrees C, 95% O2/5% CO2) for monitoring short-circuit current (Isc). Serosal application of C5a evoked a transient increase in Isc with an EC50 value of 5.0 nM indicating a potent effect. The C5a-induced increase in Isc was abolished by elimination of both Cl- and HCO3- from the Krebs' solution. Pretreatment with the cyclooxygenase inhibitor indomethacin (5 microM), the neurotoxin tetrodotoxin (0.5 microM) and the H1 receptor antagonist pyrilamine (0.5 microM) reduced the effect of C5a, but the muscarinic antagonist atropine (0.5 microM) was without effect. C5a (100 nM) also evoked the release of histamine (measured by radioimmunoassay in the serosal bathing fluid) by 282% of the control value. In conclusion, in the guinea pig ileum C5a stimulates mucosal anion secretion by releasing histamine and cyclooxygenase products of arachidonic acid. The response is also mediated, in part, via non-chloinergic enteric nerves.

Preferential binding of the novel prostaglandin SC-46275 to canine gastric versus intestinal receptors.

Prostaglandins (PGs) in the E-series exhibit potent gastric antisecretory activity, but can also cause diarrhea, which is mediated via PGE receptors. SC-46275, an omega-chain cyclopentenyl analog of the E-type PG enisoprost, was evaluated with other E-PGs for PGE receptor binding activity in gastric and intestinal tissues. SC-46275, enisoprost, misoprostol and PGE1 were first evaluated in enriched canine gastric parietal cells with [3H]misoprostol free acid binding and subsequently with [3H]PGE1 binding in canine intestinal tissues where misoprostol free acid had weak receptor binding activity. The receptor binding potency of SC-46275 (IC50, 0.013 mM) in enriched canine parietal cell preparations was found to be much greater than misoprostol and enisoprost (IC50, 10 and 8 nM), whereas PGE1 had the least potency (IC50, 37 nM). Similar relative potencies for these PGs were also obtained in the inhibition of histamine-stimulated acid secretion in enriched parietal cell preparations. In small intestinal mucosal and muscle membranes, the receptor binding potency of SC-46275 (IC50, 13 and 20 microM) was much less than misoprostol or enisoprost (IC50, 0.39-1.2 microM) and substantially less than PGE1 (IC50, 0.017 and 0.066 microM). This weak binding activity of SC-46275 in intestinal tissues is consistent with its reported weak diarrheagenic activity in the rat. These results suggest that SC-46275 binds preferentially to gastric vs. intestinal PGE receptors and is specific for the EP3 receptors.

Colitis reduces short-circuit current response to inflammatory mediators in rat colonic mucosa.

Inflammatory mediators may contribute to the diarrhea associated with colitis. Although the secretory action of such mediators is reported in normal tissue, there is little information regarding their effects on inflamed tissue. We examined the short-circuit current response (Isc) to these mediators, in mitomycin-C (MC)-induced colitis, a model with histological similarities to colitis in man. Rats were injected once with MC (3.25 mg/kg, intraperitoneally) or vehicle. The colons were removed three and seven days later and mounted, devoid of muscularis, in Ussing chambers for measurement of Isc, potential difference (PD), and resistance (Rt). MC-treated rats had diarrhea after three days, and microscopic studies revealed colonic inflammation. There were no significant differences in Rt, PD, and Isc between control and MC-treated tissues at three and seven days. Maximal increases in Isc to bradykinin, prostaglandin E1, carbachol, substance P, and serotonin were depressed at three and/or seven days after MC. The Isc response to theophylline was not affected. Theophylline activates secretion through an intracellular mechanism; the other agonists act by interaction with epithelial cell membranes. Therefore, the mechanism for the decreased Isc may result from uncoupling of receptors to second-messenger systems or desensitization of receptor-linked secretory mechanisms.

Bradykinin acting on B2 receptors contracts colon circular muscle cells by IP3 generation and adenylate cyclase inhibition.

Receptor subtypes and intracellular signaling events involved in bradykinin-evoked contraction of colonic circular muscle are unknown. We studied the roles of inositol trisphosphate (IP3) and cyclic AMP generation and the selectivity for B1 and B2 receptors in guinea pig colon. Bradykinin induced concentration-dependent contraction of circular muscle strips with an EC50 of 2 x 10(-8) M that was inhibited by the B2 antagonist D-Argo-(Hyp3,Thi5,8,D-Phe7)-bradykinin but not the B1 antagonist des-Arg9-[Leu8]bradykinin. The B1 agonist des-Arg9-bradykinin did not evoke contraction or relaxation. Bradykinin induced concentration-dependent shortening of isolated myocytes from circular muscle with an EC50 of 2 x 10(-11) M that was inhibited by the B2 but not the B1 antagonist, confirming the myogenic nature of the bradykinin receptors. Persistence of myocyte contraction in a calcium-free medium with EGTA confirmed the lack of dependence on extracellular calcium. In colon muscle tissue, bradykinin evoked concentration-dependent IP3 generation with an EC50 of 10(-7) M and a maximal level of 58 +/- 17 pmol/mg of protein at 10(-4) M that was inhibited by the B2 but not the B1 antagonist. Bradykinin, acting on B2 receptors, inhibited cyclic AMP formation after forskolin (10(-5) M) with an EC50 of 3 x 10(-8) M and maximal inhibition of 48% at 10(-5) M. In conclusion, bradykinin induces colon muscle contraction via myogenic non-B1 receptors, which are likely of the B2 subtype, with phosphoinositide turnover activation and adenylate cyclase inhibition.(ABSTRACT TRUNCATED AT 250 WORDS)

Dissociation of castor oil-induced diarrhoea and intestinal mucosal injury in rat: effect of NG-nitro-L-arginine methyl ester.

1. Castor oil (2 ml orally) produced diarrhoea in rats 1-7 h after challenge, which was associated with gross damage to the duodenal and jejunal mucosa. 2. The injury was accompanied by release of acid phosphatase into the gut lumen, indicating cellular injury. 3. Intraperitoneal injection of the nitric oxide (NO) synthase inhibitor NG-nitro-L-arginine methyl ester (L-NAME, 2.5-50 mg kg-1 twice), prevented the diarrhoea. The dose of L-NAME (50 mg kg-1) completely blocked the diarrhoea but increased the release of acid phosphatase and worsened the gross damage. 4. The NO donating compound, isosorbide-5-mononitrate (IMN, 150 mg kg-1 twice) reversed the effects of L-NAME (50 mg kg-1) on castor oil-induced diarrhoea, gross damage and acid phosphatase release. 5. The apparent dissociation of the diarrhoeal and intestinal mucosal damaging effects of castor oil suggest that NO has a protective effect on the rat duodenal and jejunal mucosa, but that NO mediates, in part, the diarrhoea effect of this laxative.

Nitric oxide involvement in sodium choleate-induced fluid secretion and diarrhoea in rats.

Bile salt-induced diarrhoea, net water and electrolyte secretion, gastrointestinal transit and nitric oxide (NO) synthase activity were studied in rats. NG-Nitro-L-arginine methyl ester (2.5-25 mg/kg i.p.), an inhibitor of NO synthase, and dexamethasone (0.03-0.3 mg/kg i.p.), an inhibitor of the inducible isoform of NO synthase, antagonized the diarrhoeal response. The NO precursor, L-arginine and isosorbide-5-mononitrate (an NO donor), reversed the inhibitory effect of NG-nitro-L-arginine methyl ester. The bile salt-stimulated fluid secretion, transit through the gut and NO synthase all were inhibited by NG-nitro-L-arginine methyl ester (but not NG-nitro-D-arginine methyl ester). NO synthase activity also was inhibited by dexamethasone. The results are consistent with bile salt induction of epithelial cell injury and concomitant synthesis of NO, mainly through activation of the inducible form of the enzyme. We believe that in this study NO is a mediator of intestinal secretion and motility changes that enhance transit of luminal contents through the gut, resulting in diarrhoea.

Nitric oxide as a mediator of bisacodyl and phenolphthalein laxative action: induction of nitric oxide synthase.

Bisacodyl and phenolphthalein are diphenylmethane laxatives that have effects on intestinal water and electrolyte transport and smooth muscle contractility. Nitric oxide (NO) is produced in the intestine, where it stimulates electrolyte secretion and relaxes smooth muscle. Therefore, we studied in rats the effect of these laxatives on diarrhea, fluid transport in vivo, gastrointestinal transit and NO synthase activity in the absence and presence of inhibitors of NO synthesis. Both laxatives (50 mg/kg p.o.) produced diarrhea, which was delayed in onset by 25 mg/kg (i.p.) of the NO synthase inhibitor NG-nitro-L-arginine methyl ester (L-NAME). The L-NAME effect was reversed by the NO donor isosorbide-5-mononitrate (30-120 mg/kg i.p.). L-Arginine (600 and 1500 mg/kg i.p.) prevented the inhibitory effect of L-NAME on diarrhea. The laxatives evoked water and electrolyte secretion and enhanced the transit of a suspension of charcoal through the gastrointestinal tract. This was inhibited by L-NAME but not D-NAME. The inhibitor of inducible NO synthase, dexamethasone (0.03-0.3 mg/kg i.p.), prevented the effects of both laxatives on electrolyte and water transport. Stimulation by these drugs of NO synthase was also inhibited by dexamethasone. The results demonstrate that bisacodyl and phenolphthalein stimulate water and electrolyte secretion, promote transit of intraluminal contents and produce diarrhea in association with enhanced production of NO. Furthermore, it appears that the NO is derived principally from activation of an inducible form of NO synthase.

Nitric oxide as a mediator of the laxative action of magnesium sulphate.

1. Magnesium sulphate was studied for its effects on diarrhoea, fluid secretion, gastrointestinal transit and nitric oxide (NO) synthase activity in rats. 2. At a dose of 2 g kg-1 orally magnesium sulphate produced diarrhoea that was delayed in onset and intensity in a dose-related manner by the NO synthase inhibitor NG-nitro-L-arginine methyl ester (L-NAME). This was prevented by the NO precursor, L-arginine and the NO donating compound, isosorbide-5-mononitrate (IMN). 3. Nitric oxide synthase activity was stimulated in gut tissue from rats given magnesium sulphate and this was inhibited by L-NAME. Dexamethasone (1 mg kg-1, i.p.), an inhibitor of inducible NO synthase, had no effect on magnesium sulphate-induced diarrhoea. 4. Magnesium sulphate stimulated fluid and electrolyte accumulation in the intestinal lumen; these effects were prevented by L-NAME but not D-NAME. 5. Gastrointestinal transit of a non-absorbable marker (charcoal suspension) was increased by oral magnesium sulphate from a mean value of 54.1% to 72.9% (P < 0.01), and this was prevented by pretreatment with L-NAME. 6. The results demonstrate that oral magnesium sulphate produces diarrhoea in rats by increasing the accumulation of fluid in the intestinal lumen and enhancing flow from the proximal to distal intestine. The mechanism involves release of NO, probably through stimulation of the constitutive form of NO synthase. Whether or not the effects of magnesium sulphate are due to an osmotic action or an intrinsic effect of the magnesium or sulphate ions cannot be determined from these experiments.

Clinical relevance of basic research in peptic ulcer disease.

Historically, the interplay between basic research and clinical observation has been essential in the development of new therapies for peptic ulcer disease. That histamine is an important regulator of acid secretion emerged from basic research, followed by the clinical development and use of the H2-receptor antagonists. Basic research contributed again by defining the importance of H+/K(+)-ATPase in acid secretion, resulting in a new class of useful antisecretory agents. Basic studies also gave us prostaglandins (PG) as mucosal protective agents. As 'replacement' therapy, clinicians have found that PG are protective against non-steroidal anti-inflammatory drug (NSAID)-induced gastric ulcer (GU). Physiologic studies established that somatostatin is a potent inhibitor of acid secretion, providing the stimulus for clinical studies in Zollinger-Ellison (ZE) Syndrome with a synthetic analog (octreotide). Work on isoforms of the parietal cell gastrin receptor has shown differences in the cytoplasmic domain for G protein coupling. This will aid in understanding how receptor changes and coupling to second messengers relate to the aetiopathogenesis of abnormal gastric secretion. Immune cells express mRNA for histamine, muscarinic and gastrin receptors, supporting the relevance of mucosal immunology in gastroenterology, especially in light of Helicobacter pylori-associated gastritis and ulcers. Lab research has revealed a potential role for basic fibroblast growth factor (bFGF), and another endogenous peptide BPC-15, in ulcer healing. The former substance may be responsible for the antiulcer efficacy of sucralfate. Intensive basic work on how H. pylori organisms attach to gastric cells and initiate inflammatory reactions in the mucosa will have unquestionable impact on improved therapy for peptic ulcer disease.

Stereospecific actions of misoprostol on rat colonic electrolyte transport.

Misoprostol (Miso) produces a mild, transient diarrhea in some patients, which is believed to be partly due to intraluminal fluid accumulation. To better understand this diarrheagenic action, we compared the effects of Miso, its 4 stereoisomers (11R16R, 11R16S, 11S16S, 11S16R), misoprostol free acid (Miso-FA), and 16,16-dimethyl PGE2 (dmPGE2) on rat colonic electrolyte transport in vitro. Increases in short-circuit current (Isc) were measured (after serosal addition) in segments of mucosa stripped of muscularis and mounted in Ussing chambers. The rank order of apparent potencies, in terms of threshold, were (microM): 11R, 16S (1.2) approximately dmPGE2 (1.0) > Miso-FA (10.0) approximately Miso >> 11R, 16R; 11S, 16R; 11S, 16S (all inactive at 100 microM). The response to dmPGE2 and Miso was attenuated in the presence of the Na+/K+/Cl- co-transport inhibitor bumetanide (100 microM). Pretreatment with atropine (0.1 microM) did not affect the Isc response to Miso, Miso-FA, or dmPGE2. Tetrodotoxin partially attenuated (39 +/- 9% inhibition) the response to Miso-FA, but did not affect Miso or dmPGE2. In conclusion, Miso increases Cl- secretion across rat colonic mucosa through a direct action on epithelial cells. The activity resides in the 11R,16S isomer, thus implying a stereospecific interaction at PGE receptors. The effect of Miso to stimulate epithelial Cl- secretion might contribute to its diarrheagenic action in vivo.

Effect of the thiol-oxidizing agent diamide on NH2Cl-induced rat colonic electrolyte secretion.

The granulocyte-derived oxidant, monochloramine (NH2Cl), is known to stimulate chloride ion secretion in rat distal colonic mucosa mounted in Ussing chambers, through mechanisms that are sensitive and insensitive to tetrodotoxin (TTX). The possible role of intracellular thiols, in the mechanism of action of NH2Cl as a secretagogue, was evaluated with the thiol-oxidizing agent diamide and by measuring tissue sulfhydryl levels in response to NH2Cl. Serosal exposure to the antioxidant glutathione (0.25 mM), 5 min before NH2Cl (50 microM) addition, decreased the maximal effect of 50 microM NH2Cl on short-circuit current (Isc). The NH2Cl-stimulated increase in Isc was not affected by mucosal amiloride (5 microM). Pretreatment with 0.1 mM diamide shortened the lag period before the increase in Isc in response to NH2Cl, but it did not affect the maximal increase in Isc. Although TTX (0.5 microM) increased the lag time for achievement of the maximal Isc response to NH2Cl, the neurotoxin did not inhibit the effect of diamide, suggesting that diamide acts primarily on the nonneural component of NH2Cl-stimulated secretion. Incubation of colonic mucosa with NH2Cl, with or without diamide, decreased cellular acid-soluble sulfhydryl concentrations. Taken together, the results support a role for epithelial cell thiols in NH2Cl-stimulated electrolyte secretion by the rat colon.

Direct evidence for nitric oxide stimulation of electrolyte secretion in the rat colon.

Nitric Oxide (NO) is synthesized in the intestinal tract and may serve as a physiological regulator of intestinal ion transport and/or a pathophysiologic mediator of secretory diarrhea associated with inflammatory mucosal diseases. Indirect approaches, employing inhibitors of nitric oxide synthase or compounds capable of donating NO in solution, have been used to demonstrate the effects on gastrointestinal muscle and the mucosa. To determine directly whether nitric oxide itself is capable of stimulating electrolyte secretion we mounted muscle-stripped rat distal colon in Ussing chambers and monitored short-circuit current (Isc), as an indicator of effects on mucosal ion transport. Comparisons were made to sodium nitroprusside (SNP). NO and SNP stimulated concentration-dependent (0.1 microM to 100 microM) increases in Isc, with NO being more potent than SNP. The EC50 for NO was approximately 8 microM compared to a value < 20 microM for SNP. The response to NO was immediate. In contrast, SNP required a mean lag-time of 41 +/- 4 seconds, and a significantly longer time was required for SNP to reach its maximum effect. The response to both of these agonists was blocked by bumetanide, indicating that they were stimulating a chloride ion secretory response. The cyclooxygenase inhibitor piroxicam, the neurotoxin tetrodotoxin and the inhibitor of guanylate cyclase, methylene blue, all inhibited the response to both agonists. These studies demonstrate that NO itself can stimulate chloride secretion by the rat colonic mucosa through a prostaglandin-dependent, and partially neural mechanism that may involve guanylate cyclase.

Oxidant-evoked release of acetylcholine from enteric neurons of the rat colon.

In the presence of halides, granulocytes generate hypochlorous acid and, subsequently, chlorinated amines (chloramines). These lipophilic, potent reactive oxygen metabolites may contribute to the mucosal pathophysiology associated with inflammatory bowel disease. A common symptom of inflammatory bowel disease is mucosal secretion of fluid and electrolytes, leading to diarrhea. Because acetylcholine (Ach) can stimulate colonic fluid secretion, we determined the effect of monochloramine (NH2Cl) on Ach release by mucosal/submucosal nerves. Mucosa from the rat colon was separated from outer muscle layers and minced before incubation with [14C]choline to label stores of Ach in cholinergic neurons. Release of [14C]Ach was evoked with NH2Cl in the absence and presence of 5-aminosalicylic acid, glutathione, nordihydroguaiarectic acid or the cyclooxygenase inhibitor piroxicam. NH2Cl produced concentration-related increases in [14C] Ach release into the medium; greater than 100% over base line was observed at 0.5 mM. Glutathione inhibited the NH2Cl-evoked release in a concentration-dependent fashion. Release induced by 0.1 mM NH2Cl was abolished by 5-aminosalicylic acid and significantly inhibited by nordihydroguaiarectic acid. Piroxicam also prevented the effect of NH2Cl on release of [14C] Ach. None of these agents alone had any effect on base line [14C]Ach release. Tetrodotoxin (5 microM) did not significantly inhibit the NH2Cl-evoked transmitter release. We conclude that NH2Cl, at concentrations believed to exist in inflamed tissue, causes the release of Ach from mucosal/submucosal nerves primarily through nonspecific neural membrane injury. Endogenous prostaglandins, possibly liberated as a consequence of the injury, may be involved in the Ach release process.

Effects of SC-41930 on leukocyte adherence and emigration in rat mesenteric venules.

This study demonstrates that SC-41930, 7-[3-(4-acetyl-3-methoxy-2-propylphenoxy)-propoxy]-3,4-dihydro-8-p ropyl-2H-1-benzopyran-2-carboxylic acid, an orally active LTB4 receptor antagonist, reduces LTB4-induced leukocyte adhesion and emigration in rat mesenteric venules. The mesentery of Sprague-Dawley rats was prepared for intravital microscopic examination and venules of 25-35 microns were chosen for evaluation. In control animals, LTB4 (20 nM) was superfused over the mesentery for 30 min. In the treatment group SC-41930 (5 microM) was superfused for 30 min, followed by a 30 min superfusion with SC-41930 and LTB4. The LTB4-induced increase in leukocyte adherence and emigration in postcapillary venules was significantly attenuated by pretreatment with SC-41930. Other experiments demonstrated that platelet-activating-factor-induced leukocyte adherence was not affected by SC-41930. These results indicate that SC-41930 is a potent inhibitor of LTB4-induced leukocyte-endothelial cell adhesive interactions in postcapillary venules.

Endothelin-1 stimulates contraction and ion transport in the rat colon: different mechanisms of action.

Endothelin-like immunoreactivity has been detected in all regions of the rat gastrointestinal tract. In the present study, we studied the effect of endothelin-1 (ET-1) on muscle contraction and ion transport in the rat colon. Isometric tension was recorded in colonic muscle strips oriented along their longitudinal axis. The effect of ET-1 on ion transport was investigated by assessing changes in short-circuit current in segments of muscle-stripped rat colon in Ussing chambers. ET-1 induced concentration-dependent contraction of the colon (EC50, 3 nM). The concentration-response curve to ET-1 was not modified by the neuronal blocker tetrodotoxin (0.1 microM) or by atropine (1 microM). Pretreatment of colon muscle strips with the calcium channel blockers diltiazem (0.1 microM) or nicardipine (1 microM) had no effect on the contractile response to ET-1. Furthermore, the response was not affected by removal of extracellular calcium. In the ion transport studies, serosal addition of ET-1 produced a transient, bumetanide (chloride secretion inhibitor) -sensitive, increase in transepithelial short-circuit current. The maximal increase was 107 +/- 13 microA/sq. cm, with an EC50 of 2.5 nM. The increase in short-circuit current evoked by ET-1 was not significantly affected by 1 microM atropine, but was reduced by 50% (P less than .05) by 1 microM tetrodotoxin, or removal of extracellular calcium. We conclude that ET-1 stimulates smooth muscle directly, whereas its effect on epithelial chloride secretion is mediated in part via the enteric nerves. Moreover, the effect of ET-1 in these two systems can be differentiated on the basis of sensitivity to extracellular calcium.