Search

Background & Aims: Diarrhea is a major cause of illness and death in children and young animals. The aim of this study was to investigate the possible therapeutic effect of bromelain, a proteolytic extract obtained from pineapple stems on bacterial toxin and second-messenger agonist-induced intestinal secretion., Methods: The effect of bromelain pretreatment on short-circuit responses to Escherichia coli heat-labile enterotoxin, heat-stable enterotoxin, and Vibrio cholerae cholera toxin was evaluated in rabbit ileum mounted in Ussing chambers., Results: Bromelain was 62% effective in preventing heat-labile enterotoxin-induced secretion, 51% effective against cholera toxin, and 35% effective against heat-stable enterotoxin [corrected]. Bromelain also prevented secretory changes caused by prostaglandin E2, theophylline, calcium-ionophore A23187, 8-bromoadenosine 3':5'-cyclic monophosphate, and 8-bromoguanosine 3':5'-cyclic monophosphate, well-known intracellular mediators of ion secretion. The efficacy of bromelain was not caused by reduced tissue viability resulting from its proteolytic effects on enterocytes, indicated by experiments measuring uptakes of nutrients into intestinal cells and experiments measuring short-circuit responses to glucose., Conclusions: Bromelain prevents intestinal fluid secretion mediated by secretagogues that act via adenosine 3':5'-cyclic monophosphate, guanosine 3':5'-cyclic monophosphate, and calcium-dependent signaling cascades. It may be clinically useful as an antidiarrheal drug.

Bicarbonate secretion by duodenal mucosa just distal to the Brunner's glands area and devoid of pancreatic secretions was titrated in situ in anesthetized rats. Intravenous injection of diazepam (0.1 and 0.5 mg/kg) significantly increased the secretion; this stimulation was abolished by proximal bilateral vagotomy. Ro 15-1788, a benzodiazepine antagonist that also has well-known intrinsic activity, caused similar stimulation of the secretion when administered IV (0.01 and 0.1 mg/kg). Intracerebroventricular infusion of Ro 15-1788 (10 micrograms/h) resulted in a greater increase in secretion; again, this stimulation was prevented by vagotomy. Adrenoceptor blockade by phentolamine increased basal alkaline secretion but did not affect the stimulation by diazepam. The tricyclic antidepressant trimipramine (2.5 mg/kg IV) did not affect the duodenal bicarbonate secretion. For comparison, effects of diazepam and Ro 15-1788 (10(-6)-10(-4) mol/L) were also tested in isolated bullfrog duodenal mucosa. Neither drug effected the alkaline secretion in vitro. The combined results strongly suggest that benzodiazepines, as previously shown for certain brain peptides, influence the central nervous control of duodenal mucosal alkaline secretion and that their stimulation of secretion is vagally mediated. This action benzodiazepines might be used in modulating mucosal protection against acid.

Ciguatera fish poisoning is a clinical syndrome consisting of a combination of gastrointestinal and neurological symptoms occurring after eating toxin-containing tropical reef fish; it is a major cause of morbidity in Hawaii, the South Pacific, Australia, and the Caribbean. In an effort to define pathophysiological mechanisms responsible for the diarrheal component of the illness, we examined the effect of crude and fractionated toxin preparations on isolated rabbit ileal tissue in a Ussing chamber model. Both the crude toxin preparation (prepared from toxic Ctenochaetus strigosus) and 10% and 50% methanol-chloroform toxin fraction (prepared from a pool of toxic fish samples) gave a striking increase in transepithelial electrical potential difference and short-circuit current. Enterotoxic activity seemed to be mediated by calcium. When examined by light microscopy, the intestinal mucosa was not damaged by the toxin preparations used. Our data demonstrate that toxins involved in ciguatera fish poisoning directly stimulate intestinal fluid secretion without accompanying tissue damage and suggest that calcium is the "second messenger" mediating the process.

To determine the effects of luminal protease inhibition on duodenal delivery and the intraluminal fate of pancreatic enzymes, six healthy subjects were intubated with an oro-ileal multilumen tube assembly. By using nonabsorbable markers, cumulative trypsin, chymotrypsin, lipase, and amylase activities were measured as delivered to duodenum, midjejunum, and distal ileum, with or without simultaneous duodenal perfusion of the protease inhibitor camostat at graded doses. Compared with saline, camostat (a) inhibited trypsin activity in the entire small intestinal lumen by up to 99%, and significantly reduced chymotrypsin activity by up to 89%; (b) significantly increased duodenal deliveries of lipase activity, amylase activity and volume; (c) did not influence plasma cholecystokinin concentrations; and (d) significantly increased jejunal and ileal deliveries of lipase but not amylase activity. Small intestinal transit and motility were not affected by camostat. In additional in vitro studies, camostat significantly reduced the spontaneous decline in lipase activity in fresh human duodenal juice incubated at 37 degrees C. These findings demonstrate that duodenal deliveries of lipase and amylase activities increase when intraluminal protease activity is decreased; they suggest that this increase is not caused by slower proteolytic destruction of enzyme protein but by stimulation of pancreatic secretion. Thus, luminal protease-mediated feedback regulation of pancreatic secretion may be operative in humans. Because plasma cholecystokinin concentrations were not affected, these effects may in part be independent of cholecystokinin. The data further suggest that proteolytic digestion plays a major role in the rapid loss of luminal lipase activity on small intestinal transit.

Although clinical diarrhea is often caused by more than one enteropathogen, it is not known whether secretagogues have additive effects on secretion. This question was examined in anesthetized, ventilated rats by exposing the ileum to secretory agents with different mechanisms of action. Four hours after inoculation of ileal loops with either cholera toxin or saline, transport was measured during perfusion with Ringer's solution, with Ringer's solution containing Escherichia coli heat-stable enterotoxin or mannitol, or with Ringer's solution containing both heat-stable enterotoxin and mannitol. We found that heat-stable enterotoxin caused similar decrements in water absorption in loops exposed to Ringer's solution, hypertonic mannitol, cholera toxin, and cholera toxin plus mannitol. By contrast, the decrement in water transport caused by hypertonic mannitol was inversely related to the level of ongoing water transport. In addition, the electrolyte transport changes caused by each enterotoxin were preserved despite the presence of other secretory agents. These data suggest that the transport effects of secretagogues with different mechanisms of action are additive in the ileum. However, the absolute decrements in water absorption may be independent of (heat-stable enterotoxin and cholera toxin) or dependent on (hypertonic mannitol) the ongoing level of water transport.

We studied the effect of intraileal administration of a 100-ml test meal, 100 ml of 5% glucose, amino acids, 10% Intralipos, or bile on canine plasma immunoreactive vasoactive intestinal polypeptide (VIP) levels in the mesenteric and femoral veins of dogs. The effects of intravenous administration of caerulein on plasma immunoreactive VIP levels were examined, with or without intraduodenal ballooning, to rule out the effect of bile and pancreatic and gastric juice. In addition, hexamethonium bromide or atropine sulfate was given intravenously before the administration of bile or intravenous caerulein to study the role of the neural pathway on the release of VIP. Basal plasma VIP levels in the mesenteric vein were significantly higher than those in the femoral vein. Intraileal administration of the test meal, 5% glucose, amino acids, or 10% Intralipos induced no variation of mesenteric VIP. Intraileal administration of bile or intravenous caerulein, with or without intraduodenal ballooning, produced a significant increase of plasma VIP in the mesenteric vein. Pretreatment with atropine completely blocked the bile-stimulated VIP release and significantly inhibited the caerulein-stimulated release of VIP. On the other hand, pretreatment with hexamethonium did not significantly decrease the VIP response by bile or caerulein. Thus, the muscarinic receptor probably plays an important role in bile- or caerulein-stimulated release of VIP.

We studied the influence of morphine on intestinal secretion induced by three different secretogogues in an attempt to elucidate the mechanism for its antisecretory activity. In rabbit ileal mucosa in vitro morphine (10(-4) M) did not influence the electrical response to the subsequent administration of prostaglandin E2 or of acetylcholine but did inhibit the net secretion of chloride provoked by both agents and prevented the reduction of sodium adsorption induced by the prostaglandin. Morphine reduced the peak electrical response to cholera toxin and reduced toxin-induced net chloride secretion. The modes of action of morphine appeared to be by both enhancing adsorption, as it does in control, nonsecreting mucosa, and by inhibiting secretogogue activity.

Methylxanthines produce intracellular accumulation of cyclic 3'5'-AMP (cAMP) by inhibition of phosphodiesterase and mucosal cAMP accumulation. Cyclic AMP is thought to mediate small intestinal secretion caused by some enterotoxins, hormones, and methylxanthines. These studies were designed to evaluate the effect of caffeine on small intestinal net fluid movement and transit times. The administration of caffeine in amounts ordinarily contained in many beverages and medications (75 to 300 mg) resulted in striking net secretion in the jejunum which lasted at least 15 minutes. This occurred in six of seven studies. Baseline net absorption of 0.5 ml per cm per hr was reversed to net secretion of 6.0 +/- 2.2 ml per cm per hr after oral caffeine ingestion (P less than 0.01). Net secretion also occurred in the ileum in seven of eight studies, but the onset of secretion appeared 35 min later than in the jejunum. These patterns of secretion correlated best with the passage of the intestinal bolus of caffeine rather than plasma caffeine levels. In contrast to other net secretory conditions, which increase the speed of transit, small intestinal transit times, as determined by dye dilution methods, were unchanged by caffeine. It is possible that methylxanthine-induced small intestinal secretion may play a role in the symptoms experienced by some patients with functional diarrhea.

In dogs with gastric fistulas and vagally denervated Heidenhain pouches, omeprazole, a benzimidazole derivative infused intravenously or given intraduodenally, dose-dependently inhibited gastric acid secretion, which had been induced by histamine, pentagastrin, or urecholine. It also suppressed gastric acid response to physiologic stimulants such as sham-feeding and gastric peptone meal without affecting serum gastrin level. The inhibition of histamine-induced acid secretion was accompanied by a parallel reduction in the mucosal blood flow, but no significant alteration in the ratio (R) value, indicating that omeprazole primarily affected gastric acid secretion but did not limit gastric mucosal microcirculation. Omeprazole, infused into the Heidenhain pouch, caused a dose-dependent inhibition of the Heidenhain pouch response to intravenous histamine without any significant change in the acid response of the main stomach and plasma concentrations of the drug. This indicates that omeprazole may exhibit local inhibitory action on the oxyntic glands. Omeprazole did not affect gastric mucosal integrity or the rate of alkaline secretion from the gastroduodenal mucosa or the pancreas stimulated by duodenal acidification or secretin.

The effect of somatostatin on colonic secretion induced by 10(-8) M vasoactive intestinal peptide (VIP), 10(-2) M theophylline, and 2 X 10(-3) M dibutyryl cyclic AMP was studied in muscle-stripped everted open rat colon sacs. The secretory response to VIP, measured as the decrease in net absorptive flow rate (microliters 30 min-1 mg-1 of dry weight), was maximal and equalled the responses to theophylline or dibutyryl cyclic AMP. Somatostatin (10(-5) M) blocked completely the secretory response to VIP but only partially the secretory response to theophylline or dibutyryl cyclic AMP. This difference in the extent of inhibition suggested that somatostatin exerted an inhibitory effect both before and after the point of generation of intracellular cyclic AMP. In order to test the hypothesis that one component of the action of somatostatin involved inhibition of the production of cyclic AMP, measurements of this nucleotide were made in isolated rat colon cells. Control levels of cyclic AMP measured by radioimmunoassay (12.6 +/- 1.6 pmoles per 10(6) cells) were not affected by 10(-5) M somatostatin. VIP (5 X 10(-8) M) increased cyclic AMP levels 2-fold (P less than 0.01) and this increase was blocked by somatostatin. The results indicated that somatostatin inhibits colonic secretion by exerting effects at two sites: one site lies at, and another beyond, the point of generation of intracellular cyclic AMP.

Perfusion studies were performed in 35 healthy volunteers to investigate further the secretory effect of prostaglandin E1 (PGE1), administered intraluminally, on the human jejunum. A perfusion system with a proximal occluding balloon and continuous aspiration of duodenal secretions was used. The influence of PGE1 (0.9 mug per kg per min) on glucose, fluid, and ion transport was entirely reproducible throughout a 6-hr perfusion experiment and was fully reversible 60 min after the end of PGE1 administration. The influence of ethacrynic acid (EA), aspirin (two agents previously reported to inhibit choleraic secretion), and furosemide was studied both on basal jejunal absorption and on PGE1-induced secretion of water and electrolytes. EA (2.5 mg per kg), administered in jejunal lumen, significantly reduced (P less than 0.001) the net secretory effect of intraluminal PGE1, and suppressed the PGE1-induced increase in plasma-to-lumen unidirectional flux of sodium. Intravenous or intraluminal aspirin (25 to 40 mg per kg) as well as intraluminal furosemide (1.5 mg per kg) did not modify the PGE1-induced secretion nor the basal absorption rates. Plasma immunoreactive levels of PGE1, calcitonin, and diverse gut hormones did not change significantly during PGE1-induced secretion. These results indicate that (1) aspirin, an inhibitor of prostaglandin synthesis, has no influence on the intestinal secretory effect of preformed PGE1; (2) EA, unlike furosemide, inhibits in man PGE1-induced jejunal secretion, an effect of EA similar to that observed in animals on cholera toxin- and cyclic AMP-mediated secretion.

Stimulation of net secretion by deoxycholic acid (DCA) in the colon and by cholera enterotoxin (CE) in the jejunum is mediated by cAMP. Propranolol (Pr) inhibits adenylate cyclase (AC) activity and net secretion induced by bile acid in the colon. The aim of this study was to assess the organ specificity of DCA and CE as well as the selectivity of Pr inhibition. Three colonic and three jejunal loops were prepared in each of 8 rabbits treated intravenously with Pr, 4 mg per kg, 1/2 hr before loop construction and in each of 10 untreated control rabbits. One milliliter of DCA, 6 mM, CE, 10 mug per ml, or heat-inactivated CE or 0.9% NaCl, as basal controls were injected in random order into each of the loops. The volume of luminal fluid and mucosal AC were measured in each intestinal loop 5 hr later. DCA in the colon stimulated AC 2-fold (P less than 0.01) and luminal fluid 15-fold (P less than 0.01). CE in the jejunum stimulated AC 2.3-fold (P less than 0.01) and luminal fluid 9-fold (P less that 0.01). No significant effects on volume or AC occurred in response to CE in the colon or to DCA in the jejunum. Pr pretreatment completely prevented the stimulation of AC and luminal fluid by DCA in the colon but did not affect the action of CE in the jejunum of the same animals. Thus, DCA and CE are organ-specific stimulants of cAMP systems, and Pr is a selective inhibitor of certain inducers of cAMP and net secretion.

Bile acids cause diarrhea by inducing colonic secretion, probably mediated through the cyclic AMP system. The aim was to determine the effects of an adenylate cyclase inhibitor, propranolol, on deoxycholic acid (DCA) stimulation of net secretion and the cyclic AMP system in the colon. In each of 30 New Zealand white rabbits, 0.9% NaC1 as control and 6 mM and 8 mM DCA were injected in random sequence into three colonic loops in situ. Propranolol, 4 mg per kg was administered intravenously to 12 of the 30 rabbits 1/2 hr before preparation of the loops, i.e., 5 1/2 hr before the rabbits were killed. In the 18 untreated animals, 6 and 8 mM DCA significantly stimulated colonic net secretion and mucosal adenylate cyclase activity; 6 mM DCA caused no change in mucosal phosphodiesterase activity, whereas 8 mM DCA caused a 25% decrease (P less than 0.01). In propranolol-treated animals compared to untreated animals, the volume of luminal fluid in controls was not different, with 6 mM DCA it was 88% less (P less than 0.01), and with 8 mM DCA it was 45% less (P less than 0.01); adenylate cyclase activity in controls was 43% less (P less than 0.01), with 6 mM DCA it was 67% less (P less than 0.01), and with 8 mM DCA it was 65% less (P less than 0.01); phosphodiesterase activity in controls and with 6 mM DCA was not different and with 8 mM DCA it was 38% greater (P less than 0.02). In conclusion, propranolol prevented DCA stimulation of colonic net secretion and inhibited the cyclic AMP system. Propranolol, therefore, warrants investigation as therapy for diarrhea caused by bile acids in the colon.

Previous studies have shown that synthetic salmon calcitonin (SCT) infused intravenously causes secretion of water and electrolytes in the jejunum of normal human subjects. The present experiments were carried out to learn more about the nature of this intestinal secretory process. During SCT- or synthetic human calcitonin (HCT)-induced intestinal secretion, the following observations were made: (1) There was no change in potential difference; (2) Cl was secreted against an electrochemical gradient; (3) unidirectional Na flux out of the lumen was decreased while the opposite flux was normal; (4) luminal pCO2 fell; (5) addition of glucose to the jejunal contents stimulated Na abdsorption, and this in turn counteracted the secretory effect of calcitonin. These findings suggest that calcitonin induces active Cl secretion and inhibits active Na absorption, and that HCO3 absorption is reduced by virtue of OH secretion; furthermore, jejunal glucose absorption and glucose-stimulated Na absorption are intact during calcitonin-induced secretion. Intravenous infusion of HCT caused intestinal secretion only when blood levels were much higher than occur physiologically; therefore, calcitonin is probably not a mediator of spontaneous variations of intestinal transport in normal people. However, because calcitonin induces secretion in the ileum as well as in the jejunum, hypercalcitonemia (within the range commonly found in patients with medullary carcinoma of the thyroid) could be a cause of severe secretory diarrhea.

Using a triple-lumen gut perfusion technique, the net flux of water and electrolytes in the jejunum of 5 healthy volunteers was measured before, during and after an intravenous infusion of gastric inhibitory polypeptide (GIP). The dose of GIP used (1 mug/min for 30 mins) has previously been shown to raise serum GIP levels to physiological levels achieved in normal subjects after a meal. During GIP infusion, net water Na, K, and HCO3 absorption was significantly reduced and chloride flux was switched from absorption to secretion when compared to pre- and post-GIP control periods (p less than 0.001).

5-Hydroxytryptamine (5-HT) has been claimed to mediate intestinal secretion in morphine withdrawal diarrhea through stimulation of local prostaglandin formation without involving cyclic adenosine monophosphate. Therefore, experiments were performed to study (a) the effects of exogenous 5-HT and the cyclic adenosine monophosphate-dependent secretagogue, vasoactive intestinal polypeptide, on intestinal prostaglandin E2 (PGE2) formation and (b) the involvement of calcium in the secretory response to close intraarterial infusion of 5-HT, PGE2, or vasoactive intestinal polypeptide in tied-off loops of rat jejunum in vivo. 5-Hydroxytryptamine and vasoactive intestinal polypeptide reversed fluid absorption to net secretion (p less than 0.01), but only 5-HT caused an increase in luminal PGE2 output (p less than 0.01). Indomethacin and d,l-verapamil prevented only the secretory effect of 5-HT. Exogenous PGE2 (1.6-160 ng/min) reversed absorption to secretion (p less than 0.01) in a dose-dependent manner, irrespective of whether the rats were pretreated with indomethacin or not. Racemic and l-verapamil, but not d-verapamil, markedly reduced (p less than 0.01) the secretory effect of physiologically low doses of PGE2 (1.6 and 16 ng/min), whereas high doses of PGE2 (160 ng/min), which caused a significant increase in mucosal cyclic adenosine monophosphate (p less than 0.005), were not inhibited by verapamil. These data suggest that PGE2 may be an important intermediate in the transduction mechanism leading to 5-HT-induced intestinal secretion, and that physiologic doses of PGE2 may act by facilitating calcium entry, rather than by increasing intracellular calcium through activation of the adenylate cyclase.

The role of microtubules in small intestinal electrolyte secretion was investigated in rats by testing the effects of colchicine and vinblastine, both of which inhibit microtubule assembly. In the intact rat, intraperitoneal injection of colchicine (5 mg/kg) inhibited cholera toxin and prostaglandin-stimulated secretion without inhibiting their effects on adenylate cyclase or cAMP concentration. Pretreatment with colchicine had no effect on fluid transport in the absence of secretory stimuli. When added to rat ileum in vitro, colchicine reduced by 60% the short-circuit current (Isc) response to dibutyryl cAMP added 4 h later, whereas its structural isomer, lumicolchicine, which does not inhibit microtubule assembly, was ineffective. Vinblastine reduced by 55% the Isc response to dibutyryl cAMP and theophylline added 2 h later. Two hour pretreatment with vinblastine also reduced by 40% the Isc response to the cholinergic agonist, carbamylcholine, a Ca-dependent secretory stimulus which does not increase cAMP concentration. In contrast to their antisecretory actions, neither colchicine nor vinblastine inhibited glucose-stimulated active Na absorption. These results suggest a role for microtubules in active electrolyte secretion in the small intestine.

Cholera enterotoxin (CT) produces intestinal secretion associated with an elevation of intestinal cyclic AMP (cAMP). Indomethacin, a potent inhibitor of prostaglandin (PG) synthesis, decreases CT-induced secretion, although a role for PG in this process has not been demonstrated. The purpose of this study was to measure the effects of indomethacin on net fluid movement and unidirectional Na fluxes in rabbit jejunal loops exposed to CT and to correlate these findings with intestinal cAMP levels. In untreated animals (no indomethacin), CT loops secreted 0.37 ml per cm per 4 hr compared to absorption in control loops of 0.23 ml per cm per 4 hr (P less than 0.001). In indomethacin-treated animals, there was a striking reduction of secretion in CT loops (0.07 ml per cm per 4 hr, P less than 0.001). Absorption in control loops in indomethacin animals was greater than in untreated animals (0.42 ml per cm per 4 hr, P less than 0.02). Unidirectional Na fluxes were greatly depressed in indomethacin animals at 1 h in both CT and control loops. This effect disappeared by 4 hr. Intestinal cAMP levels in CT loops, although not elevated at 1 hr despite the onset of secretion, were significantly elevated at 4 hr. Indomethacin did not alter cAMP levels at 1 and 4 hr in either cholera or control loops. These studies support the view that PG synthesis is not involved in CT-induced elevation of intestinal cAMP. Indomethacin may depress intestinal secretion by inhibiting a PG-mediated step beyond the generation of cAMP or by acting on some other, as yet unidentified, biological mechanism involved in intestinal secretion.

The purpose of this study was to compare the actions of shigella toxin and cholera toxin to determine if mucosal damage is a prerequisite for shigella toxin-induced secretion. The secretory response to maximal doses of cholera toxin and shigella toxin were compared. The maximal rate of secretion and the electrolyte and protein concentration of the intestinal fluid were similar for both toxins. On the other hand, the time of onset after exposure to the toxin was 105 min for shigella toxin and 15 to 30 min for cholera toxin. In addition, cholera toxin-induced secretion was associated with depletion of goblet cell mucus, whereas no change was seen in association with the response to shigella toxin. Other than goblet cell depletion, there were no histological differences between loops secreting in response to cholera toxin and to shigella toxin. Finally, the secretory effect of the toxins are not additive. These studies suggest that, in spite of apparent differences in the patterns of secretory response to the two toxins, they may share a rate-limiting step in the secretory process.

Previous in vitro experiments suggest that protonated 2,4,6-triaminopyrimidine (TAP+) inhibits passive Na+ movement across tight junctions of various epithelial tissues. So far no evidence has been found that TAP+ interferes with other mucosal transport processes. Because blockage of the tight junctions would be a promising tool in studying intestinal transport physiology, the effect of TAP+ was investigated in the dog ileum in vivo. When TAP+ was added to a sodium-free mannitol solution, the transepithelial sodium diffusion potential was significantly decreased (60% inhibition with 34 mm TAP+); this would be expected if TAP+ inhibited NA+ permeation through tight junctions. However, TAP+ was also found to diminish Na+ and Cl- absorption. Furthermore, TAP+ increased unidirectional Na+ flux from plasma to lumen; this is opposite to the expected result of tight junction blockage. In addition, TAP+ reduced glucose, fructose, and xylose absorption by about 50%. In ileal loops exposed to cholera toxin, TAP+ enhanced secretion by a rate equal to the rate by which it reduced absorption in loops not exposed to cholera toxin. All changes induced by TAP+ approached normal within 4 hr of its removal from the perfusate. TAP+ did not cause nay mucosal damage that could be detected by protein leakage or by light microscopy. These studies show that TAP+ has many effects on intestinal transport processes that cannot be explained on the basis of tight junction blockage.