Localization of transforming growth factor alpha and its receptor in gastric mucosal cells. Implications for a regulatory role in acid secretion and mucosal renewal.

Transforming growth factor alpha (TGF alpha) shares with epidermal growth factor (EGF) structural homology (35%), a common cell-surface membrane receptor (TGF alpha/EGF receptor), and a nearly identical spectrum of biological activity, including inhibition of gastric acid secretion. Herein, we report expression of TGF alpha mRNA in normal gastric mucosa of the adult guinea pig, rat, and dog. TGF alpha mRNA was also detected in matched surgically resected gastric mucosa and adjacent gastric carcinoma from 10 patients, and in gastric mucosa adjacent to a benign ulcer from an additional patient. TGF alpha protein was quantitated by radioimmunoassay and was present in tumor and adjacent mucosa. TGF alpha/EGF receptor mRNA was also detected in gastric mucosa from all species studied. Localization of TGF alpha and TGF alpha/EGF receptor mRNA expression was examined in samples of unfractionated guinea pig gastric mucosa and from chief cell-enriched and parietal cell-enriched fractions. All samples exhibited TGF alpha and TGF alpha/EGF receptor expression. The TGF alpha signal was greatest in the parietal cell fraction (5.8-fold increase), but was also enhanced in the chief cell fraction (1.9-fold increase) relative to the unfractionated gastric mucosa. Like TGF alpha expression, TGF alpha/EGF receptor mRNA expression was most intense in the parietal cell-enriched fraction (7.8-fold increase), but was also increased in the chief cell-enriched fraction (2.7-fold increase) relative to the unfractionated guinea pig gastric mucosa. We conclude that TGF alpha and TGF alpha/EGF receptor genes are expressed in normal adult mammalian gastric mucosa. These findings, when interpreted in light of described actions of TGF alpha and EGF, provide evidence that local production of TGF alpha could play an important role in the regulation of acid secretion and mucosal renewal in the stomach.

Binding of 125I-CCK-8 and 125I-gastrin-I to dispersed chief cells from guinea-pig stomach.

In dispersed gastric chief cells from guinea-pig stomach, binding of iodinated cholecystokinin octapeptide (125I-CCK-8) was relatively slow, temperature-dependent, to a single class of binding sites and inhibited by various gastrin- and CCK-related agonists and receptor antagonists. Binding of iodinated gastrin-I (125I-gastrin-I) was moderately rapid, temperature-dependent, to a single class of binding sites, and inhibited by various gastrin and CCK-related agonists and receptor antagonists. Gastrin-I as well as C-terminal fragments of CCK containing from eight amino acids (CCK-8) to four amino acids (CCK-4) stimulated pepsinogen secretion and inhibited binding of 125I-CCK-8 and 125I-gastrin-I. In addition, each of five different receptor antagonists inhibited binding of 125I-CCK-8 and 125I-gastrin-I and inhibited pepsinogen secretion stimulated by CCK-8 or gastrin-I. With each of eight different agonists and with each of five different antagonists the value of IC50 for inhibition of binding of 125I-CCK-8 was not significantly different from the value of IC50 for inhibition of binding of 125I-gastrin-I, indicating that in gastric chief cells the sites to which 125I-CCK-8 binds are the same sites to which 125I-gastrin-I binds. With the agonists as well as with the antagonists, however, there was no consistent relationship between the ability of a particular agent to inhibit binding and its ability to modify pepsinogen secretion, indicating that in gastric chief cells the sites that bind 125I-CCK-8 and 125I-gastrin-I are not the receptors that mediate stimulation of pepsinogen secretion by CCK-8 or by gastrin-I.

Comparison of the effectiveness of ranitidine and cimetidine in inhibiting acid secretion in patients with gastric hypersecretory states.

The H2-histamine receptor antagonists ranitidine and cimetidine were compared for their abilities to control gastric acid hypersecretion on a short- and long-term basis in 22 patients with gastric acid hypersecretory states. Nineteen patients had Zollinger-Ellison syndrome, one patient had systemic mastocytosis, and two patients had idiopathic hypersecretion. The rates of onset of the action of cimetidine and ranitidine were the same. The actions of both drugs were increased by anticholinergic agents, and there was a close correlation between the daily maintenance dose of each drug needed to control acid secretion. However, ranitidine was threefold more potent than cimetidine both in acute inhibition studies and in the median maintenance dose needed (1.2 g per day for ranitidine and 3.6 g per day for cimetidine). Sixty percent of the males developed breast changes or impotence while taking cimetidine and in all cases these changes disappeared when cimetidine was replaced by ranitidine. Treatment with high doses of cimetidine (one to 60 months; median, 11 months) or ranitidine (two to 31 months; median, 14 months) was not associated with hepatic or hematologic toxicity or alterations of serum gastrin concentrations, but ranitidine therapy was associated with a significantly lower serum creatinine level than seen with cimetidine therapy. The results show that both drugs can adequately inhibit acid secretion in patients with gastric hypersecretory states. Both are safe at high doses, but ranitidine is threefold more potent and does not cause the antiandrogen side effects frequently seen with high doses of cimetidine.

Atrial natriuretic factor activates membrane-bound guanylate cyclase of chief cells.

The present study examined the effect of atrial natriuretic factor (ANF) on cGMP generation by dispersed chief cells from guinea pig stomach. ANF caused a rapid dose-dependent increase in cGMP, a 7-fold increase in cGMP caused by 1 microM ANF, with or without 3-isobutyl-1-methylxanthine present. Methylene blue reduced cGMP in response to nitroprusside but not ANF. Guanylate cyclase activity of a chief cell membrane fraction doubled in response to ANF, but was not affected by nitroprusside. ANF had no effect on guanylate cyclase activity of the soluble fraction of lysed chief cells. Dose-response curves for whole cell cGMP production and membrane guanylate cyclase activity in response to ANF were closely related. These data indicate that ANF increases chief cell cGMP production by activating particulate guanylate cyclase, providing functional evidence that chief cells possess surface membrane receptors for ANF.

Benefit of resection of metastatic gastrinoma in multiple endocrine neoplasia type I.

Although gastrinoma resection is generally advocated for patients with the sporadic form of nonmetastatic Zollinger-Ellison syndrome, there is controversy regarding the surgical management of the gastrinoma among patients with multiple endocrine neoplasia type I (MEN-I). Using strict criteria, to date no biochemical cures of the Zollinger-Ellison syndrome lasting greater than 5 months have been achieved by gastrinoma resection among patients with MEN-I. Whereas resections of hepatic metastases have been performed in patients with sporadic gastrinoma, none have been reported among patients with MEN-I. The current report describes a patient with MEN-I, closely followed up for 30 years, in whom enlargement of pancreatic gastrinoma and development of hepatic gastrinoma was observed to occur over 3 years. After preoperative localization, an 80% pancreatectomy and a left lateral segmentectomy of the liver were performed. Sixteen months after the operation, secretin and calcium provocative testing showed that the patient's fasting gastrin and stimulated plasma gastrin concentrations were normal; also, results of computerized tomographic angiography, selective abdominal angiography, and hepatic venous sampling for gastrin after intra-arterial secretin injection were negative for gastrinoma. By achieving a 16-month cure of gastrinoma, this case shows that an aggressive surgical approach can benefit certain patients with gastrinoma who have MEN-I even in the presence of hepatic metastases.

Cholecystokinin-induced desensitization of pepsinogen secretion from chief cells.

When dispersed chief cells from guinea pig stomach were first incubated with cholecystokinin (CCK), washed, and then reincubated with CCK in fresh incubation solution, the stimulation of pepsinogen secretion and the rise in intracellular calcium concentration during the second incubation were reduced. CCK did not cause residual enzyme secretion but caused desensitization that was rapid, temperature dependent, dependent on extracellular calcium, reversible with time, and prevented but not reversed by CCK receptor antagonists. Cholecystokinin octapeptide (CCK-8) caused desensitization over the same range of concentrations that stimulate pepsinogen secretion, whereas the concentration of gastrin required to cause maximal desensitization was greater than that required to cause maximal stimulation of enzyme secretion. CCK-8 caused heterologous desensitization of pepsinogen secretion stimulated by agonists that interact with receptors to cause mobilization of cellular calcium and activation of protein kinase C or by agonists that bypass receptors to activate these mediators directly; however, CCK-8 did not induce desensitization of the stimulation caused by any secretagogue whose actions are mediated by adenosine 3',5'-cyclic monophosphate. Because CCK-8 caused greater desensitization of secretion stimulated by agonists that interact with receptors than by agonists that bypass receptors, it is likely that receptor modulation as well as a postreceptor action contribute to the ability of CCK to cause desensitization of pepsinogen secretion from chief cells.

Cholinergic desensitization of pepsinogen secretion and calcium mobilization of dispersed guinea pig chief cells.

When dispersed chief cells from guinea pig stomach were first incubated with carbachol, washed, and then reincubated with carbachol in fresh incubation solution, the stimulation of pepsinogen secretion and the rise in intracellular calcium concentration during the second incubation were reduced. Carbachol did not cause residual enzyme secretion, but the same range of concentrations that causes enzyme secretion caused desensitization that was rapid, temperature dependent, and reversible with time. Preincubation with carbachol caused approximately a 65% reduction in enzyme secretion stimulated during a subsequent incubation with this agonist, but the potency of carbachol was unaffected. Prior exposure to carbachol also reduced subsequent stimulation caused by cholecystokinin (CCK-8), gastrin I, ionophore A23187, or 12-O-tetradecanoylphorbol 13-acetate but did not alter stimulation by any agonist that increases cellular cAMP. Carbachol pretreatment of Fura-loaded chief cells caused a threefold increase in the EC50 for carbachol-stimulated [Ca2+]i and approximately a 30% reduction in the maximal rise in [Ca2+]i in response to carbachol or CCK-8. Inhibition of [N-methyl-3H] scopolamine binding by carbachol following carbachol pretreatment indicated that modulation of receptor affinity or number did not account for functional desensitization. These data indicate that carbachol causes heterologous desensitization of pepsinogen secretion stimulated by agonists that mobilize cellular Ca2+ or activate protein kinase C through a postreceptor action and suggest that an attenuated rise in chief cell calcium is one mechanism mediating the desensitization of enzyme secretion.

Comparison of nizatidine and cimetidine as once-nightly treatment of acute duodenal ulcer. Nizatidine Multicenter Duodenal Ulcer Study Group.

Nizatidine, a new H2-receptor antagonist for the treatment of duodenal ulcer disease, was compared with cimetidine in an 8-wk, randomized, double-blind, multicenter clinical trial. Patients were randomly allocated to receive either nizatidine 300 mg h.s. or cimetidine 800 mg h.s. Patients were treated for 8 wk, regardless of the healing status of their ulcers. An endoscopy was performed at Wk 2, 4, and 8. Healing rates with nizatidine 300 mg h.s. were numerically, but not statistically significantly, superior to those with cimetidine 800 mg h.s. at each treatment period. Ulcer healing rates at Wk 2, 4, and 8 were 41% (78/191), 73% (130/179), and 81% (145/179) for nizatidine and 33% (60/184), 67% (116/174), and 75% (126/168) for cimetidine, respectively. Symptoms of peptic ulcer disease were similarly reduced at each treatment period by nizatidine and cimetidine. Patients with healed ulcers at either Wk 2 or Wk 4 were continued on therapy and an endoscopy was performed at Wk 8. Ulcer recurrence occurred in 10% of nizatidine-treated and 19% of cimetidine-treated patients at Wk 8 (p = 0.085). The observation of recurrence of duodenal ulcer while patients were receiving full-dose H2-receptor antagonist therapy has not been reported previously.

Ability of prostaglandin to reduce ethanol injury to dispersed chief cells from guinea pig stomach.

To determine whether prostaglandin exerts a direct action on individual gastric epithelial cells that protects them from ethanol-induced injury, dispersed chief cells from guinea pig stomach were pretreated with 16,16-dimethyl-prostaglandin E2 (dmPGE2) or placebo before incubation with ethanol or control. Cell injury was assessed in terms of exclusion of Fast Green dye, release of lactate dehydrogenase, alterations of ultrastructure, and pepsinogen secretion stimulated by a variety of secretagogues. Of chief cells 60 +/- 2% were stained by Fast Green if incubated with 10% ethanol for 1 h after pretreatment with placebo, whereas only 38 +/- 1% of cells showed Fast Green staining when pretreated with 2.6 microM dmPGE2 before ethanol exposure. Similarly, 63 +/- 2% of cellular lactate dehydrogenase was released from chief cells pretreated with placebo compared with 36 +/- 4% of lactate dehydrogenase released from cells pretreated with 2.6 microM dmPGE2 (P less than 0.01). The prostaglandin's protective effect persisted throughout a 6-h incubation with ethanol. Scanning and transmission electron micrographs demonstrated disintegration of chief cells pretreated with placebo before ethanol exposure, whereas ultrastructural architecture was relatively preserved among chief cells pretreated with dmPGE2. Preincubation with 8 or 10% ethanol inhibited the subsequent stimulation of pepsinogen secretion caused by carbachol, cholecystokinin, A23187, 12-O-tetradecanoylphorbol 13-acetate, forskolin, or 8-bromoadenosine 3',5'-cyclic monophosphate. Pretreatment with dmPGE2 did not reduce the ethanol-induced inhibition of secretion stimulated by any of these secretagogues. These data indicate that dmPGE2 significantly reduces ethanol-induced damage to dispersed chief cells in terms of alterations of membrane permeability and ultrastructure but does not prevent the ethanol-induced impairment of pepsinogen secretion. These findings provide evidence that dmPGE2 exerts a direct but limited protective action on the gastric chief cell, independent of vascular, paracrine, or neural actions.

Gastrointestinal dysfunction in systemic mastocytosis. A prospective study.

In 16 consecutive patients with systemic mastocytosis, we prospectively evaluated a variety of gastrointestinal functions and examined how they relate to the occurrence of gastrointestinal symptoms. Nine patients had either a duodenal ulcer or duodenitis. Hypersecretion of gastric acid was present in 6 patients, and in these patients the mean basal acid output was 20.7 +/- 4.1 mEq/h (range 14-39 mEq/h). Impaired small intestinal absorption occurred in 5 patients, although this was usually mild. The mean fractional emptying rate of liquids for all patients (14.7% +/- 2.3% per minute) did not differ from that for controls (10.7% +/- 0.6% per minute). Mean mouth-to-cecum transit time measured by breath hydrogen testing was the same among patients (87.7 +/- 6.7 min) and controls (86.7 +/- 8.0 min). Plasma histamine concentrations were increased in all patients (mean 1886 pg/ml, range 480-7450) and correlated with the basal acid output (r = 0.64, p less than 0.02) but not maximal acid output or the presence or absence of pain or diarrhea. Mean fasting plasma concentrations of motilin, substance P, and neurotensin from 6 patients did not differ significantly from controls, whereas gastrin and vasoactive intestinal peptide were significantly less than in controls (p less than 0.01). Gastrointestinal symptoms, consisting of abdominal pain or diarrhea, occurred in 80% of patients. Abdominal pain classified as dyspeptic was usually associated with acid-peptic disease of the duodenum and hypersecretion of gastric acid, whereas abdominal pain of a nondyspeptic character was not. Only in those cases of diarrhea consisting of greater than 200 g stool/day was gastric acid hypersecretion frequently found. Neither fecal urgency nor nondyspeptic pain could be accounted for by alterations of gastrointestinal transit. These results demonstrate that gastrointestinal symptoms, peptic disease, and mild malabsorption are much more common than described previously in patients with systemic mastocytosis. Furthermore, the results provide no evidence for the contention that altered gastrointestinal transit is involved in the pathogenesis of these symptoms.

Functionally distinct receptors for cholecystokinin and gastrin on dispersed chief cells from guinea pig stomach.

Caerulein, gastrin, and C-terminal fragments of cholecystokinin (CCK) varying in length from eight (CCK-8) to four (CCK-4) amino acids stimulate pepsinogen secretion from dispersed chief cells prepared from guinea pig stomach. C-terminal fragments of CCK containing fewer than four amino acids, even when tested at concentrations as high as 3 mM, do not stimulate pepsinogen secretion. The efficacies of gastrin and the various CCK-related peptides, coupled with the pattern of action of CCK receptor antagonists, indicate that chief cells from guinea pig stomach possess two functionally distinct classes of receptors, C-receptors and G-receptors. The C-receptors can be occupied by caerulein, CCK-8, CCK-7, des(SO3)CCK-8, or des(SO3)CCK-7, and occupation of C-receptors causes full stimulation of pepsinogen secretion. G-receptors can be occupied by gastrin I, gastrin II, CCK-6, CCK-5, or CCK-4, and occupation of G-receptors causes stimulation of pepsinogen secretion that is 60% of maximal.

Selective venous sampling for gastrin to localize gastrinomas. A prospective assessment.

In 27 consecutive patients with Zollinger-Ellison syndrome, we prospectively evaluated the ability of selective venous sampling for gastrin to localize gastrinomas, then compared the results with those from imaging studies and with findings at surgery. All patients had a gastrin gradient, but in only 20 patients was it significant. Neither the magnitude of the gastrin gradient nor its presence or absence correlated with the frequency with which gastrinoma was found at surgery. A gastrinoma was found at surgery in 15 patients, of whom 12 had positive imaging studies, 11 had a significant gastrin gradient, 14 had both tests positive, and 1 had both tests negative. A gastrinoma was not found at surgery in 12 patients, of whom 8 had a significant gradient and none had a positive imaging study. Gastrin sampling has equal sensitivity with imaging studies in localizing gastrinoma, but imaging studies have higher positive and negative predictive values and higher specificity. Thus, selective venous sampling for gastrin is much less useful in localizing gastrinoma than has been suggested and should not be routinely done preoperatively in patients with Zollinger-Ellison syndrome.

Famotidine, a new, potent, long-acting histamine H2-receptor antagonist: comparison with cimetidine and ranitidine in the treatment of Zollinger-Ellison syndrome.

Famotidine, a new, potent, long-acting histamine H2-receptor antagonist was compared with cimetidine and ranitidine in 9 patients with Zollinger-Ellison syndrome. The mean minimum daily requirement of famotidine to control gastric acid hypersecretion was 0.24 g (range 0.08-0.48 g) compared with 2.1 g (range 0.6-3.6 g) for ranitidine and 7.8 g (range 1.2-13.2 g) for cimetidine. Equally potent doses of the three drugs had similar onsets of action, but the duration of action of famotidine was 30% longer than the duration of action of either ranitidine or cimetidine (p less than 0.05). Eight patients were treated for up to 9 mo (mean 6 mo) with good control of gastric acid hypersecretion and with no evidence of biochemical or hematologic toxicity. These studies demonstrate that famotidine is nine times more potent than ranitidine and 32 times more potent than cimetidine, has a longer duration of action than ranitidine or cimetidine, and is both safe and effective in the long-term therapy of Zollinger-Ellison syndrome.

Omeprazole: effective, convenient therapy for Zollinger-Ellison syndrome.

The acute and long-term effects of omeprazole on gastric acid secretion were examined in 11 patients with Zollinger-Ellison syndrome. Basal gastric acid secretion was inhibited by 50% 3 h after a single 60-mg dose of omeprazole and 78% 4 h after administration of omeprazole. Patients were treated with a single daily dose of omeprazole, and the dose requirement was defined as the lowest dose of omeprazole that would reduce gastric acid secretion to less than 10 mEq/h during the last hour before the next dose. The mean daily dose requirement was 70 mg (range 20-160 mg). Ten of the 11 patients were given omeprazole once a day and 1 patient required omeprazole every 12 h. When omeprazole was discontinued after several months of therapy, mean basal gastric acid secretion was inhibited by greater than 50% 48 h after administration of omeprazole. Omeprazole continued to inhibit gastric acid secretion during 1-9 mo of therapy and patients remained free of toxicity or side effects related to omeprazole. Omeprazole is a highly effective inhibitor of gastric acid secretion in patients with Zollinger-Ellison syndrome. Because of its potency and long duration of action, omeprazole offers an advance in convenient medical therapy for Zollinger-Ellison syndrome compared with the histamine H2-receptor antagonists.