620 Gut, 1991,32,620-623 Peptic ulcer disease: absence ofantibodies

stimulating the histamine sensitive adenylate cyclase Gut: first published as 10.1136/gut.32.6.620 on 1 June 1991. Downloaded from of gastric mucosal cells

P Burman, S Mardh, L Loof, J Naesdal, F A Karlsson

Abstract pylori and duodenal ulcer as well as gastritis has The possible presence of been proposed." This issue is under current stimulating antibodies was examined in sera debate and awaits confirmation. from 57 patients with relapsing ulcer disease. It has been suggested that some forms of The sera were obtained at the time of sympto- duodenal ulcer may be of immunological matic relapse and all patients had ulcers origin.'2 Dobi and Lenkey'3 reported in 1982 confirmed by endoscopy. A sensitive assay that immunoglobulins prepared from sera of based on adenosine 3':5' cyclic monophos- patients with relapsing ulcers increased gastric phate (cAMP) production in isolated porcine acid secretion when injected into rats. Recently gastric mucosal cells was used as a measure. deLazarri etal claimed the existence ofhistamine cAMP production increased up to four hours receptor stimulating antibodies in 13 of 30 of incubation and was histamine responsive; duodenal ulcer patients.'4 In this study we an approximately 20-fold increase was found established a sensitive assay for detection of with histamine 104 moml. Sera from both parietal cell stimulation to explore this patients and healthy control subjects showed possibility. some inhibitory effect on basal cAMP production compared with incubation in medium only, whereas immunoglobulin Methods preparations had a weaker non-specific effect. No stimulation was found when the patients' PREPARATION OF PORCINE GASTRIC MUCOSAL AND http://gut.bmj.com/ sera and immunoglobulins (up to a concen- PARIETAL CELLS tration of 6 mglml) were examined. These Gastric mucosal cell preparations containing 15- results suggest that hypersecretion 25% (mean 18%) parietal cells were obtained in duodenal ulcer disease is not an effect of from the corpus mucosa of porcine stomachs histamine receptor stimulating antibodies. after treatment with pronase and collagenase. The datathus argue against a recent hypothesis that severe chronic ulcer disease in some on September 28, 2021 by guest. Protected copyright. patients has an autoimmune origin. STIMULATION OF ADENOSINE 3' :5' CYCLIC MONOPHOSPHATE (cAMP) PRODUCTION OF GASTRIC MUCOSAL CELLS Peptic ulcer is a multifactorial disease that is not Porcine corpus mucosal cells were incubated at yet fully understood. Gastric acid is considered +37°C in Eagle's medium with 20 mmol/l Hepes to be a major aggressive factor and gastric acid pH 7-4 and 0.5 mmol/l isobutylmethylxanthine hypersection has been shown in patients with in an atmosphere of 95% 02/5% C02. In pre- duodenal ulcer disease.' A large parietal cell liminary experiments, cAMP was detectable in mass,2 enhanced sensitivity to ,' and media, but not in cell lysates, after 30 minutes of increased secretion after meals have been incubation inthe presenceofaphosphodiesterase reported in subgroups of patients.4 Besides the inhibitor (0 5 mmol/l isobutylmethylxanthine). Departments of Internal presence of gastric acid, an impairment of We and others'6 have obtained analogous results Medicine and Medical mucosal defense mechanisms may be ofrelevance with respect to cAMPin experiments stimulating and Physiological in the development of ulcers. In several studies, the adenylate cyclase of thyroid cells in culture. Chemistry, Uppsala University, Uppsala, smoking has been associated with an increased Initially, concentrations of mucosal cells up to Sweden risk of relapse and slower ulcer healing rates.56 lOx 106/ml were used. A more appropriate P Burman Smoking reduces pancreatic bicarbonate output7 concentration was later found to be about S Mardh L Loof and thereby causes a decrease in duodenal pH. 2x 106/ml, and this concentration was used in F A Karlsson Furthermore, studies on intestinal mucosal experiments with patient immunoglobulins. blood flow in dogs have shown a negative effect After incubations with sera (20% v/v) or Department of Medicine, Bispebjerg Hospital, of nicotine.8 Impaired bicarbonate secretion immunoglobulin preparations (2-4 mg/ml or 6-0 University of from the proximal duodenal mucosa at rest and mg/ml) from patients and healthy subjects the Copenhagen, Denmark in response to luminal acidification has been mixtures were spun at 8000 g for 60 seconds. J Naesdal observed in patients with a history of duodenal The cell pellets were discarded and the amount Correspondence to: Dr P Burman, Department of ulcer disease.9 Chemical compounds such as of cAMP in the supernatants was assayed with a Internal Medicine, University aspirin and non-steroidal anti-inflammatory radioimmunoassay kit (Rianen cyclic AMP '251, Hospital, S-751 85 Uppsala, Sweden drugs have been associated with a high incidence Dupont). Submaximal and maximal stimulation Accepted for publication of gastric ulcers as well as ulcer bleeding.'` More with histamine (10-6 and 10 4M) were determined 20 August 1990 recently a causal relation between Helicobacter in each separate run. The interassay coefficient Peptic ulcerdisease: absence ofantibodies stimulating the histamine sensitive adenylate cyclase ofgastric mucosalcells 621

PATIENTS 100- Fifty seven patients (31 men and 26 women) aged A 20-78 years (mean 54 years) were included in

the All but four had a of Gut: first published as 10.1136/gut.32.6.620 on 1 June 1991. Downloaded from o / * ~~~~~~Fetalcc alfserum study. patients history mean _t / 0 Mediurnalf serum relapsing ulcer disease for one to 30 years, A Fetal caalf serum, histamine 10-4mol/l 10.8 years. Two patients had earlier been treated o /A Mediu n, histamine 10-4mo1/l for Graves' disease, another two were having thyroxine substitution treatment because of E autoimmune thyroiditis, two had non- dependent diabetes mellitus, two had hyper- tension, and two patients suffered from claudi- o catio intermittens. Thirty five of the patients were smokers, the majority of whom smoked 1 4 5 20 > 10 cigarettes per day. The patients were Time (hours) referred from general practitioners, ward units, Figure 1: Time dependent adenosijine 3':5' cyclic and outpatient clinics for endoscopy because of monophosphate(cAMP)productiu on in porcinegastric mucosal cells. Cells, S.8X 10/ml (20% parrietal cells) were incubated symptomatic ulcer relapse. Forty eight were at 370 in Eagle's medium containing05 mmol/l found to have duodenal ulcers while nine had isobutylmethylxanthine with 20% fetal calfsemum. prepyloric ulcerations. Treatment was begun Maximal cAMP production was eetstimated in the presence of with either H2 receptor antagonists, omeprazole, 10-4 moll histamine. Reactions were stopped by centrifugation, see methods sectioni. The results are expressed or rioprostil. With regard to the outcome of the as the mean ofduplicate incubations. ulcer medication, the patients could be separated into two groups. In 25 patients, including the four with a first attack ofulcer disease, the ulcers of variation, derived from,six duplicates of basal healed in less than two months on daily doses of ml either g cimetidine, 600 pg rioprostil, or 40 cAMP production with 1016 parietal cells per mg omeprazole.0-8 In the remaining 32 patients was 18%. there was a poor response to treatment. None of the ulcers healed after eight weeks on ranitidine IMMUNOGLOBULIN PREPARiATIONS 300 mg daily or cimetidine 0.8-1 g daily. In this Immunoglobulins were pirecipitated from sera group blood samples were taken within two with 1.6 mol/l ammoniuim sulphate, dialysed months ofan endoscopy showing an active ulcer. In the first group, in whom ulcers responded to

against Eagle's medium plH 7.4 and studied at a http://gut.bmj.com/ serum were obtained at the concentration of 2.4 mgV/ml. In subsequent treatment, samples experiments with higher concentrations of time of the first endoscopy. Control sera were immunoglobulins, the precipitates were more drawn from 24 healthy blood donors (14 men, 10 extensively dialysed agailibtn L dg1C Z Atmli-.M women; 22 to 65 years of age, mean 41 years). with 20 mmol/l Hepes pEI 7.4 and 0.5 mmol/l isobutylmethylxanthine in order to min imse thle

non-specific inhibition of parietal1 cAMP Results on September 28, 2021 by guest. Protected copyright. In the initial we searched for formation. Final protein concentratioins ranged experiments .8 mg/mle conditions that would optimise the assay. cAMP between 14X6 and 25.5 mg/L ml, mean 19 detection in the media increased as a function of the as assessed by the BioRad dye protein incubation time and reached a plateau at about kit. four hours (Fig 1). The cAMP production was stimulated by histamine. The response was nega- tively correlated to the cell concentration during the incubations, possibly by favouring anaerobic conditions or changes in pH. Thus, with -;; 250 Histamine 10-4m( histamine 10 mmol/l and 2 or lOx 106 cells/ml, -i 20 or sevenfold increases in cAMP were found, was a m respectively. The observation confirmed in 0 second set of experiments with various concen- trations ofcells (Fig 2). In further experiments, a E 150 four hour incubation period and cell concen- trations at about 2 x 106/ml were used routinely. U _~~~~ aa The effect ofhistamine was dose dependent (Fig ,o,0 0 3). The stimulatory effect of histamine was 05 reproducible and was confirmed with each cell 50 preparation. In six consecutive experiments the response to histamine 10`6mol/1 was 1-6-2*8-fold O (2-29 (0.21) mean (SEM)), while stimulation

0 2 6 8 10|0 121'2 with histamine 10` mol/l gave a 10 to 25-fold Gastric mucosal cells (106/ml) (18-5 (2.16)) increase, and 10-3 mol/l in two Figure 2: Effect ofcell concentration on adenosine 3':5' cyclic experiments gave >25-fold stimulation. monophosphate (cAMP)production inporcine gas,tric mucosal Sera from 17 ulcer patients were tested for cells. Gastric mucosal cells (15% parietal cells) at various their ability to stimulate cAMP production. No concentrations were incubatedforfour hours and b and maximal (10 `mol/l) histamine stimulated cAMPa0slrelease was stimulation was found. By contrast, a 20% tested as described in methods. The data represent jmean of impairment compared with incubation in duplicate incubations. medium alone, was repeatedly observed. By 622 Burman, Mardh, Loof, Naesdal, Karlsson

280- muscarinic, gastrin, and histamine. In vivo most of the gastrin stimulated acid secretion can be 240- blocked by H2 receptor antagonists, indicating an action via a histamine stimulated pathway.'7 Gut: first published as 10.1136/gut.32.6.620 on 1 June 1991. Downloaded from The addition ofhistamine results in rising values 200- of cAMP in the cell.'8 In order to test the hypothesis that stimulatory antibodies could be a 160- responsible for non-regulated acid secretion in CL subgroups of patients, we adopted a sensitive in 0 vitro system for analysis of cAMP in gastric E 120- mucosal cells. When sera or immunoglobulins in concentrations up to 6 mg/ml were used no < 80- stimulatoryeffect on the histamine receptor could be detected. Thisfindingis in contrastwith that of 40- deLazarri et al, who reported stimulating immunoglobulins, tested at 2 and 4 mg/ml, in 13 out of 30 cases.'4 With reference to severity, duration ofdisease, and response to treatment the io-8 lo07 10-6 10-5 l0-4 1°-3 patients in the two studies were comparable. The Histamine (mol/l) male:female ratio was 28:2 in their study and Figure 3: Histamine stimulated adenosine 3':5' cyClic 31:26 in ours. However, the test systems do seem monophosphate (cAMP)production inporcinegastric mucosal cells, 2.5x 106/ml (18% parietal cells) were incub to differ both with respect to response to hista- p1 Eagle's medium pH 7-4 with 0-5 mmolll mine (20-fold compared with fivefold cAMP isobutylmethylxanthine at 370. 15 [1 ofhistamine was added increase with maximal stimulation) and to perfor- tofinal concentrations of10 ` molll to 10-3 mol/l. Incubations were stopped afterfour hours by centrifugation an mance. In the study of deLazarri et al the non- the media assayed as described in methods. The reesultsare specific inhibition of cAMP production in the expressed as mean and range oftriplicates. presence of immunoglobulin preparations was considerable, and basal cAMP accumulation was reported to vary up to 12 times in different cell precipitating immunoglobulin fractions from preparations. Surprisingly, the relative incre- sera this non-specific effect was reduced to about mental stimulation from basal values with such 10%. In 33 patients with recurrent severe ulcer different stomachs were reported to show no disease, immunoglobulins were incubbated with statistical difference. In our study we attempted gastric mucosal cells at a concenttration of to optimise the assay to improve precision and http://gut.bmj.com/ 6 mg/ml. No stimulatory effect on the histamine obtained a histamine response better or similar to receptors of the parietal cells could be shown that described by others.'9 23 We used porcine (Fig 4). In another 10 patients, vvhere the gastric mucosal cell preparations whereas amounts of sera were limited, immuncDglobulins deLazarri et al used guinea pig mucosa. Human tested at 2.4 mg/ml, likewise, shiowed no autoantibodies in patients with various auto- stimulatory activity compared with controls. immune disorders generally crossreact well with The effect of 10-4 mol/l histamine was not tissues of other species. For instance, rat tissues on September 28, 2021 by guest. Protected copyright. blunted by the presence of either serum or are routinely used in immunofluorescence tests immunoglobulin preparations from healthy for detection ofantibodies against nuclei, smooth controls. muscle, and parietal cells,24 and in our experience, and that ofothers,'6 thyroid stimulating hormone receptor autoantibodies are equally well detected Discussion with rat or porcine cells as with human cells. In The parietal cell is stimulated to prodiice gastric our view, this excludes the possibility that the acid via at least three types of receptors: discrepancy between our study and that of deLazarri et al could be a result of the species differences in the experimental sources. Sera lg preparations Dobi and Lenkey,'3 by injections of immuno- - 120- (20%) (2.4mg/mi) (6.0mg/r-n) globulins into rats, describedgastric acid stimula- o tory immunoglobulins in 25 sera of 51 patients - 0 0 0 with duodenal ulcer. the use of an indirect 0 By 0 O> immunofluorescence technique the same 80-- 8

and in patients with partial gastrectomy (Billroth I). Gastro- antibodies against the parietal cell antigen, enterology 1976; 71: 552-7. H+,K+-ATPase, we did notfind autoantibodies in 5 Korman MG, Hansky J, Eaves ER, Schmidt GT. Influence of smoking on healing and relapse in duodenal ulcer disease.

ulcer Gut: first published as 10.1136/gut.32.6.620 on 1 June 1991. Downloaded from a group of 20 patients with duodenal Gastroenterology 1983; 86: 871-4. disease.25 In this study, a membrane fraction of 6 Sontag S, Graham D, Becsito A, et al. Cimetidine, cigarette cells was used as antigen in an enzyme smoking and recurrence of duodenal ulcer. N Engl J Med parietal 1984; 311: 689-93. linked immunosorbent assay. 7 Bynum TE, Solomon TE, Johnson LR, Jacobsen ED. In studies of risk factors for Inhibition of pancreatic secretion in man by cigarette previous smoking. Gut 1972; 13: 361-5. developing relapsing ulcer disease, an increased 8 Gallavan RH Jr, Tsuchiya Y, Jacobsen ED. Effects ofnicotine of smoking, alcohol abuse, intake of on canine intestinal blood flow and oxygen consumption. incidence Amj Physiol 1984; 246: G195-203. analgesic drugs, and family history of peptic 9 Isenberg JI, Selling JA, Hogan DL, Koss MA. Impaired There no proximal duodenal mucosal bicarbonate secretion in patients ulcer have been proposed. is, however, with duodenal ulcers. N EngljMed 1987; 316: 374-9. clinical association between severe ulcer disease 10 Somerville K, Faulkner G, Langman M. Non-steroidal anti- inflammatory drugs and bleeding peptic ulcer. Lancet 1986; and autoimmune endocrine disorders, such as i: 462-4. Graves' disease, thyroiditis, and Addison's 11 Hornick RB. Peptic ulcer disease: a bacterial infection? N Engl auto- JMed 1987; 316: 1598-600. disease, all characterised by organ specific 12 Rotter JI, Heiner DC. Are there immunologic forms of antibodies against cytoplasmic or membrane duodenal ulcer?J Clin Lab Immunol 1982; 7: 1-6. increased 13 Dobi S, Lenkey B. Role of a secretagogue immunoglobulin in bound antigens. Neither is there an gastric acid secretion. Acta Med Acad Sci Hung 1982; 60: 9- coexistence between the autoimmune disorder 25. 14 DeLazzari F, Mirakian R, Hammond L, Venturi C, Naccarato atrophic gastritis with pernicious anaemia and R, Bottazzo GF. Gastric cell c-AMP stimulating autoanti- relapsing ulcers.26 Parietal cell antibodies are not bodies in duodenal ulcer disease. Gut 1988; 29: 94-100. an in sera from 15 Mardh S, Norberg L, Ljungstrom M, Humble E, Borg T, found in increased frequency Carlsson C. Preparation of cells from pig gastric mucosa. patients with ulcer disease.2526 Yet, deLazzari Isolations of parietal cells by isopycnic centrifugation on linear density gradients of Percoll. Acta Physiol Scand 1984; et al have suggested that ulcer disease is a new 122: 607-13. member of this group of autoimmune disorders. 16 Kasagi K, Konishi J, Arai K, Misaki T, Llda Y, Endo K, the are not com- Torizuka K. A sensitive and practical assay for thyroid- The results of present study stimulating antibodies using crude immunoglobulin patible with this hypothesis. Also, the known fractions precipitated with polyethylene glycol. J Clin EndocrinolMetab 1986; 62: 855-62. risk factors and the clinical characteristics of the 17 Malinovska DH, Sachs G. Cellular mechanisms of acid patients argue against an autoimmune origin for secretion. Clin Gastroenterol 1984; 13: 309-26. our does not 18 Soll AH, Berlindh T. Physiology ofisolated gastric glands and duodenal ulcers. Although study parietal cells: receptors and effectors regulating functions. rule out the possibility of gastric stimulatory In: Johnson LR, ed. Physiology ofgastrointestinal tract 2nd edition. New York: Raven Press, 1987: 883-909. antibodies in exceptional cases, relapsing ulcer 19 Wollin A, Soll AH, Samloff IM. Actions of histamine, disease can not, in our view, be regarded as an secretin, and PGE2 on cyclic AMP production by isolated canine fundic mucosal cells. AmJ Physiol 1979; 237: E437- http://gut.bmj.com/ organ specific autoimmune disorder. 43. 20 Soll AH, Wollin A. Histamine and cyclic AMP in isolated We thank Thomas Bjorkman, Margareta Ericsson, and Maistin canine parietal cells. Amj Physiol 1979; 237: E444-50. Lundberg for providing excellent assistance. 21 Batzri S, Gardner JD. Cellular cyclic AMP in dispersed This study was supported by grants from the Medical Research mucosal cells from guinea pig stomach. Biochim Biophys Council (project no 4996, 4965), Swedish Society of Medicine, Acta 1978; 541: 181-9. Agnes and Mac Rudbergs Fund and 'Forenade Liv' Mutual Group 22 Chew CS, Hersey SJ, Sachs G. Berglindh T. Histamine Life Insurance Company, Stockholm, Sweden. responsiveness ofisolated gastric glands. Amj Physiol 1980; Part of these results were presented in abstract form at the 238: G312-20. American Gastroenterological Association Meeting, Washington, 23 Holian 0, Ruiz C, Bombeck CT, Nyhus LM. Comparison of May 1989. cAMP system in parietal cells from rat and guinea pig. Scand on September 28, 2021 by guest. Protected copyright. JGastroenterol 1983; 16: 819-24. 24 Zweiman B, Lisak RP. Autoantibodies: autoimmunity and 1 Soll AH, Isenberg JI. Duodenal ulcer disease. In: Sleisenger immune complexes. In: Todd I, Sanford M, Davidsohn R, MH, Fordtran JS, eds. Gastrointestinal disease: patho- Henry JB, eds. Clinical diagnosis and management by labora- physiology, diagnosis and management. 3rd edition. Phila- tory methods. Philadelphia: WB Saunders Company, 1984: delphia: WB Saunders, 1983: 625-71. 924-54. 2 Cox AJ, Stomach size and its relation to chronic peptic ulcer. 25 Karlsson FA, Burman P, L6of L, Olsson M, Scheynius A, Arch Pathol 1952; 54: 407-22. Mardh S. Enzyme-linked immunosorbent assay of H+,K+- 3 Isenberg JJI, Grossman MI, Maxwell V, Walsh JH. Increased ATPase, the parietal cell antigen. Clin Exp Immunol 1987; sensitivity to stimulation ofacid secretion by pentagastrin in 70:604-10 duodenal ulcer.J Clin Invest 1975; 55: 330-7. 26 Ungar B, Francis CM, Cowling DC. Antibody to parietal cells 4 Fritsch WP, Hausaman TU, Rick W. Gastric and extragastric in patients with duodenal ulcer, and relationship to per- gastrin release in normal subjects, in duodenal ulcer patients, nicious anemia. MedjAust 1976; 2: 900-2.