GASTROENTEROLOGY
1991:101:131-137
Basophil-Bound and Serum Immunoglobulin E Directed Against Helicobacter pylori in Patients With Chronic Gastritis ANTONIO VINCENZO ANTONIO FRANCESCO ANTONIO
ACETI,
DOMENICO
CASALE, GRASSI,
CELESTINO,
FRANCESCO ANASTASIO
SCIARRETTA,
MARTA
CITARDA, GRILLI,
ORIANA
ALFRED0
LERI,
CAFERRO,
ETTORE FRANC0
M. CONTI,
PENNICA, AMEGLIO,
and
SEBASTIAN1
Institute of the Clinic of Tropical and Infectious Diseases, “Regina Elena” National Cancer Institute, Rome, Italy
The immunoglobulin (Ig) E immune response in patients with Helicobacterpylori-associated chronic gastritis has been evaluated. Of 26 patients with H. pylori infection, 22 (84%) tested positive for basophilbound specific IgE (determined by the histamine release test) and 18 (69%) for serum specific IgE (determined by an enzyme-linked immunosorbent assay). In contrast, only 1 of 17 persons in whom the bacterium was not detected presented cell-bound and serum specific IgE. In the 4 histamine release test-positive but enzyme-linked immunosorbent assay-negative patients, removal of antibody from the basophil surface by acid elution showed that histamine release occurred through an IgE-dependent mechanism. When normal basophils, passively sensitized with serum from IgE-positive patients, were exposed to the H. pylori antigen, a significant release was observed, confirming the class specificity of the response. Inhibition experiments with bacteria other than H. pylori showed that the IgE antibody was specifically directed against this organism. The percentage of antigen-induced histamine release did not correlate with serum specific IgE level. However, the response of basophils to antigenic challenge was proportional to &E-dependent cellular releasability. This finding suggests that target cell sensitivity may be the most important factor in determining the entity of biological response to the antigenic challenge. The ability of H. pylori to induce a specific IgE immune response could answer key questions regarding the mechanisms inducing gastric inflammation.
“La Sapienza”
University,
Rome; and
S
piral bacteria were originally observed in human gastric tissue many years ago (l), but early investigations to confirm this observation and to examine its significance have shown conflicting results (2,3). These observations were therefore ignored for many years until the recent investigations of Warren and Marshall (4) and Marshall and Warren (5). These authors have reported a significant association between the occurrence of gastritis and the presence in the stomach of a new bacterium, now called Helicohater pylori, implying that this organism may be a possible cause of the disease. Since then, similar results have been reported by other investigators (6). Although H. pylori is clearly associated with histologically proven gastritis, its specific role in causing this disease is still unclear. The presence of H. pylorispecific immunoglobulin (Ig) G in serum and IgA in gastric juice provides further evidence of the pathogenetic role of this organism in gastroduodenal disease (7-9). It is interesting that the presence of such antibodies does not seem able to eliminate the infection or to prevent the recrudescence (10). It could be that the antigenic load to which the gastric mucosa is exposed during H. pylori infection initiates, or at least exacerbates, the inflammatory responses, leading to gastritis. In this regard, studies on the immune mechanisms operating on this mucosal surface should be
Abbreviation
used in this paper: HRT, histamine
D 1991 by the American
Gastroenterological
00X6-5085/91/$3.00
release test. Association
GASTROENTEROLOGY Vol. 101,No. 1
132 ACETI ET AL.
conducted. Indeed, there is a close association between H. pylori colonization and acute neutrophilic inflammation, both rare in the absence of chronic lymphomononuclear inflammation (4,5,11,12). Recent observations of gastric biopsy specimens containing the bacterium also revealed eosinophil infiltration (13). Human allergic reactions and degranulation manifest similar histological features associated with clinical inflammatory signs (14,151. Cellular recruitment is regarded as a consequence of the release of chemotactic factors from mediator-containing cells, triggered by antigen-induced IgE-dependent mast cell degranulation (16,17). Thus the histological picture of H. pylori-induced gastritis may be associated with and dependent on antigen-specific IgE. At present there is no evidence of systemic allergic manifestations in patients with H. pylori infection. However, the features of an atopic condition have been observed in a large number of patients with chronic gastritis and peptic ulceration (18). Recent studies have also shown that bacteria can induce histamine release from human basophils by an &E-mediated reaction. In addition, it has been suggested that this release could be a pathogenetic mechanism in infectious diseases (19,20). These observations have led us to evaluate the immune response involving IgE in patients with chronic gastritis who are infected by I-I.pylori. Materials and Methods Forty-three adult subjects (22 male and 21 female) were submitted to gastroscopy; the indications for upper endoscopy were mild or moderate gastrointestinal symptoms (dyspepsia or pain), suspected chronic atrophic gastritis, a healed peptic ulcer, or a family history of stomach cancer. Patients with stomach neoplasm or a history of allergic diseases and those recently treated with antibiotics, aspirin or nonsteroidal antiinflammatory drugs, H, blockers, or antihistamine drugs were excluded from the study. A clinical report was completed for each patient.
Endoscopy Esophagogastroduodenoscopy was performed with forward-view Olympus endoscopes (GIF-QlO and GIFXQlO; Olympus, Lake Success, NY), and at least two biopsy specimens were obtained from the antrum. The endoscope and biopsy forceps were disinfected in glutaraldehyde after each use.
Histology Tissue samples were fixed in 10% buffered formalin and embedded in paraffin. Sections were stained with H&E for light microscopy. The criteria used to classify gastritis are those commonly applied (21).
Culture Test A tissue sample from each patient was transported to the laboratory in a sterile container without the addition of preservatives and was processed within 1 hour after collection. The biopsy specimen was cultured on 5% lysed horse blood agar, selective for H. pylori (supplement SR69; Oxoid). The plates were incubated at 37°C under reduced oxygen for 7 days and examined at 3, 5, and 7 days for growth of organisms morphologically resembling Campylobatter. The following features were suggestive of H. pylori: microaerobic growth at 37°C; curved or S-shaped cells; positive reactions for catalase, oxidase, and urease; negative reactions for hippurate hydrolysis and nitrate reduction; susceptibility acid.
to cephalothin;
and resistence
to nalidixic
Antigen Preparation The antigens used were the surface components
of
H. pylori cells harvested with an acid wash from 10 different strains according to the method described by McCoy et al. (22). The composition of this preparation has previously been described in detail (23). Antigens from unrelated bacteria (Campylobacter jejuni and Campylobacter coli) were also prepared according to the same procedure.
Histamine Release Assay After informed consent was obtained, venous blood was drawn by venipuncture at the time of the endoscopy. Mixed leukocytes containing basophils were prepared by dextran sedimentation (24). In brief, 25 mL venous blood was mixed with 6.2 mL 6% clinical dextran and 2.5 mL 0.1 mol/L ethylenediaminetetraacetic acid and allowed to sediment for 60-90 minutes at room temperature. containing plasma layer was then removed sedimented by centrifugation (150 x g for 8 washed twice with PIPES buffer (Sigma, containing human serum albumin and
The leukocyteand cells were minutes at 4”C), St. Louis, MO) glucose, resus-
pended by gently shaking, and then diluted in appropriate volume with the same buffer containing Ca2+ and Mg” for use in the histamine release assay. After preincubation at 37”C, a volume of 0.4 mL mixed leukocytes was added in prewarmed tubes containing 0.2 mL of the different bacterial antigens at various concentrations (100, 50, and 10 kg/mL) and 0.4 mL of the buffer containing Ca’+ and Mg”. To assess the IgE-dependent basophil releasability, the cell preparations were also challenged with goat anti-human IgE (Sigma). The number of leukocytes was chosen so that each tube contained 20-40 ng histamine. The tubes therefore contained approximately 3-5 x lo6leukocytes (2-4 x lo4basophils). Spontaneous release was that observed when cells were added to the buffer alone. After incubation at 37°C for 45 minutes, the tubes were centrifuged (1000 x g for 3 minutes at 23°C) and aliquots of the cell-free supernatant were
IgE TO H. PYLDRI IN CHRONIC GASTRITIS
July 1991
assessed for histamine content by the automated fluorometric assay (25). The results were based on the mean of the duplicate determinations and expressed as percent of histamine release by dividing by total cellular histamine after subtracting for spontaneous release of unstimulated basophils (typically, 2%-5% during the 45-minute assay). Total cellular histamine (100% release) was obtained by lysing cells in 2% perchloric acid. The specimens with a histamine release > 10% were considered positive (26). Bacterial antigens did not contain histamine or interfere with the histamine measurement. No functional IgG was found on the cell surface because the leukocytes did not release histamine when challenged with anti-human IgG. Enzyme-Linked
Zmmunosorben t Assay
Serum IgE specific for the H. pylori antigen was determined by indirect enzyme-linked immunosorbent assay (ELISA) performed using a microtitration procedure (27). In the present study, the optimal antigen concentration was 50 pg/mL and the optimal dilutions for the serum and conjugate (goat anti-human IgE, alkaline phosphataselabeled; Sigma) were so and ~000, respectively, according to checkerboard titration. All tests were performed in duplicate. The cut-off value (optical density above 0.200) was calculated by adding 3 SD to the mean of absorbance of negative subjects. Passive Sensitization
of Normal Basophils
To demonstrate the class specificity of the response, peripheral blood leukocytes (basophils) from a normal donor were sensitized with serum of ELISA &E-positive patients (28) and then exposed to bacterial antigens. The histamine level in the supernatant was determined as described above.
133
The IgE elution from basophils was confirmed by the lack of histamine release following anti-human IgE challenge. Statistical Analysis Paired and unpaired t tests, x2 analysis, the F test, and the correlation coefficient (r) were used for statistical analysis. Results Detection Diagnosis
ofH. pylori
and Histological
Thirty-two of the 43 antral biopsy specimens showed histological evidence of gastritis (19 superficial and 13 atrophic). Normal histological features were found in the remaining 11 samples. H. pylori was isolated by culture in 26 of the 32 patients with gastritis, showing a correlation between the H. pylori and histological gastritis. In contrast, none of the biopsy specimens with normal mucosa contained H. pylori. Detection ofanti-H. Zmmunoglobulin E
pylori
Cell-Bound
The ranges of maximal histamine release values obtained in subjects in whom H. pylori was or was not detected are shown in Figure 1. There is a highly significant difference between these two groups (F test; P < 10m5). Of the 17 individuals not infected by H. pylori, 16 had release values below lo%, and only 1 with gastritis had a 40% release value (mean ‘_ SD,
Inhibition Assay To establish whether the IgE antibody detected using ELISA test was specifically directed against H. pylori antigens, positive sera were incubated in triplicate at 37°C for 60 minutes (one with 50 p,g of H. pylori antigen and the other two with an acid extract of C. jejuni or C. coli). The specimens were then tested by ELISA, performed as described above. Acetate Elution of Basophil Zmmunoglobulin
.
. .
.
.
E
To determine whether histamine release was IgE dependent, aliquots of cells from four patients with H. pylori-induced gastritis were stimulated with the antigen before and after acetate elution of basophil IgE. According to the method described by Ishizaka and Ishizaka (29), after the second wash (see above] the cells were resuspended in 1 mL of acetate buffer, pH 3.7, and allowed to incubate for 10 minutes at 0°C. The cells were sedimented by centrifugation, washed twice with PIPES buffer containing human serum albumin and glucose, and resuspended in the same buffer containing Ca2+ and Mg”.
.
J Figure 1. Histamine release values obtained by challenging with H. pylori antigens basophils from infected (left) and noninfected (right) individuals. Dottedhe shows cutoff limit.
134
ACETI ET AL.
GASTROENTEROLOGY Vol. 101, No. 1
5 -C 9). In contrast, 3 of the infected patients had 100% release values, 6 had values between 95% and SO%, 10 between 50% and 30%, and 3 between 25% and 15%. Thus, when a cutoff value (10%) to discriminate between reactive and unreactive subjects for histamine release test (HRT) was considered, 22 (84.6%) of the 26 patients with H. pylori infection were positive, whereas only 1 (5.8%) positive result was found in the noninfected persons (x” test, P < 10e5; Table 1). Histamine release values obtained from basophils challenged with bacterial antigens other than H. pylori never reached 10%. Release due to anti-human IgE did not differ significantly between infected (mean + SD, 33 + 11) and noninfected individuals (mean * SD, 31 + 12; unpaired t test, P > 0.5). However, a significant correlation was found between the antigen and the anti-IgEinduced histamine release values (r = 0.7; P < 10w3).
Detection ofAn&H. Immunoglobulin E
pylori Serum
Specific IgE serum antibodies were detected by ELISA in 18 (69.2%) patients with H. pylori and in 1 (5.8%) noninfected individual (x” test, P < 10m3; Table 1). This was a true IgE response. In fact, normal basophils passively sensitized with ELISA-positive sera released histamine upon H. pylori antigen exposure in a significant manner (mean 4 SD, 38 r 9). The remaining bacterial antigens were ineffective in inducing basophil degranulation. The inhibition assay showed that the IgE antibody was specifically directed against H. pylori. In fact, as Figure 2 shows, antigens from this bacterium adsorbed out IgE in positive sera (paired t test, P < 10e6), whereas extracts from C. jejuni and C. coli induced only a very slight reduction of optical density values in reactive specimens (paired t test, P > 0.5). All subjects with detectable specific serum IgE had positive HRT results, although no correlation between the percentage of histamine release and the ELISA value (r = 0.14; P > 0.5) was observed.
Table 1, Results Obtained Using Histamine Release Test and an Enzyme-Linked Immunosorbent Assay for Detecting H. pylori-specific Immunoglobulin E in the Subjects Studied Group
rl
HRT-positive subjects
ELISA-positive subjects
Individuals with H. pylori infection Noninfected individuals
26 17
22 (84.6%) 1(5.8%)
18 (69.2%) 1 (5.8%)
2. H. pylori-specific IgE, expressed as optical density values ( x lo), before (left) and after (rig&) adsorption with three dillkent bacteria. Squares indicate median values.
Figure
Acetate Elution of Basophil Immunoglobulin
E
The leukocytes of the 4 patients with H. pylori infection who tested positive upon HRT without detectable specific serum IgE were challenged with the bacterial antigen both before and after acidic treatment. As reported in Table 2, a significant decrease in histamine release was found after IgE elution (paired t test, P < lo-*). Discussion The aim of the present study was to test whether patients with H. pylon’ have a specific demonstrable IgE-mediated immune response. It has previously been shown that bacteria can induce histamine release from human basophil leukocytes and mast cells by an &E-mediated pathway. This mechanism was discovered when basophils from patients with intrinsic asthma were challenged with Haemophilus influenzae and Streptococcus pneumoniae (26). An &E-dependent histamine release was also seen in patients with Escherichia coli bacteremia and fever
IgE TO H. PYLORI IN CHRONIC GASTRITIS
July 1991
Table 2. H. pylori-Induced Histamine Release Values in Four Patients Before and After Immunoglobulin E El&ion Patient A.P. G.F. G.C. V.L.
Before IgE elution (%I
After IgE elution (%I
63 70 83 94
5 6 5 3
and in patients suffering from the immune deficiency syndrome with specific IgE to Staphylococcus aureus isolated from those patients (26,30). On the other hand, a wide spectrum of both gram-positive and gram-negative bacteria can release histamine through a lectin-mediated mechanism and other nonimmunological mechanisms (26). The data of the present study strongly suggest that an immunologic response involving IgE is present in patients infected by H. pylori. Indeed, basophil-bound specific IgE was seen in 84% and serum specific IgE in 69% of the persons with H. pylori infection. In contrast, among the individuals in whom the bacterium was not detected, only one, with histologically proven gastritis, had cell-bound and serum specific IgE. In the subjects who tested positive by HRT without detectable specific circulating IgE, acidelution experiments showed that histamine release from basophils occurred through an immunologic mechanism. To demonstrate further the class specificity of the response, the H. pylori antigen was used to challenge normal basophils passively sensitized with serum from IgE-positive patients. The significant histamine release obtained upon exposure to antigen points to the reliability of the ELISA test and excludes interference with antibodies other than IgE directed against the same antigen. Bacteria other than H. pylori were ineffective in inducing histamine release from basophils of infected patients and in adsorbing out specific serum antibody. These data confirm that the IgE detected were specifically directed against H. pylori antigens. As expected, the percentage of antigen-induced histamine release was not correlated with the serum specific IgE level. It has been previously shown that although a good correlation exists between cell-bound and serum IgE in human basophils, histamine release is not related to the absolute quantity of IgE antibody present on cells challenged with the appropriate antigen (31). In fact, there is evidence that in addition to immunologic parameters, the release of chemical mediators from a target cell also depends on its releasability, i.e., the intrinsic capacity of the cell to release mediators in response to a variety of immuno-
135
logic and nonimmunologic stimuli (32). We have recently shown that infectious factors may play a role in the control of IgE-mediated releasability (33). In the present study, the anti-IgE-induced histamine release did not differ in the infected patients and in the control subjects. However, the basophil response to antigenic challenge was related to IgE-dependent cellular releasability. In fact, a close correlation was present between anti-IgE- and antigen-induced histamine release. This observation may have some clinical importance because it suggests that target cell sensitivity is the major factor determining the entity of the biological response to antigenic challenge. Although much evidence has been accumulated concerning an etiologic role for H. pylori in chronic gastritis, the pathogenetic mechanisms involved in bacterium-associated gastroduodenal inflammation are still unclear. The ability of H. pylori to induce a specific IgE immune response could answer key questions regarding the mechanisms of gastric inflammation. For many years, allergic reactions have been considered acute and short-lived, capable of causing ephemeral clinical problems with few chronic features. There have always been peculiar clinical features that have failed to fit within this concept. Moreover, it became clear in the laboratory assessment of mast cell mediators that this cell could elicit intense, prolonged inflammatory reactions (14,15). Since the early 197Os, there has been increasing clinical and experimental evidence of the capacity of allergic reactions to evolve into chronic inflammatory responses (34). In fact, in a variety of experimental models and disease states, it is clear that an initial immunological trigger, i.e., the antigen and mast cell-bound IgE, leads not only to the rapid release of vasoactive mediators but also to the so-called late-phase reaction, characterized by the involvement of inflammatory cells (neutrophils, eosinophils, mononuclear cells). The infiltrate consisting of polymorphonuclear leukocytes is an early event in such a reaction, followed by a later mononuclear cell infiltration. Cellular recruitment is regarded as a result of the release of mast cell-derived chemotactic factors (16,17). Furthermore, both inflammatory cells and mediators of hypersensitivity contribute to the late sustained response, as well as to the subsequent subacute/chronic inflammatory phase (34). The latephase reaction has not been clearly demonstrated in the gastrointestinal tract, although late-phase reactions may well occur. The histological picture of H. pylori-associated gastritis is quite consistent with that observed in allergic reactions. Indeed, the biopsy specimens containing the bacterium are characterized by the presence of a neutrophilic infiltrate in the context of a diffuse mononuclear cellularity, and eosinophil leuko-
136
ACETI ET AL.
cytes have been observed in association with the bacterium in gastric biopsy samples (4,5,11-13). Thus, it is conceivable that IgE plays a role in mediating H. pylori-induced inflammatory processes in the stomach through the activation of gastric mast cells. The presence of IgE in gastric mucosa is not a new phenomenon, as there is evidence of IgEmediated reactions in patients with erosive gastritis and peptic ulcer (35,36). In addition, IgE directed against cockroach antigens has recently been detected in the serum and gastric mucosa of patients with peptic ulcer and chronic gastritis (18). It is possible that not only histamine, but also other preformed or newly generated mediators released by mast cells, are implicated in tissue damage development. For instance, a potent ulcerogenic action of the plateletactivating factor on the stomach has recently been shown (37). In addition, a direct and/or indirect pathogenetic role for infiltrating cells cannot be excluded. The results of the present study should have therapeutic implications. However, mast cells exhibit morphological and functional heterogeneity both between species and between organs of the same species (38). Some potent pharmacological agents which inhibit the generation, release, or target organ effects of specific mast cell-dependent mediators, have recently been developed. This has permitted the pharmacological dissection of the relative contributions of individual mediators to the pathogenesis of acute and chronic allergic reactions in humans (39). Drugs that inhibit the formation or the target tissue effects of individual mast cell-dependent mediators may exert differing effects from one organ to another, and consequently the assessment of drug activity at one tissue site might not predict its effect at another.
References 1. Salomon H. Ueber das Spirillum des Saugetiermagens und sein Verhalten zu den belegzellen. Zentralbl Bakteriol1896;19:433442. 2. Freedberg AS, Barron LE. The presence of spirochetes in human gastric mucosa. Am J Dig Dis 1940;7:443-445. 3. Palmer ED. Investigation of the gastric mucosa spirochetes of the human. Gastroenterology 1954;27:218-220. 4. Warren JR, Marshall BJ. Unidentified curved bacilli on gastric epithelium in active chronic gastritis. Lancet 1983;1:12731275. 5. Marshall BJ, Warren JR. Unidentified curved bacilli in the stomach of patients with gastritis and peptic ulceration. Lancet 1984;1:1311-1315. 6. Dooley CP, Cohen H. The clinical significance of Campylobacterpylori. Ann Intern Med 1988;108:70-79. 7. Rathbone BJ, Wyatt JI, Worsley BW, Shires SE, Trejdosiewicz LK, Heatley RV, Losowsky MJ. Systemic and local antibody responses to gastric Campylobacter pyloridis in nonulcer dyspepsia. Gut 1986;27:642-647. 8. Booth L, Holdstock G, MacBride H, Hawtin P, Gibson JR,
GASTROENTEROLOGY
Vol. 101, No. 1
Ireland A, Bamforth J, DuBoulay CE, Lloyd RS, Pearson AD. Clinical importance of Campylobacferpyloridis and associated serum IgG and IgA antibody responses in patients undergoing upper gastrointestinal 219. 9. Goodwin CS, Blincow
endoscopy.
J Clin Path01 1986;39:215-
E, Peterson G, Sanderson
C, Cheng W,
Marshall B, Warren JR, McCulloch R. Enzyme-linked immunosorbent assay for Campylobacter pyloridis: correlation with presence of C. pyIoridis in the gastric mucosa. J Infect Dis 1987;155:488-494. 10. Langenberg W, Rauws EA, Widjojokusumo A, Tytgat GN, isolates by Zanen HC. Identification of Campylobacterpyloridis restriction endonuclease DNA analysis. J Clin Microbial 1986; 24:414-417. 11. Andersen LP, Holck S, Povlsen CO, Elseborg L, Justesen T. Campylobacter pyloridis in peptic ulcer disease. I. Gastric and and duodenal infection caused by C. pyloridis: histopathologic microbiologic findings. Stand J Gastroenterol 1987;22:219224. 12. Berstad A, Alexander B, Weberg R, Serck-Hanssen A, Holland
13.
14.
15. 16.
17.
S, Hirschowitz BI. Antacids reduce Campylobacter pylori colonization without healing the gastritis in patients with nonulcer dyspepsia and erosive prepyloric changes. Gastroenterology 1988;95:619-624. McGovern TW, Kephart GM, Gleich GJ, Talley NJ. Eosinophil infiltration and degranulation in Campylobacter pylori gastritis. Klin Wochenschr 1989;67(Suppl 18):46. Gleich GJ. The late phase of the immunoglobulin E-mediated reaction: a link between anaphylaxis and common allergic disease? J Allergy Clin 1mmuno11982;70:160-169. Lemanske RF, Kaliner M. Mast cell-dependent late phase reactions. Clin Immunol Rev 1982:1:547-580. Kay AB, Austen KF. The IgE-mediated release of an eosinophil leukocyte chemotactic factor from human lung. J Immunol 1971;107:899-902. Nagy L, Lee TH, Kay AB. Neutrophil chemotactic activity in antigen-induced late asthmatic reactions. N Engl J Med 1982; 306:497-501.
18, Romamski B, Bartuzi Z, Zbikowska-Gotz M, Staszynska M, Korenkiewicz J. Allergy to cockroach antigens in patients with peptic ulcers and chronic gastritis. Allergol Immunopathol 1988;16:219-224. 19. Norn S, Stahl Skov P, Jensen C, Koch C, Permin H, Bog-Hansen C, Lovenstein H, Hoiby N. Intrinsic asthma and bacterial histaminereleasevialectineffect. Agents Actions 1983;13:210212. 20. Jensen C, Norn S, Stahl Skov P, Espersen F, Koch C, Permin H. Bacterial histamine release by immunological and nonimmunological lectin-mediated reactions. Allergy 1984;39:371377. 21. Whitehead R, Truelove SC, Gear ML. The histological diagnosis of chronic gastritis in fiberoptic gastroscope biopsy specimens. J Clin Path01 1972;25:1-11. 22. McCoy EC, Doyle D, Burda L, Corbeil LB, Winter AJ. Superficial antigens of Campylobacter (Vibrio] fetus: characterization of an antiphagocitic component. Infect Immun 1975;11:517-525. 23. Aceti A, Attanasio R, Pennica A, Taliani G, Sebastiani A, Rezza G, Ippolito G, Perucci CA. Campylobacter pylori infection in homosexuals. Lancet 1987;2:154-155. LM, Osler AG. Studies on the mechanisms of 24 Lichtenstein hypersensitivity phenomena. IX. Histamine release from human leukocytes by ragweed pollen antigens. J Exp Med 1964; 120:507-530. 25, Siraganian RP. Automated histamine release. A method for in vitro allergy diagnosis. Int Arch Allergy Appl Immunol1975;49: 108-110. 26. Norn S, Stahl Skov P, Jensen C, Jarlov JO, Espersen F.
July 1991
27.
28.
29.
30.
31.
32. 33.
34.
A new mechanism in Histamine release induced by bacteria. asthma? Agents Actions 1987;20:29-34. Voller A, Draper CC, Bidwell DE, Bartlett A. Microplate enzyme-linked immunosorbent assay for Chagas’ disease. Lancet 1975;1:426-429. Aceti A, Celestino D, Caferro M, Pennica A, Sebastiani A. Total and specific IgE in sera from patients with invasive amoebiasis. Trop Geogr Med 1989;41:151-153. Ishizaka T, Ishizaka K. Mechanisms of passive sensitization. IV. Dissociation of IgE molecules from basophil receptors at acid pH. J 1mmuno11974;112:1078-1084. Norn S, Jensen C, Dahl BT, Stahl Skov P, Baek L, Permin H, Jarlow JD, Sorensen H. Endotoxins release histamine by complement activation and potentiate bacteria-induced histamine release. Agents Actions 1986;18:149-152. Conroy MC, Adkinson FN, Lichtenstein LM. Measurement of IgE on human basophils: relation to serum IgE and anti-IgEinduced histamine release. J Immunol1977;118:1317-1321. Lichtenstein LM, MacGlashan DW. The concept of basophil releasability. J Allergy Clin Immunol 1986;77:291-294. Aceti A, Celestino D, Caferro M, Teggi A, Pennica A, Adriani E, Grilli A, Paparo S, Leri 0, De Rosa F, Sebastiani A. Basophil releasability in human hydatidosis. Int Arch Allergy Appl 1mmun011990;91:111-112. Kaliner M, Lemanske R. Mast cell-derived inflammatory factors and late-phase allergic reactions. In: Kay AB, Austen KF, Lichtenstein LM, eds. Asthma: physiology, immunopharmacology, and treatment. London: Academic, 1984:229-244.
IgE TO
H. PYLORIIN CHRONIC
C, Moulinier 35. Andre anaphylactic reaction
B, Andre in peptic
GASTRITIS
F, Daniere ulcer
137
S. Evidence
and varioliform
for
gastritis.
Ann Allergy 1983;51:325-331. 36. Andre F, Andre C, Cavagna S. Role of histamine in the cell turnover changes associated with experimental gastric ulceration in the mastomys. Gastroenterology 1985;88:452-457. 37. Rosam
A-C, Wallace
JL, Whittle
tions of platelet-activating 1986;318:54-56. 38. Pearce J, Ennis
FL, Ali H, Barrett M, Flint
BJR. Potent
factor
on
the
ulcerogenic stomach.
KE, Befus AD, Bienenstock
KC, Hudspith
B, Johnson
J, Brostoff
NM, Leung
Peachell PT. Functional characteristics of mucosal tive tissue mast cells of man, the rat and other
ac-
Nature
KBP,
and connecanimals. Int
Arch Allergy Appl Immunol 1985;77:274-279. 39. Holgate ST, Rafferty P, Beasley R, Robinson C, Hovel1 CJ, Curzon NP, Church MK. In vitro and in viva studies on mast cells of human skin and airways. In: Kay AB, ed. Allergy and inflammation. London: Academic. 1987:29-52.
Received July 31, 1989. Accepted December 3,199O. Address requests for reprints to: Antonio Aceti, M.D., Clinica delle Malattie Tropicali ed Infettive, Policlinico Umberto I’, 00161 Rome, Italy. This work was supported della Pubblica Istruzione.
in part by grants
from the Minister0
1991:101:131-137
Basophil-Bound and Serum Immunoglobulin E Directed Against Helicobacter pylori in Patients With Chronic Gastritis ANTONIO VINCENZO ANTONIO FRANCESCO ANTONIO
ACETI,
DOMENICO
CASALE, GRASSI,
CELESTINO,
FRANCESCO ANASTASIO
SCIARRETTA,
MARTA
CITARDA, GRILLI,
ORIANA
ALFRED0
LERI,
CAFERRO,
ETTORE FRANC0
M. CONTI,
PENNICA, AMEGLIO,
and
SEBASTIAN1
Institute of the Clinic of Tropical and Infectious Diseases, “Regina Elena” National Cancer Institute, Rome, Italy
The immunoglobulin (Ig) E immune response in patients with Helicobacterpylori-associated chronic gastritis has been evaluated. Of 26 patients with H. pylori infection, 22 (84%) tested positive for basophilbound specific IgE (determined by the histamine release test) and 18 (69%) for serum specific IgE (determined by an enzyme-linked immunosorbent assay). In contrast, only 1 of 17 persons in whom the bacterium was not detected presented cell-bound and serum specific IgE. In the 4 histamine release test-positive but enzyme-linked immunosorbent assay-negative patients, removal of antibody from the basophil surface by acid elution showed that histamine release occurred through an IgE-dependent mechanism. When normal basophils, passively sensitized with serum from IgE-positive patients, were exposed to the H. pylori antigen, a significant release was observed, confirming the class specificity of the response. Inhibition experiments with bacteria other than H. pylori showed that the IgE antibody was specifically directed against this organism. The percentage of antigen-induced histamine release did not correlate with serum specific IgE level. However, the response of basophils to antigenic challenge was proportional to &E-dependent cellular releasability. This finding suggests that target cell sensitivity may be the most important factor in determining the entity of biological response to the antigenic challenge. The ability of H. pylori to induce a specific IgE immune response could answer key questions regarding the mechanisms inducing gastric inflammation.
“La Sapienza”
University,
Rome; and
S
piral bacteria were originally observed in human gastric tissue many years ago (l), but early investigations to confirm this observation and to examine its significance have shown conflicting results (2,3). These observations were therefore ignored for many years until the recent investigations of Warren and Marshall (4) and Marshall and Warren (5). These authors have reported a significant association between the occurrence of gastritis and the presence in the stomach of a new bacterium, now called Helicohater pylori, implying that this organism may be a possible cause of the disease. Since then, similar results have been reported by other investigators (6). Although H. pylori is clearly associated with histologically proven gastritis, its specific role in causing this disease is still unclear. The presence of H. pylorispecific immunoglobulin (Ig) G in serum and IgA in gastric juice provides further evidence of the pathogenetic role of this organism in gastroduodenal disease (7-9). It is interesting that the presence of such antibodies does not seem able to eliminate the infection or to prevent the recrudescence (10). It could be that the antigenic load to which the gastric mucosa is exposed during H. pylori infection initiates, or at least exacerbates, the inflammatory responses, leading to gastritis. In this regard, studies on the immune mechanisms operating on this mucosal surface should be
Abbreviation
used in this paper: HRT, histamine
D 1991 by the American
Gastroenterological
00X6-5085/91/$3.00
release test. Association
GASTROENTEROLOGY Vol. 101,No. 1
132 ACETI ET AL.
conducted. Indeed, there is a close association between H. pylori colonization and acute neutrophilic inflammation, both rare in the absence of chronic lymphomononuclear inflammation (4,5,11,12). Recent observations of gastric biopsy specimens containing the bacterium also revealed eosinophil infiltration (13). Human allergic reactions and degranulation manifest similar histological features associated with clinical inflammatory signs (14,151. Cellular recruitment is regarded as a consequence of the release of chemotactic factors from mediator-containing cells, triggered by antigen-induced IgE-dependent mast cell degranulation (16,17). Thus the histological picture of H. pylori-induced gastritis may be associated with and dependent on antigen-specific IgE. At present there is no evidence of systemic allergic manifestations in patients with H. pylori infection. However, the features of an atopic condition have been observed in a large number of patients with chronic gastritis and peptic ulceration (18). Recent studies have also shown that bacteria can induce histamine release from human basophils by an &E-mediated reaction. In addition, it has been suggested that this release could be a pathogenetic mechanism in infectious diseases (19,20). These observations have led us to evaluate the immune response involving IgE in patients with chronic gastritis who are infected by I-I.pylori. Materials and Methods Forty-three adult subjects (22 male and 21 female) were submitted to gastroscopy; the indications for upper endoscopy were mild or moderate gastrointestinal symptoms (dyspepsia or pain), suspected chronic atrophic gastritis, a healed peptic ulcer, or a family history of stomach cancer. Patients with stomach neoplasm or a history of allergic diseases and those recently treated with antibiotics, aspirin or nonsteroidal antiinflammatory drugs, H, blockers, or antihistamine drugs were excluded from the study. A clinical report was completed for each patient.
Endoscopy Esophagogastroduodenoscopy was performed with forward-view Olympus endoscopes (GIF-QlO and GIFXQlO; Olympus, Lake Success, NY), and at least two biopsy specimens were obtained from the antrum. The endoscope and biopsy forceps were disinfected in glutaraldehyde after each use.
Histology Tissue samples were fixed in 10% buffered formalin and embedded in paraffin. Sections were stained with H&E for light microscopy. The criteria used to classify gastritis are those commonly applied (21).
Culture Test A tissue sample from each patient was transported to the laboratory in a sterile container without the addition of preservatives and was processed within 1 hour after collection. The biopsy specimen was cultured on 5% lysed horse blood agar, selective for H. pylori (supplement SR69; Oxoid). The plates were incubated at 37°C under reduced oxygen for 7 days and examined at 3, 5, and 7 days for growth of organisms morphologically resembling Campylobatter. The following features were suggestive of H. pylori: microaerobic growth at 37°C; curved or S-shaped cells; positive reactions for catalase, oxidase, and urease; negative reactions for hippurate hydrolysis and nitrate reduction; susceptibility acid.
to cephalothin;
and resistence
to nalidixic
Antigen Preparation The antigens used were the surface components
of
H. pylori cells harvested with an acid wash from 10 different strains according to the method described by McCoy et al. (22). The composition of this preparation has previously been described in detail (23). Antigens from unrelated bacteria (Campylobacter jejuni and Campylobacter coli) were also prepared according to the same procedure.
Histamine Release Assay After informed consent was obtained, venous blood was drawn by venipuncture at the time of the endoscopy. Mixed leukocytes containing basophils were prepared by dextran sedimentation (24). In brief, 25 mL venous blood was mixed with 6.2 mL 6% clinical dextran and 2.5 mL 0.1 mol/L ethylenediaminetetraacetic acid and allowed to sediment for 60-90 minutes at room temperature. containing plasma layer was then removed sedimented by centrifugation (150 x g for 8 washed twice with PIPES buffer (Sigma, containing human serum albumin and
The leukocyteand cells were minutes at 4”C), St. Louis, MO) glucose, resus-
pended by gently shaking, and then diluted in appropriate volume with the same buffer containing Ca2+ and Mg” for use in the histamine release assay. After preincubation at 37”C, a volume of 0.4 mL mixed leukocytes was added in prewarmed tubes containing 0.2 mL of the different bacterial antigens at various concentrations (100, 50, and 10 kg/mL) and 0.4 mL of the buffer containing Ca’+ and Mg”. To assess the IgE-dependent basophil releasability, the cell preparations were also challenged with goat anti-human IgE (Sigma). The number of leukocytes was chosen so that each tube contained 20-40 ng histamine. The tubes therefore contained approximately 3-5 x lo6leukocytes (2-4 x lo4basophils). Spontaneous release was that observed when cells were added to the buffer alone. After incubation at 37°C for 45 minutes, the tubes were centrifuged (1000 x g for 3 minutes at 23°C) and aliquots of the cell-free supernatant were
IgE TO H. PYLDRI IN CHRONIC GASTRITIS
July 1991
assessed for histamine content by the automated fluorometric assay (25). The results were based on the mean of the duplicate determinations and expressed as percent of histamine release by dividing by total cellular histamine after subtracting for spontaneous release of unstimulated basophils (typically, 2%-5% during the 45-minute assay). Total cellular histamine (100% release) was obtained by lysing cells in 2% perchloric acid. The specimens with a histamine release > 10% were considered positive (26). Bacterial antigens did not contain histamine or interfere with the histamine measurement. No functional IgG was found on the cell surface because the leukocytes did not release histamine when challenged with anti-human IgG. Enzyme-Linked
Zmmunosorben t Assay
Serum IgE specific for the H. pylori antigen was determined by indirect enzyme-linked immunosorbent assay (ELISA) performed using a microtitration procedure (27). In the present study, the optimal antigen concentration was 50 pg/mL and the optimal dilutions for the serum and conjugate (goat anti-human IgE, alkaline phosphataselabeled; Sigma) were so and ~000, respectively, according to checkerboard titration. All tests were performed in duplicate. The cut-off value (optical density above 0.200) was calculated by adding 3 SD to the mean of absorbance of negative subjects. Passive Sensitization
of Normal Basophils
To demonstrate the class specificity of the response, peripheral blood leukocytes (basophils) from a normal donor were sensitized with serum of ELISA &E-positive patients (28) and then exposed to bacterial antigens. The histamine level in the supernatant was determined as described above.
133
The IgE elution from basophils was confirmed by the lack of histamine release following anti-human IgE challenge. Statistical Analysis Paired and unpaired t tests, x2 analysis, the F test, and the correlation coefficient (r) were used for statistical analysis. Results Detection Diagnosis
ofH. pylori
and Histological
Thirty-two of the 43 antral biopsy specimens showed histological evidence of gastritis (19 superficial and 13 atrophic). Normal histological features were found in the remaining 11 samples. H. pylori was isolated by culture in 26 of the 32 patients with gastritis, showing a correlation between the H. pylori and histological gastritis. In contrast, none of the biopsy specimens with normal mucosa contained H. pylori. Detection ofanti-H. Zmmunoglobulin E
pylori
Cell-Bound
The ranges of maximal histamine release values obtained in subjects in whom H. pylori was or was not detected are shown in Figure 1. There is a highly significant difference between these two groups (F test; P < 10m5). Of the 17 individuals not infected by H. pylori, 16 had release values below lo%, and only 1 with gastritis had a 40% release value (mean ‘_ SD,
Inhibition Assay To establish whether the IgE antibody detected using ELISA test was specifically directed against H. pylori antigens, positive sera were incubated in triplicate at 37°C for 60 minutes (one with 50 p,g of H. pylori antigen and the other two with an acid extract of C. jejuni or C. coli). The specimens were then tested by ELISA, performed as described above. Acetate Elution of Basophil Zmmunoglobulin
.
. .
.
.
E
To determine whether histamine release was IgE dependent, aliquots of cells from four patients with H. pylori-induced gastritis were stimulated with the antigen before and after acetate elution of basophil IgE. According to the method described by Ishizaka and Ishizaka (29), after the second wash (see above] the cells were resuspended in 1 mL of acetate buffer, pH 3.7, and allowed to incubate for 10 minutes at 0°C. The cells were sedimented by centrifugation, washed twice with PIPES buffer containing human serum albumin and glucose, and resuspended in the same buffer containing Ca2+ and Mg”.
.
J Figure 1. Histamine release values obtained by challenging with H. pylori antigens basophils from infected (left) and noninfected (right) individuals. Dottedhe shows cutoff limit.
134
ACETI ET AL.
GASTROENTEROLOGY Vol. 101, No. 1
5 -C 9). In contrast, 3 of the infected patients had 100% release values, 6 had values between 95% and SO%, 10 between 50% and 30%, and 3 between 25% and 15%. Thus, when a cutoff value (10%) to discriminate between reactive and unreactive subjects for histamine release test (HRT) was considered, 22 (84.6%) of the 26 patients with H. pylori infection were positive, whereas only 1 (5.8%) positive result was found in the noninfected persons (x” test, P < 10e5; Table 1). Histamine release values obtained from basophils challenged with bacterial antigens other than H. pylori never reached 10%. Release due to anti-human IgE did not differ significantly between infected (mean + SD, 33 + 11) and noninfected individuals (mean * SD, 31 + 12; unpaired t test, P > 0.5). However, a significant correlation was found between the antigen and the anti-IgEinduced histamine release values (r = 0.7; P < 10w3).
Detection ofAn&H. Immunoglobulin E
pylori Serum
Specific IgE serum antibodies were detected by ELISA in 18 (69.2%) patients with H. pylori and in 1 (5.8%) noninfected individual (x” test, P < 10m3; Table 1). This was a true IgE response. In fact, normal basophils passively sensitized with ELISA-positive sera released histamine upon H. pylori antigen exposure in a significant manner (mean 4 SD, 38 r 9). The remaining bacterial antigens were ineffective in inducing basophil degranulation. The inhibition assay showed that the IgE antibody was specifically directed against H. pylori. In fact, as Figure 2 shows, antigens from this bacterium adsorbed out IgE in positive sera (paired t test, P < 10e6), whereas extracts from C. jejuni and C. coli induced only a very slight reduction of optical density values in reactive specimens (paired t test, P > 0.5). All subjects with detectable specific serum IgE had positive HRT results, although no correlation between the percentage of histamine release and the ELISA value (r = 0.14; P > 0.5) was observed.
Table 1, Results Obtained Using Histamine Release Test and an Enzyme-Linked Immunosorbent Assay for Detecting H. pylori-specific Immunoglobulin E in the Subjects Studied Group
rl
HRT-positive subjects
ELISA-positive subjects
Individuals with H. pylori infection Noninfected individuals
26 17
22 (84.6%) 1(5.8%)
18 (69.2%) 1 (5.8%)
2. H. pylori-specific IgE, expressed as optical density values ( x lo), before (left) and after (rig&) adsorption with three dillkent bacteria. Squares indicate median values.
Figure
Acetate Elution of Basophil Immunoglobulin
E
The leukocytes of the 4 patients with H. pylori infection who tested positive upon HRT without detectable specific serum IgE were challenged with the bacterial antigen both before and after acidic treatment. As reported in Table 2, a significant decrease in histamine release was found after IgE elution (paired t test, P < lo-*). Discussion The aim of the present study was to test whether patients with H. pylon’ have a specific demonstrable IgE-mediated immune response. It has previously been shown that bacteria can induce histamine release from human basophil leukocytes and mast cells by an &E-mediated pathway. This mechanism was discovered when basophils from patients with intrinsic asthma were challenged with Haemophilus influenzae and Streptococcus pneumoniae (26). An &E-dependent histamine release was also seen in patients with Escherichia coli bacteremia and fever
IgE TO H. PYLORI IN CHRONIC GASTRITIS
July 1991
Table 2. H. pylori-Induced Histamine Release Values in Four Patients Before and After Immunoglobulin E El&ion Patient A.P. G.F. G.C. V.L.
Before IgE elution (%I
After IgE elution (%I
63 70 83 94
5 6 5 3
and in patients suffering from the immune deficiency syndrome with specific IgE to Staphylococcus aureus isolated from those patients (26,30). On the other hand, a wide spectrum of both gram-positive and gram-negative bacteria can release histamine through a lectin-mediated mechanism and other nonimmunological mechanisms (26). The data of the present study strongly suggest that an immunologic response involving IgE is present in patients infected by H. pylori. Indeed, basophil-bound specific IgE was seen in 84% and serum specific IgE in 69% of the persons with H. pylori infection. In contrast, among the individuals in whom the bacterium was not detected, only one, with histologically proven gastritis, had cell-bound and serum specific IgE. In the subjects who tested positive by HRT without detectable specific circulating IgE, acidelution experiments showed that histamine release from basophils occurred through an immunologic mechanism. To demonstrate further the class specificity of the response, the H. pylori antigen was used to challenge normal basophils passively sensitized with serum from IgE-positive patients. The significant histamine release obtained upon exposure to antigen points to the reliability of the ELISA test and excludes interference with antibodies other than IgE directed against the same antigen. Bacteria other than H. pylori were ineffective in inducing histamine release from basophils of infected patients and in adsorbing out specific serum antibody. These data confirm that the IgE detected were specifically directed against H. pylori antigens. As expected, the percentage of antigen-induced histamine release was not correlated with the serum specific IgE level. It has been previously shown that although a good correlation exists between cell-bound and serum IgE in human basophils, histamine release is not related to the absolute quantity of IgE antibody present on cells challenged with the appropriate antigen (31). In fact, there is evidence that in addition to immunologic parameters, the release of chemical mediators from a target cell also depends on its releasability, i.e., the intrinsic capacity of the cell to release mediators in response to a variety of immuno-
135
logic and nonimmunologic stimuli (32). We have recently shown that infectious factors may play a role in the control of IgE-mediated releasability (33). In the present study, the anti-IgE-induced histamine release did not differ in the infected patients and in the control subjects. However, the basophil response to antigenic challenge was related to IgE-dependent cellular releasability. In fact, a close correlation was present between anti-IgE- and antigen-induced histamine release. This observation may have some clinical importance because it suggests that target cell sensitivity is the major factor determining the entity of the biological response to antigenic challenge. Although much evidence has been accumulated concerning an etiologic role for H. pylori in chronic gastritis, the pathogenetic mechanisms involved in bacterium-associated gastroduodenal inflammation are still unclear. The ability of H. pylori to induce a specific IgE immune response could answer key questions regarding the mechanisms of gastric inflammation. For many years, allergic reactions have been considered acute and short-lived, capable of causing ephemeral clinical problems with few chronic features. There have always been peculiar clinical features that have failed to fit within this concept. Moreover, it became clear in the laboratory assessment of mast cell mediators that this cell could elicit intense, prolonged inflammatory reactions (14,15). Since the early 197Os, there has been increasing clinical and experimental evidence of the capacity of allergic reactions to evolve into chronic inflammatory responses (34). In fact, in a variety of experimental models and disease states, it is clear that an initial immunological trigger, i.e., the antigen and mast cell-bound IgE, leads not only to the rapid release of vasoactive mediators but also to the so-called late-phase reaction, characterized by the involvement of inflammatory cells (neutrophils, eosinophils, mononuclear cells). The infiltrate consisting of polymorphonuclear leukocytes is an early event in such a reaction, followed by a later mononuclear cell infiltration. Cellular recruitment is regarded as a result of the release of mast cell-derived chemotactic factors (16,17). Furthermore, both inflammatory cells and mediators of hypersensitivity contribute to the late sustained response, as well as to the subsequent subacute/chronic inflammatory phase (34). The latephase reaction has not been clearly demonstrated in the gastrointestinal tract, although late-phase reactions may well occur. The histological picture of H. pylori-associated gastritis is quite consistent with that observed in allergic reactions. Indeed, the biopsy specimens containing the bacterium are characterized by the presence of a neutrophilic infiltrate in the context of a diffuse mononuclear cellularity, and eosinophil leuko-
136
ACETI ET AL.
cytes have been observed in association with the bacterium in gastric biopsy samples (4,5,11-13). Thus, it is conceivable that IgE plays a role in mediating H. pylori-induced inflammatory processes in the stomach through the activation of gastric mast cells. The presence of IgE in gastric mucosa is not a new phenomenon, as there is evidence of IgEmediated reactions in patients with erosive gastritis and peptic ulcer (35,36). In addition, IgE directed against cockroach antigens has recently been detected in the serum and gastric mucosa of patients with peptic ulcer and chronic gastritis (18). It is possible that not only histamine, but also other preformed or newly generated mediators released by mast cells, are implicated in tissue damage development. For instance, a potent ulcerogenic action of the plateletactivating factor on the stomach has recently been shown (37). In addition, a direct and/or indirect pathogenetic role for infiltrating cells cannot be excluded. The results of the present study should have therapeutic implications. However, mast cells exhibit morphological and functional heterogeneity both between species and between organs of the same species (38). Some potent pharmacological agents which inhibit the generation, release, or target organ effects of specific mast cell-dependent mediators, have recently been developed. This has permitted the pharmacological dissection of the relative contributions of individual mediators to the pathogenesis of acute and chronic allergic reactions in humans (39). Drugs that inhibit the formation or the target tissue effects of individual mast cell-dependent mediators may exert differing effects from one organ to another, and consequently the assessment of drug activity at one tissue site might not predict its effect at another.
References 1. Salomon H. Ueber das Spirillum des Saugetiermagens und sein Verhalten zu den belegzellen. Zentralbl Bakteriol1896;19:433442. 2. Freedberg AS, Barron LE. The presence of spirochetes in human gastric mucosa. Am J Dig Dis 1940;7:443-445. 3. Palmer ED. Investigation of the gastric mucosa spirochetes of the human. Gastroenterology 1954;27:218-220. 4. Warren JR, Marshall BJ. Unidentified curved bacilli on gastric epithelium in active chronic gastritis. Lancet 1983;1:12731275. 5. Marshall BJ, Warren JR. Unidentified curved bacilli in the stomach of patients with gastritis and peptic ulceration. Lancet 1984;1:1311-1315. 6. Dooley CP, Cohen H. The clinical significance of Campylobacterpylori. Ann Intern Med 1988;108:70-79. 7. Rathbone BJ, Wyatt JI, Worsley BW, Shires SE, Trejdosiewicz LK, Heatley RV, Losowsky MJ. Systemic and local antibody responses to gastric Campylobacter pyloridis in nonulcer dyspepsia. Gut 1986;27:642-647. 8. Booth L, Holdstock G, MacBride H, Hawtin P, Gibson JR,
GASTROENTEROLOGY
Vol. 101, No. 1
Ireland A, Bamforth J, DuBoulay CE, Lloyd RS, Pearson AD. Clinical importance of Campylobacferpyloridis and associated serum IgG and IgA antibody responses in patients undergoing upper gastrointestinal 219. 9. Goodwin CS, Blincow
endoscopy.
J Clin Path01 1986;39:215-
E, Peterson G, Sanderson
C, Cheng W,
Marshall B, Warren JR, McCulloch R. Enzyme-linked immunosorbent assay for Campylobacter pyloridis: correlation with presence of C. pyIoridis in the gastric mucosa. J Infect Dis 1987;155:488-494. 10. Langenberg W, Rauws EA, Widjojokusumo A, Tytgat GN, isolates by Zanen HC. Identification of Campylobacterpyloridis restriction endonuclease DNA analysis. J Clin Microbial 1986; 24:414-417. 11. Andersen LP, Holck S, Povlsen CO, Elseborg L, Justesen T. Campylobacter pyloridis in peptic ulcer disease. I. Gastric and and duodenal infection caused by C. pyloridis: histopathologic microbiologic findings. Stand J Gastroenterol 1987;22:219224. 12. Berstad A, Alexander B, Weberg R, Serck-Hanssen A, Holland
13.
14.
15. 16.
17.
S, Hirschowitz BI. Antacids reduce Campylobacter pylori colonization without healing the gastritis in patients with nonulcer dyspepsia and erosive prepyloric changes. Gastroenterology 1988;95:619-624. McGovern TW, Kephart GM, Gleich GJ, Talley NJ. Eosinophil infiltration and degranulation in Campylobacter pylori gastritis. Klin Wochenschr 1989;67(Suppl 18):46. Gleich GJ. The late phase of the immunoglobulin E-mediated reaction: a link between anaphylaxis and common allergic disease? J Allergy Clin 1mmuno11982;70:160-169. Lemanske RF, Kaliner M. Mast cell-dependent late phase reactions. Clin Immunol Rev 1982:1:547-580. Kay AB, Austen KF. The IgE-mediated release of an eosinophil leukocyte chemotactic factor from human lung. J Immunol 1971;107:899-902. Nagy L, Lee TH, Kay AB. Neutrophil chemotactic activity in antigen-induced late asthmatic reactions. N Engl J Med 1982; 306:497-501.
18, Romamski B, Bartuzi Z, Zbikowska-Gotz M, Staszynska M, Korenkiewicz J. Allergy to cockroach antigens in patients with peptic ulcers and chronic gastritis. Allergol Immunopathol 1988;16:219-224. 19. Norn S, Stahl Skov P, Jensen C, Koch C, Permin H, Bog-Hansen C, Lovenstein H, Hoiby N. Intrinsic asthma and bacterial histaminereleasevialectineffect. Agents Actions 1983;13:210212. 20. Jensen C, Norn S, Stahl Skov P, Espersen F, Koch C, Permin H. Bacterial histamine release by immunological and nonimmunological lectin-mediated reactions. Allergy 1984;39:371377. 21. Whitehead R, Truelove SC, Gear ML. The histological diagnosis of chronic gastritis in fiberoptic gastroscope biopsy specimens. J Clin Path01 1972;25:1-11. 22. McCoy EC, Doyle D, Burda L, Corbeil LB, Winter AJ. Superficial antigens of Campylobacter (Vibrio] fetus: characterization of an antiphagocitic component. Infect Immun 1975;11:517-525. 23. Aceti A, Attanasio R, Pennica A, Taliani G, Sebastiani A, Rezza G, Ippolito G, Perucci CA. Campylobacter pylori infection in homosexuals. Lancet 1987;2:154-155. LM, Osler AG. Studies on the mechanisms of 24 Lichtenstein hypersensitivity phenomena. IX. Histamine release from human leukocytes by ragweed pollen antigens. J Exp Med 1964; 120:507-530. 25, Siraganian RP. Automated histamine release. A method for in vitro allergy diagnosis. Int Arch Allergy Appl Immunol1975;49: 108-110. 26. Norn S, Stahl Skov P, Jensen C, Jarlov JO, Espersen F.
July 1991
27.
28.
29.
30.
31.
32. 33.
34.
A new mechanism in Histamine release induced by bacteria. asthma? Agents Actions 1987;20:29-34. Voller A, Draper CC, Bidwell DE, Bartlett A. Microplate enzyme-linked immunosorbent assay for Chagas’ disease. Lancet 1975;1:426-429. Aceti A, Celestino D, Caferro M, Pennica A, Sebastiani A. Total and specific IgE in sera from patients with invasive amoebiasis. Trop Geogr Med 1989;41:151-153. Ishizaka T, Ishizaka K. Mechanisms of passive sensitization. IV. Dissociation of IgE molecules from basophil receptors at acid pH. J 1mmuno11974;112:1078-1084. Norn S, Jensen C, Dahl BT, Stahl Skov P, Baek L, Permin H, Jarlow JD, Sorensen H. Endotoxins release histamine by complement activation and potentiate bacteria-induced histamine release. Agents Actions 1986;18:149-152. Conroy MC, Adkinson FN, Lichtenstein LM. Measurement of IgE on human basophils: relation to serum IgE and anti-IgEinduced histamine release. J Immunol1977;118:1317-1321. Lichtenstein LM, MacGlashan DW. The concept of basophil releasability. J Allergy Clin Immunol 1986;77:291-294. Aceti A, Celestino D, Caferro M, Teggi A, Pennica A, Adriani E, Grilli A, Paparo S, Leri 0, De Rosa F, Sebastiani A. Basophil releasability in human hydatidosis. Int Arch Allergy Appl 1mmun011990;91:111-112. Kaliner M, Lemanske R. Mast cell-derived inflammatory factors and late-phase allergic reactions. In: Kay AB, Austen KF, Lichtenstein LM, eds. Asthma: physiology, immunopharmacology, and treatment. London: Academic, 1984:229-244.
IgE TO
H. PYLORIIN CHRONIC
C, Moulinier 35. Andre anaphylactic reaction
B, Andre in peptic
GASTRITIS
F, Daniere ulcer
137
S. Evidence
and varioliform
for
gastritis.
Ann Allergy 1983;51:325-331. 36. Andre F, Andre C, Cavagna S. Role of histamine in the cell turnover changes associated with experimental gastric ulceration in the mastomys. Gastroenterology 1985;88:452-457. 37. Rosam
A-C, Wallace
JL, Whittle
tions of platelet-activating 1986;318:54-56. 38. Pearce J, Ennis
FL, Ali H, Barrett M, Flint
BJR. Potent
factor
on
the
ulcerogenic stomach.
KE, Befus AD, Bienenstock
KC, Hudspith
B, Johnson
J, Brostoff
NM, Leung
Peachell PT. Functional characteristics of mucosal tive tissue mast cells of man, the rat and other
ac-
Nature
KBP,
and connecanimals. Int
Arch Allergy Appl Immunol 1985;77:274-279. 39. Holgate ST, Rafferty P, Beasley R, Robinson C, Hovel1 CJ, Curzon NP, Church MK. In vitro and in viva studies on mast cells of human skin and airways. In: Kay AB, ed. Allergy and inflammation. London: Academic. 1987:29-52.
Received July 31, 1989. Accepted December 3,199O. Address requests for reprints to: Antonio Aceti, M.D., Clinica delle Malattie Tropicali ed Infettive, Policlinico Umberto I’, 00161 Rome, Italy. This work was supported della Pubblica Istruzione.
in part by grants
from the Minister0
