Cellular Immune Responses Are Essential for the Development of felis -Associated Gastric Pathology

This information is current as Kevin A. Roth, Sharookh B. Kapadia, Steven M. Martin and of September 26, 2021. Robin G. Lorenz J Immunol 1999; 163:1490-1497; ; http://www.jimmunol.org/content/163/3/1490 Downloaded from

References This article cites 47 articles, 14 of which you can access for free at: http://www.jimmunol.org/content/163/3/1490.full#ref-list-1

Why The JI? Submit online. http://www.jimmunol.org/

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average by guest on September 26, 2021

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 1999 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Cellular Immune Responses Are Essential for the Development of -Associated Gastric Pathology1

Kevin A. Roth,*† Sharookh B. Kapadia,*‡ Steven M. Martin,*‡ and Robin G. Lorenz2*‡§

The is a major human that infects over half of the world’s population. initiates a series of changes in the gastric mucosa, beginning with atrophic and leading in some patients to , mucosa-associated lymphomas, and gastric adenocarcinoma. Although this cascade of events clearly occurs, little is known about the role of the host immune response in disease progression. We have utilized the C57BL/6 Helicobacter felis mouse model to critically analyze the role of the adaptive immune response in the development of Helicobacter-associated gastric pathology. Infection of B and T cell-deficient RAG-1؊/؊ mice or T cell-deficient TCR␤␦؊/؊ mice with H. felis resulted in high levels of colonization, but no detectable gastric pathology. Conversely, infection of B cell-deficient ␮MT mice resulted in severe gastric alterations identical with those seen in immunocompetent C57BL/6-infected mice, including gastric mucosal hyperplasia and intestinal metaplasia. These results demonstrate that the host T cell response is a critical mediator of Helicobacter-associated Downloaded from gastric pathology, and that B cells and their secreted Abs are not the effectors of the immune-mediated gastric pathology seen after H. felis infection. These results indicate that in addition to specific Helicobacter factors, the host immune response is an important determinant of Helicobacter-associated disease. The Journal of Immunology, 1999, 163: 1490–1497.

t has been 14 years since Helicobacter pylori was conclu- One potential host factor is the adaptive immune response, as

sively shown to cause acute gastritis (1, 2). Patients infected the gastric inflammatory infiltrate associated with Helicobacter in- http://www.jimmunol.org/ I with H. pylori develop a chronic gastric inflammatory infil- fection is predominantly IgG- and IgA-producing B cells and CD4 trate that primarily affects the antral region. The majority of in- and CD8-positive T cells (9, 10). Initial studies in immunodeficient fections result in no clinical symptoms; however, 10–20% (C.B-17 SCID) mice indicated that the adaptive immune response progress to serious clinical disease. Multiple studies have demon- was not required to maintain chronic gastric inflammation after strated that H. pylori is associated with the development of duo- Helicobacter infection (11). However, recent reports have now denal and gastric ulcer disease, gastric adenocarcinoma, and mu- shown that several mouse strains, including the strain used above, cosa-associated lymphoid tissue (MALT) lymphoma (3–5). The do not develop gastric pathology after Helicobacter infection. Spe- critical factors, either host, bacterial, and/or environmental, that cifically, the inbred mouse strain C57BL/6 develops a severe by guest on September 26, 2021 influence the clinical manifestations of Helicobacter infection are chronic active gastritis and intestinal metaplasia, whereas BALB/c still being elucidated. Some of the observed clinical variation in mouse strains exhibit only mild gastritis after infection with the H. the pathologic response to H. pylori is due to phenotypic and ge- pylori strain SS1 (Cag-Aϩ) and the closely related Helicobacter notypic diversity of the H. pylori bacteria itself, as strains of H. felis (Cag-AϪ) (12–15). Due to the availability of this new pylori with the cytotoxin-associated gene (cag)3 pathogenicity is- C57BL/6 mouse model of Helicobacter-associated gastric disease, land have been associated with more severe peptic ulcer disease we reevaluated the hypothesis that the host adaptive immune re- (6). However, this correlation does not account for all the mani- sponse initiated by Helicobacter infection is responsible for sub- festations of Helicobacter-induced disease, as MALT lymphomas sequent gastric pathology. are not associated with expression of the cag pathogenicity island, We established gastric infection with H. felis in C57BL/6 wild- and in Far Eastern countries no association between disease out- type and induced mutant mice. H. felis infection in C57BL/6 mice comes and cytotoxin genes has been demonstrated (7, 8). These mimics human disease in many aspects, including a predominant observations suggest that additional factors must interact with H. CD4ϩ T cell response (9, 16) that is associated with an increased pylori to influence the course of gastric disease. epithelial cell proliferative index and hyperplasia (12, 13, 17). However, H. felis does not express the cag pathogenicity island or Departments of *Pathology, †Molecular Biology and Pharmacology, and §Medicine, the vacuolating toxin (vacA), and thus does not itself damage the and ‡Center for Immunology, Washington University School of Medicine, St. Louis, gastric epithelium (18). Induced mutant mouse strains deficient in MO 63110 both B and T cells, or in B cells or T cells alone were infected with Received for publication December 3, 1998. Accepted for publication April 15, 1999. H. felis, and the subsequent gastric pathology was evaluated. In The costs of publication of this article were defrayed in part by the payment of page this report, we demonstrate that the T cell response is crucial for charges. This article must therefore be hereby marked advertisement in accordance Helicobacter-induced gastric pathology, and that post-Helicobac- with 18 U.S.C. Section 1734 solely to indicate this fact. ter infection gastric destruction and hyperproliferation are not me- 1 This study was supported in part by the Charles E. Culpeper Foundation and an Institutional Howard Hughes Medical Institute Pilot Project Award. S.B.K. is sup- diated by B cells or their secreted products. ported by the Division of Biology and Biomedical Sciences, and S.M.M. is supported by a National Institutes of Health Training Grant T32 AI07163. R.G.L. is a Charles E. Culpeper Foundation Scholar. Materials and Methods 2 Address correspondence and reprint requests to: Dr. Robin G. Lorenz, Department Mice of Pathology (Box 8118), Washington University School of Medicine, 660 South Euclid Avenue, St. Louis, MO 63110. E-mail address: [email protected] C57BL/6J-Rag-1tm1Mom, C57BL/6-Igh-6tm1Cgn, and C57BL/6J-Tcrbtm1Mom tm1Mom 3 Abbreviations used in this paper: cag, cytotoxin associated gene; BrdUrd, 5-bromo- Tcrd breeder pairs were purchased from The Jackson Laboratory 2Ј-deoxyuridine; RAG, recombinase activating gene. (Bar Harbor, ME). The strains were maintained by mating homozygous

Copyright © 1999 by The American Association of Immunologists 0022-1767/99/$02.00 The Journal of Immunology 1491 siblings. C57BL/6J control mice were purchased from The Jackson Lab- oratory. Mice were housed in a specific pathogen-free facility in microiso- lator cages under a strictly controlled light cycle and given a standard autoclaved chow diet ad libitum. All mice were maintained in accordance with the guidelines of the Animal Studies Committee of Washington University.

Generation of H. felis-induced gastritis H. felis (ATCC 49179; American Type Culture Collection, Manassas, VA) was grown on moist ATCC 260 medium (trypicase soy agar (BBL 11043; Fisher Scientific, Pittsburgh, PA)) supplemented with defibrinated calf blood (5% v/v; Colorado Serum Company, Denver, CO), trimethoprim (5 mg/L; Sigma, St. Louis, MO), vancomycin (6 mg/L; Sigma), and Fun- gizome (1% v/v; Life Technologies, Gaithersburg, MD) under microaero- philic conditions at 37°C for 2 days. Confluent plates of H. felis were harvested, and the number of bacteria were determined by adsorption at 9 OD450, with one OD unit corresponding to 10 bacteria. Bacteria were stored in a mixture of brain heart infusion broth (BHI; Difco 0037; Fisher) and glycerol (31 ml glycerol/100 ml broth) at Ϫ70°C. Per os (p.o.) infec- tion of H. felis was accomplished using a 200-␮l pipette tip. Mice were inoculated twice over a 3-day period with 0.05 ml of bacterial suspension.

Mock infected animals were inoculated with the BHI/glycerol suspension Downloaded from medium. FIGURE 1. Levels of colonization of H. felis in infected C57BL/6, Assessment of H. felis colonization and histopathology Ϫ Ϫ Ϫ Ϫ ␮MT, RAG-1 / , and TCR␤␦ / mouse strains. Colonization is repre- Mice were sacrificed and their stomachs rapidly removed. Ninety minutes sented by the H. felis (HF) colonization score, which was based on eval- before sacrifice, animals received an i.p. injection of 5-bromo-2Ј-deoxyuri- uation of immunohistochemical staining of the H. felis organisms as de- dine (BrdUrd; Sigma) (dose, 120 mg/kg) to label cells in S phase (19).

scribed in Materials and Methods. Only mouse strains with functional http://www.jimmunol.org/ Tissue samples were fixed in methacarn solution (60% methanol, 30% T-lymphocytes were able to control the level of H. felis colonization. chloroform, 10% glacial acetic acid) and embedded in paraffin, and 5-␮m- thick sections were prepared. Colonization was established by evaluation of immunohistochemical staining of the H. felis organisms in histologic sections of the entire greater and lesser curvatures of the . A semi- quantitative determination (H. felis (HF) colonization score) of the organ- ism load was made based on the number of bacteria observed per gastric kindly supplied by Dr. Steven Cohn, University of Virginia, Charlottes- glandular unit (0 ϭ no bacteria, 1 ϭ 1–2 bacteria, 2 ϭ 3–10 bacteria, 3 ϭ ville, VA) (21, 22); rabbit anti-rat intrinsic factor (1:1000) (27); rabbit ϭϾ anti-rat pepsinogen (1:500, kindly supplied by Michael Samloff, University 11–20 bacteria, 4 20 bacteria). Sections were evaluated by two pa- ϩ ϩ thologists blinded to the experimental status of the mice. of California, Los Angeles, CA) rabbit anti-H /K -ATPase ␤-subunit (1:

Gastrointestinal inflammation and destruction were assessed by micro- 250, kindly supplied by Michael Caplan, Yale University, New Haven, by guest on September 26, 2021 scopic examination of hematoxylin/eosin-stained gastric tissue. The stom- CT); rabbit anti-H. pylori (1:1000, Dako, Carpinteria, CA); Griffonia sim- achs were transected along the greater and lesser curvature to obtain two plifolica II lectin (specific for mucous neck cells; final concentration, 5 halves, with each half subsequently embedded in paraffin. Gastric pathol- ␮g/ml, EY Laboratories, San Mateo, CA) (24); Dolichos biflorus aggluti- ogy was evaluated by two pathologists blinded to the strain and experi- nin (specific for parietal cells; final concentration. 20 ␮g/ml; Sigma) (24); mental status of the mice. The scoring system used was modified from toxin B-subunit (specific for surface mucous (pit) cells; final con- Ermak et al. (20) and is based on a grade of 0–3 in each of three categories: centration, 5 ␮g/ml; List Biological Laboratories, Campbell, CA) (24). longitudinal extent of inflammation (0 ϭ no inflammation, 1 ϭ patchy, 2 ϭ Ͻ50%, 3 ϭϾ50%); vertical extent of inflammation (0 ϭ no inflammation, Stastical analysis 1 ϭ basal lamina propria only, 2 ϭ transmural, 3 ϭ both mucosa and submucosa involvement); and histopathological changes (0 ϭ none, 1 ϭ Data analysis using a standard Student’s t test or Mann-Whitney test was mild alterations in differentiated epithelial cells, 2 ϭ moderate alterations performed using GraphPad Prism (San Diego, CA). in differentiated epithelial cells, 3 ϭ severe alterations in differentiated epithelial cells). The total histological score (0–9) was determined by sum- Results mation of each subscore. All zones of the stomach were examined: squa- mous zone, zymogenic zone (parietal cells, zymogenic cells, surface and Adaptive immunity is critical for Helicobacter-associated gastric neck mucous cells), mucoparietal zone (parietal cells, surface and neck pathology mucous cells), and mucous zone (surface and neck mucous cells). To assess the contribution of the adaptive immune response to the gastric epithelial pathology seen after Helicobacter infection, Immunohistochemistry Ϫ Ϫ C57BL/6 and C57BL/6J-Rag1tm1Mom (RAG-1 / ) were infected To assess the proliferative state of the , as well as the p.o. with H. felis. RAG-1Ϫ/Ϫ mice have no mature B or T lym- infiltrating leukocytes, a polyclonal Ab against BrdUrd was used (21, 22). phocytes and consequently do not mount any Ag-specific immune Immunohistochemical methods were also used on stomach sections to as- sess parietal and zymogenic cell density, and inflammation (23). Specifi- responses (28). Infected mice and control mock-infected mice cally, a panel of lectins and Abs was used as previously described to define were sacrificed 4, 6, 8, 14, and 22 wk after inoculation (14 total Ϫ Ϫ alterations in the proliferation and differentiation programs of gastric epi- C57BL/6 and 15 RAG-1 / ). After 4 wk of infection, the RAG- thelial cell lineages (23–25). Briefly, 5-␮m-thick sections were deparaf- 1Ϫ/Ϫ mice had similar levels of Helicobacter organisms in their finized, nonspecific staining blocked using PBS containing 1% BSA, 0.2% gastric glands as C57BL/6, with colonization initially occurring in powdered skim milk, and 0.3% Triton X-100 (PBS-BB), and then incu- bated with the appropriate primary Ab diluted in PBS-BB overnight at 4°C. the pure mucous zone (corresponding to the antral region of the After extensive PBS washes, cyanine (Cy)-3 or fluorescein (FITC)-conju- human stomach) (Fig. 1). However, at all time points examined, gated secondary Abs and/or fluorescent-conjugated lectins were added, in- they displayed minimal immune infiltrates, no alterations in their cubated on the sections for1hatroom temperature, washed and cover- gastric morphology (Fig. 2, A–C) and no proliferative abnormal- slipped with PBS/glycerol (1:1), and examined using a Zeiss (Oberkochen, Germany) Axioskop fluorescent microscope. The methods used for single- ities (data not shown). This was in contrast to the C57BL/6 im- and double-label immunohistochemistry have been described previously munocompetent mice, which as early as 4 wk after infection dis- (26). Abs and lectins used were: goat anti-BrdUrd (final dilution, 1:1000; played gastric infiltrates consisting primarily of mononuclear cells 1492 IMMUNE RESPONSE AND HELICOBACTER-ASSOCIATED PATHOLOGY Downloaded from http://www.jimmunol.org/ by guest on September 26, 2021 The Journal of Immunology 1493

Table I. Total histology score and histopathological changes subscore C57BL/6 mice with an isolated deficiency in the B cell compart- in mice infected 8 or 14 wk with H. felis ment caused by a targeted disruption in the heavy chain of IgM (C57BL/6-Igh-6tm1Cgn, ␮MT) (35). This disruption results in B Histopathological cell maturation arrest at the pre-B cell stage and the absence of Weeks After Total Histology Changes Infection Scorea Subscorea mature B cells. The pathology seen after H. felis infection of B cell-deficient mice was identical with that seen in wild-type C57BL/6 mock 8 0 0 C57BL/6 (Fig. 2, E, H, and I). As early as 4 wk after inoculation ء Ϯ ء Ϯ C57BL/6 H. felis 8 7.5 0.3 1.75 0.3 with H. felis, a severe inflammatory infiltrate, consisting of both ␮MT mock 8 0 0 -monocytes and neutrophils, was seen. This infiltrate was accom ءϮ 0.3 2.3 ء␮MT H. felis 8 8.3 Ϯ 0.3 TCR␤␦Ϫ/Ϫ mock 8 1.3 Ϯ 0.6 0.3 Ϯ 0.2 panied by dramatic epithelial changes, including a marked reduc- Ϫ Ϫ Ϯ 0.2 tion in zymogenic and parietal cells (Fig. 3, A and D, and data not 0.8 ءTCR␤␦ / H. felis 8 3.5 Ϯ 0.4 shown). The differentiated gastric epithelial cells were replaced by C57BL/6 mock 14 0 0 a hyperplastic epithelium, which included large areas of intestinal ءϮ 0.3 2.8 ءC57BL/6 H. felis 14 7.5 Ϯ 0.8 RAG-1Ϫ/Ϫ mock 14 0.5 Ϯ 0.5 0 metaplasia containing mucin-producing goblet cells (Fig. 3E). RAG-1Ϫ/Ϫ H. felis 14 0.5 Ϯ 0.5 0 These epithelial cell alterations were seen in both the C57BL/6 and ␮ a Total histology score (0–9) is based on microscopic evaluation of gastrointes- MT mouse strains and were maintained throughout the length of tinal inflammation and destruction as described in detail in Materials and Methods. infection (20 wk, a total of 32 ␮MT and 29 C57BL/6 mice were The histopathological changes subscore (0–3) is the specific component of the total examined). Both mouse strains appeared to drastically reduce their histology score in which only alterations in differentiated epithelial cells (parietal cells, zymogenic cells, mucous neck cells) are evaluated. levels of H. felis colonization by 12 wk after inoculation (Fig. 1), Downloaded from p Ͻ 0.05, as compared to the respective mock infected mice. demonstrating that B cells and/or Abs are not essential for this ,ء reduction in bacterial colonization levels. H. felis infection of TCR (TCR␤␦) double knockout mice (n ϭ and scattered neutrophils. This infiltrate was accompanied by dra- 28; C57BL/6J-Tcrbtm1Mom Tcrdtm1Mom) (36–38) did not result in matic gastric epithelial changes, including a marked reduction in any significant histological changes when compared with mock- the number of parietal and zymogenic cells in the gastric epithe- Ϫ/Ϫ

infected control TCR␤␦ mice. Mice examined 4–12 wk after http://www.jimmunol.org/ lium by 14 wk after inoculation with H. felis (Fig. 2, C, D, and F). H. felis inoculation had high levels of bacterial colonization (Fig. These severe alterations in differentiated gastric epithelial cells are 1), but demonstrated minimal gastritis and no significant gastric reflected by the high histopathological changes subscore as seen in epithelial pathology (Fig. 4 and Table I). It should be noted that Table I. The differentiated gastric epithelial cells appear to be re- Ϫ Ϫ mock-infected control TCR␤␦ / have a low, but consistent level placed by two cell types. One is a rapidly proliferating cell that of gastric inflammation and histological changes. As these TCR does not display any characteristics of gastric epithelial cell dif- mutant mice have been previously described to develop spontane- ferentiation (see Fig. 5B). The second is a mucous-producing gob- ous inflammatory bowel disease, this low level gastric inflamma- let-like cell that results in a histologic picture characteristic of tion is not surprising (37). Although H. felis infection significantly

intestinal metaplasia (see Fig. 3, E and F). The absence of gastric Ϫ Ϫ by guest on September 26, 2021 increases the total histology score in the TCR␤␦ / mice; this pathology in H. felis infected RAG-1Ϫ/Ϫ mice was consistently elevated score is due to an increase in the level of inflammatory seen at all time points examined, even though all RAG-1Ϫ/Ϫ an- infiltrate, as the histopathological changes subscore is not signif- imals maintained high levels of H. felis colonization (Fig. 1). icantly different (Table I). These data indicate that T cells and/or These findings indicate that in the C57BL/6 inbred mouse strain, their secreted products are the crucial mediators of the gastric pa- the host adaptive immune response generated after infection with thology seen after Helicobacter infection. Helicobacter was causally responsible for the subsequent gastric pathology. Gastric epithelial cell proliferation was increased in response to T cells, not B cells, are the crucial mediators of H. felis- H. felis infection associated gastric pathology Intestinal-type gastric cancer develops over decades from superfi- Recent reports have demonstrated that infection with H. pylori can cial to chronic gastritis, followed by the development of atrophy, lead to the production of anti-H. pylori Abs that cross-react with intestinal metaplasia, dysplasia, and finally cancer (39). We have human gastric mucosal Ags (29–34). These autoantibodies have now demonstrated that the initial stages of this pathway—gastritis the potential to play a crucial role in the pathogenesis of gastric and atrophy (defined as the loss of parietal and chief cell popula- diseases, as Negrini et al. (29) demonstrated that mice bearing tions)—are a result of the T cell response to Helicobacter infec- hybridomas secreting a H. pylori cross-reactive mAb caused gas- tion. As gastritis and atrophy are associated clinically with an in- tric abnormalities similar to gastritis. To analyze the role of B cells crease in mucosal cell proliferation (40, 41), we wanted to directly and their secreted products (cytokines and Abs) in the develop- assess the levels of gastric epithelial cell proliferation in our mouse ment of Helicobacter-associated gastric pathology, we used model of Helicobacter gastritis. The mice were injected i.p. with

FIGURE 2. RAG-1Ϫ/Ϫ, C57BL/6, and ␮MT gastric glands after infection with H. felis. Shown are RAG-1Ϫ/Ϫ (A and B) and C57BL/6 (D and F) mice 14 wk after infection with H. felis (B and F) or suspension medium only (mock-infection, A and D). Rag-1Ϫ/Ϫ mice exhibit no gastric pathology. Note the precisely organized gastric zymogenic glands and the lack of cellular immune infiltrate. Sections were stained with hematoxylin/eosin (bar ϭ 25 ␮m). This lack of pathology was confirmed with specific immunostaining for differentiated gastric epithelial cells (data not shown). This normal epithelium is in distinct contrast to the C57BL/6 gastric glands, which demonstrate a loss of differentiated parietal and zymogenic cells, with replacement by mucous- producing and undifferentiated epithelial cells. Shown are C57BL/6 (E) and ␮MT (G and H) mice 8 wk after infection with H. felis (E and H)or suspension-media only (mock-infection, G). The pathological alterations seen after H. felis infection of B cell-deficient mice are identical with those seen in wild-type C57BL/6 mice. Rag-1Ϫ/Ϫ mice demonstrate no gastric infiltration or pathology after H. felis infection as measured by a composite histology scoring system (C; Materials and Methods). This is in contrast to C57BL/6 and ␮MT mice, which exhibit severe gastric pathology as measured by histologic score (I). 1494 IMMUNE RESPONSE AND HELICOBACTER-ASSOCIATED PATHOLOGY Downloaded from http://www.jimmunol.org/ by guest on September 26, 2021

FIGURE 3. Hϩ/Kϩ-ATPase positive parietal cells (A, red) and Griffonia simplifolica II (GSII) lectin positive mucous neck cells (B, green) are present in mock-infected ␮MT gastric zymogenic glands. C, Both cell types in a double-labeled section. D–F, Note the loss of parietal cells (D, red) at 8 wk after infection of ␮MT mice, and the dramatic expansion of GSII-positive mucous cells (E, green). F, Both cell types in a double-labeled section (bar ϭ 50 ␮m).

BrdUrd (120 mg/kg) 90 min before sacrifice. This thymidine an- alogue is incorporated into the DNA of dividing cells. Cells in S phase during the injection period can subsequently be immunohis- tochemically detected in tissue sections using an Ab specific for BrdUrd (19, 21). In normal murine stomach, BrdUrd labeling was limited to a narrow band of positive cells in the previously well described isthmus progenitor cell zone (42). In contrast, H. felis- infected C57BL/6 stomachs demonstrated a dramatic increase in the number of BrdUrd-positive cells and in the width of the pro- liferative zone (Fig. 5, A and B, and Table II). The marked pro- liferative abnormalities were seen as early as 6 wk after infection in C57BL/6 mice and continued throughout the length of our ex- periment (20 wk). Note, that after 12 wk, there were very few detectable Helicobacter organisms in the stomach (Fig. 1); thus, direct effects of H. felis itself on gastric proliferation is unlikely. FIGURE 4. TCR␤␦Ϫ/Ϫ gastric glands show no significant alterations Epithelial cell hyperplasia was not dependent on the presence of B Ϫ Ϫ after infection with H. felis. TCR␤␦ / mice 8 wk after infection with H. cells, as a marked expansion of the number of gastric epithelial felis (A) show slightly increased levels of gastric infiltration when com- cells in S phase as well as the size of the proliferative zone was pared with TCR␤␦Ϫ/Ϫ mice 8 wk after inoculation with suspension me- observed in infected B cell-deficient ␮MT mice (Fig. 5, C and D, dium only (mock infection, B). However, only slight alterations in gastric epithelial cell differentiation are noted, and these changes are not signifi- and Table II). In contrast, there was no proliferative abnormality Ϫ/Ϫ cantly different between infected and noninfected TCR␤␦Ϫ/Ϫ gastric seen at any time after infection in the RAG-1 mouse strain or glands (see also Table I). in uninfected control mice (data not shown). The Journal of Immunology 1495 Downloaded from

FIGURE 5. Immunostaining of BrdUrd labeled DNA in the gastric glands of C57BL/6 (B) and ␮MT (D) mice infected 8 wk earlier with H. felis,

compared with control mock-infected http://www.jimmunol.org/ C57BL/6 (A)or␮MT (C) mice (bar ϭ 50 ␮m). Note the greatly increased proliferation in the infected C57BL/6 and ␮MT mice. by guest on September 26, 2021

Discussion gastric atrophy, metaplasia, and cancer, this is the first demonstra- tion of the crucial role of the T cell response in this progression. In this study, we used the induced mutant mouse strains RAG- Ϫ Ϫ Ϫ Ϫ The hyperplastic epithelium and altered glandular differentiation 1 / , TCR␤␦ / , and ␮MT to evaluate the contribution of the adaptive immune response to the gastric epithelial hyperprolifera- seen in the C57BL/6 mouse are all features of preneoplastic lesions tion, altered glandular differentiation, and intestinal metaplasia seen in the recently described Mongolian Gerbil model of Helico- seen after Helicobacter infection. C57BL/6 mice demonstrated se- bacter-induced gastric cancer (43). vere chronic active gastritis with increased epithelial cell prolifer- These results confirm and extend our previous report that tar- ation and loss of specialized parietal and zymogenic cells after geted microbial attachment in a novel transgenic mouse model of infection with H. felis. Strikingly, C57BL/6 mice with either no H. pylori infection results in increased gastric infiltration and gas- mature B or T cells (RAG-1Ϫ/Ϫ) or specifically no mature T cells tric pathology (44). These data also serve as a potential explanation (TCR␤␦Ϫ/Ϫ) do not develop gastric epithelial pathology after H. for the observation of an unexpectedly low prevalence of H. py- felis infection. Although immune responses have been postulated lori-associated gastric histopathology in immunocompromised pa- to be involved in the progression from Helicobacter infection to tients with AIDS (45). However, our findings are in contrast with 1496 IMMUNE RESPONSE AND HELICOBACTER-ASSOCIATED PATHOLOGY

Table II. Proliferative abnormalities and hyperplasia in mice two major potential mechanisms of T cell-dependent gastric pa- infected Ն 8 wk with H. felis thology. One mechanism would be based on the postulate that specific immune recognition events are critical for the resulting BrdUrdϩ Cellsa Mucosal Thicknessb gastric epithelial pathology. This Ag-specific event, where a T cell C57BL/6 mock 74.5 Ϯ 10.2 409 Ϯ 38 with a TCR specific for gastric self-proteins could be stimulated to develop by molecular mimicry of a H. felis protein, or by exposure ءϮ 41 654 ءC57BL/6 H. felis 182.2 Ϯ 23.2 ␮MT mock 54.0 Ϯ 11.9 318 Ϯ 16 to a previously sequestered gastric Ag, could result in the devel- † † ␮MT H. felis 125.2 Ϯ 15.8 670 Ϯ 58 opment of a gastric autoimmune disease. A second potential mech- a Reported as the mean number of BrdUrd positive cells Ϯ SEM per ϫ 20 mag- anism is based on an indirect effect of T cells and the factors they nification field at a distance of 1 mm from the forestomach-glandular epithelial junc- secrete in response to a gastric infection. In this scenario, the T cell tion. Measurements were done by an observer blinded to the experimental protocol. b Measurements shown represent the mean thickness Ϯ SEM in microns. The response to H. felis infection results in a cytokine/chemokine mi- gastric mucosal thickness was measured 1 mm from the forestomach-glandular epi- lieu that modifies the differentiation of gastric epithelial cells from thelial junction. Measurements were done by an observer blinded to the experimental the gastric epithelial stem cell. This altered milieu could be due protocol. p Ͻ 0.05, as compared with mock infected C57BL/6. There were no significant either to the infiltration of H. felis-specific T cells or to the influx ,ء differences when compared with infected ␮MT. of a polyclonal T cell population into the gastric epithelium. This Ͻ ␮ †, p 0.05, as compared with mock infected MT. modified gastric epithelial cell differentiation program would result in an increase in proliferating undifferentiated epithelial cells and mucus-producing intestinal-like epithelial cells, and a decrease in a previously reported H. felis SCID model, which concluded that zymogenic/parietal cells. As there is currently a very limited Downloaded from the host adaptive immune response played a limited role in control knowledge of the factors involved in gastric epithelial differenti- of Helicobacter colonization (11). In the SCID study, immunode- ation from the gastric stem cell, it is unclear whether lymphocyte- ficient (SCID) mice and immunocompetent C.B-17 congenic mice derived factors could have this effect on gastric epithelial cell (identical with the BALB/c strain except at the Ig heavy-chain differentiation. locus (Igh)) (46) were both infected with H. felis. No statistical There is now evidence in human H. pylori infection of an altered

difference between the two mouse strains in the number of H. felis gastric epithelial cell turnover early in the course of the disease http://www.jimmunol.org/ organisms or the intensity of inflammation was seen; however, this (49–51). Both increased epithelial cell apoptosis and proliferation study did not evaluate Helicobacter-associated gastric pathology. have been described. The lack of epithelial cell hyperproliferation We and others have now shown that the BALB/c inbred mouse in the RAG-1Ϫ/Ϫ-infected mice indicates that in the absence of the strain does not develop extensive gastric pathology after Helico- adaptive immune response, factors are missing that crucially affect bacter infection (histological score ϭ 2.05 Ϯ 0.43, 4–28 wk after epithelial cell proliferation. These data suggest that future inves- infection, n ϭ 10), whereas the C57BL/6 mouse strain develops tigations of the infiltrating adaptive immune response will further hyperplasia and parietal and zymogenic cell destruction (see Fig. our knowledge of the mechanisms by which Helicobacter infec- 2) (12, 13). Therefore, to stringently test our hypothesis that the tion results in hyperplastic changes in the gastric epithelium. host adaptive immune response initiated by Helicobacter infection In summary, our studies have demonstrated that the host T cell by guest on September 26, 2021 is causally responsible for the subsequent gastric pathology, it was response initiated by Helicobacter infection is causally responsible necessary to infect immunodeficient mice that were of the for subsequent gastric pathology. Future studies of the interplay C57BL/6 background. The different interpretation of our results between the host cellular immune response and Helicobacter in- and the previously published data is almost certainly due to our fection should provide insights into the key immune mediators ability to detect and quantitate gastric mucosal abnormalities in the responsible for the Helicobacter-associated gastric pathology. C57BL/6 strain, as these abnormalities do not occur in the C.B- 17/BALB/c inbred strain. Acknowledgments It has been clearly shown that human infection with H. pylori induces autoantibodies reactive with gastric parietal cell Ags, spe- We thank Jacquie McDonough and Kimberly Goodwin for expert technical cifically to Hϩ/Kϩ-ATPase, which is also known to be a key au- assistance and Rodney Newberry for helpful discussions. toantigen in autoimmune gastritis (34, 47). It is unknown whether these autoantibodies are directly pathogenic for the gastric epithe- References lium or are in response to gastric epithelial cell destruction caused 1. Marshall, B. J., J. A. Armstrong, D. B. McGechie, and R. J. Glancy. 1985. by Helicobacter virulence factors such as cytotoxins (48). The Attempt to fulfil Koch’s postulates for pyloric . Med. J. Aust. significance of autoantibodies to Helicobacter-induced gastric pa- 142:436. thology has proven difficult to experimentally address in human 2. Marshall, B. J., D. B. McGechie, P. A. Rogers, and R. J. Glancy. 1985. Pyloric campylobacter infection and gastroduodenal disease. Med. J. Aust. 142:439. subjects. To assess the contribution of host B cells and their prod- 3. Cover, T. L., and M. J. Blaser. 1995. Helicobacter pylori: a bacterial cause of ucts in Helicobacter-induced gastric pathology, we have infected gastritis, peptic ulcer disease, and gastric cancer. ASM News 61:21. the B cell-deficient mouse strain, ␮MT. Our results showed that 4. Veldhuyzen van Zanten, S. J. O., and P. M. Sherman. 1994. Helicobacter pylori ␮ infection as a cause of gastritis, duodenal ulcer, gastric cancer and nonulcer the MT and C57BL/6 mice exhibited identical gastric pathology dyspepsia: a systematic overview. Can. Med. Assoc. J. 150:177. after H. felis infection, with mucosal inflammation, gastric epithe- 5. Dixon, M. F. 1995. Histological responses to Helicobacter pylori infection: gas- lial cell hyperproliferation, parietal and zymogenic cell destruc- tritis, atrophy and preneoplasia. Bailliere’s Clin. Gastroenterol. 9:467. 6. Covacci, A., S. Falkow, D. E. Berg, and R. Rappuoli. 1997. Did the inheritance tion, and metaplastic changes. This demonstrates for the first time of a pathogenicity island modify the virulence of Helicobacter pylori. Trends that B cells and their secretion of potentially autoreactive Abs are Microbiol. 5:205. not critical to the development of gastric epithelial cell destruction 7. Matsukura, N., M. Onda, S. Kato, H. Hasegawa, K.-I. Okawa, T. Shirakawa, A. Tokunaga, K. Yamashita, and A. Hayashi. 1997. Cytotoxin genes of Helico- and hyperplasia after H. felis infection. bacter pylori in chronic gastritis, gastroduodenal ulcer and gastric cancer: an age The finding that the adaptive immune system, specifically T and gender matched case-control study. Jpn. J. Cancer Res. 88:532. cells, are critical mediators of H. felis-associated gastric pathology 8. de Jong, D., R. W. M. van der Hulst, G. Pals, W. C. van Dijk, A. van der Ende, G. N. J. Tytgat, B. G. Taal, and H. Boot. 1996. Gastric non-Hodgkin lymphomas does not allow for the discrimination between direct Ag-specific of mucosa-associated lymphoid tissue are not associated with more aggressive and indirect non Ag-specific effects of T cells. There are at least Helicobacter pylori strains as identified by CagA. Am. J. Clin. Pathol. 106:670. The Journal of Immunology 1497

9. Fox, J. G., M. Blanco, J. C. Murphy, N. S. Taylor, A. Lee, Z. Zabok, and 30. Uibo, R., T. Vorobjova, K. Metskula, K. Kisand, T. Wadstrom, and T. Kivik. J. Pappo. 1993. Local and systemic immune responses in murine Helicobacter 1995. Association of Helicobacter pylori and gastric autoimmunity: A popula- felis active chronic gastritis. Infect. Immun. 61:2309. tion-based study. FEMS Immunol. Med. Microbiol. 11:65. 10. Ernst, P. B., Y. Jin, V. E. Reyes, and S. E. Crowe. 1994. The role of the local 31. Faller, G., H. Steininger, M. Eck, J. Hensen, E. G. Hahn, and T. Kirchner. 1996. immune response in the pathogenesis of peptic ulcer formation. Scand. Antigastric autoantibodies in Helicobacter pylori gastritis: prevalence, in-situ J. Gastroenterol. 29(Suppl. 205):22. binding sites and clues for clinical relevance. Virchows Arch. 427:483. 11. Blanchard, T. G., S. J. Czinn, J. G. Nedrud, and R. W. Redline. 1995. Helico- 32. Negrini, R., A. Savio, C. Poiesi, B. J. Appelmelk, F. Buffoli, A. Paterlini, bacter-associated gastritis in SCID mice. Infect. Immun. 63:1113. P. Cesari, M. Graffeo, D. Vaira, and G. Franzin. 1996. Antigenic mimicry be- 12. Sakagami, T., M. Dixon, J. O’Rourke, R. Howlett, F. Alderuccio, J. Vella, tween Helicobacter pylori and gastric mucosa in the pathogenesis of body atro- T. Shimoyama, and A. Lee. 1996. Atrophic gastric changes in both Helicobacter phic gastritis. Gastroenterology 111:655. felis and Helicobacter pylori infected mice are host dependent and separate from 33. Appelmelk, B. J., G. Faller, D. Claeys, T. Kirchner, and antral gastritis. Gut 39:639. C. M. J. E. Vandenbroucke-Grauls. 1998. Bugs on trial: the case of Helicobacter 13. Mohammadi, M., R. Redline, J. Nedrud, and S. Czinn. 1996. Role of the host in pylori and autoimmunity. Immunol. Today 19:296. pathogenesis of Helicobacter-associated gastritis: H. felis infection of inbred and 34. Claeys, D., G. Faller, B. J. Appelmelk, R. Negrini, and T. Kirchner. 1998. The congenic mouse strains. Infect. Immun. 64:238. gastric Hϩ,Kϩ-ATPase is a major autoantigen in chronic Helicobacter pylori 14. Lee, A., J. O’Rourke, M. Corazon de Ungria, B. Robertson, G. Daskalopoulos, gastritis with body mucosa atrophy. Gastroenterology 115:340. and M. F. Dixon. 1997. A standardized mouse model of Helicobacter pylori 35. Kitamura, D., J. Roes, R. Kuhn, and K. Rajewsky. 1991. A B cell-deficient mouse infection: introducing the Sydney strain. Gastroenterology 112:1386. by targeted disruption of the membrane exon of the immunoglobulin ␮ chain 15. Ferrero, R. L., J. M. Thiberge, M. Huerre, and A. Labigne. 1998. Immune re- gene. Nature 350:423. sponses of specific-pathogen-free mice to chronic Helicobacter pylori (strain SS1) infection. Infect. Immun. 66:1349. 36. Mombaerts, P., A. R. Clarke, M. A. Rudnicki, J. Iacomini, S. Itohara, J. J. Lafaille, L. Wang, Y. Ichikawa, R. Jaenisch, M. L. Hooper, and 16. Pappo, J., W. D. Thomas, Z. Zabok, N. S. Taylor, J. C. Murphy, and J. G. Fox. ␣ ␤ 1995. Effect of oral immunization with recombinant on murine Helico- S. Tonegawa. 1992. Mutations in T-cell antigen receptor genes and block bacter felis gastritis. Infect. Immun. 63:1246. thymocyte development at different stages. Nature 360:225. 17. Fox, J. G., X. Li, R. J. Cahill, K. Andrutis, A. K. Rustgi, R. Odze, and 37. Mombaerts, P., E. Mizoguchi, M. J. Grusby, L. H. Glimcher, A. K. Bhan, and T. C. Wang. 1996. Hypertrophic gastropathy in Helicobacter felis-infected wild- S. Tonegawa. 1993. Spontaneous development of inflammatory bowel disease in Downloaded from type C57BL/6 mice and p53 hemizygous transgenic mice. Gastroenterology 110: T cell receptor mutant mice. Cell 75:275. 155. 38. Itohara, S., P. Mombaerts, J. Lafaille, J. Iacomini, A. Nelson, A. R. Clarke, 18. Xiang, Z., S. Censini, P. F. Bayeli, J. L. Telford, N. Figura, R. Rappuoli, and M. L. Hooper, A. Farr, and S. Tonegawa. 1993. T cell receptor ␦ gene mutant A. Covacci. 1995. Analysis of expression of CagA and VacA virulence factors in mice: independent generation of ␣␤ T cells and programmed rearrangements of 43 strains of Helicobacter pylori reveals that clinical isolates can be divided into ␥␦ TCR genes. Cell 72:337. two major types and that CagA is not necessary for expression of the vacuolating 39. Correa, P. 1992. Human gastric carcinogenesis: A multistep and multifactorial cytotoxin. Infect. Immun. 63:94. process: first American Cancer Society award lecture on cancer epidemiology and prevention. Cancer Res. 52:6735. 19. Chwalinski, S., C. S. Potten, and G. Evans. 1988. Double labeling with bromode- http://www.jimmunol.org/ oxyuridine and [3H]thymidine of proliferative cells in small intestinal epithelium 40. Cahill, R. J., C. Kilgallen, S. Beattie, H. Hamilton, and C. O’Morain. 1996. in steady state and after irradiation. Cell Tissue Res. 21:317. Gastric epithelial cell kinetics in the progression from normal mucosa to gastric 20. Ermak, T. H., R. Ding, B. Ekstein, J. Hill, G. A. Myers, C. K. Lee, J. Pappo, carcinoma. Gut 38:177. H. K. Kleanthous, and T. P. Monath. 1997. Gastritis in urease-immunized mice 41. Lynch, D. A. F., N. P. Mapstone, A. M. T. Clarke, G. M. Sobala, P. Jackson, after Helicobacter felis challenge may be due to residual bacteria. Gastroenter- L. Morrison, M. F. Dixon, P. Quirke, and A. T. R. Axon. 1995. Cell proliferation ology 113:1118. in Helicobacter pylori associated gastritis and the effect of eradication therapy. 21. Cohn, S. M., and M. W. Lieberman. 1984. The use of antibodies to 5-bromo-2Ј- Gut 36:346. deoxyuridine for the isolation of DNA sequences containing excision-repair sites. 42. Karam, S. M., and C. P. Leblond. 1993. Dynamics of epithelial cells in the corpus J. Biol. Chem. 259:12456. of the mouse stomach. I. Identification of proliferative cell types and pinpointing 22. Cohn, S. M., T. C. Simon, K. A. Roth, E. H. Birkenmeier, and J. I. Gordon. 1992. of the stem cell. Anat. Rec. 236:259. Use of transgenic mice to map cis-acting elements in the intestinal fatty acid 43. Watanabe, T., M. Tada, H. Nagai, S. Sasaki, and M. Nakao. 1998. Helicobacter

binding protein gene (Fabpi) that control its cell lineage-specific and regional pylori infection induces gastric cancer in mongolian gerbils. Gastroenterology by guest on September 26, 2021 patterns of expression along the duodenal-colonic and crypt-villus axes of the gut 115:642. epithelium. J. Cell Biol. 119:27. 44. Guruge, J. L., P. G. Falk, R. G. Lorenz, M. Dans, H.-P. WIrth, M. J. Blaser, 23. Lorenz, R. G., and J. I. Gordon. 1993. Use of transgenic mice to study regulation D. E. Berg, and J. I. Gordon. 1998. Epithelial attachment alters the outcome of of gene expression in the parietal cell lineage of gastric units. J. Biol. Chem. Helicobacter pylori infection. Proc. Natl. Acad. Sci. USA 95:3925. 268:26559. 45. Edwards, P. D., J. Carrick, J. Turner, A. Lee, H. Mitchell, and D. A. Cooper. 24. Falk, P., K. A. Roth, and J. I. Gordon. 1994. Lectins are sensitive tools for 1991. Helicobacter pylori-associated gastritis is rare in AIDS: effect or defining the differentiation programs of mouse gut epithelial cell lineages. a consequence of immunodeficiency? Am. J. Gastroenterol. 86:1761. Am. J. Physiol. 266:G987. 46. Bosma, M. J., G. C. Bosma, and J. L. Owen. 1978. Prevention of immunoglobulin 25. Falk, P., R. G. Lorenz, N. Sharon, and J. I. Gordon. 1995. Moluccella laevis production by allotype-dependent T cells. Eur. J. Immunol. 8:562. lectin, a marker for cellular differentiation programs in mouse gut epithelium. Am. J. Physiol. 268:G553. 47. Gleeson, P. A., and B.-H. Toh. 1991. Molecular targets in pernicious anaemia. 26. Hermiston, M. L., C. B. Latham, J. I. Gordon, and K. A. Roth. 1992. Simulta- Immunol. Today 12:233. neous localization of six antigens in a single section of transgenic mouse intestine 48. Cover, T. L. 1996. The vacuolating cytotoxin of Helicobacter pylori. Mol. Mi- using a combination of light and fluorescence microscopy. J. Histochem. Cyto- crobiol. 20:241. chem. 40:1283. 49. Wagner, S., W. Beil, J. Westermann, R. P. H. Logan, C. T. Bock, C. Trautwein, 27. Lee, E. Y., B. Seetharam, D. H. Alpers, and K. DeSchryver-Kecskemeti. 1989. J. S. Bleck, and M. P. Manns. 1997. Regulation of gastric epithelial cell growth Immunohistochemical survey of cobalamin-binding proteins. Gastroenterology by Helicobacter pylori: evidence for a major role of apoptosis. Gastroenterology 97:1171. 113:1836. 28. Mombaerts, P., J. Iacomini, R. S. Johnson, K. Herrup, S. Tonegawa, and 50. Jones, N. L., P. T. Shannon, E. Cutz, H. Yeger, and P. M. Sherman. 1997. V. E. Papaioannou. 1992. RAG-1-deficient mice have no mature B and T lym- Increase in proliferation and apoptosis of gastric epithelial cells early in the nat- phocytes. Cell 68:869. ural history of Helicobacter pylori infection. Am. J. Pathol. 151:1695. 29. Negrini, R., L. Lisato, I. Zanella, L. Cavazzini, S. Gullini, V. Villanacci, 51. Yabuki, N., H. Sasano, M. Tobita, A. Imatani, T. Hoshi, K. Kato, S. Ohara, C. Poiesi, A. Albertini, and S. Ghielmi. 1991. Helicobacter pylori infection in- S. Asaki, T. Toyota, and H. Nagura. 1997. Analysis of cell damage and prolif- duces antibodies cross-reacting with human gastric mucosa. Gastroenterology eration in Helicobacter pylori-infected human gastric mucosa from patients with 101:437. gastric adenocarcinoma. Am. J. Pathol. 151:821.