Gastric Ulcer, Atrophic Gastritis, and Intestinal Metaplasia Caused byHelicobacter pyloriInfection in Mongolian Gerbils

S. HONDA, T. FUJIOKA, M. TOKIEDA, T. GOTOH, A. NISHIZONO & M. NASUSecond Dept. of Internal Medicine and Dept. of Microbiology, Oita Medical University, Oita, Japan

Honda S, Fujioka T, Tokieda M, Gotoh T, Nishizono A, Nasu M. Gastric ulcer, atrophic gastritis,and intestinal metaplasia caused byHelicobacter pyloriin Monogolian gerbils. Scand J Gastroenterol1998;33:454–460.

Background: Helicobacter pyloriinfection is associated with gastroduodenal disease in humans. In thisstudy we aimed to show this relationship directly in Mongolian gerbils.Methods:The animals werechallenged orally withH. pylori and killed 1, 2, 3, and 6 months after inoculation for histologic and anti-H.pylori antibody titer examination.Results:The spiral bacteria were observed in the mucus and gastric pitsof all infected animals. A severe infiltration of the lamina propria by polymorphonuclear and mononuclearcells was seen 1 month afterH. pylori inoculation. The submucosa was infiltrated by mainly mononuclearcells with formation of lymphoid follicles. Erosion of the gastric mucosa appeared soon after inoculation,whereas gastric ulcers, gastritis cystica profunda, and atrophy with goblet cell metaplasia occurredbetween 3 and 6 months after inoculation. In the duodenal mucosa a mild inflammatory cell infiltrationwith ballooning and diminished number of duodenal glands was seen. The IgG anti-H. pylori antibody titerincreased gradually after 2 months of inoculation.Conclusions:Since the gastritis, gastric ulcers, atrophicgastritis, and intestinal metaplasia that developed in Mongolian gerbils were similar to those observed inhumans, this model may be useful to study the therapy of gastric ulcer and, with a longer observationperiod, to confirm a possible relationship betweenH. pylori and malignancy.

Key words:Animal model; gastric atrophy; gastric ulcer;Helicobacter pylori; intestinal metaplasia;Mongolian gerbil

Shoji Honda, M.D., Second Dept. of Internal Medicine, Oita Medical University, Hasama-machi, Oita879-55, Japan (fax:�81 975-49-4245)

Since its first isolation from a patient with gastritis by Warren& Marshall in 1983 (1),Helicobacter pylorihas been thefocus of a great number of research articles from all over theworld (2–7). These studies have markedly contributed to ourunderstanding of the role ofH. pylori in gastroduodenaldiseases. It is generally believed thatH. pylori infection isassociated with chronic gastritis, peptic ulcer disease, andgastric carcinoma in humans (2–19). However, most evidenceon this association derives from epidemiologic and clinicalstudies. Since it is difficult to show directly the developmentof infection in the human environment, the use of animalmodels has been recommended. They offer the advantages ofa close follow-up and the use of a sufficient number ofanimals at a time.

Numerous animal models ofH. pylori infection andassociated gastritis have been reported (20–30). The hostresponse toH. pylori varies among different animals, and onlyprimate models have been shown to develop a gastritis similarto that observed in humans (27–30). However, the use of suchanimals is often difficult for general laboratories. Therefore,the search for a more proper animal model could be of criticalimportance. Since the Mongolian gerbil has been reported tobe colonized byH. pylori (31, 32), we used it to study thedevelopment ofH. pylori-induced gastritis and gastric ulcerand, furthermore, to investigate the relationship betweenH.pylori infection and malignancy.

MATERIALS AND METHODS

BacteriaH. pylori ATCC-43504 possessing the cagA gene and

expressing vacuolating cytotoxin was used. A 4-day cultureon blood agar at 37°C under microaerophilic conditions washarvested and incubated in brucella broth (DIFCO Labora-tories, Detroit, Mich., USA) with 10% horse serum for 24 h.Inoculum size was adjusted with sterile saline to produce theoptical density of McFarland 4 at 540 nm.

AnimalsFive-week-old male Mongolian gerbils weighing 30–40 g

(Seiwa Experimental Animals Co. Ltd., Fukuoka, Japan) wereused. They were housed five per cage, starved for 24 h, andthen fed with chow (Oriental Yeast Co., Tokyo, Japan) andwater ad libitum beginning 12 h afterH. pylori inoculation.Experiments were performed in accordance with the guide-lines of the Ethical Committee for Animal Experiments atOita Medical University, Oita, Japan.

Infection modelOn the day of infection the Mongolian gerbils were

randomly divided into 2 groups of 20 animals each andchallenged orally with vehicle or 109 colony-forming units(CFU) H. pylori in 1.0 ml of brucella broth with 10% horse

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Cal

ifor

nia

Irvi

ne o

n 10

/29/

14Fo

r pe

rson

al u

se o

nly.

Fig. 1. Gastric mucosa of Mongolian gerbilinfected with Helicobacter pylori (Giemsastain; magnification, �1200). The spiralbacteria (arrow) were seen in the gastric pit.

Fig. 2A. Pyloric mucosa of Mongolian gerbil3 months afterHelicobacter pylori inocula-tion (hematoxylin and eosin stain; magnifica-tion, �48). The mucosa was thick, and thepyloric glands dilated, with some of themdestroying the lamina muscularis and extend-ing to the submucosal layer. A marked focalinfiltration of the submucosa by mononuclearcells with formation of lymphoid follicles wasseen. 2B. Pyloric mucosa of an uninfectedanimal with normal histology.

H. pylori Gastric Lesions in Gerbils 455

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Cal

ifor

nia

Irvi

ne o

n 10

/29/

14Fo

r pe

rson

al u

se o

nly.

serum. Three or five animals in each experimental group wereweighed and killed under anesthesia with ether 1, 2, 3, and 6months after inoculation. Immediately after they had beenkilled, blood was drawn from the heart, and serum was iso-lated and stored atÿ20°C until use. The stomach was quicklyremoved, divided in two, and used for microscopic andculture examinations.

Histologic examinationHalf of the stomach was fixed in 10% neutral buffered

formalin and embedded in paraffin. Five-micrometer sectionswere stained with hematoxylin and eosin (H&E), Giemsa, andperiodic acid-Schiff (PAS) stains and examined for inflam-matory response and the presence of organisms.

Culture studyThe remaining half of the stomach was homogenized in

2 ml 20% glucose liquid in a glass homogenizer (Iwaki GlassCo. Ltd., Tokyo, Japan), and the homogenates were culturedon blood agar for 4 days at 37°C under microaerophilicconditions.

Serum anti-H. pylori antibodySerum IgG antibody toH. pylori was determined with an

enzyme-linked immunosorbent assay (ELISA) described byGoodwin et al. (33), with slight modification. In brief, thesonicated whole cells ofH. pylori strain ATCC-43504 in 0.5M carbonate buffer (pH 9.6) were coated onto ELISAmicroplates for 1 h at 37°C. After they had been washedwith phosphate-buffered saline–Tween (PBST), the excessprotein binding sites were blocked with bovine serumalbumin and kept overnight at 4°C. The diluted Mongoliangerbils serum (1:100) was then added, and the well micro-plates incubated at 37°C for 1 h. After another washing, thediluted peroxidase-conjugated anti-mouse IgG antibody(Cappel, Durham, N.C., USA) was added, and the wellmicroplates incubated at 37°C. Washing was repeated, and 2-2-azinobis (ABTS) was added for 30 min. The absorbancewas measured at 414 nm.

Statistic analysisAll values are expressed as the means� standard error of

the mean. Data were analyzed with Student’st test. Signifi-cance was defined asP� 0.05.

RESULTS

The spiral bacteria were observed in the mucus and gastricpits of almost all inoculated animals from 1 month afterinoculation throughout the whole observation period (Fig. 1).However, only about 50% of the animals were culture-positive. The bacterial counts from the stomachs of Mon-golian gerbils 1 and 6 months afterH. pylori inoculation were2.5 and 41 CFU/mg of tissue, respectively.

Infected animals carrying the spiral bacteria in their

stomachs throughout the observation period showed erosionsof the gastric mucosa and various degrees of inflammatoryresponse (Table I). A severe infiltration of the lamina propriaby polymorphonuclear (PMN) and mononuclear (MN) cellswas observed. The submucosal layer was infiltrated by mainlyMN cells, with formation of lymphoid follicles (Table II).Pyloric glands were dilated, and some of them penetrated thelamina muscularis. They extended into the submucosal layer3 months after inoculation (Fig. 2). The gastric ulcers,gastritis cystica profunda, atrophy, and goblet cell metaplasiaoccurred between 3 and 6 months after inoculation (Table I).Seven ulcers localized in the pyloric area were seen in four offive animals (Fig. 3), and all of them penetrated the laminamuscularis (Fig. 4A). Atrophic mucosa (Fig. 4B) appeared inthe pyloric area, and isolated goblet cell metaplasia (Fig. 4C)was seen in two of five animals. In the duodenal mucosa mildinflammatory cell infiltration and ballooning of duodenalglands were seen. At 6 months after inoculation the balloon-ing of duodenal glands was severe and the number of glandsdiminished. However, neither gastric metaplasia nor duodenalulcers occurred (Table I).

The results of anti-H. pylori antibody titers are shown inFig. 5. The titer of IgG anti-H. pylori antibody increasedgradually beginning 2 months after inoculation.

Table I. Histopathologic changes in the gastric and duodenal mucosain Mongolian gerbils infected withHelicobacter pyloriATCC-43504

Months (positive cases/total)

Findings 1 2 3 6

Organism 3/4 2/3 4/4 5/5Erosion 3/4 2/3 4/4 5/5Gastritis 3/4 2/3 4/4 5/5Lymph follicle 0/4 1/3 4/4 5/5Gastric ulcer 0/4 0/3 0/4 4/5Gastritis cystica profunda 0/4 0/3 0/4 3/5Atrophy 0/4 0/3 0/4 4/5Goblet cell metaplasia 0/4 0/3 0/4 2/5Inflammatory cell infiltration of

duodenal bulb 0/4 2/3 4/4 5/5Gastric metaplasia 0/4 0/3 0/4 0/5Duodenal ulcer 0/4 0/3 0/4 0/5

Table II. Inflammatory cell infiltration in the gastric mucosa ofMongolian gerbils infected withHelicobacter pyloriATCC-43504

Months/grade*

Cell Layer 1 2 3 6

PMN Lamina propria ��� �� ��� ��

Submucosal layer � � � �

MN Lamina propria �� ��� ��� ��

Submucosal layer � ��� ��� ���

* � = mild;�� = moderate;��� = severe. PMN = polymorpho-nuclear; MN = mononuclear.

456 S. Honda et al.

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Cal

ifor

nia

Irvi

ne o

n 10

/29/

14Fo

r pe

rson

al u

se o

nly.

DISCUSSION

The ability of H. pylori to colonize the stomach of variousanimals differs. Infection with this organism has beensuccessfully developed in animals such as gnotobiotic piglets,barrier-born pigs, gnotobiotic beagles, nude mice, mice,cynomolgus monkeys, rhesus monkeys, Japanese monkeys,and Mongolian gerbils (20–32). Although the gastric ulcersand erosions were produced in gnotobiotic piglets (22), mice(26), and Mongolian gerbils (32), macroscopic and deepgastric ulcers were only observed in gnotobiotic piglets andMongolian gerbils (22, 32). The experimentalH. pyloriinfection model using Mongolian gerbils was reported byYokota et al. (31). During 2 months of their observation only amild inflammatory infiltration in the gastric mucosa ofinoculated animals was seen. Hirayama et al. (32) reportedthat ulcers and intestinal metaplasia were produced 6 monthsafter inoculation withH. pylori in Mongolian gerbils (32).

In the present studyH. pylori infection caused a severeinflammatory cell infiltration of the gastric mucosa by PMNand MN cells, formation of lymphoid follicles in the sub-mucosal layer, gastric ulcers, atrophy of the gastric mucosa,and goblet cell metaplasia. These histologic findings aresimilar to those found in the human gastric mucosa infectedwith H. pylori (2, 3, 37).

Gastric ulcers at the pyloric area in proximity to the fundusappeared in 80% of animals 6 months after inoculation, andalmost all of them penetrated the lamina muscularis. Becauseno ulcers were seen in uninfected animals, it was certain thatH. pylori infection caused gastric ulcers independently. It hasbeen reported that products ofH. pylori, such as vacuolatingcytotoxin and CagA protein (13, 34, 35), ammonia (16, 17),heat shock protein (18), and active oxygen species (19) maybe ulcerogenic promoters ofH. pylori infection. In this studyPMN and MN cell infiltration and lymph follicles wereconstantly observed several months before the appearance ofgastric ulcers. We believe that the persistence of inflamma-

tory cell infiltration in the gastric mucosa was indispensablefor the appearance of gastric ulcers. On the other hand, thebacterial counts in the stomachs of infected animals hadincreased significantly at 6 months compared with that at 3months after inoculation, suggesting that the number ofbacteria also plays a role in the ulcer formation.

In humans duodenal ulcers are strongly associated withH.pylori infection (3–7). Furthermore, it has been reported thatthe major risk factor for duodenal ulceration in patientscolonized byH. pylori is the presence of gastric metaplasia inthe duodenal bulb (36). In this study a mild inflammatory cellinfiltration and ballooning of duodenal glands were seen inthe duodenal mucosa. However, neither gastric metaplasia ofthe duodenal bulb nor duodenal ulcers appeared. In addition,H. pylori was not detected in the duodenal epithelium. Thesefindings suggest that the Mongolian gerbil is not a suitableduodenal ulcer model.

The lymph follicle formation has also been observed ininfants (37), gnotobiotic piglets (21), and gnotobiotic beagles(24) infected withH. pylori but not in primates, which show apattern of inflammation similar to that in adult humans (27–30). Gnotobiotic piglets (21), gnotobiotic beagles (24), andMongolian gerbils were infected within a few weeks afterbirth, whereas the primates (30) were infected in adult age.This difference in the infection age may account for thedifferent histologic findings in the hosts.

H. pylori infection is assumed to be a cause of gastric-associated lymphoid tissue lymphoma (MALToma). Enno etal. (38) reported findings similar to the MALToma-likelesions in a mouse model ofH. felis infection. The reportedclinical experience also suggests thatH. pylori may be a causeof MALToma (39, 40). In agreement with the above reports,our findings showed a marked focal infiltration by MN in thelamina propria and lymphoid follicle formation in thesubmucosal layer. Although these histologic findings of thestomach may be suggestive of the MALToma-like lesions,they are not typical, nor is the presence of a clonal B-cellproliferation in the lymphoid follicles confirmed by means ofimmunohistochemistry. Hence, further investigation isneeded to elucidate this matter.

Atrophy of the pyloric mucosa and goblet cell metaplasiaoccurred between 3 and 6 months afterH. pylori inoculation.Mucosal atrophy has been reported to occur in Japanesemonkeys 1½ years after infection withH. pylori (41), butintestinal metaplasia has not been reported either in primatesor in other animals. At 6 months after infection withH. pylorithe goblet cell metaplasia was rare and isolated. However,intestinal metaplasia is considered a precursor of gastriccancer (42). We are continuing the observation of our modelto discern whether the gastric cancer occurs after the intestinalmetaplasia.

Serum IgG anti-H. pylori antibody was increased from 2 to6 months afterH. pylori inoculation, and all animals with highantibody titer showed inflammatory cell infiltration of thegastric epithelium and spiral bacteria in the gastric pits.

Fig. 3. Delineated schema of the gastric mucosa of Mongoliangerbils showing the ulcer location. Seven ulcers were seen in four offive animals, all of them localized in the pyloric area (P) inproximity to the fundus (F). D = duodenum.

H. pylori Gastric Lesions in Gerbils 457

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Cal

ifor

nia

Irvi

ne o

n 10

/29/

14Fo

r pe

rson

al u

se o

nly.

Therefore, the antibody titer was more useful for the diagnosisof H. pylori infection in Mongolian gerbils. We choseperoxidase-labeled anti-mouse IgG antibody for use in theELISA method because the cross-reactivity between theMongolian gerbil serum and the mouse serum could beascertained with the Ouchterlony method. Although nodifficulties appeared in determining the antibody, furtherwork is needed to produce an anti-Mongolian gerbil mono-clonal antibody.

Since the Mongolian gerbil developed gastric ulcers,atrophic gastritis, and intestinal metaplasia due toH. pyloriinfection, this model may be useful in studying the therapy of

gastric ulcers or the impact ofH. pylori eradication onprogression/regression of atrophic gastritis, and, with a longerobservation period, it may help to clarify a possible relation-ship betweenH. pylori and malignancy.

Fig. 4. Microscopic views of the pyloric mucosa of Mongolian gerbils infectedwith Helicobacter pylori. 4A. Gastric ulcer (hematoxylin and eosin (HE) stain;magnification,�30). Mucosal atrophy with a reduced number of pyloric glands(4B) (HE stain; magnification,�120) and goblet cell metaplasia (4C) (arrow)(periodic acid-Schiff; magnification,�240) were seen in the infected animals 6months after H. pylori inoculation. 4D. Pyloric mucosa of an uninfectedMongolian gerbil with normal histology.

458 S. Honda et al.

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Cal

ifor

nia

Irvi

ne o

n 10

/29/

14Fo

r pe

rson

al u

se o

nly.

ACKNOWLEDGEMENTS

We thank Miss M. Kimoto and Miss K. Ohno for theirtechnical assistance, and Dr. P. Kamberi for critical reading ofthe manuscript. This work was supported in part by a Grant-in-Aid for Science Research (98457170) from the Ministry ofEducation, Science and Culture, Japan.

REFERENCES

1. Warren JR, Marshall BJ. Unidentified curved bacilli on gastricepithelium in active chronic gastritis. Lancet 1983;1:1273–5.

2. Blaser MJ.Helicobacter pyloriand the pathogenesis of gastro-duodenal inflammation. J Infect Dis 1990;161:626–33.

3. Graham DY, Lew GM, Klein PD, Evance DG, Saeed ZA,Malaty HM. Effect of treatment ofHelicobacter pyloriinfectionon the long-term recurrence of gastric or duodenal ulcer. AnnIntern Med 1992;116:705–8.

4. Hentschel E, Brandstatter G, Dragosics B, Hirsdhl AM, NemeeH, Schutze K, et al. Effect of ranitidine and amoxicillin plusmetronidazole on the eradication ofHelicobacter pyloriand therecurrence of duodenal ulcer. N Engl J Med 1993;328:308–12.

5. Forbes GM, Glaser ME, Cullen DJ, Warren JR, Christiansen KJ,Marshall BJ, et al. Duodenal ulcer treated withHelicobacterpylori eradication: seven year follow-up. Lancet 1994;343:258–60.

6. Wilhelmsen I, Berstad A. Quality of life and relapse of duodenalulcer before and after eradication ofHelicobacter pylori. ScandJ Gastroenterol 1994;29:874–9.

7. Labenz J, Brsh G. Highly significant change of the clinicalcourse of relapsing and complicated peptic ulcer disease aftercure ofHelicobacter pyloriinfection. Am J Gastroenterol 1994;89:1785–8.

8. Correa P. A human model of gastric carcinogenesis. Cancer Res1988;48:3554–60.

9. Correa P, Fox J, Fontham E, Ruiz B, Lin YP, Zavara D, et al.Helicobacter pylori and gastric carcinoma. Serum antibodyprevalence in populations with contrasting cancer risks. Cancer1990;66:2569–74.

10. Cover TL, Dooley CP, Blaser MJ. Characterization of humanserologic response to proteins inHelicobacter pylori brothculture supernatants with vacuolizing cytotoxin activity. InfectImmun 1990;58:603–10.

11. Cover TL, Blaser MJ. Purification and characterization of the

vacuolating toxin fromHelicobacter pylori. J Biol Chem 1992;267:10570–5.

12. Xiang Z, Censini S, Bayeli PF. Analysis of expression of CagAand VacA virulence factors in 43 strains ofHelicobacter pylorireveals that clinical isolates can be divided into two major typesand that CagA is not necessary for expression of the vacuolatingcytotoxin. Infect Immun 1995;63:94–8.

13. Crabtree JE, Talor JD, Wyatt JI, Heatley RV, Shalicross TM,Tompkins DS, et al. Mucosal recognition ofHelicobacter pylori120 kDa protein, peptic ulceration and gastric pathology. Lancet1991;338:332–5.

14. Crabtree JE, Farmery SM, Lindley IJD, Figura N, Peichl P,Tompkins DS. CagA/eytotoxic strains ofHelicobacter pyloriand interleukin-8 in gastric epithelial cell lines. J Clin Pathol1994;47:945–50.

15. Figura N. Progress in defining the inflammatory cascade. Eur JGastroenterol Hepatol 1995;7:296–302.

16. Hazell S, Lee A.Campylobacter pyloridis, urease, hydrogenback diffusion and gastric ulcer. Lancet 1986;2:16–7.

17. Murakami M, Yoo JK, Mizumo M, Inada M, Miyake T. Effectof ammonia on the gastric mucosa in rats: pathophysiologicalimportance of urease in gastric ulcer disease. Jpn J Pharmacol1988;47:330–2.

18. Macchia G, Massone A, Burroni D, Covacei A, Sensini S,Rappuoli R. TheHsp60protein ofHelicobacter pylori:structureand immune response in patients with gastroduodenal diseases.Mol Microbiol 1993;9:645–52.

19. Naito Y, Yoshikawa T, Yoneta T, Vagi N, Matsuyama E, AraiM, et al. A new gastric ulcer model in rats produced by ferrousiron and ascorbic acid injection. Digestion 1995;56:472–8.

20. Krakowka S, Morgan DR, Kraft WG, Leunk RD. Establishmentof gastric Campylobacter pylori infection in the neonatalgnotobiotic piglet. Infect Immun 1987;55:2789–96.

21. Eaton KA, Morgan DR, Krakowka S. Persistence ofHelico-bacter pylori in conventionalized piglets. J Infect Dis 1990;161:1299–301.

22. Krakowka S, Eaton KA, Rings DM. Occurrence of gastric ulcersin gnotobiotic piglets colonized byHelicobacter pylori. InfectImmun 1995;63:2352–5.

23. Engstrand L, Gustavsson S, Jo¨rgensen A, Schwan A, ScheyniusA. Inoculation of barrier-born pigs withHelicobacter pyloriauseful animal model for gastritis type B. Infect Immun 1990;58:1763–8.

24. Radin MJ, Eaton KA, Krakowka S, Morgan DR, Lee A, Otto G,et al.Helicobacter pylorigastric infection in gnotobiotic beagledogs. Infect Immun 1990;58:2606–12.

25. Karita M, Koehiyama T, Okita K, Nakazawa T. New small

Fig. 5. Time course of anti-Helicobacter pyloriantibody in Mongolian gerbils infected withH. pylori.The titer increased progressively 2 months afterH. pylori inoculation. Control (*); infected group (*).*P � 0.05.

H. pylori Gastric Lesions in Gerbils 459

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Cal

ifor

nia

Irvi

ne o

n 10

/29/

14Fo

r pe

rson

al u

se o

nly.

animal model for human gastricHelicobacter pyloriinfection:success in both nude and authymic mice. Am J Gastroenterol1991;86:1596–603.

26. Marchetti M, AricoB, Burroni D, Figula N, Rappuoli R, GhiaraP. Development of a mouse model of Helicobacter pyloriinfection that mimics human disease. Science 1995;267:1655–8.

27. Takahashi S, Igarashi H, Ishiyama H, Nakano M, Ozaki M, ItoM, et al. Serial change of gastric mucosa after challenging withHelicobacter pyloriin the cynomolgus monkey. Zbl Bact 1993;280:51–7.

28. Euler AR, Zirenko GE, Moe JB, Urich RG, Vagi Y. Evaluationof two monkey species(Macaca mulattaand Macaca fascicu-laris) as possible model for humanHelicobacter pyloridisease.J Clin Microbiol 1990;28:2285–90.

29. Fukuda Y, Tamura K, Yamamoto I, Kawaura A, Tonokatsu Y,Tsuyuguchi T, et al. Inoculation of rhesus monkeys with humanHelicobacterpylori:a long term investigation on gastric mucosaby endoscopy. Dig Endosc 1992;4:19 30.

30. Shuto R, Fujioka T, Kubota T, Nasu M. Experimental gastritisinduced by Helicobacter pylori in Japanese monkeys. InfectImmun 1993;61:933–9.

31. Yokota K, Kurebayashi Y, Takayama Y, Hayashi S, Isogai H,Imai K, et al. Colonization ofHelicobacter pyloriin the gastricmucosa of Mongolian gerbils. Microbiol Immunol 1991;35:475–80.

32. Hirayama F, Takagi S, Kusuhara H, Iwao E, Yokoyama Y,Ikeda Y. Induction of gastric ulcer and intestinal metaplasia inMongolian gerbils infected withHelicobacter pylori. J Gastro-enterol 1996;31:755–7.

33. Goodwin CS, Blincow E, Peterson G, Sanderson C, Cheng W,Marshall B, et al. Enzyme-linked immunosorbent assay forCampylobacter pyloridis:correlation with presence ofC.pyloridis in the gastric mucosa. J Infect Dis 1987;155:488–94.

34. Tee W, Lambert JR, Dwyer B. Cytotoxin productionby Helico-

bacter pylori from patients with upper gastrointestinal tractdiseases. J Clin Microbiol 1995;33:1203–5.

35. Telford JL, Ghiara P, Dell’Orco M, Comanducci M, Burroni D,Bugnoli M, et al. Gene structure of theHelicobacter pyloricytotoxin and evidence of its key role in gastric disease. J ExpMed 1994;179:1653–8.

36. Gormally SM, Kierce BM, Daly LE, Bourke B, Carroll R,Durnin MT, et al. Gastric metaplasia and duodenal ulcer diseasein children infected byHelicobacter pylori. Gut 1996;38:513–7.

37. Hassall E, Dimmick JE. Unique featuresof Helicobacter pyloridisease in children. Dig Dis Sci 1991;36:417–23.

38. Enno A, O’Rourke JL, Howlett CR, Jack A, Dixon MF, Lee A.MALToma like lesions in the murine gastric mucosa after long-term infection with Helicobacter felis. A mouse modelofHelicobacter pylori-induced gastric lymphoma. Am J Pathol1995;147:217–22.

39. Greiner A, Marx A, Heeseman J, Lechmann J, Schmausser B,Muller Hermelink HK. Idiotype identity in a MALT-typelymphoma and B cells inHelicobacter pyloriassociated chronicgastritis. Lab Invest 1994;70:572–8.

40. Weber DM, Dimopoulos MA, Anandu DR. Regression ofgastric lymphoma of mucosa-associated lymphoid tissue withantibiotic therapy forHelicobacter pylori. Gastroenterology1994;107:1835–8.

41. Fujioka T, Shuto R, Kodama R, Fujiyarna K, Kubota T,Murakami K, et al. Experimental model for chronic gastritiswith Helicobacter pylori:long-term follow-up study inH. pyloriinfected Japanese macaques. Eur J Gastroenterol Hepatol 1993;5Suppl 1:S73–7.

42. Rokkas T, Filipe MI, Gladen GE. Detection of an increasedincidence of early gastric cancer in patients with intestinalmetaplasia type III who are closely followed up. Gut 1991;32:1110–3.

Received 24 November 1997Accepted 4 February 1998

460 S. Honda et al.

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Cal

ifor

nia

Irvi

ne o

n 10

/29/

14Fo

r pe

rson

al u

se o

nly.