Apoptosis and proliferation in Helicobacter pylori-associatedgastric intestinal metaplasia

W. K. LEUNG*, J . YU*, K. F. TO  , M. Y. Y. GO*, P. K. MA*, F. K. L. CHAN* & J. J . Y. SUNG*

*Department of Medicine & Therapeutics and  Department of Anatomical & Cellular Pathology, Chinese University of Hong

Kong, Prince of Wales Hospital, Shatin, Hong Kong

Accepted for publication 25 April 2001

INTRODUCTION

Based on compelling epidemiological evidence, the

International Agency for Cancer Research classi®ed

Helicobacter pylori as a class I carcinogen in 1994.1 As

yet, the mechanism by which H. pylori causes gastric

cancer remains elusive. Gastric cancer is generally

believed to be a multi-step progression from chronic

gastritis, atrophy, intestinal metaplasia, dysplasia and

ultimately to cancer. This cascade of histological

changes, which was originally proposed by Correa, is

still the most widely accepted gastric carcinogenesis

model.2 H. pylori infection is generally believed to be the

initial trigger of this cascade.

One of the pathways by which H. pylori is linked to

gastric carcinogenesis may be related to the disruption

in the balance between gastric epithelial cell proliferation

and cell death, or apoptosis. Because H. pylori infection

is invariably associated with in¯ammation of the

stomach, this is linked to gastric epithelial cell prolifer-

ation and cell death. Both in vitro and in vivo studies

have shown that H. pylori induced apoptosis in gastric

epithelium.3±8 Of interest, the heightened apoptosis can

be restored by eradication of the bacterium.3, 8 On the

other hand, studies on gastric proliferation yielded

con¯icting results.5, 6 While the level of proliferation

appears to parallel that of apoptosis in H. pylori-infected

SUMMARY

Background: Imbalance between apoptosis and prolifer-

ation may be one of the mechanisms underlying

H. pylori associated gastric carcinogenesis.

Aim: To examine the cell kinetics of gastric intestinal

metaplasia and the effect of H. pylori eradication.

Methods: Endoscopic gastric biopsies were obtained

from 100 H. pylori-infected patients. Apoptosis was

determined by triphosphate nick-end labelling (TUNEL)

and apoptotic nuclei counting, whereas proliferation

was assessed by Ki67 immunostaining. Gastric biopsies

were repeated in a sub-group of intestinal metaplasia

patients after H. pylori eradication.

Results: Antral apoptotic index was signi®cantly lower

in intestinal metaplasia than in non-intestinal meta-

plasia (0.19% vs. 0.51%; P < 0.0001) whereas the

level of proliferation was comparable (28% vs. 22%,

P � 0.15). Serial antral biopsies taken from 14 intesti-

nal metaplasia patients before and 1 year after H. pylori

eradication showed a signi®cant drop in proliferation

in both intestinal metaplasia (50% vs. 12%, P < 0.001)

and non-intestinal metaplasia area (47% vs. 9%,

P < 0.001). A similar fall in apoptosis was detected in

non-metaplastic region (0.58% vs. 0.38%, P < 0.001)

but not in intestinal metaplasia (0.24% vs. 0.27%,

P � 0.56), resulting in a signi®cant increase in the

apoptosis/proliferation ratio (0.005±0.021; P � 0.03).

Conclusions: Dysregulation in apoptosis control of gas-

tric intestinal metaplasia may contribute to gastric

carcinogenesis, which may be retarded by clearance of

H. pylori.

Correspondence to: Dr W. K. Leung, Department of Medicine & Therapeutics,

Chinese University of Hong Kong, Prince of Wales Hospital, Shatin,

Hong Kong.E-mail: wkleung@cuhk.edu.hk

Aliment Pharmacol Ther 2001; 15: 1467±1472.

Ó 2001 Blackwell Science Ltd 1467

human stomach, H. pylori inhibits gastric cell prolifer-

ation in vitro. The reason for this discrepancy remains

unclear. Based on cell line experiments, it is conceivable

to speculate that the hyperproliferative state seen in the

H. pylori infected stomach may be a compensatory

response to the excessive cell loss due to induction of

apoptosis.

Intestinal metaplasia is a well-recognized precursor of

gastric cancer. In this study, we examine the level of

apoptosis and proliferation in this pre-neoplastic gastric

lesion and compare it with non-metaplastic gastric

epithelium. The effect of H. pylori eradication on cell

kinetics of gastric intestinal metaplasia is also assessed.

METHODS

Patients and gastric tissues

A total of 100 patients with H. pylori associated gastritis

were studied. These patients were randomly selected

from patients enrolled in a previous H. pylori eradication

trial.9 All of these patients were infected by H. pylori as

con®rmed by positive rapid urease test (CLOtest, Tri-Med

Specialties, Charlottesville, VA) and visualization of the

bacteria on histological examination by Giemsa stain.

As stipulated by the initial protocol, they were given a

course of anti-helicobacter therapy that consisted of

omperzaole 20 mg, amoxicillin 1 g and clarithromycin

500 mg given twice daily for 1 week. None of these

patients had peptic ulcer disease or gastric neoplasm on

endoscopic examination. They were not receiving non-

steroidal anti-in¯ammatory drugs or had received

anti-Helicobacter therapy. All patients gave informed

consent for participation in this study and the study

protocol was approved by the Ethics Committee of the

Chinese University of Hong Kong.

During upper gastrointestinal endoscopy, three biop-

sies were obtained from the antrum (from the greater

and lesser curvatures 2±3 cm from the pylorus) and two

from the corpus (one from the lesser curvature and the

other from the greater curvature). One antral biopsy

was used in rapid urease test and the other four were

®xed in 10% formalin and embedded in paraf®n for

histological examination. Sections were stained by

haematoxylin and eosin and Giemsa method, and

were examined by a single experienced pathologist

(KFT). The following parameters were scored as stipu-

lated by the updated Sydney Classi®cation into 0

(normal) to 3 (marked): density of H. pylori; acute

in¯ammation; chronic in¯ammation; intestinal meta-

plasia; and atrophy.10

Patients with intestinal metaplasia underwent a follow-

up endoscopy at 1 year post anti-helicobacter treat-

ment. Clearance of H. pylori was con®rmed by negative

results in rapid urease test, histology and 13C urea

breath test. Serial gastric biopsies were available in 14

intestinal metaplasia patients with con®rmed H. pylori

eradication.

Apoptosis and proliferation

Apoptosis was determined by the terminal deoxynucle-

otidy transferase (TdT)-mediated deoxyuridine triphos-

phate nick-end labelling (TUNEL) technique and visual

counting of apoptotic nuclei. TUNEL (ApopTag; Intergen,

Purchase, NY) was performed as directed by the manu-

facturer. To increase the speci®city of the TUNEL,

apoptotic nuclei were only counted after con®rmation

by a parallel haematoxylin and eosin stained section

under high power view (400´) as described previously

(Figures 1A and B).8 Two independent investigators

(JY, MYYG) examined the slides in a random manner.

Apoptotic nuclei were only counted in sections with well-

orientated glands. Therefore, only 86 antral biopsies and

65 corpus biopsies could be analysed. The level of

apoptosis was then determined by the average of the

total apoptotic nuclei as detected by haematoxylin and

eosin staining by both investigators. The apoptotic index

was de®ned by the percentage of positive apoptotic cells

among the total number of epithelial cells (minimum of

1000 epithelial cells) counted. The apoptotic index was

scored separately for: (i) the antrum and corpus; and

(ii) intestinal metaplasia and non-intestinal metaplasia

area.

Proliferation was determined by Ki-67 immunohisto-

chemistry. The formalin-®xed and paraf®n-embedded

gastric tissue sections were de-waxed, re-hydrated, rinsed

with distilled water and washed in Tris buffered saline.

It was then followed by trypsin digestion and heat-

induced antigen retrieval. Endogenous peroxide activity

was blocked by quenching in 3% hydrogen peroxide.

Immunostaining was performed by incubation with

mouse anti-Ki-67 monoclonal antibody (MIB 1; Zymed

Laboratories, Inc; San Francisco, CA) in a moist chamber.

After washing with Tris buffered saline, the slides were

incubated with diluted broad spectrum biotinylated

secondary antibody (Zymed) and labelled-(strept)avidin-

biotin (LAB-SA) detection system (Zymed). The reactions

1468 W. K. LEUNG et al.

Ó 2001 Blackwell Science Ltd, Aliment Pharmacol Ther 15, 1467±1472

were visualized with diaminobenzidine substrate (Dako)

and counter-stained with methyl green solution

(Figure 1C). Proliferation index was counted in a manner

similar to apoptosis by two independent investigators.

Similarly, the proliferation index was scored separately

for: (i) the antrum and corpus; and (ii) intestinal

metaplasia and non-intestinal metaplasia area. In addi-

tion, the apoptosis/proliferation ratio was obtained by

dividing the apoptotic index by the proliferation index.

Statistics

Statistical analysis was performed by GraphPad Prism

(version 2.0; SanDiego, CA). Apoptosis and proliferation

index, from the antrum and the corpus, and from

intestinal metaplasia and non-intestinal metaplasia

area, was expressed as the mean � s.e. Comparison

between intestinal metaplasia and non-intestinal meta-

plasia area was made by the unpaired t-test because not

all patients had intestinal metaplasia on gastric biopsies.

The paired t-test was used in the comparison of

parameters before and after treatment in the same

patient. Statistical signi®cance was taken if P-values

(two-sided) were less than 0.05.

RESULTS

Patients

A total of 100 patients (male:female, 55:45; mean age

53 years, range 39±75) were examined. All patients

were infected with H. pylori, as con®rmed by histology

and rapid urease test, but were free of cancer and peptic

ulcers on endoscopy. Intestinal metaplasia was docu-

mented in antral biopsies from 41 patients and in

corpus biopsy from one patient.

Intestinal metaplasia vs. non-intestinal metaplasia area

Antral biopsies from non-metaplastic area had a

signi®cantly higher level of apoptosis than biopsies

taken from the corpus (0.51 � 0.04% vs.

0.14 � 0.02%; P < 0.0001; Figure 2A). A similar pat-

tern was maintained for proliferation, which was

signi®cantly higher in the antrum than in the corpus

(22.0 � 2.2% vs. 5.4 � 0.8%; P < 0.0001, Figure 2B).

There was no correlation between gastric histology

(including density of H. pylori, neutrophils and mono-

nuclear cells in®ltration, severity of atrophy and

intestinal metaplasia) and levels of apoptosis or prolif-

eration.

Within foci of intestinal metaplasia obtained from the

gastric antrum, the apoptosis index was signi®cantly

lower than that of non-metaplastic area (0.19 � 0.03%

vs. 0.51 � 0.04%; P < 0.0001, Figure 3A). For prolif-

eration, there was no signi®cant difference between

intestinal metaplasia and non-intestinal metaplasia

regions (27.8 � 3.2% vs. 22.0 � 2.2%, P � 0.15;

Figure 3B). The respective apoptotic index/proliferation

index ratio for the antrum was 0.02 � 0.01 in intes-

tinal metaplasia and 0.09 � 0.02 in non-metaplastic

area (P � 0.002). Because of the low detection rate of

Figure 1. The presence of apoptotic nuclei

in gastric epithelium as detected by (A)

haematoxylin and eosin section (400 ´)

and (B) TUNEL. The apoptotic bodies with

condensation of chromatin and fragmen-

tation of nuclei are highlighted by the

arrows. (C) Proliferation as demonstrated

by the Ki67 immunostaining. Imunoreac-

tivity against Ki67 in the glandular area

was depicted by the dark colour.

CELL KINETICS IN H. PYLORI-ASSOCIATED INTESTINAL METAPLASIA 1469

Ó 2001 Blackwell Science Ltd, Aliment Pharmacol Ther 15, 1467±1472

intestinal metaplasia from the corpus (n � 1), compar-

ison between intestinal metaplasia and non-intestinal

metaplasia in the corpus was not possible.

Changes in cell kinetics after H. pylori eradication

Serial gastric biopsies, obtained 1 year after H. pylori

eradication, were available from 14 patients with

intestinal metaplasia. Although there was no appreci-

able regression of intestinal metaplasia among these

patients (medium score 2 vs. 2; P � 0.6), there was a

signi®cant reduction in proliferation index in both

metaplastic and non-metaplastic regions, after clear-

ance of the organism (intestinal metaplasia:

50.2 � 4.0% vs. 11.8 � 2.5%, P < 0.0001; non-

intestinal metaplasia: 47.1 � 6.6% vs. 9.1 � 2.6%,

P � 0.0002). In contrast, the level of apoptosis dropped

in the non-intestinal metaplasia region (0.58 � 0.11 vs.

0.38 � 0.06, P < 0.0001) and remained unaltered in

intestinal metaplasia (0.24 � 0.06 vs. 0.27 � 0.05,

P � 0.6). Thus, the resultant apoptotic index/prolifer-

ation index ratio of intestinal metaplasia increased from

0.008 � 0.002 to 0.13 � 0.10 after H. pylori eradica-

tion (P � 0.03). A similar increase was observed in the

non-metaplastic region, but the difference did not reach

statistical signi®cance (0.01 � 0.003 to 0.05 vs. 0.02,

P � 0.07).

DISCUSSION

Although there was mounting evidence to support

H. pylori infection inducing apoptosis and proliferation

in gastric epithelium, the exact underlying mechanism

and its link with cancer remains largely unknown.3±8

The present study offered new insight into the

tumorigenesis mechanism of gastric intestinal meta-

Figure 2. The difference in apoptosis (A) and proliferation (B)

index between the antrum and corpus (P < 0.001). Figure 3. The difference in apoptosis (A) and proliferation (B)

between intestinal metaplasia and non-intestinal metaplasia

regions.

1470 W. K. LEUNG et al.

Ó 2001 Blackwell Science Ltd, Aliment Pharmacol Ther 15, 1467±1472

plasia, a pre-neoplastic lesion, from the cellular

perspective. Our results demonstrated a marked differ-

ence in cell kinetics between metaplastic and non-

metaplastic gastric epithelium. While proliferation was

increased in both metaplastic and non-intestinal

metaplasia regions, the level of apoptosis was signi®-

cantly lower in intestinal metaplasia. As a conse-

quence, the apoptotic index/proliferation index ratio

was markedly reduced in intestinal metaplasia when

compared to the adjacent non-metaplastic area, which

favours cellular accumulation and possibly neoplasm

formation.

Our ®nding was in keeping with a recent report by

Scotiniotis et al., which showed diminished apoptosis in

nine patients with gastric intestinal metaplasia.11 In

that study, apoptosis was evaluated by the TUNEL

technique, but without con®rmation by apoptotic nuclei

counting. It may overestimate the apoptotic activity,

especially in the presence of in¯ammatory cells. In

contrast, we con®rmed the results of the TUNEL assay

with visual apoptotic nuclei counting. Additionally, the

present study looked into the changes in cell kinetics of

gastric intestinal metaplasia after H. pylori eradication.

Although there was a remarkable reduction in cellular

proliferation after the eradication of H. pylori in both

intestinal metaplasia and non-metaplastic mucosa,

apoptotic activity remained unaltered in intestinal

metaplasia. The reason for this divergent response to

antibiotic treatment between intestinal metaplasia and

non-intestinal metaplasia remains unknown. However,

several recent observations in gastric intestinal meta-

plasia have shed new light on this phenomenon. First,

aberrant expression of bcl-2, an anti-apoptotic protein,

has been described in intestinal metaplasia.12 Second,

we have recently demonstrated cyclooxygenase-2

(COX-2) over-expression in H. pylori-associated intesti-

nal metaplasia.13 Because both bcl-2 and COX-2 have

been linked to resistance to apoptosis, the ®nding of

reduced apoptosis in gastric intestinal metaplasia is not

unexpected.14 In addition, despite clearance of H. pylori

in gastric intestinal metaplasia, there is persistence of

COX-2 in the gastric foveolar epithelium, which may

explain the responses of apoptosis in intestinal meta-

plasia to H. pylori eradication therapy.13 Taken together,

these ®ndings suggest the autonomous nature of

apoptosis regulation in gastric intestinal metaplasia

that is independent of in¯ammation or the presence of

H. pylori.

In this study, we failed to show any correlation

between apoptosis or proliferation with severity of

gastric in¯ammation. Although this ®nding may appear

to contradict previous studies, it is generally cognizant

that in¯ammation per se may not have any direct effect

on the level of apoptosis in gastric epithelium.11 First,

increased apoptosis is not observed in certain cases of

gastric in¯ammation, such as non-steroidal anti-in¯am-

matory drugs and Crohn's disease.6 Second, induction

of apoptosis by H. pylori was detected in cell line

experiments even in the absence of in¯ammatory cells.5

Thus, many investigators favour the importance of

H. pylori or its products rather than in¯ammation in the

induction of apoptosis. There are recent data to

implicate the involvement of Fas-mediated apoptosis

signalling pathways in this process. Jones et al. demon-

strated that infections with cagA-positive, cagE-positive

and vacA-positive H. pylori isolates enhanced the

expression of Fas receptor in gastric cells in vitro and

hence resulted in the induction of apoptosis.15 In this

context, blocking of the CD95 (APO-1/Fas) receptor, a

TNF receptor/nerve growth receptor super-family and a

type I trans-membrane protein that plays a crucial role

in the initiation of apoptosis, abolishes the induction of

apoptosis by H. pylori.16

In this study, we have demonstrated a disturbance in

cell kinetics, with diminished apoptosis and elevated

proliferation in gastric intestinal metaplasia, which

favours cell accumulation. Of note, eradication of

H. pylori results in a marked increase in the apoptotic

index/proliferation index ratio that may potentially

reverse this tendency towards tumorigenesis. This is

supported by the results of two recently published

intervention studies, which demonstrate signi®cant

regression of pre-malignant gastric lesions after eradi-

cation of H. pylori.9, 17

ACKNOWLEDGEMENTS

This study was supported by Strategic Research

Programme (SRP/9605) of the Chinese University of

Hong Kong.

REFERENCES

1 IARC Working Group on the Evaluation of Carcinogenic Risks

to Humans. Helicobacter pylori. In: Schistosomes, Liver Flukes

and Helicobacter pylori: Views and Expert Opinions of an IARC

CELL KINETICS IN H. PYLORI-ASSOCIATED INTESTINAL METAPLASIA 1471

Ó 2001 Blackwell Science Ltd, Aliment Pharmacol Ther 15, 1467±1472

Working Group on the Evaluation of Carcinogenic Risks to

Humans. Lyon: IARC, 1994: 177±240.

2 Correa P. Human gastric carcinogenesis: a multistep and

multifactorial process. First American Cancer Society Award

Lecture on Cancer Epidemiology and Prevention. Cancer Res

1992; 52: 6735±40.

3 Moss SF, Calam J, Agarwal B, et al. Induction of gastric

epithelial apoptosis by Helicobacter pylori. Gut 1996; 38:

498±501.

4 Peek RM, Moss SF, Tham KT, et al. Helicobacter pylori cagA+

strains and dissociation of gastric epithelial cell proliferation

from apoptosis. J Natl Cancer Inst 1997; 89: 863±8.

5 Wagner S, Beil W, Westermann J, et al. Regulation of gastric

epithelial cell growth by Helicobacter pylori: evidence for

a major role of apoptosis. Gastroenterology 1997; 113:

1836±47.

6 Jones NL, Shannon PT, Cutz E, et al. Increase in proliferation

and apoptosis of gastric epithelial cells early in the natural

history of Helicobacter pylori infection. Am J Pathol 1997; 151:

1695±703.

7 Rokkas T, Ladas S, Liatsos C, et al. Relationship of Helicobacter

pylori cagA status to gatric cell proliferation and apoptosis. Dig

Dis Sci 1999; 44: 487±93.

8 Leung WK, To KF, Chan FKL, Lee TL, Chung SCS, Sung JJY.

Interaction of H. pylori and NSAID on gastric epithelial cell

apoptosis and proliferation: Implications on ulcerogenesis.

Aliment Pharmacol Ther 2000; 14: 879±85.

9 Sung JJY, Lin SR, Ching JYL, et al. Atrophy and intestinal

metaplasia one year after cure of H. pylori infection: a Pros-

pective, randomized study. Gastroenterology 2000; 119: 7±14.

10 Dixon MF, Genta RM, Yardley JH, Correa P. Classi®cation and

grading of gastritis. The updated Sydney System. Inter-

national Workshop on the Histopathology of Gastritis, Hous-

ton 1994. Am J Surg Pathol 1996; 20: 1161±81.

11 Scotiniotis IA, Rokkas T, Furth EE, Rigas B, Shiff SJ. Altered

gastric epithelial cell kinetics in Helicobacter pylori-associated

intestinal metaplasia: Implications for gastric carcinogenesis.

Int J Cancer 2000; 85: 192±200.

12 Saegusa M, Takano Y, Okayasu I. Bcl-2 expression and its

association with cell kinetics in human gastric carcinomas

and intestinal metaplasia. J Cancer Res Clin Oncol 1995; 121:

357±63.

13 Sung JJY, Leung WK, Go MYY, et al. Cyclooxygenase-2

expression in Helicobacter pylori-associated premalignant and

malignant gastric lesions. Am J Pathol 2000; 157: 729±35.

14 Tsujii M, DuBois RN. Alterations in cellular adhesion and

apoptosis in epithelial cells overexpressing prostaglandin

endoperoxide synthase-2. Cell 1995; 83: 493±501.

15 Jones NL, Day AS, Jennings HA, Sherman PM. Helicobacter

pylori induces gastric epithelial cell apoptosis in association

with increase Fas receptor expression. Infect Immunol 1999;

67: 4237±42.

16 Rudi J, Kuck D, Strand S, et al. Involvement of the CD95

(APO-1/Fas) receptor and ligand system in Helicobacter pylori-

induced gastric epithelial apoptosis. J Clin Invest 1998; 102:

1506±14.

17 Correa P, Fonthan ETH, Bravo JC, et al. Chemoprevention of

gastric dysplasia: randomized trial of antioxidant supplements

and anti-Helicobacter pylori therapy. J Natl Cancer Inst 2000;

92: 1881±8.

1472 W. K. LEUNG et al.

Ó 2001 Blackwell Science Ltd, Aliment Pharmacol Ther 15, 1467±1472