Pathogenesis of the Carcinogenic Bacterium, Helicobacterpylori
by
Chung-Wei Lee
M.D., Medicine (1998)
National Taiwan University
Submitted to the Division of Biological Engineering
in Partial Fulfillment of the Requirement for the Degree of
Doctor of Philosophy Degree in Molecular and Systems Bacterial Pathogenesis
at the
ARCHvES
Massachusetts Institute of Technology
MASSACHUSETTS INSTMFrE
OF TECHNOLOGY
June 2007
AUG 0 2 2007
© 2007 Massachusetts Institute of Technolog
LIBRARIES
All rights reserved
Signature of Author
Division of Biological Engineering
Division of Comparative Medicine
May 17, 2007
A
Certified by
James G.Fox
Professor, Division of Biological Engineering
Director, Division of Comparative Medicine
Thesis Supervisor
Accepted by
Professor David B. Schauer
//i n
Accepted by
Chairman
- 1
-
Professor James L. Sherley
-
Accepted hv
1
!
/9
Professor Timothy C. Wang
Accepted by
Professor Alan J. Grodzinsky
Chairman, Graduate Committee
Pathogenesis of the Carcinogenic Bacterium, Helicobacterpylori
By
Chung-Wei Lee
Submitted to the Division of Biological Engineering on May 25, 2007 in Partial Fulfillment of
the Requirement for the Degree of Doctor of Philosophy Degree in Molecular and Systems
Bacterial Pathogenesis
ABSTRACT
Gastric cancer is the second most common malignancy in the digestive system and the second
leading cause of cancer-related death worldwide. Epidemiological data and experimental studies
have identified several risk factors for gastric cancer, including Helicobacterpylori infection,
low fruit and vegetable intake, N-nitrosoamine exposure, high salt diet, and smoking. Among
these risk factors, H. pylori infection is the major cause of gastric cancer. Therefore, H. pylori
has been classified as a type 1 (definite) carcinogen for gastric cancer by the World Health
Organization (WHO) in 1994.
H. pylori colonizes the human stomach and has been definitively linked to chronic gastritis.
Infection in some susceptible individuals results in serious gastric disease such as peptic ulcer or
gastric cancer. The first aim of this thesis was to examine the role of different T cell
subpopulations in H. pylori gastritis. Using a murine adoptive transfer model, adoptive transfer
of wildtype (wt) effector T cells (TE) into H. pylori-infected lymphopenic Rag2Y recipient mice
resulted in H. pylori-associated corpus gastritis superimposed with non-specific gastroduodenitis.
Cotransfer with TE and regulatory T cells (TR) from wt or IL10' mice reduced gastroduodenitis,
but only wt TR cells reduced corpus gastritis. The second aim of this thesis was to evaluate the
effect of vitamin C on H. pylori gastritis in vitamin C-deficient gulo/ mice. It was found that a
high vitamin C supplementation (3300 mg/L) in drinking water did not protect H. pylori gastritis,
while a low vitamin C supplementation (33 mg/L) reduced the severity of H. pylori gastritis via
an attenuated cellular immune response to H. pylori. The third aim of this thesis was to examine
the role of DNA repair in H. pylori-associated gastric disease. We found that H. pylori-associated
premalignant gastric atrophy was more severe in infected mice lacking DNA repair protein 3alkyladenine DNA glycosylase or 0 6 -methylguanine DNA methyltransferase in comparison to
infected wt control mice. The forth aim of this thesis was to examine whether antimicrobial H.
pylori eradication therapy could prevent gastric cancer development in INS-GAS mice, a model
of gastric cancer. We found that antimicrobial H. pylori eradication therapy prevented the
progression to gastric cancer in H. pylori-infected INS-GAS mice when treatment was instituted
at an early stage of H. pylori infection.
In conclusion, these studies provide further insight into the role of host immune responses in H.
pylori pathogenesis. Additionally, information was garnered regarding the roles of vitamin C
supplementation, DNA repair proteins, and H. pylori eradication therapy in H. pylori-associated
gastric disease using genetically manipulated mice.
Thesis Supervisor: James G. Fox
Title: Professor, Division of Biological Engineering
Director, Division of Comparative Medicine
ACKNOWLEDGEMENTS
First of all, I would like to thank my advisor, Dr. James G. Fox, who has made everything
possible for me at MIT. He gave me visionary guidance and an optimistic spirit even during the
most difficult time. I am very grateful for all the time and effort he has devoted to me. Dr. Fox is
Director of the Division of Comparative Medicine (DCM) and Professor in the Division of
Biological Engineering at the Massachusetts Institute of Technology, and a member of the
Institute of Medicine of the National Academies of Science. He has provided an exceptional
research and training environment, providing equipment and resources to perform experiments
that could not be accomplished at many other facilities. Finally, I want to thank Dr. Fox for
introducing me to the world of Helicobacterspp. and a variety of laboratory animal models. I am
so proud to have been mentored by Dr. Fox.
I also would like to thank my Committee Chairman Dr. David B. Schauer. He is a wonderful
mentor who always points out the strengths and weaknesses of experiments and experimental
results. Discussing the literature with Dr. Schauer is always a pleasant and fruitful experience.
Dr. James L. Sherley is an expert in adult stem cells and cell kinetics. He has generously spent
his precious time with me to practice proposal writing. He has encouraged me to think deeply
about my studies. I appreciate that Dr. Sherley has prayed for my success in my thesis work. I am
grateful to Dr. Timothy C. Wang for his advice and insightful discussions regarding the clinical
importance of H.pylori infection and gastric cancer. Dr. Wang has guided my research direction
and been a supportive collaborator of Dr. Fox's lab. He is an expert in gastric cancer and cancer
stem cells. Without Dr. Wang's INS-GAS mouse model, I would not been able to finish part of
my thesis work.
I would like to thank Dr. Leona D. Samson and Dr. Lisiane B. Meira for providing Aag~'~ and
Mgmf'~ mice to examine the roles of these two DNA repair proteins in H. pylori gastritis. I would
also like to thank Dr. Pete C. Dedon and his post-docs Drs. Bo Pang and Xinfeng Zhou. Their
tremendous knowledge in DNA adduct formation and repair, and insightful discussions have
helped us to delineate future research directions for the mechanism of H. pylori carcinogenesis. I
would like to thanks Dr. Xiang-Dong Wang, Professor at the Nutrition and Cancer Biology
Laboratory and the Jean Mayer USDA Human Nutrition Research Center on Aging at Tufts
University. He has shared his knowledge in vitamin C and expertise in vitamin C measurements.
He has spent time discussing the experimental design and results with us, providing a clear
direction in future studies of vitamin C and its role in H. pylori gastritis. I am grateful to Dr.
Kuo-Liong Chien, a colleague of mine and Professor at the Institute of Preventive Medicine at
the National Taiwan University. Dr. Chien has shared his statistics expertise in analyzing our
vitamin C studies. Dr. Andrea Varro, Professor of Physiology at the University of Liverpool,
helped by measuring gastrin levels. Dr. David E. Cohen, Professor of Medicine at Harvard
Medical School and Director of Hepatology at Brigham and Women's Hospital, has shared his
expertise in lipid metabolism and provided equipment and resources to perform experiments. His
help has provided us the chance to explore the mechanisms of H. pylori-associated gastric and
extra-gastric pathologies.
Dr. Arlin B. Rogers and Dr. Barry Rickman are great researchers, pathologists, and friends. I can
always knock on their door and get answers to solve my puzzle. Dr. Mark T. Whary is a
wonderful and patient mentor. He has done a great job critiquing my scientific writing. Dr. Susan
E. Erdman and Dr. Varada P. Rao have always provided important advice in my experimental
design. I appreciate that they have tolerated my scientific "flight of thought". Nancy S. Taylor is
a wonderful microbiologist and listener. She trained me how to grow Helicobacterspp. as well
as how to cultivate faith in God. I am grateful to Dr. Yan Feng, Zhongming Ge, Zeli Shen, and
Sandy Xu, who have shared their scientific expertise and experience throughout my thesis work.
I would like to thank Kathy Cormier for her professional histological preparations and
techniques. Additionally, there are many hard-working DCMers who have helped take care of the
laboratory animals in supporting these pioneering studies. Their efforts help make DCM the best
facility in the world for animal studies. I would like to thank those I neglected to mention for all
of their efforts in helping me reach my goals.
I would like to thank my parents and in-laws. My parents have cultivated my curiosity in the
natural sciences since my childhood. Their influences on my early development cast my life-long
interest in seeking and finding the truths behind nature. My in-laws are not traditional Taiwanese
parents who need to keep children nearby; instead, they have provided my family every
opportunity and the best support for my pursuits of science.
I am grateful to God, the greatest creator, for opening a window for me to peek into this
wonderful world. I also thank God for blessing us with Jean and John, my two lovely children,
while I was a graduate student. Their smiling faces, sweet voices, and heartfelt prayers have
encouraged me throughout my studies.
Most importantly, I am grateful to my wife Jui-Lin. In 2000, she gave up her position as a child
psychiatrist in Taiwan including her promotion to Director of the department, and came to the
United States with me. She continued to sacrifice her academic pursuit of advanced studies to
provide the best care to our family, especially Jean and John. Her prayers have always supported
me. I could not have completed this degree without her love. I am so fortunate to have such a
wonderful wife. I dedicate this MIT Ph.D. to her.
Table of Contents
Title page
Abstract
Acknowledgements
Table of Contents
List of Figures
List of Tables
Chapter 1 - Background and literature review
Chapter 2 - Wildtype and IL1O~'~ T regulatory cells reduced T effector cell-mediated
gastroduodenitis in Rag2~'~ mice but only wildtype T regulatory cells
suppressed Helicobacter pylori gastritis.
Chapter 3 - High vitamin C intake does not protect vitamin C-deficient L-gulono-ylactone oxidase knockout mice from Helicobacterpylori-induced
premalignant gastric lesions.
Chapter 4 - Dysfunction of DNA repair genes 3-alkyladenine DNA glycosylase and
0 6 -methylguanine DNA methyltransferase enhanced development of
preneoplastic oxyntic gland atrophy in the stomach of chronically
Helicobacterpylori-infected mice.
Chapter 5 - Helicobacterpylori eradication prevents progression of gastric cancer in
hypergastrinemic INS-GAS Mice
Chapter 6 - Summary
1
3
5
8
14
15
16
55
84
120
144
181
Appendix 1 - Lee, CW., Rao, VP., Rogers, AB., Ge, Z., Erdman, SE., Whary, WT., and Fox,
JG. Wild-Type and Interleukin-10-Deficient Regulatory T Cells Reduce Effector T-CellMediated Gastroduodenitis in Rag2~'~ Mice, but Only Wild-Type Regulatory T Cells Suppress
Helicobacter pylori Gastritis. Infect Immun 2007 Jun;75(6):2699-707
Table of Contents
Chapter 1: Background and literature review
Gastric cancer
Pathology
Epidemiology
Risk factors
Helicobacterpylori
History
Animal models of H. pylori infection
Virulence factors and pathogenesis of H.pylori
Host immune responses to H. pylori
Persistent infection, chronic inflammation, and cancer
Diet and environmental factors
Ascorbic acid
N-nitroso compounds
Host genetic factors
Polymorphisms of inflammation-related genes
Polymorphisms of DNA repair genes
Intervention of H. pylori infection and H. pylori-relateddisease
Antimicrobial H. pylori eradication therapy
Other chemopreventive drugs
References
17
18
19
19
21
24
26
31
33
35
37
38
40
41
44
Chapter 2: Wildtype and IL10'" T regulatory cells reduced T effector cell-mediated
gastroduodenitis in Rag2-'' mice but only wildtype T regulatory cells suppressed
Helicobacterpylorigastritis
Abstract
Introduction
Materials and Methods
Experimental groups
Helicobacterpyloriinfection
Cell sorting and adoptive transfer
Histological evaluation
Special stains and immunohistochemistry
RNA extraction and quantitative PCR for cytokine expression and Foxp3
Quantitative culture of H. pylori
Statistical analysis
Results
Rag2-'- TE-recipient mice developed morbidity and gastroduodenitis independent
of H. pylori infection
Adaptive immunity is required to develop Hp-associated gastritis
56
57
58
59
59
60
60
61
62
62
62
63
Hp infection exacerbated corpus gastritis in Rag2~'~ TE-recipient mice
Adoptive transfer of wt or IL10~'~ TR cells reduced morbidity and gastroduodenitis
but only wt TR cells reduced severity of Hp corpus gastritis
Th1 cytokine responses were down-regulated to a greater extent by wt than IL10~'TR cells
Foxp3 expression in gastric tissues was promoted to a greater extent by wt than
IL10~'~ TR cells
Hp colonization levels were reduced by TE cells and maintained when wt TR cells
were co-transferred
Discussion
Acknowledgements
References
Legends
Figures
Table
64
65
65
66
67
67
72
73
76
78
82
Chapter 3: High vitamin C intake and normal vitamin C levels in plasma and gastric tissue
do not protect vitamin C-deficient L-gulono-y-lactone
oxidase knockout mice from
Helicobacterpylori-induced premalignant gastric lesions
Abstract
Introduction
Materials and Methods
Mice
Experimental design
Experimental infection with H. pylori
Vitamin C measurements
Histological evaluation
RNA extraction and quantitative PCR for cytokine mRNA
Plasma IgG isotypes measurement
Quantitative PCR for H. pylori colonization
Quantification of plasma levels of gastrin
Statistical Analysis
Results
Vitamin C content in plasma and gastric tissues correlated with vitamin C
supplementation levels in gulo~'~ mice
High vitamin C supplementation did not reduce H. pylori gastritis
Less severe gastritis in H. pylori-infected gulo'- mice supplemented with low
vitamin C was associated with lower gastric mRNA levels of IFN-y and TNF-a
Gulo~'~ mice supplemented with low vitamin C had lower H. pylori-specific IgG2c
responses
H. pylori colonization levels were higher in gulo-'- mice supplemented with low
vitamin C at 32 WPI
Gastrin amidation was not impaired in vitamin C-supplemented gulo-~ mice
Discussion
Acknowledgements
References
Legends
Figures
Tables
98
99
105
106
110
113
118
Chapter 4: Dysfunction of DNA repair genes 3-alkyladenine DNA glycosylase and O6
methylguanine DNA methyltransferase enhanced development of preneoplastic oxyntic
gland atrophy in the stomach of chronically Helicobacterpylori-infectedmice.
Abstract
Introduction
Materials and Methods
Mice
Bacteria and experimental infection
Experimental design
Histological evaluation
Serum IgG isotype measurement
Quantitative PCR for H. pylori colonization
Measurements of DNA adducts
Statistical Analysis
Results
H. pylori infection resulted in more severe oxyntic gland atrophy and hyperplasia
in Aag~'~ and Mgmt'~ mice than in wt mice
Aag~'~ and Mgmt~' mice developed H. pylori-specific, Th1-predominant IgG2c
responses comparable to infected wt mice
H. pylori colonization levels were comparable between Aag~'~ or Mgmt- mice and
their wt counterparts
DNA adducts in gastric tissues
Discussion
Acknowledgements
References
Legends
Figures
Tables
121
122
124
124
124
125
125
125
126
126
127
128
128
128
129
132
132
136
138
141
Chapter 5: Helicobacter pylori Eradication Prevents Progression of Gastric Cancer in
Hypergastrinemic INS-GAS Mice
Abstract
Introduction
Materials and Methods
Mice
Experimental design
Tissue collection and histological analysis
Confirmation of H. pyloi eradication by quantitative PCR
Serum H. pylori-specific antibodies
Quantitative analysis of mRNA expression
Measurements of DNA adducts
Analysis of lipids in serum
Statistical analysis
Results
Eradication of H. pylori was confirmed in INS-GAS mice that received
eradication therapy
H. pylori infection promoted developed of gastric cancer in INS-GAS mice
H. pylori eradication at 8 WPI reduced gastritis and premalignant lesions to the
greatest extent
H. pylori eradication at 8 WPI prevented progression to low-grade and high-grade
gastrointestinal intraepithelial neoplasia
H. pylori-specific antibody responses
Gastric IFN-y, TNF-a, and iNOS mRNA levels were reduced in all mice that
received H.pylori eradication therapy
Gastric mucosal cell proliferation was also reduced in all mice that received H.
pylori eradication therapy
DNA adducts in gastric tissues
H. pylori-infected INS-GAS mice had significantly higher serum levels of
triglyceride, total cholesterol, and elevated fraction of very-low-density
lipoprotein
Discussion
Acknowledgements
References
Legends
Figures
Tables
Chapter 6: Summary
145
146
148
148
148
149
150
150
151
151
152
153
153
154
155
156
157
158
159
159
161
166
168
171
175
179
181
Appendix 1:
Lee, CW., Rao, VP., Rogers, AB., Ge, Z., Erdman, SE., Whary, WT., and Fox, JG..
Wild-Type and Interleukin-10-Deficient Regulatory T Cells Reduce Effector T-CellMediated Gastroduodenitis in Rag2~'~ Mice, but Only Wild-Type Regulatory T Cells
Suppress Helicobacter pylori Gastritis. Infect Immun 2007 Jun;75(6):2699-707
List of Figures
1-1
1-2
2-1
2-2
2-3
2-4
3-1
3-2
3-3
3-4
3-5
3-6
3-7
4-1
4-2
4-3
4-4
5-1
5-2
5-3
5-4
5-5
5-6
5-7
H. pylori activates NF-KB in epithelial cells and macrophages.
Pathogen-host interactions in the pathogenesis of Helicobacterpyloriinfection.
Transfer of wildtype (wt) T effector (TE) cells into Rag2~'- mice resulted in
immune-mediated pangastritis and gastroduodenitis.
Histopathology of H. pylori (Hp)-infected Rag2'~ mice cotransferred with wt TE
cells alone or in combination with wt or IL10~'~ T regulatory (TR) cells.
Expression levels of mRNA for IFN-y, TNF-a, IL4, IL10, and Foxp3.
Mean H. pylori (Hp) colonization in the corpus and antrum.
Vitamin C levels in plasma and gastric tissue.
Gastric histopathology of H. pylori infection in gulo~'~ and wt C57BL/6 controls
Histological scores and total gastric indices in H.pylori (Hp)-infected mice.
Gastric mRNA levels of IFN-y and TNF-a in gulo'~ and wt mice at 32 WPI.
H. pylori-specific IgG2c and IgG1 levels in H. pylori-infected mice.
H. pylori colonization levels.
Plasma levels of amidated gastrin at 32 WPI.
Histopathology of gastric disease and mucous metaplasia induced by H. pylori
infection in the fundic mucosa of C57BIU6 wt or Aag-'~ mice.
Serum IgG1 and IgG2c responses to H. pylori in Aag~'~ or Mgmf'~ mice.
H. pylori colonization levels in the stomach.
Levels of etheno-dA and etheno-dC in gastric DNA
Histopathology of gastric disease induced by H. pylori infection in the corpus
mucosa of INS-GAS mice with and without eradication therapy.
Histological scores and incidences of non-gastrointestinal intraepithelial
neoplasia
Relative mRNA levels of IFN-y, TNF-a, and iNOS in the gastric tissue.
H. pylori-specific, Thl-associated IgG2a levels.
Immunohistochemical staining of Ki67 in the corpus.
DNA adducts of 2'-deoxyinosine (dI), N6-ethenodeoxyadenosine (etheno-dA),
and 8-oxo-7,8-dihydro-2'-deoxyguanosine (8-oxodG) in the stomach
Triglyceride and total cholesterol concentrations and distribution of lipoprotein
cholesterol in serum.
27
28
78
79
80
81
113
114
115
116
116
117
117
138
139
139
140
175
176
177
177
177
178
178
List of Tables
1-1
2-1
3-1
3-2
4-1
4-2
5-1
5-2
Summary of rodent models of helicobacter gastritis and gastric cancer.
Gastric lesions at 16 and 20 weeks post H. pylori (Hp) infection (WPI).
H. pylori infection at 16 weeks post infection.
Number of H. pylori-infected mice at 32 WPI.
Gastric lesions in H. pylori-infected wild type (wt), Aag~' and Mgmt'~ mice.
Gastric lesions in H. pylori-infected wild type (wt), Aag~- and Mgmt'~ mice.
Numbers of mice in each treatment group.
Gastric lesions at 22 and 28 weeks post H.pylori infection (WPI).
23
82
118
118
141
142
179
179
Chapter 1: Background and literature review
Gastric cancer
Pathology
Epidemiology
Risk factors
17
18
19
Helicobacterpylori
History
Animal models of H. pylori infection
Virulence factors and pathogenesis of H. pylori
Host immune responses to H.pylori
Persistent infection, chronic inflammation, and cancer
Diet and environmental factors
Ascorbic acid
N-nitroso compounds
Host genetic factors
Polymorphisms of inflammation-related genes
Polymorphisms of DNA repair genes
Intervention of H. pylori infection and H. pylori-relateddisease
Antimicrobial H.pylori eradication therapy
Other chemopreventive drugs
References
19
21
24
26
31
33
35
37
38
40
41
44
§ Gastric cancer
Pathology
Adenocarcinoma (referred to as gastric cancer in this paper) accounts for about 85% of gastric
malignancies (Harrison's Principles of Internal Medicine,
16 th
ed, McGraw-Hill). Gastric cancers
are classified as cardiac and noncardiac according to anatomical sites. Histologically, gastric
cancers are subdivided into two types: an intestinal type characterized by well-differentiated
neoplasitic cells that form a tubular, glandular structure mimicing the intestinal glands and a
diffuse type characterized by profound cell infiltration and poorly-differentiated cells lacking any
glandular structure [1, 2]. The intestinal type is more common in males and older age groups. In
contrast, the diffuse type is the same in both genders and more common in younger age groups.
There has been no germline mutation linked to the intestinal type of gastric cancer. The diffuse
type of gastric cancer is sometimes familial in distribution and has been associated with germline
mutations of E-cadherin (CDH1) [3]. Thus, different etiologic factor(s) may be involved in each
type of gastric cancer [4]. Histologically, the intestinal type of gastric cancer is usually found in
areas of tissue with chronic inflammation and atrophy. Correa proposed a human model of
gastric carcinogenesis with sequential changes beginning with superficial gastritis and
progressing to chronic inflammation, atrophic gastritis, intestinal metaplasia, dysplasia, and
finally gastric cancer [5].
Epidemiology of gastric cancer
Gastric cancer is the second most common malignancy in the digestive system and the
14th
leading cause of death worldwide [6]. The incidence and mortality of gastric cancer have
declined sharply in the past few decades. For example, there was an approximate 50% decrease
in the incidence of gastric cancer and gastric cancer-related mortality in the United States
between 1975 to 2002 (http://crchd.cancer.gov/definitions/statistics.html). Despite this, gastric
cancer remains the second leading cause of cancer-related death worldwide and is still a serious
epidemiological problem in the 210t century [7]. It has been estimated that there were more than
870,000 deaths from gastric cancer in the year 2000, accounting for approximately 12% of all
cancer-related mortality [4]. The survival rate for gastric cancer is poor and depends on the stage
of disease at the time of diagnosis. Because early gastric cancer usually does not present
clinically, screening and treatment are difficult. Thus, the stage of gastric cancer at diagnosis is
usually advanced, making the prognosis poor with an overall 5-year survival rate less than 25%
[8].
There is a marked geographic variation in the incidence of gastric cancer. Gastric cancer
incidence is relatively low in Western industrialized countries such as the United States and the
United Kingdom. Despite the decrease in gastric cancer incidence worldwide, the incidence of
gastric cancer remains high in Japan, China, and Chile [4]. The International Agency for
Research of Cancer (IARC) reported in 1997 that the age-adjusted incidence of gastric cancer
death in males ranged from the lowest in Caucasians in the United States (7.5 per 100,000) to the
highest in Yamagata, Japan (95.5 per 100,000) [7].
There are significant differences in the risks of gastric cancer in different ethnic groups in the
same regions [7]. In the United States, gastric cancer incidence in blacks was approximately
twice that in whites between 1998-2002 (Ries LAG, Eisner MP, Kosary CL, Hankey BF, Miller
BA, Clegg L, Mariotto A, Feuer EJ, Edwards BK (eds). SEER Cancer Statistics Review, 1975-
2002, National Cancer Institute. Bethesda, MD, http://seer.cancer.gov/csr/1975_2002/,based on
November 2004 SEER data submission, posted to the SEER web site 2005.). Gender variations
in the incidence of gastric cancer have also been observed; both cardia and noncardia gastric
cancers are more common in males than females [7].
Risk factors
Epidemiological data and experimental studies have identified risk factors for gastric cancer,
including Helicobacterpylori infection, N-nitrosoamine exposure, high salt diet, low fruit and
vegetable intake, and smoking [9]. Since H. pylori was first reported in 1983 [10], the role of this
microorganism in gastric inflammation and carcinogenesis has been extensively studied.
Helicobacter pylori, which infects approximating 50% of the population worldwide [11], is
highly associated with gastric carcinogenesis. It has been demonstrated in an international study
that there is a direct correlation between high incidence of gastric cancer and high prevalence of
H. pylori infection [12]. Therefore, H. pylori has been classified as a type 1 (definite) carcinogen
for gastric cancer by the World Health Organization (WHO) in 1994 [13]. Interestingly, the
incidence of gastric cancer is low in some areas of Africa despite a high prevalence of H. pylori
infection. This has been referred to as the "African enigma" [14].
§ Helicobacterpylori
History
Spiral-shaped microorganisms have been observed in gastric tissue for more than a century [15,
16]. Doenges reported gastric spiral bacteria in 43% of human samples and 100% of rhesus
macaque samples in 1939 [16]. In a 1975 report, spiral-shaped bacteria were observed in
inflamed human gastric mucosa and were associated with polymorphonuclear leukocyte
infiltration [17]. A gram-negative, spiral-shaped, microaerophilic, and flagellate microorganism
was first isolated and identified in the antrum of patients with chronic gastritis, gastric ulceration
and duodenal ulceration by Marshall and Warren in 1982 [10]. This bacterium was initially
named Camphylobacterpyloridis but was later renamed Camphylobacterpylori [18]. Due to its
differing 16S rRNA gene sequence and sheathed flagella, nomenclature was further changed in
1989 to Helicobacterpylori to reflect a new genus [19]. H. pylori experimental infection in
human volunteer, who had histologically normal gastric mucosa before infection, resulted in
acute pyloric gastritis within 10 days post infection [20]. These findings were subsequently
reproduced when an individual dosed himself with H. pylori and developed chronic gastritis as a
result of the bacterial infection [21]. Thus, H. pylori satisfies Koch's postulates as a causative
microbe of human gastritis.
H. pylori infection is mainly acquired in childhood and usually persists for life within the gastric
mucosa [22]. Many gastric diseases, including gastritis, peptic ulcer, atrophic gastritis, intestinal
metaplasia, gastric adenocarcinoma, and gastric mucosa-associated lymphoid tissue lymphoma
(MALToma), have been linked to H. pylori infection epidemiologically in humans and
experimentally in mice and other laboratory animal models [9]. Gastric cancer develops
predominantly in H. pylori-infected humans. Among these infected individuals, those with
duodenal ulcer are not at risk to develop gastric cancer; those with gastric ulcer or corpuspredominant gastritis, severe gastric atrophy, or intestinal metaplasia are at increased risk of
gastric cancer [23]. Most infected humans remain asymptomatic; only 15% will develop peptic
ulcers and less than 1%will develop gastric cancer [24].
Animal models of H. pylori infection
Neonatal gnotobiotic piglets were the first reported animal model of H. pylori infection [25]. H.
pylori was also reported to colonize and cause gastritis in Japanese monkeys [26] and Mongolian
gerbils [27]. In early experiments, H. pylori did not consistently colonize mice and other
laboratory animals [27, 28]. As a result, Helicobacterfelis, a gastric helicobacter closely related
to H. pylori that reliably colonized mice, was used and continues to be used to study
helicobacter-associated gastric disease in mouse models [29]. H. felis causes oxyntic gland
atrophy, and gastric cancer in C57BIJ6 mice by 15 months of infection, [30, 31]. A standardized
mouse model of H. pylori experimental infection was introduced in 1997 [32]. H. pylori Sydney
strain (SS1) contains the entire cytotoxin associated gene pathogenicity island (cag-PAI),
reproducibly colonizes the inbred mouse strains C57BIJ6, BALB/c, DBA/2, and C3H/He, and
causes chronic active gastritis in C57BIU6 mice [32]. Additionally, guinea pigs colonized with H.
pylori Sydney strain SS1 developed antral gastritis and mucosa associated lymphoid tissue
(MALT) [33]. Experimental infection with another rodent-adapted, cag-PAI positive H. pylori
strain, B128, resulted in gastritis, ulceration, and atrophy in Mongolian gerbils [34] and our
substrain of B128 caused rapid onset of gastric cancer in male Mongolian gerbils [35]. This
strain of H. pylori also and promoted gastric carcinogenesis in INS-GAS mice [36]. Recently, a
third mouse-adapted H. pylori strain, Sydney 2000, that lacks the entire cag-PAI was reported to
induce chronic gastritis in mice [37]. (Table 1)
Due to the difficulty of longitudinally following the carcinogenic process of gastric cancer in
humans, various rodent models have been used to study the pathogenesis of helicobacter-
mediated gastric disease [38, 39]. H. pylori infection in Mongolian gerbils recapitulated gastric
inflammation in humans and resulted in the sequential changes proposed by Correa [5, 40-42].
H.pylori (SS1) infection in C57BIJ6 mice results in chronic and atrophic gastritis, but H. pylori
infection alone does not reliably induce gastric cancer in these mice [37]. In contrast, H. pylori
experimental infection has been used to study helicobacter-associated gastric carcinogenesis in
the susceptible INS-GAS transgenic mice (on a FVB background) [43] and in B6129 mice [44].
(Table 1)
Table 1. Summary of rodent models of helicobacter gastritis and gastric cancer
Rodent
C57BIJ6 mice
C57BIJ6 mice
INS-GAS FVB
mice
B6129 mice
BALB/c mice
Infectious agent
H. pylori
H.felis
H.pylori and H.felis
Disease
Atrophic gastritis
Gastric cancer
Gastric cancer
Comment
No cancer
18-24 MPI
7 MPI
References
[31, 35]
[28]
[41]
H. pylori
Helicobacterspp.
15 MPI
18-24 MPI
[42]
[32]
Mongolian gerbils
H. pylori
Gastric cancer
Gastric MALT
lymphoma
Gastric cancer
[5, 38-40]
Guinea pigs
H.pylori
Gastritis
In combination
with MNU or
MNNG
15 WPI
[32]
Virulence factors and pathogenesis of H. pylori
The H. pylori genome (1.65 Mbp) encodes about 1500 proteins [45, 46]. The genome changes
continuously during chronic colonization in an individual host by importing small pieces of
foreign DNA from other H. pylori strains by natural transformation [47] during persistent or
transient mixed infection [48-51], theoretically facilitating adaptation of the organism to the
gastric mucosa. The acidic environment, secretory antimicrobial peptides, and mucosal barrier
provide protection for gastric tissue against bacterial infection [52]. However, after being
ingested, H. pylori can evade the acidic, bactericidal environment of the gastric lumen by
entering the mucosal layer. It does this by using urease to hydrolyze urea into carbon dioxide and
ammonia, thereby neutralizing the acidic milieu and allowing colonization [27, 53]. In
colonizing the gastric mucosa, multiple bacterial surface proteins aid the organism's ability to
attach tightly to epithelial cells [52]. For example, the well characterized adhesins, BabA and
SabA, bind to the fucosylated Lewis B blood group antigen and sialyl-dimeric-Lewis x
glycosphingolipid, respectively, on the gastric epithelial surface [54, 55]. Additionally, flagella of
H. pylori are essential for motility and facilitate colonization of the gastric mucosa [56].
Most strains of H. pylori produce the vacuolating cytotoxin VacA, a secreted exotoxin [52]. The
toxin is targeted to the cell membrane as well as to mitochonidrial membranes, where it forms an
anion-selective, voltage-gated channel [57]. Cytosolic bicarbonate, organic anions and
mitochondrial cytochrome c can be released through the VacA channel to induce apoptosis and
subsequently increase the turnover rate of gastric epithelium [58-60]. The role of VacA in the
pathogenesis of H. pylori is unclear. Although a VacA-positive strain of H. pylori was reported to
have some growth advantage in initial host colonization over a VacA-negative isogenic strain,
VacA was not essential for H. pylori survival and had no effect on gastric inflammation [61].
Additionally, both VacA-negative and -positive H. pylori strains have been isolated from humans
[62]. Some VacA-positive H. pylori strains are associated with more severe disease in Western
countries, but this association has not been noted in Asian countries [63]. These reports suggest
that VacA may facilitate the transmission of H. pylori by promoting the initial colonization in the
host. Additionally, VacA blocks T cell proliferation by inducing a GUS cell cycle arrest and
results in down-regulation of IL-2 transcription [64], suggesting VacA-positive H. pylori strains
evade host immune response and may reduce H. pylori-associated gastric inflammation.
Some strains of H. pylori have a pathogenicity island containing a cytotoxin-associated antigen A
(cag-PAI). Pathogenicity islands are a distinct class of linked genes on the bacterial chromosome
that correlate with pathogenicity [65]. The H. pylori cag-PAI is a 37-kb genomic fragment
containing 29 genes and encoding a type IV secretion system [66]. Cytotoxin-associated antigen
A (CagA) is a marker for the cag-PAI [66]. Most H. pylori strains are classified into two types:
type I strains possess an entire cag-PAI and have been associated with more severe gastric
disease; type II strains do not contain a complete cag-PAI and have been associated with milder
gastric lesions [52].
H. pylori delivers the CagA protein into gastric epithelial cells through the type IV secretion
system encoded by cag-PAI [67]. Translocation of CagA is essential for the induction of IL-8 upregulation, formation of stress fibers, and a growth change phenotypically similar to that induced
by hepatocyte growth factor [68, 69]. Additionally, CagA is tyrosine phosphorylated within the
epithelium at the EPIYA domains by Src family kinases. Phosphorylated CagA binds to SHP-2
tyrosine phosphatase and provokes abnormal activation of SHP-2 and Erk kinase [70, 71].
Perturbation of the SHP-2 oncoprotein and other signaling molecules by CagA may predispose
epithelial cells to gastric cancer progression [72]. Polymorphisms in the EPIYA domain of CagA
are associated with differences in the magnitude and duration of CagA activities [73]. CagA
polymorphisms may be useful markers to classify H. pylori strains into "benign" and
"malignant" strains [74]. Due to the in vitro data showing that H. pylori CagA is essential for
bacteria-mediated transformation of host cells [68, 72], the understanding of CagA
polymorphisms may help to determine clinical guidelines for H. pylori treatment [73]
Host immune responses to H. pylori
H. pylori induces expression of proinflammatory cytokines including IL-8 through activation of
NF-idB in epithelial cells and macrophages [75, 76]. H. pylori-induced NF-KB activation in
epithelial cells is mediated by NODI and dependedent on a functional cag-PAI and bacteriaepithelia contact. In contrast, H. pylori-induced NF-KB activation in macrophages is mediated by
the TLR4 pathway and independent of cag PAI or contact [77, 78] (Figure 1). In the acute phase
of H. pylori infection, bacterial antigens stimulate maturation of antigen presenting cells like
macrophages and dendritic cells, which then induce the differentiation of naive T lymphocytes
into ThI cells [79].
Figure 1. H. pylori activates NF-KB in epithelial cells and macrophages via different
mechanisms.
H. pylori activates epithelial cells depending on direct contact and through NODI pathway. H.
pylori activates macrophages in a contact-independent manner and through TLR4 pathway [77,
78]. NF-KB activation is the common down-stream pathway that medicates H. pylori activation
of epithelial cells and macrophages.
cag independent
eag dependent
p_4CDIp
Ontact
TLR4
NODI
IRAK
TRAF2, 6
TRAF6
Epithelial
Macrophage cell
tKK
liB
Maeda S, et al, 2001
NF-KB activation
Figure 1.
Though innate immunity is activated during H. pylori infection, the innate immune system alone
is not sufficient to induce H. pyloi-associated gastritis. H. pylori-infected wildtype (wt)
C57BIJ6 mice, that have functional B and T cells, develope robust helicobacter gastritis [32].
Lymphopenic B6.129S7-RaglmJom and B6.CB17-PrkdcC"'/SzJ (SCID) mice, and T celldeficient B6.129P2-TcrblomTcrd"om mice are colonized at a high density with H. pylori and
H. felis, but developed only minimal gastritis [80, 81). Conversely, H. pylori-infected, B celldeficient C57BU6-Igh-6"mlcs" mice developed identical gastric lesions observed in infected wt
C57BIU6 mice [81]. These data indicate that T lymphocytes are critical in inducing helicobacterassociated gastric inflammation.
H. pylori induces a Thl-predominant cellular immune response in humans and mice
characterized by activation of CD4* T helper lymphocytes, production of pro-inflammatory Th I
cytokines such as IFN-y, and down-regulated Th2 cytokine expression such as IL-4 [82, 83].
Reconstitution of T lymphocytes in helicobacter-infected SCID mice with Thl-predominant
lymphocytes induced more severe gastritis than those reconstituted with Th2-predominant
lymphocytes [84, 85]. (Figure 2)
Figure 2. Pathogen-Host Interactions in the Pathogenesis of HelicobacterpyloriInfection.
The host response to H. pylori participates in the induction of damage to the gastric epithelium
and therefore has an integral role in H. pylori pathogenesis. During the early phase of the
infection, binding of H. pylori to gastric epithelial cells, in particular through BabA and by
strains harboring the cag pathogenicity island, results in the production of interleukin-8 and other
chemokines, such as epithelial-cell-derived neutrophil-activating peptide 78 (ENA-78) and
growth-related oncogene r.r(GRO-tr), by epithelial cells. Nuclear factor-B (NF-nB) and the earlyresponse transcription-factor activator protein 1 (AP-1) are the intracellular messengers involved
in this process. The chemokines secreted by epithelial cells bind to the proteoglycan scaffolding,
generating a gradient along which polymorphonuclear cells (PMN) are recruited. The chronic
phase of H. pylori gastritis combines an adaptive lymphocyte response with the initial innate
immune response. Lymphocyte recruitment is facilitated by chemokine-mediated expression of
vascular addressins such as vascular-cell adhesion molecule 1 (VCAM-1) and intercellular
adhesion molecule 1 (ICAM-1) that are required for lymphocyte extravasation. Macrophages that
participate in interleukin-8 production produce proinflammatory cytokines involved in the
activation of the recruited cells, in particular T helper cells (ThO, Thl, Th2), that respond with a
biased Th1 response to H. pylori. In turn, Thl-type cytokines such as interferon-7 (INF-T) induce
the expression of class II major histocompatibility complexes (MHC) and accessory molecules
B7-1 and B7-2 by epithelial cells, making them competent for antigen presentation. The
cytotoxin VacA- and Fas-mediated apoptosis induced by tumor necrosis factor m(TNF-2) leads to
disruption of the epithelial barrier, facilitating translocation of bacterial antigens and leading to
further activation of macrophages. Cytokines produced by macrophages can also alter the
secretion of mucus, contributing to H. pylori-mediated disruption of the mucous layer. Cytokines
produced in the gastric mucosa induce changes in gastric-acid secretion and homeostasis (dashed
lines). TNF-x, interleukin-1 3 , and interferon-T increase gastrin release, stimulating parietal and
enterochromaffin cells and thus acid secretion. TNF-q also induces a decrease in the number of
antral D cells, leading to decreased somatostatin production and indirectly enhancing acid
production. LPS denotes lipopolysaccharide [52].
Figure 1.Pathogen-Host Interactions in the Pathogenesis of Helicobacterpylori
Infection. (Suerbaum and Michetti, N Engl J Med, 2002)
Recent data has demonstrated that different subpopulations of CD4* T lymphocytes play diverse
roles in mediating and regulating H. pylori-induced gastritis. In SCID mice, adoptive transfer of
wt CD4+CD45RBHi T effector
(TE)
cells from naYve donors causes severe gastritis in H. pylori-
infected recipients, while co-transfer of wt CD4*CD45RB"' T regulatory
(TR)
cells protects
against development of gastritis [84]. TR cells have also been defined by expression of the IL-2
receptor a chain and Foxp3 [86], the forkhead transcription factor critical for thymic selection of
CD4*CD25*
TR
cells [87]. Depletion of CD25*Foxp3* TR cells in H. pylori-infected C57BL/6
mice is reported to cause loss of immune regulation and more severe gastritis [88] as does
adoptive transfer of lymphocytes depleted of CD4*CD25* cells into H. pylori-infected B6.CgFoxnl"u (nu/nu) recipients [89]. Of the many T cell subsets with ascribed regulatory function
[90], cell sorting experiments commonly use CD4*CD25*CD45RB" as naturally occurring
TR
cells. However, the mechanism(s) for regulation by this type of TR cell are not fully understood.
Reconstitution of CD4+CD25-CD45RB i
TE
cells and CD4+CD25*CD45RB'O
TR
cells in H.
pylori-infected B6.129S6-Rag2m1Fwa (Rag2'~) mice demonstrates that wt TE cells mediate a
gastroduodenitis in Rag2"~ mice independently of H. pylori infection and that co-transfer of wt
TR cells suppresses this lesion to a greater extent than IL10~'- TR cells. Only wt TR cells suppress
additive corpus gastritis attributable to H. pylori infection [91]. These results indicate that 1L10competent
TR
cells are pivotal to contain immune response to H. pylori-associated gastritis
(discussed in Chapter 2).
Persistent infection, chronic inflammation, and cancer
The link between inflammation and cancer has been recognized since 1863 [92]. The paradigm
of persistent infection leading to chronic inflammation and subsequent carcinogenesis is
supported by some chronic infection-associated malignancies, including Schistosoma-induced
squamous cell carcinoma of the urinary bladder, liver fluke-induced cholangiocarcinoma or liver
cancer, viral hepatitis-induced hepatocellular carcinoma, and H. pylori-induced gastric cancer
[13, 93].
Chronic inflammation results in persistent damage to tissue and biomolecules of all types
including DNA, protein, lipid, and carbohydrate due to the reactive oxygen and nitrogen species
(RONS) from inflammatory cells [94]. The DNA damage, cell proliferation, and tissue
remodeling in an inflammatory microenvironment may increase the risk of neoplasia due to
tumor initiation. Moreover, inflammatory cells also secrete proinflammatory cytokines,
chemokines, and growth factors that may affect survival, growth, proliferation, and
differentiation of cells [92]. Thus, persistent infection leads to a chronic inflammatory mileu
affording promotion and progression of initiated cells [95].
The mechanism of H. pylori-associated carcinogenesis has not been well characterized. Since H.
pylori infection usually persists lifelong and causes chronic active inflammation, the missing link
between H. pylori infection and gastric cancer may reside in the helicobacter-induced
inflammation. The paradigm of persistent infection, chronic inflammation, and cancer may
extend to H. pylori-associated gastric carcinogenesis and provide another approach for gastric
cancer prevention. Since host ThI, but not Th2, immune responses are responsible for H. pyloriassociated gastritis [82-85], Fox et al examined the effect of modulating the Th1 response of
mice to H. felis infection using a mouse model of parasitic infection that causes a Th2 immune
response [96]. In C57BLJ6 mice coinfected with helminths and H. felis, reduced systemic Thi
later observed that acute H. pylori infection resulted in transient hypochlorhydria and a persistent
decrease of vitamin C in gastric juice. Furthermore, the total vitamin C concentration in gastric
juice became elevated after successful H. pylori antimicrobial eradication therapy in patients
with intestinal metaplasia or duodenal ulcer [102, 103]. It is not clear whether the decreased
levels of ascorbic acid in gastric juice is due to dietary vitamin C, impaired secretion of vitamin
C by gastric mucosa, hypochlorhydria, increased oxidative stress in the stomach, or H. pylori
infection per se. It appears that a very complex interaction exists between H. pylori infection,
severity of gastric lesions, and ascorbic acid metabolism [104].
Using the vitamin C-deficient B6.129P2-GulomlIUncmemcd (gulo') mice lacking L-gulono-ylactone oxidase, we observed that high vitamin C supplementation (3300 mg/L) in drinking
water correlated with physiologically high vitamin C levels in plasma and gastric tissue. These
levels of vitamin C did not protect H. pylori-infected gulo'1 mice from helicobacter-associated
gastric disease. Low vitamin C (33 mg/ml)-supplemented H. pylori-infected gulo' mice had
physiologically low vitamin C levels in plasma and gastric tissue, and less severe gastric lesions
compared to high vitamin C-supplemented infected gulo' mice (discussed in Chapter 3). The
systemic H. pylori-specific IgG isotype responses and gastric proinflammatory cytokine
responses directly correlated with vitamin C levels in plasma and gastric tissue, indicating that a
physiologically low vitamin C supplementation may impair the immune response against H.
pylori infection and thus reduce severity of gastritis.
In a human epidemiological study comparing populations in New Orleans, LA and Colombia,
S.A., populations in Colombia had higher H. pylori prevalence, gastric cancer incidence, and
immune responses and lower levels of Thl-mediated gastric cytokines were associated with
increased H. felis colonization levels, attenuated gastric inflammation, and less severe
premalignant lesions [96]. The findings strongly suggest that H. pylori infection results in
chronic gastric inflammation that is largely responsible for the early stages of disease progression
and may promote gastric carcinogenesis [97]. These results in the C57BIJ6 mice coinfected with
helminths and H. felis may also in part explain the "African enigma" where the incidence of
gastric cancer is low in some African counties where parasitic infections are common despite a
high prevalence of H. pylori infection [14].
§ Diet and environmental factors
Ascorbic acid
Vitamin C, the water soluble ascorbic acid, is essential for maintaining health. Deficiency in
vitamin C causes scurvy which presents as spongy gums, subcutaneous and mucosal bleeding,
and eventually death [98]. Dietary ascorbic acid is usually oxidized to dehydroascorbic acid
(DHAA) in the stomach, absorbed in the small intestine, and reduced to ascorbic acid in the
circulation. Circulating ascorbic acid is accumulated in leukocytes and actively secreted into
gastric juice; excessive ascorbic acid is excreted into urine [99].
It has been speculated for a long time that vitamin C might be important in preventing
progression from gastritis to gastric cancer because vitamin C is an antioxidant that inhibits
formation of gastric carcinogenic N-nitroso compounds in vitro (discussed below) [100]. Sobala
GM and et al reported that levels of ascorbic acid in gastric juice were significantly reduced in
patients with chronic gastritis and directly correlated with hypochlorhydria [101]. This group
vitamin C levels in serum, but significantly lower vitamin C levels in gastric juice compared to
populations in New Orleans. Vitamin C levels in gastric juice were rinversely correlated with
severity of gastritis, atrophy, and gastric juice pH [104]. Our results in the gulo~'~ mouse model
may provide an explanation for the findings in Colombia [104]. High serum vitamin C levels
afford robust immune and inflammatory responses to H. pylori and result in more severe gastric
lesions and hypochlorhydria. Severe gastric mucosal damage may impair secretion of vitamin C
into gastric juice; hypochlorhydria allows overgrowth of bacteria that catalyzes formation of
more N-nitroso compounds that further consume vitamin C. Our results in the gulo~'~ mouse
model may also explain why protection against gastric cancer development was not observed in
several clinical trials of vitamin C supplementation alone or in combination with other
micronutrients in humans [105-107].
N-nitroso compounds
More than one hundred N-nitroso compounds are carcinogenic in laboratory animals [108-111].
N-nitroso compound-associated tumors have been observed in at least 39 animal species [112114]. Although the causal link between N-nitroso compound exposure and human cancer has not
been established [115, 116], these findings in laboratory animals strongly suggest that N-nitroso
compounds may be carcinogenic to humans. Thus, N-nitroso compounds have been classified as
carcinogenic for humans by IARC in 1987 [117-119].
N-nitroso compounds in the Swedish variant of moist oral smokeless tobacco (snus) have been
directly associated with gastric cancer development in mouse models [120]. Exposure to snus
alone increased the risk of gastric carcinoma in situ (CIS) in INS-GAS mice on a FVB
background. Exposure to snus and H. pylori infection synergistically increased the risk of CIS in
wt FVB and INS-GAS mice [120]. These results suggest that snus is a gastric carcinogen in
mice.
N-nitroso compounds are mainly from two sources: exogenous and endogenous. Exogenous
sources are from intake of preformed N-nitroso compounds in the diet or environment.
Endogenous sources are resulted from the formation of N-nitroso compounds in vivo from
nitrosatable precursors or nitrosating agents [114]. The main site for endogenous N-nitroso
compound formation is the stomach. N-nitroso compounds in the stomach are formed by two
mechanisms: non-enzymatic and enzymatic [114]. Non-enzymatic formation of N-nitroso
compounds requires the presence of nitrite, nitrosatable substrates, and acid (pH 1-3) [121]. In
the United States, the average dietary intake of nitrate is 75 mg/day, mainly from vegetables or
drinking water [122]. About 25% of the ingested nitrate is secreted into saliva, reduced to nitrite
by oral bacteria, and then enters the stomach [122, 123]. The remainder of ingested nitrate is
reduced to nitrite in the stomach by gastric bacteria [114, 124]. Nitrite and most secondary
amines can form N-nitroso compounds in the presence of gastric acid [114]. Bacterial catalysis
of N-nitrosation is another means of forming N-nitroso compounds. Bacteria, such as H. pylori,
that possess nitrate reductase or nitrosating enzymes such as cytochrome cdl-nitrite reductase
contribute to formation of N-nitroso compounds in the achlorhydric stomach [125, 126].
Additionally, nitrosating compounds can also be synthesized by enzymatic reaction in LPSactivated macrophages [127, 128]. Upregulated inducible nitric oxide synthase (iNOS) in
activated macrophages mediates the generation of nitric oxide (NO) that reacts with superoxide
(02-)
to form peroxynitrite (ONOO~) that could be converted to nitrous anhydride (N20 3), a
potent nitrosating agent [129]. It has been demonstrated that H. pylori extract stimulates
production of NO [130]. H. pylori infection leads to up-regulation of iNOS in the antrum
accompanied with increased levels of nitrotyrosine, a marker for peroxynitrite [131]. These in
vivo and in vitro findings strongly suggest that H. pylori infection may increase the production of
N-nitroso compounds in the stomach. Due to continual exposure to N-nitroso compounds in food
and water and synthesis of N-nitroso compounds by H. pylori catalysis or production of NO
from inflammatory cells in the gastric mucosa, N-nitroso compounds have been proposed to be
gastric carcinogenic in humans [132].
§ Host genetic factors
Although H. pylori is the major etiology of gastric cancer, most evidence suggests the
importance of host genetic factors in the progression of gastric cancer [133]. Approximately 10%
of gastric cancer cases show familial clustering [134, 135], and relatives of gastric cancer
patients have a two- to three-fold increased risk of developing gastric cancer in both genders
after controlling for H. pylori infection [136]. E-cadherin germline mutations were identified in
familial, early-onset, poorly-differentiated diffuse-type gastric cancer [3]. Nevertheless, few
specific genetic abnormalities responsible for the increased risk of intestinal-type gastric cancer
have been identified. Some polymorphisms in inflammation-related or DNA repair genes have
been associated with a high risk of gastric cancer in many populations throughout the world
[133]. The association between gastric cancer and these polymorphisms is discussed below.
Polymorphisms of inflammation-related genes
Antrum-predominat gastritis or duodenitis retains normal or higher acid secretion, while corpus-
predominant gastritis is associated with gastric atrophy and hypochlorhydria. Since IL-16 is a
pro-inflammatory cytokine and a powerful inhibitor of gastric acid secretion [137], these factors
were examined for their association with gastric cancer. In Scotland and Poland, polymorphisms
of IL-10 (IL-1@ -31*C or -511*T) and its receptor antagonist IL-1RN (IL-1RN*2/*2) were first
identified in populations with increased risk of gastric atrophy and cancer [138]. A higher risk of
gastric cancer from populations in China, Japan and Taiwan has been associated with certain
polymorphisms of genes in the IL-i pathway, including IL-1, IL-IRN, and type 1 IL-i receptor
(IL-IRI) [139-141]. Polymorphisms of other pro-inflammatory cytokines, such as IL-2 (IL-2 330 T to G), IL-4 (IL-4 -33 C to T), IL-8 (IL-8 -251 T to A), and TNF-aL ( TNF-ca -308 G to A),
also have been associated with a higher risk of gastric cancer [142-147]. Studies also identified
the association between polymorphisms of IL-10 (low IL-10 haplotype base on polymorphisms 592, -819, or -1082), an anti-inflammatory cytokine, and a higher incidence of gastric cancer in
Taiwan and other countries [142, 145, 147-150]. A combination of different genetic
polymorphisms may have synergistic effects on gastric cancer risk. A recent study observed that
polymorphisms of TNF-a and IL-10, when combined with polymorphisms of genes in IL-I
pathway, lead to a 27-fold or greater risk of gastric cancer [142]. These data indicate that
polymorphisms of inflammation-related genes are important determinants of susceptibility to and
severity of H. pylori-associated gastritis and gastric cancer [151].
Polymorphisms of DNA repair genes
H. pylori gastritis may lead to increased oxidative DNA damage due to the production of RONS
during chronic inflammation [94]. RONS cause DNA damage via (i) direct base oxidation and
deamination and (ii) indirect alkylation to form E-base lesions by 4-hydroxyalkenals through
lipid peroxidation [152-154]. Increased levels of oxidative DNA damage, double-strand DNA
breaks, and DNA fragmentation have been observed in human and mouse gastric mucosa with H.
pylori infection [131, 155-160]. Meanwhile, proteins related to DNA damage and repair, such as
Ku, poly (ADP-ribosyl) polymerase, 8-hydroxyguanine glycosylase (OGG1), and MSH2 were
selectively up-regulated in human gastric mucosa with acute gastritis [158, 159]. These results
suggest the importance of DNA repair enzymes in H. pylori-related carcinogenesis by removing
oxidative DNA damage.
Recent studies have identified several DNA repair proteins associated with gastric cancer,
including
0 6-methylguanine
DNA
methyltransferase
(Mgmt) and 8-oxoguanine
DNA
glycosylase (Oggl) [161, 162]. Mgmt is the DNA repair protein capable of repairing specific
types of mutagenic DNA adducts including 0 6 -methylguanine and 0 4-methylthymine [163, 164].
Transcription of Mgmt is frequently inactivated by promoter hypermethylation in precancerous
intestinal metaplasia and gastric cancer [165, 166]. Additionaly, Mgmt Ile143Val polymorphism
results in reduced enzyme activity [167] and has been associated with a higher risk of gastric
cancer in patients with low fruit and vegetable intake [161]. Ogg1 gene encodes the enzyme to
remove 8-hydroxy-2'-deoxyguanosine (8-oxo-dG) from DNA. Ogg1 Ser326Cys polymorphism
results in impaired enzyme activity [168, 169] and has been associated with accumulation of 8oxo-dG in H. pylori-infected gastric mucosa and a higher risk of gastric cancer [170, 171]. These
observations suggest that activities of DNA repair proteins may affect the natural course and
outcome of chronic H. pyloi gastritis. The roles of DNA repair proteins Mgmt and 3alkyladenine DNA glycosylase (Aag) in H. pylori-associated gastric disease are the subject of
Chapter 4.
§ Intervention of H. pylori infection and H. pylori-related disease
Antimicrobial H. pyloi eradication therapy
Since H. pylori was identified in patients with chronic gastritis and peptic ulcers in 1982 [10],
antimicrobial H. pylori eradication therapy has been the first-line therapy for peptic ulcer disease
[97]. Additionally, H. pylori is the major etiology of gastric cancer. According to Correa's model
of gastric cancer, H. pylori infection results in gastritis and initiates the progression to atrophy,
intestinal metaplasia, dysplasia, and finally gastric cancer [97, 133]. However, whether
antimicrobial H. pylori eradication therapy reduces the risk of gastric cancer is not conclusive.
Eradication of H. pylori in humans has been associated with preventing the progression of
preneoplastic lesions [105, 107, 172-174]. The optimal effect of antibiotic eradication therapy in
preventing gastric cancer has been observed in H. pylori infected patients who did not have
precancerous lesions prior to intervention with antimicrobial H. pylori eradication (p<0.05)
[175]. Once the preneoplastic lesions developed, antimicrobial H. pylori eradication therapy
failed to reduce the combined prevalence of dysplasia or gastric cancer [107]. These clinical
trials continue to pose key questions; in which patients would H. pylori eradication be beneficial
in preventing gastric cancer and at what stage in the progression of gastric disease would H.
pyloi antimicrobial eradication prevent progression of gastric lesions.
In human antimicrobial H. pylori eradication studies, confounding variables include duration of
H. pylori infection, stage of H. pylori gastritis, development of preneoplastic lesions prior to H.
pylori eradication therapy, diet, host genetic and environmental factors. Therefore, animal
models have been used to analyze the effect of antimicrobial H. pylori eradication therapy on
gastric cancer development. Antimicrobial H. pylori eradication therapy reverses the histologic
progression of dysplasia in H. pylori-infected Mongolian gerbils [176, 177]. In H. felis-infected
C57BIJ6 mice, antibiotics given at early time points of helicobacter infection prevents
development of dysplasia; eradication of H. felis during late stage of disease arrests progression
of dysplasia and prevents gastric cancer-related deaths [31]. These in vivo data support the
hypothesis that eradication therapy, depending on the timing of antibiotic treatment, may be
effective in preventing helicobacter gastritis from progressing to gastric cancer. Efficacy of
antimicrobial H. pylori eradication therapy in the development of H. pylori-associated gastric
cancer using INS-GAS mice is the subject of Chapter 5.
Other chemopreventive drugs
Cancer chemoprevention targeting the malignant cells or tumor microenvironment has been
extensively examined [178]. Some molecules or oncoproteins, including cyclooxygenase 1 or 2
(COXI or COX2), epidermal growth factor receptor (EGFR), Akt, NF-kB, and histone
deacetylase, have been identified as potential targets for chemopreventive treatment [178].
Nonsteroidal anti-inflammatory drugs (NSAIDs) inhibit prostaglandin E2 synthesis of COX-1
and COX-2 and have been widely used as anti-inflammatory and antipyretic agents. The safety
of NSAIDs has been confirmed in the long-term use. NSAIDs are known to reduce the risk of
colon cancer and breast cancer [179-182]. Despite the increased cardiovascular risk associated
with long-term use, selective COX-2 inhibitors have a chemopreventive effect on colon cancer
and do not have the gastrointestinal complications of non-selective NSAIDs [182]. Moreover,
short-term (2 weeks) treatment with selective COX-2 inhibitor, rofecoxib, enhances proapoptotic
proteins-dependent apoptosis and reduces gastrin-mediated cell growth in H. pylori-infected
patients, suggesting that inhibition of COX-2 may prevent the development of gastric cancer
[183]. Additionally, a retrospective study demonstrated that long-term use (> 3 years) of nonaspirin NSAIDs, but not aspirin, correlates with a reduced risk of gastric cancer [184]. Several
laboratory animal models have been used to examine the effect of NSAIDs on gastric cancer.
The selective COX-2 inihibor, celecoxib, reduced the multiplicity and size of gastric tumors in
N-methyl-N'-nitro-N-nitrosoguanidine (MNNG)-treated Wistar rats [185]. Other selective COX2 inhibitors were also reported to reduced gastric cancer incidence in H. pylori-infected, Nmethyl-N-nitrosourea (MNU)-treated C57BL/6 mice or Mongolian gerbils [186, 187]. The
efficacy of longterm NSAIDs treatment as chemopreventive drugs merits further studies.
Gastric cancer prevention with other chemopreventive drugs has also been examined in animal
models. Fox example, the combination of gastrin- and histamine H2-receptor antagonists (YF476
and loxitidine) has a synergistic, inhibitory effect on gastric cancer in H.felis-infected INS-GAS
mice [188]. Curcumin, extracted from the root of Curcuma longa Linn, inhibits the PI3K/Akt
pathway and induces apoptosis in vitro [189, 190] and reduces the multiplicity of gastric cancer
in N-methyl-N'-nitro-N-nitrosoguanidine (MNNG)-treated Wistar rats [191]. Further human
studies are necessary to verify the chemopreventive effects of these compounds.
REFERENCES
1.
Correa, P., Human gastric carcinogenesis: a multistep and multifactorialprocess--First
American Cancer Society Award Lecture on Cancer Epidemiology and Prevention.
Cancer Res, 1992. 52(24): p. 6735-40.
2.
Lauren, P., The Two Histological Main Types of Gastric Carcinoma: Diffuse and SoCalled Intestinal-Type Carcinoma. an Attempt at a Histo-Clinical Classification. Acta
Pathol Microbiol Scand, 1965. 64: p. 31-49.
3.
Guilford, P., et al., E-cadherin germline mutations in familial gastric cancer Nature,
4.
Kelley, J.R. and J.M. Duggan, Gastric cancer epidemiology and risk factors. J Clin
1998. 392(6674): p. 402-5.
Epidemiol, 2003. 56(1): p. 1-9.
5.
Correa, P., A human model of gastric carcinogenesis. Cancer Res, 1988. 48(13): p. 3554-
60.
6.
7.
Konturek, P.C., et al., Helicobacter pylori associated gastric pathology. J Physiol
Pharmacol, 1999. 50(5): p. 695-710.
Parkin, D., S. Whelan, and J. Ferlay, Cancer incidence in five continents. Vol. VII. 1997,
Lyon: International Agency for Research on Cancer.
8.
Chan, A.O., B.C. Wong, and S.K. Lam, Gastric cancer: past, present and future. Can J
Gastroenterol, 2001. 15(7): p. 469-74.
9.
Houghton, J., J.G Fox, and T.C. Wang, Gastric cancer: laboratory bench to clinic. J
Gastroenterol Hepatol, 2002. 17(4): p. 495-502.
10.
11.
Marshall, B.J. and J.R. Warren, Unidentified curved bacilli in the stomach of patients
with gastritis and peptic ulceration. Lancet, 1984. 1(8390): p. 1311-5.
Pisani, P., et al., Estimates of the worldwide mortality from 25 cancers in 1990. Int J
Cancer, 1999. 83(1): p. 18-29.
12.
13.
An international association between Helicobacter pylori infection and gastric cancer
The EUROGAST Study Group. Lancet, 1993. 341(8857): p. 1359-62.
IARC Working Group on the Evaluation of Carcinogenic Risk to Humans., in
Schistosomes liver flukes and Helicobacter pylori. 1994, International Association for
Research against Cancer: Lyon. p. 1-241.
14.
15.
Holcombe, C., Helicobacterpylori: the African enigma. Gut, 1992. 33(4): p. 429-31.
Fox, J.G, Non-human reservoirs of Helicobacterpylori. Aliment Pharmacol Ther, 1995.
16.
Fox, J.G and A. Lee, Gastric campylobacter-like organisms: their role in gastric disease
of laboratoryanimals. Lab Anim Sci, 1989. 39(6): p. 543-53.
Steer, H.W., Ultrastructure of cell migration throught the gastric epithelium and its
relationship to bacteria. J Clin Pathol, 1975. 28(8): p. 639-46.
Marshall, B.J., et al., Rapid urease test in the management of Campylobacter pyloridis-
9 Suppl 2: p. 93-103.
17.
18.
associatedgastritis.Am J Gastroenterol, 1987. 82(3): p. 200-10.
19.
Goodwin, C.S., A. Gordon, and V. Burke, Helicobacterpylori (Campylobacter pylori)
and duodenal ulcer Med J Aust, 1990. 153(2): p. 66-7.
20.
Marshall, B.J., et al., Attempt to fulfil Koch's postulatesfor pyloric Campylobacter.Med J
Aust, 1985. 142(8): p. 436-9.
21.
Morris, A. and G Nicholson, Ingestion of Campylobacter pyloridis causes gastritis and
raisedfastinggastricpH. Am J Gastroenterol, 1987. 82(3): p. 192-9.
22.
Rowland, M. and B. Drumm, Clinical significance of Helicobacter infection in children.
Br Med Bull, 1998. 54(1): p. 95-103.
23.
24.
25.
26.
Uemura, N., et al., Changes in Helicobacterpylori-induced gastritis in the antrum and
corpus during long-term acid-suppressive treatment in Japan. Aliment Pharmacol Ther,
2000. 14(10): p. 1345-52.
Kuipers, E.J., et al., Long-term sequelae of Helicobacterpylori gastritis. Lancet, 1995.
345(8964): p. 1525-8.
Krakowka, S., et al., Establishment of gastric Campylobacter pylori infection in the
neonatal gnotobioticpiglet. Infect Immun, 1987. 55(11): p. 2789-96.
Shuto, R., et al., Experimental gastritis induced by Helicobacter pylori in Japanese
monkeys. Infect Immun, 1993. 61(3): p. 933-9.
27.
Tsuda, M., et al., Essential role of Helicobacterpylori urease in gastric colonization:
definite proof using a urease-negative mutant constructed by gene replacement. Eur J
Gastroenterol Hepatol, 1994. 6 Suppl 1: p. S49-52.
28.
29.
Karita, M., et al., New small animal model for human gastric Helicobacter pylori
infection: success in both nude and euthymic mice. Am J Gastroenterol, 1991. 86(11): p.
1596-603.
Lee, A., et al., A small animal model of human Helicobacter pylori active chronic
gastritis.Gastroenterology, 1990. 99(5): p. 1315-23.
30.
Wang, T.C., et al., Mice lacking secretoryphospholipaseA2 show altered apoptosis and
differentiation with Helicobacterfelis infection. Gastroenterology, 1998. 114(4): p. 675-
.89.
31.
32.
Cai, X., et al., Helicobacterfelis eradication restores normal architecture and inhibits
gastriccancerprogressionin C57BL6 mice. Gastroenterology, 2005. 128(7): p. 1937-52.
Lee, A., et al., A standardizedmouse model of Helicobacterpylori infection: introducing
the Sydney strain. Gastroenterology, 1997. 112(4): p. 1386-97.
33.
Shomer, N.H., et al., Experimental Helicobacterpylori infection induces antral gastritis
and gastric mucosa-associated lymphoid tissue in guinea pigs. Infect Immun, 1998.
66(6): p. 2614-8.
34.
Israel, D.A., et al., Helicobacter pylori strain-specific diferences in genetic content,
identified by microarray, influence host inflammatory responses. J Clin Invest, 2001.
35.
Franco, A.T., et al., Activation of beta-catenin by carcinogenicHelicobacterpylori. Proc
107(5): p. 611-20.
Natl Acad Sci U S A, 2005. 102(30): p. 10646-51.
36.
Fox, J.G, et al., Host and microbial constituents influence Helicobacterpylori-induced
cancer in a murine model of hypergastrinemia.Gastroenterology, 2003. 124(7): p. 1879-
90.
37.
Thompson, L.J., et al., Chronic Helicobacterpylori infection with Sydney strain 1 and a
newly identifed mouse-adapted strain (Sydney strain 2000) in C57BUI6 and BALBIc
mice. Infect Immun, 2004. 72(8): p. 4668-79.
38.
Rogers, A.B. and J.G Fox, Inflammation and Cancer. L Rodent models of infectious
gastrointestinaland liver cancer Am J Physiol Gastrointest Liver Physiol, 2004. 286(3):
p. G361-6.
39.
Whary, M.T. and J.G. Fox, Detection, eradication, and research implications of
Helicobacter infections in laboratoryrodents. Lab Anim (NY), 2006. 35(7): p. 25-7, 30-
6.
40.
41.
Honda, S., et al., Development of Helicobacter pylori-induced gastric carcinoma in
Mongolian gerbils. Cancer Res, 1998. 58(19): p. 4255-9.
Watanabe, T., et al., Helicobacterpylori infection induces gastric cancer in mongolian
gerbils. Gastroenterology, 1998. 115(3): p. 642-8.
42.
Ogura, K., et al., Virulencefactors of Helicobacterpylori responsiblefor gastric diseases
in Mongolian gerbil.J Exp Med, 2000. 192(11): p. 1601-10.
43.
gastric cancer in INS-GAS mice is gender
Fox, J.G., et al., Helicobacterpylori-associated
specific. Cancer Res, 2003. 63(5): p. 942-50.
44.
Rogers, A.B., et al., Helicobacterpylori but not high salt induces gastric intraepithelial
neoplasia in B6129 mice. Cancer Res, 2005. 65(23): p. 10709-15.
45.
Tomb, J.F., et al., The complete genome sequence of the gastric pathogen Helicobacter
pylori. Nature, 1997. 388(6642): p. 539-47.
46.
47.
Alm, R.A., et al., Genomic-sequence comparison of two unrelated isolates of the human
gastricpathogenHelicobacterpylori. Nature, 1999. 397(6715): p. 176-80.
Hofreuter, D., S. Odenbreit, and R. Haas, Natural transformation competence in
Helicobacterpylori is mediated by the basic components of a type IV secretion system.
Mol Microbiol, 2001. 41(2): p. 379-9 1.
48.
Falush, D., et al., Recombination and mutation during long-term gastric colonization by
Helicobacterpylori: estimates of clock rates, recombination size, and minimal age. Proc
Natl Acad Sci U S A, 2001. 98(26): p. 15056-61.
49.
Suerbaum, S., et al., Free recombinationwithin Helicobacterpylori. Proc Nat. Acad Sci
U S A, 1998. 95(21): p. 12619-24.
50.
51.
Achtman, M., et al., Recombination and clonal groupings within Helicobacter pylori
from different geographicalregions. Mol Microbiol, 1999. 32(3): p. 459-70.
Suerbaum, S. and M. Achtman, Evolution of Helicobacter pylori: the role of
52.
recombination.Trends Microbiol, 1999. 7(5): p. 182-4.
Suerbaum, S. and P. Michetti, Helicobacter pylori infection. N Engl J Med, 2002.
347(15): p. 1175-86.
53.
Ferrero, R.L., S.L. Hazell, and A. Lee, The urease enzymes of Campylobacterpylori and
a relatedbacterium. J Med Microbiol, 1988. 27(1): p. 33-40.
54.
Mahdavi, J., et al., Helicobacterpylori SabA adhesin in persistent infection and chronic
inflammation. Science, 2002. 297(5581): p. 573-8.
55.
Rad, R., et al., The Helicobacterpylori blood group antigen-binding adhesinfacilitates
bacterial colonization and augments a nonspecific immune response. J Immunol, 2002.
168(6): p. 3033-41.
56.
Kim, J.S., et al., Molecular cloning and characterizationof the Helicobacterpylori fliD
gene, an essentialfactorinflagellarstructure and motility. J Bacteriol, 1999. 181(22): p.
6969-76.
57.
Tombola, F., et al., Blockers of VacA provide insights into the structure of the pore.
58.
Biophys J, 2000. 79(2): p. 863-73.
Lehmann, F.S., et al., In situ correlation of cytokine secretion and apoptosis in
Helicobacterpylori-associatedgastritis.Am J Physiol Gastrointest Liver Physiol, 2002.
283(2): p. G481-8.
59.
Fukui, T., et al., Significance of apoptosis induced by tumor necrosisfactor-alphaand/or
interferon-gamma against human gastric cancer cell lines and the role of the p53 gene.
Surg Today, 2003. 33(11): p. 847-53.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
Galmiche, A., et al., The N-terminal 34 kDa fragment of Helicobacterpylori vacuolating
cytotoxin targets mitochondria and induces cytochrome c release. Embo J, 2000. 19(23):
p. 6361-70.
Salama, N.R., et al., Vacuolating cytotoxin of Helicobacter pylori plays a role during
colonization in a mouse model of infection. Infect Immun, 2001. 69(2): p. 730-6.
Xiang, Z., et al., Analysis of expression of CagA and VacA virulence factors in 43 strains
of Helicobacterpylori reveals that clinical isolates can be divided into two major types
and that CagA is not necessaryfor expression of the vacuolating cytotoxin. Infect Immun,
1995. 63(1): p. 94-8.
Atherton, J.C., et al., Clinical and pathological importance of heterogeneity in vacA, the
vacuolating cytotoxin gene of Helicobacterpylori.Gastroenterology, 1997. 112(1): p. 929.
Gebert, B., et al., Helicobacter pylori vacuolating cytotoxin inhibits T lymphocyte
activation. Science, 2003. 301(5636): p. 1099-102.
Blum, G., et al., Gene clusters encoding the cytotoxic necrotizing factor type 1, Prsfimbriae and alpha-hemolysin form the pathogenicity island II of the uropathogenic
Escherichiacoli strain J96. FEMS Microbiol Lett, 1995. 126(2): p. 189-95.
Censini, S., et al., cag, a pathogenicity island of Helicobacterpylori, encodes type Ispecific and disease-associated virulence factors. Proc Natl Acad Sci U S A, 1996.
93(25): p. 14648-53.
Odenbreit, S., et al., Translocation of Helicobacterpylori CagA into gastric epithelial
cells by type IV secretion. Science, 2000. 287(5457): p. 1497-500.
Segal, E.D., et al., Altered states: involvement ofphosphorylated CagA in the induction of
host cellular growth changes by Helicobacterpylori. Proc Natl Acad Sci U S A, 1999.
96(25): p. 14559-64.
Fischer, W., et al., Systematic mutagenesis of the Helicobacterpylori cag pathogenicity
island: essential genes for CagA translocationin host cells and induction of interleukin8. Mol Microbiol, 2001. 42(5): p. 1337-48.
Higashi, H., et al., Helicobacter pylori CagA induces Ras-independent morphogenetic
response through SHP-2 recruitment and activation. J Biol Chem, 2004. 279(17): p.
17205-16.
Higashi, H., et al., SHP-2 tyrosine phosphataseas an intracellulartarget of Helicobacter
pylori CagA protein. Science, 2002. 295(5555): p. 683-6.
Higuchi, M., et al., Conditionalgene silencing utilizing the lac repressor reveals a role of
SHP-2 in cagA-positive Helicobacterpylori pathogenicity. Cancer Sci, 2004. 95(5): p.
442-7.
Naito, M., et al., Influence of EPIYA-repeat polymorphism on the phosphorylationdependent biological activity of Helicobacter pylori CagA. Gastroenterology, 2006.
130(4): p. 1181-90.
Hatakeyama, M., The role of Helicobacterpylori CagA in gastric carcinogenesis. Int J
Hematol, 2006. 84(4): p. 301-8.
Crabtree, J.E., et al., CagA/cytotoxic strains of Helicobacterpylori and interleukin-8 in
gastric epithelial cell lines. J Clin Pathol, 1994. 47(10): p. 945-50.
Harris, P.R., et al., Recombinant Helicobacterpylori urease activates primary mucosal
macrophages. J Infect Dis, 1998. 178(5): p. 1516-20.
Maeda, S., et al., Distinct mechanism of Helicobacter pylori-mediated NF-kappa B
78.
79.
80.
81.
82.
activation between gastric cancer cells and monocytic cells. J Biol Chem, 2001. 276(48):
p. 44856-64.
Viala, J., et al., Nodi responds to peptidoglycan delivered by the Helicobacterpylori cag
pathogenicity island.Nat Immunol, 2004. 5(11): p. 1166-74.
Hafsi, N., et al., Human dendriticcells respond to Helicobacterpylori,promoting NK cell
and Th1-effector responses in vitro. J Immunol, 2004. 173(2): p. 1249-57.
Eaton, K.A., S.R. Ringler, and S.J. Danon, Murine splenocytes induce severe gastritis
and delayed-type hypersensitivity and suppress bacterial colonization in Helicobacter
pylori-infected SCID mice. Infect Immun, 1999. 67(9): p. 4594-602.
Roth, K.A., et al., Cellular immune responses are essential for the development of
Helicobacterfelis-associatedgastricpathology. J Immunol, 1999. 163(3): p. 1490-7.
Karttunen, R., Blood lymphocyte proliferation, cytokine secretion and appearance of T
cells with activation surface markers in cultures with Helicobacterpylori. Comparison of
the responses of subjects with and without antibodies to H. pylori. Clin Exp Immunol,
1991. 83(3): p. 396-400.
83.
Mohammadi, M., et al., Helicobacter-specificcell-mediated immune responses display a
predominantThi phenotype andpromote a delayed-type hypersensitivity response in the
stomachs of mice. J Immunol, 1996. 156(12): p. 4729-38.
84.
Eaton, K.A., M. Mefford, and T. Thevenot, The role of T cell subsets and cytokines in the
pathogenesisof Helicobacterpylorigastritisin mice. J Immunol, 2001. 166(12): p. 7456-
61.
85.
Mohammadi, M., et al., Murine CD4 T-cell response to Helicobacterinfection: TH1 cells
enhance gastritisand TH2 cells reduce bacterial load. Gastroenterology, 1997. 113(6): p.
86.
Banham, A.H., F.M. Powrie, and E. Suri-Payer, FOXP3+ regulatory T cells: Current
controversies andfutureperspectives. Eur J Immunol, 2006. 36(11): p. 2832-6.
Hori, S., T. Nomura, and S. Sakaguchi, Control of regulatory T cell development by the
transcriptionfactorFoxp3. Science, 2003. 299(5609): p. 1057-6 1.
Rad, R., et al., CD25+/Foxp3+ T cells regulate gastric inflammation and Helicobacter
1848-57.
87.
88.
pylori colonization in vivo. Gastroenterology, 2006. 131(2): p. 525-37.
89.
Raghavan, S., et al., Absence of CD4+CD25+ regulatory T cells is associatedwith a loss
of regulation leading to increasedpathology in Helicobacterpylori-infected mice. Clin
Exp Immunol, 2003. 132(3): p. 393-400.
90.
Jiang, S., et al., Regulatory T cells and transplantationtolerance. Hum Immunol, 2006.
91.
Lee, C.W., et al., Wildtype and IL10-/- T regulatorycells reduced T effector cell-mediated
gastroduodenitis in Rag2-/- mice but only wildtype T regulatory cells suppressed
Helicobacterpylorigastritis.Infect Immun, 2007.
Balkwill, F. and A. Mantovani, Inflammation and cancer: back to Virchow? Lancet, 2001.
357(9255): p. 539-45.
Chen, C.J. and D.S. Chen, Interaction of hepatitis B virus, chemical carcinogen, and
genetic susceptibility: multistage hepatocarcinogenesis with multifactorial etiology.
67(10): p. 765-76.
92.
93.
Hepatology, 2002. 36(5): p. 1046-9.
94.
Dedon, P.C. and S.R. Tannenbaum, Reactive nitrogen species in the chemical biology of
95.
inflammation. Arch Biochem Biophys, 2004. 423(1): p. 12-22.
Coussens, L.M. and Z. Werb, Inflammation and cancer Nature, 2002. 420(6917): p. 860-
7.
96.
97.
Fox, J.G., et al., Concurrent enteric helminth infection modulates inflammation and
gastric immune responses and reduces helicobacter-inducedgastric atrophy. Nat Med,
2000. 6(5): p. 536-42.
Fox, J.G and T.C. Wang, Helicobacterpylori--not a good bug after all! N Engi J Med,
2001. 345(11): p. 829-32.
98.
Institute of Medicine. Dietary reference intakes for vitamin C, vitamin E, selenium and
99.
Levine, M., et al., Vitamin C pharmacokinetics in healthy volunteers: evidence for a
carotenoids.2000, Washington DC: National Academy Press.
recommended dietary allowance. Proc Natl Acad Sci U S A, 1996. 93(8): p. 3704-9.
100.
101.
102.
103.
104.
Schmahl, D. and G Eisenbrand, Influence of ascorbic acid on the endogenous
(intragastral)formation of N-nitroso compounds. Int J Vitam Nutr Res Suppl, 1982. 23:
p. 91-102.
Sobala, GM., et al., Ascorbic acid in the human stomach. Gastroenterology, 1989. 97(2):
p. 357-63.
Sobala, GM., et al., Effect of eradicationof Helicobacterpylori on gastricjuice ascorbic
acid concentrations.Gut, 1993. 34(8): p. 1038-41.
Sobala, GM., et al., Acute Helicobacter pylori infection: clinical features, local and
systemic immune response, gastric mucosal histology, and gastricjuice ascorbic acid
concentrations.Gut, 1991. 32(11): p. 1415-8.
Correa, P., et al., Review article: Antioxidant micronutrientsand gastric cancer Aliment
Pharmacol Ther, 1998. 12 Suppl 1: p. 73-82.
105.
Mera, R., et al., Long term follow up of patients treatedfor Helicobacterpylori infection.
Gut, 2005. 54(11): p. 1536-40.
106.
Plummer, M., et al., Chemoprevention of precancerousgastric lesions with antioxidant
vitamin supplementation: a randomized trial in a high-risk population. J Natl Cancer
Inst, 2007. 99(2): p. 137-46.
107.
You, W.C., et al., Randomized double-blindfactorial trial of three treatments to reduce
the prevalence ofprecancerous gastric lesions. J Natl Cancer Inst, 2006. 98(14): p. 974-
83.
108.
Hoffmann, D., et al., Assessment of tobacco-specific N-nitrosamines in tobacco products.
109.
Hoffmann, D., et al., Formationand analysis of N-nitrosamines in tobacco products and
their endogenousformationin consumers. IARC Sci Publ, 1984(57): p. 743-62.
110.
Lijinsky, W.,
111.
nitrosodiethanolaminein rats. Nature, 1980. 288(5791): p. 589-90.
Lijinsky, W., Significance of in vivo formation of N-nitroso compounds. Oncology, 1980.
Cancer Res, 1979. 39(7 Pt 1): p. 2505-9.
M.D.
Reuber, and W.B.
Manning,
Potent carcinogenicity of
37(4): p. 223-6.
112.
Lijinsky, W., Comparative carcinogenicity of N-nitroso compounds in diferent species.
IARC Sci Publ, 1980(31): p. 745-53.
113.
Bogovski, P. and S. Bogovski, Animal Species in which N-nitroso compounds induce
cancer Int J Cancer, 1981. 27(4): p. 471-4.
114.
115.
Brambilla, G and A. Martelli, Genotoxic and carcinogenicrisk to humans of drug-nitrite
interactionproducts. Mutat Res, 2007. 635(1): p. 17-52.
Bartsch, H. and R. Montesano, Relevance of nitrosamines to human cancer
Carcinogenesis, 1984. 5(11): p. 1381-93.
116.
117.
Tenth internationalmeeting on N-nitroso compounds, mycotoxins and tobacco smoke:
relevance to human cancer IARC Sci Publ, 1991(105): p. 1-594.
Tobacco habits other than smoking; betel-quid and areca-nutchewing; and some related
nitrosamines. IARC Working Group. Lyon, 23-30 October 1984. IARC Monogr Eval
Carcinog Risk Chem Hum, 1985. 37: p. 1-268.
118.
Betel-quid and areca-nut chewing and some areca-nut derived nitrosamines. IARC
Monogr Eval Carcinog Risks Hum, 2004. 85: p. 1-334.
119.
Monographs on the Evolution of the Carcinogenic Risks to Humans, in Overall
Evaluation of Carcinogenicity:An Updating of IARC Monographs. 1987, International
Association for Research against Cancer: Lyon. p. 1-42.
120.
Stenstrom, B., et al., Swedish moist snuff accelerates gastric cancer development in
Helicobacter pylori-infected wild-type and gastrin transgenic mice. Carcinogenesis,
2007.
121.
122.
N-nitroso compounds: occurrence, biological effects and relevance to human cancer
Proceedings of the VHIth International Symposium on N-Nitroso Compounds. Banff
Canada,5-9 September 1983. IARC Sci Publ, 1984(57): p. 1-1011.
Mirvish, S.S., The etiology of gastric cancer Intragastricnitrosamide formation and
123.
Tannenbaum, S.R., et al., Nitrite in human saliva. Its possible relationshipto nitrosamine
other theories. J Nati Cancer Inst, 1983. 71(3): p. 629-47.
formation.J Natl Cancer Inst, 1974. 53(1): p. 79-84.
124.
125.
Kyrtopoulos, S.A., et al., Studies in gastric carcinogenesis. II. Absence of elevated
concentrations of N-nitroso compounds in the gastric juice of Greek hypochlorhydric
individuals.Carcinogenesis, 1985. 6(8): p. 1135-40.
Ziebarth, D., B. Spiegelhalder, and H. Bartsch, N-nitrosation of medicinal drugs
catalysed by bacteria from human saliva and gastro-intestinal tract, including
Helicobacterpylori.Carcinogenesis, 1997. 18(2): p. 383-9.
126.
Calmels, S., et al., Characterizationof bacterial cytochrome cd(J)-nitrite reductase as
one enzyme responsiblefor catalysis of nitrosation of secondary amines. Carcinogenesis,
127.
1996. 17(3): p. 533-6.
Miwa, M., et al., Nitrosation of amines by stimulated macrophages. Carcinogenesis,
1987. 8(7): p. 955-8.
128.
Miwa, M., et al., N-nitrosamineformation by macrophages. IARC Sci Publ, 1987(84): p.
129.
Williams, D.L., Nitrosating agents: is peroxynitrite a likely candidate? Nitric Oxide,
130.
Tsuji, S., et al., Helicobacter pylori extract stimulates inflammatory nitric oxide
131.
production. Cancer Lett, 1996. 108(2): p. 195-200.
Mannick, E.E., et al., Inducible nitric oxide synthase, nitrotyrosine, and apoptosis in
132.
Helicobacterpylori gastritis: effect of antibiotics and antioxidants. Cancer Res, 1996.
56(14): p. 3238-43.
Palli, D., et al., 06-alkylguanines, dietary N-nitroso compounds, and theirprecursors in
340-4.
1997. 1(6): p. 522-7.
gastriccancer Nutr Cancer, 2001. 39(1): p. 42-9.
133.
Fox, J.G and T.C. Wang, Inflammation, atrophy, and gastric cancer J Clin Invest, 2007.
117(1): p. 60-9.
134.
La Vecchia, C., et al., Family history and the risk of stomach and colorectal cancer
Cancer, 1992. 70(1): p. 50-5.
135.
136.
137.
138.
139.
140.
141.
142.
143.
144.
145.
146.
147.
148.
149.
150.
151.
152.
Ignasi Elizalde, J. and J.M. Pique, Risk assessment in relatives of gastric cancer patients:
hyperproliferation, genetics, and Helicobacter pylori infection. Eur J Gastroenterol
Hepatol, 2006. 18(8): p. 877-9.
Imsland, A.K., et al., Genetic epidemiologic aspects of gastric cancer in Iceland. J Am
Coll Surg, 2002. 195(2): p. 181-6; discussion 186-7.
Beales, I.L. and J. Calam, Interleukin 1 beta and tumour necrosisfactor alpha inhibit
acid secretion in cultured rabbitparietal cells by multiple pathways. Gut, 1998. 42(2): p.
227-34.
El-Omar, E.M., et al., Interleukin-1 polymorphisms associated with increased risk of
gastriccancer Nature, 2000. 404(6776): p. 398-402.
Hamajima, N., PersistentHelicobacterpylori infection and genetic polymorphisms of the
host. Nagoya J Med Sci, 2003. 66(3-4): p. 103-17.
Liou, J.M., et al., IL-IB-511 C-->T polymorphism is associated with increased host
susceptibility to Helicobacterpylori infection in Chinese. Helicobacter, 2007. 12(2): p.
142-9.
Seno, H., et al., Novel interleukin-4 and interleukin-1 receptorantagonistgene variations
associated with non-cardia gastric cancer in Japan: Comprehensive analysis of 207
polymorphisms of11 cytokine genes. J Gastroenterol Hepatol, 2007. 22(5): p. 729-37.
El-Omar, E.M., et al., Increased risk of noncardia gastric cancer associated with
proinflammatorycytokine gene polymorphisms. Gastroenterology, 2003. 124(5): p. 1193201.
Garza Gonzalez, E., et al., Assessment of the toll-like receptor4 Asp299gly, thr399Ile and
interleukin-8 -251 polymorphisms in the risk for the development of distal gastric cancer
BMC Cancer, 2007. 7(1): p. 70.
Lee, W.P., et al., The -251T allele of the interleukin-8 promoter is associated with
increased risk of gastric carcinoma featuring diffuse-type histopathology in Chinese
population. Clin Cancer Res, 2005. 11(18): p. 6431-41.
Lu, W., et al., Genetic polymorphisms of interleukin (IL)-JB, IL-1RN, IL-8, IL-10 and
tumor necrosis factor [alpha] and risk of gastric cancer in a Chinese population.
Carcinogenesis, 2005. 26(3): p. 631-6.
Togawa, S., et al., Interleukin-2 gene polymorphisms associated with increased risk of
gastric atrophyfrom Helicobacterpylori infection. Helicobacter, 2005. 10(3): p. 172-8.
Kato, I., et al., Genetic polymorphisms in anti-inflammatory cytokine signaling and the
prevalence of gastric precancerous lesions in Venezuela. Cancer Causes Control, 2006.
17(9): p. 1183-91.
Rad, R., et al., Cytokine gene polymorphisms influence mucosal cytokine expression,
gastric inflammation, and host specific colonisation during Helicobacterpylori infection.
Gut, 2004. 53(8): p. 1082-9.
Wu, M.S., et al., Tumor necrosis factor-alpha and interleukin-10 promoter
polymorphisms in Epstein-Barr virus-associated gastric carcinoma. J Infect Dis, 2002.
185(1): p. 106-9.
Wu, M.S., et al., Interleukin-10 genotypes associate with the risk of gastric carcinoma in
Taiwanese Chinese. Int J Cancer, 2003. 104(5): p. 617-23.
El-Omar, E.M., Role of host genes in sporadic gastric cancer Best Pract Res Clin
Gastroenterol, 2006. 20(4): p. 675-86.
Bartsch, H. and J. Nair, Ultrasensitive and specific detection methods for exocylic DNA
153.
154.
155.
156.
157.
158.
159.
160.
161.
162.
163.
164.
165.
166.
167.
168.
169.
170.
adducts: markersfor lipid peroxidation and oxidative stress. Toxicology, 2000. 153(1-3):
p. 105-14.
Bartsch, H. and J. Nair, Potential role of lipid peroxidation derived DNA damage in
human colon carcinogenesis:studies on exocyclic base adducts as stable oxidative stress
markers. Cancer Detect Prev, 2002. 26(4): p. 308-12.
Bartsch, H., J. Nair, and R.W. Owen, Exocyclic DNA adducts as oxidative stress markers
in colon carcinogenesis:potential role of lipid peroxidation, dietaryfat and antioxidants.
Biol Chem, 2002. 383(6): p. 915-21.
Baik, S.C., et al., Increased oxidative DNA damage in Helicobacter pylori-infected
human gastricmucosa. Cancer Res, 1996. 56(6): p. 1279-82.
Papa, A., et al., Role of Helicobacterpylori CagA+ infection in determining oxidative
DNA damage in gastric mucosa. Scand J Gastroenterol, 2002. 37(4): p. 409-13.
Miyazawa, M., et al., Suppressed apoptosis in the inflamed gastric mucosa of
Helicobacterpylori-colonized iNOS-knockout mice. Free Radic Biol Med, 2003. 34(12):
p. 1621-30.
Jang, J., et al., Malgun (clear) cell change in Helicobacter pylori gastritis reflects
epithelial genomic damage and repairAm J Pathol, 2003. 162(4): p. 1203-11.
Lee, H., et al., "Malgun"(clear) cell change of gastric epithelium in chronic Helicobacter
pylori gastritis.Pathol Res Pract, 2000. 196(8): p. 541-5 1.
Arabski, M., et al., DNA damage and repair in Helicobacter pylori-infected gastric
mucosa cells. Mutat Res, 2005. 570(1): p. 129-35.
Huang, W.Y, et al., Selected DNA repairpolymorphisms and gastric cancer in Poland.
Carcinogenesis, 2005. 26(8): p. 1354-9.
Gonzalez, C.A., N. Sala, and G. Capella, Genetic susceptibility and gastric cancer risk.
Int J Cancer, 2002. 100(3): p. 249-60.
Singer, B. and B. Hang, What structuralfeatures determine repairenzyme specificity and
mechanism in chemically modified DNA? Chem Res Toxicol, 1997. 10(7): p. 713-32.
Sandercock, L.E., et al., Mutational-reportertransgenes rescuedfrom mice lacking either
Mgmt, or both Mgmt and Msh6 suggest that 06-alkylguanine-induced miscoding does
not contribute to the spontaneous mutational spectrum. Oncogene, 2004. 23(35): p. 593140.
Oue, N., et al., Promotermethylation status of the DNA repairgenes hMLH1 and MGMT
in gastriccarcinoma and metaplastic mucosa. Pathobiology, 2001. 69(3): p. 143-9.
Oue, N., et al., Promoter hypermethylation of MGMT is associated with protein loss in
gastric carcinoma. Int J Cancer, 2001. 93(6): p. 805-9.
Deng, C., et al., Genetic polymorphism of human 06-alkylguanine-DNA alkyltransferase:
identification of a missense variation in the active site region. Pharmacogenetics, 1999.
9(1): p. 81-7.
Hill, J.W. and M.K. Evans, Dimerization and opposite base-dependent catalytic
impairment of polymorphic S326C OGG1 glycosylase. Nucleic Acids Res, 2006. 34(5): p.
1620-32.
Smart, D.J., J.K. Chipman, and N.J. Hodges, Activity of OGGi variants in the repairof
pro-oxidant-induced 8-oxo-2'-deoxyguanosine. DNA Repair (Amst), 2006. 5(11): p.
1337-45.
Takezaki, T., et al., hOGG1 Ser(326)Cys polymorphism and modification by
environmentalfactors of stomach cancer risk in Chinese. Int J Cancer, 2002. 99(4): p.
624-7.
171.
Izzotti, A., et al., Interplay between Helicobacterpylori and host gene polymorphisms in
inducing oxidative DNA damage in the gastric mucosa. Carcinogenesis, 2007. 28(4): p.
892-8.
172.
Ley, C., et al., Helicobacterpylori eradication and gastric preneoplastic conditions: a
randomized, double-blind, placebo-controlledtrial. Cancer Epidemiol Biomarkers Prev,
2004. 13(1): p. 4-10.
173.
Ito, M., et al., Helicobacterpylori eradication therapy improves atrophic gastritis and
intestinal metaplasia: a 5-year prospective study of patients with atrophic gastritis.
Aliment Pharmacol Ther, 2002. 16(8): p. 1449-56.
174.
Correa, P., et al., Chemoprevention of gastric dysplasia: randomized trial of antioxidant
supplements and anti-helicobacterpylori therapy. J Natl Cancer Inst, 2000. 92(23): p.
1881-8.
175.
176.
Wong, B.C., et al., Helicobacterpylori eradicationto prevent gastric cancer in a highrisk region of China: a randomized controlled trial.Jama, 2004. 291(2): p. 187-94.
Cao, X., et al., Eradication of Helicobacter pylori induces apoptosis and inhibits
proliferation of heterotopic proliferative glands in infected Mongolian gerbils. Cancer
Sci, 2004. 95(11): p. 872-7.
177.
178.
Nozaki, K., et al., Effect of early eradication on Helicobacter pylori-related gastric
carcinogenesisin Mongolian gerbils. Cancer Sci, 2003. 94(3): p. 235-9.
Albini, A. and M.B. Sporn, The tumour microenvironment as a target for
179.
chemoprevention. Nat Rev Cancer, 2007. 7(2): p. 139-47.
Backlund, M.G, J.R. Mann, and R.N. Dubois, Mechanisms for the prevention of
gastrointestinalcancer: the role of prostaglandinE2. Oncology, 2005. 69 Suppl 1: p. 28-
32.
180.
182.
Mann, J.R., M.G Backlund, and R.N. DuBois, Mechanisms of disease: Inflammatory
mediators and cancerprevention. Nat Clin Pract Oncol, 2005. 2(4): p. 202-10.
Harris, R.E., J. Beebe-Donk, and G.A. Alshafie, Reduction in the risk of human breast
cancer by selective cyclooxygenase-2 (COX-2) inhibitors.BMC Cancer, 2006. 6: p. 27.
Brown, J.R. and R.N. DuBois, COX-2: a molecular target for colorectal cancer
183.
prevention. J Clin Oncol, 2005. 23(12): p. 2840-55.
Konturek, P.C., et al., Influence of COX-2 inhibition by rofecoxib on serum and tumor
181.
progastrin and gastrin levels and expression of PPARgamma and apoptosis-related
proteins in gastriccancerpatients.Dig Dis Sci, 2003. 48(10): p. 2005-17.
184.
Lindblad, M., J. Lagergren, and L.A. Garcia Rodriguez, Nonsteroidal anti-inflammatory
drugs and risk of esophageal and gastric cancer Cancer Epidemiol Biomarkers Prev,
185.
186.
2005. 14(2): p. 444-50.
Hu, P.J., et al., Chemoprevention of gastriccancer by celecoxib in rats. Gut, 2004. 53(2):
p. 195-200.
Magari, H., et al., Inhibitory effect of etodolac, a selective cyclooxygenase-2 inhibitor; on
stomach carcinogenesis in Helicobacter pylori-infected Mongolian gerbils. Biochem
Biophys Res Commun, 2005. 334(2): p. 606-12.
187.
Nam, K.T., et al., The selective cyclooxygenase-2 inhibitor nimesulide prevents
Helicobacter pylori-associated gastric cancer development in a mouse model. Clin
Cancer Res, 2004. 10(23): p. 8105-13.
188.
Takaishi, S., et al., Synergistic inhibitory effects of gastrin and histamine receptor
antagonists on Helicobacter-inducedgastric cancer. Gastroenterology, 2005. 128(7): p.
1965-83.
189.
190.
Hussain, A.R., et al., Curcumin induces apoptosis via inhibition of PI3'-kinase/AKT
pathway in acute T cell leukemias. Apoptosis, 2006. 11(2): p. 245-54.
Woo, J.H., et al., Molecular mechanisms of curcumin-induced cytotoxicity: induction of
apoptosis through generation of reactive oxygen species, down-regulation of Bcl-XL and
IAP, the release of cytochrome c and inhibition of Akt. Carcinogenesis, 2003. 24(7): p.
1199-208.
191.
Ikezaki, S., et al., Chemopreventive effects of curcumin on glandular stomach
carcinogenesis induced by N-methyl-N'-nitro-N-nitrosoguanidineand sodium chloride in
rats. Anticancer Res, 2001. 21(5): p. 3407-11.
54
Chapter 2:: Wildtype and IL10' T regulatory cells reduced T effector cell-mediated
gastroduodenitis in Rag2'' mice but only wildtype T regulatory cells suppressed
Helicobacterpylori gastritis
Abstract
Introduction
Materials and Methods
Experimental groups
Helicobacterpylori infection
56
57
58
59
Cell sorting and adoptive transfer
59
Histological evaluation
60
Special stains and immunohistochemistry
60
RNA extraction and quantitative PCR for cytokine expression and Foxp3
61
Quantitative culture of H. pylori
62
Statistical analysis
-62
Results
Rag2-'~ TE-recipient mice developed morbidity and gastroduodenitis independent
62
of H. pylori infection
Adaptive immunity is required to develop Hp-associated gastritis
63
Hp infection exacerbated corpus gastritis in Rag2~'~ TE-recipient mice
64
Adoptive transfer of wt or IL10~'~ TR cells reduced morbidity and gastroduodenitis
65
but only wt TR cells reduced severity of Hp corpus gastritis
Th1 cytokine responses were down-regulated to a greater extent by wt than 1L10-'
65
TR cells
Foxp3 expression in gastric tissues was promoted to a greater extent by wt than
66
IL10-' TR cells
Hp colonization levels were reduced by TE cells and maintained when wt TR cells
67
were co-transferred
Discussion
67
Acknowledgements
72
References
73
Legends
76
Figures
78
Table
82
A version of this chapter has been previously published and is reprinted here with the permission
of the publisher:
Lee, CW., Rao, VP., Rogers, AB., Ge, Z., Erdman, SE., Whary, WT., and Fox, JG.. Wild-Type
and Interleukin-10-Deficient Regulatory T Cells Reduce Effector T-Cell-Mediated
Gastroduodenitis in Rag2~'~ Mice, but Only Wild-Type Regulatory T Cells Suppress Helicobacter
pylori Gastritis. Infect Immun 2007 Jun;75(6):2699-707
ABSTRACT
CD4+CD45RBHiCD25- T effector cells (TE) promote Helicobacterpylori (Hp) gastritis in mice
and CD4*CD45RB"'CD25* T regulatory cells (TR) are anti-inflammatory. Using adoptive
transfer into Hp-infected Rag2~'~ mice, we evaluated effects of wildtype C57BIJ6 (wt) or
congenic IL10-'~ TR cells on gastritis, gastric cytokines and Hp colonization. Infected Rag2~'~ mice
colonized in the corpus and antrum with 105 to 106 Hp CFU/gram without associated gastritis. TE
cell transfer
caused morbidity
and
a Hp-independent
pangastritis
and
duodenitis
(gastroduodenitis) associated with increased expression of IFN-y and TNF-a. TE cell transfer to
Hp-infected mice led to additive corpus gastritis associated with inflammatory cytokine
expression and reduced colonization. wt
TR
cells reduced morbidity, Hp corpus gastritis,
gastroduodenitis, inflammatory cytokine expression and reversed the decline in Hp colonization
attributable to
TE
Cells. Although less effective than wt TR cells, IL10~' TR cells also reduced
morbidity and gastroduodenitis but did not reduce Hp corpus gastritis or impact TE cell inhibition
of colonization. Gastric tissues from mice receiving wt TR cells expressed higher levels of Foxp3
compared to recipients of IL10~'~ TR cells, consistent with lower regulatory activity of IL10~'- TR
cells. These results demonstrate that wt
TR
cells suppressed
TE
cell-mediated Hp-independent
gastroduodenitis and Hp-dependent corpus gastritis more effectively than IL10~'~
TR
cells.
Compartmental differences in TE cell- and Hp-mediated inflammation and the differences in
regulatory T cell effects between wt TR and IL10'' TR cells suggest that ILlO expression by wt TR
cells is important to regulatory suppression of gastric inflammation.
INTRODUCTION
Helicobacter pylori (Hp) infects the human stomach and causes gastritis, with a subset of
patients developing peptic ulcer disease, gastric carcinoma and gastric mucosa-associated
lymphoid tissue lymphoma [11, 186]. Infection of C57BL/6 mice with Hp or H. felis results in
chronic active gastritis [28, 31] and has been used to study the immune basis of Hp-induced
gastritis. Helicobacter infections in humans and mice induce a Thl-predominant immune
response with activation of CD4* T lymphocytes and expression of pro-inflammatory cytokines
such as IFN-y [79, 80]. This cell-mediated immunity results in gastritis characterized by
mononuclear cell infiltrates, mucosal hyperplasia and intestinal metaplasia. In contrast,
immunodeficient B6.129S7-Ragl'
mice that lack mature B and T cells colonized at high
density with H. felis but developed only minimal gastritis [78], indicating the importance of the
adaptive immune response to helicobacter-associated gastric disease.
Recent data have demonstrated that different subpopulations of CD4* T lymphocytes play
diverse roles in mediating and regulating Hp-induced gastritis. In B6.CB17-Prkdc"id (SCID)
mice, adoptive transfer of wildtype (wt) CD4+CD45RBHi T effector (TE) cells from naYve donors
caused severe gastritis in Hp-infected recipients, while co-transfer of wt CD4*CD45RB'
T
regulatory (TR) cells protected against development of gastritis [81]. TR cells have also been
defined by expression of the IL-2 receptor x chain and Foxp3 [83], the forkhead transcription
factor critical to thymic selection of CD4*CD25* TR cells [84]. Depletion of CD25*Foxp3* TR
cells in Hp-infected C57BIJ6 mice led to loss of immune regulation and more severe gastritis
[85] as did adoptive transfer of lymphocytes depleted of CD4*CD25* cells into Hp-infected
B6.Cg-Foxn1"" (nu/nu) recipients [86]. Of the many T cell subsets with ascribed regulatory
function [87], cell sorting experiments commonly use CD4*CD25*CD45RBO as naturally
occurring TR cells. However, the mechanism(s) for regulation by this type of TR cell are not fully
understood.
In adoptive transfer models using co-administration of wt TE cells and wt or I10-deficient
(EI10'~) TR cells, IL10 has been shown to be essential for the function of TR cells in suppressing
inflammatory bowel disease, dysplasia, and cancer of the colon [187, 188]. Based on this
evidence, we evaluated Hp gastritis in B6.129S6-Rag2tm1Fwa (Rag2'-) mice that had received wt
CD4+CD25-CD45RBHi TE cells and CD4+CD25*CD45RBL" TR cells from either wt C57BIJ6 or
congenic EI10~'~ mice. The results demonstrate that TE cells mediated a gastroduodenitis in Rag2mice independently of H. pylori infection and that wt TR cells suppressed this lesion to a greater
extent than EI10-' TR cells. Only wt TR cells suppressed additive corpus gastritis attributable to
Hp infection. The data support compartmental differences in TE and H. pylori-mediated
inflammation of the stomach and the differences in regulatory T cell effects between wt and
IL10-' TR cells suggest that EI10 expression by wt TR cells is important to regulatory suppression
of gastric inflammation.
MATERIALS and METHODS
Experimental groups
C57BIJ6 wildtype (wt) cell donor mice and adoptive transfer recipient Rag2 gene knockout mice
(B6.129S6-Rag2tm1Fwa or Rag2-'~), backcrossed 12 generations to B6 wt mice, were originally
from Taconic Farms (Germantown, NY). EI10 knockout cell donor mice (B6. 1 2 9 P2-1 l 10iicgJ
or IL10~'~), backcrossed 10 generations to B6 wt mice, were originally from the Jackson
Laboratory (Bar Harbor, ME). Mice were bred and maintained in an Association for Assessment
and Accreditation of Laboratory Animal Care-accredited facility in static microisolator cages
under specific-pathogen-free conditions as previously described [41]. Male and female, 6 to 8
week old, helicobacter-free Rag2-' mice were randomly assigned to uninfected or Hp-infected
groups that were further subdivided 4 weeks later into groups that received either no T cells, wt
TE cells, or wt TE cells in combination with wt or IL10-' TR cells. Helicobacter-free C57BL/6
mice were also dosed with the same inoculum of Hp to confirm the mouse-adapted strain
induced robust gastritis in wt mice. Some mice died acutely or were euthanized at earlier time
points due to declining body condition. Only data from mice surviving to the predetermined
necropsy time points of 16 and 20 weeks post Hp infection (12 and 16 weeks post adoptive
transfer, respectively) were analyzed. The protocol was approved by the Committee on Animal
Care of the Massachusetts Institute of Technology.
Helicobacterpyloriinfection
Helicobacterpylori Sydney strain 1 (SS1) was used for oral inoculation as described previously
[34, 41]. Hp was grown for 24 hours at 37*C under microaerobic conditions in Brucella broth
with 10% fetal bovine serum. The inoculum was suspended in PBS to an OD600=1.000 and then
assessed by Gram stain and phase microscopy for purity, morphology, and motility as well as for
urease, catalase, and oxidase activity. Mice were gavaged with 0.2 ml every other day for three
doses. Control groups were given 0.2 ml of PBS.
Cell sorting and adoptive transfer
Single cell suspensions from spleens and mesenteric lymph nodes of wt or IL10~' mice were
prepared as described previously [189]. In brief, CD4* cells were isolated by using CD4
Dynabeads and CD4 DETACHaBEAD (Dynal, Oslo, Norway). Cells were then labeled with
anti-CD4-Cy, anti-CD45RB-FITC and anti-CD25-PE antibodies (Pharmingen, La Jolla, CA) and
then sorted by flow cytometry (Model Mo-flo, Cytomation Inc., Fort Collins, CO) to a purity of
>95%
for
wt
TE
cells
(CD4*CD25~CD45RBHi)
and
wt
or
IL10-
TR
cells
(CD4*CD25*CD45RBL"). Anesthetized Rag2- mice were injected in the retro-orbital sinus with
3 x 105 TE cells alone or in combination with 3 x 105 wt or IL10'~ TR cells suspended in 200 d
HBSS.
Histological evaluation
At necropsy, the stomach and proximal duodenum were removed and cut along the greater
curvature. Linear gastric strips from the lesser curvature were fixed overnight in 10% neutralbuffered formalin, embedded, cut at 4 pm, and stained with hematoxylin and eosin (H & E).
Lesions were scored by a veterinary pathologist blinded to sample identity as described
previously [42]. Total lesion indices were calculated by the addition of individual scores for each
assessment described in Table 1 except for mucous metaplasia and hyalinosis, which have been
observed to develop spontaneously in mice [190] as well as from H.felis infection [29, 191].
Special stains and immunohistochemistry
Selective tissues were characterized using special stains and immunohistochemistry. Acidic
(intestinal type) mucins were demonstrated using pH 2.5 Alcian blue followed by periodic acidSchiff (AB-PAS) to stain remaining neutral (gastric type) mucins [42]. Macrophages were
stained with monoclonal antibody F4/80 (Caltag Laboratories, Burlingame, CA, 1:150) and with
an avidin-biotin-peroxidase complex kit (Vector Laboratories, Burlingame, CA) according to the
manufacturer instructions.
RNA extraction and quantitative PCR for cytokine expression and Foxp3
Stomach tissue was harvested and snap-frozen in liquid nitrogen. Total RNA was extracted with
Trizol reagent (Invitrogen, Carlsbad, CA). cDNA was synthesized from 5 pg of total RNA using
the High Capacity cDNA Archive kit (Applied Biosystems, Forster City, CA). Levels of IFN-y,
TNF-a, IL4, IL10 and Foxp3 were quantified with SYBR Green PCR reagent (Qiagen, Valencia,
CA) in an ABI Prism Sequence Detection System 7700 (Applied Biosystems) per the
manufacturer instructions. Primers were designed by Lasergene software (DNASTAR, Madison,
WI).
Sequences
of
primers
were
as
follows:
INFy:
Forward
(F):
CATggCTgTTTCTggCTgTTACTg, Reverse (R): gTTgCTgATggCCTgATTgTCTTT, Tm: 56 C;
TNFa: F: CATCTTCTCAAAATTCgAgTgACAA, R: TgggAgTAgACAAggTACAACCC,
Tm:
600 C; IL4: F: ACAggAgAAgggACgCCAT, R: gAAgCCCTACAgACgAgCTCA, Tm: 600 C;
IL10: F: ggTTgCCAAgCCTTATCggA, R: ACCTgCTCCACTGCCTTgCT, Tm: 600 C; Foxp3:
F: CCCAggAAAgACAgCAACCTT, R: CTCACAACCAggCCACTTgCA, Tm: 600 C; GAPDH:
F: TCCATgACAACTTTggCATTg, R: TCACgCCACAgCTTTCCA, Tm: 600 C. The final
concentration of each primer was 0.3 pM. Ten-fold dilutions (10 to 101 copies) of each cytokine
cDNA plasmid were used to generate standard curves. Expression levels were calculated as
mRNA copies of each cytokine per 105 copies of GAPDH or as ratio of Foxp3 to the internal
control of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (User Bulletin #2, Applied
Biosystems).
Quantitative culture of H. pylori
Colonization of Hp in the stomach was assessed by quantitative culture as described previously
[41]. Briefly, tissues from corpus and antrum were weighed and homogenized in 250 pLL of
Brucella broth using a sterile glass tissue grinder. The homogenate was serially diluted 10- and
100-fold in Brucella broth and plated onto selection plates containing 5% horse blood, 250 tg/ml
amphotericin B, 7 gg/ml bacitracin, 10.7 gg/ml nalidixic acid, 3.3 pg/ml polymyxin, and 100
pg/ml vancomycin. Plates were incubated microaerobically at 37*C for 7 days. Bacterial colonies
were counted and the CFU per gram of tissue calculated.
Statistical analysis
Morbidity between groups was compared by the Logrank test. Lesion scores were compared by
the Mann-Whitney U test or by the Kruskal-Wallis one-way analysis of variance with Dunnett's
test. Cytokine and Foxp3 expression levels and H. pylori colonization data (post log
transformation) were compared using the Newman-Keuls test. Statistical analysis was performed
using commercial software (Graphpad Prism 4.0, GraphPad Software, Inc., San Diego, CA) with
significance at p<0.05.
RESULTS
Rag2~'' TE-recipient mice developed morbidity and gastroduodenitis independent of H.
pylori infection
Hp-infected and helicobacter-free control Rag2~'' mice that did not receive cell transfers were
clinically normal throughout the 20 week study period. In contrast, 7 of 10 uninfected
TE-
recipient and 3 of 12 Hp-infected TE-recipient Rag2~'~ mice died acutely or developed diarrhea
and declining body condition, necessitating early euthanasia between 4 and 12 weeks post T cell
transfer (Table 1). All remaining uninfected and Hp-infected Rag2' mice that had received TE
cells alone were necropsied at 12 weeks post T cell transfer which was also 16 weeks post Hp
infection (WPI).
In all uninfected TE-recipient Rag24 mice, the entire alimentary tract including the stomach
through the colon was grossly edematous. The stomach and proximal duodenum from all mice
were evaluated histologically and there was severe pangastritis that involved the squamous and
glandular compartments of the stomach with inflammation extending into the proximal
duodenum (Table 1 and Fig. 1). This TE cell-mediated gastroduodenitis was characterized by
extensive infiltration of the mucosa and submucosa with lymphocytes, macrophages, eosinophils
and neutrophils. Other histological changes
included hypertrophy
and orthokeratotic
hyperkeratosis of the squamous portion of the stomach, as well as epithelial defects, hyalinosis
and mucous metaplasia of the corpus, along with mild blunting, atrophy and fusion of duodenal
villi (Table 1 and Fig. 1). A consistent feature of the gastroduodenitis was loss of Brunner's
glands through a combination of atrophy and dysplasia, accompanied by metaplasia to a
tubuloductular phenotype (Fig. 1). The bowel distal to the proximal duodenum was examined
histologically in 3 of these mice and had inflammatory infiltrates in the cecum and colon (data
not shown) which is consistent with a previous report of TE cell transfer-associated colitis in
Rag24 mice [192].
Adaptive immunity is required to develop Hp-associated gastritis
When evaluated at 20 WPI, the stomach and intestinal tract of Hp-infected Rag2' mice appeared
grossly normal and were similar to those from uninfected Rag2~'- mice. Histologically, Hpinfected Rag2~'- mice developed only minimal corpus gastritis with scattered neutrophils (Table
1, Fig. 2) and F4/80* macrophages in the submucosa (not shown). Hp-infected wt mice were
evaluated to confirm robust gastritis from infection with Hp SS1. These mice had moderate gross
thickening of the stomach, accompanied by histological lesions consisting of significant
infiltration with mononuclear cells and neutrophils, epithelial defects, oxyntic atrophy, intestinal
metaplasia, and hyperplasia (Table 1, p<0.001), as previously reported [31]. These findings
indicated that Hp-associated gastritis in mice was promoted by adaptive immunity and that
inflammation observed in Hp-infected, TE recipient Rag2~'~ mice was attributable to the
inflammatory activity of donor TE cells.
Hp infection exacerbated corpus gastritis in Rag2'~ TE-recipient mice
Hp-infected Rag2~'~ mice that received TE cells at 4 WPI and were necropsied at 16 WPI were
clinically affected similar to that observed in uninfected Rag2-'- mice that received TE cells alone.
Unexpectedly, Hp-infected, TE-recipient Rag2~'~ mice had less mortality than uninfected TErecipients, with 9 of 12 mice surviving to the time point of 16 WPI (Table 1, p<0.05). The
infected TE-recipient mice developed similar gross changes in the stomach and intestinal tract
with similar histological evidence of gastroduodenitis (Table 1, Fig. 2). However, corpus gastritis
was notably more severe in the Hp-infected TE-recipient mice compared to the uninfected TErecipients (Table 1, p<0.05). Otherwise, total lesion indices were similar (Table 1, p=0.20)
between Hp-infected and uninfected TE cell-recipient Rag2~'~ mice, suggesting that TE cells
mediated the antral gastritis, duodenitis and destruction of Brunner's glands. These results also
indicate that Hp infection further promoted the corpus gastritis that was superimposed on Hp-.
independent gastroduodenitis.
Adoptive transfer of wt or IL10-' T cells reduced morbidity and gastroduodenitis but only
wt TR cells reduced severity of Hp corpus gastritis
Clinical morbidity post transfer of TE cells into Hp-infected Rag2~'~ mice was less severe when
either wt (p<0.001) or IL10~'~ TR cells (p<0.01) were co-transferred (Table 1). Total lesion indices
for TE cell-associated gastroduodenitis that developed independently of Hp infection were lower
in Rag2-'~ mice co-transferred with either wt TR (p<0.001) or 1L10~' TR cells (p<0.01) (Table 1).
wt TR cells were more efficacious than IL10-' TR cells in ameliorating antral gastritis (p<0.01,
p<0.10, respectively), Brunner's gland destruction (p<0.01, p<0.05, respectively) and antral
dysplasia (p<0.01, p<0.05, respectively). Comparable to Hp-infected wt mice, gastritis in the Hpinfected Rag2~'~ mice that received TE and either type of TR cells was concentrated in the corpus
(Table 1, Fig. 2), with mild inflammation in adjacent compartments. Notably, the TE cellmediated corpus pathology in Hp-infected Rag2~'~ mice was diminished to the greatest extent in
mice that were co-transferred with wt TR cells compared to recipients of IL10~' TR cells (corpus
gastritis p<0.001, epithelial defects p<0.05, hyperplasia p<0.05; total lesion indices p<0.001).
None of the individual pathology parameters characterizing corpus gastritis were significantly
different in Hp infected TE-recipient mice given IL10~'~ TR cells from infected mice that received
TE cells alone (p=0.26 and higher, Table 1, Fig. 2). Interestingly, co-transfer of wt TR cells
resulted in significantly greater mucous metaplasia of parietal cells than the Hp-infected Rag2'
mice that received TE cells alone (p< 0.01, Fig. 2).
Th1 cytokine responses were down-regulated to a greater extent by wt than IL10-' TR cells
Consistent with the absence of gastritis, low levels of mRNA expression for IFN-y, TNF-ax, IL4
and IL10 were observed in gastric samples from uninfected and Hp-infected Rag2-'~ mice that
had not received T cells (Fig. 3).
TE
cell-associated gastroduodenitis was accompanied by a
significant transcriptional up-regulation of IFN-y, TNF-a, IL4 and IL10 mRNA (p<0.001)
regardless of Hp infection status (p=0.2 and higher); indicating that
TE
cells were the main
stimulus for cytokine expression. Notably, Thl-associated IFN-y and TNF-z expression levels in
gastric samples from TE-recipient, Hp-infected Rag2-'~ mice were two logs higher than Th2associated 1L4 and IL10 mRNA levels, consistent with the proinflammatory response to gastric
helicobacters [81, 93]. Hp-infected Rag2'~ mice that received TE plus wt
TR
cells had
significantly lower expression levels of IFN-y, TNF-a, IL4 and IL1O (p<0.01 and lower)
compared to infected Rag2~'~ mice that received TE cells only. Consistent with amelioration of TE
cell-mediated gastroduodenitis, co-transfer of ILiO~- TR cells also lowered mRNA expression for
IFN-y, TNF-a and IL4 (p<0.005 and lower) but a decrease in IL1O was not observed (p=0.51).
The ability of wt TR cells, and not ILIO'~ TR cells, to suppress corpus gastritis was consistent
with higher expression levels for IFN-y. (p<0.05, Fig. 3a) in recipients of IL1O''
TR
cells.
Additionally, suppression of TNF-a mRNA levels was more significant in mice receiving wt
compared to IL10~1
TR
cells (p<0.01, p<0.05, respectively, Fig. 3b). Thus, Th1-predominant
cytokine responses in Hp-infected TE cell-recipients were suppressed by both wt and IL10~'- TR
cells but suppression was greatest in recipients of wt TR cells.
Foxp3 expression in gastric tissues was promoted to a greater extent by wt than IL10~'
TR
cells
Uninfected and Hp-infected Rag2~'~ mice that did not receive TE cells had no detectable mRNA
of Foxp3 in gastric tissues (Fig. 3e). Foxp3 expression levels were elevated in mice that received
TE
cells (p<0.05') and Hp infection did not further stimulate expression. Levels of Foxp3 were
significantly higher in Hp-infected mice that received wt TR cells (p<0.001). Foxp3 expression in
tissues from recipients of L1O-' TR cells was not further elevated over levels observed for Hpinfected mice that received TE cells alone.
Hp colonization levels were reduced by TE cells and maintained when wt TR cells were cotransferred
Hp colonization levels have been reported to be inversely related to the severity of associated
chronic gastritis in mice [193]. Consistent with the absence of an inflammatory response, Hpinfected Rag2' mice that did not receive T cells were colonized at levels 3 to 5 logs higher in the
corpus and antrum, respectively, than in infected, TE cell-recipient Rag2' mice (Fig. 4, p<0.001).
Hp-infected mice that received TE and wt TR cells maintained colonization in the corpus at levels
similar to Hp-infected mice that did not receive TE cells (Fig. 4a, p=O.18). In contrast, Hpinfected mice that received TE plus IL10' TR cells had reduced corpus colonization similar to
infected mice that received TE cells alone (p=0.43), suggesting IL10' TR cells failed to influence
a TE cell-mediated reduction of Hp colonization. In the antrum, wt TR cells maintained higher
colonization levels than in infected mice that received TE cells alone (p<0.05), whereas
colonization in mice receiving TE and 1L10' TR cells was similar to infected mice receiving TE
without TR cells (p=O.19, Fig. 4b).
DISCUSSION
Using T cell transfer into Hp-infected Rag2' mice, this study demonstrated that wt TR cells
reduced TE cell-mediated morbidity, Hp-dependent corpus gastritis and Hp-independent
gastroduodenitis. The Hp-independent gastroduodenitis was characterized by antral gastritis,
duodenitis, Brunner cell loss, and epithelial dysplasia. These lesions have not been previously
described in mice with helicobacter infections and are consistent with autoimmune-like disease
[87]. I1O~'~ TR cells also reduced TE cell-mediated morbidity and to a lesser extent, Hpindependent gastroduodenitis. Notably, IL10-' TR cells did not reduce corpus gastritis
exacerbated by Hp infection, nor did IL10'~ TR cells reverse the TE cell-mediated reduction in Hp
colonization levels. TE cells mediated a sufficiently robust pro-inflammatory cytokine response
that IFN-y and TNF-a expression levels were similar irrespective of Hp infection status. This
Thl-predominant response was suppressed by both wt and IL1O~'- TR cells but reduction in
proinflammatory cytokine levels were greatest in recipients of wt TR cells. Gastric tissues from
mice receiving wt TR cells expressed higher levels of Foxp3 compared to recipients of ILl 0-' TR
cells, consistent with lower regulatory activity of IL10~'~ TR cells. These results suggest
compartmental differences in TE cell and Hp-mediated gastritis and that IL1O expression by wt
TR cells is important to regulatory suppression of gastric inflammation.
Similar to H.felis infection in Rag1~'~ mice [78] and despite high Hp colonization levels, Rag2~
mice that did not receive TE cells did not develop morbidity or gastrointestinal lesions. TErecipient mice developed morbidity and gastroduodenitis independently of Hp infection,
consistent with previous studies of TE cell transfer into Rag2 -- mice [192, 194] and mouse
models of TR deficiency [195, 196]. Observation of TE cell-mediated inflammation in the upper
gastrointestinal tract has been reported [81, 197] but not as frequently as colitis [195, 196] and
may be impacted by the number of cells transferred (VP Rao, personal communication). Extra-
intestinal inflammation attributable to TE cell infiltrates in sites such as the liver and Harderian
gland have also been observed [198] and may contribute to morbidity. At these sites,
TE
cells
likely proliferate in response to host, or more likely, bacterial antigens either absorbed through
the local epithelium or distributed systemically by vascular and lymphatic circulation.
TE
cell-
mediated morbidity may also be attributable to systemic effects of up-regulation of IFN-y, TNFa, IL4 and IL10 and other mediators released from gastrointestinal and extra-intestinal tissues.
Interestingly, pre-existing Hp infection resulted in a higher survival rate in mice that
subsequently received TE cells, potentially by promoting homing of TE cells to mucosal sites,
resulting in less extra-intestinal inflammation.
In uninfected TE cell-recipient Rag2~'- mice, reactive TE cells in the gastrointestinal mucosa most
likely expanded in response to dietary antigens and antigens of normal microbiota known to
colonize all regions of the bowel, including the upper gastrointestinal tract of mice [199].
TE-
mediated pangastritis involved the squamous and glandular compartments of the stomach,
extending into the duodenum. Lymphocytic infiltration of the antrum and duodenum promoted
antral dysplasia, villus atrophy, and destruction of Brunner glands and was an overlapping but
distinct disease pattern compared to the Hp corpus gastritis as described in this study and by
others [192, 194, 200]. Severity of corpus gastritis was inversely correlated with Hp colonization
levels which is indirect evidence that TE cells were responding to Hp antigens. Both wt and IL10~
~TR cells reduced the Hp-independent gastroduodenitis but only wt TR cells reduced the Hp
corpus gastritis and maintained higher levels of Hp colonization, suggesting compartmental or
other differences in the stomach that may favor a protective role for IL1O. The anti-inflammatory
effect of IL1O has been implicated by observation that IL10~'- mice developed more severe H.
felis [201] and Hp gastritis [202, 203] than congenic wildtype mice, as did Hp-infected SCID
mice injected with IL1O-'- splenocytes [81]. Additionally, a requirement for IL1O-competency of
TR cells has been shown in a mouse model of inflammatory bowel disease using H. hepaticus
infection. In this model only wt TR but not IL10-'~ TR cells reduced inflammation, dysplasia and
cancer lesions [188].
IL10~'~ TR cells reduced morbidity and attenuated TE cell-mediated gastroduodenitis along with
lower IFN-y and TNF-a expression. This observation is also consistent with the absence of this
distinct gastroduodenitis in uninfected or Hp-infected IL10~'- mice [203] that lack IL1Ocompetent TR cells. Interestingly, IL10 expression in the Rag2~- gastric tissues was not reduced
in IL1O'~ TR recipient mice as observed in mice that received wt TR cells. This result suggests
that IL1O mRNA was expressed by donor TE or host epithelial or other inflammatory cells [204],
possibly as a compensatory response of inflamed gastric tissue. Similarly, IL1O protein levels in
the Hp-infected gastric mucosa of humans have been positively associated with the severity of
gastritis [204]. Although the gastric IL1O level was not sufficient to dampen Hp-mediated corpus
gastritis, our findings suggest that IL10-competency of
TR
cells, but not local IL1O per se,
directly or indirectly reduces severity of Hp gastritis. Importantly, IL1O competence appears
important for
TR
cell suppression of Hp gastritis but our results do not rule out other
mechanisms. For example, CTLA-4 engagement induced and maintained anergy of Hp-specific
T cells in an TR-independent manner [205]. Our data also show that Foxp3 expression in gastric
tissues was promoted by TE cells, suggesting differentiation of TE cells into TR cells as reported
by others [206]. Lastly, Foxp3 expression was promoted only by wt and not IL10~'~
TR
cells,
consistent with the more potent regulatory activity of wt compared to IL10'- TR cells. Further
studies are necessary to clarify these issues.
In humans, Hp gastritis and subsequent progression to carcinoma has been associated with
atrophy and intestinal metaplasia of fundic glands [102], consistent with the hypothesis that
gastritis, gastric atrophy, and gastric cancer represent a continuum of progressive disease [127].
Irrespective of Hp infection, Rag2' mice that received TE cells developed marked mucous
metaplasia in the corpus which was more prominent in mice that received co-transfer of wt TR
cells. Fundic atrophy from chronic H. felis infection in mice has been associated with mucous
metaplasia characterized by increased expression of trefoil factor 2 (TFF2), also known as
spasmolytic polypeptide, by gastric mucous neck cells [29, 191]. This lesion appears to be a
replacement of parietal and chief cells with TFF2-secreting mucous cells that are similar in
morphology to antral or Brunner's glands and has been suggested to be a precursor lesion of
gastric cancer in mice [34] and in humans [207, 208]. Alternatively, because TFF2" mice
infected with H. pylori [209] or H. felis [210] developed more severe gastritis and hyperresponsiveness to IL-1B compared to wildtype mice, TFF2 may be a negative regulator of
gastritis and the promotion of mucous metaplasia by wt TR cells may reflect a protective TFF2associated response through an unknown mechanism.
In summary, our data demonstrate the role of TE cells as one component of adaptive immunity
that can trigger and sustain gastroduodenal inflammation in the absence of TR cells. Hp infection
coupled with TE cell transfer resulted in a more severe corpus gastritis that was compartmentally
distinct from TE cell-mediated gastroduodenitis. Further, our results demonstrate that IL10competence of TR cells appeared to be a requirement for suppression of Hp-induced gastritis, but
was not as critical for amelioration of TE-mediated gastroduodenitis or protection against
associated morbidity. These results are compatible with previous reports that Hp-induced
gastritis in mice is regulated by an IL10-dependent, Thl-type immune response [201, 203], likely
via the function of CD25*Foxp3* TR cells [85] and is consistent with increased Foxp3 expression
in wt TR-recipients. Compartmental differences in TE and Hp-mediated inflammation and
differential regulation by wt or IL10~'- TR cells should prove valuable for study of bacterial and
immune-mediated gastritis.
ACKNOWLEDGEMENTS
Grant support: R01AI37750, R01A150952, P01CA26T31, P30ES02109 (JGF).
We thank Vivian Ng and Philip Lee for animal care and necropsy; Kathy Cormier for histology
and immunohistochemistry; Sandy Xu and Nancy Taylor for quantitative culture; Kuo-Liong
Chien for assistance with statistical analysis; and Glenn A. Paradis and Michele Perry for their
expert assistance with cell sorting.
REFERENCES
1.
Pisani, P., et al., Estimates of the worldwide mortality from 25 cancers in 1990. Int J
Cancer, 1999. 83(1): p. 18-29.
2.
3.
Hansson, L.E., et al., The risk of stomach cancerin patients with gastric or duodenal
ulcer disease. N Engl J Med, 1996. 335(4): p. 242-9.
Lee, A., et al., A small animal model of human Helicobacterpyloriactive chronic
gastritis.Gastroenterology, 1990. 99(5): p. 1315-23.
4.
Lee, A., et al., A standardizedmouse model of Helicobacterpyloriinfection: introducing
the Sydney strain.Gastroenterology, 1997. 112(4): p. 1386-97.
5.
Karttunen, R., Blood lymphocyte proliferation,cytokine secretion and appearanceof T
cells with activationsurface markers in cultures with Helicobacterpylori. Comparison of
the responses of subjects with and without antibodiesto H. pylori. Clin Exp Immunol,
6.
Mohammadi, M., et al., Helicobacter-specificcell-mediated immune responses display a
predominant Thi phenotype and promote a delayed-type hypersensitivity response in the
stomachs of mice. J Immunol, 1996. 156(12): p. 4729-38.
Roth, K.A., et al., Cellularimmune responses are essentialfor the development of
Helicobacterfelis-associatedgastricpathology. J Immunol, 1999. 163(3): p. 1490-7.
Eaton, K.A., M. Mefford, and T. Thevenot, The role of T cell subsets and cytokines in the
pathogenesis of Helicobacterpylorigastritisin mice. J Immunol, 2001. 166(12): p. 7456-
1991. 83(3): p. 396-400.
7.
8.
61.
9.
10.
11.
Banham, A.H., F.M. Powrie, and E. Suri-Payer, FOXP3+ regulatoryT cells: Current
controversiesandfuture perspectives. Eur J Immunol, 2006. 36(11): p. 2832-6.
Hori, S., T. Nomura, and S. Sakaguchi, Control of regulatory T cell development by the
transcriptionfactor Foxp3. Science, 2003. 299(5609): p. 1057-6 1.
Rad, R., et al., CD25+/Foxp3+T cells regulate gastric inflammation and Helicobacter
pylori colonization in vivo. Gastroenterology, 2006. 131(2): p. 525-37.
12.
Raghavan, S., et al., Absence of CD4+CD25+ regulatory T cells is associatedwith a loss
of regulationleading to increasedpathology in Helicobacterpylori-infected mice. Clin
Exp Immunol, 2003. 132(3): p. 393-400.
13.
Jiang, S., et al., Regulatory T cells and transplantationtolerance. Hum Immunol, 2006.
67(10): p. 765-76.
14.
15.
16.
17.
Asseman, C., et al., An essential rolefor interleukin 10 in the function of regulatoryT
cells that inhibitintestinal inflammation. J Exp Med, 1999. 190(7): p. 995-1004.
Erdman, S.E., et al., CD4+CD25+ regulatory lymphocytes require interleukin 10 to
interruptcolon carcinogenesisin mice. Cancer Res, 2003. 63(18): p. 6042-50.
Fox, J.G., et al., Helicobacterpylori-associatedgastriccancer in INS-GAS mice is gender
specific. Cancer Res, 2003. 63(5): p. 942-50.
Fox, J.G, et al., Host andmicrobial constituents influence Helicobacterpylori-induced
cancerin a murine model of hypergastrinemia.Gastroenterology, 2003. 124(7): p. 1879-
90.
18.
19.
Erdman, S.E., et al., CD4+ CD25+ regulatory T lymphocytes inhibit microbially induced
colon cancer in Rag2-deficient mice. Am J Pathol, 2003. 162(2): p. 691-702.
Rogers, A.B., et al., Helicobacterpyloribut not high salt induces gastric intraepithelial
neoplasia in B6129 mice. Cancer Res, 2005. 65(23): p. 10709-15.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
Kang, W., et al., Interferon gamma induction of gastricmucous neck cell hypertrophy.
Lab Invest, 2005. 85(5): p. 702-15.
Fox, J.G., et al., Hypertrophicgastropathy in Helicobacterfelis-infectedwild-type
C57BL16 mice andp53 hemizygous transgenic mice. Gastroenterology, 1996. 110(1): p.
155-66.
Wang, T.C., et al., Mice lacking secretoryphospholipaseA2 show altered apoptosis and
differentiation with Helicobacterfelis infection. Gastroenterology, 1998. 114(4): p. 67589.
Corazza, N., et al., Nonlymphocyte-derived tumor necrosisfactor is requiredfor
induction of colitis in recombinationactivating gene RAG2-/- mice upon transfer of
CD4+CD45RBhiT cells. J Exp Med, 1999. 190(10): p. 1479-92.
Fox, J.G, et al., Concurrententeric helminth infection modulates inflammation and
gastric immune responses and reduces helicobacter-inducedgastricatrophy. Nat Med,
2000. 6(5): p. 536-42.
Sakagami, T., et al., Atrophic gastric changes in both Helicobacterfelisand Helicobacter
pylori infected mice are host dependent and separatefrom antral gastritis.Gut, 1996.
39(5): p. 639-48.
Annacker, 0., et al., CD25+ CD4+ T cells regulate the expansion of peripheral CD4 T
cells through the production of IL-10. J Immunol, 2001. 166(5): p. 3008-18.
Morrissey, P.J., et al., CD4+ T cells that express high levels of CD45RB induce wasting
disease when transferredinto congenic severe combined immunodeficient mice. Disease
development is prevented by cotransferof purified CD4+ T cells. J Exp Med, 1993.
178(1): p. 237-44.
Powrie, F., et al., Phenotypically distinct subsets of CD4+ T cells induce or protectfrom
chronic intestinalinflammation in C. B-17 scid mice. Int Immunol, 1993. 5(11): p. 146 171.
Dohi, T., et al., CD4+CD45RBHiinterleukin-4 defective T cells elicit antralgastritisand
duodenitis.Am J Pathol, 2004. 165(4): p. 1257-68.
Theve, E., J. Fox, and A. Rogers, Multiple autoimmune syndrome in 129/SvEv Rag2-/mice receiving adoptive transfer of T effector cells. Vet Pathol 2005. 42: p. 707.
Sarma-Rupavtarm, R.B., et al., Spatial distributionand stability of the eight microbial
species of the alteredschaedlerflora in the mouse gastrointestinaltract. Appl Environ
Microbiol, 2004. 70(5): p. 2791-800.
Suri-Payer, E. and H. Cantor, Differential cytokine requirementsfor regulationof
autoimmune gastritisand colitis by CD4+CD25+ T cells. J Autoimmun, 2001. 16(2): p.
115-23.
Berg, D.J., et al., Rapid development of severe hyperplastic gastritiswith gastric
epithelial dedifferentiationin Helicobacterfelis-infectedIL-10-/- mice. Am J Pathol,
1998. 152(5): p. 1377-86.
Matsumoto, Y, et al., Eradicationof Helicobacterpyloriand Resolution of Gastritisin
the GastricMucosa of IL-10-Deficient Mice. Helicobacter, 2005. 10(5): p. 407-15.
Whary, M., et al., IL-10 deficiency reduces Helicobacterpylorigastric colonization in
C57B16 mice. . Gastroenterology, 2001. 120(5): p. A99-A 100.
Bodger, K., J.I. Wyatt, and R.V. Heatley, Gastricmucosal secretion of interleukin-10:
relationsto histopathology,Helicobacterpylori status, and tumour necrosisfactor-alpha
secretion. Gut, 1997. 40(6): p. 739-44.
37.
Anderson, K.M., et al., Induction of CTLA-4-mediated anergy contributesto persistent
colonization in the murine model of gastric Helicobacterpyloriinfection. J Immunol,
2006. 176(9): p. 5306-13.
38.
Chen, W., et al., Conversion of peripheralCD4+CD25- naive T cells to CD4+CD25+
regulatory T cells by TGF-beta induction of transcriptionfactor Foxp3. J Exp Med, 2003.
39.
Mera, R., et al., Long term follow up of patients treatedfor Helicobacterpylori infection.
198(12): p. 1875-86.
Gut, 2005. 54(11): p. 1536-40.
40.
Fox, J.G. and T.C. Wang, Inflammation, atrophy,and gastric cancer J Clin Invest, 2007.
41.
Dhar, D.K., et al., Expression of cytoplasmic TFF2 is a marker of tumor metastasisand
negativeprognosticfactorin gastriccancer Lab Invest, 2003. 83(9): p. 1343-52.
Schmidt, P.H., et al., Identification of a metaplasticcell lineage associatedwith human
gastric adenocarcinoma.Lab Invest, 1999. 79(6): p. 639-46.
Fox, J.G., et al., H. pylori acceleratesthe progressionof dysplasia of the pylori antrum in
Tff2-/- C57BI6 x Sv129 infected mice. Gastroenterology, 2006. 130: p. A9.
Kurt-Jones, E.A., et al., Trefoilfamilyfactor 2 is expressed in murine gastricand immune
cells and controls both gastrointestinalinflammation and systemic immune responses.
117(1): p. 60-9.
42.
43.
44.
Infect Immun, 2007. 75(1): p. 471-80.
Legends
Figure 1. Transfer of wildtype (wt) T effector (TE) cells into Rag2~'~ mice resulted in immunemediated pangastritis and gastroduodenitis. (a) Normal squamous gastric compartment of Rag2'
mouse that did not receive TE cells. (b) TE cell transfer resulted in inflammation of the squamous
stomach with reactive epithelial cell hypertrophy and plication of the surface mucosa. (c) TE cell
transfer also caused inflammation in the cardia and corpus. (d) TE cell-mediated gastritis of the
glandular and oxyntic mucosa resulted in oxyntic atrophy characterized by loss of parietal and
chief cells. (e) Normal Brunner's glands (arrow) in the proximal duodenum. (f) TE cell transfer
produced duodenitis with near complete loss of Brunner's glands due to atrophy and
tubuloductular metaplasia (arrow). H&E; bar = 200 pm.
Figure 2. H. pylori (Hp) infection in Rag2''~ mice with or without subsequent transfer of
wildtype (wt) T effector (TE) cells alone or in combination with wildtype (wt) or IL10~'- T
regulatory (TR) cells. (a) In the absence of T cells, Hp infection produced minimal to no gastritis.
(b) Hp infection followed by transfer of TE cells produced moderate to severe gastritis. (c) Cotransfer of wt TR cells ameliorated severity of gastric inflammation in Hp infected, TE-recipient
mice, but for unknown reasons, marked mucous metaplasia of parietal cells developed (arrow).
(d) Co-transfer of IL10~'~ TR cells into Hp-infected TE-recipient mice decreased TE cell-mediated
antral gastritis, Brunner cell loss and epithelial dysplasia but did not reduce Hp-induced corpus
gastritis. H&E; bar = 200 pm.
Figure 3. Mean expression levels of mRNA for IFN-y (a), TNF-a (b), IL4 (c), IL10 (d), and
Foxp3 (e) were determined by quantitative PCR in gastric tissues from uninfected and H. pylori
(Hp)-infected Rag2'~ mice subsequently transferred with wildtype (wt) T effector (TE) cells alone
or in combination with wildtype (wt) or IL10~'~ T regulatory (TR) cells. TE cells up-regulated the
expression of Th 1 cytokines (IFN-y, TNF-c) and Th2 cytokine (IL4) with mRNA for Th 1
cytokines expressed 1-2 logs higher than for Th2 cytokines. Co-transfer of TE plus wt or IL10
TR cells suppressed expression of FN-y, TNF-c, and IL4. IL10 expression was up-regulated in
TE-recipient mice while concurrent transfer of wt TR but not IL10~' TR cells down-regulated IL1O
expression. Foxp3 expression was significantly higher in mice that received TE cells. Co-transfer
of wt TR cells resulted in 4-fold up-regulation of Foxp3 expression which was significantly
higher than induced by IL10~'~ TR cells (p<0.05) which was unchanged from mice receiving TE
cells alone. Bars represent standard error. (* p<0.05, ** p< 0.01; *** p< 0.001).
Figure 4. Mean H. pylori (Hp) colonization in the corpus (a) and antrum (b) of Rag2~- mice
receiving either no T cells, wildtype (wt) T effector (TE) cells or TE cells in combination with T
regulatory (TR) cells from wildtype (wt) or IL10~'- mice. Hp colonization in the corpus was lower
in mice receiving TE or TE plus IL10- TR cells. Co-transfer of wt TR cells maintained
colonization levels in the corpus (only) whereas IL10~- TR cells significantly diminished Hp
colonization in the corpus and antrum. Bars represent standard error. (* p<0.05, ** p< 0.01; ***
p< 0.001).
Fig. 1
Fig. 2
Fig. 3
**
**
400
~ F1
*
F1
1500
r
fl1
1
-~
0
0
==I
300
1000
o
200
CL
0
S100
500
z
L
0
0
H. pylori
TE
TR
+
-
+
-
+
+
+
+
+
+
wt 1L10--
H. pylori
TE
TR
+
-
+
-
+
+
+
+
+
+
wt IL10-/
C
**
**
0
7.5
?
**
15
II
0
10
0
5
CL
2.5 =
0H. pylori
5
u
+
+
IL1 0-/-
-
+
+
+
+
+
-
-
wt
-
+
+
+
+
w+
+
-
-
wt
IL10-/-
+
-
TE
TR
0
2
5
a.
4
Cf,
x
0
U-
3
2
0
H. pylori
-
+
-
-
TE
TR
+
H. pylori
TE
TR
IL10-'-
Fig. 4
I
I
(n
C.)
LiL
C-)
T
0~
H. pylori
TE
TR
+
+
+
+
H. pylori
+
-
+
+
+
-
-
wt
IL10-/-
TE
TR
-
+
-
+
+
+
+
+
-
wt
L10--
Table 1. Gastric lesions at 16 and 20 weeks post H. pylori (Hp) infection (WPI)
Median (Range) of Lesion Indices:
Antrum
Corpus
Group (na)
WPj Hp TE
WT (2/2)
20
-
WT (3/3)
20
Rag (7/7)
Ep Defc
Hyper
Int
Atrop
Muc
Total
Infm
Brnr
Dys
Total
TR
Infm
0
0
0
0
0
0
0
0
0
0
0
-
-
(0)
(0)
(0)
(0)
(0)
(0)
(0)
(0)
(0)
(0)
(0)
2
1
1
+
-
-
(2-2.5)
(1-2)
(0.5-1)
0
0
0
0
0
0
0
0
0
0
0
20
-
-
-
(0-0.5)
(0)
(0)
(0)
(0)
(0)
(0-0.5)
(0)
(0)
(0)
(0)
Hp (11/11)
20
+
-
-
0.5
(0-1.5)
1.5
0
(0-0.5)
0.5
0
(0)
1.5
0
(0)
1
0
(0-1)
3
0
(0-1.5)
2
0.5
(0-2.5)
8
0
(0)
2
0
(0)
4
0
(0)
1
0
(0)
9
TE (3/10)
16
-
+
-
(1.5-2)e
(0.5)e
(0.5-2)
(0.5-1)
(2-3)
(2)
(5-8)
(2-2.5)
(4)
(1)e
(8-9)
2.5
1
1
0.5
3
1.5
8.5
1.5
3.5
0.5
7.5
(2-3)c'e
( 0 .5- 2 )'e
(1-2)c
(0-1.5)
(2-3.5)
(0.5-3)c
( 6 - 1 1 .5 )d
( 1 .5-2. 5 )b
( 2 .5- 4 )'c
( 0 .5- 1 )b"ce
( 6 .5- 10 )'cd
Hp+TE (9/12)
16
+
+
-
20
+
+
wt
20
+
+
Hp+TE+WtTR
(6/7)
1
Hp+TE+IL1O-T,
(4/6)
IL10-
0.5
(0.5-1.5)c ( 0 .5- 1 )'
1
( 0 .5- 1 )"''
1
(0.5-1)
2.5
0
8
0
0
0
0
(2.5)
(0-1)
(6.5-9)
(0)
(0)
(0)
(0)
0.5
3
3
5
1
2.5
0
4.5
(0-0.5)
(2.5-2.5)
(2.5-3.5)c
( 4 .5-5. 5 )d
( 0 .5- 2 )b
(0.5-3)c
(0-0.5)c
( 3 -6. 5 )d
2
0.5
1.25
0.25
2.75
2.75
7
1.25
2.25
0
5.25
(1.5-3)
(0.5-1)
( 1 -2 )'
(0-0.5)
(2-3)
(2-3)
(5-9)
(1-1.5)
(2 - 3 )b
( 0 -0. 5 )b
(4-6)c
Abbreviation: TE: T effector cells; TR:T regulatory cells; Infm: inflammation; Ep Defc: epithelial defect; Hyper: hyperplasia; Int: intestinal metaplasia; Atrop:
atrophy;
Muc: mucous metaplasia; Brnr: Bruner's gland destruction; Dys: dysplasia.
"Number necropsied/Group size. Differences represent early morbidity
bSignificant influence of TR cells at a p of <0.05
"Significant influence of TR cells at a p of <0.01
"Significant influence of TR cells at a p of <0.001
eSignificant influence of Hp infection at a p of <0.05
fSignificant difference between wt and IL1O~'- TR cells at a p of <0.05
83
Chapter 3: High vitamin C intake and normal vitamin C levels in plasma and gastric tissue
do not protect vitamin C-deficient L-gulono-y-lactone
oxidase knockout mice from
Helicobacterpylori-induced premalignant gastric lesions
Abstract
Introduction
Materials and Methods
Mice
Experimental design
Experimental infection with H. pylori
Vitamin C measurements
Histological evaluation
RNA extraction and quantitative PCR for cytokine mRNA
Plasma IgG isotypes measurement
Quantitative PCR for H.pylori colonization
Quantification of plasma levels of gastrin
Statistical Analysis
Results
Vitamin C content in plasma and gastric tissues correlated with vitamin C
supplementation levels in gulo-'~ mice
High vitamin C supplementation did not reduce H. pylori gastritis
Less severe gastritis in H. pylori-infected gulo' mice supplemented with low
vitamin C was associated with lower gastric mRNA levels of IFN-y and TNF-a
Gulo~' mice supplemented with low vitamin C had lower H. pylori-specific IgG2c
responses
H. pylori colonization levels were higher in gulo~'~ mice supplemented with low
vitamin C at 32 WPI
Gastrin amidation was not impaired in vitamin C-supplemented gulo-'- mice
Discussion
Acknowledgements
References
Legends
Figures
Tables
85
86
88
89
89
90
90
91
91
92
92
93
93
94
96
97
97
98
99
105
106
110
113
118
1
ABSTRACT
2
In human studies, low vitamin C intake has been associated with more severe Helicobacter
3
pylori gastritis and a higher incidence of gastric cancer. However, vitamin C supplementation has
4
not been definitively shown to protect against gastric cancer. Using vitamin C-deficient
5
B6.1 29 P2 -Gulotm'me/d (gulo'~) mice lacking L-gulono-y-lactone oxidase, we compared
6
gastric lesions and Th1 immune responses in H. pylori-infected gulo~'~ mice supplemented with
7
low (33mg/L) or high (3300mg/L) vitamin C in drinking water for 16 or 32 weeks. Vitamin C
8
content in plasma and gastric tissue correlated with the vitamin C supplementation levels in gulo-
9
'
mice. H. pylori infection resulted in comparable gastritis and premalignant lesions in wildtype
10
C57BLJ6 and gulo~'- mice supplemented with high vitamin C, but lesions were less severe in
11
gulo~'~ mice supplemented with low vitamin C at 32 weeks post infection. The reduced gastric
12
lesions in infected gulo~'~ mice supplemented with low vitamin C correlated with reduced Thl-
13
associated IgG2c, gastric IFN-y and TNF-a mRNA and higher H. pylori colonization levels.
14
Additionally, a trend toward higher amidated gastrin levels may have contributed to more severe
15
premalignant lesions in H. pylori-infected, high vitamin C-supplemented gulo'- mice. These
16
results in the H. pylori-infected gulo~'~ mouse model suggest that although supplementation with
17
a high level of vitamin C achieved physiologically normal vitamin C levels in plasma and gastric
18
tissue, this dose of vitamin C did not protect gulo~' mice from H. pylori-induced premalignant
1
gastric lesions. However, low vitamin C supplementation correlated with an attenuated Th 1
2
inflammatory response and less severe gastric lesions.
3
4
INTRODUCTION
5
Helicobacterpylori infects the human stomach [1] and has been definitively linked to chronic
6
gastritis, which in some individuals results in serious gastric disease such as peptic ulcer, gastric
7
adenocarcinoma or gastric MALToma [2]. Multiple factors have been evaluated for impact on
8
helicobacter-associated gastric disease [2, 3]. Dietary factors, including nitrosamines, high salt,
9
and low dietary vitamin C (ascorbic acid), have been proposed to negatively influence the
10
clinical outcome of H.pylori infection in epidemiological and animal studies [4-8].
11
12
Vitamin C, a water-soluble antioxidant, reduces the formation of carcinogenic N-nitroso
13
compounds in gastric juice and scavenges reactive oxygen metabolites in the gastric mucosa [9,
14
10]. Vitamin C is also important for carboxyamidation of gastrin and cross linkage of collagen
15
and elastin [11, 12]. Epidemiological studies in humans have linked vitamin C deficiency to
16
more severe H. pylori-associated gastritis and a higher risk for gastric cancer [10, 13]. It has also
17
been reported that reduced vitamin C levels in gastric juice and plasma in H. pylori-infected
18
patients returned to normal levels after H. pylori eradication [10, 13-16]. Supplementation of
1
vitamin C has been associated with reduced gastric cancer risk in some human studies [7, 13].
2
Despite initial promising results in a prospective trial in a very high-risk population
3
supplemented with vitamin C at 2 g per day for 6 years [17], a followup study on this population
4
indicated that vitamin C supplementation over a 12 year period did not provide any lasting
5
protection against gastric cancer [18]. These results are consistent with other studies that did not
6
observe a correlation between severity of chronic gastritis or gastric cancer risk and vitamin C
7
levels [19-21].
8
9
In human vitamin C intervention studies, confounding variables include diet, vitamin C status,
10
genetic polymorphisms, duration of infection with specific or unknown H. pylori strains, and
11
degree of gastritis. Therefore, animal models have been used to analyze the effects of vitamin C
12
on H. pylori gastritis and gastric cancer [22-24]. Mice and Mongolian gerbils have been used to
13
evaluate H. pylori gastritis and vitamin C oral intake [25, 26]. However, like most laboratory
14
rodents, a major limitation for using these animal models for vitamin C studies is their ability to
15
endogenously synthesize vitamin C. Thus results in these rodents are difficult to interpret. In
16
contrast, the vitamin C-deficient gulo'
17
Gulotm0me/mmcd)
18
vitamin C [27]. Using this model, we were able to modulate low and high vitamin C levels in
mouse on a C57BIJ6 background (B6.129P2-
lacks L-gulono-y-lactone oxidase, and thus cannot endogenously synthesize
1
plasma and gastric tissue during H. pylori infection to specifically analyze whether dietary
2
vitamin C would influence the outcome of H. pylori infection. We hypothesized that high dietary
3
levels of vitamin C would reduce the severity of H. pylori gastritis, while physiologically lower
4
levels of vitamin C would exacerbate disease.
5
6
MATERIALS and METHODS
7
Mice
8
Mice were maintained in an Association for Assessment and Accreditation of Laboratory Animal
9
Care-accredited facility in static microisolator cages under specific-pathogen-free (SPF) status
10
including free of Helicobacterspp. as previously described [28]. Wildtype (wt) helicobacter-free
11
C57BIJ6 mice were obtained from Taconic Farms (Germantown, NY). Gulo*'~ mice (B6.129P2-
12
Gulo'm
13
Mouse Regional Resource Center (University of California at Davis, CA), rederived to SPF
14
status, and bred to maintain a homozygous state [27]. For breeding and maintenance, gulo~'~ mice
15
were weaned at 3 weeks of age and fed ad libitum with regular mouse chow (Prolab RMH 3000,
16
PMI Nutrition International, Richmond, Indiana) and water supplemented with 330 mg/L of L-
17
ascorbic acid (Sigma-Aldrich Co., St. Louis, MO) and 0.01 mM EDTA (Sigma-Aldrich Co.).
18
Supplemented water was changed weekly [27]. Animal experiments were approved by the
Ome/mmea,
back-crossed to C57BIU6 for 10 generations) were obtained from the Mutant
1
Committee on Animal Care of the Massachusetts Institute of Technology.
2
Experimental design
3
Male and female, 6-8 week old gulo~'~ mice supplemented with vitamin C (330 mg/L) were
4
experimentally infected with H. pylori and then randomly subdivided into low and high vitamin
5
C supplemented groups. The low vitamin C group was supplemented with vitamin C in water at
6
33 mg/L and the high vitamin C group was supplemented with vitamin C in water at 3,300 mg/L
7
[29]. Control uninfected gulo' mice were supplemented with low or high vitamin C. Age-
8
matched control uninfected wt mice and wt mice dosed with the same inoculum of H. pylori
were used to confirm the mouse-adapted strain induced robust gastritis in wt mice.
10
Approximately half of the mice of each group were euthanatized with CO 2 at 16 or 32 weeks
11
post infection (WPI) (Table 1 and 2). To confirm the.results observed at 32 WPI, a second
12
experiment evaluated the same vitamin C treatment groups of uninfected and H. pylori-infected
13
gulo~'~ mice along with uninfected and H. pylori-infected wt mice. In addition, a group of H.
14
pylori-infected wt mice were supplemented with high vitamin C in the water (Table 2).
15
Experimental infection with H. pylori
16
H. pylori Sydney strain (SS1) was used for oral inoculation as described previously [28, 30].
17
After incubation for 24 hours at 37*C while shaking under microaerobic conditions in Brucella
18
broth with 10% fetal bovine serum, H. pylori was harvested, resuspended in PBS and assessed by
1
Gram stain and phase microscopy for purity, morphology, and motility. The bacteria]
2
concentration was adjusted to OD 600=1 .000 in PBS. This is approximately 109 organisms/ml.
3
Mice were dosed with 0.2 ml of the H. pylori suspension in PBS by gavage every other day for
4
three doses. Control mice were dosed with PBS.
5
Vitamin C measurements
6
Total vitamin C levels were measured by high performance liquid chromatography and UV detection
7
as described previously with some modifications [31]. In brief, blood was collected at necropsy in
8
EDTA and centrifuged immediately at 4*C. Vitamin C was extracted from plasma by adding an
9
equal volume of cold perchloric acid (PCA) solution (1 L contained 50 ml of perchloric acid and 95
10
mg of EDTA in ddH 2O), followed by vortexing and centrifugation at 2,500 g at 4'C for 10 minutes.
11
The supernatant was frozen at -70*C pending analysis. To measure vitamin C levels in tissue, a
12
longitudinal strip of the gastric greater curvature was weighed, frozen in liquid nitrogen, and stored
13
at -70*C. Frozen tissue was added to cold PCA solution at a ratio of 1:9 (weight/weight) followed by
14
homogenization and centrifugation at 2,500 g at 4*C for 10 minutes, and then frozen at -70 0C until
15
analysis. Throughout the vitamin C extraction process, samples were kept on ice, protected from
16
light, and measured within one week post processing.
17
Histological evaluation
18
At necropsy, the stomach and proximal duodenum were removed and opened along the greater
1
curvature. Linear gastric strips from the lesser curvature were fixed overnight in 10% neutral-
2
buffered formalin, embedded, cut at 4 pLm, and then stained with hematoxylin and eosin. Using
3
criteria described previously [32], gastric lesions were scored for inflammation, epithelial
4
defects, mucous metaplasia, atrophy, hyperplasia, intestinal metaplasia, and dysplasia by board
5
certified veterinary pathologists (BR and ABR) blinded to sample identity. Total gastric indices
6
were calculated by the addition of individual scores for each type of gastric lesion.
7
RNA extraction and quantitative PCR for cytokine mRNA
8
A longitudinal strip of gastric tissue from the anterior wall was harvested and snap-frozen in
9
liquid nitrogen. Total RNA was extracted with Trizol reagent (Invitrogen, Carlsbad, CA). cDNA
10
was synthesized from 5 pg of total RNA with the High Capacity cDNA Archive kit (Applied
11
Biosystems, Forster City, CA). mRNA levels of interferon-y (IFN-y) and tumor necrosis factor-a
12
(TNF-a) were quantified with TaqMan gene expression assays and TaqMan Fast Universal PCR
13
Master Mix in a 7500 Fast Real-Time PCR system (Applied Biosystems) per manufacturer's
14
instructions. mRNA levels of each cytokine were normalized to the mRNA level of internal
15
control glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and compared to data from
16
uninfected wt mice using the AACT method (Bulletin 2, Applied Biosystems).
17
Plasma IgG isotypes measurement
18
Plasma was evaluated for H. pylori-specific IgG2c and IgG1 by ELISA using an outer membrane
1
protein preparation from H. pylori (SS1 strain) as described previously [33]. In brief, 96 well
2
flat-bottom plates were coated with 100 pl of antigen (10 gg/ml) overnight at 4'C, and sera were
3
diluted 1:100. Biotinylated secondary antibodies for detecting IgG2c and IgG1 were from clone
4
5.7 and A85-1 (BD Pharmingen, San Jose, CA). Incubation with extravidin peroxidase (Sigma-
5
Aldrich) was followed by treatment with 2,2'-azinobis (3-ethylbenzthiazolinesulfonic acid)
6
(ABTS) substrate (Kirkegaard & Perry Laboratories, Gaithersburg, MD) for color development.
7
The optical density was recorded by a plate reader per manufacturer's protocol (Power WaveX
8
Select, Biotek Instruments, Winooski, VT).
9
Quantitative PCR for H. pylori colonization
10
A longitudinal strip of gastric tissue from the greater curvature was proteinase K digested at 550C
11
for 8 hours followed by DNA extraction with phenol:chloroform:isoamyl alcohol (25:24:1) and
12
ethanol precipitation. H. pylori colonization levels (DNA copy numbers) were quantified by a
13
fluorogenic quantitative PCR assay using urease B primers and probe [34]. H. pylori copy
14
numbers were normalized to the amount of murine genomic DNA as determined by quantitative
15
PCR using a eukaryotic 18S endogenous control (Applied Biosystems) (Bulletin 2, Applied
16
Biosystems). Data is presented and compared as log-transformed copy numbers per pg host
17
DNA.
18
Quantification of plasma levels of gastrin
1
Plasma concentrations of glycine-extended gastrin (G-gly) and amidated gastrin were determined
2
by radioimmunoassay using the antibodies 109-21 and L2 specific to G-gly and amidated gastrin,
3
respectively, as described previously [35].
4
Statistical Analysis
5
Histological scores from both 32 WPI experiments were pooled for analysis because they were
6
performed under identical conditions. Pathology scores were analyzed by variance (ANOVA)
7
followed by post-hoc comparison when ANOVA results were significant. Because cytokine
8
mRNA levels increased progressively across the H. pylori-infected groups from low vitamin C to
9
high vitamin C to wt, we assigned a numerical value from 1 to 3 to each group in the model and
10
estimated the significance level of gastric lesions using a general linear model that tests for
11
trends. Levels of gastric cytokine, plasma IgG isotypes, H. pylori colonization, and amidated
12
gastrin were compared by the Student t test. All statistical analyses were two-sided tests at a
13
significance level of 0.05 performed with SAS software version 9.1 (SAS Institute Inc., Cary,
14
NC). The mean and standard error of all data are presented in the figures using Graphpad Prism
15
4.0 (Graphpad software, San Diego, CA).
16
17
RESULTS
18
Vitamin C content in plasma and gastric tissues correlated with vitamin C supplementation
1
levels in gulo~' mice
2
Both uninfected and H. pylori-infected gulo'
3
significantly lower plasma (p<0.01) and gastric tissue (p<0.001) vitamin C levels relative to the
4
mice supplemented with high vitamin C at 16 and 32 weeks (Fig. 1). Uninfected and H. pylori-
5
infected gulo-'- mice supplemented with high vitamin C had plasma and gastric tissue vitamin C
6
levels comparable to those in unsupplemented wt mice at the same time points. H. pylori
7
infection had no significant effect on plasma vitamin C levels in gulo~' mice supplemented with
8
low or high vitamin C or in wt mice (p=0.37 and higher). At 16 and 32 WPI, there was a trend
9
for H. pylori infection to reduce gastric tissue vitamin C levels in gulo~'~ mice supplemented with
10
low vitamin C (p=0.071,=0.069, respectively). This reduction in gastric tissue vitamin C
11
associated with H. pylori was significant in gulo' mice supplemented with high vitamin C at 16
12
WPI (p<0.05), however, paradoxically, H. pylori infection was associated with increased gastric
13
tissue vitamin C levels at 32 WPI (p<0.05). Infection in wt mice did not significantly affect
14
gastric tissue vitamin C levels at 32 WPI (p=0.24).
mice supplemented with low vitamin C had
15
16
High vitamin C supplementation did not reduce H.pylori gastritis
17
Gastritis was not observed in uninfected wt mice and gulo' mice supplemented with low or high
18
vitamin C. Consistent with previous reports [8, 26], H. pylori-infected wt mice developed robust
1
chronic atrophic gastritis at 16 and 32 WPI characterized by corpus inflammation with
2
lymphocytic and granulocytic infiltration, epithelial defects, hyperplasia, intestinal metaplasia,
3
and atrophy (Fig. 2, 3).
4
5
At 16 and 32 WPI, H. pylori-infected gulo~'~ mice supplemented with high vitamin C developed
6
gastric lesions of comparable severity relative to infected wt mice. At 16 WPI, there was a trend
7
for H. pylori-infected gulo~'~ mice supplemented with low vitamin C to have fewer epithelial
8
defects (p=0.094), hyperplasia (p=0.069), intestinal metaplasia (p=O. 15), atrophy (p=O. 15), and
9
dysplasia (p=0.078) relative to infected gulo~'- mice supplemented with high vitamin C. These
10
trends were also reflected in slightly lower total gastric indices at 16 WPI in H. pylori-infected
11
gulo' mice supplemented with low vitamin C (p=0.15).
12
13
Compared to H. pylori-infected gulo~'~ mice supplemented with high vitamin C, by 32 WPI H.
14
pylori-infected gulo~'~ mice supplemented with low vitamin C had developed less severe gastritis
15
(p=0.11), atrophy (p=0.06), and dysplasia (p=0.12) accompanied by significantly less extensive
16
epithelial defects (p<0.01), mucous metaplasia (p<0.05), and hyperplasia (p<0.05). When
17
comparing total gastric indices of disease between low and high vitamin C supplementation
18
levels, infected gulo~' mice supplemented with low vitamin C had statistically significant lower
1
total gastric indices compared to those mice supplemented with high vitamin C (p<0.05). Based
2
on differences identified in cytokine gene expression (below), one-way trend analysis of H.
3
pylori-infected gulo~'- mice with low vitamin C compared to gulo-'~ mice with high vitamin C
4
followed by comparison with wt mice demonstrated a difference in atrophy (p=0.053) and
5
significant differences in epithelial defects (p<0.001), hyperplasia (p<0.05), mucous metaplasia
6
(p<0.05), and total gastric indices (p<0.05).
7
8
Less severe gastritis in H.pylori-infected gulo-' mice supplemented with low vitamin C was
9
associated with lower gastric mRNA levels of IFN-y and TNF-a
10
Control gulo~'~ mice supplemented with low or high vitamin C and uninfected wt mice had
11
similar background gastric mRNA levels of IFN-y and TNF-a at 32 WPI. H. pylori infection
12
significantly up-regulated the mRNA levels of IFN-y and TNF-c in the stomachs of gulo-'~ and
13
wt mice at 32 WPI. IFN-y and TNF-x expression levels were elevated by H. pylori infection to a
14
greater extent in gulo- mice supplemented with high vitamin C (p<0.001) and wt mice
15
(p<0.001) compared to gulo~'- mice given low vitamin C supplementation (IFN-y p<0.01, TNF-a
16
p=0.088) (Fig. 4). The H. pylori-infected gulo~'~ mice supplemented with low vitamin C had
17
lower gastric mRNA levels of IFN-y and TNF-u compared to infected gulo'- mice supplemented
18
with high vitamin C or wt mice (p<0.05 or lower). There were no significant differences in
1
gastric mRNA levels of IFN-y and TNF-a between H. pylori-infected gulo~' mice supplemented
2
with high vitamin C and infected wt mice (p=0.31, p= 0.23, respectively).
3
4
Gulo-'' mice supplemented with low vitamin C had lower H.pylori-specific IgG2c responses
5
H. pylori infection resulted in a Thl-predominant IgG2c response in wt and gulo-' mice as
6
previously reported [33, 36] (Fig. 5). At 32 WPI, H. pylori-specific IgG2c levels were
7
significantly higher in H. pylori-infected wt mice than in infected gulo~' mice supplemented with
8
low vitamin C (p<0.05) but IgG2c responses were similar between gulo~'~ mice given low or high
9
vitamin C (p=0.44). There was a trend for infected wt mice to have higher H. pylori-specific
10
IgG2c levels than gulo~' mice supplemented with high vitamin C (p=0.051). However, in the 6
11
H. pylori-infected wt mice given vitamin C supplementation, there was no difference in H.
12
pylori-specific IgG2c responses compared to the 10 infected wt mice that didn't receive oral
13
vitamin C supplementation (p=0.51, data not shown). H. pylori infection induced a low IgG1
14
response in all infected groups.
15
16
H. pylori colonization levels were higher in gulo-' mice supplemented with low vitamin C at
17
32 WPI
18
H. pylori colonization levels in the stomach were comparable among infected gulo-'~ mice
1
supplemented with low or high vitamin C and wt mice at 16 WPI (Fig. 6). By 32 WPI, infected
2
gulo~'~ mice supplemented with low vitamin C maintained higher H. pylori colonization levels
3
than gulo'~ mice supplemented with high vitamin C (p<0.05) or wt mice (p=0.082) where H.
4
pylori colonization levels decreased over time in concert with development of more severe
5
gastritis.
6
7
Gastrin amidation was not impaired in vitamin C-supplemented gulo~' mice
8
Amidation of the intermediate glycine-extended gastrin (G-gly) is vitamin C dependent [11, 37]
9
and amidated gastrin levels are elevated in H. pylori infection in C57BIU6 mice [8]. At 32 WPI,
10
levels of the intermediate G-gly in plasma from most wt and gulo-'~ mice supplemented with low
11
or high vitamin C were below the assay detection limit (data not shown). Uninfected wt mice and
12
gulo~' mice supplemented with low or high vitamin C had similar amidated gastrin levels
13
(p=0.13) (Fig. 7). Uninfected and H. pylori-infected gulo~' mice supplemented with low vitamin
14
C had comparable plasma amidated gastrin levels (p=0.82). There was a trend toward higher
15
plasma amidated gastrin levels in H. pylori-infected gulo~'~ mice supplemented with high vitamin
16
C compared to uninfected counterparts (p=0.2 1). In contrast, H. pylori infection was associated
17
with a significant increase in plasma amidated gastrin levels in wt mice compared to uninfected
18
wt controls (p<0.01). Among the H. pylori infected gulo~'~ and wt mice, gulo'
mice
mi
1
supplemented with low vitamin C had lower amidated gastrin levels compared to infected gulo
2
mice supplemented with high vitamin C and wt mice (p=0.065 and <0.05, respectively).
3
4
DISCUSSION
5
Using the gulo~' mouse model, we were able to accurately assess and correlate low and high
6
vitamin C supplementation with corresponding levels in plasma and gastric tissue, thus enabling
7
in vivo evaluation of the role of dietary vitamin C in H. pylori-associated gastric disease. We
8
observed that plasma and gastric tissue vitamin C levels in gulo"- mice directly correlated with
9
vitamin C intake, consistent with human studies [19, 38]. Gulo'- mice supplemented with high
10
vitamin C and chronically infected with H. pylori over a period of 32 weeks developed chronic
11
gastritis and atrophy comparable to wt mice, indicating high vitamin C supplementation did not
12
protect infected gulo~' mice from gastritis nor the development of premalignant lesions. In
13
contrast, low vitamin C-supplemented gulo'~ mice tended to have lower degrees
14
inflammation, atrophy and dysplasia and significantly less severe epithelial defects, intestinal
15
metaplasia and hyperplasia relative to high vitamin C-supplemented gulo' and wt mice.
17
Plasma vitamin C levels in gulo'~ mice directly correlated with the level of dietary vitamin C and
18
were not affected by H. pylori infection. Gastric tissue vitamin C levels also correlated with
1
dietary vitamin C intake but in contrast to plasma levels, gastric tissue vitamin C levels were
2
reduced in H. pylori-infected gulo' mice at 16 WPI which is consistent with reduced vitamin C
3
levels in gastric juice during acute H. pylori infection in humans [39]. However, for unknown
4
reasons gastric tissue vitamin C levels were increased in H. pylori-infected gulo- mice at 32
5
WPI. Notably, gastric tissue vitamin C levels were not impacted by H. pylori infection in wt
6
mice which may be confounded by endogenous synthesis of vitamin C in normal mice.
7
Acknowledging that the mechanism of vitamin C secretion into gastric juice is unclear [40], and
8
that it has not been. conclusively shown whether H. pylori infection affects vitamin C levels in
9
human gastric mucosa [20, 41], additional studies are necessary to delineate the influence of H.
10
pylori gastritis on vitamin C levels in the gastric compartment.
11
12
Several clinical trials have examined nutritional interventions using vitamin C alone or in
13
combination with H. pylori eradication or other nutrients in preventing gastric carcinogenesis in
14
humans; none of these studies demonstrated a reproducible benefit of vitamin C supplementation
15
[18, 21, 42, 43]. However, the results of these studies are difficult to interpret given the
16
probability of preexisting H. pylori-associated premalignant lesions in the target populations
17
when vitamin C intervention or H. pylori eradication therapies were initiated [18, 21, 42-44]. To
18
obviate these variables; in the gulo~'- mouse model vitamin C supplementation was administered
100
1
orally at lower and higher levels than the recommended level for maintenance, allowing us to
2
establish dietary levels of vitamin C prior to dosing with H. pylori and subsequent development
3
of H. pylori-associated gastric lesions.
4
5
Despite the limited interpretation of data using rodents that endogenously synthesize vitamin C,
6
rodent studies have reported that 7 to 10 days of vitamin C supplementation reduced gastric
7
inflammation, H. pylori colonization levels, and lipid peroxidation in BALB/c mice given 400
8
mg vitamin C / kg / day [23] and decreased H. pylori colonization levels in Mongolian gerbils
9
given 10 mg vitamin C / day [22]. However, long-term vitamin C supplementation at 50 mg / kg
10
/ day over a period of 52 weeks had no protective effect on severity of gastritis, bacterial
11
colonization levels, and mucosal 8-hydroxydeoxyguanosine levels in H. pylori-infected gerbils
12
[24]. Supplementation with vitamin C in the diet at 10-fold higher levels than a maintenance
13
dose (3102 mg / kg) for 6 weeks in guinea pigs, which lack L-gulono-y-lactone oxidase like
14
humans and gulo~'~ mice, did not impact H. pylori colonization levels nor severity of gastritis
15
[45]. Similar to the guinea pig study, H. pylori-infected gulo~- mice supplemented with high
16
vitamin C had comparable gastric inflammation and premalignant lesions compared to the
17
infected wt mice. Thus, high level supplementation with vitamin C in the gulo~'~ mouse model as
18
well as in guinea pigs did not prevent the development of H. pylori gastric disease or impact
101
1
disease progression.
2
3
Helicobacter-induced gastric disease in humans and mice is mediated by a Thl-predominant host
4
immune response [46-48]. Gulo~'~ mice supplemented with high vitamin C had a similar Thl-
5
promoted inflammatory response to H. pylori as infected wt mice. We unexpectedly observed
6
that low vitamin C-supplemented, H. pylori-infected gulo~' mice had attenuated gastric disease
7
associated with suppressed Th1 responses as evidenced by reduced levels of plasma IgG2c,
8
reduced expression of gastric IFN-y and TNF-a mRNA, and increased H. pylori colonization
9
levels. These data suggest that low dietary vitamin C impairs the host's ability to sustain an
10
inflammatory response and conversely, that adequate vitamin C levels in the infected host may
11
be important for maintaining a robust Th1 immune response to chronic H. pylori infection.
12
Consistent with our data, others have reported attenuated Th1 immune responses in vitamin C-
13
deprived gulo~'~ mice. This was manifested with less severe pneumonia and suppressed
14
pulmonary expression of proinflammatory IL-1$ and TNF-a mRNA in vitamin C-deprived gulo~
15
-mice during the first few days of acute viral influenza [49]. In hosts with low or no vitamin C
16
intake, attenuated Th1 responses to other types of infection, such as tuberculosis in humans or
17
Klebsiella pneumonia sepsis in gulo~'~ mice, may also be important in predicting survival [50,
18
51].
102
2
Using a mouse model of intestinal parasitic infection that causes a Th2 immune response, Fox et
3
al examined the effect of modulating the ThI-associated response to H.felis infection [52]. In the
4
C57BIU6 mice coinfected with helminths and H.felis, reduced systemic Thi immune responses
5
and lower levels of Thl-mediated gastric cytokines were associated with increased H. felis
6
colonization levels and less severe premalignant lesions [52]. These results may in part explain
7
the "African enigma" where the incidence of gastric cancer is low in some African countries,
8
where parasitic infections are common, despite a high prevalence of H. pylori infection [53]. Our
9
findings using the H. pylori-infected, low vitamin C-supplemented gulo-' mouse model may
10
offer another possible explanation to the African enigma. In Gambia, due to the impact of
11
drought on the food supply, mean daily intake of vitamin C approaches zero for 7 months of the
12
year, and is accompanied by low.plasma vitamin C levels [54]. It is tempting to speculate based
13
on our animal studies, that minimal vitamin C intake during H. pylori infection may be one of
14
the factors contributing to a lower incidence of gastric cancer in some populations in Africa due
15
to attenuated gastric inflammation. This hypothesis is also consistent with a report of diminished
16
mitogen responses of peripheral blood mononuclear cells from pigs affected by heritable vitamin
17
C deficiency [55].
18
103
1
Plasma amidated gastrin levels are increased in H. pylori-infected C57BIJ6 mice [8].
2
Overexpression of amidated gastrin promotes the progression of H. pylori-associated gastritis
3
and gastric cancer [28]. In a guinea pig model, vitamin C-deprivation impaired the amidation of
4
gastrin and levels of G-gly were 30-fold higher than normal in the gastric antra [37]. In our
5
study, uninfected gulo-'~ mice supplemented with low vitamin C had undetectable G-gly but
6
comparable levels of amidated gastrin relative to uninfected gulo~'- mice supplemented with high
7
vitamin C. Consistent with previous report [8], H. pylori infection significantly increased plasma
8
amidated gastrin levels in wt mice. H. pylori slightly increased amidated gastrin levels in gulo-'
9
mice supplemented with high vitamin C but not in those supplemented with low vitamin C.
10
Additionally, amidated gastrin levels in H. pylori-infected gulo-'~ mice supplemented with low
11
vitamin C were lower than in infected wt mice and gulo-'- mice supplemented with high vitamin
12
C. These results suggests that the dose used for low vitamin C supplementation is sufficient for
13
gastrin amidation in uninfected animals but not sufficient to sustain increased amidated gastrin
14
levels during H. pylori infection. Higher levels of admidated gastrin in H. pylori-infected gulo
15
mice supplemented with high vitamin C, compared to those with low vitamin C, may in part
16
explain the higher degree of premalignant lesions in the high vitamin C supplemented gulo'
17
mice.
18
104
1
In summary, our study indicates that gulo~' mice, unlike other rodent models, provide a reliable
2
model to study the role of dietary vitamin C in H. pylori-associatedgastric disease. High vitamin
3
C supplementation in this model, similar to previous epidemiological studies in humans, did not
4
prevent progression of H. pylori-induced gastritis and development of premalignant gastric
5
lesions. In contrast, low dietary vitamin C resulted in less severe gastric disease by down-
6
regulating gastric and systemic Th1 immune responses to H. pylori infection. Additional studies
7
to examine more chronic phases of H. pylori gastritis in gulo~' mice and investigation of how low
8
vitamin C intake impacts immune function in this model are warranted.
9
10
ACKNOWLEDGEMENTS
11
Grant support: R01A137750, P01CA26731, P30ES02109 (JGF).
12
We thank Liz Horrigan for animal care; Sandy Xu and Nancy Taylor for culturing H. pylori;
13
Kristen Clapp and Juri Miyamae for performing necropsies; Kathy Cormier for providing
14
histology expertise; Barry Rickman and Arlin B. Rogers for pathological expertise; Zhongming
15
Ge for assisting quantitative PCR; Mark T. Whary for performing ELISA; Xiang-Dong Wang
16
and Bella Gindelsky for expert assistance with vitamin C measurements.
105
I
REFERENCES
2
1.
with gastritisandpeptic ulceration.Lancet, 1984. 1(8390): p. 1311-5.
3
4
2.
5
6
Marshall, B.J. and J.R. Warren, Unidentified curved bacilli in the stomach of patients
Houghton, J., J.G. Fox, and T.C. Wang, Gastric cancer: laboratory bench to clinic. J
Gastroenterol Hepatol, 2002. 17(4): p. 495-502.
3.
IARC Working Group on the Evaluation of Carcinogenic Risk to Humans., in
7
Schistosomes liver flukes and Helicobacterpylori. 1994, International Association for
8
Research against Cancer: Lyon. p. 1-241.
9
4.
10
11
Howson, C.P., T. Hiyama, and E.L. Wynder, The decline in gastric cancer: epidemiology
of an unplanned triumph. Epidemiol Rev, 1986. 8: p. 1-27.
5.
Food, Nutrition and the Prevention of Cancer: a global perspective. 1997, Washington,
12
D.C.: World Cancer Research Fund and American Institute for Cancer Research. 148-
13
175.
14
6.
Harrison, L.E., et al., The role of dietaryfactors in the intestinal and diffuse histologic
15
subtypes of gastric adenocarcinoma: a case-control study in the U.S. Cancer, 1997.
16
80(6): p. 1021-8.
17
7.
18
19
Su, L., et al., Association of dietary antioxidantson the severity of gastritis in a high risk
population.Ann Epidemiol, 2000. 10(7): p. 468.
8.
Fox, J.G, et al., High-salt diet induces gastric epithelial hyperplasia and parietal cell
20
loss, and enhances Helicobacterpylori colonization in C57BL6 mice. Cancer Res, 1999.
21
59(19): p. 4823-8.
22
9.
23
Mannick, E.E., et al., Inducible nitric oxide synthase, nitrotyrosine, and apoptosis in
Helicobacterpylori gastritis: effect of antibiotics and antioxidants. Cancer Res, 1996.
24
56(14): p. 3238-43.
25
10.
Zhang, Z.W., et al., The relation between gastric vitamin C concentrations, mucosal
histology, and CagA seropositivity in the human stomach. Gut, 1998. 43(3): p. 322-6.
11.
Stepan, V., et al., Gastrin biosynthesis in canine G cells. Am J Physiol Gastrointest Liver
26
27
28
29
30
31
Physiol, 2002. 282(5): p. G766-75.
12.
Pinnell, S.R., Regulation of collagen biosynthesis by ascorbic acid: a review. Yale J Biol
13.
Med, 1985. 58(6): p. 553-9.
Correa, P., et al., Review article:Antioxidant micronutrientsand gastric cancer Aliment
32
33
34
35
Pharmacol Ther, 1998. 12 Suppi 1: p. 73-82.
14.
Dabrowska-Ufniarz, E., et al., Vitamin C concentration in gastricjuice in patients with
precancerous lesions of the stomach and gastric cancer Med Sci Monit, 2002. 8(2): p.
CR96-103.
106
1
15.
Jenab, M., et al., Plasma and dietary vitamin C levels and risk of gastric cancer in the
2
European Prospective Investigation into Cancer and Nutrition (EPIC-EURGAST).
3
Carcinogenesis, 2006. 27(11): p. 2250-7.
4
16.
Ekstrom, A.M., et al., Dietary antioxidant intake and the risk of cardia cancer and
5
noncardia cancer of the intestinal and diffuse types: a population-based case-control
6
study in Sweden. Int J Cancer, 2000. 87(1): p. 133-40.
Correa, P., et al., Chemoprevention of gastric dysplasia: randomized trial of antioxidant
7
17.
8
supplements and anti-helicobacterpylori therapy. J Natl Cancer Inst, 2000. 92(23): p.
9
1881-8.
10
18.
11
12
Gut, 2005. 54(11): p. 1536-40.
19.
13
14
15
20.
21.
Bjelakovic, G., et al., Antioxidant supplementsfor prevention of gastrointestinalcancers:
22.
a systematic review and meta-analysis. Lancet, 2004. 364(9441): p. 1219-28.
Zhang, H.M., et al., Vitamin C inhibits the growth of a bacterial risk factorfor gastric
18
19
20
21
Waring, A.J., et al., Ascorbic acid and total vitamin C concentrations in plasma, gastric
juice, and gastrointestinalmucosa: effects of gastritis and oral supplementation. Gut,
1996. 38(2): p. 171-6.
Phull, P.S., et al., Gastroduodenal mucosal vitamin-C levels in Helicobacter pylori
infection. Scand J Gastroenterol, 1999. 34(4): p. 361-6.
16
17
Mera, R., et al., Long term follow up of patients treatedfor Helicobacterpylori infection.
carcinoma:Helicobacterpylori.Cancer, 1997. 80(10): p. 1897-903.
23.
Wang, X., R. Willen, and T. Wadstrom, Astaxanthin-richalgal meal and vitamin C inhibit
22
Helicobacterpylori infection in BALB/cA mice. Antimicrob Agents Chemother, 2000.
23
44(9): p. 2452-7.
24
24.
Sun, YQ., et al., Effects of antioxidant vitamin supplements on Helicobacterpyloriinduced gastritisin Mongolian gerbils. Helicobacter, 2005. 10(1): p. 33-42.
25.
Lee, A., et al., A small animal model of human Helicobacter pylori active chronic
25
26
27
28
gastritis.Gastroenterology, 1990. 99(5): p. 1315-23.
26.
29
30
the Sydney strain.Gastroenterology, 1997. 112(4): p. 1386-97.
27.
31
32
36
Maeda, N., et al., Aortic wall damage in mice unable to synthesize ascorbic acid. Proc
Natl Acad Sci U S A, 2000. 97(2): p. 841-6.
28.
33
34
35
Lee, A., et al., A standardizedmouse model of Helicobacterpylori infection: introducing
Fox, J.G., et al., Helicobacterpylori-associatedgastric cancer in INS-GAS mice is gender
specific. Cancer Res, 2003. 63(5): p. 942-50.
29.
Nakata, Y and N. Maeda, Vulnerable atherosclerotic plaque morphology in
apolipoprotein E-deficient mice unable to make ascorbic Acid. Circulation, 2002.
105(12): p. 1485-90.
107
1
30.
Lee, C.W., et al., Wildtype and IL10-/- T regulatory cells reduced T effector cell-mediated
2
gastroduodenitis in Rag2-/- mice but only wildtype T regulatory cells suppressed
3
Helicobacterpylori gastritis.Infect Immun, 2007.
4
31.
McKay, D.L., et al., The effects of a multivitamin/mineralsupplement on micronutrient
5
status, antioxidantcapacity and cytokine production in healthy older adults consuming a
6
fortified diet. J Am Coll Nutr, 2000. 19(5): p. 613-21.
7
32.
8
9
Rogers, A.B., et al., Helicobacterpylori but not high salt induces gastric intraepithelial
neoplasia in B6129 mice. Cancer Res, 2005. 65(23): p. 10709-15.
33.
Ihrig, M., et al., Gastric helicobacter infection induces a Th2 phenotype but does not
10
elevate serum cholesterol in mice lacking inducible nitric oxide synthase. Infect Immun,
11
2005. 73(3): p. 1664-70.
12
34.
Maurer, K.J., et al., Helicobacterpylori and cholesterol gallstone formation in C57UJ
35.
mice: a prospective study. Am J Physiol Gastrointest Liver Physiol, 2006. 290(1): p.
G175-82.
Varro, A., S. Voronina, and GJ. Dockray, Pathways ofprocessing of the gastrinprecursor
13
14
15
16
17
in rat antralmucosa. J Clin Invest, 1995. 95(4): p. 1642-9.
36.
Martin, R.M., J.L. Brady, and A.M. Lew, The need for IgG2c specific antiserum when
18
isotyping antibodiesfrom C57BUJ6 andNOD mice. J Immunol Methods, 1998. 212(2): p.
19
20
187-92.
Hilsted, L., J.F. Rehfeld, and T.W. Schwartz, Impairedalpha-carboxyamidationof gastrin
37.
21
22
in vitamin C-deficientguinea pigs. FEBS Lett, 1986. 196(1): p. 151-4.
38.
23
24
Levine, M., et al., A new recommended dietary allowance of vitamin Cfor healthy young
women. Proc Natl Acad Sci U S A, 2001. 98(17): p. 9842-6.
39.
25
Sobala, G.M., et al., Acute Helicobacterpylori infection: clinical features, local and
systemic immune response, gastric mucosal histology, and gastric juice ascorbic acid
26
27
concentrations.Gut, 1991. 32(11): p. 1415-8.
Sobala, GM., et al., Ascorbic acid in the human stomach. Gastroenterology, 1989. 97(2):
40.
28
29
p. 357-63.
41.
30
31
32
Drake, I.M., et al., Reactive oxygen species activity and lipid peroxidation in
Helicobacter pylori associated gastritis: relation to gastric mucosal ascorbic acid
33
concentrationsand effect of Hpylori eradication.Gut, 1998. 42(6): p. 768-71.
You, W.C., et al., Randomized double-blindfactorial trial of three treatments to reduce
the prevalence of precancerousgastric lesions. J Natl Cancer Inst, 2006. 98(14): p. 974-
34
83.
35
36
42.
43.
Plummer, M., et al., Chemoprevention of precancerous gastric lesions with antioxidant
vitamin supplementation: a randomized trial in a high-risk population. J Natl Cancer
108
1
2
3
44.
4
45.
Inst, 2007. 99(2): p. 137-46.
Taylor, P.R., Prevention of gastric cancer: a miss. J Natl Cancer Inst, 2007. 99(2): p. 1013.
5
Sjunnesson, H., et al., High intake of selenium, beta-carotene,and vitamins A, C, and E
reduces growth of Helicobacterpylori in the guinea pig. Comp Med, 2001. 51(5): p. 418-
6
23.
7
46.
Mohammadi, M., et al., Murine CD4 T-cell response to Helicobacterinfection: THJ cells
8
enhance gastritisand TH2 cells reduce bacterialload. Gastroenterology, 1997. 113(6): p.
9
1848-57.
10
47.
11
12
117(1): p. 60-9.
48.
49.
50.
51.
19
20
Gaut, J.P., et al., Vitamin C fails to protect amino acids and lipids from oxidation during
acute inflammation. Free Radic Biol Med, 2006. 40(9): p. 1494-501.
17
18
Li, W., N. Maeda, and M.A. Beck, Vitamin C deficiency increases the lung pathology of
influenza virus-infectedgulo-/- mice. J Nutr, 2006. 136(10): p. 2611-6.
15
16
DElios, M.M., et al., T helper 1 effector cells specific for Helicobacterpylori in the
gastricantrum ofpatients with peptic ulcer disease. J Immunol, 1997. 158(2): p. 962-7.
13
14
Fox, J.G and T.C. Wang, Inflammation, atrophy, and gastric cancer J Clin Invest, 2007.
Hemila, H., et al., Vitamin C and other compounds in vitamin C rich food in relation to
risk of tuberculosis in male smokers. Am J Epidemiol, 1999. 150(6): p. 632-41.
52.
Fox, J.G, et al., Concurrent enteric helminth infection modulates inflammation and
21
gastric immune responses and reduces helicobacter-inducedgastric atrophy. Nat Med,
22
2000. 6(5): p. 536-42.
23
53.
Holcombe, C., Helicobacterpylori: the African enigma. Gut, 1992. 33(4): p. 429-31.
24
54.
Bates, C.J., A.M. Prentice, and A.A. Paul, Seasonal variations in vitamins A, C,
25
riboflavin and folate intakes and status of pregnant and lactating women in a rural
26
Gambian community: some possible implications.Eur J Clin Nutr, 1994. 48(9): p. 660-8.
27
28
29
30
55.
Schwager, J. and J. Schulze, Modulation of interleukin production by ascorbic acid. Vet
Immunol Immunopathol, 1998. 64(1): p. 45-57.
109
LEGENDS
Figure 1. Vitamin C levels in plasma (a) and gastric tissue (b). White bars are 16 WPI; Black
bars are 32 WPI. Compared to gulo~'~ mice supplemented with high vitamin C (H), gulo~'~ mice
supplemented with low vitamin C (L) had significantly lower plasma and gastric tissue vitamin C
levels (**, p<0.01; ***, p<0.001). Vitamin C levels in plasma and gastric tissue in gulo-'~ mice
supplemented with high vitamin C were comparable to those in wt mice. H. pylori (Hp) infection
had no significant effect on plasma vitamin C levels or gastric vitamin C levels in gulo~'~ mice
supplemented with low vitamin C or in wt mice. However, gastric tissue vitamin C levels were
reduced at 16 WPI but increased at 32 WPI by H. pylori infection in gulo-'~ mice supplemented
with high vitamin C (*, p<0.05).
Figure 2. Gastric histopathology of H. pylori infection in Gulo-/- with low vitamin C
supplementation (a, d, g) or high vitamin C supplementation (b, e, h) versus wt C57BL/6
controls (c, f, i). (a-c) No significant lesions developed in uninfected mice in any group at 32
weeks. (d-f) Equivalent gastric lesions developed in all H. pylori-infected groups at 16 weeks
post infection (WPI), although gulo~4 mice with low vitamin C supplementation (d) exhibited a
trend for less severe epithelial defects, hyperplasia and dysplasia. (g-i) Equivalent gastric
inflammation developed in all groups at 32 WPI, although gulo~'- mice with low vitamin C
supplementation (g) had less severe epithelial defects, mucous metaplasia and hyperplasia.
Figure 3. Histological scores and total gastric indices in H. pylori (Hp)-infected mice. Gulo'
mice supplemented with high vitamin C had similar gastric lesions and total gastric indices
compared to wt mice at 16 and 32 WPI. (a) At 16 WPI, there was a trend for gulo-'~ mice
110
supplemented with low vitamin C (white bars) compared to those supplemented with high
vitamin C (black bars) for less severe epithelial defects, hyperplasia, and dysplasia (#,
0.05<p<0.10). (b) At 32 WPI, H. pylori-infected gulo-' mice supplemented with low vitamin C
had significantly lower degrees of epithelial defects, mucous metaplasia, and hyperplasia relative
to those supplemented with high vitamin C (*, p<0.05; **, p<0.01). There was a trend toward
less severe atrophy in gulo'~ mice supplemented with low vitamin C compared to those mice
supplemented with high vitamin C (#, 0.05<p<0.10). (c) At 32 WPI, total gastric indices were
significantly lower in gulo~' mice that received low vitamin C than those mice receiving high
vitamin C (*, p<0.05).
Figure 4. Gastric mRNA levels of IFN-y (a) and TNF-a (b) in gulo- and wt mice at 32 WPI.
Fold changes in expression levels were normalized using data from uninfected wt mice. There
were no differences in IFN-y and TNF-a mRNA levels among uninfected wt or gulo-'~ mice that
received low or high vitamin C. H. pylori infection up-regulated IFN-y mRNA levels in gulo'
mice supplemented with low vitamin C (white bar, p<0.01) and to the greatest extent in gulomice supplemented with high vitamin C (black bar) and wt mice (dotted bar) (p<0.001). H.
pylori-infected gulo~'~ mice supplemented with low vitamin C had significantly lower IFN-y
mRNA levels compared to gulo-'~ mice supplemented with high vitamin C or wt mice (p<0.01
and 0.05, respectively). H. pylori infection significantly up-regulated TNF-a mRNA levels in
gulo~'~ mice supplemented with high vitamin C and wt mice (p<0.001), but not in gulo-' mice
supplemented with low vitamin C. Therefore, TNF-cc mRNA levels in gulo-'- mice supplemented
with low vitamin C were significantly lower than the other two groups (p<0.01). (Compared to
uninfected wt mice, ##, p<0.01; ###, p<0.001. Compared to H. pylori-infected gulo-'~ mice
111
supplemented with low vitamin C, *, p<0.05; **, p<0.01).
Figure 5. H. pylori-specific IgG2c and IgG1 levels in H. pylori-infected mice. At 32 WPI, there
was no difference in IgG2c levels among gulo~'~ mice supplemented with low vitamin C (white
bar) and high vitamin C (black bar). Gulo~'' mice supplemented with low or high vitamin C had
lower IgG2c levels than did wt mice (dotted bar) but differences were significant only for gulomice supplemented with low vitamin C (*, p<0.05). IgGI levels among these three groups of
mice were low and similar at 32 WPI (p=0.44).
Figure 6. H. pylori (Hp) colonization levels (log CFU / pg host DNA). At 16 WPI, there were no
differences in H. pylori colonization levels across groups. At 32 WPI, gulo-'~ mice supplemented
with low vitamin C had higher H. pylori colonization levels than gulo~'~ mice supplemented with
high vitamin C and wt mice. (*, p<0.05; #, 0.05<p<0.10).
Figure 7.
Plasma levels of amidated gastrin at 32 WPI. Among uninfected mice, wt mice (dotted bar) and
gulo~'~ mice supplemented with low (white bar) or high (black bar) vitamin C had similar levels
of amidated gastrin (p=O.13) H. pylori infection significantly up-regulated amidated gastrin
levels in wt mice (p<0.01), but not in gulo~'~ mice that received low or high vitamin C.Among H.
pylori-infected mice, gulo~'~ mice supplemented with low vitamin C had amidated gastrin levels
lower than gulo~'- mice supplemented with high vitamin C and wt mice (p=0.065 and <0.05,
respectively). (*, p<0.05; **, p<0.01; #, p=0.65).
112
Fig. 1
250
200
150
100
50
0
gulo~~
wt
Vitamin C L L H H Hp
-+
-++
L
-
gulo~~
L HH
+
-+
gulo~ -
wt
-
+
Hp
-+
-++
gulo~ ~
wt
Vitamin C L L H H -
wt
L L H H-
+
-+
-
+
113
Fig. 2
114
C) Il.)2
0)
000C
Total gastric indices 0
Dys
Hyper
Muc meta
Atrop
Int meta
Infm
Epi defc
Dys
Muc mete
Atrop
Hyper
-.
~.
Int meteC).
Infrr
Epi defc
--4
4...m
--
.........
Scores
Scores
e--]*a
]
p
*
.-
-- '
0
0-4
~jJ
Fig. 4
IFN-y
01)
CP
#*
40
#"
##
TNF-c
**
**
-
30
-C
20
-
10
-
C-
0
LL
rL6ieF
H.pylori
H. pylori
Fig. 5
*
O0 .6-
0
.4
o.20
_
_
_
IgG2c
IgG1
116
o
0
11
o
<4
r Admidated gastrin (pM)
0
ii
*
0:
...............
..............
..
............
Z111111
CO
W~
4N)
.
n(
Log # Hp per pg host DNA
Table 1. H. pylori infection at 16 weeks post infection (WPI)
Vitamin C
Number of mice
Genotype
H. pylori
guloLow
5
gulo'
+
gulo~'~
gulo-'~
wt
+
+
Table 2. H.pylori infection at 32 WPI
Genotype
H. pylon
gulo~~
gulo-'~
+
gulo~'gulo~'~
+
wt
wt
wt
+
+
Low
13
High
5
High
None
8
6
Vitamin C
Low
Low
High
High
None
Number of micea
7/13
11/15
7/13
12/14
3/9
None
High
5/10
0/6
aFirst experiment/second experiment
118
119
Chapter 4: Dysfunction of DNA repair genes 3-alkyladenine DNA glycosylase and O6
methylguanine DNA methyltransferase enhanced development of preneoplastic oxyntic gland
atrophy in the stomach of chronically Helicobacterpylori-infectedmice.
Abstract
Introduction
Materials and Methods
Mice
Bacteria and experimental infection
Experimental design
Histological evaluation
Serum IgG isotype measurement
Quantitative PCR for H.pylori colonization
Measurements of DNA adducts
Statistical Analysis
Results
H. pyloi infection resulted in more severe oxyntic gland atrophy and hyperplasia
in Aag" and Mgmtf mice than in wt mice
Aag4 and Mgmt~' mice developed H. pylori-specific,Thl-predominant IgG2c
responses comparable to infected wt mice
H. pylori colonization levels were comparable between Aag~- or Mgmt* mice and
their wt counterparts
DNA adducts in gastric tissues
Discussion
Acknowledgements
References
Legends
Figures
Tables
121
122
124
124
124
125
125
125
126
126
127
128
128
128
129
132
132
136
138
141
120
ABSTRACT
Dysfunction and polymorphisms of DNA repair enzymes have been associated with a higher risk of
gastric cancer. Due to the fact that Helicobacterpylori causes gastritis and promotes progression to
gastric cancer, we designed studies to examine whether defects in the 3-alkyladenine DNA
glycosylase (Aag) or 0 6-methylguanine DNA methyltransferase (Mgmt) genes would have an effect
on H. pylori-associated gastritis and gastric cancer. To access this question, Aag or Mgmt knockout
mice on a C57BL/6 background were used. Aag~'- mice infected with H. pylori Sydney strain for 32
weeks developed significantly more severe gastric lesions including foveolar hyperplasia (p<0.05),
mucous metaplasia (p<0.001), and precancerous oxyntic gland atrophy (p<0.01) relative to wild type
(wt) mice. H. pylori-infected Mgmt~' mice developed higher degrees of oxyntic atrophy (p<0.05) and
mucous metaplasia (p<0.05) than infected wt mice. No significant differences in inflammation,
epithelial defects, intestinal metaplasia, and dysplasia were observed between H. pylori-infected wt
and Aag~- or Mgmt-' mice. Both H. pylori-infected Aag-' and Mgmt4~ mice had similar levels of H.
pylori-specific, Thl-associated IgG2c response and H. pylori colonization compared to infected wt
mice. Levels of etheno-forms of DNA adducts were statistically similar between H. pylori-infected
wt and Aag-'- mice. Dysfunction of Aag or Mgmt may potentiate H. pylori-associated gastric
carcinogenesis by promoting onset of premalignant gastric atrophy.
121
INTRODUCTION
Helicobacterpylori infects the human stomach [1] and has been strongly associated with gastritis,
peptic ulcer, atrophic gastritis, gastric cancer, and gastric MALToma [2]. H. pylori was classified as a
group I carcinogen by a Working Group of the International Agency for Research on Cancer, a
branch of the World Health Organization in 1994 based on accumulated epidemiologic studies [3].
Infection of C57BIJ6 mice with H. pylori or H. felis results in chronic active gastritis which can
progress to severe dysplasia or in situ gastric adenocarcinoma [4, 5]. In addition to H. pylori
infection, host and environmental factors contribute to the multi-factorial gastric carcinogenesis [2,
3]. Recent studies have identified several genetic factors associated with gastric cancer, including
DNA repair proteins 0 6 -methylguanine DNA methyltransferase (Mgmt) and 8-oxoguanine DNA
glycosylase (Ogg1) [6, 7]. Leukocytes and other phagocytic cells generate reactive oxygen and
nitrogen species (RONS) that damage DNA and other macromolecules during inflammation [8].
RONS cause DNA damage via (i) direct base oxidation and deamination and (ii) indirect alkylation
to form E-base lesions by 4-hydroxyalkenals through lipid peroxidation [9-11]. It has been
demonstrated that gastric helicobacter infection resulted in chronic inflammation accompanied with
higher oxidative stress, and increased levels of 8-hydrocy-deoxyguanine and some alkylated bases
were found in the DNA from H. pylori-infected gastric tissues in vitro and in vivo [12, 13]. The role
of oxidative DNA damage and DNA repair proteins in chronic H. pylori infection has not been
studied.
Aag gene encodes 3-alkyladenine DNA glycosylase (Aag), the major DNA glycosylase that removes
3-methyladenine (3-MeA), deaminated and etheno adducts and hypoxanthine from DNA to initiate
the base excision repair pathway [14]. Mouse embryonic stem cells defective for Aag are more
sensitive to alkylating agents [14] and mitomycin-C [15]. In contrast, bone marrow cells are more
122
resistant to alkylation agents [16]. Aag activity is higher in stomach compared to other organs in
mice, suggesting Aag is important in the gastric compartment [17]. Higher 3-MeA levels were
observed in gastric mucosa with chronic H.pylori infection [13]. However, the role of Aag in chronic
H. pylori gastritis and development of gastric cancer has not been examined. Since Aag
overexpression was associated with chronic active inflammation in patients suffering from ulcerative
colitis and correlated with increased microsatellite instability [18], it is possible that Aag may play an
important role in H. pylori-associated chronic gastritis and gastric cancer.
Mgmt is the DNA repair protein capable of repairing specific types of mutagenic DNA adducts
including 0 6-methylguanine and 0 4-methylthymine [19, 20]. Transcription of Mgmt is frequently
inactivated by promoter hypermethylation in gastric cancer [21]. However, mice lacking Mgmt [22]
do not develop gastric cancer spontaneously, and there is no evidence that 06 -alkylguanine-induced
DNA adducts contribute to the spontaneous mutational spectrum in liver and small intestinal
epithelium from 6-8 week-old and 12 month-old Mgmt~'~ mice [20]. There is no increase in
spontaneous malignancy in the aged Mgmt'~ mice up to 24 months old (unpublished observation).
Therefore, it is not clear whether transcriptional inactivation of Mgmt is the cause or consequence of
gastric carcinogenesis in humans.
To investigate the possible roles of the DNA repair proteins Aag and Mgmt in H. pylori-associated
gastric cancer, C57BL/6 mice lacking functional Aag [16] or Mgmt [22] were infected with H.
pylori. We hypothesized that dysfunction of DNA repair proteins Aag or Mgmt would promote
progression of H. pylori-associated gastric cancer.
MATERIALS AND METHODS
123
Mice
Mice were maintained in an Association for Assessment and Accreditation of Laboratory Animal
Care-accredited facility in static microisolator cages as previously described [23]. Aag-'- mice on a
C57BU6 background were rederived to barrier specific-pathogen-free (SPF) status; Mgmt- mice on
a C57B16 background were housed in a non-barrier SPF facility which is free of human
Helicobacterspp with the exception of H. rodentium. Mice were weaned at 3 weeks of age and fed
ad libitum with regular mouse chow and water. Wild-type (wt) C57BU6 mice were obtained from
Taconic Farms (Germantown, NY) and divided equally into two groups as controls. Dirty bedding
material from cages of Mgmt~ or Aag~'~ mice was mixed with that from cages of corresponding wt
mice to ensure all groups of mice had similar commensal microbial flora. The protocol was approved
by Committee on Animal Care of the Massachusetts Institute of Technology.
Bacteria and experimental infection
H. pylori Sydney strain (SS1) was used for oral inoculation as described previously [23, 24]. The
bacteria were grown for 24 hours at 37*C under microaerobic conditions in Brucella broth with 10%
fetal bovine serum, harvested, and resuspended in PBS. H. pylori was assessed by Gram's stain and
phase microscopy for purity, morphology, and motility. The bacterial concentration was adjusted to
OD60 0=1.000 in PBS. Mice were dosed with 0.2 ml of H. pylori suspension in PBS by gavage every
other day for three doses. Control mice were dosed with PBS.
Experimental design
Both genders of 6-8 week old Mgmt-'-, Aag-, and wt mice were experimentally infected with H.
pylori. Mice were euthanatized with CO 2 at 32 weeks post infection (WPI) (Table 1 and 2).
124
Histological evaluation
At necropsy, the stomach and proximal duodenum were removed and cut along the greater curvature.
Linear gastric strips from the lesser curvature were fixed overnight in 10% neutral-buffered formalin
or 70% ethanol, embedded, cut at 4 pm, and stained with hematoxylin and eosin. Selected tissues
from infected and control mice were characterized with special stains and immunohistochemistry.
Acidic (intestinal type) mucins were shown with Alcian blue at pH 2.5 and neutral (gastric type)
mucins were stained by periodic acid-Schiff [25]. Expression of trefoil factor 2 (TFF2) protein was
stained by immunohistochemistry [26]. Lesions were scored by a veterinary pathologist (BR) blinded
to sample identity as described previously [25].
Serum IgG isotype measurement
Serum collected at necropsy was evaluated for H. pylori-specific IgG1 (Th2-associated isotype) and
IgG2c (Thl-associated isotype) by enzyme-linked immunosorbent assay (ELISA) using an outer
membrane protein (OMP) preparation from H. pylori (SS1 strain) as described previously [27]. In
brief, 96 well flat-bottom plates were coated with 100 g1 OMP (10 pg/ml) overnight at 4*C, and sera
were diluted to a ratio of 1:100. Biotinylated secondary antibodies (clones A85-1 and 5.7, BD
Pharmingen, San Jose, CA) were used for detecting IgG1 and IgG2c. Incubation with extravidin
peroxidase
(Sigma-Aldrich)
followed
by
treatment
with
ABTS
[2,2'-azinobis(3-
ethylbenzthiazolinesulfonic acid)] substrate (Kirkegaard & Perry Laboratories, Gaithersburg, MD.)
was used for color development. The optical density was recorded by a plate reader per
manufacturer's protocol (Power WaveX select, Biotek Instruments, Winooski, VT).
Quantitative PCR for H. pylori colonization
H. pylori colonization levels were quantified using a fluorogenic quantitative PCR assay. In brief,
125
frozen gastric tissues were first digested by proteinase K at 55"C for 8 hours followed by DNA
extraction with phenol: chloroform: isoamyl alcohol (25:24:1) followed by ethanol precipitation.
Genomic DNA was subjected to a quantitative PCR assay using UreB primers [24]. Copy numbers of
gastric H. pylori DNA (colonization forming units, CFU) were normalized to the amount of murine
chromosomal DNA determined by quantitative PCR using a eukaryotic 18S endogenous control
(Applied Biosystems) [24].
Measurements of DNA adducts
Etheno adducts of DNA, 1, N6-ethenodeoxyadenosine (etheno-dA) and 3, N4-ethenodeoxycytosine
(etheno-dC), were quantified from selected tissues of H. pylori-infected or control Aag-'~ and wt
mice. DNA was isolated from stomach using a Genomic DNA isolation Kit for Cells and Tissues
(Roche, Indianapolis, IN) per manufacturer's protocol with some modification. In brief, an
antioxidant (0.1 mM desferrioxamine) and a combination of deaminase inhibitors (5 pg/ml
coformycin, 50
sg/nl
tetrahydrouridine) were added into the lysis buffer to reduce artifactual
oxidative DNA damage during DNA isolation. Etheno adducts were analyzed using an LC-MS/MS
method described previously [28].
Statistical Analysis
Direct comparison of mean histological grades was made by Mann-Whitney U test. Group means of
log-transformed bacterial colonization and serum antibody titers were compared by t test. Graphpad
Prism 4.0 (Graphpad software, San Diego, CA) software was used for all statistic analysis. Statistical
significance was set at p <0.05.
RESULTS
126
H. pylori infection resulted in more severe oxyntic gland atrophy and hyperplasia in Aag"' and
Mgmt' mice than in wt mice
There were no significant gross or microscopic gastric lesions in H. pylori-free Aag'~, Mgmt', or wt
mice (Tale 1, Fig. 1). H. pylori infection in wt mice at 32 weeks post infection (WPI) resulted in
chronic atrophic gastritis of the corpus characterized by infiltration of lymphocytes and granulocytes,
parietal and chief cell loss, and foveolar hyperplasia (Fig. 1). Both H. pylori-infected male and
female wt mice developed similar degrees of gastric lesions (Table 2).
H. pylori-infected Aag-'~ mice developed more severe oxyntic gland atrophy (p<0.01), foveolar
hyperplasia (p<0.05), and mucous metaplasia (p<0.001) in the corpus than did wt mice with H.
pylori infection at 32 WPI (Table 1, Fig. 1). H. pylori infection resulted in similar degrees of
inflammation, epithelial defects, intestinal metaplasia, and dysplasia in both Aag~ and wt mice. H.
pylori-associated gastric lesions were comparable between infected male and female Aag'~ or wt
mice, and both genders of Aag-'- mice with H. pylori infection developed more severe oxyntic gland
atrophy and mucous metaplasia than did their infected wt counterparts (p<0.05 for all comparisions)
(Table 2). Higher degrees of foveolar hyperplasia were observed in H. pylori-infected male Aag-~
mice than in infected male wt mice (p<0.05), but differences in foveolar hyperplasia were not
observed between H. pylori-infected female Aag~'and wt mice (Table 2).
H. pylori-infected Mgmf'~ mice developed more severe oxyntic gland atrophy (p<0.05) and mucous
metaplasia (p<0.05) in the corpus than did wt mice with H. pylori infection at 32 WPI (Table 1). H.
pylori infection resulted in similar degrees of inflammation, epithelial defects, foveolar hyperplasia,
intestinal metaplasia, and dysplasia in both Mgmt~'- and wt mice. However, H. pylori-infected female
Mgmf'~ mice developed more severe oxyntic atrophy (p<0.05) and mucous metaplasia (p<0.05) than
127
did infected female wt mice, irrespective of comparable degrees of other histological parameters
including corpus inflammation, epithelial defects, epithelial hyperplasia, intestinal metaplasia, and
dysplasia (Table 2). Due to the limited number (n=2) of H. pylori-infected male Mgmtf~ mice, gastric
lesions in these mice were not compared with wt male mice.
Aag''' and Mgmt' 1 mice developed H. pyloi-specific, Thi-predominant IgG2c responses
comparable to infected wt mice
At 32 WPI, H. pylori infection resulted in robust Thl-associated IgG2c responses in infected wt
(p<0.01), Aag-l~ (p<0.01) and Mgmf' (p<0.001) mice relative to uninfected mice (Fig. 2). Although
H. pylori induced less robust Th2-associated IgG1 responses than IgG2c responses in wt, Aag~ and
Mgmt4' mice (p<0.001), H. pylori-specific IgGI antibody responses were significantly elevated in
infected Mgmt~'~ mice (p< 0.05) but not in infected Aag~'~ mice (p= 0.13) or wt mice (p=0.10).
Additionally, there were no significant differences in H. pylori-specific IgG2c or IgGI responses
between Aag~'~ or Mgmt-~ mice and their wt counterparts.
H. pyloi colonization levels were comparable between Aag~'- or Mgmt 4 - mice and their wt
counterparts
The mean densities of H. pylori (log CFU/pg genomic DNA) at 32 WPI were comparable between
Aag'~ or Mgrtm'l mice and their wt counterparts (p=0.99 and 0.31, respectively) (Fig. 3).
DNA adducts in gastric tissues
Levels of etheno-dA in H. pylori-uninfected Aag~'~ and wt mice were comparable (Fig. 4). H. pylori
infection resulted in slight but not significant elevation in levels of etheno-dA (wt mice, p=0.28; Aag'mice, p=O.15). Etheno-dC levels were slightly higher in H. pylori-infected wt mice than in infected
128
Aag' mice (p=0.08). Etheno-dC levels were not compared in H. pylori-uninfected mice due to the
limited numbers of DNA samples.
DISCUSSION
The present study investigated the roles of DNA repair proteins Aag and Mgmt in the immune
response and gastric lesions in mice with chronic H. pylori infection. Aag' and Mgmt' mice
infected with H. pylori developed more severe oxyntic gland atrophy, mucous metaplasia, and/or
foveolar hyperplasia, but had similar degrees of inflammation and epithelial defects compared to the
infected wt mice at 32 WPI. These results suggest that Aag and Mgmt play a role in attenuating the
progression of H. pyloi-induced gastric gland atrophy, a premalignant lesion in humans and mice
[29, 30], but these DNA repair proteins have no protective effect on gastric inflammation.
It has been demonstrated that the Th1 proinflammatory response to H. pylori or H. felis is required
for the development of corpus gastritis and progression to parietal cell loss and oxyntic gland atrophy
[31, 32]. Modulation of the balance between Thi and Th2 immune responses elicited by gastric
helicobacter infection varying by host background or by concurrent parasite infection can
dramatically affect the progression of H. pylori-induced gastric diseases [29, 33]. In this study, the
degrees of corpus inflammation, epithelial defects, intestinal metaplasia and dysplasia were
comparable between Aag' or MgmtA mice and their wt controls, whereas more severe gastric
atrophy was noted in mice lacking functional Aag or Mgmt. H. pylori colonization levels are
documented to be inversely correlate with the intensity of H. pylori- and Thl-mediated inflammation
[23, 33]. This observation is consistent with comparable degrees of corpus inflammation, similar
levels of H. pylori colonization and of H. pylori-specific IgG2c responses in Aag' or Mgmt
mice
and their wt control mice. As a result, the more severe gastric atrophy in Aag' or Mgmt
mice
129
suggests that dysfunction of DNA repair proteins Aag or Mgmt may be partially responsible for the
progression of preneoplastic gastric gland atrophy associated with chronic H. pylori infection.
Several previous observations demonstrated that gastric helicobacter infection resulted in chronic
inflammation accompanied with increased oxidative stress and oxidative DNA damage in vitro and
in vivo [12, 13], which may contribute to gastric mutagenic effects of H. pylori and H. felis [34]. It is
unclear, however, how these gastric mutations contribute to gastric carcinogenesis. Whether the
helicobacter-associated oxidative DNA damage is repaired by DNA repair proteins and whether these
DNA mutations reside in epithelial or inflammatory cells is not known. Mice defective in Ogg1 DNA
glycosylase developed less severe H. pylori-associated gastritis and had lower gastric mutant
frequency than the H. pylori-infected wt mice [35]. Our current study demonstrates that H. pylori
infection resulted in more severe preneoplastic gastric atrophy in mice lacking functional Aag or
Mgmt, the proteins that can repair H. pylori-associatedDNA adducts including 3-methyladenine [13]
or 0 6 -alkylguanine
[36]. Importantly, transcriptional inactivation of Mgmt by promoter
hypermethylation is observed in the presence of severe chronic gastric inflammation and intestinal
metaplasia, and has been associated with gastric cancer in humans [7, 21, 37]. Transcriptional
inactivation of gastric Mgmt by DNA hypermethylation in H. pylori-associated gastritis results in
incapability to repair certain mutagenic oxidative DNA adducts, accumulation of gastric mutations,
and progression to gastric cancer. In addition, certain polymorphisms of Mgmt gene have been
associated with a higher risk of gastric and other malignancies [6, 38, 39], which may be related to
differences in Mgmt activity. The role of Aag in human gastritis and gastric cancer may be similar,
since the CpG island within its promoter region [40] may be important for epigenetic regulation in H.
pylori gastritis that predisposes to inactivation of Aag. In this study, we observed slightly higher
gastric levels of etheno-dC in H. pylori-infected wt mice than infected Aag' mice. The elevated
130
etheno-dC might bind to Aag protein, "hijack" its enzyme activity to excise etheno-dA, and resulted
in accumulation of etheno-dA observed in H. pylori-infected wt mice [41]. Further studies are
necessary to examine the roles of DNA repair proteins in the progression to gastric cancer.
TFF2, also known as spasmolytic polypeptide (SP), is mainly secreted by gastric mucous neck cells
[42]. TFF2 is upregulated during acute parietal cell loss induced in rodents treated with DMP-777
and in mice chronically infected with gastric helicobacter [26, 43-46]. These helicobacter-infected
mice developed atrophic gastritis and mucous metaplastic cells that are TFF2-positive. The
spasmolytic polypeptide expressing metaplasia (SPEM) lineages are associated with a strong Thlpredominant immune response and upregulation of IFN-y during chronic helicobacter infection [47,
48]. This TFF2-expressing SPEM lineages are noted in helicobacter-infected mice with severe
dysplasia and in the dysplasic glands of gastric adenocarcinoma in humans [24, 49, 50]. In fact, it has
been documented that SPEM cells are recruited from bone marrow-derived cells (BMDCs) during
chronic H. felis infection. BMDCs recruitment to gastric mucosa is not observed in acute
helicobacter infection or transient mucosal damage [51]. Therefore, SPEM lineages appear to be the
precursor of gastric cancer; chronic inflammation is an essential factor to create a milieu for
recruitment and repopulation of BDMCs in the stomach and for progression to metaplasia, dysplasia,
and cancer. A higher degree of preneoplastic oxyntic gland atrophy in H. pylori-infected Aag~* and
Mgmtr~ mice, consistent with more TFF2-expressing cells, suggests that dysfunction of Aag and
Mgmt potentiates H. pylori-associated gastric carcinogenesis.
In summary, our data demonstrate the roles of DNA repair enzymes Aag and Mgmt in H. pyloriassociated gastric diseases. Compared to wt mice, H. pylori infection in Aag'~ and Mgmt-'- mice
resulted in similar degrees of corpus gastritis and host immune responses against H. pylori
131
manifested as comparable levels of anti-H. pylori IgG2c and H. pylori colonization. Inactivation of
Aag or Mgmt promoted the development of more severe mucous metaplasia and preneoplastic
gastric gland atrophy in the corpus, and may potentiate H. pylori-associated gastric tumorogenicity
through the accumulation of un-repaired mutagenic DNA damage associated with chronic H. pylori
gastritis. Importantly, these results in H. pylori-infected Mgmt~ mice are compatible with previous
reports in humans that transcriptional inactivation of Mgmt in chronic gastritis may predispose to
development of gastric cancer [21, 37]. Aag-~ and Mgmtf- mice provide useful models to dissect the
mechanisms of oxidative DNA damage and DNA repair in carcinogenicity of chronic H. pylori
infection in humans.
ACKNOWLEDGEMENTS
We thank Lisiane B. Meira and Leona D. Samson for providing Aag'~ and Mgmt~ mice; Kathy
Cormier for providing histology expertise; Barry Rickman for pathological expertise; Bo Pang and
Peter C. Dedon for expert assistance with DNA adduct measurements.
REFERENCES
1.
2.
Marshall, B.J. and J.R. Warren, Unidentified curved bacilli in the stomach of patients with
gastritisand peptic ulceration. Lancet, 1984. 1(8390): p. 1311-5.
Houghton, J., J.G. Fox, and T.C. Wang, Gastric cancer: laboratory bench to clinic. J
Gastroenterol Hepatol, 2002. 17(4): p. 495-502.
3.
IARC Working Group on the Evaluation of Carcinogenic Risk to Humans., in Schistosomes
liver flukes and Helicobacter pylori. 1994, International Association for Research against
Cancer: Lyon. p. 1-241.
4.
Lee, A., et al., A small animal model of human Helicobacterpylori active chronic gastritis.
Gastroenterology, 1990. 99(5): p. 1315-23.
5.
Lee, A., et al., A standardizedmouse model of Helicobacterpylori infection: introducing the
Sydney strain. Gastroenterology, 1997. 112(4): p. 1386-97.
6.
7.
8.
Huang, W.Y, et al., Selected DNA repair polymorphisms and gastric cancer in Poland.
Carcinogenesis, 2005. 26(8): p. 1354-9.
Gonzalez, C.A., N. Sala, and G Capella, Genetic susceptibility and gastric cancer risk. Int J
Cancer, 2002. 100(3): p. 249-60.
Coussens, L.M. and Z. Werb, Inflammation and cancer Nature, 2002. 420(6917): p. 860-7.
132
9.
10.
11.
Bartsch, H. and J. Nair, Ultrasensitive and specific detection methods for exocylic DNA
adducts: markersfor lipidperoxidation and oxidative stress. Toxicology, 2000. 153(1-3): p.
105-14.
Bartsch, H. and J. Nair, Potential role of lipid peroxidation derived DNA damage in human
colon carcinogenesis:studies on exocyclic base adducts as stable oxidative stress markers.
Cancer Detect Prev, 2002. 26(4): p. 308-12.
Bartsch, H., J. Nair, and R.W. Owen, Exocyclic DNA adducts as oxidative stress markers in
colon carcinogenesis:potential role of lipid peroxidation, dietaryfat and antioxidants. Biol
Chem, 2002. 383(6): p. 915-21.
12.
Baik, S.C., et al., Increased oxidative DNA damage in Helicobacterpylori-infected human
gastricmucosa. Cancer Res, 1996. 56(6): p. 1279-82.
13.
14.
Arabski, M., et al., DNA damage and repair in Helicobacterpylori-infected gastric mucosa
cells. Mutat Res, 2005. 570(1): p. 129-35.
Engelward, B.P., et al., Base excision repairdeficient mice lacking the Aag alkyladenine DNA
glycosylase. Proc Natl Acad Sci U S A, 1997. 94(24): p. 13087-92.
15.
Allan, J.M., et al., Mammalian 3-methyladenine DNA glycosylase protects against the toxicity
and clastogenicity of certain chemotherapeuticDNA cross-linking agents. Cancer Res, 1998.
58(17): p. 3965-73.
16.
17.
18.
19.
20.
21.
Roth, R.B. and L.D. Samson, 3-Methyladenine DNA glycosylase-deficient Aag null mice
display unexpected bone marrow alkylation resistance. Cancer Res, 2002. 62(3): p. 656-60.
Washington, W.J., et al., Tissue-specific variation in repair activity for 3-methyladenine in
DNA in two stocks of mice. Mutat Res, 1988. 207(3-4): p. 165-9.
Hofseth, L.J., et al., The adaptive imbalance in base excision-repair enzymes generates
microsatelliteinstability in chronic inflammation. J Clin Invest, 2003. 112(12): p. 1887-94.
Singer, B. and B. Hang, What structuralfeatures determine repair enzyme specificity and
mechanism in chemically modifed DNA? Chem Res Toxicol, 1997. 10(7): p. 713-32.
Sandercock, L.E., et al., Mutational-reportertransgenes rescued from mice lacking either
Mgmt, or both Mgmt and Msh6 suggest that 06-alkylguanine-inducedmiscoding does not
contribute to the spontaneous mutationalspectrum. Oncogene, 2004. 23(35): p. 5931-40.
Oue, N., et al., Promoter hypermethylation of MGMT is associated with protein loss in
gastric carcinoma. Int J Cancer, 2001. 93(6): p. 805-9.
22.
23.
Glassner, B.J., et al., DNA repairmethyltransferase(Mgmt) knockout mice are sensitive to the
lethal effects of chemotherapeuticalkylating agents. Mutagenesis, 1999. 14(3): p. 339-47.
Fox, J.G, et al., Helicobacterpylori-associatedgastric cancer in INS-GAS mice is gender
specific. Cancer Res, 2003. 63(5): p. 942-50.
24.
25.
Fox, J.G., et al., Host and microbial constituents influence Helicobacter pylori-induced
cancer in a murine model of hypergastrinemia.Gastroenterology, 2003. 124(7): p. 1879-90.
Rogers, A.B., et al., Helicobacter pylori but not high salt induces gastric intraepithelial
neoplasia in B6129 mice. Cancer Res, 2005. 65(23): p. 10709-15.
26.
27.
Wang, T.C., et al., Mice lacking secretory phospholipase A2 show altered apoptosis and
differentiation with Helicobacterfelisinfection. Gastroenterology, 1998. 114(4): p. 675-89.
Ihrig, M., et al., Gastric helicobacter infection induces a Th2 phenotype but does not elevate
serum cholesterol in mice lacking inducible nitric oxide synthase. Infect Immun, 2005. 73(3):
p. 1664-70.
28.
29.
Pang, B., et al., Lipid peroxidation dominates the chemistry of DNA adduct formation in a
mouse model of inflammation. Carcinogenesis, 2007.
Fox, J.G, et al., Concurrent enteric helminth infection modulates inflammation and gastric
immune responses and reduces helicobacter-inducedgastric atrophy. Nat Med, 2000. 6(5): p.
133
30.
536-42.
Correa, P., et al., Gastriccancer in Colombia. IlL Natural history of precursorlesions. J Nat]
Cancer Inst, 1976. 57(5): p. 1027-35.
31.
32.
Roth, K.A., et al., Cellular immune responses are essential for the development of
Helicobacterfelis-associatedgastricpathology. J Immunol, 1999. 163(3): p. 1490-7.
Mohammadi, M., et al., Helicobacter-specific cell-mediated immune responses display a
predominant Thi phenotype and promote a delayed-type hypersensitivity response in the
stomachs of mice. J Immunol, 1996. 156(12): p. 4729-38.
33.
Sakagami, T., et al., Atrophic gastric changes in both Helicobacterfelis and Helicobacter
pylori infected mice are host dependent and separatefrom antral gastritis.Gut, 1996. 39(5):
35.
p. 639-48.
Touati, E., et al., Chronic Helicobacterpylori infections induce gastric mutations in mice.
Gastroenterology, 2003. 124(5): p. 1408-19.
Touati, E., et al., Deficiency in OGG1 protects against inflammation and mutagenic effects
36.
associatedwith H. pylori infection in mouse. Helicobacter, 2006. 11(5): p. 494-505.
Palli, D., et al., 06-alkylguanines, dietary N-nitroso compounds, and their precursors in
34.
gastriccancer Nutr Cancer, 2001. 39(1): p. 42-9.
37.
Kang, G.H., et al., Profile of aberrant CpG island methylation along multistep gastric
carcinogenesis.Lab Invest, 2003. 83(4): p. 519-26.
38.
Han, J., et al., Polymorphisms in 06-methylguanine DNA methyltransferase and breast
cancer risk. Pharmacogenet Genomics, 2006. 16(7): p. 469-74.
39.
Tranah, G.J., et al., 06-methylguanine-DNA methyltransferase Leu84Phe and llel43Val
polymorphisms and risk of colorectal cancer in the Nurses' Health Study and Physicians'
Health Study (United States). Cancer Causes Control, 2006. 17(5): p. 721-3 1.
40.
41.
Vickers, M.A., et al., Structure of the human 3-methyladenine DNA glycosylase gene and
localization close to the 16p telomere. Proc Natl Acad Sci U S A, 1993. 90(8): p. 3437-41.
Gros, L., et al., Hijacking of the human alkyl-N-purine-DNA glycosylase by 3,N4ethenocytosine, a lipid peroxidation-inducedDNA adduct. J Biol Chem, 2004. 279(17): p.
17723-30.
42.
Taupin, D. and D.K. Podolsky, Trefoil factors: initiatorsof mucosal healing. Nat Rev Mol
Cell Biol, 2003. 4(9): p. 721-32.
43.
44.
45.
46.
Nomura, S., et al., Alterations in gastric mucosal lineages induced by acute oxyntic atrophy
in wild-type and gastrin-deficient mice. Am J Physiol Gastrointest Liver Physiol, 2005.
288(2): p. G362-75.
Goldenring, J.R., et al., Reversible drug-induced oxyntic atrophy in rats. Gastroenterology,
2000. 118(6): p. 1080-93.
Fox, J.G, et al., Hypertrophic gastropathy in Helicobacterfelis-infected wild-type C57BI6
mice and p53 hemizygous transgenicmice. Gastroenterology, 1996. 110(1): p. 155-66.
Ogawa, M., et al., Altered metaplastic response of waved-2 EGF receptor mutant mice to
acute oxyntic atrophy.Am J Physiol Gastrointest Liver Physiol, 2006. 290(4): p. G793-804.
47.
Kang, W., et al., Interferon gamma induction of gastric mucous neck cell hypertrophy. Lab
Invest, 2005. 85(5): p. 702-15.
48.
49.
50.
Sawai, N., et al., Role of gamma interferon in Helicobacter pylori-induced gastric
inflammatory responses in a mouse model. Infect Immun, 1999. 67(1): p. 279-85.
Wang, T.C., et al., Synergistic interaction between hypergastrinemia and Helicobacter
infection in a mouse model of gastriccancer Gastroenterology, 2000. 118(1): p. 36-47.
Schmidt, P.H., et al., Identification of a metaplastic cell lineage associated with human
gastricadenocarcinoma.Lab Invest, 1999. 79(6): p. 639-46.
134
51.
Houghton, J., et al., Gastric cancer originatingfrom bone marrow-derived cells. Science,
2004. 306(5701): p. 1568-71.
135
LEGENDS
Fig. 1.
Histopathology of gastric disease and mucous metaplasia induced by H. pylori infection in the fundic
mucosa of C57BIJ6 wt or Aag" mice. Representative H&E, AB/PAS -stained, and anti-TFF2
immunohistochemistry (DAB) of stomachs from 38-week old wt (a,b,c), Aag~'- infected (d,e,f) and
wt infected (g,h,i) mice. Uninfected mouse stomach showing normal microscopic architecture on
H&E stain (a) and a thin surface lining of gastric-type neutral mucins (red) by AB/PAS (b). AntiTFF2 antibody stained mucous neck cells in the fundic mucosa with high specificity (c). Tissue from
H-pylori-infected Aag-'~ mice (32 WPI) exhibited loss of oxyntic cell mass due to mucous metaplasia
characterized by foamy change of oxyntic cell cytoplasm on H&E stain (d) and prominent
coexpression of both intestinal-type acid mucins (blue) and gastric type mucins (red) in mucous
metaplastic glands shown by AB/PAS stain (e). Anti-TFF2 antibody showed similar staining pattern
in the mucous metaplastic glands (f) as shown in (e). Tissue from H-pylori-infected wt mice (32
WPI) exhibited mild loss of oxyntic cell mass with minimal mucous metaplasia of oxyntic cell
cytoplasm on H&E stain (g) and mild coexpression of both intestinal-type acid mucins (blue) and
gastric type mucins (red) in mucous metaplastic glands shown by AB/PAS stain (h). Anti-TFF2
antibody showed fewer metaplastic glands but similar staining pattern in the mucous metaplastic
glands (i) as shown in (h).
Fig. 2.
Serum IgGI and IgG2c responses to H. pylori (H.pylori) in Aag-'~ (a) or Mgmt~'~ (b) mice and their
wt controls at 32 weeks post infection. Data were presented as mean values + standard errors. H.
pylori infection resulted in more robust Thl-associated IgG2c responses than Th2-associated IgGI
responses in Aag~'-, Mgmf', and wt mice (p<0.001). Both mutant mice developed comparable IgG2c
136
responses against H. pylori compared to infected wt mice, suggesting that dysfunction of Aag or
Mgmt had no effect on H. pylori-specific immunity.
Fig. 3.
H. pylori colonization levels in the stomach. Data were presented as log transformed colony
formatting unit (CFU) per gg of genomic DNA. H. pylori colonization levels were not different
between Aag~'- and wt mice or Mgmt'~ and wt mice. ns, non-significant.
Fig. 4.
Levels of etheno-dA and etheno-dC in gastric DNA. Data were presented as number of adducts per
107 nucleoside. There were trends toward elevated levels of etheno-dA in H. pylori-infected wt and
Aag~'~ mice, but the differences were not significant. H. pylori-infected wt mice had slightly higher
levels of etheno-dC than infected Aag~ mice (p=0.0 8).
137
Fig. 1
wt infected
Aag~/~ infected
wt
o6
".
Ul)
"
:"~~
r
~jf~
LL
LL
-
~
it
'
V
A
~
.~.
C
s#'~~...rV' ~tV;V~
4
138
Fig. 2
1.0
1.0
-
T
T
~T
0.5 .
0.5
77~
0
w
0
IgG2c
IgG1
wt
Aag'
0-
~ -
-
IgG2c
IgG1
Mgmt-/-
Aag-/-
-
wt
Mgmt-/-
wt
Fig. 3
ns
5I
ns
II
I
432-
Aag'
wt
Mgmt
139
Fig. 4
p=0. 08
b
---
.
0
4:
(D
-
-
+
Aag
wt
Aag-/-
wt
H. pylori
C
H. pylori
-
-
+
140
TABLE 1. Gastric lesions in Helicobacterpylori(H.pylori)-infected wild type (wt), Aag-'- and Mgmt~'~ mice
Median (range) of lesion scores ina_
no.
7
Infm
0
Epi defc
0
Int meta
0
Atroph
0
Muc meta
Hyper
Dys
0
0
0
+
11
1(0-2)
0(0-1)
0
1 (0-2)c
0.5 ( 0 -3 )d
0.5 (0-1)b
0
None
6
0
0
0
0
0
0
0
0 (0-1)
Genotype H. pylori
None
wt
wt
Aag
+
13
1.5 (1-2.5)
1 (0-2)
0 (0-0.5)
2 (1.5-3)c
2.5 ( 2 -3. 5 )d
1 (0-2)b
None
+
5
7
0
1 (0.5-2.5)
0
0.5 (0-1)
0
0 (0-1)
0
1.5 (0.5-3)e
0
0
0
2 (1-3)e
1 (0-1)
0 (0-0.5)
Ngmt-
None
0
0
0
0
0
0
0
Mgmt-
+
13
11
1(0.5-1.5)
0.5 (0-2)
0(0-0.5)
2.5 (1.5-3)e
3 (2.5-3)e
1(0.5-1.5)
0(0-1)
Aag
wt
wt
Abbreviations: Infm, inflammation; Epi defc, epithelial defects; Int meta, intestinal metaplasia; Atroph, oxyntic atrophy; Muc meta,
mucous metaplasia; Hyper, foveolar hyperplasia; and Dys, dysplasia.
aComparision was made between Aag~'~ or Mgmt~'~ mice and wt mice by Mann-Whitney U test.
bSignificant influence of genotype (Aag~'~ vs wt) at a p<0.05.
'Significant influence of genotype (Aag~'~ vs wt) at a p<0.01.
dSignificant influence of genotype (Aag~'- vs wt) at a p<0.001.
eSignificant influence of genotype (Mgmt'- vs wt) at a p<0.05.
141
TABLE 2. Gastric lesions in Helicobacterpylori (H. pylori)-infected wild type (wt), Aag~'~ and Mgmt-' mice
Median (range) of lesion scores ina
Genotype H. pylori Sex' (no.)
Infm
Epi defc
Int meta
Atroph
Muc meta
Hyper
Dys
+
M (6)
1(0-2)
0(0-1)
0
0.75 (0-2)"
1 (0-3)"
F (5)
0.5 (0-1)
0.5 (0-2)
0
0 (0-0.5)
1 (0-1.5)b
2 (1.5-3)"
0.5 (0-0.5)b
2.5 (2-3.5)b
0
0
0 (0-1)
0(0-0.5)
0 (0-0.5)
0
wt
Aag
wt
Mgmt-,-
+
M (9)
1.5 (1-2)
1.5 (1-2.5)
+
F (4)
M (4)
1.25 (1-2)
1 (0.5-2.5)
1.5 (0-1.5)
0.75 (0-1)
0(0-0.5)
0.25 (0-0.5)
1.75 (1.5-2)b
2 (1-3)
2 (2-2.5)
2.5 (1.5-3)
0.25 (0-1)"
0.5 (0-1)
1 (0-2)b
0.75 (0-1)
1
F (3)
1(0.5-1.5)
0.5 (0-1)
0 (0-1)
1 (0.5-2)c
1.5 (1-2)'
0 (0-1)
M (2)
1.25 (1-1.5)
0.75 (0.5-1)
0
2.25 (2-2.5)
3
0.75 (0.5-1)
0
0.5 (0-2)
0.5 (0-0.5)
3 (1.5-3)'
3 (2.5-3)d
1 (0.5-1.5)
0 (0-1)
+
F (9)
1 (0.5-1.5)
Abbreviations: Infm, inflammation; Epi defc, epithelial defects; Int meta, intestinal metaplasia; Atroph, oxyntic atrophy; Muc meta,
mucous metaplasia; Hyper, foveolar hyperplasia; and Dys, dysplasia.
aComparision was made between same gender of Aag-/- or Mgmt- mice and wt mice by Mann-Whitney U test.
bSignificant influence of genotype (Aag~'~ vs wt) at a p<0.05.
cSignificant influence of genotype (Mgmt~' vs wt) at a p<0.05.
dSignificant influence of genotype (Mgmt~-i vs wt) at a p<0.01.
eM, male; F, female.
142
143
Chapter 5: Helicobacter pylori Eradication Prevents Progression of Gastric Cancer in
Hypergastrinemic INS-GAS Mice
Abstract
Introduction
Materials and Methods
Mice
Experimental design
Tissue collection and histological analysis
Confirmation of H. pylori eradication by quantitative PCR
Serum H. pylori-specific antibodies
Quantitative analysis of mRNA expression
Measurements of DNA adducts
Analysis of lipids in serum
Statistical analysis
Results
Eradication of H. pylori was confirmed in INS-GAS mice that received
eradication therapy
H. pylori infection promoted developed of gastric cancer in INS-GAS mice
H.pylori eradication at 8 WPI reduced gastritis and premalignant lesions to the
greatest extent
H. pylori eradication at 8 WPI prevented progression to low-grade and high-grade
gastrointestinal intraepithelial neoplasia
H. pylori-specific antibody responses
Gastric IFN-y, TNF-a, and iNOS mRNA levels were reduced in all mice that
received H. pylori eradication therapy
Gastric mucosal cell proliferation was also reduced in all mice that received H.
pylori eradication therapy
DNA adducts in gastric tissues
H. pylori-infected INS-GAS mice had significantly higher serum levels of
triglyceride, total cholesterol, and elevated fraction of very-low-density
lipoprotein
Discussion
Acknowledgements
References
Legends
Figures
Tables
145
146
148
148
148
149
150
150
151
151
152
153
153
154
155
156
157
158
159
159
161
166
168
171
175
179
144
ABSTRACT
Helicobacterpylori infection results in chronic gastritis which may progress to gastric cancer. In
the H. pyloi-infected INS-GAS mouse, a model of gastric cancer, omeprazole, metronidazole,
and clarithromycin were administered at 8, 12, or 22 weeks post H. pylori infection (WPI) to
evaluate the efficacy of eradication in preventing the development of gastric cancer. H. pylori
infection resulted in severe dysplasia classified as high- and low-grade gastrointestinal
intraepithelial neoplasia (GIN) in INS-GAS mice at 28 WPI. H. pylori antimicrobial therapy at 8,
12 and 22 WPI significantly reduced the severity of dysplsia (p<0.01). Moreover, H. pylori
eradication at 8 WPI completely prevented development of GIN (p<0.001) with the return of
mucosal architecture to that of uninfected mice. Although not as effective as early antimicrobial
treatment, prevention of progression to high-grade GIN was achieved by H. pylori eradication at
12 and 22 WPI (p<0.05 or less). Consistent with reduced gastric pathology, H. pylori eradication
at all time points significantly down-regulated gastric IFNy, TNFx, and iNOS mRNA levels
(p<0.05) and reduced epithelial proliferation in the corpus (p<0.01). Additionally, elevated
triglyceride and total cholesterol concentrations in serum were observed in infected INS-GAS
mice compared to other treatment groups. We conclude that H. pylori eradication may prevent
progression to gastric cancer to the greatest extent when antibiotics are given at an early point of
H. pylori infection. Mice with severe dysplastic lesions may benefit from H. pylori eradication
therapy given at a later time point although it is not as effective as early H. pylori eradication.
145
INTRODUCTION
Helicobacterpylori was first identified in the antrum of patients with active chronic gastritis and
peptic ulcers [1]. Base on epidemiological evidence, H. pylori was identified as the major cause
of gastric cancer and has been classified as a group I carcinogen of gastric cancer by a Working
Group of the International Agency for Research on Cancer, a branch of the World Health
Organization in 1994 [2, 3]. H. pylori infection causes a persistent chronic gastritis, which in
susceptible individuals may progress to atrophy, intestinal metaplasia, dysplasia, and finally
intestinal-type gastric cancer [3, 4]. Although a clinical trial failed to demonstrate a protective
effect of H. pylori eradication against progression of preneoplastic lesions such as gastric atrophy
and intestinal metaplasia [5], eradication of H. pylori in humans has been associated with
prevention or regression of preneoplastic lesions [6-10]. Some epidemiological studies observed
that antibiotic treatment during knee or hip replacement surgery reduced the incidence of gastric
cancer in the following years, probably due to eradication of H. pylori [11, 12]. In intervention
studies, the optimal effect of antibiotic eradication therapy in preventing gastric cancer has been
observed in H. pylori infected patients who did not have precancerous lesions prior to
antimicrobial H. pylori eradication therapy (p<0.05) [13]. However, H. pylori eradication did not
reduce the overall prevalence of dysplasia or gastric cancer in another studies [7, 10, 13, 14].
Since it takes decades for gastric cancer to develop in susceptible hosts acquiring H. pylori
infection at an early age [4], these H. pylori eradication trials continue to pose key questions; in
which patients would H. pylori eradication be beneficial in preventing gastric cancer and at what
stage of gastric disease would H. pylori antimicrobial eradication prevent the progression of
gastric lesions to gastric cancer.
146
Several animal models have been used to examine whether H. pylori eradication is effective in
reversal of preneoplasitic gastric lesions and preventing the progression of these preneoplastic
lesions to gastric cancer. Antibiotic eradication therapy reversed the histologic progression of
dysplasia in H. pylori-infected Mongolian gerbils [15, 16]. In Helicobacter felis-infected
C57BIJ6 mice that developed gastric cancer within 24 months post infection, antibiotic therapy
given at 2 or 6 months post infection (MPI) led to a regression of inflammation and reversion of
premalignant lesions. Antibiotics given at 12 MPI arrested progression of dysplasia and reduced
the risk of gastric cancer [17]. These in vivo data support the hypothesis that eradication therapy,
depending on the timing of antibiotic treatment, may be effective in preventing helicobacter
gastritis from progressing to gastric cancer.
Recent studies suggest an association with hypergastrinemia, helicobacter infection and gastric
cancer in humans and mice [18-21]. In the absence of helicobacter infection, transgenic INSGAS mice that over-express amidated gastrin have a higher gastric acid secretion and an
increased parietal cell number when they are 1-3 months old. As the mice age, they lose parietal
cell mass and develop hypochlorhydria, gastric atrophy, metaplasia, and dysplasia. At 20 months
of age, helicobacter-uninfected INS-GAS mice develop invasive gastric cancer [20, 22]. The
development of gastric cancer is accelerated by helicobacter infection and lesion severity is more
profound in male INS-GAS mice [18, 20, 21]. The purpose of this study was to examine the
effect of H. pylori eradication at different stages of progression from gastritis to gastric cancer in
INS-GAS mice.
147
MATERIALS AND METHODS
Mice
The animal protocol was reviewed and approved by the Massachusetts Institute of Technology
Committee on Animal Care. Specific pathogen-free (including Helicobacter spp.) male INSGAS mice on a FVB/N background were used in this study [22]. Mice were maintained in an
AAALAC accredited facility and housed on hard wood bedding in microisolator, solid-bottomed
polycarbonate cages, and given a normal rodent diet (Prolab RMH 3000, PMI Nutrition
International, Richmond, Indiana) and water ad libitum. Age-match male wt FVB/N mice were
from retired breeding colony.
Experimental design
Fifty-four 6-8 week old mice were infected by oral gavage with H. pylori (SS1 strain) on
alternate days for a total of 3 doses [23, 24]. The H. pylori inoculum for oral gavage was adjusted
with phosphate-buffered saline (PBS) to an optical density of 1.0 at 600 nm [23]. Helicobacteruninfected mice were sham-dosed with PBS. Infected mice were orally dosed with omeprazole
(400 pmol/kg/day, Sigma-Aldrich, St. Louis, MO), metronidazole (14.2 mg/kg/day, SigmaAldrich), and clarithromycin (7.15 mg/kg/day, Abbott Laboratory, North Chicago, IL) in 0.2 ml
twice a day for 7 days [25]. This antimicrobial therapy was previously demonstrated to eradicate
100% of mice infected with H. pylori was instituted [25]. Treatment was administered at 8, 12, or
22 weeks post H.pylori infection (WPI).
Tissue collection and histological analysis
Mice were euthanized at 28 WPI. After CO 2 asphyxia, blood was collected by cardiac puncture.
148
The stomach and proximal duodenum were removed and cut along the greater curvature. Linear
gastric strips from the lesser curvature were fixed overnight in 10% neutral-buffered formalin,
embedded, cut at 4 pm, and stained with hematoxylin and eosin (H&E). Tissue sections were
scored for gastric lesions using previously published criteria by veterinary pathologists (B.R and
A.B.R) blinded to sample identities [26]. A dysplasia score of 3.0 is considered carcinoma in situ
or low-grade gastrointestinal intraepithelial neoplasia (GIN). Dysplasia scores equal or higher
than 3.5 represent intramucosal carcinoma or herniated, high-grade GIN [27]. Both low- and
high-grade GIN in mouse stomachs are diagnosed as gastric cancer. Ki67 immunostaining (BD
Biosciences, San Diego, CA) was used to measure epithelial proliferation of gastric mucosa. The
ratio of positive to total epithelial nuclei in glands occupying the full length of the proximal
corpus was quantified manually for the Ki67 labeling index (LI), and results were averaged. The
remainder of the gastric tissue was snap-frozen in liquid nitrogen and stored in -70 'C for DNA
and RNA analysis.
Confirmation of H. pylori eradication by quantitative PCR
A longitudinal strip of gastric tissue from the greater curvature was digested with proteinase K at
550C overnight followed by DNA extraction with phenol:chloroform:isoamyl alcohol (25:24:1)
and ethanol precipitation. H. pylori copy numbers were quantified using a fluorogenic
quantitative PCR assay with urease B primers [28]. H. pylori copy numbers were normalized to
microgram of murine genomic DNA as determined by quantitative PCR using a eukaryotic 18S
endogenous control (Applied Biosystems) (User Bulletin #2, Applied Biosystems, Foster City,
CA).
149
Serum H. pylori-specific antibodies
Sera collected at necropsy were evaluated for H. pylori-specific, Th2-associated IgG1 and ThIassociated IgG2a by enzyme-linked immunosorbent assay (ELISA) using an outer membrane
protein preparation from H. pylori (SS1 strain) as described previously [29]. In brief, 96 well
flat-bottom plates were coated with 100 pl antigen (10 pg/ml) overnight at 4*C; sera were diluted
to a ratio of 1:100 and added to the wells. Biotinylated secondary antibodies included goat antimouse antibodies were used for detecting IgG1 and IgG2a (clones A85-1 and 5.7, BD
Pharmingen, San Jose, CA). Incubation with extravidin peroxidase (Sigma-Aldrich) was
followed by treatment with ABTS [2,2'-azinobis(3-ethylbenzthiazolinesulfonic acid)] substrate
(Kirkegaard & Perry Laboratories, Gaithersburg, MD.) for color development. The optical
density was recorded at OD405 and OD595 by a plate reader per manufacturer's protocol (Power
WaveX select, Biotek Instruments, Winooski, VT)
Quantitative analysis of mRNA expression
Two mice in which H. pylori eradication was unsuccessful were excluded from further analysis
(Table 1). A longitudinal strip of gastric tissue from the anterior wall was harvested and snapfrozen in liquid nitrogen. Total RNA was extracted with Trizol reagent (Invitrogen, Carlsbad,
CA). cDNA was synthesized from 5 pg of total RNA with High Capacity cDNA Archive kit
(Applied Biosystems). IFN-y and TNF-a mRNA levels were quantified with SYBR Green PCR
reagent (Qiagen, Valencia, CA) using primers for INFy: CATGGCTGTTTCTGGCTGTTACTG
(Forward [F]) and GTTGCTGATGGCCTGATTGTCTTT (Reverse [R]) annealing at 560C; for
TNFa: CATCTTCTCAAAATTCGAGTGACAA (F) and TGGGAGTAGACAAGGTACAACCC
(R) annealing at 60*C. The final concentration of each primer was 0.3 gM. iNOS and
150
glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA levels were quantified with
TaqMan gene expression assays (Applied Biosystems) in an ABI Prism Sequence Detection
System 7700. mRNA levels of each cytokine were normalized to the mRNA level of internal
control glyceraldehyde-3-phosphate
dehydrogenase (GAPDH)
and compared to data of
uninfected mice using the AACT method (User Bulletin #2, Applied Biosystems).
Measurements of DNA adducts
Adducts of DNA, 2'-deoxyinosine (dl), N6 -ethenodeoxyadenosine (etheno-dA), and 8-oxo-7,8dihydro-2'-deoxyguanosine (8-oxodG), were quantified from selected tissues of H. pyloriinfected or control INS-GAS and wt mice. DNA was isolated from stomach using a Genomic
DNA isolation Kit for Cells and Tissues (Roche, Indianapolis, IN) per manufacturer's protocol
with some modification. In brief, an antioxidant (0.1 mM desferrioxamine) and a combination of
deaminase inhibitors (5 pg/ml coformycin, 50 pg/ml tetrahydrouridine) were added into the lysis
buffer to reduce artifactual oxidative DNA damage during DNA isolation. DNA adducts were
analyzed using an LC-MS/MS method described previously [30].
Analysis of lipids in serum
Sera collected from unfasted animals at necropsy were evaluated for triglyceride and total
cholesterol concentrations enzymatically using reagents from Roche (Indianapolis, IN) and
Sigma, respectively [31]. Lipoproteins were separated by fast-protein liquid chromatography
after equal volumes of serum from 5 mice per group were pooled. Cholesterol concentrations in
individual fractions were determined enzymatically [32].
151
Statistical analysis
Infected mice with successful H. pylori eradication were included in the analysis. Gastric lesion
scores and Ki67 LI for proliferation indices were compared by the Mann-Whitney U test.
Cytokines and iNOS expression levels were compared using the student t test. Incidences of no,
low-grade, and high-grade GIN were compared by Fisher's exact t test. DNA adduct levels and
triglyceride and total cholesterol concentrations were compared by Student t test. Statistical
analysis was performed using a commercial software (Graphpad Prism 4.0, GraphPad Software,
Inc., San Diego, CA) with significance at p<0.05.
152
RESULTS
Eradication of H. pylori was confirmed in INS-GAS mice that received eradication therapy
To assess the effect of H. pylori eradication on progression of gastric cancer in INS-GAS mice,
eradication therapy using the combination of omeprazole, metronidazole, and clazithromycin
was administered orally to mice at 8, 12, or 22 WPI. Eradication of H. pylori was confirmed at
necropsy using quantitative PCR. H. pylori was successfully eradicated in all mice treated at 8 or
12 weeks post infection (WPI). Twelve of fourteen (85.7%) animals that underwent eradication
therapy at 22 WPI were successfully H. pylori eradicated (Table 1).
H. pylori infection promoted developed of gastric cancer in INS-GAS mice
Helicobacter infection promotes gastric carcinogenesis in INS-GAS mice particularly in male
animals [20, 21]. Consistent with a previous study [21], uninfected INS-GAS mice at 28 to 34
weeks of age developed progressive gastric lesions including atrophy, epithelial defect, and
dysplasia (p<0.05) accompanied with minimal inflammation, severe hyperplasia and intestinal
metaplasia (Table 2). Corpus hypertrophy in infected INS-GAS mice at 22 and 28 WPI was
noted at necropsy as thickened mucosal folds. The H. pylori-infected INS-GAS mice had chronic
atrophic gastritis with profound changes in mucosal architecture. These histologic changes were
mainly restricted to the corpus and were characterized by loss of parietal and chief cells (Fig 1).
Compared to age-matched uninfected mice, infected INS-GAS mice had more severe
inflammation (p<0.001), oxyntic atrophy (p<0.05), hyperplasia (p<0.05), epithelial defects
(p<0.001), intestinal metaplasia (p<0.05) and dysplasia (p<0.001) at 22 WPI, and higher degrees
of inflammation (p<0.001), dysplasia (p<0.001), and hyperplasia (p<0.01) at 28 WPI (Fig. 2).
Compared to infected mice at 22 WPI, infected mice at 28 WPI had more severe dysplasia and
153
atrophy (p<0.05).
H. pylori eradication at 8 WPI reduced gastritis and premalignant lesions to the greatest
extent
H. pylori-infected INS-GAS mice received antimicrobial eradication therapy at 8, 12, or 22 WPI
and were euthanized at 28 WPI. Infected mice that received H. pylori eradication therapy at 8
WPI had a gastric architecture indistinguishable from that of uninfected age-matched mice (Fig.
1). Compared to H. pylori-infected INS-GAS mice that did not receive antimicrobial eradication
therapy, H. pylori antimicrobial eradication therapy at 8 WPI inhibited corpus dysplasia,
inflammation, atrophy, hyperplasia, epithelial defects (all p<0.001), and intestinal metaplasia
(p<0.05) (Fig. 2). Additionally, these mice had milder corpus inflammation, atrophy, and
epithelial defects when compared to uninfected mice (p<0.05, <0.01 and <0.01, respectively)
(Fig. 3).
H. pylori-infected male INS-GAS mice that received H. pylori antimicrobial eradication therapy
at 12 WPI and were euthanized 16 weeks later had corpus hyperplasia and thickened mucosal
folds. Microscopically, these mice had distorted mucosal architecture with dilated glands and
glandular dysplasia (Fig. 1). Compared to untreated H. pylori-infected INS-GAS mice, infected
male INS-GAS mice that received H. pylori antimicrobial eradication therapy at 12 WPI had
statistically less severe dysplasia, inflammation, and intestinal metaplasia (p<0.01, <0.05, and
<0.05, respectively) but similar degrees of atrophy, hyperplasia, and epithelial defects (Fig. 2).
Compared to uninfected mice, infected male INS-GAS mice that received H. pylori antimicrobial
eradication therapy at 12 WPI had more sever dysplasia and inflammation (p<0.001) but
154
comparable atrophy, hyperplasia, epithelial defects, and intestinal metaplasia.
H. pylori-infected INS-GAS mice that received H. pylori antimicrobial eradication at 22 WPI
and were euthanized 8 weeks later had thickened mucosal folds and developed corpus
hyperplasia and dysplasia (Fig. 1). Microscopically, these mice developed statistically less severe
dysplasia, inflammation, atrophy, and epithelial defects compared to untreated H. pylori-infected
INS-GAS mice (p<0.01 or less) (Fig. 2). Compared to uninfected mice, infected INS-GAS mice
that received H. pylori antimicrobial eradication therapy at 22 WPI had more severe dysplasia,
hyperplasia, and epithelial defects (p<0.05 and <0.01).
Among the infected INS-GAS mice that received H. pylori antimicrobial eradication therapy, the
mice receiving antimicrobial therapy at 8 WPI had significantly lower scores of dysplasia,
inflammation, atrophy, and hyperplasia compared to those mice receiving antimicrobial therapy
at 12 or 22 WPI (p<0.05 or less), and less severe epithelial defects compared to the mice
receiving antimicrobial therapy at 12 WPI (p<0.001). Most gastric lesions in infected mice that
received eradication therapy at 12 and 22 WPI were comparable, except for epithelial defects that
were more severe in the 12 WPI group (p<0.01).
H. pylori eradication at 8 WPI prevented progression to low-grade and high-grade
gastrointestinal intraepithelial neoplasia
Uninfected INS-GAS mice develop spontaneous gastric cancer at 20 months of age [20]; none of
the uninfected INS-GAS mice between 34 to 36 weeks old developed gastrointestinal
intraepithelial neoplasia (GIN). In contrast, all untreated H. pylori-infected INS-GAS mice
155
(n=10) at 28 WPI developed gastric cancer; two with low-grade GIN (20%) and eight with highgrade GIN (80%) (Fig. 3). In the infected mice that received eradication therapy at 8 WPI
(n=11), ten of them did not have GIN (91%) and one had low-grade GIN (9%). In the infected
mice that received eradication therapy at 12 WPI (n=9), one did not have GIN (11%), seven had
low-grade and one had high-grade GIN (78 and 11%, respectively). Six of the 12 infected mice
that received eradication therapy at 22 WPI did not have GIN (50%), and the remainder had lowgrade GIN (50%).
Compared to uninfected mice, infected mice that received eradication therapy at 8 WPI had a
similar incidence of GIN (p=0.38). In contrast, infected mice that received eradication therapy at
12 or 22 WPI and those untreated H. pylori-infected mice with had a higher incidence of lowand high-grade GIN (p<0.05 or less). Compared to H. pylori-infected INS-GAS mice that did not
receive eradication therapy, the incidences of low-grade and high-grade GIN were statistically
lower in the infected mice receiving eradication therapy at 8, 12, or 22 WPI (p<0.05 or less).
Among the infected mice receiving eradication therapy, eradication therapy at 8 WPI resulted in
significantly lower incidences of GIN compared to eradication therapy at 12 or 22 WPI (p<0.05).
The incidences of low-grade and high-grade GIN were statistically similar between infected mice
receiving eradication therapy at 12 and 22 WPI (p=O.12).
H. pylori-specific antibody responses
The H. pylori-specific, Thl-associated IgG2a levels were significantly induced in all infected
mice that received eradication therapy and untreated H. pylori-infected mice compared to
uninfected mice (p<0.01) (Fig. ). H. pylori eradication therapy at all time points had no effect on
156
H. pylori-specific IgG2a levels compared to untreated H. pylori-infected mice. However,
infected mice that received eradication therapy at 22 WPI had statistically higher H. pylorispecific IgG2a levels compared to those receiving eradication therapy at 8 or 12 WPI (p<0.05
and <0.01). H. pylori-specific, Th2-associated IgG1 responses were induced in all infected mice
that received eradication therapy and untreated H. pylori-infected mice compared to uninfected
mice (p<0.05). Compared to untreated H. pylori-infected mice, H. pylori-specific IgG1 levels
were reduced in mice that received eradication therapy at 8 WPI (p<0.01) but not affected by
eradication therapy at 12 or 22 WPI (p=0.11 or higher). H. pylori-specific IgG1 levels in mice
that received eradication therapy at 8 WPI were also lower than those in mice that received
eradication therapy at 12 or 22 WPI (p<0.01 and <0.05)
Gastric IFN-y, TNF-a, and iNOS mRNA levels were reduced in all mice that received H.
pylori eradication therapy
Given the importance of the inflammatory response in the pathogenesis of H. pylori gastritis, we
analyzed selected pro-inflammatory cytokines and iNOS mRNA levels in the gastric tissue at 28
WPI. Compared to uninfected mice, IFN-y mRNA levels were down-regulated in H. pyloriinfected mice receiving antimicrobial therapy at 8 WPI (p<0.05), but significantly up-regulated
in the infected mice that received H. pylori antimicrobial therapy at 12 (p<0.001) or 22 WPI
(p<0.05) as well as untreated H. pylori-infected INS-GAS mice (p<0.001) (Fig. 5). Compared to
untreated H. pylori-infected mice, H. pylori antimicrobial eradication significantly reduced
mRNA levels of IFN-y in infected mice receiving eradication therapy at 8, 12 (p<0.001), and 22
WPI (p<0.05). Among the infected mice receiving eradication therapy, those receiving
eradication therapy at 8 WPI had lower IFN-y mRNA levels relative to those receiving
157
eradication therapy at 12 or 22 WPI (p<0.01).
Compared to uninfected mice, gastric TNF-a mRNA levels were down-regulated in H. pyloriinfected mice that received H. pylori eradication therapy at 8 WPI, but up-regulated in infected
mice that received H. pylori eradication therapy at 12 (p<0.01) and untreated H. pylori-infected
mice (p<0.001). Compared to untreated H. pylori-infected mice, gastric TNF-a mRNA levels
were significantly reduced in all mice that received eradication therapy (p<0.001). Among the
infected mice receiving eradication therapy, infected mice that received eradication therapy at 8
WPI had significantly lower TNF-a mRNA levels compared to those mice that received
eradication therapy at 12 or 22 WPI (p<0.001 and <0.05). Infected mice that received eradication
therapy at 12 WPI had higher TNF-a mRNA levels compared to the mice that received
eradication therapy at 12 WPI (p<0.05).
Compared to uninfected mice, gastric iNOS mRNA levels were down-regulated in infected mice
that received eradication therapy at 8 WPI (p<0.01), but up-regulated in H. pylori-infected mice
that received eradication therapy at 12 and 22 WPI and untreated H. pylori-infected INS-GAS
mice (p<0.05 or less). Compared to untreated H. pylori-infected mice, gastric iNOS mRNA
levels were significantly down-regulated by H. pylori eradication therapy at all time points
(p<0.05 or less). Among the infected mice that received H. pylori eradication therapy, those
receiving eradication therapy at 8 WPI had lower iNOS mRNA levels compared to the mice
receiving eradication therapy at 12 and 22 WPI (p<0.001).
Gastric mucosal cell proliferation was also reduced in all mice that received H. pylori
158
eradication therapy
Epithelial proliferating cells detected by Ki67 immunohistological staining were mainly in the
isthmus regions of corpus mucosa in H. pylori-uninfected INS-GAS mice and infected mice that
received eradication therapy at 12 WPI (Fig. 5). Proliferating cells in the corpus expanded from
isthmus regions to hypertrophic foveolar regions in INS-GAS mice infected by H. pylori.
Compared to uninfected mice, corpus epithelial proliferation measured by Ki67 labeling indices
(LI) were comparable in infected mice that received H. pylori eradication at 8 WPI (p=0.093),
but higher in infected mice that received H. pylori eradication at 12 or 22 WPI, and untreated H.
pylori-infected mice (p<0.05 or lower). H. pylori eradication at 8, 12 and 22 WPI significantly
reduced Ki67 LI in the corpus compared to untreated H. pylori-infected mice (p<0.01 or lower).
Among the infected mice receiving H. pylori eradication therapy, mice receiving eradication
therapy at 8 WPI had lower corpus epithelial LI than those receiving eradication therapy at 12 or
22 WPI (p<0.05 and =0.058, respectively)
DNA adducts in gastric tissues
Levels of dI and etheno-dA in uninfected wt mice, INS-GAS, H. pylori-infected INS-GAS mice
receiving eradication at 12 WPI, and untreated H. pylori-infected INS-GAS mice were
comparable (Fig. 6). Levels of 8-oxodG were higher in uninfected INS-GAS mice than H.
pylori-infected INS-GAS mice receiving eradication at 12 WPI and untreated infected INS-GAS
mice (p<0.05).
H. pylori-infected INS-GAS mice had significantly higher serum levels of triglyceride, total
cholesterol, and elevated fraction of very-low-density lipoprotein
159
Triglyceride concentrations in untreated H. pylori-infected INS-GAS mice were significantly
higher than uninfected mice or H. pylori-infected mice that received eradication therapy at 12
WPI (p<0.001) (Fig. 7). There were no differences in triglyceride concentrations between
uninfected mice and H. pylori-infected mice that received eradication therapy at 12 WPI
(p=0.39). Total cholesterol concentrations in untreated in infected mice that received eradication
therapy at 12 WPI were significantly low than in uninfected or H. pylori-infected INS-GAS mice
(p<0.05). Uninfected or H. pylori-infected INS-GAS mice has similar concentrations of total
cholesterol levels (p=0.20). H. pylori infection increased the proportion of very-low-density
lipoprotein (VLDL) cholesterol in serum compared to uninfected or infected INS-GAS mice that
received eradication therapy at 12 WPI.
160
DISCUSSION
In this study, we used the well-described INS-GAS mouse model to study the effect of H. pylori
eradication at different stages of H. pylori-associated gastric pathology. All INS-GAS mice
infected with H. pylori for 28 WPI developed gastric cancer accompanied by inflammation, loss
of parietal and chief cells, and hypertrophy of foveolar glands. When eradication therapy was
instituted at an early point of infection, these histopathologic changes were reversible and
completely prevented progression to gastric cancer. Infected mice receiving H. pylori eradication
therapy at 8 WPI had a gastric cancer risk comparable to that of uninfected mice. Additionally,
reduced gastric inflammation in these mice was accompanied by lower proinflammatory
cytokine mRNA levels and decreased epithelial cell proliferation, which may contribute to lower
dysplasia and reduced gastric cancer risk. Eradication therapy at 12 and 22 WPI also resulted in a
statistically lower degree of gastric dysplasia, inflammation, and prevented progression to highgrade GIN when compared to the data recorded in untreated H. pylori-infected mice. The
attenuated inflammation was accompanied by reduced cytokine mRNA levels and epithelial cell
proliferation. Eradication therapy at later time points of infection did not reverse some
histopathologic changes such as atrophy, hyperplasia, and dysplasia. Thus, eradication at these
time points did not completely prevent progression to gastric cancer compared to uninfected
mice or infected mice that received treatment at 8 WPI.
H. felis-infected C57BIJ6 mice developed invasive carcinoma at 15 MPI [33]. Helicobacter
eradication therapy in H. felis-infected C57BIJ6 mice at 2 or 6 MPI prevented development of
gastric cancer; eradication therapy at 12 MPI reduced the risk of gastric cancer [17]. Consistent
with the H. felis C57BL/6 model, our data demonstrate that H. pylori eradication therapy at 8
161
WPI in INS-GAS mice restores mucosal architecture and prevented development of GIN. Also
eradication therapy at 12 or 22 WPI prevents progression to high-grade GIN compared to
untreated H. pylori-infected INS-GAS mice. However, the incidence of low-grade and highgrade GIN was statistically similar between H. pylori-infected mice that received eradication
therapy at 12 and 22 WPI. These results suggest that H. pylori infection in INS-GAS mice at 12
WPI results in certain irreversible changes in the gastric mucosa and that H. pylori eradication
may have little effect on preventing the progression to gastric cancer after such a "point of no
return". Mature parietal cells are necessary to maintain the integrity of gastric mucosa. Parietal
cell loss results in the dysregulation of gastric stem cell homeostasis and migration-associated
differentiation of pit and zymogenic cells [34]. The hypergastrinemic INS-GAS mice have
increased parietal cell mass before 3 months of age, then start developing gastric atrophy,
metaplasia, dysplasia, and finally cancer [20]. Independent studies from our laboratory
confirmed parietal cell loss in uninfected INS-GAS mice after 5 months of age (approximately
equivalent to 12-14 weeks post H. pylori infection) [20]. Therefore, H.pylori eradication therapy
given at 8 WPI in infected INS-GAS mice might reverse H. pylori-associated parietal cell loss
concomitant with the restoration of normal epithelial architecture. In contrast, eradication
therapy given at 12 WPI or later might fail to restore parietal cell mass and subsequent epithelial
differentiation.
H. pylori eradication in INS-GAS mice at all time points was associated with attenuated gastric
inflammation manifested as lower degrees of inflammatory scores and down-regulated IFN-y,
TNF-a, and iNOS mRNA levels compared to untreated H. pylori-infected mice. Gastric
inflammatory scores and IFN-y mRNA levels in INS-GAS mice correlated with epithelial
162
proliferation that appears to be mediated by up-regulated inflammatory cytokines such as IFNy [35]. Increased cell proliferation is a biomarker of gastric cancer risk. Reversal to a normal
epithelial proliferation has been associated with a reduce cancer risk [36, 37]. H. pylori
eradication therapy at 8 WPI returned epithelial proliferation rates to the levels of uninfected
mice. Eradication therapy at 12 or 22 WPI significantly reduced epithelial proliferation compared
to untreated H. pylori-infected mice but did not completely restore proliferation rates to the
levels of uninfected mice, suggesting that a higher gastric cancer risk existed. This notion is
supported by the elevated incidence of low-grade and high-grade GIN despite a successful H.
pylori eradication at 12 or 22 WPI. Another possibility is that with accumulated genetic damage
in mice as a result of long-standing H. pylori infection, H. pylori infection may no longer be
necessary for sustained dysplasia and progression to gastric cancer.
It has been proposed in human clinical trials that the regression of premalignant lesions in H.
pylori eradication may follow a sigmoid curve, where statistically significant results may not be
observed during the first a few years after treatment [9, 38]. The protective effect of H. pylori
eradication at 12 or 22 WPI in INS-GAS mice may not be seen at necropsy at 28 WPI.
Intervention at late stages of H. pylori infection may require a longer period post treatment to
observe benefit.
Unexpectedly, H. pylori-infected mice receiving eradication therapy at 8 WPI had significantly
lower degrees of gastric inflammation accompanied by less severe atrophy and down-regulated
IFN-y, TNF-a, and iNOS mRNA levels relative to those levels recorded in uninfected mice. The
mechanism by which antimicrobial eradication therapy exerted a protective effect in the H.
163
pylori-infected INS-GAS mice at 8 WPI is unknown. However, there is a complex ecosystem of
bacterial microbiota in mouse and human stomachs, including Proteobacteria,Firmicutes,
Actinobacteria, Bacteroidetes, and Fusobacteriaphyla, which may be associated with gastric
inflammation particularly when they overgrow in elevated gastric pH [39, 40]. Experimental
infection with Acinetobacter lwoffli in C57BU6 mice causes chronic gastritis as does H. pylori
[41]. Moreover, studies in humans and in various animal models have shown that bacteria are
important in triggering inflammation and epithelial damage in the stomach [42], suggesting the
potential interaction between H. pylori and other bacteria in the induction of gastric disease. A
triple antimicrobial therapy regimen with omeprazole, metronidazole and clarithromycin in H.
pylori-infected dyspeptic human patients resulted in H. pylori eradication and a dynamic change
of gastric microbiota; certain bacteria were eradicated (ex. Fusobacterium spp.) while some
bacteria overgrew (ex. Haemophilus spp. and Neisseria spp) [43]. Thus, it is conceivable that
part of the gastric phenotype in INS-GAS mice receiving antimicrobial therapy at 8 WPI could
be attributed to eradication of H. pylori and other non-helicobacter bacteria from the stomach
particularly in the hypochlorhydric state [40]. Our laboratory observed that among the 6 bacterial
species of altered Schaedler flora in H. pylori-infected INS-GAS mice, antimicrobial agents
eradicated gastric Mucispirillum schaedleri (unpublished data) [40]. These findings in humans
and mice suggest that antibiotic treatment potentially changes gastric microbiota, and this may
impact gastric carcinogenesis. Further experimentation is necessary to investigate the role of
microbiota on H.pylpri-associated gastric disease.
H. pylori infection in INS-GAS mice induced both IgG2a and IgG1 responses [21]. Compared to
untreated H. pylori-infected mice, antibiotic eradication therapy did not significantly reduce H.
164
pylori-specific, Thl-associated IgG2a levels. However, there was a trend toward lower levels of
H. pylori-specific, Th2-associated IgG1 in mice receiving eradication therapy. Our data suggest
that IgG subtypes may not be sensitive markers for H.pylori eradication.
Several previous observations demonstrated that gastric helicobacter infection resulted in chronic
inflammation accompanied with increased oxidative stress and oxidative DNA damage in vitro
and in vivo [44, 45]. Unexpectedly, no significant differences in the levels of dI and etheno-dA,
and 8-oxodG were observed among uninfected and H. pylori-infected INS-GAS mice with or
without antibiotic treatment. However, levels of 8-oxodG were higher in uninfected INS-GAS
mice than H. pylori-infected INS-GAS mice receiving eradication at 12 WPI and untreated
infected INS-GAS mice. These results are probably due to tissue specificity or detection limit.
Further efforts are necessary to examine DNA adducts in epithelial cells.
Although the association of H. pylori with hyperlipidemia has been speculated, no correlation
between H pylori seropositivity and hyperlipidemia has been observed [46]. However, elevated
concentrations of serum triglyceride and total cholesterol were observed in the untreated, H.
pylori-infected INS-GAS mice compared to uninfected mice and H. pylori-infected mice that
received eradication therapy at 12 WPI. Increased proportions of VLDL accompanied with serum
triglyceride and total cholesterol concentrations correlated with dysplasia and incidence of
gastric cancer in H. pylori-infected INS-GAS mice. One possible explanation for elevated VLDL
proportions in H. pylori-infected INS-GAS mice is that lipoprotein lipase activity is downregulated. Lipoprotein lipase deficiency results in hypertriglycemia and accumulation of VLDL
[47]. Elevated triglyceride levels were previously observed in aged APC1309 and APC""" mice,
165
animal model of human colon cancer [48, 49]. Down-regulated lipoprotein lipase mRNA levels
in the liver and small intestine were observed in both APC 3 09 and APC'"i" mice [48], suggesting
that reduced lipoprotein lipase activity may result in decreased clearance of VLDL from serum in
mice with colon cancer.
Many clinical trials in humans demonstrated that H. pylori eradication results in regression of
gastric atrophy and inflammation [5-10, 13]. H. pylori eradication has failed to reduce the
incidence of gastric cancer in most intervention studies in humans except in subpopulations that
did not develop premalignant lesions prior to antibiotic treatment [6-10, 13]. These results
suggest that the stage of H. pylori gastric disease is a determinant factor for the outcome of H.
pylori eradication therapy. Our results in H. pylori-infected INS-GAS mice support the premise
that H. pylori eradication therapy may be most beneficial in preventing gastric cancer when
antibiotics are given at the early stages of H. pylori infection. In the mice where metaplasia and
dysplasia existed, H. pylori eradication therapy substantially prevented progression to high-grade
GIN. In conclusion, our study using INS-GAS mice strongly suggests that to prevent the
development of gastric cancer H. pylori infection should be treated, especially in patients with
symptomatic gastritis or family history of gastric cancer. Those H. pylori-infected patients who
have developed premalignant lesions might benefit from H. pylori eradication to slow
progression to advanced disease.
ACKNOWLEDGEMENTS
We thank Kristen Clapp and Juri Miyamae for performing necropsies; Kathy Cormier for
providing histology expertise; Barry Rickman and Arlin B. Rogers for scoring gastric lesions; Bo
166
Pang and Peter C. Dedon for expert assistance with DNA adduct measurements; and Wen-Jun
Wang and David E. Cohen for providing expertise and instrument for serum lipid measurements.
167
REFERENCES
1.
2.
Marshall, B.J. and J.R. Warren, Unidentified curved bacilli in the stomach of patients
with gastritisandpeptic ulceration.Lancet, 1984. 1(8390): p. 1311-5.
IARC Working Group on the Evaluation of Carcinogenic Risk to Humans., in
Schistosomes liver flukes and Helicobacter pylori. 1994, International Association for
Research against Cancer: Lyon. p. 1-241.
3.
Fox, J.G and T.C. Wang, Helicobacterpylori--not a good bug after all! N Engl J Med,
2001. 345(11): p. 829-32.
4.
Correa, P., A human model of gastric carcinogenesis.Cancer Res, 1988. 48(13): p. 3554-
60.
5.
6.
Sung, J.J., et al., Atrophy and intestinal metaplasia one year after cure of H. pylori
infection: a prospective, randomizedstudy. Gastroenterology, 2000. 119(1): p. 7-14.
Ley, C., et al., Helicobacterpylori eradication and gastric preneoplastic conditions: a
randomized, double-blind,placebo-controlledtrial. Cancer Epidemiol Biomarkers Prev,
2004. 13(1): p. 4-10.
7.
You, W.C., et al., Randomized double-blindfactorial trial of three treatments to reduce
the prevalence ofprecancerousgastric lesions. J Natl Cancer Inst, 2006. 98(14): p. 974-
83.
8.
Ito, M., et al., Helicobacterpylori eradicationtherapy improves atrophic gastritis and
intestinal metaplasia: a 5-year prospective study of patients with atrophic gastritis.
Aliment Pharmacol Ther, 2002. 16(8): p. 1449-56.
9.
Correa, P., et al., Chemoprevention of gastric dysplasia: randomized trial of antioxidant
supplements and anti-helicobacterpylori therapy. J Natl Cancer Inst, 2000. 92(23): p.
10.
Mera, R., et al., Long term follow up of patients treatedfor Helicobacterpylori infection.
11.
Gut, 2005. 54(11): p. 1536-40.
Nyren, 0., et al., Cancer risk after hip replacement with metal implants: a population-
12.
based cohort study in Sweden. J Natl Cancer Inst, 1995. 87(1): p. 28-33.
Akre, K., et al., Risk for gastriccancerafter antibioticprophylaxis in patients undergoing
1881-8.
hip replacement. Cancer Res, 2000. 60(22): p. 6376-80.
13.
14.
Wong, B.C., et al., Helicobacterpylori eradicationto prevent gastric cancer in a highrisk region of China: a randomized controlled trial. Jama, 2004. 291(2): p. 187-94.
Leung, W.K., et al., Factors predicting progression of gastric intestinal metaplasia:
results of a randomised trial on Helicobacterpylori eradication. Gut, 2004. 53(9): p.
1244-9.
15.
Cao, X., et al., Eradication of Helicobacter pylori induces apoptosis and inhibits
proliferation of heterotopic proliferative glands in infected Mongolian gerbils. Cancer
Sci, 2004. 95(11): p. 872-7.
16.
17.
18.
Nozaki, K., et al., Effect of early eradication on Helicobacter pylori-related gastric
carcinogenesisin Mongolian gerbils. Cancer Sci, 2003. 94(3): p. 235-9.
Cai, X., et al., Helicobacterfelis eradication restores normal architecture and inhibits
gastric cancerprogressionin C57B16 mice. Gastroenterology, 2005. 128(7): p. 1937-52.
Fox, J.G, et al., Host and microbial constituents influence Helicobacterpylori-induced
cancer in a murine model of hypergastrinemia.Gastroenterology, 2003. 124(7): p. 1879-
90.
168
19.
Konda, Y., et al., Gastrin stimulates the growth of gastric pit with less-differentiated
features.Am J Physiol, 1999. 277(4 Pt 1): p. G773-84.
20.
21.
Wang, T.C., et al., Synergistic interaction between hypergastrinemia and Helicobacter
infection in a mouse model of gastriccancer: Gastroenterology, 2000. 118(1): p. 36-47.
Fox, J.G, et al., Helicobacterpylori-associatedgastriccancer in INS-GAS mice is gender
specific. Cancer Res, 2003. 63(5): p. 942-50.
22.
Wang, T.C., et al., Processing andproliferative effects of human progastrinin transgenic
mice. J Clin Invest, 1996. 98(8): p. 1918-29.
23.
24.
Lee, C.W., et al., Wildtype and IL10-/- T regulatory cells reduced T effector cell-mediated
gastroduodenitis in Rag2-/- mice but only wildtype T regulatory cells suppressed
Helicobacterpylori gastritis.Infect Immun, 2007.
Lee, A., et al., A standardizedmouse model of Helicobacterpylori infection: introducing
the Sydney strain. Gastroenterology, 1997. 112(4): p. 1386-97.
25.
van Zanten, S.J., et al., Gastric transitionalzones, areas where Helicobacter treatment
fails: results of a treatment trial using the Sydney strain mouse model. Antimicrob Agents
Chemother, 2003. 47(7): p. 2249-55.
26.
27.
Rogers, A.B., et al., Helicobacterpylori but not high salt induces gastric intraepithelial
neoplasia in B6129 mice. Cancer Res, 2005. 65(23): p. 10709-15.
Boivin, G.P., et al., Pathology of mouse models of intestinal cancer: consensus report and
recommendations. Gastroenterology, 2003. 124(3): p. 762-77.
28.
Maurer, K.J., et al., Helicobacterpylori and cholesterol gallstone formation in C57LIJ
mice: a prospective study. Am J Physiol Gastrointest Liver Physiol, 2006. 290(1): p.
G175-82.
29.
Ihrig, M., et al., Gastric helicobacter infection induces a Th2 phenotype but does not
elevate serum cholesterol in mice lacking inducible nitric oxide synthase. Infect Immun,
2005. 73(3): p. 1664-70.
30.
31.
32.
Pang, B., et al., Lipidperoxidation dominates the chemistry of DNA adduct formation in a
mouse model of inflammation. Carcinogenesis, 2007.
Wang, W.J., et al., Homozygous disruption of Pctp modulates atherosclerosis in
apolipoproteinE-deficient mice. J Lipid Res, 2006. 47(11): p. 2400-7.
Hyogo, H., et al., Leptin promotes biliary cholesterol elimination during weight loss in
ob/ob mice by regulating the enterohepatic circulation of bile salts. J Biol Chem, 2002.
277(37): p. 34117-24.
33.
Houghton, J., et al., Gastriccanceroriginatingfrom bone marrow-derivedcells. Science,
2004. 306(5701): p. 1568-71.
34.
35.
Li, Q., S.M. Karam, and J.I. Gordon, Diphtheriatoxin-mediated ablation of parietalcells
in the stomach of transgenicmice. J Biol Chem, 1996. 271(7): p. 3671-6.
Kang, W., et al., Interferon gamma induction of gastric mucous neck cell hypertrophy.
36.
Lab Invest, 2005. 85(5): p. 702-15.
Biasco, G, et al., Cell proliferation biomarkers in the gastrointestinal tract. J Cell
Biochem Suppl, 1992. 16G: p. 73-8.
37.
Biasco, G, et al., Cell proliferation kinetics as a marker of gastric cancer risk. Ital J
Gastroenterol, 1991. 23(1): p. 24-8.
38.
39.
Mera, R., et al., Re: Randomized double-blindfactorial trial of three treatments to reduce
the prevalence ofprecancerousgastric lesions. J Natl Cancer Inst, 2006. 98(19): p. 1426.
Bik, E.M., et al., Molecular analysis of the bacterial microbiota in the human stomach.
169
Proc Natl Acad Sci U S A, 2006. 103(3): p. 732-7.
40.
Sarma-Rupavtarm, R.B., et al., Spatial distribution and stability of the eight microbial
species of the altered schaedlerflora in the mouse gastrointestinaltract. Appl Environ
Microbiol, 2004. 70(5): p. 2791-800.
41.
Zavros, Y, et al., Gastritis and hypergastrinemia due to Acinetobacter Iwoffli in mice.
Infect Immun, 2002. 70(5): p. 2630-9.
42.
Fox, J.G and T.C. Wang, Inflammation, atrophy, and gastric cancer J Clin Invest, 2007.
43.
Adamsson, I., et al., Comparative effects of omeprazole, amoxycillin plus metronidazole
versus omeprazole, clarithromycinplus metronidazole on the oral, gastric and intestinal
microflora in Helicobacter pylori-infected patients. J Antimicrob Chemother, 1999.
44(5): p. 629-40.
Baik, S.C., et al., Increased oxidative DNA damage in Helicobacter pylori-infected
117(1): p. 60-9.
44.
human gastricmucosa. Cancer Res, 1996. 56(6): p. 1279-82.
45.
Arabski, M., et al., DNA damage and repair in Helicobacter pylori-infected gastric
mucosa cells. Mutat Res, 2005. 570(1): p. 129-35.
46.
47.
Danesh, J. and R. Peto, Risk factors for coronary heart disease and infection with
Helicobacterpylori:meta-analysisof 18 studies. Bmj, 1998. 316(7138): p. 1130-2.
Patsch, J., Influence of lipolysis on chylomicron clearance and HDL cholesterol levels.
Eur Heart J, 1998. 19 Suppl H: p. H2-6.
48.
Niho, N., et al., Concomitant suppression of hyperlipidemia and intestinal polyp
formation in Apc-deficient mice by peroxisome proliferator-activatedreceptor ligands.
Cancer Res, 2003. 63(18): p. 6090-5.
49.
Niho, N., et al., Dose-dependent suppression of hyperlipidemia and intestinal polyp
formation in Min mice by pioglitazone, a PPAR gamma ligand.Cancer Sci, 2003. 94(11):
p. 960-4.
170
Legends
Figure 1.
Histopathology of gastric disease induced by H. pylori infection in the corpus mucosa of INS-GAS
mice with and without eradication therapy. Representative H&E of stomachs from uninfected INSGAS mice (a), treated mice at 8, 12, and 22 weeks post-infection (b, c, d) and untreated infected mice
(e, f). (a) Uninfected mouse stomach showing typical INS-GAS background lesions of mild to
moderate dysplasia, mild inflammation, severe atrophy, and mild to moderate hyperplasia on H&E
stain. (b) Tissue from a treated INS-GAS mouse 8 weeks post-infection exhibiting mild to moderate
dysplasia, mild inflammation, moderate to severe atrophy, and mild to moderate hyperplasia on H&E
stain. (c) Tissue from a treated INS-GAS mouse 12 weeks post-infection exhibiting moderate
dysplasia, moderate inflammation, severe atrophy, and moderate hyperplasia on H&E stain. (d) Tissue
from a treated INS-GAS mouse 22 weeks post-infection exhibiting moderate dysplasia, mild to
moderate inflammation, severe atrophy, and moderate hyperplasia on H&E stain. (e) Tissue from an
untreated INS-GAS mouse 22 weeks post-infection exhibiting severe dysplasia, moderate
inflammation, severe atrophy, and moderate to severe hyperplasia on H&E stain. (f) Tissue from (e) at
higher magnification demonstrating apoptotic bodies, dilated glands and glandular dysplasia
characterized by loss of columnar orientation, elongation, branching, irregular cell shapes and sizes,
and stratification.
Figure. 2.
Histological scores of (a) dysplasia, (b) inflammation, (c) atrophy, (d) hyperplasia, (e) epithelial
defects, (f) intestinal metaplasia, and (g) incidences of non-gastrointestinal intraepithelial neoplasia
(GIN), low-grade GIN (a dysplasia score of 3.0), and high-grade GIN (dysplasia scores equal or
higher than 3.5). (Statistical significance is presented as *, p<0.05; **, p<0.01; ***, p<0.001 when
171
compared to uninfected mice; #, p<0.05; ##, p<0.01; ###, p<0.001 when compared to H. pyloriinfected mice which did not received antimicrobial therapy; §, p<0.05; §§, p<0.01; §§§, p<0.001;
comparison as indicated).
Figure 3.
Relative mRNA levels of IFN-y (a), TNF-a (b), and iNOS (c) in the gastric tissue. Data are presented
as fold change relative to uninfected INS-GAS mice. H. pylori-infected mice without eradication
therapy had significantly higher levels of IFN-y, TNF-a, and iNOS mRNA (p<0.001). H. pylori
eradication at 8 WPI had statistically lower IFN-y, TNF-a, and iNOS mRNA levels compared to
uninfected mice and infected mice without eradication therapy. H. pylori eradication at 12 or 22 WPI
had IFN-y, TNF-a, and iNOS mRNA levels which were significantly higher than uninfected mice, but
lower than infected control mice that did not receive eradication therapy. (Compared to uninfected
mice: *, p<0.05; **, p<0.01; ***, p<0 .0 0 1. Compared to infected mice which did not receive
antimicrobial eradication: #, p<0.05; ##, p<0.01; ###, p<0.001).
Figure 4.
H. pylori-specific, Thl-associated IgG2a levels were significantly elevated in infected mice at 28
WPI irrespective of antimicrobial eradication when compared to uninfected controls (**, p<0.01) (a).
H. pylori eradication did not significantly reduce serum levels of H. pylori-specific IgG2a.
Comparing to untreated H. pylori-infected mice, mice received H. pylori eradication at 22 WPI had
higher IgG2a responses than those received H. pylori eradication at 8 or 12 WPI (#, p<0.05; ##,
p<0.01). H. pylori-specific, Th2-associated IgG1 levels were elevated in infected mice that received
H. pylori eradication at 12 or 22 WPI and mice that actively infected with H. pylori (*, p<0.05; §,
172
0.05<p<0.10) compared to uninfected mice (b). Infected mice receiving eradication therapy at 8 WPI
had significantly lower H. pylori-specific IgG1 levels compared to infected mice given eradication
therapy at 12 or 22 WPI (#, p<0.05; ##, p<0.01) and infected mice which did not receive eradication
therapy (p<0.01).
Figure 5.
Immunohistochemical staining of Ki67 in the corpus. Proliferating cells are positive stained for Ki67.
In helicobacter-uninfected INS-GAS mice, proliferating cells were restricted to the isthmus regions
(a). Proliferating cells were restricted to the isthmus regions irrespective of the hyperplastic foveolar
glands in the infected mice that received H. pyloni eradication at 12 WPI (b). Proliferating cells
expanded from isthmus regions to foveolar regions in the infected INS-GAS mice that did not receive
eradication therapy (c). Ki67 labeling indices (LI) of Ki67 were significantly increased in H. pyloriinfected mice that received eradication therapy at 12 or 22 WPI, and infected control mice compared
to uninfected mice (d). H. pylori eradication at 8, 12, and 22 WPI significantly reduced corpus LI
compared to infected mice without eradication therapy. Additionally, infected mice receiving H.
pylori eradication at 8 WPI had lower corpus LI compared to the infected mice receiving eradication
therapy at 12 WPI. (Statistical significance is presented as *, p<0.05; **, p<0.01; ***, p<0.001 when
compared to uninfected mice; #, p<0.05; ##, p<0.01; ###, p<O.001 when compared to H. pyloriinfected mice which did not received antimicrobial therapy; §, p<0.05; comparison as indicated).
Figure 6.
DNA adducts of 2'-deoxyinosine (dI), N6-ethenodeoxyadenosine (etheno-dA), and 8-oxo-7,8dihydro-2'-deoxyguanosine (8-oxodG) in the stomach. White bars are wt mice; black bars are INS-
173
GAS mice. Levels of (a) dI and (b) etheno-dA were similar among all groups. (c) Levels of 8-oxodG
were significantly higher in uninfected INS-GAS mice than H. pylori infected mice that received
eradication therapy at 12 WPI and untreated infected INS-GAS mice. (Statistical significance is
presented as *, p<0.05).
Figure 7.
Triglyceride and total cholesterol concentrations and distribution of lipoprotein cholesterol in serum.
(a) Triglyceride concentrations in untreated H. pylori-infected INS-GAS mice were significantly
higher than uninfected mice and H. pylori-infected mice receiving eradication therapy at 12 WPI. (b)
Total cholesterol levels in H. pylori-infected mice receiving eradication therapy at 12 WPI were lower
than uninfected mice and untreated H. pyloi-infected INS-GAS mice. Proportion of VLDL
cholesterol in serum was increase in untreated H. pylori-infected INS-GAS mice (d) but not in
uninfected INS-GAS mice (c).
174
Fig. 1
a
175
Fig. 2
***
*
*
***
4
0
-a
*
***
4
e*
dan 2
0
cD
0
4
0
3
C
0
em.
E
E
##
#
56§
4
o
m 2
pr a
0
emm
§§
c
0m
SUMS
3
0
0
0
sem
"'*
ago
---..--
...
...
0
0
H. pylori
Eradication
(WPI)
-
+
+
+
+
-
8
12
22
-
*
)
***
###
###
##
###
4
H. pylon
Eradication
(WPI)
mm
0*
-
cation
+
+
+
***
#U
-
-
50
-
25
-
-
8
12
22
-
(WPI)
4-
E
3.**
,
-
+
+
+
+
-
8
12
22
-
#
U
**
n9H5
.!".
.0
.
m
"
-
-
"..m
...
""*
""*
-
+
8
+
12
+
22
+
-
H. pylori
Eradication
(WPI)
-
+
+
+
+
8
12
22
-
*
non-GIN
CM Low-grade GIN
M High-grade GIN
-a -
u
H. pylori
Eradication
(WPI)
H. pylon
Eradication
#
100 '
75
+
0
o 0
H. pylor
Eradication
(WPI)
+
22
12
8
+
06
aWP) 0
H. pylori-
+
**
0
4010
+
**
-
3 - eo
-
I
nE
-
+
+
-
8
12
HI
i_
+
+
22
-
176
Fig. 3
*
***
*
100-
**
*****
#
#
#
bz
***
r
§
H§
555
75
50
u
2
25-
0
L
0
0
H. pylori
Eradication
(WPI)
***
55
C 505
2
*
100-
75.
E
**
E-
-
-
+
+
8
12
**
**
+
+
22
25
p
0 'H. pylon
Eradication
(WPI)
+
+
+
12
8
22
+
H. pylori
-
+
+
+
-
Eradication
(WPI)
-
8
12
22
+
Fig. 4
§
*
**
§
-r
~~1
0 I-mm
H. pylori
Eradication
(WPI)
+
+
+
8
12
22
+I
0 1
H. pylori
Eradication
(WPI)
-
+
+
+
+
8
12
22
-
Fig. 5
b
**
*
***
60p=0.058
50.u 40-E 0-300
(~10
H. pylori
Eradication
(WPI)
+
+
0
H pylon
Eradication
(WPI)
.
+
+
+
+
8
12
22
-
177
Fig. 6
10
8
8
Ii
0.5
4 6 -~
2
.LL-~.
0
S2
0
H. pyWor
-
-
+
+
Eradication
(WPI)
-
-
12
-
H. pyloi
Eradication
0
-
+
.
.
H.pylon
Eradication
-
-
.
(WPI)
(WPI)
Fig. 7
*
II
500-
E 300-
400-
-* 3001
j!
8 200]
0
.C
200-
i
1u
00 1-
100-
0 .J
H. pylon
Eradication
(WPI)
C
0
-
+
+
0
Wfl
VLDL
LDL/HDL
HDL
CD30
5
20
10
Z
10
+
+.
Eradication
(WPI)
Snu12
%ifni lfI ifi.r
H. pylon
40
30
*
0
*1
e#
.'
*
'+
+4
*
+'
.'
q
p
p
p
Fractons
Table 1. Numbers of mice in each treatment group
H.pyloria
Eradication
(WPI)
Necropsy
(WPI)
Nb
-
-
28
naI8
+
+
+
8
12
22
28
28
28
11/11
9/9
12/14
+
-
28
na/10
Abbreviation: WPI, weeks post infection; na, not-applicable
aH. pylori inoculation at the beginning of experiment.
bNumbers of mice with successful H.pylori eradication / Group size.
Table 2. Gastric lesions at 22 and 28 weeks post H. pylori infection (WPI)
Weeksa H.pylori
22
-
N
D
9
1.5
22
+
10
28
-
8
28
+
10
Median (Range) of Lesion indices
H
I
A
1
2.5
2
( 1 .5- 2 )b"e
(0.5-2)e
( 1 .5- 3 )'d
( 1 .5- 3 )d
3
2.5
3
3
(2.5-3.5)c'e
(2-2.5)e
( 3 )c'd
2
1
3
( 1 .5-2. 5 )b.'
3.5
( 3 -3. 5)g
( 1 -1.
2
( 1 .5-2. 5 )'e
IM
3
( 2 - 3 )d
3
3.5
( 2 - 3 )d
(3)e
(3-3.5)d
2.25
2.75
3
(3 -3. 5 )b
( 1 .5- 3 )f
( 2 .5- 3 )b
(3-3.5)
.2.25
3.5
3
3
3.5
( 2 -2. 5 )g
(3.5)c
(3 -3. 5 )f
(3)
(3-3.5)
5
)g
ED
Abbreviations: D, dysplasia; I, inflammation; A, Atrophy; H, hyperplasia; ED, epithelial defects;
IM, intestinal metaplasia.
aEquivilant to WPI
bSignificant difference between 22 and 28 WPI in uninfected mice, p<0.05.
cSignificant difference between 22 and 28 WPI in infected mice, p<0.05.
dSignificant difference between uninfected and infected mice at
22 WPI, p<0.05.
*Significant difference between uninfected and infected mice at 22 WPI, p<0.001.
Significant difference between uninfected and infected mice at 28 WPI, p<0.01.
5Significant difference between uninfected and infected mice at 28 WPI, p<0.001.
179
180
Chapter 6: Summary
181
This thesis reported studies on pathogenesis of the carcinogenic bacterium Helicobacterpylori
and its inter-relationship with host and environmental factors contributing to gastric cancer.
Using mouse models of experimental H. pylori infection, these studies focused on 4 areas: (1)
IL10 requirement of CD4* regulatory T (TR) cells function in suppressing H. pylori-associated
gastritis using adoptive transfer model; (2) whether vitamin C supplementation influences H.
pylori gastritis; (3) the roles of DNA repair proteins Aag and Mgmt in H. pylori gastritis; and (4)
whether H.pylori eradication prevents progression of H.pylori gastritis to gastric cancer.
The results of adoptive transfer model of H. pylori gastritis demonstrate that: (1) H. pyloriinfected C57BIJ6 Rag2~'~ mice that do not have functional B and T cells do not develop gastritis;
(2) transfer of wildtype (wt) CD4* effector T (TE) cells into Rag2'~ mice results in TE-mediated
gastroduodenitis involving both stomach and proximal duodenum, while H. pylori infection
superimposes severity of corpus gastritis; and (3) cotransfer of wt TE plus wt TR cells suppresses
both TE-mediated and H. pylori-associated inflammation, while cotransfer of wt TE plus IL 10TR cells suppresses TE-mediated gastroduodenitis but not H. pylori-associated corpus gastritis,
suggesting that IL1O is required for TR cells function in suppressing H. pylon-associated
gastritis. These data are consistent with previous studies that TE cells are required to induce H.
pylori gastritis [1, 2], while TR cells suppress H. pylori gastritis [3]. ILlO is critical for TR cell
function in suppressing H. pylori-associated corpus gastritis [4], consistent with the ILlO
requirement for TR cell suppression of inflammatory bowel disease [5]. Importantly, our data
supports the findings in humans that polymorphisms of IL1O promoter are associated with a
higher risk of gastric cancer [6].
182
Chronic H. pylori infection results in increased oxidative stress and oxidative DNA damage in
human gastric mucosa [7, 8]. Since vitamin C is a strong antioxidant, it has been speculated
whether vitamin C supplementation reduces H. pylori-induced oxidative damage and severity of
gastritis. Consistent with human clinical trials of vitamin C [9-12], a vitamin C supplementation
10-fold higher than maintenance dose in gulo~'~ mice does not reduce H. pylori gastritis. In
contrast, a low vitamin C supplementation in gulo~' mice results in less severe H. pylori gastritis
due to reduced Th1 immune responses, consistent with impaired mitogen responses of peripheral
blood mononuclear cells from pigs with heritable vitamin C deficiency [13]. Our data also
supports the "African enigma" where the incidence of gastric cancer is low in some African
countries despite a high prevalence of H. pylori infection [14]. In Gambia, people have very low
dietary vitamin C and plasma vitamin C levels during raining season for 7 months/year [15]. The
low vitamin C state may reduce Th1 responses to H. pylori and cause less severe H. pylori
gastric lesions.
The molecular mechanism of H. pylori-induced gastric cancer remains unclear. Dysfunction and
polymorphisms of DNA repair genes have been associated with a higher risk of gastric cancer
[16, 17]. The roles of DNA repair proteins 3-alkyladenine DNA glycosylase (Aag) or 06methylguanine DNA methyltransferase (Mgmt) in H. pylori gastritis were examined using Aag
or Mgmt knockout mice on a C57BIJ6 background. Both infected knockout mice developed
more preneoplastic gastric atrophy, hyperplasia, and mucous metaplasia compared to infected wt
mice, suggesting Aag and Mgmt may reduce the development of H. pylori-associated gastric
cancer by preventing development of preneoplastic lesions.
183
H. pylori causes life-ling infection and chronic gastritis. In a susceptible host, H. pylori chronic
gastritis may progress to gastric cancer [18]. The natural course of H. pylori infection raises the
possibility that eradication of H. pylori may prevent the development of gastric cancer. The key
question is that at what stage of gastric disease would H. pylori eradication therapy prevent
disease progression to caner. Eradication of H. pylori in humans prevents development of gastric
cancer in the patients without precancerous lesions [19], suggesting that disease stage when
treatment is given may be important. Results of H. pylori eradication therapy in hypergastrinemic
INS-GAS mice demonstrate that untreated H. pylori-infected mice develop gastric cancer at 28
WPI. H. pylori eradication at all time points (8, 12, and 22 WPI) reduces gastric inflammation,
accompanied by down-regulated proinflammatory cytokine mRNA and lower cell proliferation
which is associated with decrease dysplasia. Optimal effect on cancer prevention was observed
when antibiotics are given at 8 WPI. These results support the Maastricht consensus III that H.
pylori infection should be treated early, especially in patients with gastric ulcer or a family
history of gastric cancer [20].
H. pylori infection is a serious problem for public health; more than 50% of population are
infected worldwide [18]. H. pylori-associated gastric cancer is the
4th
most common cancer and
causes 2 "dcancer-related mortality [21]. Although evidences support early H. pylori eradication
reduces gastric cancer risk [19], it is impractical to treat all infected individuals. Preventive and
therapeutic H. pylori vaccination looks promising [22], but large pharmaceutical firms to date
have not been interested in pursuing this strategy. Data reported in diet studies and experiments
conducted with parasitic infections demonstrate immune modulation from Thi to Th2 or overall
lower Th1 responses may reduce severity of H. pylori gastritis [23]. Chemoprevention trials to
184
date, however, have been disappointing in using this strategy to prevent gastric cancer. Our
vitamin C study in gulo' mice seems to support these results. It is well recognized those
individuals with family history of gastric cancer should be treated for H. pylori infection [20].
Due to the genetic association, it is recommended that efforts be launched to develop rapid
screening for polymorphisms in key cytokine genes that are associated with a higher risk of
gastric cancer [6, 18]. Individuals with polymorphisms associated with increased risk for cancer
should be identified and treated for H. pylori.
REFERENCES
1.
2.
Eaton, K.A., S.R. Ringler, and S.J. Danon, Murine splenocytes induce severe gastritis
and delayed-type hypersensitivity and suppress bacterial colonization in Helicobacter
pylori-infected SCID mice. Infect Immun, 1999. 67(9): p. 4594-602.
Eaton, K.A., M. Mefford, and T. Thevenot, The role of T cell subsets and cytokines in the
pathogenesis of Helicobacterpylorigastritisin mice. J Immunol, 2001. 166(12): p. 7456-
61.
3.
4.
5.
6.
Rad, R., et al., CD25+/Foxp3+ T cells regulate gastric inflammation and Helicobacter
pylor colonization in vivo. Gastroenterology, 2006. 131(2): p. 525-37.
Lee, C.W., et al., Wild-Type and Interleukin-10-Deficient Regulatory T Cells Reduce
Effector T-Cell-Mediated Gastroduodenitis in Rag2-/- Mice, but Only Wild-Type
Regulatory T Cells Suppress HelicobacterpyloriGastritis.Infect Immun, 2007. 75(6): p.
2699-707.
Erdman, S.E., et al., CD4(+)CD25(+) regulatory lymphocytes require interleukin 10 to
interrupt colon carcinogenesisin mice. Cancer Res, 2003. 63(18): p. 6042-50.
El-Omar, E.M., et al., Increased risk of noncardia gastric cancer associated with
proinflammatorycytokine gene polymorphisms. Gastroenterology, 2003. 124(5): p. 1193-
201.
7.
Baik, S.C., et al., Increased oxidative DNA damage in Helicobacter pylori-infected
human gastricmucosa. Cancer Res, 1996. 56(6): p. 1279-82.
8.
Papa, A., et al., Role of Helicobacterpylori CagA+ infection in determining oxidative
DNA damage in gastricmucosa. Scand J Gastroenterol, 2002. 37(4): p. 409-13.
9.
Mera, R., et al., Long term follow up of patients treatedfor Helicobacterpylori infection.
Gut, 2005. 54(11): p. 1536-40.
10.
Waring, A.J., et al., Ascorbic acid and total vitamin C concentrations in plasma, gastric
juice, and gastrointestinalmucosa: effects of gastritis and oral supplementation. Gut,
1996. 38(2): p. 171-6.
185
11.
Phull, P.S., et al., Gastroduodenal mucosal vitamin-C levels in Helicobacter pylori
infection. Scand J Gastroenterol, 1999. 34(4): p. 361-6.
12.
13.
Bjelakovic, G., et al., Antioxidant supplementsfor prevention of gastrointestinalcancers:
a systematic review and meta-analysis.Lancet, 2004. 364(9441): p. 1219-28.
Schwager, J. and J. Schulze, Modulation of interleukin production by ascorbic acid. Vet
Immunol Immunopathol, 1998. 64(1): p. 45-57.
14.
15.
16.
Holcombe, C., Helicobacterpylori: the African enigma. Gut, 1992. 33(4): p. 429-3 1.
Bates, C.J., et al., The effect of vitamin C supplementation on lactating women in Keneba,
a West African ruralcommunity. Int J Vitam Nutr Res, 1983. 53(1): p. 68-76.
Huang, W.Y, et al., Selected DNA repairpolymorphisms and gastric cancer in Poland.
17.
Oue, N., et al., Promoterhypermethylation of MGMT is associated with protein loss in
Carcinogenesis, 2005. 26(8): p. 1354-9.
gastric carcinoma.Int J Cancer, 2001. 93(6): p. 805-9.
18.
Fox, J.G. and T.C. Wang, Inflammation, atrophy, and gastric cancer J Clin Invest, 2007.
19.
Wong, B.C., et al., Helicobacterpylori eradication to prevent gastric cancer in a highrisk region of China:a randomizedcontrolledtrial.Jama, 2004. 291(2): p. 187-94.
Malfertheiner, P., et al., Current concepts in the management of Helicobacter pylori
infection: the MaastrichtIII Consensus Report. Gut, 2007. 56(6): p. 772-781.
117(1): p. 60-9.
20.
21.
Parkin, D., S. Whelan, and J. Ferlay, Cancer incidence in five continents. Vol. VII. 1997,
Lyon: International Agency for Research on Cancer.
22.
Kabir, S., The current status of Helicobacter pylori vaccines: a review. Helicobacter,
2007. 12(2): p. 89-102.
23.
Fox, J.G., et al., Concurrent enteric helminth infection modulates inflammation and
gastric immune responses and reduces helicobacter-inducedgastric atrophy. Nat Med,
2000. 6(5): p. 536-42.
186
187