See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/7970510

The Story So Far: Helicobacter Pylori and Gastric Autoimmunity

Article  in  International Reviews Of Immunology · August 2009

DOI: 10.1080/08830180590884648 · Source: PubMed

CITATIONS READS

60 294

8 authors, including:

M.P. Bergman Christina M J E Vandenbroucke-Grauls

Vrije Universiteit Amsterdam Amsterdam University Medical Center
32 PUBLICATIONS   1,441 CITATIONS    491 PUBLICATIONS   25,182 CITATIONS   

SEE PROFILE SEE PROFILE

Ben J Appelmelk Mario Milco D’Elios

Amsterdam University Medical Center University of Florence
155 PUBLICATIONS   9,125 CITATIONS    235 PUBLICATIONS   10,151 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

"The Halifax Project" led by Getting to Know Cancer (NGO) View project

IS-pro validation and implementation in clinical diagnostics View project

All content following this page was uploaded by M.P. Bergman on 30 November 2015.

The user has requested enhancement of the downloaded file.

 International Reviews of Immunology

ISSN: 0883-0185 (Print) 1563-5244 (Online) Journal homepage: http://www.tandfonline.com/loi/iiri20

The Story So Far: Helicobacter Pylori and Gastric

Autoimmunity

Mathijs P. Bergman, Christina M. J. E. Vandenbroucke-Grauls, Ben J.

Appelmelk, Mario M. D'Elios, Amedeo Amedei, Annalisa Azzurri, Marisa
Benagiano & Gianfranco Del Prete

To cite this article: Mathijs P. Bergman, Christina M. J. E. Vandenbroucke-Grauls, Ben

J. Appelmelk, Mario M. D'Elios, Amedeo Amedei, Annalisa Azzurri, Marisa Benagiano &
Gianfranco Del Prete (2005) The Story So Far: Helicobacter Pylori and Gastric Autoimmunity,
International Reviews of Immunology, 24:1-2, 63-91

To link to this article: http://dx.doi.org/10.1080/08830180590884648

Published online: 03 Aug 2009.

Submit your article to this journal

Article views: 78

View related articles

Citing articles: 1 View citing articles

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=iiri20

Download by: [Hong Kong Polytechnic University] Date: 30 October 2015, At: 13:24
 International Reviews of Immunology, 24: 63
91, 2005
Copyright # Taylor & Francis Inc.
ISSN: 0883-0185 print/1563-5244 online
DOI: 10.1080/08830180590884648

THE STORY SO FAR: HELICOBACTER PYLORI AND

GASTRIC AUTOIMMUNITY

MATHIJS P. BERGMAN, CHRISTINA M.J.E.

Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

VANDENBROUCKE-GRAULS, and BEN J. APPELMELK

Department of Medical Microbiology and Infection Control,
VU Medical Center, Amsterdam, The Netherlands

MARIO M. D’ELIOS, AMEDEO AMEDEI, ANNALISA AZZURRI,

MARISA BENAGIANO, and GIANFRANCO DEL PRETE
Department of Internal Medicine and Immunoallergology,
University of Florence, Florence, Italy

The gastric mucosal pathogen Helicobacter pylori induces autoantibodies directed

against the gastric proton pump Hþ,Kþ-ATPase in 20
30% of infected patients. The
presence of these autoantibodies is associated with severity of gastritis, increased atro-
phy, and apoptosis in the corpus mucosa, and patients with these autoantibodies
infected with H. pylori display histopathological and clinical features that are similar
to those of autoimmune gastritis (AIG). This review will focus on the T helper cell
responses, cytokines, and adhesion molecules involved in corpus mucosal atrophy in
chronic H. pylori gastritis and in AIG, and the role of H. pylori in the onset of AIG.

Keywords: Helicobacter pylori, antiparietal cell autoantibodies, Hþ,Kþ-ATPase,

autoimmune gastritis, Th1 cells, cytokines, Fas, apoptosis

INTRODUCTION
Helicobacter pylori is a gram-negative human gastric pathogen that
causes persistent infection in half of the world’s population. H. pylori
infection leads to chronic gastric inflammation and peptic ulcer dis-
ease, and it is associated with mucosa-associated lymphoid tissue
(MALT) lymphoma and with loss of gastric glands in the antrum
and corpus, which is referred to as atrophy. Atrophy is considered

Address correspondence to Mathijs P. Bergman, Department of Medical Micro-

biology, Erasmus MC, P.O. Box 1738, 3000 DR Rotterdam, The Netherlands.
E-mail: m.bergman@erasmusmc.nl

63
 64 M. P. Bergman et al.

a precursor of gastric adenocarcinoma, the second most frequent cause

of cancer-related death, and there is strong evidence that H. pylori
infection increases the risk of gastric cancer [1]. In a recent prospec-
tive study of 1526 Japanese subjects, gastric cancer developed in
2.9% of 1246 H. pylori
infected patients over 7.8 years, whereas gas-
tric cancer was not observed in 280 noninfected control subjects, nor in
a subgroup of 253 individuals that received H. pylori eradication ther-
apy early during follow-up [2].
The trigger of atrophy and mechanisms involved in its development
Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

are still poorly understood. Although H. pylori is not invasive and

usually resides in the antrum, glands located deep in the mucosa of
the antrum and corpus disappear. Experimentally infected mice
develop atrophy in the corpus, while the colonization takes place
mainly in the antrum [3]. Accumulating data on humans suggest that
gastric corpus atrophy is caused by an H. pylori
driven autoimmune
process.
In this review, we describe the difference between antigastric anti-
body responses upon H. pylori infection in animals and in humans,
and discuss the histomorphological and functional features of the
gastric mucosa that are associated with the presence of gastric auto-
antibodies in H. pylori
associated autoimmunity. Secondly, the
pathogenesis of H. pylori
associated atrophic gastritis is reviewed
and compared with the pathogenesis of classical autoimmune gastritis
and pernicious anemia (AIG=PA), which is discussed by D’Elios and
colleagues elsewhere in this issue. Based on the striking similarities
between AIG=PA and H. pylori
gastritis accompanied by corpus atro-
phy and autoantibodies, it seems plausible to consider an initiating
role of H. pylori in gastric autoimmunity. In the third part of this
review, models that may explain the role of H. pylori in gastric auto-
immune responses are discussed.

H. PYLORI INFECTION INDUCES AUTOANTIBODIES TO

GASTRIC MUCOSAL ANTIGENS
H. pylori induces autoantibodies reactive with gastric mucosal anti-
gens in approximately half of the infected individuals (range
49
64%) [4
8]. Immunohistochemistry on healthy, noninfected
human gastric tissue incubated with serum from H. pylori
infected
patients, reveals two different binding sites for such autoantibodies
[6,7]: one on the luminal membranes of the foveolar epithelial cells
in the antrum and corpus mucosa, and the other on the canalicular
membranes of parietal cells in the gastric corpus mucosa (Figure 1).
Parietal cells in the corpus secrete gastric acid, and their apical
 Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

FIGURE 1 Helicobacter pylori
infected patient serum reacts with a section from the glandular region of the corpus of an
uninfected person. The in situ binding sites of antigastric autoantibodies induced by H. pylori infection are shown in both
panels. A and B (a) (Immunohistochemistry, original magnification 40) Antiluminal antibodies bind to the luminal mem-
branes (arrows) of the foveolar epithelial cells in the antrum as well as the corpus (shown here). (b) (Immunohistochemistry,
original magnification 100) Anticanalicular antibodies bind to parietal cell canaliculi (black structures, arrows) containing

65
the predominant canalicular antigen, i.e., gastric H þ, K þ-ATPase. Presence of anticanalicular antibodies is associated with
H. pylori atrophic corpus gastritis (see text).
 Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

66
FIGURE 1. Continued.
 Helicobacter pylori and Gastric Autoimmunity 67

secretory canaliculi are rich in Hþ,Kþ-ATPase, the gastric proton pump.

The gastric Hþ,K þ-ATPase has been identified as the single major
autoantigen in chronic H. pylori gastritis with corpus atrophy [9].
In mice, H. pylori infection induces autoantibodies through mimicry
with Lewis antigens present on the gastric proton pump (reviewed in
Appelmelk et al. [10]). H. pylori expresses in its lipopolysaccharide
(LPS) Lewis blood group (Le), antigens that are similar to those
expressed on human cells, including gastric epithelial cells [11,12].
Mice, that were experimentally infected with H. pylori developed anti-
Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

bodies to H. pylori LPS, Hþ,Kþ-ATPase and Lewis x (Lex) and Ley. In

humans, Lex and Ley are expressed on gastric mucin, whereas Ley is
expressed on the b-subunit of gastric Hþ,Kþ-ATPase. The murine gas-
tric Hþ,K þ-ATPase b-subunit contains Lex, and Ley, or both, and
serum of H. pylori
infected mice has been shown to cross-react with
the human gastric mucosa. These murine cross-reactive antibodies
can be removed by preincubation of the serum with H. pylori cells
[11,12], and therefore they are the result of molecular mimicry
between H. pylori and its host. Human sera, however, also react with
recombinant Hþ,Kþ-ATPase that lacks Lewis antigens [9], and their
reactivity to gastric parietal cells cannot be removed by preincubation
of the serum with H. pylori cells [13]. In humans, anti-Hþ,Kþ-ATPase
autoantibodies associated with H. pylori infection are directed
against protein epitopes, and their formation does not involve Lewis
mimicry.
Ferrets are naturally colonized with Helicobacter mustelae and
develop diseases similar to those that are associated with H. pylori
infection in humans, including gastritis and ulcer disease [14,15], gas-
tric adenocarcinoma [16], and MALT lymphoma [17]. Like H. pylori,
H. mustelae expresses in its LPS antigenic structures that are also
present on host epithelial cells, in this case blood group A [18]. How-
ever, as in humans natural infection of ferrets with H. mustelae
induces autoantibodies that react with ferret parietal cells, and these
antibodies cannot be removed with H. mustelae cells or red blood cells
expressing blood group A [19]. Thus, H. mustelae
induced gastric
autoantibodies in ferrets are not due to molecular mimicry and may
arise via similar mechanisms that lead to gastric autoantibody pro-
duction in humans. These findings clearly show the difference in
immune response seen with natural infection models, compared to
the response that occurs in animals after experimental infection or
immunization, and indicate that the naturally H. mustelae
infected
ferret may be a suitable model to investigate the pathogenic mechan-
isms underlying H. pylori
associated corpus atrophy in humans. The
target of antiparietal cell autoantibodies in H. mustelae
infected
 68 M. P. Bergman et al.

ferrets is likely the gastric Hþ,Kþ-ATPase, which is also the target of

antiparietal cell autoantibodies in H. pylori
infected humans.

HISTOLOGICAL AND CLINICAL FEATURES ASSOCIATED

WITH ANTICANALICULAR AUTOANTIBODIES IN
H. PYLORI GASTRITIS
Both antiluminal and anticanalicular antibodies are significantly
Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

associated with H. pylori infection. However, only the presence of

anticanalicular antibodies is significantly correlated with morphologi-
cal features and functional state of the gastric mucosa and clinical
parameters in H. pylori
induced gastritis [7] (Table I). Severity of
gastritis in the corpus mucosa of H. pylori
infected patients is posi-
tively correlated with the presence of anticanalicular autoantibodies,
and anticanalicular autoantibodies also correlate with atrophy and
increased apoptosis in the gastric corpus mucosa [5
7]. This suggests
that H. pylori
associated corpus atrophy may result from an auto-

TABLE I Similarities between Classical AIG and H. pylori
Associated

Corpus Atrophic Gastritis
Antigen
. Gastric Hþ,Kþ-ATPase is the autoantigen recognized by circulating PCA [9,39].
. Circulating antibodies against Hþ,Kþ-ATPase in H. pylori gastritis are suggestive of
the presence of Hþ,Kþ-ATPase-specific T helper cells, such as in AIG.
Cellular immune responses
. Increased B=T cell influx around glands and into the epithelium [9,39].
. Putative antigen presentation by gastric epithelium; increased levels of MHC class II
[46,83,84] and B7.1 and B7.2 molecules on gastric epithelial cells [84,85].
. Increased expression of MAdCAM-1 on gastric endotheliuma [54,59,88] in association
with homing of a4b7 T lymphocytes to the gastric mucosa [54,88].
. Increased Fas expression on parietal cells [53, 56] in EAIG and on gastric epithelial
cellsb in H. pylori infection [90
92].
Histological and clinical features
. Increased apoptosis in the corpus mucosa [6
8].
. Increased body mucosa atrophy [7,9].
. Higher serum gastrin levels [6,7,20].
. Lower pepsinogen I:II ratio [7].
. Lower gastric acid secretion [20,21].
. Decreased incidence of duodenal ulcer [7].
a
In H. felis
infected mice. In humans, MAdCAM-1 is equally expressed on the gastric
epithelium of H. pylori
infected and noninfected individuals [54].
b
Data regarding Fas expression on parietal cells, compared with other epithelial cells,
during H. pylori infection are absent.
 Helicobacter pylori and Gastric Autoimmunity 69

immune attack against the gastric glands, leading to an increased rate

of cell death.
Hþ,Kþ-ATPase is the major target of anticanalicular autoantibo-
dies, and patients with autoreactivity against the a- or b-subunit of
Hþ,Kþ-ATPase show the highest prevalence of corpus atrophy [9]. In
addition, several clinical features of H. pylori
associated antigastric
autoimmunity are also common to glandular atrophy in classical
AIG=PA. H. pylori
infected patients with anticanalicular autoanti-
bodies have significantly higher fasting serum gastrin levels and
Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

lower pepsinogen I to II ratios, which is a sensitive marker for corpus

mucosa atrophy, as compared to patients without anticanalicular auto-
antibodies [7,20]. The presence of anticanalicular autoantibodies also
correlates with a decreased gastric acid output in both H. pylori

infected patients with nonulcer dyspepsia [20] and those with duodenal
ulcer [21].
Based on the similar histopathological and clinical characteristics
of H. pylori
associated atrophic corpus gastritis with anticanalicular
antibodies and classical AIG=PA, it is hypothesized that H. pylori is a
causative agent for the development of autoimmune gastritis [10,22].
This concept is supported by the observations that antigastric auto-
antibodies in some patients—but not all—decrease after cure of infec-
tion [23], and that histologically defined preclinical stages of
autoimmune gastritis can be successfully treated by H. pylori eradi-
cation [24
26]. However, conflicting data regarding the beneficial
effects of eradication therapy have been reported. Eradication of
H. pylori in individuals with atrophic body gastritis had no effect
on body atrophy and intestinal metaplasia, but cure of infection
was associated with increased gastric acid secretion [27,28] and with
a reduction of hypergastrinemia [28]. In a recent study, H. pylori
eradication in infected patients with corpus atrophy did restore both
histomorphology and function of the corpus mucosa, but in only 20%
of the cases [29].
Initially, a negative association between H. pylori and pernicious
anemia was reported, and it was concluded that H. pylori is uniquely
associated with chronic, nonspecific gastritis, and that the pathologi-
cal process in pernicious anemia protects against infection with
H. pylori [30]. In contrast, an association between H. pylori infection
and gastric autoimmunity is supported by a number of later studies
that indicate that a substantial portion of patients with autoimmune
gastritis have or have had H. pylori infection [31
33]. The apparent
controversial reports regarding the prevalence of H. pylori infection
in AIG=PA, and in less strictly defined atrophic body gastritis, are
largely based on the methods used for detection of infection, i.e.,
 70 M. P. Bergman et al.

histology versus serology. Although actual colonization could only be

detected in 10
14% of PA patients and in 33% of individuals with
atrophic body gastritis [31,34,35], antibodies to H. pylori were shown
in 51
83% of patients with PA and in 53
86% of atrophic body gas-
tritis patients [31,32,35,36]. Furthermore, in a group of 150 patients
with atrophic body gastritis, 23% were positive for H. pylori in the
corpus mucosa, both in histology and serology. In this subpopulation
of atrophic body gastritis patients, the presence of PA, as well as
the grade of corpus atrophy, were lower than in patients that were
Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

only positive in serology, whereas the group of patients that were

negative for H. pylori in both serology and histology had the highest
prevalence of PA and corpus mucosal atrophy [36]. Thereby, the group
of serologically and histologically H. pylori
positive patients repre-
sents an intermediate stage between atrophic body gastritis and full
blown AIG, suggesting that H. pylori infection can induce gastric
autoimmune disease and that a majority of patients with atrophic
body gastritis had H. pylori gastritis before the bacteria were cleared
by the development of atrophy. This supports the hypothesis that in
at least a subpopulation of AIG=PA, e.g., patients with a certain
genetic background, H. pylori may be or have been involved in the
etiology of the disease [10,35].
Interestingly, such a proposed association between H. pylori and
gastric autoimmunity seems to be absent in type 1 diabetes patients
[37]. Type 1 diabetes mellitus results from autoimmune destruction
of pancreatic b-cells, and in 15
20% of type 1 diabetic patients
parietal cell antibodies are also present, which is associated with mani-
festations of gastric autoimmunity [38]. Except, in these patients the
presence of Hþ,Kþ-ATPase antibodies, serum gastrin levels, and acid
secretion parameters do not differ between H. pylori
positive and
H. pylori
negative individuals. However, upon division of patients
into groups based on their level of circulating parietal cell antibodies,
it was suggested that low titers (between 1=20 and 1=40) of PCA reflect
H. pylori infection, whereas higher PCA titers appear to be indicative
of gastric autoimmune disease. In addition, in type 1 diabetics,
H. pylori infection or parietal cell antibodies are associated with differ-
ent HLA-DQ haplotypes, i.e., HLA-DQA1 0501-B1 0201 or HLA-
DQA1 0501-B1 0301, respectively [37]. Together, these observations
suggest that H. pylori gastritis and autoimmune gastritis may be sep-
arate entities in type 1 diabetes patients. Longitudinal and post-H.
pylori eradication studies are needed to assess the role of H. pylori
in gastric autoimmune responses in type 1 diabetes patients or to
clarify whether type 1 diabetes mellitus predisposes to gastric auto-
immunity independent of H. pylori infection.
 Helicobacter pylori and Gastric Autoimmunity 71

SIMILAR PATHOGENIC MECHANISMS IN

H. PYLORI–ASSOCIATED ATROPHIC CORPUS GASTRITIS AND
CLASSICAL AUTOIMMUNE GASTRITIS
Striking similarities between corporal glandular autoimmunity asso-
ciated with H. pylori and the gastric mucosal damage observed in
classical AIG suggest that the development of these diseases involves
similar pathogenic mechanisms and may even share the same trigger,
i.e., H. pylori.
Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

Human Autoimmune Gastritis and Pernicious Anemia

AIG is characterized by a chronic inflammatory infiltrate in the gas-
tric mucosa accompanied by loss of gastric parietal cells and zymo-
genic cells; it affects the fundus and the corpus of the stomach but
spares the antrum, and may eventually proceed to PA. PA is con-
sidered to be the most common cause of vitamin B12 deficiency in
Western populations [39]. In AIG, autoantibodies to gastric parietal
cells are present in the circulation, and clinical characteristics of
patients with AIG are hyperplasia of gastrin-producing cells, resulting
in high serum gastrin concentration, decreased acid secretion, and
decreased pepsinogen I:II ratio [39]. As mentioned above, this clinical
spectrum is also significantly correlated with the presence of anticana-
licular autoantibodies in H. pylori infection [7,20].
In addition, H. pylori has been suggested as a causative agent in the
development of adult vitamin B12 deficiency. H. pylori was detected in
57% of patients with vitamin B12 deficiency, and eradication of infec-
tion improved anemia and serum vitamin B12 levels in a subgroup
(40%) of the infected patients [40]. These findings further support
the possibility that H. pylori plays a role in the development of gastric
autoimmunity, i.e., AIG=PA.

T Cells, H+,K+-ATPase, and Cytokines in AIG/EAIG

Current knowledge of the pathogenesis of human AIG is derived
largely from studies of experimental autoimmune gastritis (EAIG)
mouse models, reviewed in Alderuccio et al. [41]. Like AIG, EAIG is
characterized by a chronic inflammatory infiltrate that extends into
the gastric mucosa with loss of acid-secreting parietal cells and zymo-
genic cells, followed by appearance of circulating autoantibodies direc-
ted against the a- and b-subunits of Hþ,Kþ-ATPase [42,43]. Again,
these features are similar to those seen in H. pylori
infected patients
with corpus atrophy.
 72 M. P. Bergman et al.

In EAIG and AIG, the gastric inflammatory infiltrate contains both

CD4þ and CD8þ T cells, macrophages, and B cells [44], and the histo-
pathological lesions in mouse models are similar to those observed in
humans with chronic AIG [45,46]. EAIG is mediated by Hþ,Kþ-
ATPase-specific CD4þ T cells [47], and T cell epitopes in EAIG have
been identified [48
50]. Also, human AIG is mediated by autoreactive
CD4þ T cells that are specific for Hþ,Kþ-ATPase [51], and recently it
has been shown that a significant number (36%) of Hþ,Kþ-ATPase
peptide epitopes recognized by human gastric T cell clones are rel-
Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

evant in EAIG [52]. Autoreactive T cells in human AIG and murine

EAIG are predominantly of the T helper 1 phenotype, secreting high
levels of IFN-c and TNF-a in the absence of IL-4, and express cytotoxic
functions [51,53,54], as do H. pylori
specific CD4þ T cells derived from
the gastric mucosa of infected patients [55].

Destruction of Parietal Cells in AIG/PA and EAIG

Accumulating evidence points to the Fas (CD95)-Fas ligand
(FasL=CD95L) pathway as a major effector mechanism in the mucosal
damage observed in AIG. An Hþ,Kþ-ATPase-specific T cell clone from
an EAIG mouse has been reported to express FasL upon activation
and to induce Fas-FasL-mediated apoptosis of epitope-loaded target
cells in an antigen-dependent way [53]. Likewise, gastric mucosal
Hþ,Kþ-ATPase-specific T cell clones derived from AIG patients exert
their cytotoxic effects via Fas=FasL-induced apoptosis and are capable
of perforin-mediated lysis of target cells [51]. A definite requirement
for Fas in the development of murine EAIG was recently demonstrated
by the failure of lpr=lpr mice, deficient in Fas expression, to develop
destructive gastritis and autoantibodies upon neonatal thymectomy
[56]. In the same study, Fas expression was colocalized with gastric par-
ietal cells in EAIG mice, whereas nonthymectomized, nongastritic mice
did not express Fas in their gastric mucosa. This indicates upregulated
Fas expression on parietal cells in EAIG and may partially explain the
selective destruction of parietal cells observed in murine and human
AIG. Recently it has been shown, again in mice, that dendritic cells
(DC) constitutively acquire Hþ,Kþ-ATPase from the gastric mucosa,
after which they migrate to the draining lymph node and present
the autoantigen to T cells [57]. This process is markedly increased upon
induction of murine EAIG and suggests a major role for DC as antigen-
presenting cells in T cell activation during murine and human AIG.
Increased expression of Major Histocompatibility Complex (MHC) class
II and costimulatory molecule B7.1 on DC from EAIG mice compared to
normal animals, further supports this view [57]. Furthermore, the
 Helicobacter pylori and Gastric Autoimmunity 73

observations of increased MHC class II expression on gastric epithelial

cells [46] suggest that gastric mucosal cells are also capable of antigen
presentation to T cells, a view that is supported by the functional
expression of antigen-processing cathepsins by human gastric epithelial
cells [58]. In addition, expression of MAdCAM-1, the endothelial counter
receptor for integrin a4b7, on the mucosal endothelium is increased
in murine EAIG [54,59], and plays a role in selective extravasation
(‘‘homing’’) of a4b7-integrin positive, autoreactive CD4þ Th1cells to
lesions in the gastric mucosa in EAIG [54].
Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

The above-mentioned observations are in line with the model for

destruction of parietal cells in murine EAIG and human AIG=PA, as
originally proposed by Toh et al. [44]: DC in the gastric mucosa become
activated, capture and process gastric Hþ,Kþ-ATPase that may be
released during normal turnover of parietal cells, and migrate to a
draining lymph node. DC present Hþ,Kþ-ATPase-peptides to naive,
potentially pathogenic Hþ,Kþ-ATPase-specific CD4 þ T cells, which
become activated, proliferate, and acquire adhesion molecules that
mediate their homing to the gastric mucosa. Subsequently, inflamma-
tory cells comprising other T cells, either CD4þ or CD8 þ, monocytes,
and B cells are recruited through release of chemokines and cytokines
by the Hþ,Kþ-ATPase-specific CD4þ T cells [44]. In the gastric corpus
mucosa, CD4þ Hþ,Kþ-ATPase-specific T cells directly mediate the
loss of parietal cells in EAIG=AIG by Fas-FasL-mediated induction
of apoptosis of parietal cells that express Fas, and they have MHC II=
Hþ,Kþ-ATPase complexes on their surface, by Hþ,Kþ-ATPase-specific T
cells expressing Fas ligand [44]. In human AIG, Human Leukocyte Anti-
gen (HLA)-DR is aberrantly expressed on glandular epithelium in the
vicinity of T cell infiltrates [45], indicating that similar mechanisms
may be responsible for parietal cell loss.

Gastric T Cell and Cytokine Responses in H. pylori Infection

H. pylori infection leads to a lymphocytic influx, similar to the one
observed in AIG, and H. pylori
specific cellular immune responses
have been studied in human and animal models for several years.
Many different animal models, such as mice, ferrets, dogs, and cats,
have been used for studying helicobacter infection, and particularly
Helicobacter felis
infected mice provided a useful model resembling
gastric pathological changes in human H. pylori infection. H. felis
infection of B and T cell-deficient (RAG-1-=-) mice and T cell-deficient
(TCRbd-=-) mice [60] does not result in gastric pathology despite high
levels of colonization, whereas infection of B cell-deficient (mMT)
mice results in severe gastric alterations, identical to those seen in
 74 M. P. Bergman et al.

immunocompetent mice [60]. These results demonstrate that immune-

mediated gastric damage that is seen after Helicobacter infection
requires cellular immune responses and is mediated by T cells, not
by B cells or their secreted antibodies.
The outcome of H. pylori infection is dependent on both bacterial
strain and host factors [61,62], and in particular the host cytokine
response is considered an important determinant in the onset of gas-
tritis and the development of disease. Data regarding early stages in
H. pylori infection are scarce, but experimental infection of mice and
Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

nonhuman primates revealed that inflammatory responses to acute

H. pylori infection are of the Th1 phenotype [63,64]. Also, in humans
infected with H. pylori gastric mucosal lymphocytes have a predomi-
nant Th1 phenotype [65,66], which is associated with pathology
[67,68]. However, in individuals with asymptomatic chronic H. pylori
gastritis, the majority—up to 64%—of the gastric mucosal T cells are
of the Th0 phenotype, i.e., able to secrete both IFN-c and IL-4 [55,68].
In vivo neutralization of IFN-c reveals the existence of Th2
responses in H. felis
immunized challenged C57BL=6 mice [63].
In addition, in BALB=c mice oral vaccination with recombinant
urease induces a Th2 response and cures H. felis infection [69]. These
observations suggest that a shift to a Th2 response would be required
for protection from H. pylori infection. However, vaccination with
H. pylori antigen in the presence of either AlOH or Freund’s complete
adjuvant, which induces IL-5-secreting or IFN-c-secreting antigen-
specific T cells, respectively, both confer protection from H. pylori
infection [70]. Furthermore, recent vaccination studies in mice show
that protection against H. pylori does not require IL-4 [71]; instead
it is dependent on IL-12 and is mediated by Th1 cells [72,73].
Together, these observations indicate that Th1 and Th2 cytokines
are both involved in the adaptive immune response against H. pylori,
and that protection from disease may require the ability of the host
to mount a balanced Th1=Th2 response upon infection.
It has been recognized that the BALB=c and C57BL=6 inbred mouse
strains profoundly differ in susceptibility and outcome of H. pylori
infection [3,61,74]. BALB=c mice, which have a genetic tendency to
favor Th2 responses, are much more resistant to infection and
H. pylori
induced gastrointestinal pathology when compared to Th1
prone C57BL=6 mice. With regard to the observations that Th1 and
Th2 are both involved in protection from H. pylori infection, a recent
study of Panthel and coworkers [75] may be of special interest. In this
study, the influence of the host genetic background on the role of
Th1=Th2 responses during helicobacter infection, was assessed by
using IL-4 and IL-12 knockout mutants of BALB=c and C57BL=6J
 Helicobacter pylori and Gastric Autoimmunity 75

mice. It was demonstrated that IL-4 and IL-12—cytokines marking a

Th2 or a Th1 response, respectively—play opposing roles in H. pylori
colonization of the BALB=c and C57BL=6J mouse strains. Disruption
of the IL-12 gene in BALB=c mice, thereby enhancing the pre-existing
Th2 background of these mice, results in strong colonization by
H. pylori compared to BALB=c wild-type mice. In striking contrast,
the same mutation in C57BL=6J mice, making these mice more Th2
skewed than C57BL=6J wild-type mice, results in animals that can
hardly be colonized at all. Similar, but less strong, effects were seen
Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

after disruption of IL-4 in BALB=c and C57BL=6J backgrounds,

respectively. Except for the C57BL=6J IL-12 knockout mice, disrup-
tion of IL-4 or IL-12 in both mouse strains does not affect the response
of the mice to vaccination, compared to the cognate wild-type animals
[75]. These findings further confirm the importance of host factors
on outcome of H. pylori infection, and they may provide a basis for
further studies to elucidate the determinants that are associated
with susceptibility for colonization and subsequent development of
H. pylori
induced pathology.
Like Hþ,Kþ-ATPase-specific T cell clones [51], H. pylori
specific
gastric T cell clones produce high levels of TNF-a and IFN-c upon anti-
genic stimulation [55,76]. In a recent study, a highly significant in situ
correlation was shown between TNF-a and IFN-c production and
grade of gastritis, H. pylori colonization density, epithelial cell apopto-
sis, and Fas=FasL-expression [76].
In conclusion, both Th1 and Th2 cytokines are involved in protec-
tion from H. pylori infection, but a predominant Th1 response is asso-
ciated with disease, whereas the ability of the host to mount a
combined Th1=Th2 response is associated with protection from
H. pylori
induced pathology.

CHANGES IN THE GASTRIC EPITHELIUM IN

H. PYLORI –ASSOCIATED GASTRIC AUTOREACTIVITY
The autoantigen recognized by H. pylori
induced antigastric autoan-
tibodies is the gastric proton pump Hþ,Kþ -ATPase [9]. Gastric Hþ,Kþ-
ATPase is also the major autoantigen recognized by autoantibodies in
AIG=PA [77
80], and is the target for autoreactive T cells that
mediate the destruction of glandular epithelium in murine and human
AIG [48,51,81]. Furthermore, H. pylori gastritis patients with antica-
nalicular autoantibodies develop gastric pathology and symptoms
similar to those seen in AIG=PA patients [7,20]. These observations
suggest that a similar autoimmune response, involving parietal cell
autoantigens may be responsible for corpus atrophy in H. pylori
 76 M. P. Bergman et al.

gastritis. Several findings suggest a role for autoantigen-specific

destruction of the gastric glandular epithelium by CD4 þ T cells in
H. pylori infection with gastric corpus atrophy, similar to the destruction
of glands seen in AIG=PA.
Sustained predominance of the acute Th1 cytokine profile is thought
to play a prominent role in the chronicity of H. pylori
induced gastric
inflammation. In vivo neutralization of IFN-c in H. felis
infected mice
dramatically reduces the severity of gastric inflammation, further sup-
porting the contribution of Th1 cell-mediated immune responses in dis-
Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

ease pathogenesis [63]. IFN-c also plays a key role in the onset of murine
EAIG, and induction of this disease can also be abrogated by neutrali-
zing this Th1 cytokine in vivo [82].
Like in EAIG=AIG, gastric epithelial cells may function as antigen-
presenting cells in the H. pylori
infected gastric mucosa. H. pylori
infection leads to de novo expression of MHC class II [83,84] and costi-
mulatory molecules B7.1 and B7.2 [84,85] on gastric epithelium.
Increase in MHC class II expression correlates with an increased
gastric mucosal influx of CD4þ T cells. Furthermore, IFN-c
induced
expression of B7.2 on epithelial cell lines increases after cognate inter-
action of surface MHC molecules [85]. In addition, human gastric epi-
thelial cells are capable of antigen processing. Gastric epithelial cell
lines, as well as epithelial cells isolated from gastric biopsies, function-
ally express several cathepsins that are capable of processing both
H. pylori and dietary antigens, as well as degrading the class II
MHC-associated invariant chain [58], which is required to allow
peptide-binding by MHC class II molecules [86]. Together, these obser-
vations indicate that gastric epithelial cells may play a role in antigen
presentation and subsequent activation of gastric intraepithelial CD4þ
T cells. Both bacterial antigens and autoantigens that are released
during cell turnover, such as gastric Hþ,Kþ-ATPase, could be pre-
sented to gastric T cells. Increased expression of ICAM-1 on gastric
epithelial cells is found in H. pylori
infected patients [84] and may
play a role in adhesion of T cells to the epithelial cells during antigen
presentation.
Similar to EAIG mice, mice infected with H. pylori express
increased levels of MAdCAM-1 in their lamina propria and submu-
cosa, as compared to noninfected control animals. Most of the lamina
propria infiltrating CD4þ lymphocytes in infected animals coexpress
b7-integrin and are localized in the vicinity of MAdCAM-1 rich areas
6 months after infection [87]. However, when measured 2 weeks
after, H. felis infection does not induce a4b7hi T lymphocytes in
mice, whereas oral immunization with H. felis sonicate does [88].
MAdCAM-1 and a4b7-integrin are involved in the homing of CD4þ
 Helicobacter pylori and Gastric Autoimmunity 77

T cells to the gastric mucosa and protection against H. felis infection in

mice [88], but their roles in human H. pylori infection remain to be
elucidated. MAdCAM-1 was found to be equally expressed on the gastric
epithelium of H. pylori-infected and noninfected individuals [89].

Destruction of Glandular Epithelium in H. pylori Gastritis

with Corpus Mucosa Atrophy
Several studies have shown the involvement of Fas=FasL signaling in
Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

apoptosis of epithelial cells in H. pylori infection [90
92]. Fas receptor

and FasL are expressed on foveolar and glandular epithelial cells in
the antrum and corpus [84,92]. Whereas FasL is constitutively
expressed [84], Fas expression on epithelial cells is increased during
H. pylori infection [92]. Activated FasLþ T cells can lyse Fas-expressing
antigen-presenting cells in an antigen-dependent way (direct lysis), as
well as nearby Fasþ cells in an antigen-independent way (bystander
lysis) [93]. As compared to Th2 cells, Th1 cells express higher levels
of FasL [94], and Th1 cells have been reported to contribute to apop-
tosis of gastric epithelial cells by Fas=FasL interaction [92]. The
expression of Fas on gastric epithelial (cell line) cells is increased
by H. pylori and by the Th1 cytokines IFN-c and TNF-a, either alone
or in any combination, and by IL-1b [95]. Gastric epithelial cell
lines expressing Fas can be killed by T cells that express FasL,
and antibodies that block Fas=FasL interaction inhibit this cytotoxi-
city [92]. These findings are in line with the above-mentioned in situ
correlation between TNF-a and IFN-c secretion and apoptosis in
H. pylori
associated gastritis [76]. Thus, during H. pylori infection
Fasþ gastric parietal cell may be killed upon presentation of autoan-
tigen, e.g., Hþ,Kþ-ATPase, to autoantigen-specific CD4þ T cells
expressing FasL. Gastric parietal cells might also be killed via
bystander lysis by circulating activated FasLþ CD4þ T cells specific
for other antigens, including H. pylori.
In addition to elevated Fas expression on gastric epithelium, H. pylori
is also able to induce Fas expression and Fas=FasL-mediated apoptosis
in T cells. This process is dependent on the presence of the cag patho-
genicity island (PAI) in the infecting H. pylori strain, and it has been
proposed that induction of apoptosis in T cells enhances the survival
of H. pylori and contributes to persistence of infection [96].
Fas=FasL signaling has been shown to be involved in H. pylori

associated apoptosis in several studies [90
92]. Recently the regulat-
ory role of this apoptotic pathway has been addressed in a murine
H. pylori infection model [97]. Mucosal atrophic changes, including
mucus cell hyperplasia and parietal cell loss, are detected more
 78 M. P. Bergman et al.

frequently in the corpus regions of H. pylori
infected C57BL=6 gld

(FasL-deficient) mice compared to infected C57BL=6 wild-type mice,
despite comparable levels of bacterial colonization and gastritis [97].
In addition, apoptotic cells are virtually absent in the mucosa of
H. pylori
infected gld mice, whereas increased apoptosis is observed
in the gastric mucosa of H. pylori
infected C57BL=6 wild-type
mice [97]. The authors suggest that dysregulation of Fas=FasL sign-
aling may either promote carcinogenesis, in the case of the absence
of Fas-triggered programmed cell death, or ulcerogenesis due to
Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

enhanced apoptosis, which would lead to disruption of the gastric

epithelial barrier [97]. It is tempting to hypothesize that the observed
parietal cell loss in H. pylori
infected FasL-deficient mice [97] may be
due to Fas-independent destruction, e.g., perforin-mediated lysis, of
parietal cell by Th1 cells that have escaped Fas-triggered programmed
cell death [98]. However, the observations that splenocytes from
neither H. pylori
infected nor uninfected gld mice produce detectable
levels of IFN-c, whereas they produce low levels of IL-5 [97], seem to
contradict this idea.
In addition to cytokine-induced upregulation of Fas antigen
expression and apoptosis mediated by H. pylori infection, direct con-
tact with the bacterium induces both apoptosis through cytochrome
c release from the mitochondria (‘‘the mitochondrial pathway’’) and
antiapoptotic signals via and NFjB in human gastric cancer cell lines
[99]. However, the significance of induction of apoptosis in parietal
cells by direct contact with H. pylori may be questioned, since H. pylori
usually resides in the human antrum [1]. When inoculated in mice,
H. pylori colonizes gastric niches devoid of parietal cells [100]. This
restriction of bacterial tropism is not seen in H. pylori
infected
transgenic mice that lack parietal cells [100].
Excessive expression of Bcl-2, a 26 kDa protein that usually coun-
teracts apoptotic effects of another protein involved in apoptosis, i.e.,
Bax, is found in AIG and H. pylori
associated atrophic gastritis, but
not in H. pylori antral gastritis without atrophy [101]. Similarities
in increased Fas and FasL expression, and in excess expression of
Bcl-2, between AIG=PA and H. pylori
associated atrophic corpus
gastritis, may indicate similar pathogenic pathways in both diseases.

Models for the Pathogenesis of H. pylori –Associated

Antigastric Autoimmunity
Despite the strikingly similar pathophysiological changes of the gastric
corpus mucosa in AIG=PA and in H. pylori
associated atrophic
gastritis, that suggest an initiating role for H. pylori in both these
 Helicobacter pylori and Gastric Autoimmunity 79

gastric autoimmune disorders, there is often a negative association

between current H. pylori infection and AIG=PA. Studies on the associ-
ation between AIG=PA and H. pylori infection indicate that patients
with PA are less often infected with H. pylori than age-matched controls
[102]. However, as indicated before, a majority of patients with PA show
serological evidence of previous H. pylori infection [32,35], while actual
infection with H. pylori, as shown by histology, is detected in only a min-
ority of PA patients [34,35]. H. pylori infection may have been present
before the development of pernicious anemia and may be cleared during
Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

the period of transition from normal antral and corpus mucosa to prea-
trophic corpus mucosa and established corporal gastric atrophy. This
course of infection is supported by observations from a long-term fol-
low-up study, in which a subgroup of H. pylori
infected patients
initially had gastritis of the antrum that later progressed to severe
atrophy of the corpus [103]. This process was often accompanied by
normalization of the antral mucosa, disappearance of H. pylori, and
appearance of parietal cell antibodies. These data strongly support a
causative role of H. pylori in at least a subpopulation of patients suffer-
ing from gastric autoimmunity.
H. pylori
associated corpus mucosal atrophy, may develop via the
following processes (visualized in Figures 2a
2c). In the noninflamed
gastric mucosa, only a few gastric periglandular lymphocytes are
present. Infection with H. pylori leads to an inflammatory influx of
T and B lymphocytes and macrophages. Subsequently, epithelial cells
acquire properties that are required for antigen presentation
(described above) under the influence of the predominant Th1 cytokine
milieu resulting directly from H. pylori infection [64,66]. Epithelial
cells may then present gastric Hþ,Kþ-ATPase to CD4þ Hþ,Kþ -
ATPase-specific Th1 cells that have escaped negative selection in the
thymus, which may result in killing of Fasþ parietal cells (Figure
2a), and to autoreactive T cells that provide help for B cell stimulation
and autoantibody production. Activated Th1 cells expressing FasL
could kill Fasþ parietal cells by either antigen-dependent Fas
FasL
interaction or by Fas-mediated bystander lysis, resulting in destruc-
tion of glands (atrophy). The abundance of Hþ,Kþ -ATPase in parietal
cells, together with the lack of expression of the Hþ,Kþ-ATPase b-sub-
unit in the thymus [104,105], at least in mice, might favor presen-
tation of Hþ,Kþ-ATPase over presentation of other autoantigens that
are present in the gastric mucosa. During the course of the chronic
antigastric cell-mediated inflammatory response, epitope spreading
to other autoantigens such as intrinsic factor and pepsinogen may
occur. These antigens are targets of autoantibodies in full-blown
AIG=PA [39,106]. Loss of H. pylori and normalization of the antrum
 Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

80
FIGURE 2 Autoimmune destruction of parietal cells during H. pylori infection may involve several mechanisms (for
details see text). (a) H. pylori infection leads to increase in local levels of Th1 cytokines, resulting in both breakdown of
 Helicobacter pylori and Gastric Autoimmunity 81

may parallel increasing corpus atrophy. In the final stage, the auto-
immune T cell responses that were initially triggered by local inflam-
matory factors resulting from H. pylori infection do not require the
presence of bacteria anymore but are maintained by the predominant
Th1 response and cytokine milieu, and they are indistinguishable from
the gastric inflammatory responses seen in AIG=PA (Figure 2a). This
model is supported by the observation in mice that local expression of
GM-CSF, a proinflammatory cytokine, is sufficient to break tolerance
and initiate gastric autoimmunity mediated by Hþ,Kþ-ATPase specific
Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

CD4þ T cells [107].

Alternatively, or parallel to antigen presentation by gastric
epithelial cells, Hþ,Kþ-ATPase may also be presented by DCs. DCs
have been shown to constitutively acquire Hþ,Kþ-ATPase from the
gastric mucosa, after which they migrate to the draining lymph node
and present the autoantigen to T cells [57].
In addition to autoantigen (i.e., Hþ,Kþ-ATPase)-dependent Fas-
mediated lysis, epithelial=parietal cells may be killed by bystander effect
of FasLþ gastric T cells that are H. pylori
specific [92,93]. Presentation
of bacterial antigens by gastric epithelial cells and professional APCs,
such as macrophages and DCs, may result in activation of H. pylori

specific T cells that kill the antigen-presenting epithelial cells by
H. pylori
antigen-dependent (i.e., direct) mechanisms, e.g., perforin-
mediated lysis. Nearby Fasþ parietal cells may be subsequently killed
by antigen-independent bystander lysis (Figure 2b).

3
self-tolerance and survival of negative selection by autoreactive (e.g., Hþ, Kþ-
ATPase-specific) T cells. Increased antigen-presentation function of epithelial
cells, including parietal cells, and elevated levels of Fas expression on parietal
cells as compared to other epithelial cells, may result in selective destruction
of parietal cells by Hþ,Kþ-ATPase-specific autoreactive T cells, via antigen-
dependent Fas=FasL-mediated apoptosis. (b) H. pylori infection induces
activation and proliferation of H. pylori
specific T cells that directly target
epithelial cells presenting H. pylori antigens by antigen-dependent
mechanisms, like perforin-mediated lysis. Nearby Fasþ parietal cells may be
destroyed via ‘‘bystander’’ apoptosis by H. pylori
specific T cells that express
FasL. The selective disappearance of parietal and zymogenic cells may be due
to disruption of mucosal developmental pathways during gastric autoimmun-
ity (see text). (c) The population of H. pylori
specific T cells, induced by
infection, may contain a subpopulation of T cells that cross-react with peptide
epitopes of Hþ,Kþ-ATPase. H. pylori=Hþ,Kþ-ATPase cross-reactive T cells
could mediate destruction of both nonparietal epithelial cells and parietal cells
via antigen-dependent, perforin-mediated lysis, and may also target Fasþ
parietal cells by ‘‘bystander’’ apoptosis. (d) Legend to (a)
(c).
 Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

82
FIGURE 2. Continued.
 Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

FIGURE 2. Continued.

83
 Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

84
FIGURE 2. Continued.
 Helicobacter pylori and Gastric Autoimmunity 85

The increased expression of Fas on parietal cells compared with

other cells in the gastric epithelium may explain the selective destruc-
tion of parietal cells in (E)AIG and the H. pylori
associated corpus
atrophy models described above. However, zymogenic cells also disap-
pear during these diseases. It has been proposed that in EAIG the nor-
mal developmental pathways of the gastric mucosa are disrupted,
resulting in the accumulation of mucous neck cells and immature stem
cells (hyperplasia), that do not differentiate to end-stage cells. This
may explain the depletion of zymogenic cells, although they are not
Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

directly targeted by the immune system [108].

A third model of how H. pylori infection may mediate autoimmune
destruction of the gastric corpus mucosa, and finally result in AIG=PA,
PA, involves molecular mimicry at the level of T cell epitopes (Figure
2c). The detection of anti-Hþ,Kþ-ATPase antibodies in the circulation
of H. pylori
infected patients is suggestive of the presence of T helper
cells specific for the same autoantigen. It has been proposed that
sequence homology between H. pylori polypeptides and Hþ, Kþ-ATPase
may underlie the induction of autoreactive antibodies and T cells by H.
pylori [109]. In such a setting, infection induces activation and expan-
sion of H. pylori
specific T cells that recognize not only their bacterial
peptide epitope but also a peptide derived from Hþ,Kþ-ATPase.
Upon activation, these cross-reactive T cells may directly kill antigen-
presenting gastric epithelial cells that present the appropriate peptide
(i.e., derived from H. pylori or Hþ,Kþ-ATPase), and indirectly induce
apoptosis in Fasþ parietal cells. Such events may provide an additional
explanation for the selective destruction of parietal cells, because
Hþ,Kþ-ATPase is exclusively expressed in parietal cells, and also Fas
expression seems to be predominantly upregulated in parietal cells dur-
ing gastric autoimmunity as compared to other epithelial cells.
This model, in which molecular mimicry forms the basis of the
pathogenesis of autoimmune gastric atrophy, is strongly supported
by the recent discovery of in vivo
activated cross-reactive autoreac-
tive T cells in patients with both AIG and concurrent H. pylori infec-
tion [110]. In addition to CD4þ T cell clones reactive to H. pylori
lysate, and CD4þ T cell clones that recognized Hþ,Kþ-ATPase, these
patients harbor gastric Th1 cells that recognize H. pylori antigens as
well as Hþ,Kþ-ATPase. The different submolecular specificities of such
gastric T cells have been characterized, and cross-reactive epitopes
from nine H. pylori proteins have been identified. Of note, recognition
of the cross-reactive peptides is dependent on a particular HLA-DR
haplotype, and none of the identified cross-reactive proteins belong
to the known immunodominant antigens of H. pylori. Instead they
belong to products of H. pylori household genes, and include histidine
 86 M. P. Bergman et al.

kinase, acetate kinase, and penicillin-binding protein as well as a LPS

biosynthesis protein. Recognition of these H. pylori cross-reactive pep-
tides leads to functional activation of cross-reactive T cell clones and
primes mechanisms that are associated with gastric pathology in
AIG [51,110].
The novel data provided by our laboratories fit into the following
model for the role of H. pylori in the development of gastric corpus
mucosal atrophy and subsequently chronic AIG: H. pylori-induced
chronic inflammation may result in breakdown of tolerance for
Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

self-antigens, including Hþ,Kþ-ATPase, in genetically predisposed

individuals, leading to escape from negative selection of self-
reactive=cross-reactive T cells, i.e., both Hþ,Kþ-ATPase-specific and
H. pylori=Hþ,Kþ-ATPase cross reactive. These potentially autoreactive
T cells become activated and expand due to constitutive presentation
of Hþ,Kþ-ATPase by professional APC, such as DCs and macrophages.
In addition, in individuals expressing certain but as yet unidentified
HLA haplotypes, appropriate presentation of H. pylori antigens may
result in expansion and activation of H. pylori=Hþ,Kþ-ATPase
cross-reactive, autoreactive T cells. As a result from continuous
inflammation, epitope spreading to other autoantigen epitopes of
Hþ,Kþ-ATPase (that are not mimicked by bacterial peptides), may
occur, and the presence of H. pylori antigens to maintain the popu-
lation of autoreactive T cells is not further required. With ongoing cor-
pus atrophy and increase in gastric pH due to disappearance of
parietal cells, H. pylori may be deprived from its preferred niche and
be eventually lost. After this step, the gastric mucosal functions and
clinical characteristics of an individual with H. pylori-triggered
atrophic corpus gastritis will resemble classical AIG.
In conclusion, we have discussed the striking similarities between
H. pylori-induced gastric autoimmunity and classical AIG=PA with
respect to both humoral and cell-mediated immune responses.
Although there are still many questions regarding the exact role of
H. pylori in gastric autoimmunity and the pathogenic mechanism(s)
involved in the development of gastric corpus atrophy, there is
increasing evidence that supports the idea that in some hosts
H. pylori may initiate AIG via molecular mimicry.

REFERENCES
[1] S. Suerbaum and P. Michetti, N. Engl. J. Med., 347: 1175
1186, 2002.
[2] N. Uemura, S. Okamoto, S. Yamamoto, N. Matsumura, S. Yamaguchi, M.
Yamakido, K. Taniyama, N. Sasaki, and R.J. Schlemper, N. Engl. J. Med., 345:
784
789, 2001.
 Helicobacter pylori and Gastric Autoimmunity 87

[3] T. Sakagami, M. Dixon, J. O’Rourke, R. Howlett, F. Alderuccio, J. Vella,

T. Shimoyama, and A. Lee, Gut, 39: 639
648, 1996.
[4] R. Negrini, L. Lisato, I. Zanella, L. Cavazzini, S. Gullini, V. Villanacci, C. Poiesi,
A. Albertini, and S. Ghielmi, Gastroenterology, 101: 437
445, 1991.
[5] R. Negrini, A. Savio, C. Poiesi, B.J. Appelmelk, F. Buffoli, A. Paterlini, P. Cesari,
M. Graffeo, D. Vaira, and G. Franzin, Gastroenterology, 111: 655
665, 1996.
[6] G. Faller, H. Steininger, M. Eck, J. Hensen, E.G. Hann, and T. Kirchner, Virchows
Arch., 427: 483
486, 1996.
[7] G. Faller, H. Steininger, J. Krnzlein, H. Maul, T. Kerkau, J. Hensen, E.G. Hahn,
and T. Kirchner, Gut, 41: 619
623, 1997.
Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

[8] H. Steininger, G. Faller, E. Dewald, T. Brabletz, A. Jung, and T. Kirchner,

Virchows Arch., 433: 13
18, 1998.
[9] D. Claeys, G. Faller, B.J. Appelmelk, R. Negrini, and T. Kirchner, Gastroentero-
logy, 115: 340
347, 1998.
[10] B.J. Appelmelk, G. Faller, D. Claeys, T. Kirchner, and C.M. Vandenbroucke-Grauls,
Immunol. Today, 19: 296
299, 1998.
[11] B.J. Appelmelk, I. Simoons-Smit, R. Negrini, A.P. Moran, G.O. Aspinall,
J.G. Forte, T. De Vries, H. Quan, T. Verboom, J.J. Maaskant, P. Ghiara, E.J.
Kuipers, E. Bloemena, T.M. Tadema, R.R. Townsend, K. Tyagarajan, J.M.J.
Crothers, M.A. Monteiro, A. Savio, and J. de Graaff, Infect. Immunol., 64:
2031
2040, 1996.
[12] J.L. Guruge, P.G. Falk, R.G. Lorenz, M. Dans, H.P. Wirth, M.J. Blaser, D.E. Berg,
and J.I. Gordon, Proc. Natl. Acad. Sci. USA, 95: 3925
3930, 1998.
[13] G. Faller, H. Steininger, B. Appelmelk, and T. Kirchner, J. Clin. Pathol., 51:
244
245, 1998.
[14] J.G. Fox, P. Correa, N.S. Taylor, A. Lee, G. Otto, J.C. Murphy, and R. Rose,
Gastroenterology, 99: 352
361, 1990.
[15] J.G. Fox, G. Otto, N.S. Taylor, W. Rosenblad, and J.C. Murphy, Infect. Immunol.,
59: 1875
1880, 1991.
[16] J.G. Fox, C.A. Dangler, W. Sager, R. Borkowski, and J.M. Gliatto, Vet. Pathol., 34:
225
229, 1997.
[17] S.E. Erdman, P. Correa, L.A. Coleman, M.D. Schrenzel, X. Li, and J.G. Fox, Am.
J Pathol., 151: 273
280, 1997.
[18] O. Croinı́n, M. Clyne, and B. Drumm, Gastroenterology, 114: 690
696, 1998.
[19] T.O. Croinin, M. Clyne, B.J. Appelmelk, and B. Drumm, Infect. Immunol., 69:
2708
2713, 2001.
[20] G. Faller, M. Winter, H. Steininger, P. Konturek, S.J. Konturek, and T. Kirchner,
Scand. J. Gastroenterol., 33: 276
282, 1998.
[21] F. Parente, R. Negrini, V. Imbesi, G. Maconi, C. Cucino, and G. Bianchi Porro, The
presence of gastric autoantibodies impairs gastric secretory function in patients
with H. pylori-positive duodenal ulcer. Gut, 45 (Suppl. 3). A40, 1999.
[22] G. Faller and T. Kirchner, Microsc. Res. Tech., 48: 321
326, 2000.
[23] G. Faller, M. Winter, H. Steininger, N. Lehn, A. Meining, E. Bayerd’’rffer, and
T. Kirchner, Pathol. Res. Pract., 195: 243
246, 1999.
[24] M. Stolte, K. Baumann, B. Bethke, M. Ritter, E. Lauer, and H. Eidt, Z. Gastroenterol.,
30: 729
735, 1992.
[25] M. Stolte, E. Meier, and A. Meining, Z. Gastroenterol., 36: 641
643, 1998.
[26] A. Tucci, L. Poli, C. Tosetti, G. Biasco, W. Grigioni, O. Varoli, C. Mazzoni,
G.F. Paparo, V. Stanghellini, and G. Caletti, Am. J. Gastroenterol., 93:
1425
1431, 1998.
 88 M. P. Bergman et al.

[27] E.M. El Omar, K. Oien, A. El Nujumi, D. Gillen, A. Wirz, S. Dahill, C. Williams,

J.E. Ardill, and K.E. McColl, Gastroenterology, 113: 15
24, 1997.
[28] B. Annibale, M.R. Aprile, G. D’Ambra, P. Caruana, C. Bordi, and F.G. Delle,
Aliment. Pharmacol. Ther., 14: 625
634, 2000.
[29] B. Annibale, E. Di Giulio, P. Caruana, E. Lahner, G. Capurso, C. Bordi, and
G. delle Fave, Aliment. Pharmacol. Ther., 16: 1723
1731, 2002.
[30] T.L. Fong, C.P. Dooley, M. Dehesa, H. Cohen, R. Carmel, P.L. Fitzgibbons,
G.I. Perez-Perez, and M.J. Blaser, Gastroenterology, 100: 328
332, 1991.
[31] W.E. Karnes, Jr., I.M. Samloff, M. Siurala, M. Kekki, P. Sipponen, S.W. Kim, and
J.H. Walsh, Gastroenterology, 101: 167
174, 1991.
Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

[32] J.Y. Ma, K. Borch, S.E. Sjöstrand, L. Janzon, and S. Mårdh, Scand. J. Gastroenterol.,
29: 961
965, 1994.
[33] S. Eidt, G. Oberhuber, A. Schneider, and M. Stolte, Pathol. Res. Pract., 192:
101
106, 1996.
[34] O. Varis, J. Valle, and M. Siurala, Scand. J. Gastroenterol., 28: 705
708, 1993.
[35] B. Annibale, E. Lahner, C. Bordi, G. Martino, P. Caruana, C. Grossi, R. Negrini,
and F.G. Delle, Dig. Liver Dis., 32: 756
762, 2000.
[36] B. Annibale, R. Negrini, P. Caruana, E. Lahner, C. Grossi, C. Bordi, and G. delle
Fave, Helicobacter, 6: 225
233, 2001.
[37] C.E. De Block, I.H. De Leeuw, J.J. Bogers, P.A. Pelckmans, M.M. Ieven, E.A.
Van Marck, H. Van, V.E. Maday, K.L. Van Acker, and L.F. Van Gaal, Aliment.
Pharmacol. Ther., 16: 281
289, 2002.
[38] C.E. De Block, I.H. De Leeuw, and L.F. Van Gaal, J. Clin. Endocrinol. Metab., 84:
4062
4067, 1999.
[39] B.H. Toh, I.R. van Driel, and P.A. Gleeson, N. Engl. J. Med., 337: 1441
1448,
1997.
[40] K. Kaptan, C. Beyan, A.U. Ural, T. Cetin, F. Avcu, M. Gulsen, R. Finci, and
A. Yalcin, Arch. Intern. Med., 160: 1349
1353, 2000.
[41] F. Alderuccio, J.W. Sentry, A.C. Marshall, M. Biondo, and B.H. Toh, Clin. Immu-
nol., 102: 48
58, 2002.
[42] Y. Mori, K. Fukuma, Y. Adachi, K. Shigeta, R. Kannagi, H. Tanaka, M. Sakai,
K. Kuribayashi, H. Uchino, and T. Masuda, Gastroenterology, 97: 364
375, 1989.
[43] C.M. Jones, J.M. Callaghan, P.A. Gleeson, Y. Mori, T. Masuda, and B.H. Toh,
Gastroenterology, 101: 287
294, 1991.
[44] B.H. Toh, J.W. Sentry, and F. Alderuccio, Immunol. Today, 21: 348
354, 2000.
[45] P. Burman, O. Kampe, W. Kraaz, L. Loof, A. Smalka, A. Karlsson, and A.
Karlsson-Parra, Gastroenterology, 103: 934
942, 1992.
[46] T.M. Martinelli, I.R. van Driel, F. Alderuccio, P.A. Gleeson, and B.H. Toh,
Gastroenterology, 110: 1791
1802, 1996.
[47] H.D. De Silva, I.R. van Driel, N. La Gruta, B.H. Toh, and P.A. Gleeson, Immu-
nology, 93: 405
408, 1998.
[48] A. Nishio, M. Hosono, Y. Watanabe, M. Sakai, M. Okuma, and T. Masuda,
Gastroenterology, 107: 1408
1414, 1994.
[49] E. Suri-Payer, A.Z. Amar, R. McHugh, K. Natarajan, D.H. Margulies, and
E.M. Shevach, Eur. J. Immunol., 29: 669
677, 1999.
[50] H.D. De Silva, P.A. Gleeson, B.H. Toh, I.R. van Driel, and F.R. Carbone,
Immunology, 96: 145
151, 1999.
[51] M.M. D’Elios, M.P. Bergman, A. Azzurri, A. Amedei, M. Benagiano, J.J. De Pont,
F. Cianchi, C.M. Vandenbroucke-Grauls, S. Romagnani, B.J. Appelmelk, and
G. Del Prete, Gastroenterology, 120: 377
386, 2001.
 Helicobacter pylori and Gastric Autoimmunity 89

[52] M.P. Bergman, A. Amedei, M.M. D’Elios, A. Azzurri, M. Benagiano, C. Tamburini,

R. van der Zee, C.M. Vandenbroucke-Grauls, B.J. Appelmelk, and G. Del Prete,
Eur. J. Immunol., 33: 539
545, 2003.
[53] A. Nishio, T. Katakai, C. Oshima, S. Kasakura, M. Sakai, S. Yonehara, T. Suda,
S. Nagata, and T. Masuda, Gastroenterology, 111: 959
967, 1996.
[54] T. Katakai, K.J. Mori, T. Masuda, and A. Shimizu, Int. Immunol., 14: 167
175,
2002.
[55] M.M. D’Elios, M. Manghetti, M. De Carli, F. Costa, C.T. Baldari, D. Burroni, J.L.
Telford, S. Romagnani, and G. Del Prete, J. Immunol., 158: 962
967, 1997.
[56] A.C. Marshall, F. Alderuccio, and B.H. Toh, Gastroenterology, 123: 780
789, 2002.
Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

[57] C. Scheinecker, R. McHugh, E.M. Shevach, and R.N. Germain, J Exp. Med., 196:
1079
1090, 2002.
[58] C. Barrera, G. Ye, R. Espejo, S. Gunasena, R. Almanza, J. Leary, S. Crowe,
P. Ernst, and V.E. Reyes, Hum. Immunol., 62: 1081
1091, 2001.
[59] S.P. Barrett, A. Riordon, B.H. Toh, P.A. Gleeson, and I.R. van Driel, J. Leukoc.
Biol., 67: 169
173, 2000.
[60] K.A. Roth, S.B. Kapadia, S.M. Martin, and R.G. Lorenz, J. Immunol., 163:
1490
1497, 1999.
[61] M. Mohammadi, R. Redline, J. Nedrud, and S. Czinn, Infect. Immunol., 64:
238
245, 1996.
[62] N.E. van Doorn, F. Namavar, M. Sparrius, J. Stoof, E.P. van Rees, L.J. van Doorn,
and C.M. Vandenbroucke-Grauls, Infect. Immunol., 67: 3040
3046, 1999.
[63] M. Mohammadi, S. Czinn, R. Redline, and J. Nedrud, J. Immunol., 156:
4729
4738, 1996.
[64] J.J. Mattapallil, S. Dandekar, D.R. Canfield, and J.V. Solnick, Gastroenterology,
118: 307
315, 2000.
[65] K.B. Bamford, X. Fan, S.E. Crowe, J.F. Leary, W.K. Gourley, G.K. Luthra, E.G.
Brooks, D.Y. Graham, V.E. Reyes, and P.B. Ernst, Gastroenterology, 114:
482
492, 1998.
[66] F. Sommer, G. Faller, P. Konturek, T. Kirchner, E.G. Hahn, J. Zeus, M. Röllinghoff,
and M. Lohoff, Infect. Immunol., 66: 5543
5546, 1998.
[67] M. Mohammadi, J. Nedrud, R. Redline, N. Lycke, and S.J. Czinn, Gastroenterology,
113: 1848
1857, 1997.
[68] M.M. D’Elios, M. Manghetti, F. Almerigogna, A. Amedei, F. Costa, D. Burroni,
C.T. Baldari, S. Romagnani, J.L. Telford, and G. Del Prete, Eur. J. Immunol,
27: 1751
1755, 1997.
[69] P.F. Saldinger, N. Porta, P. Launois, J.A. Louis, G.A. Waanders, H. Bouzourene,
P. Michetti, A.L. Blum, and I.E. Corthesy-Theulaz, Gastroenterology, 115:
891
897, 1998.
[70] J.M. Gottwein, T.G. Blanchard, O.S. Targoni, J.C. Eisenberg, B.M. Zagorski, R.W.
Redline, J.G. Nedrud, M. Tary-Lehmann, P.V. Lehmann, and S.J. Czinn, J. Infect.
Dis., 184: 308
314, 2001.
[71] C.A. Garhart, J.G. Nedrud, F.P. Heinzel, N.E. Sigmund, and S.J. Czinn, Infect.
Immunol., 71: 3628
3633, 2003.
[72] C.A. Garhart, F.P. Heinzel, S.J. Czinn, and J.G. Nedrud, Infect. Immunol., 71:
910
921, 2003.
[73] A.A. Akhiani, J. Pappo, Z. Kabok, K. Schon, W. Gao, L.E. Franzen, and N. Lycke,
J. Immunol., 169: 6977
6984, 2002.
[74] A.E. Kamradt, M. Greiner, P. Ghiara, and S.H. Kaufmann, Microbes Infect.,
2: 593
597, 2000.
[75] K. Panthel, G. Faller, and R. Haas, Infect. Immunol., 71: 794
800, 2003.
 90 M. P. Bergman et al.

[76] F.S. Lehmann, L. Terracciano, I. Carena, C. Baeriswyl, J. Drewe, L. Tornillo,

G. De Libero, and C. Beglinger, Am. J. Physiol. Gastrointest. Liver Physiol, 283:
G481
G488, 2002.
[77] F.A. Karlsson, P. Burman, L. Loof, and S. Mardh, J. Clin. Invest., 81: 475
479,
1988.
[78] I. Goldkorn, P.A. Gleeson, and B.H. Toh, J. Bio Chem., 264: 18768
18774, 1989.
[79] B.H. Toh, P.A. Gleeson, R.J. Simpson, R.L. Moritz, J.M. Callaghan, I. Goldkorn,
C.M. Jones, T.M. Martinelli, F.T. Mu, and D.C. Humphris, Proc. Natl. Acad.
Sci. USA, 87: 6418
6422, 1990.
[80] J.M. Callaghan, M.A. Khan, F. Alderuccio, I.R. van Driel, P.A. Gleeson, and
Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

B.H. Toh, Autoimmunity, 16: 289
295, 1993.

[81] E. Suri-Payer, P.J. Kehn, A.W. Cheever, and E.M. Shevach, J. Immunol., 157:
1799
1805, 1996.
[82] S.P. Barrett, P.A. Gleeson, H. de Silva, B.H. Toh, and I.R. van Driel, Eur. J.
Immunol., 26: 1652
1655, 1996.
[83] L. Engstrand, A. Scheynius, C. Pahlson, L. Grimelius, A. Schwan, and S.
Gustavsson, Infect. Immunol., 57: 827
832, 1989.
[84] A. Archimandritis, S. Sougioultzis, P.G. Foukas, M. Tzivras, P. Davaris, and
H.M. Moutsopoulos, Clin. Exp. Immunol., 119: 464
471, 2000.
[85] G. Ye, C. Barrera, X. Fan, W.K. Gourley, S.E. Crowe, P.B. Ernst, and V.E. Reyes,
J. Clin. Invest., 99: 1628
1636, 1997.
[86] H.A. Chapman, Curr. Opin. Immunol., 10: 93
102, 1998.
[87] K. Hatanaka, R. Hokari, K. Matsuzaki, S. Kato, A. Kawaguchi, S. Nagao,
H. Suzuki, K. Miyazaki, E. Sekizuka, H. Nagata, H. Ishii, and S. Miura, Clin.
Exp. Immunol., 130: 183
189, 2002.
[88] M. Michetti, C.P. Kelly, J.P. Kraehenbuhl, H. Bouzourene, and P. Michetti,
Gastroenterology, 119: 109
118, 2000.
[89] M. Quiding-Jarbrink, I. Ahlstedt, C. Lindholm, E.L. Johansson, and H. Lonroth,
Gut, 49: 519
525, 2001.
[90] J. Rudi, D. Kuck, S. Strand, A. von Herbay, S.M. Mariani, P.H. Krammer,
P.R. Galle, and W. Stremmel, J. Clin. Invest., 102: 1506
1514, 1998.
[91] N.L. Jones, A.S. Day, H.A. Jennings, and P.M. Sherman, Infect. Immunol., 67:
4237
4242, 1999.
[92] J. Wang, X. Fan, C. Lindholm, M. Bennett, J. O’Connoll, F. Shanahan, E.G. Brooks,
V.E. Reyes, and P.B. Ernst, Infect. Immunol., 68: 4303
4311, 2000.
[93] K.A. Sabelko-Downes and J.H. Russell, Curr. Opin. Immunol., 12: 330
335, 2000.
[94] F. Ramsdell, M.S. Seaman, R.E. Miller, K.S. Picha, M.K. Kennedy, and D.H.
Lynch, Int. Immunol., 6: 1545
1553, 1994.
[95] J. Houghton, L.S. Macera-Bloch, L. Harrison, K.H. Kim, and R.M. Korah, Infect.
Immunol., 68: 1189
1195, 2000.
[96] J. Wang, E.G. Brooks, K.B. Bamford, T.L. Denning, J. Pappo, and P.B. Ernst,
J. Immunol., 167: 926
934, 2001.
[97] N.L. Jones, A.S. Day, H. Jennings, P.T. Shannon, E. Galindo-Mata, and
P.M. Sherman, Infect. Immunol., 70: 2591
2597, 2002.
[98] M.J. Lenardo, J. Exp. Med., 183: 721
724, 1996.
[99] S. Maeda, H. Yoshida, Y. Mitsuno, Y. Hirata, K. Ogura, Y. Shiratori, and
M. Omata, Gut, 50: 771
778, 2002.
[100] A.J. Syder, J.D. Oh, J.L. Guruge, D. O’Donnell, M. Karlsson, J.C. Mills,
B.M. Bjorkholm, and J.I. Gordon, Proc. Natl. Acad. Sci. USA, 100: 3467, 2003.
[101] Y. Maor-Kendler, G. Gabay, J. Bernheim, T. Naftali, I.U. Lesin, G. Leichtman,
I. Pomeranz, and B. Novis, Dig. Dis. Sci., 44: 680
685, 1999.
 Helicobacter pylori and Gastric Autoimmunity 91

[102] G.I. Perez-Perez, Clin. Infect. Dis., 25: 1020
1022, 1997.

[103] J. Valle, M. Kekki, P. Sipponen, T. Ihamaki, and M. Siurala, Scand. J.
Gastroenterol., 31: 546
550, 1996.
[104] F. Alderuccio, B.H. Toh, S.S. Tan, P.A. Gleeson, and I.R. van Driel, J. Exp. Med.,
178: 419
426, 1993.
[105] F. Alderuccio, P.A. Gleeson, S.P. Berzins, M. Martin, I.R. van Driel, and B.H. Toh,
Autoimmunity, 25: 167
175, 1997.
[106] S. Mardh and Y.H. Song, Acta Physiol. Scand., 136: 581
587, 1989.
[107] M. Biondo, Z. Nasa, A. Marshall, B.H. Toh, and F. Alderuccio, J. Immunol., 166:
2090
2099, 2001.
Downloaded by [Hong Kong Polytechnic University] at 13:24 30 October 2015

[108] L.M. Judd, P.A. Gleeson, B.H. Toh, and I.R. van Driel, Am. J. Physiol., 277:
G209
G218, 1999.
[109] B.J. Appelmelk, R. Negrini, A.P. Moran, and E.J. Kuipers, Trends Microbiol., 5:
70
73, 1997.
[110] A. Amedei, M.P. Bergman, B.J. Appelmelk, A. Azzurri, M. Benagiano, C. Tamburini,
R. van der Zee, J.L. Telford, C.M.J.E. Vandenbroucke-Grauls, M.M. D’Elios,
G. Del Prete, J. Exp. Med., 198: 1147
1156, 2003.

View publication stats