helicobacter pylori adhesion and patho-adaptation142899/fulltext01.pdf · 2. microbiology of...
TRANSCRIPT
UMEÅ UNIVERSITY MEDICAL DISSERTATIONSAbstract No. 83, ISSN 0345-7532, ISBN 91-7305-548-4
From the Department of Odontology, Division of Oral Microbiology,Faculty of Medicine, Umeå University, Umeå, Sweden
Helicobacter pylori Adhesion and Patho-adaptation
The role of BabA and SabA adhesins in
persistent infection and chronic inflammation
by
Jafar Mahdavi
Umeå, 2004
Print & Media
To my lovely family;
Vina, Melissa and Mona
Print & Media
Vision without action is a daydream,
Action without vision is a nightmare.
Derived from The Analects of Confucius
Print & Media
3
Contents
Page
Preface 5
Abbreviations 6
Abstract 8
Introduction 9
1. A brief history of H. pylori infection 9
2. Microbiology of Helicobacter 9
3. Epidemiology 10
4. The Lewis and ABO blood group systems and H. pylori infection 11
5. Biodiversity 13
5.1 Habitat diversity 13
5.2 Genetic diversity 13
5.3 Tropism and patho-adaptation 16
5.4 Biofilm formation 17
6. Virulence factors and persistence of infection 17
6.1 Adhesins 19
6.1.1 Consequences of bacterial adhesion 19
6.1.1.1 Effect on the bacterial cells 20
6.1.1.2 Effect on the host cells 21
6.2 Invasion of host cells 22
6.3 H. pylori enzymes 23
6.4 VacA and CagA 23
6.5 H. pylori LPS 24
Print & Media
4
7. Diseases associated with H. pylori infection 25
7.1 Gastritis 25
7.2 Peptic ulcer disease 27
7.3 Gastric adenocarcinoma 27
7.4 Barrett’s esophagus 28
Aims 29
Results and Discussion 30
8. The role of MUC5AC in H. pylori adherence 30
9. Gastric mucosa sialylation and inflammation 30
10. Identification of inflammation associated H. pylori adhesive activities 32
11. The role of BabA and SabA in persistent infection and 32
chronic inflammation
11.1 phase variation and biofilm formation 33
11.2 SabA and BabA in relation to patho-adaptation 34
12. The role of LPS in H. pylori adherence 35
Specific conclusions 37
Future perspectives 38
Acknowledgements 40
References 41
Appendix I-IV 53
Print & Media
5
Preface
This thesis is based on the following papers, which are referred to in the textby their Roman numerals:
I. Van de Bovenkamp J.H., Mahdavi J., Korteland-Van Male A., Büller H.,
Einerhand A., Borén T., and Dekker J. The MUC5AC glycoprotein is the primary
receptor for Helicobacter pylori in the human stomach.
Helicobacter 2003, 8, 521-532.
II. Mahdavi J., Sondén B., Hurtig M., Olfat O.F., Forsberg L., Roche N., Ångström
J., Larsson T., Teneberg S., Karlsson K.A., Altraja S., Wadström T., Kersulyte
D., Berg E.D., Dubois A., Petersson C., Magnusson K.M., Norberg T., Lindh F.,
Lundskog B.B., Arnqvist A., Hammarström L., and Borén T. Helicobacter pylori
SabA adhesin in persistent infection and chronic inflammation.
Science 2002, 297, 573-578.
III. Lindén S., Mahdavi J., Olsen C., Borén T., Carlstedt I., and Dubois A. Effects of
Helicobacter pylori inoculation on host glycosylation and H. pylori adhesion sites
in rhesus monkey.
Submitted.
IV. Mahdavi J., Borén T., Vandenbroucke-Grauls C., and Appelmelk B.J. Limited
role of lipopolysaccharide Lewis antigens in adherence of Helicobacter pylori to
the human gastric epithelium.
Infection & Immunity 2003, 71, 2876-2880.
Print & Media
6
Abbreviations
ALeb: blood group A derivative of Lewis b.
BLeb: blood group B derivative of Lewis b.
BabA: blood group antigen binding adhesin.
cagA: cytotoxin-associated gene A.
cag-PAI: cag-pathogenicity island.
FimH: adhesive subunit of type I fimbriae.
GERD: gastro-esophageal reflux disease.
GI: gastrointestinal.
IARC: International Agency for Research on Cancer.
ICAM: intercellular adhesion molecule.
IL-8: interleukin-8.
IVEC: in vitro explant culture.
LPS: lipopolysaccharide.
Le a, b, x and y: Lewis blood group antigen a, b, x and y.
MALT: mucosa-associated lymphoid tissue.
MAP: mitogen-activated protein kinase.
NADH: nicotinamide adenine dinucleotide (co-enzyme).
NAP: neutrophil activating protein.
NF-�B: nuclear factor-�B.
NMR: nuclear magnetic resonance.
NIH: National Institutes of Health.
OMP: outer membrane protein.
P-(pap)-fimbriae: pyelonephritis-associated pilus.
PMN: polymorphonuclear leukocyte.
RBC: red blood cell.
ROS: reactive oxygen species.
SabA: sialic acid binding adhesin.
sLex: sialyl-Lewis x antigen.
sLea: sialyl-Lewis a antigen.
sdiLex: sialyl-dimeric Lewis x antigen.
TLR: toll-like receptor.
TNF�: tumor necrosis factor alpha.
Print & Media
7
UTI: urinary tract infection.
VacA: vacuolating cytotoxin A.
Print & Media
8
Helicobacter pylori Adhesion and Patho-adaptation
The role of BabA and SabA adhesins in
persistent infection and chronic inflammation
Helicobacter pylori (H. pylori) is a human-specific gastric pathogen which is responsiblefor a spectrum of diseases ranging from superficial gastritis to gastric and duodenalulceration, and which is also highly associated with gastric cancer. The pathogenesis ofsevere gastric disorders caused by H. pylori is multifactorial and involves complexinteractions between the microbe and the gastric mucosa. H. pylori expresses severaladhesion proteins. These molecules have important roles in the establishment ofpersistent infection and chronic inflammation, which cause tissue damage.
The aim of this thesis was to study the attachment of this bacterium to human gastricepithelium, mediated by blood group antigens in both health and disease. One of thebest-characterized H. pylori adhesins is the histo-blood group antigen binding adhesin(BabA), which binds specifically to the Lewis b antigen (Leb) in the gastric mucosa.
A protective mucus layer lines the stomach. The mucosal glycosylation patterns(GPs) vary between different cell lineages, different locations along the gastrointestinal(GI) tract and different developmental stages. In addition, GPs undergo changes duringmalignant transformation. MUC5AC is a mucin molecule produced by the surfaceepithelium. Three distinctly different types of human gastrointestinal tissue were studiedby bacterial adherence analysis in situ. MUC5AC is the most important carrier of Leband the new results demonstrate that it forms major receptors for H. pylori adherence.
By analysing an H. pylori babA-deletion mutant, a novel adhesin-receptor bindingmode was found. Surprisingly, the mutant bound efficiently to both human gastricmucosa and to gastric mucosa of Leb transgenic mice. The sialylated and fucosylatedblood group antigen, sialyl-dimeric-Lewis x (sdiLex), was structurally identified as thenew receptor. A positive correlation was found between adherence of H. pylori to sialyl-Lewis x (sLex) and elevated levels of inflammation response in the human gastricmucosa. These results were supported by detailed analysis of sialylated and fucosylatedblood group antigen glycosylation patterns and, in addition, in situ bacterial adherence togastric mucosa of experimentally challenged Rhesus monkey. The cognate sialic acid-binding adhesin (SabA) was purified by the retagging technique, and the correspondingsabA-gene was identified.
H. pylori lipopolysaccharide (LPS) contains various Lewis blood group antigens suchas Lewis x (Lex) and Lewis y (Ley). Additional bacterial adherence modes, which areindependent of the BabA and/or SabA adhesins, could possibly be mediated by Lexinteractions. Adherence of a clinical isolate and its corresponding Lex mutant to humangastric mucosa with various gastric pathologies was studied in situ. The results suggestthat H. pylori LPS plays a distinct but minor role in promotion of bacterial adhesion.
Taken together, the results suggest mechanisms for continuous selection of H. pyloristrains, involving capacity to adapt to changes in the local environment such as shifts incell differentiation and associated glycosylation patterns. Adherence of H. pylori isdependent on both the BabA and the SabA adhesin. Multi-step dependent attachmentmechanisms may direct the microbes to distinct ecological niches during persistentinfections, driving the chronic inflammation processes further toward the development ofpeptic ulcer disease and/or malignant transformation.
Key words: H. pylori, BabA, adhesin, Lewis b, MUC5AC, sialyl-dimeric-Lewis x, chronicinflammation, SabA, Lewis x, LPS.
Print & Media
9
Introduction
1. A brief history of H. pylori Infection
In 1982, Barry Marshall and Robin Warren managed to culture Helicobacter pylori
from stomach biopsies. The material came from patients with gastritis, but their initial
reports suggested that gastric inflammation and peptic ulcers might be the
consequences of bacterial infections were initially met with considerable skepticism
by the medical community (Marshall & Warren, 1984).
Isolation of H. pylori (or Campylobacter pylori as it was first named) was a significant
scientific result, but still did not establish whether the bacteria were the cause of the
inflammation with which they were associated, or whether they occurred as a result
of it. However, the scientific field has rapidly moved forward during the last two
decades of H. pylori research, and today we know that merely 10% of infected
people develop serious illness such as gastritis, peptic ulcer disease and gastric
cancer, i.e. the majority (90%) of people infected with H. pylori suffer no symptoms
related to their infection In 1994, the National Institutes of Health (NIH) concluded
that there is a strong association between H. pylori and peptic ulcer disease and the
International Agency for Research on Cancer (IARC), part of the World Health
Organization (WHO), classified H. pylori as a class I carcinogen.
The pathophysiology of this infection can be understood better by considering five
central concepts; heterogeneity of strains, persistence of infection, immunological
downregulation, physiological consequences and variability in outcome. Microbial,
host and environmental factors probably all contribute to the variation in outcome
(reviewed by Blaser et al., 1996).
2. Microbiology of Helicobacter
Campylobacter, Helicobacter, Wolinella, Arcobacter and Flexispira belong to a single
phylogenetic group which is distinct from other Gram-negative bacteria. This
classification is based on 16S rRNA sequencing, DNA hybridization, genus-specific
probes, cell wall protein and lipid characterization, and also serological and
biochemical analysis (Romaniuk et al., 1987). More than 100 Helicobacter species
have now been identified, the majority being gastric organisms. H. pylori is Gram-
negative, microaerophilic, motile and has a helical (spiral or curved) cell body with 4-
6 sheathed flagellae (Solnick et al., 2001).
Print & Media
10
In 1997, the first complete genome sequences of H. pylori (laboratory strain H. pylori
26695) was released by the TIGR Institute and the second genome based on the
clinical isolate H. pylori J99 was released 2 years later (Tomb et al., 1997; Alm et al.,
1999). Together, these two genomes have provided considerable insight into H.
pylori biology. Among the 1590 predicted ORFs in the strain 26695 genome, 594
lack homology to E. coli or H. influenzae genes (Le Bouder-Langevin et al., 2002),
and 499 of the ORFs even lack obvious homology to any of the sequences in the
public genome databases.
In addition, the strain-specific genes may play distinct roles in the pathophysiology of
H. pylori infections. Both the variable gene content in the plasticity zone and strain-
specific genes outside the plasticity zone in other H. pylori strains (Aras et al., 2003)
may have an important role in this regard.
In support of this, H. pylori has a very plastic genome, reflecting its high rate of
recombination and point mutation. This plasticity promotes divergence of the
population by the development of sub-clones and presumably enhances adaptation
to host niches (Blaser et al., 2004).
3. Epidemiology
H. pylori infection is highly prevalent in many developing countries, including those of
Eastern Europe (up to 100% in some age groups). Acquisition of the organisms
during one’s lifetime is characteristic of groups living under poor hygienic conditions,
and it appears to be dependent of geographic area, age, race, and ethnicity; e.g. in
higher socioeconomic groups, the rate of colonization is lower and the prevalence
increases with age to eventually reach 40% or more. In addition, the low frequency
of H. pylori infection in Western countries compared to developing countries may be
partly explained by high consumption of antibiotics against e.g. nasopharyngeal
infection during childhood (Brown, 2000).
The means of transmission of H. pylori infection are not well known yet. No
established reservoir of the microorganism has been identified outside the human
stomach, except in non-human primates. Nevertheless, H. pylori is just one of an
increasing number of Helicobacter species being isolated from a broad range of
animals as diverse as the shrew, cheetah, tern and dolphin (Solnick et al., 2001).
It has been suggested that H. pylori is mainly transmitted directly from one human to
Print & Media
11
another (Graham, 1991). Three likely routes of transmission have been put forward:
1) fecal-oral transmission is the most important route in developing countries; 2)
person to person transmission via aspiration of H. pylori from vomit is also a
possibility, but has not been well documented (1 and 2 have been reviewed by
Brown et al., 2000); 3) H. pylori infection appears to be preferentially intrafamilial in
industrialized countries (Drumm et al., 1990; Roma-Giannikou et al., 2003). Recent
observations indicate that dental plaque or saliva is a reservoir of H. pylori and that
the bacterium can be transmitted in this way by person-to-person contact (Brown et
al., 2000; Ndip et al., 2003).
4. The Lewis and ABO blood group systems, and H. pylori infection
Carbohydrate antigens are not primary gene products, but are instead the enzymatic
reaction products of glycosyltransferase enzymes. The ABO blood group antigens
are complex carbohydrate structures on glycoproteins and glycolipids expressed at
the surface of red blood cells (RBCs or erythrocytes) and in the entire gastro-
intestinal epithelium (reviewed by Clausen et al., 1989). In addition, they are present
in secretions, as mucin glycans. In a review, Marionneau and co-workers concluded
that the highly polymorphic genes of each family provide intraspecies diversity that
allows coping with diverse and rapidly evolving pathogens (Marionneau et al., 2001).
Immunodominant structures of blood group A and B antigens, GalNAc�1.3(Fuc
�1.2)Gal- and Gal�1.3(Fuc�1.2)Gal-, respectively, are synthesized by a series of
reactions. The A and B transferases encoded by functional alleles catalyze the last
step of the synthesis, while the O allele is non-functional. Thus, the acceptor
substrate, (H antigen: Fuc�1.2Gal-) remains without further modification (Larsen et
al., 1990; reviewed by Clausen et al., 1989).
FUT III (the “Lewis type” enzyme) has the broadest �3FUT substrate specificity
since it can incorporate fucose in either �1.3- linkage to Gal(�1.4)GlcNAc (to make
Lewis x (Lex) or �1.4- linkage to Gal(�1.3)GlcNAc to make Lewis a (Lea).
The Lewis y (Ley) and Lewis b (Leb) antigens constitute the di-fucosylated
equivalents (Kukowska-Latallo et al., 1990; Henry et al., 1995; Clausen et al., 1989).
The Lewis blood groups share some genetic links with ABO i.e. H antigen can be
modified by ABO and Lewis transferases. The secretor gene (FUTII) transferase
controls the expression of the H antigen in the secretory compartments (Kelly et al.,
Print & Media
12
1995). This gene is active in 80% of the European population and will confer
expression of ABH antigens in secretions such as saliva, sweat, tears, semen and
serum (Henry et al., 1995). There is also the Le a+b+ phenotype in Asian
(Broadberry et al., 1991) and Polynesian populations (Henry et al., 1993). This
phenotype relates to poor expression of salivary ABH secretions and is known as
weak or partial-secretor phenotype (Henry et al., 1993; Henry et al., 1996). In several
studies, individuals of blood group O (Rotter, 1983) and non-secretor (Sipponen et
al., 1989) phenotypes are more prone to peptic ulcer disease.
Sialyl Lewis x, a sialylated (charged) blood group antigen, is of major importance.
Cancer cells are usually heavily sialylated (Reglero et al., 1993; Kageshita et al.,
1995). Among enteropathogenic bacteria, H. pylori is recognized to bind the H1 and
Leb antigen (Borén et al., 1993) in both healthy and inflamed gastric tissues (Paper
II). An adhesin from a strain of Staphylococcus aureus that binds to Lea has been
described (Saadi et al., 1994). Since Lea is present in large amounts on epithelial
cells from non-secretors, but in more limited amounts on those from secretors, the
former individuals are expected to be more sensitive to this strain of bacteria than
the secretor. Stapleton and colleagues reported that uropathogenic E. coli recur two
to three times more often in non-secretor women than secretors (Stapleton et al.,
1992). Similarly, the prevalence of dental caries is higher among non-secretors than
secretors (Arneberg et al., 1976). Later, Holbrook et al. (1989) confirmed the high
prevalence of dental caries in Icelanders, who have one of the highest proportions of
non-secretors recorded in European countries. This may be related to the lack of
ABH blood-group substances in saliva, which are suggested to interfere with
adherence of Streptococcus mutans to the dental surface (Holbrook et al., 1989).
Conversely, pathogens that bind to H-antigens present on epithelial cells are
expected to preferentially infect secretor individuals. Indeed, Campylobacter jejuni
strains have been shown to attach to H type 2 (i.e., the H antigen, which is the
monofucosylated carbohydrate structure of the blood group O phenotype in the ABO
system, based on the lacto series type-2 core chains (Gal�1.4GlcNAc). This
attachment could be inhibited by human milk oligosaccharides such as fucosyl-
lactose (Newburg, 2000).
It should be most useful for a population to present a diversity of blood group
antigens so that a single virulent microorganism cannot infect all individuals within
the population. This would be favorable for the population in the long run, although it
Print & Media
13
might be harmful for a fraction of individuals facing one given type of infectious
agent.
5. Biodiversity
Here are three general kinds of biodiversity discussed: habitat diversity, genetic
diversity, and species diversity. The survival of each is linked to the safety of the
other two, and together they comprise the resources of bacterial ecosystems.
5.1 Habitat diversity
Here, habitat diversity refers to the variety of organs and/or pathological alterations
caused by bacterial infection; i.e. Helicobacter species exist in the stomach for H.
pylori (Marshall & Warren, 1984), the intestinal tract for H. canadensis (Fox et al.,
2000), the liver for. H. hepaticus (Nilsson et al., 2001), and the gall bladder for H.
bilis (Chen et al., 2003).
The natural habitat of H. pylori appears to be the gastric mucus and the mucus-
producing epithelium (Shimizu et al., 1996). In the duodenum of H. pylori infected
patients, the bacterium is always found closely associated with gastric metaplastic
cells (Wyatt et al., 1990; Gisbert et al., 2000), which is a precancerous condition and
relatively common in the upper GI tract. The formation of gastric-type epithelium in
the duodenum seems to be related to increased gastric acid output (Harris et al.,
1996). This new habitat could be essential for colonization of H. pylori when gastric
changes such as chronic active atrophic gastritis or cancer induced by bacteria take
place in the natural habitat.
5.2 Genetic diversity and species diversity
The genetic diversity within a species is mainly the variety of populations that
comprise it. The more variation within an H. pylori population in an infected
individual, the better the chance that some of the phenotypes will have an allelic
variant that is suited to the changing environment in the stomach, and that it will
produce offspring with variants that will in turn reproduce and continue the population
into subsequent generations (illustrated in Figure 1).
Compared to most other organisms in the human biosphere (Suerbaum et al., 1998),
H. pylori is highly diverse. The identification of genetically divergent sub-clones
Print & Media
14
within individual hosts indicates that H. pylori diversification continues during its
decades-long colonization of the host (Israel et al., 2001). This genetic diversity is
generated through multiple mechanisms, including spontaneous point mutations,
recombination with other bacterial cells, and intragenomic rearrangements involving
mobile genetic elements (Woolhouse et al., 2001) or repetitive DNA sequences
(Aras et al., 2001).
The babA and babB genes code for a family of H. pylori outer membrane proteins
(OMPs), which have appreciable N and C-terminal identity. Recently, Pride et al.,
(2002) analyzed the nucleotide sequences of babA and babB and showed that
geographic origin was the major determinant of phylogenetic relationships (Pride et
al., 2002).
Simultaneous colonization of the human stomach with more than one strain of H.
pylori can be detected in about 5-10% of patients in the United States (Fujimoto et
al., 1994), and this may occur more commonly in other populations (Jorgensen et al.,
1996). Such mixed infections, even if transient, provide an opportunity for genetic
exchange between strains. Kersulyte et al. (1999) could show different genetic
exchanges between single cell clones from a patient who was naturally infected with
two different H. pylori strains. One exchange resulted in deletion of the cag-PAI,
while other recombination involved a region encoding outer membrane proteins that
could be involved in adherence activities. Genetic exchange may play an important
role in the biology of H. pylori by generating new genotypes much more rapidly than
is possible by mutation alone, thereby allowing genera of pathogenic bacteria to
adapt to other organs (Cover et al., 2001).
Print & Media
15
Figure 1. Schematic drawing of stages hypothetically involved in biofilm formation byH. pylori. A protective and continuous mucus layer lines the stomach. Mucins are afamily of large, heavily glycosylated proteins. Bicarbonate is secreted from theepithelium into the mucus layer, where it neutralizes acid that is back-diffused fromthe lumen of the stomach and forms a pH gradient, with a higher pH at the epithelialcell surface (Ross et al., 1981; Schade et al., 1994). In the healthy stomach,MUC5AC is produced in the foveolar epithelium and MUC6 in the glands. The biofilmis colonized by a mixture of H. pylori bacterial cells with different adhesioncapabilities, although one phenotype may predominate. Here, planktonic H. pyloricells adhere to the MUC5AC mediated by Leb. Further growth, cell replication andphase/antigenic variations lead to occurrence of different phenotypes with differentadhesion properties, which contributes to establishment of the biofilm.
Print & Media
16
5.3 Tropism and patho-adaptation
It has long been recognized that a particular bacterial species will tend to adhere to
colonize, and possibly induce pathology in only one preferred type of tissue. This
phenomenon is referred to as tissue tropism. This kind of tropism in associated with
the ability of the bacterium to rupture mucosal barriers and invade the host, and
distinguishes pathogens from commensals. Two major selective factors, adhesion
and environment, are essential for colonization of a particular type of tissue. Bacteria
use a large number of surface components such as lipopolysaccharide (LPS),
lipoteichoic acid, outer membrane proteins and adhesive structures such as fibrils,
pili and fimbriae for interaction with specific receptors on the host cell surface
(Hultgren et al., 1993; Sauer et al., 2000).
The pathogenic microorganisms affect the host cells, which results in alteration in
cell morphology, glycosylation pattern and cytokine production, i.e. environmental
changes to counteract the colonization of the pathogenic organism. In response to
these environmental changes, pathogenic organisms might adapt functionally by
bacterial evolution towards virulence is the allelic micro-variation of existing genes,
which has been called patho-adaptation (Sokurenko et al., 1999).
As a case in point, H. pylori is able to attach to the regularly fucosylated Lewis
(i.e. sLex and sLea) during the course of infection. Such differences display the
responsive adjustment of H. pylori to the environment.
Furthermore, distribution of ABO blood groups varies between different populations.
Recently, BabA adhesin molecules from H. pylori strains isolated from different parts
of the world were found to bind the A, B, and O blood group antigens, with the
exception of H. pylori strain from the South American Amerindian population. Here
H. pylori strains bind preferentially the blood group O antigens, which is of particular
interest since this population is of blood group O phenotype only (Aspholm-Hurtig et
al. submitted). The BabA adhesin produced by most H. pylori strains is a rather
conserved protein. Lack of substantial variation could possible relate to the invariant
histo-blood group antigen receptors recognized by H. pylori. However, minor
aminoacid changes have been described to result in drastic shifts in adhesive
properties such as binding of E. coli mediated by FimH to 3-mannose in the large
Print & Media
elimination of some genes or activation of silent genes. One of the mechanisms in
antigens (i.e. Leb, ALeb, BLeb) during health, and to the sialylated Lewis antigens
17
intestine as primary niche, and to mono-mannose in the urinary bladder as
secondary niche. Such changes in tissue tropism could be correlated with decreased
or increased affinity for binding to sugars (Sokurenko, et al., 1994). This can drive
the adaptive evolution of pathogenic organisms during their spread and growth in
diverse host environments.
5.4 Biofilm formation
Does H. pylori exist as isolated, single cells that are independent of their neighbors –
or as a community of cells living together in an ordered structure known as a biofilm?
In general, three stages can be recognized in biofilm formation on a water-insoluble
surface: adhesion, colonization and growth. The first phase involves adhesion of
planktonic bacteria to the surface. If an adequate supply of nutrients is available and
environmental factors are favorable, adherent bacteria may then colonize. The
community grows and develops a characteristic structure, the nature of which will
depend on many factors including the species involved, the availability of nutrients,
pH, and the hydrodynamic and mechanical shear forces present (Figure 1). It has
been established that by adaptive mechanisms, H. pylori changes its gene
expression in response to acid exposure (Wen et al., 2003). In addition, H. pylori
cells differ in length, shape, and the stage at which they convert from common shape
to spherical (with different adhesion capacity and phenotypes), which may affect the
bacterial survival in the gastric mucus (Enroth et al., 1999). Consequently, the use of
a thick protective and stabilizing biofilm by H. pylori may be important in enhancing
resistance to host defense factors and antibiotics, and in maintaining micro-
environmental pH homeostasis, both of which facilitate its growth and survival in vivo
(Stark et al., 1999).
6. Virulence factors and persistence of infection
Our recent results on the interactions between gastric epithelial cells and H. pylori
have highlighted the complex cell-to-cell communication which underlies the
pathology produced by this microorganism (paper II). Bacteria-host interactions are
clearly two-way processes, with factors from eukaryotic cells being able to interact
and modulate the activity of bacteria and vice versa.
Detailed information about how bacteria redirect and manipulate the cytokine
Print & Media
18
networks that result in tissue pathology is essential for a better understanding of the
development of infectious disease (Figure 2).
Figure 2. This figure shows one proposed mechanism involving chemokines (e.g. IL-8) for chemo-attracting PMNs to the lumen of the H. pylori infected stomach. Itshows also the adherence of H. pylori mediated by binding to Leb and sLex whichare expressed in the surface of gastric epithelial cells by the BabA and SabAadhesins (i.e. signaling between H. pylori, mucosal epithelial cells and underlyingimmune systems) in conditions of both health and inflammation. Attachment of H.pylori to gastric epithelial cells activates the type IV secretion system, which resultsin the injection of effector proteins into host cells. The effectors are thought toupregulate the activity of Nuclear Factor NF-�B, which, together with AP-I, leads toinduction of IL-8 expression. This triggers transduction pathways and pedestalformation as a result of cytoskeletal rearrangements. The leukocytes areprogrammed to reveal the healing process, but they can also enhance inflammation,which can proceed to tissue damage. The rolling and attaching phase in thecapillaries is mediated by the selectin family, the E- and L-selectins, and theirsialylated Lewis type of ligands. H. pylori adhesion stimulates the release of the IL-8and induces expression of the intercellular adhesion molecule ICAM-1, whichfacilitates the migration of PMN cross the gastric epithelium.
Print & Media
19
6.1 Adhesins
A wide variety of molecules present on adherent structures of bacteria can function
as adhesins. In bacterial-eukaryotic interfaces, there are apparently general systems
for recognition of certain microbial cell surface carbohydrate characteristics (Wilson
et al., 2002; chapter 7).
The normal habitat of E. coli is the gastrointestinal tract, but some uropathogenic
strains are capable of causing UTIs (urinary tract infections). A large number of
potential adhesins have been identified in E. coli, such as different classes of P-
fimbriae, which bind to the Gal�(1.4)-Gal residues present in glycolipids in kidney
tissue (Strömberg et al., 1991). Like other microorganisms, H. pylori requires
adhesive molecules for colonization and persistence. H. pylori has a wide range of
adhesion properties and has been suggested to bind to many different
carbohydrates, (Karlsson 1998; reviewed by Evans, 2000) mediated by various
bacterial components. The Leb and sLex antigens binding adhesins, BabA and
SabA, respectively, are the best described (Ilver et al., 1998 and paper II).
Hemagglutination of RBCs by bacteria has been used to study bacterial binding
specificities and to identify the cognate bacterial adhesins. The H. pylori sialic acid-
dependent hemagglutination was a subject for researchers for more than a decade
but recently, this sialic acid-dependent binding activity has been shown to be
mediated by SabA, since the corresponding sabA deletion mutant lacks all
hemagglutination properties (Olfat et al., manuscript).
6.1.1 Consequences of bacterial adhesion
It is becoming increasingly obvious that the act of adhering to the host cells induces
intense changes in the adherent organism, regardless of the origin of the host cells.
It is also known that adhesion of an organism to a host receptor greatly affects its
rate of growth, carbohydrate utilization, protein synthesis and energy generation, as
well as its cell wall composition and production of adhesive molecules. The host cell
and its adherent bacterium exchange signals (i.e. they engage in molecular “cross-
talk”), which brings about changes in both cell types (Diagram 1) (Soto et al., 1999;
Tesh, 1998).
Print & Media
20
Diagram 1. Adhesion of Helicobacter pylori to gastric epithelial cells induces amultitude of changes to the host cell, the most common of which are shown in thediagram.
6.1.1.1 Effect on the bacterial cells
There are a number of reasons why the bacterial cell has to adjust its phenotype
once it has adhered to a cell surface. Prior to any interaction with its host, H. pylori is
probably exposed to different environmental conditions (e.g. different pH, high
turnover, shear force, osmolality, nutrient concentration). Then, adherence is
required for microbial survival in this hostile environment. Attachment of H. pylori to
the gastric mucosa (e.g. mediated by Leb) activates the type IV secretion system
which results in the translocation of CagA protein into the host cells (Odenbreit et al.,
2000) and triggers inflammation. Having reached an inflamed epithelial surface with
the new carbohydrate structure as a ligand (which may not necessarily be its
Print & Media
21
ultimate destination), H. pylori will then have to adapt to this new environment and
this will require the upregulation or suppression of various genes such as SabA.
Such changes will certainly be induced in response to alterations in a number of
environmental factors, but some are recognized to be triggered by the adhesion
process itself. Recently, changes in H. pylori gene expression induced by bacterial
adhesion to AGS (cultured gastric adenocarcinoma) cells were reported. Several
genes, and those for two outer membrane proteins in particular, were upregulated
(Kim et al., 2004).
Attachment to the host cell may be only the first step in a sequence of events that
could involve invasion of the cell. Adhesion of H. pylori mediated by SabA may be
used as a signal for the microbe to begin synthesis of those molecules essential to
initiate invasion of the host cell. There have been many reports suggesting that
bacteria do respond in this way to contact with the host cell. As an example of this,
similar crosstalk has been described for Porphyromonas gingivalis adherence to
epithelial cells, which induces the secretion of a multitude of proteins and
downregulation of the production of extracellular proteases (Park et al., 1998).
6.1.1.2 Effect on the host cells
A number of bacterial species capable of inducing pathology in humans also adhere
to epithelial cells without actually inducing any apparent changes in these cells.
However, although the adhesion process itself may not affect the structure or
function of the host cell, microorganisms also produce toxins – enzymes that may
ultimately damage host cells.
H. pylori induces the morphological alterations in gastric epithelial cells (Segal et al.,
1999), (Figure 2). H. pylori-related chronic inflammation in gastric tissue has also
been reported to modulate the glycosylation patterns of epithelial cells. Ota et al.,
(1998), have shown increased levels of sLea antigen in response to H. pylori
infection and recently, we have shown that sLex was upregulated in gastric epithelial
cells due to H. pylori infection (paper II). In contrast, normal human gastric epithelial
cells are essentially devoid of such sialylated glycoconjugates (Filipe, 1979; Madrid
et al., 1990). In addition, recent studies have indicated that adherence of H. pylori
induces cell proliferation and apoptosis during the early phase of chronic
inflammation of the gastric mucosa (Ebert et al., 2002).
Print & Media
22
A characteristic feature of any bacterial infection is the migration of PMNs towards
the site colonized by the infecting organism. Binding of H. pylori to the gastric
epithelial cells induces the expression and secretion of IL-8 (Crabtree, 1996). In
response to a chemotactic gradient from the site of infection, PMNs first adhere to,
and then traverse, the vascular endothelium (McCormick et al., 1998; Bevilacqua et
al., 1991). This process involves the interaction between E-selectin, intercellular
adhesion molecule (ICAM) 1 and 2 on the surface of the endothelium cell, and sLex
and �2-integrins on the PMN surface (Slattery et al., 2003) (Figure 2). PMNs are rich
in various sialylated glycoconjugates, which H. pylori can bind to. An important
consequence would be the bacterial interaction with, and activation of PMNs
(Rautelin et al., 1993). Very recently, Petersson et al. (in manuscript) have
demonstrated that SabA is essential for attachment and activation of human PMNs.
Nutrients released by the inflamed and damaged cells might be used by H. pylori as
an energy source (Blaser, 1992; 1996). See Diagram 1 for how H. pylori adhesion
affects the host cells.
6.2 Invasion of host cells
A characteristic feature of H. pylori-induced inflammation is massive attraction of
phagocytes (particularly PMNs) to the site of infection. This can be achieved by the
production of H. pylori neutrophil-activating protein (HP-NAP). The latter was found
to promote the adhesion of PMNs to endothelial cells by upregulating adhesion
receptors of the 2-integrin family (Satin et al., 2000). Satin and colleagues showed
that HP-NAP stimulates (NADPH) oxidase assembly and production of reactive
oxygen species (ROS). Also, as PMNs consistently outnumber macrophages in H.
pylori infected stomach, it induces a state of chronic acute inflammation. Previous
reports (Björkholm et al., 2000; Amieva et al., 2002; Kwok et al., 2002) have
suggested that H. pylori is capable of invading epithelial cells in the gastric mucosa;
however, the role of invasion in H. pylori pathogenesis remains unclear.
The following mechanisms have been postulated to explain how H. pylori evades
phagoctosis:
Print & Media
23
1- Strong binding between H. pylori and phagocytes correlates with a rise in the
level of urease on the surface of H. pylori, thus retarding phagocytosis and
strong respiratory burst in the phagocytes (Telford, 1997).
2- Catalase, alkyl hydroperoxide reductase, and factors (cag pathogenicity island
type IV secretion apparatus) that are unique to type I strains allow bacteria to
resist phagocytic killing (Allen, 2001).
3- Delayed phagocytosis is linked to intracellular survival, since type I H. pylori
persists inside macrophages within a novel vacuole called the megasome
(Allen et al., 2000).
6.3 H. pylori enzymes
H. pylori produces catalase and urease enzymes. Urease (which converts urea to
ammonia) is an abundantly produced enzyme. H. pylori has developed a unique
mechanism to control urease-dependent pH buffering. The urea channels (UreI)
present in the inner membrane are opened at pH below 6.5, to allow for delivery of
urea to intracellular urease. The NH3 produced can then buffer the periplasm
(Weeks et al., 2000; Bury-Mone et al., 2001). Urease activity is therefore essential
for maintenance of cytosolic pH levels compatible with efficient metabolism and
survival of H. pylori in the acidic environment of the stomach.
6.4 CagA and VacA
Although most strains possess the vacA gene, some strains also exhibit another
virulence marker called cagA (cytotoxin-associated gene) and are defined as Type I
strains (Xiang et al., 1995). Those that are vacA positive but cagA negative are
considered low-virulence strains (Type II strains). Fifty to sixty per cent of isolates
carry the cagA locus, which acts as a marker for the cag pathogenicity island (an
assembly of some 40 genes that are linked to virulence of the organism, coding to a
large degree for a type IV bacterial secretion system) which is involved in the
stimulation of pro-inflammatory cytokines from the gastric epithelium.
A number of the proteins encoded by the cagpai type IV secretion system are similar
to those used by Bordetella pertussis to secrete pertussis toxin (Farizo et al., 2000).
The type IV system and the phosphorylated effector molecule CagA confer
cytoskeletal rearrangments and pronounced inflammatory responses through the
Print & Media
24
induction of IL-8 expression and, in some cases, cancer (Stein et al., 2000; Censini
et al., 2001; Nilsson et al., 2003). Since cagA (-) cells are relatively unimpaired in IL-
8 induction (Nilsson et al., 2003), other cagpai-encoded effectors may also be
translocated by the type IV secretion system (Odenbreit et al., 2000). CagA has
been shown to activate mitogen-activated protein (MAP) kinase cascades, leading to
the induction of expression of c-fos and phophorylation of c-Jun, which together
comprise the transcription factor AP-I (Stein et al., 2000) (Figure 2).
Unlike the cag pathogenicity island, all H. pylori strains possess the vacA gene and
approximately 50% of strains express the VacA protein, which causes structural and
functional alterations in epithelial cells (Leunk et al., 1988; Telford et al., 1994).
VacA binds to multiple epithelial cell-surface receptors, and partly remains
associated to the bacterial outer membrane (reviewed in Papini et al., 2001). Upon
bacterial contact with host cells, toxin clusters are transferred directly from the
bacterial surface to the host cell surface at the bacteria-cell interface, followed by
uptake and intoxication (Ilver et al., 2004). VacA is considered to be an important
virulence factor that induces vacuole formation in host cells, stimulates epithelial-cell
apoptosis and also plays a role in the pathogenesis of peptic ulcer disease (Boquet
et al., 2003).
6.5 H. pylori LPS
The LPSs of Gram-negative bacteria are among the dominant surface features in
terms of their abilities to induce a host response, which stimulates macrophages to
produce and release tumor necrosis factor (TNF-�), (Holler, 2002). However, the
LPS of H. pylori have far lower biological activity than those of E. coli or Salmonella
(Muotiala et al., 1992). One characteristic peculiar to H. pylori is that almost all
strains express Lewis blood group antigens (Lea, Leb, Lex, sialyl Lex, Ley, H-type I,
i-antigen and blood group A antigens) in their surface-expressed LPS O-antigen
(reviewed by Monteiro et al., 2001). H. pylori has been suggested to induce gastric
autoimmune responses by inducing anti-Lex antibodies that bind to gastric
epithelium (Appelmelk et al., 1996). H. pylori LPS has also been shown to mediate
adherence of the bacterium to laminin in the basement membrane, with sialic acid
binding specificity (Valkonen et al., 1994; Valkonen, 1997), and to gastric epithelium
(Edwards et al., 2000). However, deletion of glycosyltransferase genes, required for
Print & Media
25
LPS biosynthesis, yields Lewis antigen-negative mutants that colonize well in
experimental infections.
Galustian and colleagues have suggested that H. pylori LPSs contribute to the
chronic inflammation through interactions with leukocyte-endothelial adhesion
molecules of the host (i.e. E- and P-selectins). They have also suggested that H.
pylori isolates from patients with chronic gastritis, duodenal ulcer and gastric cancer
interact with E- and L-selectins (Galustian et al., 2003).
7. Diseases associated with H. pylori infection
The healthy state is the normal condition in an organ, but sometimes the organ
becomes diseased, either due to physiology dysfunction or to infections. In classical
histopathology, diagnosis is mostly based on morphological and structural features
of cells and tissues, and to a lesser extent on the biochemistry of the cell (Rosai et
al., 1989). The histology of upper gastrointestinal tract is illustrated in Figure 3.
7.1 Gastritis or gastric inflammation
Gastritis means inflammation of the stomach. It means that white blood cells move
into the wall of the stomach as a response to injury. Gastritis does not mean that
there is an ulcer or cancer. It is simply inflammation, either acute or chronic.
The mucosal surfaces of the body are programmed to signal to the immune system
to induce the chemo-attraction of leukocytes. This has been termed the “the watch-
dog” function of the mucosal epithelium. It has been shown that direct contact of H.
pylori with the epithelial cell is necessary for optimal IL-8 production (Rieder et al.,
1997). Recently, Olfat et al. (2002) have developed an in vivo model system, IVEC
(in vitro explant culture), which closely mimics the natural environment of human
stomach, and they could show that adhesion of H. pylori mediated by Leb to the
gastric explants increases IL-8 production. Here, the microbial components activate
Toll-like receptors (TLRs), thereby leading to NF-�B-dependent transcription and
resulting in production of proinflammatory cytokines such as IL-8 (Su et al., 2003)
(illustrated in Figure 2).
Print & Media
26
Figure 3. The anatomy/histology of the upper gastrointestinal tract. Celldifferentiation is diagramed in the mucous pit-gland unit of the pyloric antrum andcorpus. Esophagus: It is covered by non-keratinized stratified squamous epithelium.In the sub-mucosa there are groups of small mucus-secreting glands, theesophageal glands. In the lamina propria of the region near the stomach there aregroups of glands, the esophageal cardiac glands, that also secrete mucus.Ventricle: The stomach is anatomically divided into four regions; fundus, cardia,corpus (or body), and pyloric antrum. In general, the gastric mucosa consists of asurface epithelium that invaginates to varying extents into the lamina propria, forminggastric pits. The gastric pits are branched, tubular glands that are characteristic ofeach region of the stomach, and composed of isthmus, neck and base regions. Thepluripotential stem cells are anchored in the isthmus region and give rise to severalcommitted precursors, which undergo migration-associated differentiation to replacethe degenerating mucous, parietal (secreting hydrochloric acid), zymogenic(secreting pepsinogen), neuroendocrine cells such as histamine releasingenterochromaffin-like (ECL) cells and D-cells which secrete somatostatin. In theseepithelia, over 90% of human cancers occur via a multi-step process which starts byalteration in the proliferation/differentiation program of the stem cells (Karam, 1999).The pre-pit cell precursor (P1), pre-parietal cell precursor (P2) and pre-neck cellprecursor (P3) evolve into corpus pit, parietal and zymogenic cell lineages,respectively (Karam, 1999). The antrum-type mucosa is characterized by morebranched, predominantly mucus glycoprotein secreting glands with neuroendocrineG-cells secreting gastrin and D-cells secreting somatostatin. Duodenum: Theepithelium of the villi is continuous with that of the glands. In the intestinal glands,different types of cells such as stem cells, some absorptive cells, and goblet cells arefound.
Print & Media
27
7.2 Gastric and duodenal ulcers
H. pylori has been found to be present in 85-100% of patients with duodenal ulcers
and in 70-90% of patients with gastric ulcers (Staat et al., 1996; Drumm, 1990). This
has strongly suggested that H. pylori plays a significant role in the development of
duodenal and gastric ulcer disease. The parietal cell mass of the host (Testino et al.,
1999), and the effect of the inflammatory response on acid secretion by damaging
the stomach cells (Kersulyte et al., 1999), play an important role in the outcome of
disease. Studies of pathological tissues have shown abnormalities in the thickness of
the mucus layer among people with H. pylori infection (Morgenstern et al., 2001).
Gastric ulcers usually lead to pan-gastritis and low acid secretion, in which H. pylori
colonizes the corpus (Björkholm et al., 2003) and may result in atrophy of parietal
cells. When the patient’s immune response succeeds in suppressing the bacteria,
acid secretion returns to normal and acid digests the damaged cells, leading to
chronic gastritis and ulcer.
Duodenal ulcer is more common than gastric ulcer. The rate of acid secretion is
higher among patients with duodenal ulcer than in healthy people. The duodenum is
not protected as well as the stomach against acid and the local stress can cause
cells in the duodenum to undergo morphological changes named gastric metaplasia.
The mechanisms of targeted tissue tropism for H. pylori (as described above) will
promote transfer of the H. pylori infection from its natural niche in the antrum to the
islands of gastric metaplasia, and promote a much higher risk of peptic ulcer
disease.
7.3 Gastric adenocarcinoma, a silent killer
Gastric cancer is cancer of the stomach. The most common type of stomach cancer
is “gastric adenocarcinoma”, or cancer of the glandular tissue of the stomach. During
the last century, the incidence of gastric cancer has decreased in Western countries,
in contrast to the developing countries (Parkin, 1988). Gastric carcinogenesis is a
multistep and multifactorial process (Correa, 1992), but bacterial infection is one of
the most important risk factors for development of cancer (Correa, 2003). Based on
the WHO’s conclusion, H. pylori is a definite carcinogen leading to adenocarcinoma
of the stomach (Logan et al., 1994). Interestingly, the high rate of atrophic gastritis
and gastric cancer is accompanied with a low rate of duodenal ulcer (Blaser et al.,
Print & Media
28
1995), presumably because of the impaired ability to secrete gastric acid after the
development of atrophy. In contrast, there appear to be factors associated with
duodenal ulcer disease that protect against gastric cancer (Hansson et al., 1996).
Host genetic factors also appear to influence the chances of developing gastric
cancer following infection with H. pylori. Pro-inflammatory polymorphisms within the
interleukin IL-1�, IL-10 and TNF-� genes are associated with a higher risk of gastric
carcinoma in H. pylori-infected individuals (El-Omar et al., 2000; Machado et al.,
2003).
7.4 Barrett’s esophagus
Barrett’s esophagus (BE) is the condition in which columnar epithelium replaces the
squamous epithelium. The condition develops when gastroesophageal reflux
disease damages the squamous esophageal mucosa, through a metaplastic process
in which columnar cells replace squamous epithelium. The abnormal columnar
epithelium that characterizes Barrett’s esophagus is an incomplete form of intestinal
metaplasia (called specialized intestinal metaplasia) that predisposes patients to
adenocarcinoma (Spechler et al., 2002). Inflammation of gastric cardiac mucosa
decreases in prevalence from controls to patients with GERD (Wieczorek et al.,
2003), and H. pylori and cagA-positive strains have been suggested to protect
against the development of Barrett's esophagus (Ackermark et al., 2003). It is
important to distinguish BE from intestinal metaplasia related to carditis, because
these conditions have a different natural history, risk of malignancy, and treatment.
Glickman et al. (2003) have reported recently that MUC1 and MUC6 expression in
BE is distinct from that of the cardia and antrum with intestinal metaplasia. In
addition, Barrett´s esophagus is the most important risk factor for the development of
esophageal adenocarcinoma. The risk of developing esophageal adenocarcinoma
from BE is estimated to be 0.5% per patient year.
Print & Media
29
Aims
The main goal of my thesis has been to study H. pylori adhesion to host cells and in
particular how bacterial-ligand interactions influence the development of disease:
- To investigate the role of MUC5AC as a receptor for H. pylori in different types
of tissues in conditions of both health and disease from the gastrointestinal
tract.
- To study the H. pylori adhesion properties that are independent of BabA.
- To characterize the relationship between sLex expression in the gastric
surface epithelium and persistent infection by H. pylori in molecular terms.
- To identify and characterize the novel adhesion molecule sialic acid-binding
adhesin (SabA) from H. pylori and its corresponding receptor (sdiLex).
- To investigate the molecular impact of H. pylori infection on the glycosylation
patterns in Rhesus monkey gastric mucosa, which is a primate model with the
most similar expression pattern of blood group antigens to that of humans. In
particular, we aimed to study the shifts in glycosylation pattern due to H. pylori
infection in a time-dependent manner.
- To examine the role of H. pylori LPS Lex structure with regard to adhesion
properties in various gastric tissues, and with respect to pathological
background.
Print & Media
30
Results and Discussion
8. The role of MUC5AC in H. pylori adherence
In paper I, we tested the hypothesis that Leb antigen presented on MUC5AC
molecules functions as a H. pylori receptor. Generally, MUC5AC is expressed in the
gastric antrum, but it is produced by the metaplastic cells of gastric type in both
duodenum and Barrett’s esophagus, which indicates distinct and specific function of
this mucin. Previously, Van der Brink and co-workers demonstrated, that H. pylori in
the antrum of infected patients co-localizes specifically with MUC5AC and MUC5AC-
producing cells (Van den Brink et al., 2000). Here, a series of H. pylori strains were
overlaid on tissue sections of Barrett’s esophagus, normal stomach and duodenum
with gastric metaplasia. We could demonstrate that H. pylori co-localized very well
with both MUC5AC and Leb antigen. Thus, the results suggest that MUC5AC
mediates H. pylori adhesion and that MUC5AC is an important site for bacterial
infection and/or colonization.
More than 99% of H pylori positive bacterial cells were associated with either
secreted MUC5AC or the apical domain of MUC5AC producing cells (Van den Brink
et al., 2000). Binding to secreted MUC5AC occurs both at neutral and acidic pH.
BabA-dependent binding at neutral pH and the charge-dependent binding at acidic
pH allows these microbes to bind mucins over a broad pH range (Lindén et al.,
submitted). It seems likely that the adherence of H. pylori to the host Leb structures
mediated by BabA-adhesin effectively facilitates infection, but is not an absolute
necessity for colonization. Thus, the composition of diverse carbohydrate structures
in MUC5AC serves as receptors for different H. pylori phenotypes at different pH,
which probably makes it a suitable location for H. pylori to form the biofilm.
Therefore, H. pylori with different adhesive properties may only be able to survive in
biofilm (Figure 1). This is probably one of the reasons why successful eradication of
H. pylori necessitates consumption of a combination consisting of different antibiotics
together with acid suppressing drugs, where the pH gradient in the biofilm will be
broken and make the microbe more susceptible to antibiotics.
9. Gastric mucosa sialylation and inflammation
In paper II, there is evidence for the participation of sialylated structures in the
recognition and adherence of H. pylori to the epithelial cells in an inflamed gastric
Print & Media
31
tissue. In addition, we found positive correlations for H. pylori binding, inflammatory
cell infiltrations and expression of sLex antigen levels in the gastric mucosa.
Biological molecules containing sialic acid in the peripheral position of their
oligosaccharide chain are glycolipids and glycoproteins, as well as mucin molecules.
They have a structural function and the components are linked to cell surface
receptors (e.g. integrins) (Pochec et al., 2003). They also affect a number of cellular
activities including migration, proliferation and differentiation (Rosner 1982; Ertl et al.,
1992). Certain bacteria have evolved to take advantage of the host cell’s
requirement for such receptors and have produced adhesions that can bind to these
receptors, thereby usurping their natural function.
Chronic infection of the gastric mucosa by H. pylori results in an intense
inflammatory response. One associated host response is a chronic hyper-
proliferative state, in which there is increased cell turnover and also increased
apoptosis of the gastric epithelial cells (Chiou et al., 2001). Shirin and colleagues
explained the effects of H. pylori infection on cell cycle progression (Shirin et al.
2001) and the expression of cell cycle regulatory proteins. They concluded that
overexpression of the tumor suppressor gene 16 in gastric epithelial cells of H. pylori
infected patients was associated with an increase in apoptosis (Shirin et al., 2003).
It has been shown that sLex has an important role in inflammation processes, and
there also appears to be a correlation between cancer and degree of cell sialylation
(Renkonen et al., 1997; Alper, 2001). Overexpression of sLea and sLex may
correspond to 1) increased expression of a �2.3-sialyl-transferases, competing with
the fucosyltransferases for the same substrate, the terminal Gal�1.3GlcNAc and
Gal�1.4GlcNAc chains (type 1 and 2 chains), and 2) increased expression of �1.3-
and �1.4-fucosyltransferase activity.
Several histopathological features such as hyperproliferation of epithelial stem cells,
loss of parietal and zymogenic cells common in the stomach of patients suffering
from chronic atrophic gastritis are associated with elevated production of sialylated
glycans (Lipkin et al., 1985). Such alterations in the cell differentiation program is
very similar to that in transgenic mice lacking parietal cells, recently described by
Syder et al., (1999). Interestingly, expression of sialylated glycoconjugates in the
parietal cell ablated mice does not require the microflora, i.e. no H. pylori
Print & Media
32
infection/colonization, but is rather the result of cell differentiation in the parietal cell
ablated animal model (Syder et al., 1999).
10. Identification of inflammation-associated H. pylori adhesiv activities
We also demonstrated in adherence experiments in situ (paper II) that a babA-
mutant devoid of Lewis b antigen binding properties activated an additional
complementary adhesin-receptor interaction in human gastric mucosa. A novel
sialylated and fucosylated carbohydrate receptor, the sialyl dimeric Lewis x
glycosphingolipid (sdiLex), and the corresponding sialic acid binding adhesin (SabA)
were identified. In addition, a correlation between SabA-mediated bacterial binding
and the level of gastric inflammation was found. A panel of European clinical isolates
was analyzed for sLex binding (paper II, online material). Close to 43% of cagA-
positive clinical isolates (33 of the 77) bound to sLex, while only 11% of cagA-
negative strains (2 out of 18) were positive for sLex.
These results were supported by analyses of bacterial adherence in situ to gastric
mucosa of experimentally challenged Rhesus monkeys (paper III). Here, H. pylori
infection triggers chronic mucosal inflammation, which leads to expression of sLex
antigen in the gastric cells. The increased expression of sLea and sLex is most
prominent during the first 2 months after inoculation. Changes in levels of sLea
before and after eradication have been reported, but these modifications may differ
according to the early or late phase of chronic inflammatory response or the stage of
the disease, as explained by the change being time-dependent. Notably, in situ H.
pylori density and in vitro adherence of the babA mutant was significantly correlated
with sLex/a expression in the antrum, supporting the idea that these antigens
function as receptors for the SabA adhesin.
11. The role of BabA and SabA for induction of inflammation
Expression of BabA and SabA must be under tight control (necessitating that the
organism have a detailed “knowledge” of its environment), so BabA and SabA are
produced only at the appropriate point during the journey of the organism from its
initial site of colonization to its ultimate target receptors. When the pathological
events have occurred because of the BabA-Leb binding activity, the SabA-sLex
interaction mediates the next level of attachment. Such stepwise attachment
Print & Media
33
mechanisms could direct the microbes to distinct ecological niches during the
chronic inflammation process, driving the inflammation process further to
development of peptic ulcer disease and/or malignant transformation. There could be
a continuous selection for H. pylori strains with the capacity to adapt to changes in
the environment such as local shifts in cell-differentiation and glycosylation patterns.
As discussed above, leukocytes that are targeted to enter inflamed tissues require
an initial cellular interaction with the endothelial lining mediated by selectin cell
adhesion molecules and their cognate sLex and sLea ligands. Since H. pylori strains
express SabA adhesins on their surface in order to adhere to inflamed gastric
mucosa, this renders them vulnerable to such host defense mechanisms. The sLex
binding activity of SabA can be regarded as a mechanism of selectin mimicry. So,
how then can H. pylori differentiate between sLex presented on the surface of
leukocytes versus inflamed gastric epithelium? Below, I will discuss two different
approaches to this contradiction:
I. Phase variation and biofim formation. II. Change of function with respect to patho-
adaptation.
11.1 Phase variation and biofilm formation
Surface components such as pili, OMPs, S layer proteins, flagellae,
lipopolysaccharides (LPSs) and capsules are involved in a number of host-bacteria
interactions including adhesion and resistance to phagocytosis, and many contribute
towards the pathogenic effects of the bacterium. Many of these components that are
antigenic and exposed to the mucosal immune system have the means to switch
expression on or off by mechanisms of phase variation.
Similarly, switching on and off adhesins such as SabA alters the binding properties.
Screening of all SabA and BabA-positive H. pylori strains examined indicated that
1% of colonies spontaneously lost their ability to bind sLex, in contrast to the almost
constitutive Leb antigen binding properties. The observation suggested instability in
sLex-binding activity. By switching off expression of SabA, bacteria may be able to
“de-camp” from a particular site and retreat into areas of non-inflamed gastric
environment. Such mechanisms of phase variation could provide the infection with
means of avoiding immune defenses.
Print & Media
34
An alternative explanation of phase variation would be to promote flexible survival
within the biofilm of organisms with a diverse set of nutrient and physicochemical
requirements (Figure 1). If the biofilm consists of only a single species, the different
microhabitats (pH, nutrition, availability of functional receptor) will enable changes to
gene expression patterns.
In addition, H. pylori populations infecting one individual can undergo significant
divergence, creating stable sub-clones with substantial genotypic and phenotypic
differences (Björkholm et al., 2001). Those events have been illustrated in a
schematically drawn picture in Figure 1.
11.2 SabA and BabA in relation to patho-adaptation
Expression of BabA and SabA by H. pylori results from adaptation to different micro-
niches and environments. The change of gene functions and programmed
expression patterns of genes during infection is termed patho-adaptation.
A variety of individual host conditions are critical to the susceptibility of an individual
to H. pylori infectious diseases. Virulent strains often share their primary habitat with
commensals and rarely spread into tissues where their growth will induce
symptomatic disease.
The phenotypic properties acquired during the evolution of pathogenic bacteria are
traits that provide strategies for: 1) tropism for tissues of a virulence niche, 2)
mechanisms to consume available nutrients, and/or 3) mechanisms to evade or
overcome antibacterial defenses. Recently, genomic changes were shown in H.
pylori infected Rhesus monkeys. In some cases, the babA gene was replaced by
babB, which is a functionally uncharacterized OMP but closely related to babA.
Bacteria which lack the babA gene could not bind to Leb antigen (Solnick et al.,
2004). Similarly, by these mechanisms SabA can change and/or expand the binding
properties and overcome the PMNs in an inflamed niche, and adapt to the new
environment during the course of infection.
In contrast, activation of PMNs (ROS-activities) by adherent bacteria and invasion of
both epithelial cells and PMNs might be the strategy to escape and simultaneously
employ the host immune defense to capture the other bacterial phenotypes and be
the primary pathogen. This evolutionary pathway is expected to be important for
Print & Media
35
microorganisms that are considered to be primary pathogens for their long-term
persistence in nature.
12. The role of LPS in H. pylori adherence
In paper IV, we studied the adhesive properties of H. pylori strains with Lex antigens
in their LPS, and their corresponding mutants, in situ, to gastric tissues of different
pathology. The results suggested that LPS did indeed demonstrate adhesive
properties, but only for a subset of the patient material studied.
However, the results imply a wide range of possible interactions, such as
carbohydrate-carbohydrate, protein-protein or protein-carbohydrate. Previously, Lex-
Lex interactions have been described to provide good affinity (Hakomori, 1994).
When comparing carbohydrate-carbohydrate interactions with protein-protein or
protein-carbohydrate ones, the first type of interaction is certainly generally
considered weak. One more characteristic feature of the carbohydrate-carbohydrate
interaction is its dependence on bivalent ions (calcium magnesium and manganese)
(Eggens et al., 1989). The carbohydrate-binding protein (lectin) mediates targeting
mainly by hydrogen and/or electrostatic interactions in conserved hydrophilic
residues. Though each hydrogen bond forms a relatively weak interaction, a group of
interactions can make a force strong enough for a tight and specific binding. In
particular, water molecules often bridge hydrogen bonds. The carbohydrates are as
a whole hydrophilic, and they also have significant hydrophobicity. This could also
possibly make a second class of interactions between lectins and carbohydrates,
such as protein-protein interactions between bacteria and eukaryotic cells mediated
by bacterial LPS. Thus, a model for interaction between an unknown H. pylori lectin
“B face” and unknown human lectin “A face” mediated by multimeric and multivalent
Lex antigens must be considered. We studied a series of trypsin-treated gastric
tissues and bacterial strains and the results more likely indicate a protein-protein
interaction, results which might argue for a protein-protein interaction mediated by
bacterial LPS side chains. Molecular interactions mediated by protein-LPS-protein
interaction are illustrated in Figure 4. In addition, it could be speculated that the initial
recognition and binding of H. pylori may occur through LPS and that subsequently a
more specific interaction with a lectin-like adhesin on the bacterial surface occurs.
Print & Media
36
Figure 4. A representative model showing protein-protein interaction between H.pylori lectins and host outer membrane proteins (A and B faces) and a ligandmolecule (here, trimeric Lewis x, galactose and fucose residues). Two kinds ofinteraction dominate the binding; i.e. hydrogen bonding and van der Waals contact.Hydrogen bonds could be formed between hydroxyl groups of a ligand saccharideand the side chains of hydrophilic residues, but might also involve backbonecarbonyl or amide groups. Note that in this case the "A face" is relatively hydrophilicwhile on the other hand the "B face" is relatively hydrophobic, to which anappropriate aromatic residue (most frequently tryptophan) may make efficientcontact.
Print & Media
37
Specific conclusions
• The H. pylori BabA adhesin binds to MUC5AC mucin structures, which carry
Le b antigen in different tissues and pathologies in the gastrointestinal tract.
• As a consequence of inflammation, sLex/a antigens are expressed in gastric
epithelium. Correlations between sLex expression, chronic inflammation and
H. pylori binding were found.
• The complementary H. pylori sialic acid-binding adhesin (SabA) that mediates
binding to sLex/a antigens was identified.
• Alterations in glycosylation patterns (Leb versus sLex/a) in gastric mucosa
challenged by H. pylori infection modulate and re-program H. pylori adhesion
over time in Rhesus monkey. The H. pylori BabA and SabA adhesins together
constitute a balanced selection where adapted phenotypes are functionally
selected for in different microenvironments.
• The adherence mode mediated by H. pylori LPS has a limited role in adhesion
processes. The data presented here were obtained from in vitro adhesion
experiments, and the functional impact of the H. pylori Lewis antigens for in
vivo colonization remains to be elucidated.
• Our new results on the functional prerequisites for H. pylori adherence have
contributed to elucidation of the molecular mechanisms by which H. pylori
infection causes disease. The new results have implications both for design of
novel therapies and for clinical strategies to identify individuals at risk.
Print & Media
38
Future perspectives
A vaccine, which would make prevention of this disease a feasible public health goal,
could prevent infection in millions of children and greatly reduce the worldwide
incidence of gastritis, ulcers, and stomach cancer.
Today, antibiotic therapy is the only efficient treatment to eradicate H. pylori
infections. However, antibiotic therapy results in the generation and selection of
antibiotic resistant bacteria throughout the body. In addition, complex drugs such as
antibiotics are not readily available in the developing parts of the world. Identification
of molecular targets is the first but essential step in the development of novel
antibacterial strategies. Blocking the primary stages of infection, namely bacterial
attachment to host cell receptors and colonization of the mucosal surface, may be
the most effective strategy for prevention of bacterial infections. Different strategies
could be exploited for inhibiting adhesion of H. pylori to gastric mucosa (i.e. BabA
and SabA), such as antibodies against adhesion, receptor analogs and adhesin
analogs. One or more of these could form the basis of new approaches for
preventing and/or treating H. pylori infection.
The second approach is to tailor-design oligosaccharides that mimic host receptors
to antagonize bacterial adhesion. One of the most promising animal (primate)
models for detailed analyzes of H. pylori infection is the Rhesus monkey, since the
Rhesus monkey closely mimics both human immune responses and glycosylation
patterns. Also, simple oliogosaccharides, such as sialyl-lactose, were recently shown
to have the potential to clear H. pylori infections in two of six animals, when given
orally to Rhesus monkeys persistently colonized by H. pylori (Mysore et al., 1999).
Recently, Roche et al, (2004) investigated the binding of SabA positive H. pylori
strains and corresponding mutants to gangliosides and concluded that the factors
affecting binding affinity were: (i) the length of the N-acetyl-lactosamine carbohydrate
chain, (ii) the branches of the carbohydrate chain, and (iii) fucose substitution of the
N-acetyl -lactosamine core chain. Thus, it is of importance to consider the length,
density and multivalancy of synthetic oligosaccharides for efficient inhibition.
There is growing evidence that whole grain cereals, especially when bioprocessed,
play important roles in the prevention of chronic diseases, including bacterial
infections (Kushi et al., 1999; Björck et al., 2000). In a parallel project, we found that
fermented rye-bran (FRB) demonstrated anti-microbial activities. These novel
Print & Media
39
properties of FRB could interfere with Lewis b antigen mediated adherence of H.
pylori both in radioimmunoanalysis and to the stomach epithelium in situ (to
histological tissue sections). Finally, to corroborate the in vitro results, volunteer
individuals with H. pylori infections were treated with FRB. The results were
promising and showed that the biological activity of the H. pylori infection decreased
during therapy, as evidenced by a steep reduction in infectious load. The combined
results also suggest that FRBs contain specific receptor inhibitors for binding
mediated by Leb, which could be used for treatment of infections caused by the
virulent H. pylori strains, and for establishment of a protective gastric microflora.
We will next identify the anti-adhesive compounds in FRB and study their activities
towards SabA. The findings may thus help to develop new intragastric therapeutics
against H. pylori to suppress chronic inflammation and possibly other perturbations
in the gastric mucosa that lead to ulceration and malignancy (Mahdavi et al., in
preparation).
Print & Media
40
Acknowledgements
I would like to acknowledge the contributions of a number of people who directed the
construction of this book. I deeply appreciate their valuable contribution. I am
particularly grateful to my supervisor Thomas Borén for providing innovative
research during a journey of continuous change and personal challenge. Thank you
to Ben Appelmelk, Sven Bergström, Göran and Hanna Marklund for
criticism of my book and for the time you invested in helping me put my best foot
forward. I also thank my special friend and colleague, Anders Johansson, for
meeting the challenge presented when parts of my project didn’t function the way
they usually did. I needed to find new ways of doing things, and although some have
been of little help, others have changed my life. He also gave me a view of where I
have been, and should serve as a guide on how I should plan my future.
Finally, special thanks to my family who provided me with strength and
encouragement – giving me strength by understanding the specifics of the
challenges I might need to face. Encouragement from knowing that although
researching requires many moments of absence, the spirit remains firmly surrounded
by the love of families and those who care.
Print & Media
Ha lmansl
41
References
Ackermark P, Kuipers EJ, Wolf C, Breumelhof R, Seldenrijk CA, Timmer R,Segeren KC, Kusters JG, Smout AJ. (2003). Colonization with cagA-positiveHelicobacter pylori strains in intestinal metaplasia of the esophagus and theesophagogastric junction. Am. J. Gastroenterol. 98:1719-1724.
Allen LA. (2001). The role of the neutrophil and phagocytosis in infection caused byHelicobacter pylori. Curr Opin Infect Dis. 14: 273-277.
Allen LA, Schlesinger LS, Kang B. (2000). Virulent strains of Helicobacter pyloridemonstrate delayed phagocytosis and stimulate homotypic phagosome fusion inmacrophages. J Exp Med. 191:115-128.
Alm RA, Ling LS, Moir DT, King BL, Brown ED, Doig PC, Smith DR, Noonan B,Guild BC, deJonge BL, Carmel G, Tummino PJ, Caruso A, Uria-Nickelsen M,Mills DM, Ives C, Gibson R, Merberg D, Mills SD, Jiang Q, Taylor DE, Vovis GF,Trust TJ. (1999). Genomic-sequence comparison of two unrelated isolates of thehuman gastric pathogen Helicobacter pylori. Nature. 397:176-180.
Alper J. (2001). Searching for medicine's sweet spot. Science. 2001 291:2338-2343.
Amieva MR, Salama NR, Tompkins LS, Falkow S. (2002). Helicobacter pylorienter and survive within multivesicular vacuoles of epithelial cells. Cell Microbiol.4:677–690.
Appelmelk BJ, Simoons-Smit I, Negrini R, Moran AP, Aspinall GO, Forte JG, DeVries T, Quan H, Verboom T, Maaskant JJ, Ghiara P, Kuipers EJ, Bloemena E,Tadema TM, Townsend RR, Tyagarajan K, Crothers JM Jr, Monteiro MA, SavioA, De Graaff J. (1996). Potential role of molecular mimicry between Helicobacterpylori lipopolysaccharide and host Lewis blood group antigens in autoimmunity.Infect. Immun. 64:2031-2040.
Aras RA, Takata T, Ando T, van der Ende A, Blaser MJ. (2001). Regulation of theHpyII restriction-modification system of Helicobacter pylori by gene deletion andhorizontal reconstitution. Mol Microbiol. 42:369–382.
Aras RA, Kang J, Tschumi AI, Harasaki Y, Blaser MJ. (2003). Extensive repetitiveDNA facilitates prokaryotic genome plasticity. Proc. Natl. Acad. Sci. USA.100:13579-81354.
Arneberg P, Kornstad L, Nordbo H, Gjermo P. (1976). Less dental caries amongsecretors than among non-secretors of blood group substance. Scand J Dent Res.84:362-366.
Print & Media
42
Aspholm-Hurtig M, Dailide G, Lahmann M, Kalia A, Ilver D, Roche N, VikströmS, Sjöström R, Lindén S, Bäckström A, Arnqvist A, Mahdavi J, Nilsson UJ,Velapatiño B, Gilman RH, Gerhard M, Alarcon T, López-Brea M, Nakazawa T,Fox JG, Correa P, Dominguez-Bello MG, Perez-Perez GI, Blaser MJ, Normark S,Carlstedt I, Oscarson S, Teneberg S, Berg DE, Borén T. Functional Adaptation ofBabA, the Helicobacter pylori ABO Blood-Group Antigen Binding Adhesin.(submitted).
Bevilacqua MP, Nelson RM, Mannori G, Cecconi O. (1994). Endothelial- leukocyteadhesion molecules in human disease. Annu. Rev. Med. 45:361–378.
Björck S, Bosaeus I, Ek E, Jennische E, Lonnroth I, Johansson E, Lange S.(2000). Food induced stimulation of the antisecretory factor can improve symptomsin human inflammatory bowel disease: a study of a concept. Gut. 46:824-829.
Björkholm B, Zhukhovitsky V, Lofman C, Hulten K, Enroth H, Block M, Rigo R,Falk P, Engstrand L. (2000). Helicobacter pylori entry into human gastric epithelialcells: a potential determinant of virulence, persistence, and treatment failures.Helicobacter. 5:148–155.
Björkholm B, Lundin A, Sillen A, Guillemin K, Salama N, Rubio C, Gordon JI,Falk P, Engstrand L. (2001). Comparison of genetic divergence and fitnessbetween two subclones of Helicobacter pylori. Infect. Immun. 69:7832-7838.
Björkholm B, Falk P, Engstrand L, Nyren O. (2003). Helicobacter pylori:resurrection of the cancer link. J Intern Med. 253:102-119.
Blaser MJ. (1992). Hypotheses on the pathogenesis and natural history ofHelicobacter pylori-induced inflammation. Gastroenterology. 102:720-727.
Blaser MJ, Chyou PH, Nomura A. (1995). Age at establishment of Helicobacterpylori infection and gastric carcinoma, gastric ulcer, and duodenal ulcer risk. CancerRes. 55:562–565.
Blaser MJ. (1996). The bacteria behind ulcers. Sci. Am. 274:104-107.
Blaser MJ, Atherton JC. (2004). Helicobacter pylori persistence: biology anddisease. J. Clin. Invest. 113:321-333.
Borén T, Falk P, Roth KA, Larson G, Normark S. (1993). Attachment ofHelicobacter pylori to human gastric epithelium mediated by blood group antigens.Science 262:1892-1895.
Boquet P, Ricci V, Galmiche A, Gauthier NC. (2003). Gastric cell apoptosis and H.pylori: has the main function of VacA finally been identified? Trends Microbiol.11:410-413.
Broadberry RE, Lin-Chu M. (1991). The Lewis blood group system among Chinesein Taiwan. Hum. Hered. 41:290-294.
Print & Media
43
Brown LM. (2000). Helicobacter pylori: epidemiology and routes of transmission.Epidemiol. Rev. 22:283-297.
Bury-Mone S, Skouloubris S, Labigne A, De Reuse H. (2001). The Helicobacterpylori UreI protein: role in adaptation to acidity and identification of residues essentialfor its activity and for acid activation. Mol. Microbiol. 42:1021-1034.
Censini S, Stein M, Covacci A. (2001). Cellular responses induced after contactwith Helicobacter pylori. Curr. Opin. Microbiol. 4:41-46.
Chen W, Li D, Cannan RJ, Stubbs RS. (2003). Common presence of HelicobacterDNA in the gallbladder of patients with gallstone diseases and controls. Dig LiverDis. 35:237-243.
Chiou CC, Chan CC, Sheu DL, Chen KT, Li YS, Chan EC. (2001). Helicobacterpylori infection induced alteration of gene expression in human gastric cells. Gut.48:598-604.
Clausen H, Hakomori S.i. (1989). ABH and related histo-blood group antigens;immunochemical differences in carrier isotypes and their distribution. Vox Sang.56:1-20.
Correa P. (1992). Human gastric carcinogenesis: a multistep and multifactorialprocess. First American Cancer Society Award Lecture on Cancer Epidemiology andPrevention. Cancer Res. 52:6735-6740.
Correa P. (2003). Bacterial infections as a cause of cancer. J. Natl. Cancer Inst.95:E3.
Crabtree JE. (1996). Gastric mucosal inflammatory responses to Helicobacter pylori.Aliment Pharmacol Ther. Suppl 1:29-37.
Cover TL, Berg DE, Blaser MJ, Mobley HL. (2001). Principles of BacterialPathogenesis, chapter 11; H. pylori pathogenesis. pp:509-558.
Drumm B, Perez-Perez GI, Blaser MJ, Sherman PM. (1990). Intrafamilialclustering of Helicobacter pylori infection. N. Engl. J. Med. 322:359-363.
Ebert MP, Schandl L, Malfertheiner P. (2002). Helicobacter pylori infection andmolecular changes in gastric carcinogenesis. J. Gastroenterol. Suppl 13: 45-49.
Edwards NJ, Monteiro MA, Faller G, Walsh EJ, Moran AP, Roberts IS, High NJ.(2000). Lewis X structures in the O-antigen side-chain promote adhesion ofHelicobacter pylori to the gastric epithelium. Mol. Microbiol. 35: 1530-1539.
Eggens I, Fenderson BA, Toyokuni T, Hakomori S. (1989). A role ofcarbohydrate-carbohydrate interaction in the process of specific cell recognitionduring embryogenesis and organogenesis: a preliminary note. Biochem. Biophys.Res. Commun. 158:913-920.
Print & Media
44
El-Omar EM, Carrington M, Chow WH, McColl KE, Bream JH, Young HA,Herrera J, Lissowska J, Yuan CC, Rothman N, Lanyon G, Martin M, FraumeniJF Jr, Rabkin CS. (2000). Interleukin-1 polymorphism associated with increase riskof gastric cancer. Nature. 404:398–402.
Enroth H, Wreiber K, Rigo R, Risberg D, Uribe A, Engstrand L. (1999). In vitroaging of Helicobacter pylori: changes in morphology, intracellular composition andsurface properties. Helicobacter. 4:7-16.
Ertl C, Wrobel KH. (1992). Distribution of sugar residues in the bovine testis duringpostnatal ontogenesis demonstrated with lectin-horseradish peroxidase conjugates.Histochemistry. 97:161-171.
Evans D. (2000). Helicobacter pylori adhesins. Rev Perspect Helicobacter5:183–195.
Farizo KM, Huang T, Burns DL. (2000). Importance of holotoxin assembly in Ptl-mediated secretion of pertussis toxin from Bordetella pertussis. Infect. Immun. 68:4049–4054.
Filipe MI. (1979). Mucins in the human gastrointestinal epithelium: a review. Invest.Cell Pathol. 2:195-216.
Fox JG, Chien CC, Dewhirst FE, Paster BJ, Shen Z, Melito PL, Woodward DL,Rodgers FG. (2000). Helicobacter canadensis sp. nov. isolated from humans withdiarrhea as an example of an emerging pathogen. J. Clin. Microbiol. 38:2546-2549.
Fujimoto S, Marshall B, Blaser MJ. (1994). PCR-based restriction fragment lengthpolymorphism typing of Helicobacter pylori. J. Clin. Microbiol. 3:331-334.
Galustian C, Elviss N, Chart H, Owen R, Feizi T. (2003). Interactions of thegastrotropic bacterium Helicobacter pylori with the leukocyte-endothelium adhesionmolecules, the selectins--a preliminary report. FEMS Immunol Med Microbiol.36:127-134.
Gisbert JP, Blanco M, Cruzado AI, Pajares JM. (2000). Helicobacter pyloriinfection, gastric metaplasia in the duodenum and the relationship with ulcerrecurrence. Eur. J. Gastroenterol. Hepatol. 12:1295-1298.
Glickman JN, Shahsafaei A, Odze RD. (2003). Mucin Core Peptide ExpressionCan Help Differentiate Barrett's Esophagus From Intestinal Metaplasia of theStomach. Am. J. Surg. Pathol. 27:1357-1365.
Graham DY. (1991). Helicobacter pylori: its epidemiology and its role in duodenalulcer disease. J. Gastroenterol. Hepatol. 6:105-113.
Hakomori S.i. (1994). Role of carbohydrates in cell adhesion and recognition. InComplex Carbohydrates in Drug Research (Bock K, Clausen H eds) pp 337-349,Copenhagen:Munksgaard.
Print & Media
45
Hansson LE, Nyren O, Hsing AW, Bergstrom R, Josefsson S, Chow WH,Fraumeni JF Jr, Adami HO. (1996). The risk of stomach cancer in patients withgastric or duodenal ulcer disease. N. Engl. J. Med. 335:242-249.
Harris AW, Gummett PA, Walker MM, Misiewicz JJ, Baron JH. (1996). Relationbetween gastric acid output, Helicobacter pylori, and gastric metaplasia in theduodenal bulb. Gut. 39:513-520.
Henry S, Mollicone R, Fernandez P, Samuelsson B, Oriol R, Larson G. (1996).Molecular basis for erythrocyte Le(a+ b+) and salivary ABH partial-secretorphenotypes: expression of a FUT2 secretor allele with an A�T mutation atnucleotide 385 correlates with reduced alpha(1,2) fucosyltransferase activity.Glycoconj. J. 13:985-993.
Henry S, Oriol R, Samuelsson B. (1995). Lewis histo-blood group system andassociated secretory phenotypes. Vox. Sang. 69:166-182.
Henry S, Woodfield DG, Samuelsson BE, Oriol R. (1993). Plasma and red-cellglycolipid patterns of Le(a+b+) and Le(a+b-) Polynesians as further evidence of theweak secretor gene Se(w). Vox. Sang. 65:62-69.
Holbrook WP. (1989). Secretor status and dental caries in Iceland. FEMS Microbiol.Immunol. 1:397-399.
Holler E. (2002). Cytokines, viruses, and graft-versus-host disease. Curr. Opin.Hematol. 9:479-484.
Hultgren SJ, Abraham S, Caparon M, Falk P, St. Geme III JW and Normark S.(1993). Pilus and non-pilus bacterial adhesins: assembly and function in cellrecognition. Cell 73: 887-901.
Ilver D, Arnqvist A, Ogren J, Frick IM, Kersulyte D, Incecik ET, Berg DE,Covacci A, Engstrand L, Boren T. (1998). Helicobacter pylori adhesin bindingfucosylated histo-blood group antigens revealed by retagging. Science. 279:373-377.
Ilver D, Barone S, Mercati D, Lupetti P, Telford JL. (2004). Helicobacter pyloritoxin VacA is transferred to host cells via a novel contact-dependent mechanism.Cell Microbiol. 6:167-174.
Israel DA, Salama N, Krishna U, Rieger UM, Atherton JC, Falkow S, Peek RM Jr.(2001). Helicobacter pylori genetic diversity within the gastric niche of a singlehuman host. Proc. Natl. Acad. Sci. USA. 98:14625–14630.
Jorgensen M, Daskalopoulos G, Warburton V, Mitchell HM, Hazell SL. (1996).Multiple strain colonization and metronidazole resistance in Helicobacter pylori-infected patients: identification from sequential and multiple biopsy specimens.Infect. Dis. 174:631-635.
Print & Media
46
Kageshita T, Hirai S, Kimura T, Hanai N, Ohta S, Ono T. (1995). Associationbetween sialyl Lewis(a) expression and tumor progression in melanoma. CancerRes. 55:1748-1751.
Karlsson KA. (1998). Meaning and therapeutic potential of microbial recognition ofhost glycoconjugates. Mol. Microbiol. 29:1-11.
Karam SM. (1999) Lineage commitment and maturation of epithelial cells in the gut.Frontiers in Bioscinces. 4:D286- 298.
Kukowska-Latallo JF, Larsen RD, Nair RP, Lowe JB. (1990). A cloned humancDNA determines expression of a mouse stage-specific embryonic antigen and theLewis blood group alpha(1,3/1,4)fucosyltransferase. Genes Dev. 4:1288-1303.
Kelly RJ, Rouquier S, Giorgi D, Lennon GG, Lowe JB. (1995). Sequence andexpression of a candidate for the human Secretor blood groupalpha(1,2)fucosyltransferase gene (FUT2). Homozygosity for an enzyme-inactivatingnonsense mutation commonly correlates with the non-secretor phenotype. J. Biol.Chem. 270:4640-4649.
Kersulyte D, Chalkauskas H, Berg DE. (1999). Emergence of recombinant strainsof Helicobacter pylori during human infection. Mol. Microbiol. 31:31-43.
Kim N, Marcus EA, Wen Y, Weeks DL, Scott DR, Jung HC, Song IS, Sachs G.(2004). Genes of Helicobacter pylori Regulated by Attachment to AGS Cells. Infect.Immun. 72:2358-2368.
Kushi LH, Meyer KA, Jacobs DR. (1999). Cereals, Legumes, and chronic diseaserisk reduction: evidence from epidemiologic studies. Am. J. Clin. Nutr. 70(suppl),451s-8s
Kwok T, Backert S, Schwarz H, Berger J, Meyer TF. (2002). Specific entry ofHelicobacter pylori into cultured gastric epithelial cells via a zipper-like mechanism.Infect. Immun. 70:2108-2120.
Larsen RD, Ernst LK, Nair RP, Lowe JB. (1990). Molecular cloning, sequence, andexpression of a human GDP-L-fucose:beta-D-galactoside 2-alpha-L-fucosyltransferase cDNA that can form the H blood group antigen. Proc. Natl. Acad.Sci USA. 87:6674-6678.
Leunk RD, Johnson PT, David BC, Kraft WG, Morgan DR. (1988). Cytotoxicactivity in broth-culture filtrates of Campylobacter pylori. J. Med. Microbiol. 26:93-99.
Le Bouder-Langevin S, Capron-Montaland I, De Rosa R, Labedan B. (2002). Astrategy to retrieve the whole set of protein modules in microbial proteomes.Genome Res. 12:1961-1973.
Lindén S, Mahdavi J, Hedenbro J, Boren T, Carlstedt I. (2004). Effects of pH onHelicobacter pylori binding to human gastric mucins – identification of binding to non-MUC5AC mucins. (submitted).
Print & Media
47
Lipkin M, Correa P, Mikol YB, Higgins PJ, Cuello C, Zarama G, Fontham E,Zavala D. (1985). Proliferative and antigenic modifications in human epithelial cellsin chronic atrophic gastritis. J. Natl. Cancer Inst. 75:613-619.
Logan SM. (1994). IARC Working Group on the Evaluation of Carcinogenic Risks toHumans. Lancet. 344:1078.
Machado JC, Figueiredo C, Canedo P, Pharoah P, Carvalho R, Nabais S, CastroAlves C, Campos ML, Van Doorn LJ, Caldas C, Seruca R, Carneiro F, Sobrinho-Simoes M. (2003). A proinflammatory genetic profile increases the risk for chronicatrophic gastritis and gastric carcinoma. Gastroenterology. 125:364-371.
McCormick BA, Parkos CA, Colgan SP, Carnes DK, Madara JL. (1998). Apicalsecretion of a pathogen-elicited epithelial chemoattractant activity in response tosurface colonization of intestinal epithelia by Salmonella typhimurium. J. Immunol.160:455-466.
Madrid JF, Ballesta J, Castells MT, Hernandez F. (1990). Glycoconjugatedistribution in the human fundic mucosa revealed by lectin- and glycoprotein-goldcytochemistry. Histochemistry. 95:179-187.
Mahdavi J, Johansson A, Olfat FO, Danielsson Å, Kamal-Eldin A, Andersson R,Åman P, Hallmans G, Borén T. Adherence of Helicobacter pylori to Gastric MucosaPrevented by Fermented Rye-Bran – A Novel Therapeutic Approach. (manuscript).
Marionneau S, Cailleau-Thomas A, Rocher J, Le Moullac-Vaidye B, Ruvoen N,Clement M, Le Pendu J. (2001). ABH and Lewis histo-blood group antigens, amodel for the meaning of oligosaccharide diversity in the face of a changing world.Biochimie. 83:565-573.
Marshall BJ, Warren JR. (1984). Unidentified curved bacilli in the stomach ofpatients with gastritis and peptic ulceration. Lancet. 1:1311-1315.
Monteiro MA. (2001). Helicobacter pylori: a wolf in sheep’s clothing: the glycotypefamilies of Helicobacter pylori lipopolysaccharides expressing histoblood groups:structure, biosynthesis, and role in pathogenesis. Adv. Carbohydr. Chem. Biochem.57:99-158.
Morgenstern S, Koren R, Moss SF, Fraser G, Okon E, Niv Y. (2001). DoesHelicobacter pylori affect gastric mucin expression? Relationship between gastricantral mucin expression and H. pylori colonization. Eur. J, Gastroenterol. Hepatol.13:19-23.
Muotiala A, Helander IM, Pyhala L, Kosunen TU, Moran AP. (1992). Lowbiological activity of Helicobacter pylori lipopolysaccharide. Infect. Immun. 60:1714-1716.
Mysore JV, Wigginton T, Simon PM, Zopf D, Heman-Ackah LM, Dubois (1999).A.Treatment of Helicobacter pylori infection in rhesus monkeys using a novelantiadhesion compound. Gastroenterology. 117:1316-1325.
Print & Media
48
Ndip RN, MacKay WG, Farthing MJ, Weaver LT. (2003). Culturing Helicobacterpylori from clinical specimens: review of microbiologic methods. J. Pediatr.Gastroenterol. Nutr. 36:616-622.
Newburg DS. (2000). Oligosaccharides in human milk and bacterial colonization,Pediatr. Gastroenterol. Nutr. 30:S8–S17.
Nilsson C, Sillen A, Eriksson L, Strand ML, Enroth H, Normark S, Falk P,Engstrand L. (2003). Correlation between cag pathogenicity island composition andHelicobacter pylori-associated gastroduodenal disease. Infect. Immun. 71: 6573-6581.
Nilsson HO, Mulchandani R, Tranberg KG, Stenram U, Wadstrom T. (2001).Helicobacter species identified in liver from patients with cholangiocarcinoma andhepatocellular carcinoma. Gastroenterology. 120:323-324.
Odenbreit S, Puls J, Sedlmaier B, Gerland E, Fischer W, Haas R. (2000).Translocation of Helicobacter pylori CagA into gastric epithelial cells by type IVsecretion. Science. 287:1497-1500.
Olfat FO, Naslund E, Freedman J, Boren T, Engstrand L. (2002). Cultured humangastric explants: a model for studies of bacteria-host interaction during conditions ofexperimental Helicobacter pylori infection. J. Infect. Dis. 186:423-427.
Olfat FO, Aspholm M, Sondén B, Engstrand L, Teneberg S, Borén T. TheNeuraminyllactose-Binding Hemagglutinin of Helicobacter pylori is Identical to theSialic Acid Binding Adhesin, SabA. (manuscript).
Ota H, Nakayama J, Momose M, Hayama M, Akamatsu T, Katsuyama T, GrahamDY, Genta RM. (1998). Helicobacter pylori infection produces reversibleglycosylation changes to gastric mucins. Virchows Arch. 433:419-426.
Papini E, Zoratti M, Cover TL. (2001) In search of the Helicobacter pylori VacAmechanism of action. Toxicon. 39:1757-1767.
Parkin M D, Stjhernsward J, Muir C. (1988). Estimates of the worldwide frequencyof twelve major cancers. Int. J. Cancer. 41:184-197.
Park Y, Lamont RJ. (1998). Contact-dependent protein secretion in Porphyromonasgingivalis. Infect. Immun. 66:4777-4782.
Pride DT, Blaser MJ. (2002). Concerted evolution between duplicated geneticelements in Helicobacter pylori. J Mol Biol. 316:629-642.
Petersson C, Forsberg M, Aspholm M, Olfat FO, Forslund T, Borén T,Magnusson KE. Helicobacter pylori evokes a strong inflammatory response inhuman neutrophils by SabA-adhesin mediated selectin-mimicry. (manuscript).
Print & Media
49
Pochec E, Litynska A, Amoresano A, Casbarra A. (2003). Glycosylation profile ofintegrin alpha 3 beta 1 changes with melanoma progression. Biochim. Biophys. Acta.1643:113-123.
Rautelin H, Blomberg B, Fredlund H, Jarnerot G, Danielsson D. (1993).Incidence of Helicobacter pylori strains activating neutrophils in patients with pepticulcer disease. Gut. 34:599-603.
Reglero A, Rodriguez-Aparicio LB, Luengo JM. (1993). Polysialic acids. Int JBiochem. 25:1517-1527.
Renkonen R, Mattila P, Majuri ML, Rabina J, Toppila S, Renkonen J, Hirvas L,Niittymaki J, Turunen JP, Renkonen O, Paavonen T. (1997). In vitro experimentalstudies of sialyl Lewis x and sialyl Lewis a on endothelial and carcinoma cells: crucialglycans on selectin ligands. Glycoconj. J. 14:593-600.
Rieder G, Hatz RA, Moran AP, Walz A, Stolte M, Enders G. (1997). Role ofadherence in interleukin-8 induction in Helicobacter pylori-associated gastritis. Infect.Immun. 65: 3622-3630.
Roche N, Angstrom J, Hurtig M, Larsson T, Boren T, Teneberg S. (2004).Helicobacter pylori and complex gangliosides. Infect. Immun. 72:1519-1529.
Roma-Giannikou E, Karameris A, Balatsos B, Panayiotou J, Manika Z, Van-VlietC, Rokkas T, Skandalis N, Kattamis C. (2003). Intrafamilial spread of Helicobacterpylori: a genetic analysis. Helicobacter. 8:15-20.
Romaniuk PJ, Zoltowska B, Trust TJ, Lane DJ, Olsen GJ, Pace NR, Stahl DA.(1987). Campylobacter pylori, the spiral bacterium associated with human gastritis, isnot a true Campylobacter sp. J. Bacteriol. 169:2137-2141.
Rosner H. (1982). Ganglioside changes in the chicken optic lobes as biochemicalindicators of brain development and maturation. Brain Res. 236:49-61.
Rosai J. (1989). Basal cell carcinoma with follicular differentiation. Am. J.Dermatopathol. 11:479-497.
Ross IN, Bahari HM, Turnberg LA. (1981). The pH gradient across mucusadherent to rat fundic mucosa in vivo and the effect of potential damaging agents.Gastroenterology. 81:713-718.
Rotter JI. (1983). Peptic ulcer. In: The principles and practice of Medical Genetics.Emery AEH, Rimoin DL, eds. New York: Churchill Livingstone. 863-878.
Saadi AT, Weir DM, Poxton IR, Stewart J, Essery SD, Blackwell CC, Raza MW,Busuttil A. (1994). Isolation of an adhesin from Staphylococcus aureus that bindsLewis a blood group antigen and its relevance to sudden infant death syndrome,FEMS Immunol. Med. Microbiol. 8:315–320.
Print & Media
50
Satin B, Del Giudice G, Della Bianca V, Dusi S, Laudanna C, Tonello F, KelleherD, Rappuoli R, Montecucco C, Rossi F. (2000). The neutrophil-activating protein(HP-NAP) of Helicobacter pylori is a protective antigen and a major virulence factor.J. Exp. Med. 191:1467-1476.
Sauer FG, Barnhart M, Choudhury D, Knight SD, Waksman G, Hultgren SJ.(2000). Chaperone-assisted pilus assembly and bacterial attachment. Curr. Opin.Struct Biol. 10:548-556.
Schade C, Flemstrom G, Holm L. (1994). Hydrogen ion concentration in the mucuslayer on top of acid-stimulated and -inhibited rat gastric mucosa. Gastroenterology.107:180-188.
Segal ED, Cha J, Lo J, Falkow S, Tompkins LS. (1999). Altered states:involvement of phosphorylated CagA in the induction of host cellular growth changesby Helicobacter pylori. Proc. Natl. Acad. Sci. USA. 96:14559-14564.
Shirin H, Weinstein IB, Moss SF. (2001). Effects of H. pylori infection of gastricepithelial cells on cell cycle control. Front. Biosci. 6:E104-18.
Shirin H, Hibshoosh H, Kawabata Y, Weinstein IB, Moss SF. (2003). p16Ink4a isoverexpressed in H. pylori-associated gastritis and is correlated with increasedepithelial apoptosis. Helicobacter. 8:66-71.
Shimizu T, Akamatsu T, Ota H, Katsuyama T. (1996). Immunohistochemicaldetection of Helicobacter pylori in the surface mucous gel layer and itsclinicopathological significance. Helicobacter. 1:197-206.
Sipponen P, Aarynen M, Kaariainen I, Kettunen P, Helske T, Seppala K. (1989).Chronic antral gastritis, Lewis(a+) phenotype, and male sex as factors in predictingcoexisting duodenal ulcer. Scand. J. Gastroenterol. 5:581-588.
Slattery MJ, Dong C. (2003). Neutrophils influence melanoma adhesion andmigration under flow conditions. Int. J. Cancer. 106:713-722.
Sokurenko EV, Hasty DL, Dykhuizen DE. (1999). Pathoadaptive mutations: geneloss and variation in bacterial pathogens. Trends Microbiol. 7:191-195.
Sokurenko EV, Courtney HS, Ohman DE, Klemm P, Hasty DL. (1994). FimHfamily of type 1 fimbrial adhesins: functional heterogeneity due to minor sequencevariations among fimH genes. J. Bacteriol. 176:748-755.
Solnick JV, Schauer DB. (2001). Emergence of diverse Helicobacter species in thepathogenesis of gastric and enterohepatic diseases. Clin. Microbiol. Rev. 14:59-97.
Solnick JV, Hansen LM, Salama NR, Boonjakuakul JK, Syvanen M. (2004).Modification of Helicobacter pylori outer membrane protein expression duringexperimental infection of rhesus macaques. Proc. Natl. Acad. Sci. USA. 101:1206-1211.
Print & Media
51
Soto GE, Hultgren SJ. (1999). Bacterial adhesins: common themes and variationsin architecture and assembly. J. Bacteriol. 181:1059-1071.
Spechler SJ. (2002). Barrett's esophagus and esophageal adenocarcinoma:pathogenesis, diagnosis, and therapy. Med. Clin. North. Am. 86:1423-1445.
Stapleton A, Nudelman E, Clausen H, Hakomori S, Stamm WE. (1992). Bindingof uropathogenic Escherichia coli R45 to glycolipids extracted from vaginal epithelialcells is dependent on histo-blood group secretor status. J. Clin. Invest. 90:965-972.
Staat MA, Kruszon-Moran D, McQuillan GM, Kaslow RA. (1996). A population-based serologic survey of Helicobacter pylori infection in children and adolescents inthe United States. J. Infect. Dis. 174:1120-1123.
Stark RM. (1999). Biofilm formation by Helicobacter pylori. Lett. Appl. Microbiol.28:121-126.
Stein M, Rappuoli R, Covacci A. (2000). Tyrosine phosphorylation of theHelicobacter pylori CagA antigen after cag-driven host cell translocation. Proc. Natl.Acad. Sci. USA. 97:1263-1268.
Strömberg N, Nyholm PG, Pascher I, Normark S. (1991). Saccharide orientationat the cell surface affects glycolipid receptor function. Proc. Natl. Acad. Sci. USA.88:9340-9344.
Suerbaum S, Smith JM, Bapumia K, Morelli G, Smith NH, Kunstmann E, DyrekI, Achtman M. (1998). Free recombination within Helicobacter pylori. Proc. Natl.Acad. Sci. USA. 95:12619-12624.
Su B, Ceponis PJ, Lebel S, Huynh H, Sherman PM. (2003). Helicobacter pyloriactivates Toll-like receptor 4 expression in gastrointestinal epithelial cells. Infect.Immun. 71:3496-3502.
Syder AJ, Guruge JL, Li Q, Hu Y, Oleksiewicz CM, Lorenz RG, Karam SM, FalkPG, Gordon JI. (1999). Helicobacter pylori attaches to NeuAc alpha 2,3Gal beta 1,4glycoconjugates produced in the stomach of transgenic mice lacking parietal cells.Mol. Cell. 3:263-274.
Tesh VL. (1998). Virulence of enterohemorrhagic Escherichia coli: role of molecularcrosstalk. Trends Microbiol. 6:228-233.
Telford JL, Covacci A, Rappuoli R, Chiara P. (1997). Immunobiology ofHelicobacter pylori infection. Curr. Opin. Immunol. 9:498-503.
Telford JL, Ghiara P, Dell'Orco M, Comanducci M, Burroni D, Bugnoli M, TecceMF, Censini S, Covacci A, Xiang Z, et al. (1994). Gene structure of theHelicobacter pylori cytotoxin and evidence of its key role in gastric disease. J. Exp.Med. 179:1653-1658.
Print & Media
52
Testino G, Cornaggia M, Valentini M. (1999). Gastric cyto-secretory correlations inpeptic ulcer. Hepatogastroenterology. 46:2710-2712.
Tomb JF, White O, Kerlavage AR, et al. (1997). The complete genome sequenceof the gastric pathogen Helicobacter pylori . Nature 388:539-547.
Valkonen KH, Wadstrom T, Moran AP. (1994). Interaction of lipopolysaccharidesof Helicobacter pylori with basement membrane protein laminin. Infect. Immun.62:3640-3648.
Valkonen KH, Wadstrom T, Moran AP. (1997). Identification of the N-acetylneuraminyllactose-specific laminin-binding protein of Helicobacter pylori. Infect.Immun. 65:916-923.
Van den Brink GR, Tytgat KM, Van der Hulst RW, Van der Loos CM, EinerhandAW, Buller HA, Dekker J. (2000). H pylori colocalises with MUC5AC in the humanstomach. Gut. 46:601-607.
Wyatt JI, Rathbone BJ, Sobala GM, Shallcross T, Heatley RV, Axon AT, DixonMF. (1990). Gastric epithelium in the duodenum: its association with Helicobacterpylori and inflammation. J. Clin. Pathol. 43:981-986.
Wen Y, Marcus EA, Matrubutham U, Gleeson MA, Scott DR, Sachs G. (2003).Acid-adaptive genes of Helicobacter pylori. Infect. Immun. 71:5921-5939.
Weeks DL, Eskandari S, Scott DR, Sachs G. (2000). A H+-gated urea channel: thelink between Helicobacter pylori urease and gastric colonization. Science.287:482–485.
Wilson M, McNab R, Henderson B. (2002). Bacterial Disease Mechanisms; Anintroduction to cellular microbiology;. Cambridge University press,(www.cambridge.org). chapter 7.
Wieczorek TJ, Wang HH, Antonioli DA, Glickman JN, Odze RD. (2003).Pathologic features of reflux and Helicobacter pylori-associated carditis: acomparative study. Am. J. Surg. Pathol. 27:960-968.
Woolhouse ME, Taylor LH, Hayden DT. (2001). Population biology of multihostpathogens. Science. 292:1109–1112.
Xiang Z, Censini S, Bayeli PF, Telford JL, Figura N, Rappuoli R, Covacci A.(1995). Analysis of expression of CagA and VacA virulence factors in 43 strains ofHelicobacter pylori reveals that clinical isolates can be divided into two major typesand that CagA is not necessary for expression of the vacuolating cytotoxin. Infect.Immun. 63:94-98.
Print & Media
I
Print & Media
© 2003 Blackwell Publishing Ltd, Helicobacter, 8, 521–532 521
Volume 8 • Number 5 • 2003H E L I C O B AC T E R
Blackwell Publishing Ltd.The MUC5AC Glycoprotein is the Primary Receptor for Helicobacter pylori in the Human Stomach
Jeroen H. B. Van de Bovenkamp,* Jafar Mahdavi,† Anita M. Korteland-Van Male,* Hans A. Büller,*Alexandra W. C. Einerhand,* Thomas Borén† and Jan Dekker*
*Laboratory of Pediatrics, Erasmus MC/Sophia Children’s Hospital, Rotterdam, the Netherlands; †Department of Odontology/Oral Microbiology, Umeå University, Umeå, Sweden
A B S T R A C T
Background and objectives. Helicobacter pylorishows a characteristic tropism for the mucus-producing gastric epithelium. In infected patients,H. pylori colocalizes in situ with the gastric secretorymucin MUC5AC. The carbohydrate blood-groupantigen Lewis B (LeB) was deemed responsible forthe adherence of H. pylori to the gastric surfaceepithelium. We sought to determine if MUC5AC isthe carrier of LeB, and thus if MUC5AC is theunderlying gene product functioning as the mainreceptor for H. pylori in the stomach.Methods. We studied three types of human tissueproducing MUC5AC: Barrett’s esophagus (BE),normal gastric tissue, and gastric metaplasia of theduodenum (GMD). Tissue sections were immuno-fluorescently stained for MUC5AC or LeB, andsubsequently incubated with one of three strains ofTexas red-labeled H. pylori, one of which was unableto bind to LeB. We determined the colocalization ofMUC5AC or LeB with adherent H. pylori.
Results. The binding patterns for the two LeB-binding strains to all tissues were similar, whereasthe strain unable to bind to LeB did not bind to anyof the tissues. In normal gastric tissue, the LeB-binding strains always bound to MUC5AC- and LeB-positive epithelial cells. In four nonsecretor patients,colocalization of the LeB-binding strains was foundto MUC5AC-positive gastric epithelial cells. In BE,the LeB-binding H. pylori strains colocalized veryspecifically to MUC5AC-positive cells. MUC5AC-producing cells in GMD contained LeB. Yet, LeB-binding H. pylori not only colocalized to MUC5ACor LeB present in GMD, but also bound to the LeB-positive brush border of normal duodenal epithelium.Conclusions. Mucin MUC5AC is the most importantcarrier of the LeB carbohydrate structure in normalgastric tissue and forms the major receptor for H. pylori.Keywords. Adhesion, Helicobacter pylori, mucin,MUC5AC, Barrett’s esophagus, gastric metaplasia,Lewis B blood group antigen.
Since the discovery of Helicobacter pylori [1,2],chronic infection with this bacterium has
been identified as the major etiological factor ingastritis, gastric ulcers, gastric atrophy, and gas-tric carcinoma [3,4]. As a result, in 1994, theInternational Agency for Research on Cancer(IARC, Lyon, France), classified H. pylori infec-tion as a carcinogenic agent class I.
Within the human body, H. pylori residesprimarily in the gastric mucus layer. This mucuslayer consists mainly of the gel-forming mucinMUC5AC, which is, like all gel-forming mucins,a large, heavily O-glycosylated glycoprotein [5].
The gene encoding MUC5AC has been cloned[6–8], and is expressed by the surface mucous cellsof the gastric glands [9]. In addition to MUC5AC,the normal antral mucosa of the stomach expressesthe secretory mucin MUC6, and these two mucinshave distinct expression patterns, being expressedby different subsets of cells within the antralglands. MUC6 is mainly expressed in the deeperantrum gland epithelium, whereas MUC5AC isexpressed by the surface epithelium [10].
Different blood-group antigens, especially Lewis(Le) antigens, are correlated with the distributionof normal gastric mucin expression. Le blood-group antigens are carbohydrate structures carriedon both glycoproteins and glycolipids. They arewidely distributed in the body, and found notonly on erythrocytes, but also in secretions and
Reprint requests to: Jan Dekker, Laboratory of Pediatrics,Erasmus MC, Rm EE-15-71A, Dr Molewaterplein 50, 3015GE Rotterdam, the Netherlands.
Print & Media
522 Van de Bovenkamp et al.
© 2003 Blackwell Publishing Ltd, Helicobacter, 8, 521–532
particularly on epithelial cells. The antigenic spe-cificity is determined by relatively small differ-ences in the sugar composition and interresiduelinkage of the oligosaccharide moieties [11].These antigenic structures are carried by twotypes of O-glycan backbone structures, namedtype 1 [containing Gal(β1-3)GlcNAc] and type2 [containing Gal(β1-4)GlcNAc]. Fucosylationof the type 1 backbone structure leads to theexpression of the blood group antigens Lewis A(LeA), Lewis B (LeB), and H-1, whereas fuco-sylation of type 2 chain leads to the expressionof the blood group antigens Lewis X (LeX),Lewis Y (LeY), and H-2 [12]. In the humanstomach, the expression of MUC5AC is associ-ated with the type 1 blood group structures LeAand LeB, and MUC6 expression is tightly asso-ciated with type 2 antigens LeX and LeY [13].
A number of molecules have been implicatedas receptors for H. pylori adhesins (reviewed inEvans et al. [14]). The LeB structure was shownto mediate the attachment of H. pylori to humangastric epithelium, when assayed on tissue sections[15]. The H. pylori adhesin, which interacts withLeB, has been cloned and designated ‘blood groupantigen-binding adhesin’ (BabA). It belongs to agene family with approximately 30 members withinthe H. pylori genome [16,17]. Separate studieswith transgenic mice expressing the human LeBstructure in their gastric epithelial cells indicatedthat LeB can function as a receptor for H. pyloriadhesins and mediates attachment of H. pylori togastric pit cells and surface mucous cells [18].
We previously demonstrated that H. pylori inthe antrum of infected patients colocalizes in situwith extracellular MUC5AC as well as with theapical domain of MUC5AC-producing cells ofthe superficial epithelium [10]. As the expressionof MUC5AC is tightly associated with the pres-ence of the LeB antigen in the human stomach,we suggest that MUC5AC, as the main carrierof LeB, constitutes the primary receptor forH. pylori attachment in the human stomach. Totest this experimentally, we determined the colo-calization of LeB and MUC5AC in differenttissues of the upper gastrointestinal tract: nor-mal stomach, esophagus and duodenum. We alsoexamined metaplastic tissue of the esophagus(Barrett’s esophagus, BE) and of the duodenum(gastric metaplasia of the duodenum, GMD),which were both demonstrated to expressMUC5AC within specific cells [19,20].
Adherence of H. pylori to MUC5AC or LeBwas studied by an in vitro binding assay with
fluorescently labeled H. pylori on sections ofeach of these tissues. We used three strains ofH. pylori with well-defined LeB-binding char-acteristics: (1) CCUG17875, which avidly bindsLeB [16]; (2) P466, strongly binding LeB [15];(3) H. pylori TIGR26695, which is unable to bindto LeB [16].
With the data described, we will show that theMUC5AC and LeB antigens usually stronglycolocalize in the epithelia of the human uppergastrointestinal tract. The two LeB-bindingH. pylori strains applied to the tissue sectionscolocalized to both MUC5AC and LeB, whereasthe non-LeB-binding strain did not bind. Thus,it is probable that MUC5AC, as the main carrierof LeB in the gastric epithelium, is the mainreceptor for H. pylori in the human stomach.
Materials and Methods
Patients and Tissue
Biopsy specimens of normal (i.e. non-H. pylori-infected, noninflamed) antrum (n = 20) werecollected per endoscopy, as part of routine diag-nostic endoscopy for upper gastrointestinalcomplaints. Biopsy specimens from the duo-denal bulb (n = 12) were retrieved from archive,and were selected for the presence of gastricmetaplasia, as determined previously [19]. Thebiopsy specimens were collected in the Aca-demic Medical Center in Amsterdam, withpermission of the Medical Ethics Committee.Two biopsy specimens were available from theantrum and duodenal bulb per patient, and wereimmediately fixed in phosphate-buffered saline(PBS)/4% (w/v) paraformaldehyde solution for4 hours, and then processed in paraffin blocksaccording to standard procedures. Tissue samplesof normal esophagus and Barrett’s esophagus(BE, n = 14) were retrieved from the archive ofthe Gastrointestinal Pathology Unit, CatholicUniversity Leuven, Belgium. Per patient, four tonine endoscopic biopsy specimens were avail-able, fixed in Carnoy’s fixative, and paraffin-embedded. H. pylori infection was assessed onantrum biopsy specimens for all patients, and thebacterium was detected using both standard his-tologic staining and microbiology.
H. pylori Strains
The following H. pylori strains were used(Table 1): 1, strain CCUG17875 from the
Print & Media
Adhesion of H. pylori to Human Gastric Mucin MUC5AC 523
© 2003 Blackwell Publishing Ltd, Helicobacter, 8, 521–532
Culture Collection at the University of Göteborg(which binds strongly to LeB); 2, strain P466(which binds strongly to LeB) [21]; 3, strainTIGR26695 from The Institute for GenomicResearch (which is unable to bind to LeB), thegenome of which has been fully sequenced[17].
H. pylori strains were grown at 37°C on Bru-cella agar supplemented with 10% (v/v) bovineblood and 1% (v/v) IsoVitalex (Becton Dickin-son, Franklin Lakes, NJ) under microaerophilicconditions (5% O2/10% CO2/85% N2) and98% humidity. Two days after inoculation, thebacteria were harvested from a plate and gentlyresuspended in 500 µl PBS/0.2% (v/v) Tween 20(Sigma, St. Louis, MO), and subsequently usedin the fluorescent labeling procedure.
Binding of 125I-labeled LeB Glycoconjugate H. pyloriStrains
The binding assay was performed as previouslydescribed by Ilver et al. [16]. In short, LeB antigenglycoconjugate was prepared by the conjugationof purified fucosylated oligosaccharide to humanserum albumin (IsoSep AB, Tullinge, Sweden).LeB glycoconjugate was labeled with 125I bythe chloramine-T method. One milliliter ofH. pylori bacteria (OD600 = 0.10) was incubatedwith 400 ng of 125I-labeled LeB glycoconjugatefor 30 minutes in PBS containing 0.5% (w/v)bovine serum albumin (BSA)/0.05% (v/v) Tween20. After centrifugation, the radioactivity boundto the H. pylori pellet was measured with a gamma-counter. Binding experiments were reproducibleand performed in triplicate.
Texas Red Labeling of H. pylori
H. pylori suspensions were centrifuged at3000 × g for 5 minutes. Each pellet of bacteriawas resuspended in 500 µl of 0.15 mol/l NaCl
and 0.1 mol/l of sodium carbonate, pH 9.0, bygently pipetting. Ten microliters of a 20 mg/mlsolution of Texas Red (TR; Sigma), freshly pre-pared in acetonitrile, was added to 1 ml of bacte-rial cells (OD600 = 1.0), which was followed byincubation for 10 minutes at room temperaturein the dark. The bacteria were recovered bycentrifugation at 3000 × g for 5 minutes, resus-pended in 1 ml PBS/0.2% (v/v) Tween 20, andpelleted by centrifugation as above. The washingcycle was repeated three times. After the lastwashing cycle, the bacteria were resuspendedin PBS/1% (w/v) BSA, and were diluted to1.00 OD600. The intensity of TR labeling of allbacterial strains in each preparation was similar,as determined by inspection of comparable numbersof organisms by fluorescence microscopy.
Immunohistochemistry and H. pylori Binding to Tissue Sections
Tissue sections were deparaffinized throughthree changes of xylene and then rehydratedthrough a series of decreasing concentrationsof ethanol solution to distilled water. Antigenretrieval was performed by heating the sectionsfor 10 minutes at 100
°C in 10 mmol/l citratebuffer, pH 6.0, and then they were left to cool toroom temperature for 20 minutes. Sections werewashed three times for 5 minutes in PBS andincubated with blocking buffer [10 mmol/l Tris,5 mmol/l EDTA, 150 mmol/l NaCl, 0.25% (w/v) gelatin and 0.05% (v/v) Tween 20], pH 8.0, for30 minutes, followed by washing three times for5 minutes in PBS/0.2% (v/v) Tween 20. Primarymonoclonal antibody diluted in PBS wasadded, either anti-MUC5AC (45M1, Novocastra,Newcastle, UK; diluted 1:50), or anti-Lewis B(Signet Laboratories, Dedham, MA; diluted1:40), and incubated for 16 hours at 4°C. Sec-tions were washed three times for 5 minutes inPBS/0.2% (v/v) Tween 20, and then incubatedfor 60 min in the dark with PBS-diluted (1:50),fluorescein isothiocyanate (FITC)-labeled anti-mouse secondary antibody (DAKO, Glostrup,Denmark). Again, the sections were washedthree times for 5 minutes in PBS/0.2% (v/v)Tween 20. Per section, 40 µl [1.0 × 108 colony-forming units (cfu)/ml] of TR-labeled H. pyloriwas added, and then incubated for 60 minutesin the dark at room temperature. Slides werewashed 5 times for 5 minutes in PBS. Finally,sections were covered in Vectashield (VectorLaboratories, Burlingame, UK), and were mounted
Table 1 Characteristics of H. pylori strains used in this study
H. pylori strainType I
(CagA/VacA) BabA References
CCUG17875 + + [16,50]P466 + + [16,51]TIGR26695 + – [16,17]
Each of the three H. pylori strains is a type I strain, and carries the CagApathogenicity islet as well as the VacA toxin gene (indicated by +). Thepresence of a functional BabA protein is also indicated (+, present;–, absent), which constitutes the structural gene for the BabA adhesinprotein, the babA1 gene being a nonfunctional pseudo-gene [16].
Print & Media
524 Van de Bovenkamp et al.
© 2003 Blackwell Publishing Ltd, Helicobacter, 8, 521–532
under cover-slips. Colocalization of immuno-fluorescent staining of tissue antigens and bacteriawas analyzed by fluorescent microscopy using aTR/FITC double filter block (Nikon, Tokyo,Japan). Control stainings were performed leav-ing each of the primary antibodies out of theprocedure, resulting in an absence of FITCstaining. All binding assays on the tissue of eachpatient were performed at least in duplicate. Theresults were highly reproducible.
Separately, closely adjacent sections of eachtissue sample of each patient were stained forthree other major secretory mucins, MUC2(monoclonal antibody: WE9, a gift of Prof. D.K. Podolsky, Massachusetts General Hospital,Boston, MA; diluted 1:50) [22], MUC5B (poly-clonal antibody: BGBM, a gift of Dr G. Offner,Boston University Medical Center, Boston, MA;diluted 1:3000) [23], and MUC6 (polyclonalantibody: 6.1, a gift of Dr C. de Bolos, IMIMHospital del Mar, Barcelona, Spain; diluted1:200) [13]. Single immunohistochemical stain-ing was performed using specific antibodiesagainst each of these mucins and by standardhorseradish peroxidase/diamino-benzidine-basedstaining according to previously reported proce-dures [19,20].
Results
Binding of Soluble LeB Conjugate to H. pylori
To ascertain the expression of functional LeBantigen-binding adhesin (BabA), each of theH. pylori strains was examined for binding ofsoluble LeB conjugate. Strains CCUG17875 andP466 were able to bind soluble LeB conjugate,whereas TIGR26695 showed only very weakbinding (Table 2). This binding experiment wasperformed in triplicate and showed high repro-ducibility, which confirmed expression of func-tional BabA for strains CCUG17875 and P466.
Binding of H. pylori to Gastric Epithelium
Antrum biopsies were collected from 20patients. None of these patients suffered fromH. pylori infection as determined by standardhistochemistry and microbiology, and the tissueshowed no signs of acute or chronic inflamma-tion as assessed by standard hematoxylin andeosin staining (data not shown). Abundantexpression of the MUC5AC glycoprotein in theepithelial cells of the surface and of the neck ofthe antrum glands could be identified in eachpatient. In addition, extracellular remnants of thegastric mucus layer were observed, and thesewere also MUC5AC-positive. As reported ear-lier [24], staining of adjacent tissue sections ofeach patient with antibodies against the othermajor secretory mucins, MUC2, MUC5B andMUC6, showed that the MUC5AC-positivecells did not contain any of these other secretorymucins (data not shown).
Sixteen of the 20 patients showed expressionof the blood group antigen LeB in the gastricsurface epithelium, as determined with immuno-fluorescent staining. If a patient expressed LeB,the expression colocalized very strongly withMUC5AC-positive cells as well as with extra-cellular MUC5AC-positive mucus (Fig. 1). In all20 patients, binding of strains CCUG17875 andP466 was observed to the MUC5AC-positiveepithelial cells of the surface and the neckregion of the glands and also to the MUC5AC-positive extracellular mucus (Fig. 1A,B). StrainTIGR26695, which was unable to bind to solubleLeB conjugate, showed no binding to theepithelium or the mucus in any of the patients(Fig. 1C). As LeB showed a very strong colocal-ization to MUC5AC, strains CCUG17875 andP466 bound very specifically to the LeB-containingcells and mucus, while strain TIGR26695 didnot bind to LeB-positive structures in the tissue(Fig. 1D–F). Interestingly, in four patients, inwhom we were not able to detect LeB by immuno-fluorescence, the strains CCUG17875 and P466bound to MUC5AC-positive cells and mucus,while again no binding was found for strainTIGR26695 (data not shown).
When bacteria were added to the tissue sec-tions without prior staining by antibodiesdirected toward either MUC5AC or LeB, morebacteria were able to bind to MUC5AC-containing cells and mucus (data not shown).This effect was only found for the LeB-bindingstrains CCUG17875 and P466, while the
Table 2 Binding of soluble LeB glycoconjugate to different H. pylori strains
H. pylori strain Binding (%) Range (%)
CCUG17875 28.6 27.2–30.0P466 21.8 20.7–22.9TIGR26695 0.7 0.6–0.8
Binding of H. pylori in suspension to soluble 125I-labeled LeB glycoconjugatewas measured by gamma counting of bacteria that were pelleted after1 hour of incubation and extensive washing. Binding is expressed aspercentage of the total amount of 125I conjugate added to each of thebacterial suspensions. Data are presented as mean percentage of maximalradioactivity and range of three separate experiments.
Print & Media
Adhesion of H. pylori to Human Gastric Mucin MUC5AC 525
© 2003 Blackwell Publishing Ltd, Helicobacter, 8, 521–532
binding behavior of TIGR26695 for the tissuesections was not changed by previous stainingwith these antibodies.
Binding of H. pylori to Esophagus and BE Epithelium
Recently we showed that BE epithelium expressedthe gastric-type mucin MUC5AC [20]. We wereinterested as to whether H. pylori was also ableto bind to MUC5AC or LeB in sections of BEtissue and normal esophagus epithelium. Thenormal stratified epithelium of the esophagus didnot express MUC5AC or LeB, and, moreover,this epithelium showed no adherence of any bac-teria upon addition of the TR-labeled H. pyloristrains to the tissue sections (data not shown).
MUC5AC was expressed abundantly in tissuesections of all BE patients (n = 14), and waslocalized in the surface epithelium of BE, and toa variable extent deeper into the glandular struc-tures of the BE epithelium. In the patients inwhom LeB expression was detected by immuno-fluorescence (seven of 14 patients), the expres-sion pattern of LeB colocalized very stronglywith the MUC5AC expression pattern (Fig. 2).In each BE patient, binding to the BE epithelium
was observed for both strains CCUG17875 andP466. The binding of these strains of H. pyloriperfectly colocalized with MUC5AC-positivecells in the BE epithelium, whereas strainTIGR26695 showed no specific binding to BEepithelium (Fig. 2A–C). As LeB stronglycolocalized with MUC5AC in these BE tissues,the same binding pattern was found as forMUC5AC, i.e. strains CCUG17875 and P466were found to specifically bind to the LeB-positive epithelium, whereas strain TIGR26695showed no specific binding to the LeB-positiveepithelium (Fig. 2D–F).
Interestingly, in the tissue of seven individualsidentified with BE in whom LeB could not bedemonstrated by immunofluorescence, specificbinding of strains CCUG17875 and P466 wasfound to the MUC5AC-positive epithelialcells of BE (Fig. 3A–D). In these LeB-negativetissues, strain TIGR26695 showed no specificbinding to the BE epithelium (data not shown).
As in normal gastric tissues, it appeared thatBE tissues, to which no antibodies were bound,bound more bacteria than sections that wereeffectively stained with antibodies. For example,when BE tissue sections did not bind any anti-LeB
Figure 1 Binding of H. pylori togastric epithelium, showing an exampleof double immuno-fluorescence onhuman gastric antrum epithelium, ofclosely adjacent sections of a LeB-positive patient. Colocalization ofTR-labeled H. pylori (in red) addedto sections immunohistochemicallystained for either MUC5AC (ingreen; A–C) or LeB (in green; D–F)was studied. MUC5AC-positive cellscolocalized very strongly with LeB-positive cells in the surface epithelium.H. pylori strains (A) CCUG17875and (B) P466 bound specifically toMUC5AC-positive cells and toextracellular MUC5AC-positive mucus(arrows in A and B, respectively) andalso to LeB-positive cells and LeB-positive extracellular mucus (arrowsin D and E, respectively). StrainTIGR26695 showed no specificbinding to any particular structurein the antrum tissue (C and F).Magnification ×40.
Print & Media
526 Van de Bovenkamp et al.
© 2003 Blackwell Publishing Ltd, Helicobacter, 8, 521–532
antibodies (as in Figs 3B,D), these sectionsbound more bacteria than after effective bindingof the anti-MUC5AC antibodies (compareFig. 3A with B, and Fig. 3C with D). This effectwas found for both the LeB-binding strainsCCUG17875 and P466 (Fig. 3A–D), while thebinding characteristics of TIGR26695 were notchanged by previous incubation with either ofthese antibodies (data not shown).
Binding of H. pylori to Small Intestine Tissue and GMD
Previously, we and other authors demonstratedthat MUC5AC in the duodenum was restrictedto cells of the GMD and occasional goblet cellswithin the normal small intestine epithelium[19,25]. Immunohistochemical staining for othersecretory mucins on tissue sections of the GMDpatients in this study revealed that MUC2was expressed in all goblet cells of the smallintestine epithelium, and that Brunner’s glands,which were present in most of the biopsy speci-mens, produced MUC5B and MUC6 but notMUC5AC or MUC2 (data not shown), as wedemonstrated previously for a large number ofindividuals [19].
Twelve patients were selected from the archivein whom gastric metaplasia was detected in duo-denal bulb biopsy specimens. MUC5AC expres-sion was unequivocally found in these patientsin patches of metaplastic cells, and in a few gobletcells within the villi of the normal epithelium(Fig. 4). In contrast to both antrum and BE tis-sue, H. pylori only adhered to duodenal tissuesamples in which we were able to demonstrateLeB by immunofluorescence (i.e. in nine of 12patients). In these tissue samples, LeB was foundin the cells of the GMD, in the brush border ofthe normal enterocytes and in a subset of thenormal goblet cells. In nine (of 12) patients inwhom we detected LeB-positive small intestinetissue, there was colocalization between LeB andMUC5AC in the epithelial cells of the GMD andin occasional goblet cells of the normal duodenalepithelium (Fig. 4). In the small intestine tissuefrom three patients in which we could detect noLeB, no specific adherence of H. pylori couldbe demonstrated with any of the three strains(data not shown). In LeB-positive tissue, strainsCCUG17875 and P466 showed specific bindingto MUC5AC-positive cells of the GMD epithe-lium and to the occasional MUC5AC-positive
Figure 2 Binding of H. pylori toBarrett’s esophagus (BE) epithelium.(A–F) An example of double immuno-fluorescence on LeB-positive humanmetaplastic esophagus epithelium.Colocalization of TR-labeled H. pylori(in red) added to sectionsimmunohistochemically stained foreither MUC5AC (in green; A–C) orLeB (in green; D–F) was studied.MUC5AC-positive cells colocalizedvery strongly with LeB-positive cellsin the BE epithelium. H. pylori strainsCCUG17875 and P466 boundspecifically to MUC5AC-positivecells (A and B, respectively) andalso to LeB-positive cells (D and E,respectively). Strain TIGR26695 (Cand F) showed no specific binding toany particular structure in this LeB-positive BE tissue. Magnification ×40.
Print & Media
Adhesion of H. pylori to Human Gastric Mucin MUC5AC 527
© 2003 Blackwell Publishing Ltd, Helicobacter, 8, 521–532
goblet cells (Fig. 4A,B,G,H). No specific bind-ing of strain TIGR26695 to MUC5AC-positivecells of the GMD or to MUC5AC-positivegoblet cells in the normal epithelium was found(Fig. 4C,I).
In GMD, there was a strong colocalization ofLeB with MUC5AC, and therefore the adher-ence of the three H. pylori strains to the LeB-positive cells of the GMD was very similar tothat described above for MUC5AC (Fig. 4D–F). However, the most extensive LeB expressionwas found in the brush border of the intestinalenterocytes, and in the goblet cells of normalduodenum epithelium (Fig. 4J–L). Binding of
strains CCUG17875 and P466 was found to thebrush border of enterocytes, and to the gobletcells in those tissue samples in which we wereable to show LeB, colocalizing specifically withLeB in the brush border and these goblet cells(Fig. 4J,K). Strain TIGR26695 showed no spe-cific adherence to the LeB-positive brush borderof the enterocytes or to goblet cells (Fig. 4L).Upon immunohistochemical staining of theduodenal tissue sections with antibodies againstMUC5AC or LeB, the Brunner’s glands in themucosa of the duodenum were not stained forMUC5AC or LeB, respectively. None of thethree H. pylori strains showed any adherence tothe epithelial cells of the Brunner’s glands, bothin LeB-positive and -negative small intestine tis-sue samples, as detected by immunofluorescence(data not shown).
Discussion
Helicobacter pylori is known to reside in the gas-tric mucus layer, often close to or attached to themucus-producing epithelium [10,26,27]. There-fore, it is of great interest to establish what isso special about this gastric mucus and thesegastric mucus-producing cells that H. pylori isable to colonize this niche. As we demonstratedpreviously, H. pylori in the antrum of infectedpatients always colocalizes very specifically withMUC5AC and MUC5AC-producing cells [10].MUC5AC is a secretory, mucus-forming mucinthat is characteristically produced in very largeamounts in the gastric mucosa, making it a likelycandidate for a H. pylori receptor. Another well-characterized gastric receptor for H. pylori is theLewis B blood group structure, which is a carbo-hydrate structure primarily present on O-linkedglycans [15]. Mucins are heavily O-glycosylatedstructures and were demonstrated to carry,among others, Lewis antigens, including LeB[28]. In order to make the two ends of this storymeet, our hypothesis is that MUC5AC is the pri-mary gene product that carries the LeB carbohy-drate structures, which may form the footholdfor H. pylori in one of the steps necessary to col-onize the human stomach. To test this theory, westudied the colocalization of H. pylori added totissue sections in the presence of both MUC5ACand LeB within the tissue. Colocalization ofH. pylori to both these entities would supportour hypothesis.
First, we have to consider the uniqueness ofMUC5AC as the mucus-forming mucin in the
Figure 3 Binding of H. pylori to LeB-negative Barrett’sesophagus (BE) epithelium. (A–D) An example of adherenceof H. pylori to sections from a patient with LeB-negativeBE epithelium. These sections were immunohistochemicallystained for either MUC5AC (in green; A and C) or LeB (ingreen; B and D), and then incubated with the TR-labeledH. pylori (in red). The sections in (A) and (B) were incubatedwith CCUG17875, and the sections in (C) and (D) wereincubated with P466. Strain TIGR26695 showed no specificbinding to any structure in this LeB-negative BE tissue (data notshown). Magnification ×40.
Print & Media
528 Van de Bovenkamp et al.
© 2003 Blackwell Publishing Ltd, Helicobacter, 8, 521–532
human stomach. Gastrointestinal mucus isgenerally composed of secretory mucins of theMUC family, of which four members are gener-ally considered to have the ability to form mucusgels: MUC2, MUC5AC, MUC5B, and MUC6[29]. Each of these mucins shows a very distinctexpression pattern along the human gastrointes-tinal tract [30,31], indicating distinct and specificfunctions for each of these four secretorymucins. In the normal esophagus, only MUC5Bis produced in the submucosal glands, and thesurface epithelium does not produce gel-formingMUCs [32]. However, within Barrett’s esopha-gus (BE), which constitutes the gastric-type
and intestinal-type metaplastic epithelium of theesophagus, all four mucus-forming MUCs areexpressed, although MUC2 is restricted tointestinal-type metaplasia in BE [20]. The epi-thelium of the gastric antrum, which forms thenatural habitat of H. pylori in infected patients,produces MUC5AC in the superficial epithelialcells and the cells of the gastric pits [10,13,33].MUC5B and MUC6 are produced in the epithe-lial cells lower in the antral glands [10,13,33,34],whereas MUC2 is not produced at all in thenormal stomach. At the exit of the stomach, theMUC expression pattern again changes dramat-ically. The small intestine epithelium contains
Figure 4 Binding of H. pylori tonormal and gastric metaplasticepithelium (GMD) of the duodenum,showing an example of doubleimmuno-fluorescence on GMDepithelium (A–F) and normal villusepithelium of the duodenum (G–L).Colocalization of TR-labeled H. pylori(in red) added to sections immuno-histochemically stained for eitherMUC5AC (in green; A–C and G–I) orLeB (in green; D–F and J–L) wasstudied. H. pylori strains added tothe sections were: CCUG17875(A, D, G, J), P466 (B, E, H, K) andTIGR26695 (C, F, I, L). Some gobletcells in the normal duodenal villiwere MUC5AC-positive, indicated byarrows in (G–I). Magnification ×40.
Print & Media
Adhesion of H. pylori to Human Gastric Mucin MUC5AC 529
© 2003 Blackwell Publishing Ltd, Helicobacter, 8, 521–532
characteristic goblet cells expressing MUC2[35]. The Brunner’s glands, which are character-istic for the proximal duodenum, produce bothMUC5B and MUC6 [24,33,34]. MUC5AC isonly found in rare goblet cells in the normalintestinal villi; however, MUC5AC is expressedunequivocally in all cells of the gastric metaplasiathat are found in the proximal duodenum[19,25]. Thus, MUC5AC is very characteristic ofthe normal habitat of H. pylori, that is the surfaceepithelium of the normal stomach, whereas it isnormally absent from both the esophagus andthe duodenum.
Secondly, we have to consider mucins as car-riers of carbohydrates, in particular the humanLewis-type blood group antigens. The expres-sion of Lewis antigens is related to the ABOblood groups, and is dependent on the ‘secretor’status of the individual. LeB, like Lewis Y, is onlyexpressed in those individuals (‘secretors’) thatproduce the α(1,2)-fucosyltransferase necessaryto confer the H-structure to the glycans [36,37].In the Western population, about 80% of indi-viduals are ‘secretors’, and are able to generate LeBantigens on their carbohydrates. Nonsecretorswho lack the respective α(1,2)-fucosyltransferasehave only very low amounts of LeB on theircarbohydrates, in the range of 1–5% of the levelsof secretors [12]. LeB is present abundantly inthe gastric mucus of secretors, as a carbohy-drate side-chain of MUC5AC [28], but it is alsoexpressed in the normal esophagus [38], theduodenum [38,39], and in some areas of the meta-plasia in the esophagus and intestine [40,41].Therefore, the distribution of LeB in the uppergastrointestinal tract is more widespread thanthe distribution of MUC5AC. The affinity ofH. pylori for LeB has been demonstrated in anin situ adherence assay. In this assay, freshH. pylori adhered to fixed sections of humangastric tissue, whereas both soluble LeB and amonoclonal antibody to LeB interfered with thisadhesion [15]. The bacterial adhesin responsiblefor LeB antigen binding, BabA, was morerecently identified [16]. Although the exactpathophysiological role remains to be elucidatedfor infection of the human stomach, the BabA–LeB interaction may be an important factor inthe outcome of infection, as it worsened theseverity of gastritis in transgenic mice expressinghuman LeB [42]. Nevertheless, both the trans-genic animals carrying the LeB structure andtheir LeB-negative controls were eventuallycolonized by H. pylori. Moreover, it was shown
very recently that the H. pylori babA2 gene isstrongly associated with H. pylori strains thathave caused severe disease [43]. This suggeststhat H. pylori strains that colonize the epithe-lium without the expression of BabA adhesincause a less severe gastric disease. Some workersshowed that LeB may not play an essential rolein the colonization process of H. pylori in vivoin patients [44] and in vitro experiments in amodel system using isolated human gastric epi-thelial cells [45]. It seems likely that the adher-ence of H. pylori via its BabA receptor to thehost LeB structures effectively facilitates infec-tion, but is not an absolute necessity for eventualcolonization.
In normal stomach and in the epithelium ofBE, MUC5AC and LeB colocalize to the sameepithelial cells and to the extracellular mucus.The H. pylori bacteria added to the tissue sec-tions of antrum and BE colocalized very well toboth MUC5AC and LeB. Only the two LeB-binding strains (CCUG17875 and P466) boundto the cells and mucus, whereas the strain that isnot able to bind to LeB (TIGR26695) did notbind to the tissue. This strongly suggests thatLeB present on the MUC5AC molecules func-tions as a receptor for H. pylori.
We found indications in both the antrum andthe BE epithelium that H. pylori was sometimesable to bind to epithelial cells that could bestained for MUC5AC, whereas we were unableto detect LeB through immunofluorescence. Inthese seemingly LeB-negative tissues, strainsCCUG17875 and P466 bound to MUC5AC-positive cells, which may indicate that otherstructures of MUC5AC may also function asreceptors. Yet, this binding could only be dem-onstrated for the two strains carrying a func-tional BabA adhesin. The individuals in whomwe were not able to detect LeB were probablynonsecretors, who nevertheless are known toexpress 1–5% of the levels of LeB found in secre-tors [12]. H. pylori is possibly able to bind to thisresidual LeB in the nonsecretor tissues.
In the duodenum, H. pylori is, in infectedpatients, always found closely associated withgastric metaplastic cells [46–49]. Indeed, wewere able to demonstrate colocalization of H.pylori added to duodenal tissue sections withMUC5AC as well as LeB present in metaplasticcells. However, in LeB-positive individualsthe H. pylori bacteria of strains P466 andCCUG17875 also bound to the LeB-positivebrush border on the normal villus enterocytes.
Print & Media
530 Van de Bovenkamp et al.
© 2003 Blackwell Publishing Ltd, Helicobacter, 8, 521–532
These strains demonstrated, at least in theseexperimental circumstances in vitro, the abilityto bind to LeB structures that are not positionedon MUC5AC. Thus, there is a notable discrep-ancy between the binding of H. pylori in vitroand that in vivo when binding to the brushborder is considered. A likely explanation is thatthe H. pylori bacteria that may become attachedto the brush border of the duodenum in vivo donot survive on this surface. It was previouslynoted that the density of infection of the patchesof metaplastic cells in the duodenum was verylow, when compared to population densitiesin the stomach [46–49]. This indicates that thecircumstances for survival of H. pylori in theduodenum are far from optimal. Therefore,H. pylori might only be able to survive in veryclose association with the MUC5AC-producing,gastric-type metaplastic epithelial cells, asdemonstrated by electron microscopy [49].Although H. pylori may be able to attach to theLeB structures of the brush border in vivo, thisis not sufficient for survival, and thereforeH. pylori is not detected in biopsy specimens ofH. pylori-infected individuals in any significantnumbers.
It was found consistently in each of the sepa-rate experiments that the bacteria of the LeB-binding strains, CCUG17875 and P466, boundbetter to the tissue sections when these sectionswere not previously stained with antibodiesagainst MUC5AC or LeB. The antibodies,which bound to either MUC5AC or LeB withinthe tissue sections, apparently hindered thebinding of the bacteria in the second step of ourcolocalization protocol. Although these bindingassays are not quantitative, it is very likely thatH. pylori and the anti-MUC5AC and anti-LeBantibodies compete for binding to the samestructures in the tissue sections. Apparently,these respective antibodies are able to blockthe interaction between the BabA protein onthe H. pylori strains and LeB or MUC5AC in thetissue sections.
In summary, we demonstrated H. pylori bindingto mucin MUC5AC in gastric tissue, Barrett’sesophagus epithelium and gastric metaplasticcells in the duodenum. Only H. pylori strainsthat were able to bind to LeB adhered toMUC5AC in the tissue sections. We haveshown that MUC5AC is probably the mostimportant carrier of LeB structures in thesetissues and forms a major receptor for thebacterium.
This research was financially supported by the Nether-lands Digestive Diseases Foundation (Nieuwegein),Irene Children’s Hospital Foundation (Arnhem), JanDekker/Ludgardine Bouman Foundation (Amster-dam) and Foundation ‘De Drie Lichten’ (Leiden), allin the Netherlands (JHBvdB), and the Swedish Med-ical Research Council (11218) and Swedish CancerSociety (4101-B00–03XAB) (TB). The authors thankProf. Dr N. L. E. Y. Ectors (Leuven, Belgium), DrE. H. Van Beers (Amsterdam, the Netherlands) andT. M. Rolf (Amsterdam, the Netherlands) for obtainingthe human biopsy specimens.
References
1 Marshall BJ. Unidentified curved bacilli on gas-tric epithelium in active chronic gastritis (letter).Lancet 1983;1:1273–5.
2 Warren JR. Unidentified curved bacilli on gastricepithelium in active chronic gastritis (letter). Lancet1983;1:1273.
3 Marshall BJ, Warren JR. Unidentified curvedbacilli in the stomach of patients with gastritis andpeptic ulceration. Lancet 1984;1:1311–5.
4 Parsonnet J, Friedman GD, Vandersteen DP, et al.Helicobacter pylori infection and the risk of gastriccarcinoma. N Engl J Med 1991;325:1127–31.
5 Strous GJ, Dekker J. Mucin-type glycoproteins.Crit Rev Biochem Mol Biol 1992;27:57–92.
6 Klomp LW, Van Rens L, Strous GJ. Cloning andanalysis of human gastric mucin cDNA revealstwo types of conserved cysteine-rich domains.Biochem J 1995;308:831–8.
7 Van de Bovenkamp JH, Hau CM, Strous GJ,Büller HA, Dekker J, Einerhand AW. Molecularcloning of human gastric mucin MUC5ACreveals conserved cysteine-rich d-domains and aputative leucine zipper motif. Biochem BiophysRes Commun 1998;245:853–9.
8 Escande F, Aubert JP, Porchet N, Buisine MP.Human mucin gene MUC5AC. organization ofits 5′-region and central repetitive region. Bio-chem J 2001;358:763–72.
9 Nordman H, Davies JR, Lindell G, Carlstedt I.Human gastric mucins – a major population iden-tified as MUC5. Biochem Soc Trans 1995;23:533S.
10 Van den Brink GR, Tytgat KM, Van Der HulstRW, et al. H. pylori colocalises with MUC5AC inthe human stomach. Gut 2000;46:601–7.
11 Lloyd KO. Philip Levine award lecture. Bloodgroup antigens as markers for normal differentia-tion and malignant change in human tissues. AmJ Clin Pathol 1987;87:129–39.
12 Sakamoto S, Watanabe T, Tokumaru T, Takagi H,Nakazato H, Lloyd KO. Expression of Lewis A,Lewis B, Lewis X, Lewis Y, sialyl Lewis A, andsialyl Lewis X blood group antigens in humangastric carcinoma and in normal gastric tissue.Cancer Res 1989;49:745–52.
Print & Media
Adhesion of H. pylori to Human Gastric Mucin MUC5AC 531
© 2003 Blackwell Publishing Ltd, Helicobacter, 8, 521–532
13 De Bolos C, Garrido M, Real FX. MUC6 apomu-cin shows a distinct normal tissue distributionthat correlates with Lewis antigen expressionin the human stomach. Gastroenterology 1995;109:723–34.
14 Evans D. Helicobacter pylori adhesins. Rev Per-spect Helicobacter 2000;5:183–95.
15 Boren T, Falk P, Roth KA, Larson G, NormarkS. Attachment of Helicobacter pylori to humangastric epithelium mediated by blood group anti-gens. Science 1993;262:1892–5.
16 Ilver D, Arnqvist A, Ogren J, et al. Helicobacterpylori adhesin binding fucosylated histo-bloodgroup antigens revealed by retagging. Science1998;279:373–7.
17 Tomb JF, White O, Kerlavage AR, et al. The com-plete genome sequence of the gastric pathogenHelicobacter pylori. Nature 1997;388:539–47.
18 Falk PG, Bry L, Holgersson J, Gordon JI. Expres-sion of a human alpha-1,3/4-fucosyltransferasein the pit cell lineage of FVB/N mouse stomachresults in production of Leb-containing glyco-conjugates: a potential transgenic mouse modelfor studying Helicobacter pylori infection. ProcNatl Acad Sci USA 1995;92:1515–9.
19 Van de Bovenkamp JH, Korteland-Van MaleAM, Buller HA, Einerhand AW, Dekker J. Meta-plasia of the duodenum shows a Helicobacterpylori-correlated differentiation into gastric-typeprotein expression. Hum Pathol 2003;34:156–65.
20 Warson C, Van De Bovenkamp JH, Korteland-VanMale AM, et al. Barrett’s esophagus is charac-terized by expression of gastric-type mucins(MUC;5AC:MUC6) and TFF peptides (TFF1and TFF2), but the risk of carcinoma develop-ment may be indicated by the intestinal-typemucin, MUC2. Hum Pathol 2002;33:660–8.
21 Falk P, Roth KA, Boren T, Westblom TU, Gordon JI,Normark S. An in vitro adherence assay revealsthat Helicobacter pylori exhibits cell lineage-specific tropism in the human gastric epithelium.Proc Natl Acad Sci USA 1993;90:2035–9.
22 Tytgat KM, Klomp LW, Bovelander FJ, et al.Preparation of anti-mucin polypeptide antiserato study mucin biosynthesis. Anal Biochem1995;226:331–41.
23 Van Klinken BJ, Dekker J, van Gool SA, vanMarle J, Buller HA, Einerhand AW. MUC5B isthe prominent mucin in human gallbladder and isalso expressed in a subset of colonic goblet cells.Am J Physiol 1998;274:G871–8.
24 Van de Bovenkamp JHB, Korteland-Van MaleAM, Büller HA, Einerhand AWC, Dekker J.Gastric H. pylori infection alters the expressionpatterns of all major secretory proteins in theantrum (Abstract). Gastroenterology 2001;120:A669.
25 Shaoul R, Marcon P, Okada Y, Cutz E, Forstner G.The pathogenesis of duodenal gastric metaplasia:
the role of local goblet cell transformation. Gut2000;46:632–8.
26 Shimizu T, Akamatsu T, Sugiyama A, Ota H,Katsuyama T. Helicobacter pylori and the surfacemucous gel layer of the human stomach. Helico-bacter 1996;1:207–18.
27 Shimizu T, Akamatsu T, Ota H, Katsuyama T.Immunohistochemical detection of Helicobacterpylori in the surface mucous gel layer and itsclinicopathological significance. Helicobacter1996;1:197–206.
28 Nordman H, Davies JR, Lindell G, De Bolos C,Real F, Carlstedt I. Gastric MUC5AC and MUC6are large oligomeric mucins that differ in size,glycosylation and tissue distribution. Biochem J2002;364:191–200.
29 Dekker J, Rossen JW, Buller HA, Einerhand AW.The MUC family: an obituary. Trends BiochemSci 2002;27:126–31.
30 Van Klinken BJ, Dekker J, Buller HA, de BolosC, Einerhand AW. Biosynthesis of mucins(MUC2-6) along the longitudinal axis of thehuman gastrointestinal tract. Am J Physiol1997;273:G296–302.
31 Reid CJ, Harris A. Developmental expression ofmucin genes in the human gastrointestinal sys-tem. Gut 1998;42:220–6.
32 Arul GS, Moorghen M, Myerscough N, AldersonDA, Spicer RD, Corfield AP. Mucin gene expres-sion in Barrett’s oesophagus: an in situ hybridisa-tion and immunohistochemical study. Gut 2000;47:753–61.
33 Ho SB, Roberton AM, Shekels LL, Lyftogt CT,Niehans GA, Toribara NW. Expression cloningof gastric mucin complementary DNA and local-ization of mucin gene expression. Gastroenter-ology 1995;109:735–47.
34 Longman RJ, Douthwaite J, Sylvester PA, et al.Coordinated localisation of mucins and trefoilpeptides in the ulcer associated cell lineage andthe gastrointestinal mucosa. Gut 2000;47:792–800.
35 Chang SK, Dohrman AF, Basbaum CB, et al.Localization of mucin (MUC2 and MUC3) mes-senger RNA and peptide expression in humannormal intestine and colon cancer. Gastroenter-ology 1994;107:28–36.
36 Lopez-Ferrer A, de Bolos C, Barranco C, et al.Role of fucosyltransferases in the associationbetween apomucin and Lewis antigen expressionin normal and malignant gastric epithelium. Gut2000;47:349–56.
37 Kelly RJ, Rouquier S, Giorgi D, Lennon GG,Lowe JB. Sequence and expression of a candidatefor the human secretor blood group alpha (1,2)fucosyltransferase gene (FUT2). Homozygosityfor an enzyme-inactivating nonsense mutationcommonly correlates with the non-secretor phe-notype. J Biol Chem 1995;270:4640–9.
Print & Media
532 Van de Bovenkamp et al.
© 2003 Blackwell Publishing Ltd, Helicobacter, 8, 521–532
38 Davidson JS, Triadafilopoulos G. Blood group-related antigen expression in normal and meta-plastic human upper gastrointestinal mucosa.Gastroenterology 1992;103:1552–61.
39 Mollicone R, Bara J, Le Pendu J, Oriol R. Immuno-histologic pattern of type 1 (Lea, Leb) and type2 (X:Y, H) blood group-related antigens in thehuman pyloric and duodenal mucosae. LabInvest 1985;53:219–27.
40 Murata K, Egami H, Shibata Y, Sakamoto K,Misumi A, Ogawa M. Expression of blood group-related antigens, ABH, Lewis (a), Lewis (b),Lewis (x), Lewis (y), CA19-9, and CSLEX1 inearly cancer, intestinal metaplasia, and uninvolvedmucosa of the stomach. Am J Clin Pathol 1992;98:67–75.
41 Kobayashi K, Sakamoto J, Kito T, et al. Lewisblood group-related antigen expression in normalgastric epithelium, intestinal metaplasia, gastricadenoma, and gastric carcinoma. Am J Gastro-enterol 1993;88:919–24.
42 Guruge JL, Falk PG, Lorenz RG, et al. Epithelialattachment alters the outcome of Helicobacterpylori infection. Proc Natl Acad Sci USA 1998;95:3925–30.
43 Gerhard M, Lehn N, Neumayer N, et al. Clinicalrelevance of the Helicobacter pylori gene forblood-group antigen-binding adhesin. Proc NatlAcad Sci USA 1999;96:12778–83.
44 Oberhuber G, Kranz A, Dejaco C, et al. Bloodgroups Lewis (b) and ABH expression in gastricmucosa: lack of inter-relation with Helicobacterpylori colonisation and occurrence of gastricMALT lymphoma. Gut 1997;41:37–42.
45 Clyne M, Drumm B. Absence of effect of LewisA and Lewis B expression on adherence of Heli-cobacter pylori to human gastric cells. Gastroen-terology 1997;113:72–80.
46 Wyatt JI, Rathbone BJ, Sobala GM, et al. Gastricepithelium in the duodenum: its association withHelicobacter pylori and inflammation. J ClinPathol 1990;43:981–6.
47 Gisbert JP, Blanco M, Cruzado AI, Pajares JM.Helicobacter pylori infection, gastric metaplasiain the duodenum and the relationship withulcer recurrence. Eur J Gastroenterol Hepatol2000;12:1295–8.
48 Khulusi S, Badve S, Patel P, et al. Pathogenesis ofgastric metaplasia of the human duodenum: roleof Helicobacter pylori, gastric acid, and ulcera-tion. Gastroenterology 1996;110:452–8.
49 Noach LA, Rolf TM, Tytgat GN. Electron micro-scopic study of association between Helicobacterpylori and gastric and duodenal mucosa. J ClinPathol 1994;47:699–704.
50 Xiang Z, Censini S, Bayeli PF, et al. Analysis ofexpression of CagA and VacA virulence factorsin 43 strains of Helicobacter pylori reveals thatclinical isolates can be divided into two majortypes and that CagA is not necessary for expres-sion of the vacuolating cytotoxin. Infect Immun1995;63:94–8.
51 Su B, Hellstrom PM, Rubio C, Celik J, GranstromM, Normark S. Type I Helicobacter pylori showsLewis (b)-independent adherence to gastriccells requiring de novo protein synthesis inboth host and bacteria. J Infect Dis 1998;178:1379–90.
Print & Media
II
Helicobacter pylori SabA Adhesinin Persistent Infection
and Chronic InflammationJafar Mahdavi,1* Berit Sonden,1*† Marina Hurtig,1*Farzad O. Olfat,1,2 Lina Forsberg,1 Niamh Roche,3
Jonas Ångstrom,3 Thomas Larsson,3 Susann Teneberg,3
Karl-Anders Karlsson,3 Siiri Altraja,4 Torkel Wadstrom,5
Dangeruta Kersulyte,6 Douglas E. Berg,6 Andre Dubois,7
Christoffer Petersson,8 Karl-Eric Magnusson,8 Thomas Norberg,9
Frank Lindh,10 Bertil B. Lundskog,11 Anna Arnqvist,1,12
Lennart Hammarstrom,13 Thomas Boren1‡
Helicobacter pylori adherence in the human gastric mucosa involves specificbacterial adhesins and cognate host receptors. Here, we identify sialyl-dimeric-Lewis x glycosphingolipid as a receptor for H. pylori and show that H. pyloriinfection induced formation of sialyl-Lewis x antigens in gastric epithelium inhumans and in a Rhesus monkey. The corresponding sialic acid–binding adhesin(SabA) was isolated with the “retagging” method, and the underlying sabA gene( JHP662/HP0725) was identified. The ability of manyH. pylori strains to adhereto sialylated glycoconjugates expressed during chronic inflammation mightthus contribute to virulence and the extraordinary chronicity of H. pyloriinfection.
Helicobacter pylori persistently infects thegastric mucosa of more than half of allpeople worldwide, causes peptic ulcer dis-ease, and is an early risk factor for gastriccancer (1). Many H. pylori strains expressadhesin proteins that bind to specific host-cell macromolecule receptors (2). This ad-herence may be advantageous to H. pyloriby helping to stabilize it against mucosal
shedding into the gastric lumen and ensur-ing good access to nourishing exudate fromgastric epithelium that has been damagedby the infection. The best defined H. pyloriadhesin-receptor interaction found to dateis that between the Leb blood group antigenbinding adhesin, BabA, a member of afamily of H. pylori outer membrane pro-teins, and the H, Lewis b (Leb), and relatedABO antigens (3–5). These fucose-contain-ing blood group antigens are found on redblood cells and in the gastrointestinal mu-cosa (6 ). Blood group–O individuals sufferdisproportionately from peptic ulcer dis-ease, suggesting that bacterial adherence tothe H and Leb antigens affects the severityof infection (7 ). Additional H. pylori– hostmacromolecule interactions that do not in-volve Leb-type antigens have also beenreported (8).
H. pylori is a genetically diverse species,with strains differing markedly in virulence.Strains from persons with overt disease gener-ally carry the cag-pathogenicity island (cag-PAI) (9, 10), which mediates translocation ofCagA into host cells, where it is tyrosine phos-phorylated and affects host cell signaling (11).Leb antigen binding is most prevalent amongcag� strains from persons with overt disease (4,12). Separate studies using transgenic mice thatexpress the normally absent Leb antigen sug-gest that H. pylori adherence exacerbates in-flammatory responses in this model (13). Takentogether, these results point to the pivotal role ofH. pylori adherence in development of severedisease.
Leb antigen–independent binding. Ear-lier studies identified nearly identical babAgenes at different H. pylori chromosomalloci, each potentially encoding BabA. ThebabA2 gene encodes the complete adhesin,whereas babA1 is defective because sequenc-es encoding the translational start and signalpeptide are missing (4). Our experiments be-gan with analyses of a babA2–knockout mu-tant derivative of the reference strainCCUG17875 (hereafter referred to as 17875).Unexpectedly, this 17875 babA2 mutantbound to gastric mucosa from an H. pylori–infected patient with gastritis.
A 17875 derivative with both babA genesinactivated (babA1A2) was constructed (14).This babA1A2 mutant also adhered (Fig. 1, Band C), which showed that adherence was notdue to recombination to link the silent babA1gene with a functional translational start andsignal sequence. Pretreatment with soluble Lebantigen (structures in table S1) resulted in�80% lower adherence by the 17875 parentstrain (Figs. 1E and 3C) but did not affectadherence by its babA1A2 derivative (Figs. 1Fand 3C).
In contrast to binding to infected gastricmucosa (Fig. 1, A to I), the babA1A2 mu-tant did not bind to healthy gastric mucosafrom a person not infected with H. pylori(Fig. 1, J to M), whereas the 17875 parentstrain bound avidly (Fig. 1, M versus L).These results implicated another adhesinthat recognizes a receptor distinct from theLeb antigen and possibly associated withmucosal inflammation.
Adherence was also studied in tissuefrom a special transgenic mouse that pro-duces Leb antigen in the gastric mucosa,the consequence of expression of a human-derived �1,3/4 fucosyltransferase (FT)(15). Strain 17875 and the babA1A2 mutanteach adhered to Leb mouse gastric epithe-lium (Fig. 2A, ii and iii), whereas bindingof each strain to the mucosa of nontrans-genic (FVB/N) mice was poor and waslimited to the luminal mucus. Thus, Lebmice express additional oligosaccharidechains (glycans), possibly fucosylated, butdistinct from the Leb antigen that H. pyloricould exploit as a receptor.sdiLex antigen–mediated binding. To
search for another receptor, thin-layer chro-matography (TLC)–separated glycosphingo-lipids (GSLs) were overlaid with H. pyloricells and monoclonal antibodies (mAbs), asappropriate. These tests (i) showed that thebabA1A2 mutant bound acid GSLs (Fig. 2B,iv, lanes 2 and 4), (ii) confirmed that it did notbind Leb GSL (lane 9), (iii) showed that itsbinding was abrogated by desialylation (lanes3 and 5), and (iv) revealed that it did not bindsialylated GSLs of nonhuman origin (lane 1)[table S1, numbers 2 to 5, in (16)] (indicatingthat sialylation per se is not sufficient for
1Department of Odontology/Oral Microbiology,Umeå University, SE-901 87 Umeå, Sweden. 2TheSwedish Institute for Infectious Disease Control, SE-171 82 Solna, Sweden. 3Institute of Medical Biochem-istry, Goteborg University, Box 440, SE-405 30 Gote-borg, Sweden. 4Institute of Molecular and Cell Biolo-gy, Tartu University, EE-51010 Tartu, Estonia. 5De-partment of Infectious Diseases and MedicalMicrobiology, Lund University, SE-223 62 Lund, Swe-den. 6Department of Molecular Microbiology, Wash-ington University Medical School, St. Louis, MO63110, USA. 7Laboratory of Gastrointestinal and LiverStudies, Department of Medicine, USUHS, Bethesda,MD 20814–4799, USA. 8Department of Molecularand Clinical Medicine, Division of Medical Microbiol-ogy, Linkoping University, SE-581 85 Linkoping, Swe-den. 9Department of Chemistry, Swedish Universityof Agricultural Sciences, SE-750 07 Uppsala, Sweden.10IsoSep AB, Dalkarrsv. 11, SE-146 36 Tullinge, Swe-den. 11Department of Medical Biosciences/ClinicalCytology, Umeå University, SE-901 87 Umeå, Swe-den. 12Department of Molecular Biology, Umeå Uni-versity, SE-901 87 Umeå, Sweden. 13Center for Bio-technology, Karolinska Institute, Novum, SE-141 57Huddinge, Sweden.
*These authors contributed equally to this work.†Present address: Unite de Genetique Mycobacteri-enne, Institut Pasteur, 25 rue du Dr Roux, 75724 ParisCedex 15, France.‡To whom correspondence should be addressed. E-mail: [email protected]
R E S E A R C H A R T I C L E
www.sciencemag.org SCIENCE VOL 297 26 JULY 2002 573
Print & Media
adherence). In addition, the binding pattern ofthe babA1A2 mutant matched that of the mAbagainst sLex (Fig. 2B, ii versus iv), exceptthat sLex-mono-GSL was bound more weak-ly by the babA1A2 mutant than by the mAb(Fig. 2B, iv) and no binding to sialyl-Lewisa-GSL was detected (table S1). Thus, thebabA1A2 mutant preferably binds sialylatedgangliosides, possibly with multiple Lex (fu-cose-containing) motifs in the core chain(thus, slower migration on TLC).
The babA1A2-mutant strain was next usedto purify a high-affinity binding GSL fromhuman adenocarcinoma tissue (14). The H.pylori–binding GSL was identified by massspectrometry and 1H nuclear magnetic reso-nance (NMR) as the sialyl-dimeric-Lewis xantigen, abbreviated as the sdiLex antigen(Fig. 2, B and C) (14).
Further tests with soluble glycoconjugatesshowed that the 17875-parent strain bound bothsLex and Leb antigens, whereas its derivativebound only sLex (Fig. 3A). Pretreatment of thebabA1A2 mutant with sLex conjugate reducedits in situ adherence by more than 90% (Figs. 1Iand 3C) (14) but did not affect adherence of the17875 parent strain (Figs. 1H and 3C). Similar-ly, pretreatment of tissue sections with an mAbthat recognizes sdiLex reduced adherence bythe babA1A2 mutant by 72% (Fig. 3E).
Titration experiments showed that thebabA1A2 mutant exhibited high affinity for thesdiLex GSLs, with a level of detection of 1pmol (Fig. 2B, iv) (table S1, number 8). At least2000-fold more (2 nmol) of the shorter sialyl-(mono)-Lewis x GSL was needed for binding(table S1, number 7). In contrast, its affinity(Ka) for soluble conjugates was similar for
mono and dimeric forms of sLex, 1 � 108 M�1
and 2 � 108 M�1, respectively (Fig. 3B) (16).These patterns suggest that the sialylated bind-ing sites are best presented at the termini ofextended core chains containing multiple Lewisx motifs, such as GSLs in cell membranes.Such optimization of steric presentation wouldbe less important for soluble receptors.
The Scatchard analyses (16) also estimat-ed that 700 sLex-conjugate molecules werebound per babA1A2-mutant bacterial cell(Fig. 3B), a number similar to that of Lebconjugates bound per cell of strain 17875.Clinical isolates. A panel of 95 Europe-
an clinical isolates was analyzed for sLexbinding (14). Thirty-three of the 77 cagA�
strains (43%) bound sLex, but only 11% (2out of 18) of cagA– strains (P � 0.000).However, deletion of the cagPAI from strainG27 did not affect sLex binding, as had alsobeen seen in studies of cagA and Leb-antigenbinding. Out of the Swedish clinical isolates,39% (35 out of 89) bound sLex, and the greatmajority of these sLex-binding isolates, 28out of 35 (80%), also bound the Leb antigen.Fifteen of the 35 (43%) sLex-binding isolatesalso bound the related sialyl-Lewis a antigen(sLea) (table S1, number 6), whereas none ofthe remaining 54 sLex-nonbinding strainscould bind to sLea. The clinical isolatesSMI65 and WU12 (the isolate from the in-fected patient in Fig. 1, A to M) illustratesuch combinations of binding modes (Fig.3A). Of the two strains with sequenced ge-nomes, strain J99 (17) bound sLex, sLea, andLeb antigens, whereas strain 26695 (18)bound none of them (Fig. 3A).Binding to inflamed tissue. Gastric
tissue inflammation and malignant transfor-mation each promote synthesis of sialylatedglycoconjugates (19), which are rare inhealthy human stomachs (20). Immunohisto-chemical analysis showed that the sialylatedantigens were located to the apical surfaces ofthe surface epithelial cells (fig. S2). To studygastric mucosal sialylation in an H. pyloricontext, gastric biopsies from 29 endoscopypatients were scored for binding by thebabA1A2 mutant and by the 17875 strain insitu, and for several markers of inflammation(table S2A) (14). Substantial correlationswere found between babA1A2-mutant adher-ence and the following parameters: (i) levelsof neutrophil (PMN) infiltration, 0.47 (P �0.011); (ii) lymphocyte/plasma cell infiltra-tion, 0.46 (P � 0.012); (iii) mAb staining forsLex in surface epithelial cells and in gastricpit regions, 0.52 (P � 0.004); and (iv) histo-logical gastritis score, 0.40 (P � 0.034). Incontrast, there was no significant correlationbetween babA1A2-mutant binding in situ andH. pylori density in biopsies from naturalinfection (0.14, P � 0.47), nor between anyinflammatory parameter and in situ adher-ence of strain 17875 (Leb and sLex binding).
Fig. 1. The sLex antigen confers adherence of H. pylori to the epithelium of H. pylori–infected(strain WU12) human gastric mucosa (Fig. 3A). H/E staining reveals mucosal inflammation (A). The17875 parent strain (B) and the babA1A2 mutant (C) both adhere to the gastric epithelium. Thesurface epithelium stains positive (arrows) with both the Leb mAb (D) and sLex mAb (G) [AISdescribed in (14 )]. The 17875 strain and babA1A2 mutant responded differently after pretreatment(inhibition) with soluble Leb antigen [(E) and (F), respectively)], or with soluble sLex antigen [(H)and (I), respectively)]. In conclusion, the Leb antigen blocked binding of the 17875 strain, whereasthe sLex antigen blocked binding of the babA1A2 mutant. H/E-stained biopsy with no H. pyloriinfection (J). No staining was detected with the sLex mAb (K). Here, strain 17875 adhered (L), incontrast to the babA1A2 mutant (M), because noninflamed gastric mucosa is low in sialylation (K).
R E S E A R C H A R T I C L E
26 JULY 2002 VOL 297 SCIENCE www.sciencemag.org574
Print & Media
The series of 29 patient biopsies was thencompared to a series of six biopsies of H.pylori–noninfected individuals, and a consid-erable difference was found to be due tolower adherence of the babA1A2 mutant (P �0.000) (table S2B), whereas strain 17875showed no difference.Binding to infected tissue. Biopsy ma-
terial from a Rhesus monkey was used todirectly test the view that H. pylori infectionstimulates expression of sialylated epithelialglycosylation patterns that can then be ex-ploited by H. pylori for adherence [monkeybiopsies in (14)]. This monkey (21) had beencleared of its natural H. pylori infection, andgastritis declined to baseline. Gastric biopsiestaken at 6 months post-eradication showedexpression of sLex in the gastric gland region(Fig. 4A) and no expression in the surfaceepithelium (Fig. 4A). In situ adherence of thebabA1A2 mutant was limited to gastricglands and was closely matched to the sLexexpression pattern (Fig. 4B). No specific ad-herence to the surface epithelium was seen(Fig. 4B).
At 6 months post-therapy, this gastritis-freeanimal had been experimentally infected with acocktail of H. pylori strains, of which the J166strain (a cagA-positive, Leb and sLex bindingisolate) became predominant a few months lat-er. This led to inflammation, infiltration bylymphocytes [Fig. 4C, bluish due to hematoxy-line/eosine (H/E)–staining], and microscopicdetection of H. pylori infection (Fig. 4, E and F)(21). The virulent H. pylori infection led tostrong sLex-antigen expression in the surfaceepithelium (Fig. 4C) and maintained expressionin the deeper gastric glands (Fig. 4C); thus, abi-layered expression mode supported strongbinding of the babA1A2 mutant to both regions(Fig. 4D). Bacterial pretreatment with sLexconjugate eliminated surface epithelial adher-ence and reduced gastric gland adherence by88%. Thus, persistent H. pylori infection up-regulates expression of sLex antigens, which H.pylori can exploit for adherence to the surfaceepithelium.Binding to Leb transgenic mouse gas-
tric epithelium. We analyzed gastric mu-cosa of Leb mice for sLex antigen–depen-dent H. pylori adherence in situ [AIS, in(14 )]. Pretreatment with the sLex conjugatereduced binding by babA1A2-mutant bac-teria by more than 90% (Figs. 2A, vi, and3D) but did not affect binding by its 17875parent (Figs. 2A, v, and 3D). In compari-son, pretreatment with soluble Leb antigendecreased adherence by �80% of the17875 strain (Leb and sLex binding),whereas binding by the babA1A2 mutantwas not affected. mAb tests demonstratedsLex antigen in the gastric surface epithe-lium and pits of Leb mice (Fig. 2A, iv).That is, these mice are unusual in produc-ing sialyl as well as Leb glycoconjugates
(even without infection), which each serveas receptors for H. pylori.
A finding that H. pylori pretreatment withLeb conjugate blocked its adherence to Lebmouse tissue, whereas sLex pretreatment didnot, had been interpreted as indicating that H.pylori adherence is mediated solely by Lebantigen (22). However, because strain 17875and its babA1A2 mutant bound similar levels ofsoluble sLex conjugate (Fig. 3A), soluble Lebseems to interfere sterically with interactionsbetween sLex-specific H. pylori adhesins andsLex receptors in host tissue (see also Fig. 1, Eversus F). Because an excess of soluble sLexdid not affect H. pylori binding to Leb receptors[Figs. 1H and 3C (in humans) and Figs. 2A, v,and 3D (in Leb-mice)], the steric hindrance isnot reciprocal. Further tests showed that liquidphase binding is distinct. A 10-fold excess ofsoluble unlabeled Leb conjugate (3 �g) alongwith 300 ng of 125I-sLex conjugate did notaffect strain 17875 binding to soluble sLexglycoconjugate.
Thus, soluble Leb conjugates interferewith sLex-mediated H. pylori binding spe-cifically when the sLex moieties are con-strained on surfaces. The lack of reciproc-
ity in these Leb-sLex interference interac-tions contributes to our model of receptorpositioning on cell surfaces.SabA identified. We identified the sLex-
binding adhesin by “retagging.” This tech-nique exploits a receptor-bound multifunc-tional biotinylated crosslinker, and ultraviolet(UV) irradiation to mediate transfer of thebiotin tag to the bound adhesin (4). Here, weadded the crosslinker to sLex conjugate andused more UV exposure (14) than had beenused to isolate the BabA adhesin (4) to com-pensate for the lower affinity of the sLex-than the Leb-specific adhesin for cognate sol-uble receptors (Fig. 3B). Strain J99 was used,and a 66-kDa protein was recovered (Fig.5A). Four peptides identified by mass spec-trometry (MS)–matched peptides encoded bygene JHP662 in strain J99 (17) (geneHP0725 in strain 26695) (18). Two of thefour peptides also matched those from therelated gene JHP659 (HP0722) (86% proteinlevel similarity to JHP662) (fig. S1).
To critically test if JHP662 or JHP659 en-codes SabA, we generated camR insertion al-leles of each gene in strain J99. Using radiola-beled glycoconjugates, we found that both sLex
Fig. 2. (A) The sLex antigen con-fers adherence of H. pylori to thetransgenic Leb mouse gastricmucosa. H/E-stained biopsy ofLeb mouse gastric mucosa (i).The 17875 strain (ii) andbabA1A2 mutant (iii) both ad-hered to the surface epithelium,which stained positive with thesLex mAb (iv). Adherence by thebabA1A2 mutant was lost (vi),while adherence by the 17875strain was unperturbed (v), afterpretreatment of the bacteria with soluble sLex antigen. (B) H. pylori binds to fucosylatedgangliosides such as the sLex GSL. GSLs were separated on TLC and chemically stained (i) (14).Chromatograms were probed with sLex mAb (ii), the 17875 strain (iii), and the babA1A2 mutant(iv). Lanes contained acid (i.e., sialylated) GSLs of calf brain, 40 �g; acid GSLs of human neutrophilgranulocytes, 40 �g; sample number 2 after desialylation, 40 �g; acid GSLs of adenocarcinoma, 40�g; sample number 4 after desialylation, 40 �g; sdiLex GSL, 1 �g; sLea GSL, 4 �g; s(mono)Lex GSL,4 �g; and Leb GSL, 4 �g. Binding results are summarized in table S1. (C) The high-affinity H. pyloriGSL receptor was structurally identified by negative ion fast atom bombardment mass spectrom-etry. The molecular ion (M-H�)– at m/z 2174 indicates a GSL with one NeuAc, two fucoses, twoN-acetylhexosamines, four hexoses, and d18 :1-16:0; and the sequence NeuAcHex(Fuc)HexNAc-Hex(Fuc)HexNAcHexHex was deduced. 1H-NMR spectroscopy resolved the sdiLex antigen;NeuAc�2.3Gal�1.4(Fuc�1.3)GlcNAc�1.3Gal�1.4(Fuc�1.3) GlcNAc�1.3Gal�1.4Glc�1Cer (14).
R E S E A R C H A R T I C L E
www.sciencemag.org SCIENCE VOL 297 26 JULY 2002 575
Print & Media
and sLea antigen–binding activity was abol-ished in the JHP662 (sabA) mutant, but not inthe JHP659 (sabB) mutant. Thus, the SabAadhesin is encoded by JHP662 (HP0725). Thisgene encodes a 651-aa protein (70 kDa) andbelongs to the large hop family of H. pyloriouter membrane protein genes, including babA(17, 18). The sabA gene was then identified byPCR in six sLex-binding and six non–sLex-binding Swedish isolates, which suggests thatsabA is present in the majority of H. pyloriisolates (14).
Parallel studies indicated that the sabAinactivation did not affect adherence mediat-ed by the BabA adhesin (Fig. 5B, i) and thatpretreatment of the J99sabA mutant with sol-uble Leb antigen prevented its binding togastric epithelium (Fig. 5B, ii). This impliesthat the SabA and BabA adhesins are orga-nized and expressed as independent units.
Nevertheless, Leb conjugate pretreatment ofJ99 (BabA� and SabA�) might have inter-fered with sLex antigen–mediated adherence(see Fig. 1E). To determine if this was a stericeffect of the bulky glycoconjugate on expo-sure of SabA adhesin, single babA and sabAmutant J99 derivatives were used to furtheranalyze Leb and sLex adherence (Fig. 5C, i toiv). Both single mutants adhered to the in-flamed gastric epithelial samples, whereasthe babAsabA (double) mutant was unable tobind this same tissue.Instability of sLex binding. When
screening for SabA mutants, we also ana-lyzed single colony isolates from cultures ofparent strain J99 (which binds both sLex andLeb antigens) (Fig. 3A), which indicated that1% of colonies had spontaneously lost theability to bind sLex. Similar results wereobtained with strain 17875 (Fig. 3A) with an
OFF (non–sLex-binding) variant called17875/Leb. In contrast, each of several hun-dred isolates tested retained Leb antigen–binding capacity.
Upstream and within the start of the sabAgene are poly T/CT tracts that should behotspots for ON/OFF frameshift regulation[see HP0725 in (18)], which might underliethe observed instability of sLex-binding ac-tivity. Both strains’ genome sequences,26695 and J99, demonstrate CT repeats thatsuggest sabA to be out of frame (six and nineCTs, respectively) (17, 18). Four sLex-bind-ing and four non–sLex-binding Swedish iso-lates, and in addition strain J99, were ana-lyzed by PCR for CT repeats, and differencesin length were found between strains. Ten CTrepeats [as compared to nine in (17)] were
Fig. 3. H. pylori bindssialylated antigens. (A)H. pylori strains andmutants (14) were an-alyzed for binding todifferent 125I-labeledsoluble fucosylated andsialylated (Lewis) anti-gen-conjugates [RIA in(14)]. The bars givebacterial binding, andconjugates used aregiven in the diagram.(B) For affinity analyses(16), the sLex conju-gate was added in titra-tion series. ThebabA1A2 mutant wasincubated for 3 hourswith the s(mono)Lexconjugate to allow forequilibrium in binding,which demonstrates anaffinity (Ka) of 1 � 108
M�1. (C) Adherence ofstrain 17875 (sLex andLeb-antigen binding)and babA1A2 mutantto biopsy with inflam-mation and H. pylori in-fection, as scored bythe number of boundbacteria after pretreat-ment with soluble Leb(Fig. 1, E and F) or sLexantigen (Fig. 1, H and I)(14). The Leb antigenreduced adherence ofstrain 17875 by �80%,whereas the sLex anti-gen abolished adher-ence of the babA1A2mutant. (D) Adherenceof strain 17875 andbabA1A2 mutant to biopsy of Leb mouse gastric mucosa was scored by the number of bound bacteriaafter pretreatment with sLex conjugate. Adherence by the babA1A2 mutant was abolished (Fig. 2A, vi),while adherence by strain 17875 was unperturbed (Fig. 2A, v). (E) Adherence of spontaneous phasevariant strain 17875/Leb (Leb-antigen binding only in Fig. 3A) and babA1A2 mutant was analyzed afterpretreatment of histo-sections of human gastric mucosa with mAbs recognizing the Leb antigen or thesdiLex antigen (FH6), with efficient inhibition of adherence of both strain 17875/Leb and the babA1A2mutant, respectively. Value P � 0.001 (***), value P � 0.01 (**), value P � 0.05 (*).
Fig. 4. Infection induces expression of sLexantigens that serve as binding sites for H. pyloriadherence. (A) and (B) came from a Rhesusmonkey with healthy gastric mucosa. (C) to (F)came from the same Rhesus monkey 9 monthsafter established reinfection with virulent clin-ical H. pylori isolates. Sections were analyzedwith the sLex mAb (A) and (C) and by adher-ence in situ with the babA1A2 mutant (B) and(D). In the healthy animal, both the sLex mAbstaining and adherence of the babA1A2 mutantis restricted to the deeper gastric glands [(A)and (B), arrows] [i.e., no mAb-staining of sLexantigens and no babA1A2-mutant bacteriapresent in the surface epithelium (A) and (B),arrowhead]. In contrast, in the H. pylori–in-fected and inflamed gastric mucosa, the sLexantigen is heavily expressed in the surface ep-ithelium [(C), arrowhead with brownish immu-nostaining], in addition to the deeper gastricglands [(C), arrow]. The up-regulated sLex an-tigen now supports massive colocalized adher-ence of the babA1A2 mutant in situ [(D), ar-rowhead] in addition to the constant binding tothe deeper gastric glands [(B) and (D), arrow].The established infection was visualized byGenta stain, where the H. pylori microbeswere present in the surface epithelium [(E)and (F), arrowhead; (F) is at a higher magni-fication], while no microbes were detected inthe deeper gastric glands [(E), arrow].
R E S E A R C H A R T I C L E
26 JULY 2002 VOL 297 SCIENCE www.sciencemag.org576
Print & Media
found in the sLex-binding J99 strain, whichputs this ORF in frame, and could thus ex-plain the ON bindings. These results furthersupport the possibility for a flexible locus thatconfers ON/OFF binding properties (23).Dynamics of sialylation during health
and disease. Our analyses of H. pylori ad-herence provide insight into human re-sponses to persistent infections, where gas-tritis and inflammation elicit appearance ofsdiLex antigens and related sialylated car-bohydrates in the stomach mucosa, whichcag� (virulent) H. pylori strains by adap-tive mechanisms exploit as receptors inconcert with the higher affinity binding toLeb. These two adherence modes may eachbenefit H. pylori by improving access tonutrients leached from damaged host tis-sues, even while increasing the risk of bac-tericidal damage by these same host defens-es (Fig. 6). In the endothelial lining, sialy-lated Lewis-glycans serve as receptors for
selectin cell adhesion proteins that helpguide leukocyte migration and thus regu-late strength of response to infection orinjury (24 ). For complementary attach-ment, the neutrophils themselves also ex-press sialylated Lewis glycans, and suchneutrophil glycans allow binding and infec-tion by human granulocytic erlichiosis (25).However, sialylated glycoconjugates arelow in healthy gastric mucosa but are ex-pressed during gastritis. This sialylationwas correlated with the capacity for SabA-dependent, but not BabA-dependent, H. py-lori binding in situ. Our separate tests onRhesus monkeys for experimental H. pyloriinfection confirmed that gastric epithelialsialylation is induced by H. pylori infec-tion. In accord with this, high levels ofsialylated glycoconjugates have been foundin H. pylori–infected persons, which de-creased after eradication of infection andresolution of gastritis (26 ). Thus, a sialy-
lated carbohydrate used to signal infectionand inflammation and to guide defense re-sponses can be co-opted by H. pylori as areceptor for intimate adherence.
High levels of sialylated glycoconjugatesare associated with severe gastric disease,including dysplasia and cancer (27, 28). Sia-lylated glycoconjugates were similarly abun-dant in parietal cell–deficient mice (22). sLexwas also present in the gastric mucosa oftransgenic Leb mice (Fig. 2A). Whether thisreflects a previously unrecognized pathologystemming from the abnormal (for mice) gas-tric synthesis of �1.3/4FT or Leb antigen, orfrom fucosylation of already sialylated car-bohydrates, is not known.
Persons with blood group O and “nonsecre-tor” phenotypes (lacking the ABO bloodgroup–antigen synthesis in secretions such assaliva and milk) are relatively common (e.g.,45 and 15%, respectively, in Europe), andeach group is at increased risk for peptic ulcerdisease (29). The H1 and Leb antigens areabundant in the gastric mucosa of secretors (ofblood group O) (6), but not in nonsecretors,where instead the sLex and sLea antigens arefound (30). The blood group O–disease asso-ciation was postulated to reflect the adherenceof most cag� H. pylori strains to H1 and Lebantigens (3). We now suggest that H. pyloriadherence to sialylated glycoconjugates con-tributes similarly to the increased risk of pepticulcer disease in nonsecretor individuals.Adaptive and multistep–mediated at-
tachment modes of H. pylori. Our findingsthat the SabA adhesin mediates binding to thestructurally related sialyl-Lewis a antigen(sLea, in table S1) is noteworthy becausesLea is an established tumor antigen (31) andmarker of gastric dysplasia (27), which mayfurther illustrate H. pylori capacity to exploita full range of host responses to epithelialdamage. The H. pylori BabA adhesin bindsLeb antigen on glycoproteins (32), whereasits SabA adhesin binds sLex antigen in mem-brane glycolipids, which may protrude lessfrom the cell surface. Thus, H. pylori adher-
Fig. 5. Retagging of SabA and identifi-cation of the corresponding gene, sabA.After contact-dependent retagging (ofthe babA1A2 mutant) with crosslinkerlabeled sLex conjugate, the 66-kDa bi-otin-tagged adhesin SabA [(A), lane 1]was identified by SDS–polyacrylamidegel electrophoresis/steptavidin-blot,magnetic-bead-purified, and analyzedby MS for peptide masses (14). As acontrol, the Leb conjugate was used toretag 17875 bacteria, which visualizedthe 75-kDa BabA adhesin [(A), lane 2](4). (B) The J99/JHP662::cam (sabA)mutant does not bind the sLex antigenbut adheres to human gastric epitheli-um [(B), i], due to the BabA-adhesin,because pretreatment with soluble Lebantigen inhibited binding to the epithe-lium [(B), ii]. (C) The J99 wild-typestrain (which binds both Leb and sLexantigens) [(C), i] and the J99 sabA mu-tant (which binds Leb antigen only) [(C),ii] both exhibit strong adherence to thegastric mucosa. In comparison, the J99babA mutant binds by lower-affinity interactions [(C), iii], while the J99 sabAbabA mutant, whichis devoid of both adherence properties, does not bind to the gastric mucosa [(C), iv].
Fig. 6. H. pylori adher-ence in health and dis-ease. This figure illus-trates the proficiencyof H. pylori for adap-tive multistep mediat-ed attachment. (A) H.pylori (in green) adher-ence is mediated bythe Leb blood groupantigen expressed in glycoproteins (blue chains) in the gastric surfaceepithelium (the lower surface) (3, 32). H. pylori uses BabA (green Y’s) forstrong and specific recognition of the Leb antigen (4). Most of the sLex-binding isolates also bind the Leb antigen (SabA, in red Y’s). (B) Duringpersistent infection and chronic inflammation (gastritis), H. pylori triggersthe host tissue to retailor the gastric mucosal glycosylation patterns toup-regulate the inflammation-associated sLex antigens (red host, triangles).Then, SabA (red Y structures) performs Selectin-mimicry by bindingthe sialyl-(di)-Lewis x/a glycosphingolipids, for membrane close
attachment and apposition. (C) At sites of vigorous local inflamma-tory response, as illustrated by the recruited activated white bloodcell (orange “bleb”), those H. pylori subclones that have lost sLex-binding capacity due to ON/OFF frameshift mutation might havegained local advantage in the prepared escaping of intimate contactwith (sialylated) lymphocytes or other defensive cells. Such adapta-tion of bacterial adherence properties and subsequent inflammationpressure could be major contributors to the extraordinary chronicityof H. pylori infection in human gastric mucosa.
R E S E A R C H A R T I C L E
www.sciencemag.org SCIENCE VOL 297 26 JULY 2002 577
Print & Media
ence during chronic infection might involvetwo separate receptor-ligand, interactions—oneat “arm’s length” mediated by Leb, and anoth-er, more intimate, weaker, and sLex-mediatedadherence. The weakness of the sLex-mediatedadherence, and its metastable ON/OFF switch-ing, may benefit H. pylori by allowing escapefrom sites where bactericidal host defense re-sponses are most vigorous (Fig. 6C). In sum-mary, we found that H. pylori infection elicitsgastric mucosal sialylation as part of the chron-ic inflammatory response and that many viru-lent strains can exploit Selectin mimicry andthus “home in” on inflammation-activated do-mains of sialylated epithelium, complementingthe baseline level of Leb receptors. The spec-trum of H. pylori adhesin–receptor interactionsis complex and can be viewed as adaptive,contributing to the extraordinary chronicity ofH. pylori infection in billions of people world-wide, despite human genetic diversity and hostdefenses.
References and Notes1. T. L. Cover et al., in Principles of Bacterial Pathogen-
esis, E. A. Groisman, Ed. (Academic Press, New York,2001), pp. 509–558.
2. M. Gerhard et al., in Helicobacter pylori: Molecular andCellular Biology, S. Suerbaum and M. Achtman, Eds.(Horizon Scientific Press, Norfolk, UK, 2001), chap. 12.
3. T. Boren et al., Science 262, 1892 (1993).4. D. Ilver et al., Science 279, 373 (1998).5. M. Hurtig, T. Boren, in preparation.6. H. Clausen, S. i. Hakomori, Vox Sang 56, 1 (1989).7. T. Boren, P. Falk, Science 264, 1387 (1994).
8. K. A. Karlsson, Mol. Microbiol. 29, 1 (1998).9. S. Censini et al., Proc. Natl. Acad. Sci. U.S.A. 93, 14648
(1996).10. N. S. Akopyants et al., Mol. Microbiol. 28, 37 (1998).11. E. D. Segal et al., Proc. Natl. Acad. Sci. U.S.A. 96,
14559 (1999).12. M. Gerhard et al., Proc. Natl. Acad. Sci. U.S.A. 96,
12778 (1999).13. J. L. Guruge et al., Proc. Natl. Acad. Sci. U.S.A. 95,
3925 (1998).14. Strains and culture, adherence in situ (AIS), H. pylori
overlay to TLC, isolation and identification of thes-di-Lex GSL, apical localization of sLex antigen ex-pression (fig. S2), retagging and identification ofSabA/sabA, alignment of JHP662/JHP659 (fig. S1),construction of adhesin gene mutants, analyses ofsabA by PCR, RIA and Scatchard analyses, gastricbiopsies from patients and monkeys analyzed for H.pylori binding activity and inflammation, and a sum-mary of H. pylori binding to glycosphingolipids (tableS1) are available as supporting online material.
15. P. G. Falk et al., Proc. Natl. Acad. Sci. U.S.A. 92, 1515(1995).
16. G. Scatchard, Ann. N.Y. Acad. Sci. 51, 660 (1949).17. R. A. Alm et al., Nature 397, 176 (1999).18. J. F. Tomb et al., Nature 388, 539 (1997).19. S. i. Hakomori, in Gangliosides and Cancer, H. F.
Oettgen, Ed. (VCH Publishers, New York, 1989), pp.58–68.
20. J. F. Madrid et al., Histochemistry 95, 179 (1990).21. A. Dubois et al., Gastroenterology 116, 90 (1999).22. A. J. Syder et al., Mol. Cell 3, 263 (1999).23. A. Arnqvist et al., in preparation.24. J. Alper, Science 291, 2338 (2001).25. M. J. Herron et al., Science 288, 1653 (2000).26. H. Ota et al., Virchows Arch. 433, 419 (1998).27. P. Sipponen, J. Lindgren, Acta Pathol. Microbiol. Im-
munol. Scand. 94, 305 (1986).28. M. Amado et al., Gastroenterology 114, 462 (1998).29. P. Sipponen et al., Scand. J. Gastroenterol. 24, 581
(1989).30. J. Sakamoto et al., Cancer Res. 49, 745 (1989).
31. J. L. Magnani et al., Science 212, 55 (1981).32. P. Falk et al., Proc. Natl. Acad. Sci. U.S.A. 90, 2035
(1993).33. We thank R. Clouse, H.M.T. El-Zimaity, D. Graham, and I.
Anan for human biopsy material; J. Gordon and P. Falkfor sections of Leb and FVB/N mouse stomach; L. Eng-strand and A. Covacci for H. pylori strains; H. Clausen forthe FH6 mAb; the Swedish NMR Centre, Goteborg Uni-versity; P. Martin and O. Furberg for digital art/moviework; and J. Carlsson for critical reading of the manu-script. Supported by the Umeå University BiotechnologyFund, Swedish Society of Medicine/Bengt Ihre’s Fund,Swedish Society for Medical Research, Lion’s CancerResearch Foundation at Umeå University, County Coun-cil of Vasterbotten, Neose Glycoscience Research AwardGrant ( T.B.), Swedish Medical Research Council [11218( T.B.), 12628 (S.T.), 3967 and 10435 (K.-A.K.), 05975(L.H.), and 04723 ( T.W.)], Swedish Cancer Society[4101-B00-03XAB ( T.B.) and 4128-B99-02XAB (S.T.)],SSF programs “Glycoconjugates in Biological Systems”( T.B., S.T., and K.-A.K.), and “Infection and Vaccinology”( T.B. and K.-A.K.), J. C. Kempe Memorial Foundation(J. M., L. F., and M.H.), Wallenberg Foundation (S.T. andK.-A.K), Lundberg Foundation (K.-A.K.), ALF grant fromLund University Hospital (T.W.), and grants from theNIH [RO1 AI38166, RO3 AI49161, RO1 DK53727 (D.B.)]and from Washington University (P30 DK52574). Theseexperiments were conducted according to the principlesin the “Guide for the Care and Use of Laboratory Ani-mals,” Institute of Laboratory Animal Resources, NRC,HHS/NIH Pub. No. 85–23. The opinions and assertionsherein are private ones of the authors and are not to beconstrued as official or reflecting the views of the DOD,the Uniformed Services University of the Health Scienc-es, or the Defense Nuclear Agency.
Supporting Online Materialwww.sciencemag.org/cgi/content/full/297/5581/573/DC1Materials and MethodsFigs. S1 and S2Tables S1 and S2
17 December 2001; accepted 17 June 2002
R E P O R T S
An Aligned Stream ofLow-Metallicity Clusters in the
Halo of the Milky WaySuk-Jin Yoon* and Young-Wook Lee
One of the long-standing problems in modern astronomy is the curious divisionof Galactic globular clusters, the “Oosterhoff dichotomy,” according to theproperties of their RR Lyrae stars. Here, we find that most of the lowestmetallicity ([Fe/H] � –2.0) clusters, which are essential to an understanding ofthis phenomenon, display a planar alignment in the outer halo. This alignment,combined with evidence from kinematics and stellar population, indicates acaptured origin from a satellite galaxy. We show that, together with thehorizontal-branch evolutionary effect, the factor producing the dichotomycould be a small time gap between the cluster-formation epochs in the MilkyWay and the satellite. The results oppose the traditional view that the metal-poorest clusters represent the indigenous and oldest population of the Galaxy.
More than 60 years ago, Oosterhoff (1) dis-covered that Galactic globular clusters couldbe divided into two distinct groups according
to the mean period of type ab RR Lyraevariables (Pab�). This dichotomy was one ofthe earliest indications of systematic differ-ence among globular clusters, whose realityhas been strengthened by subsequent investi-gations (2). Given that most characteristics ofGalactic globular clusters appear to be dis-tributed in a continuous way, it is unusual
that a quantity used to characterize variablestars falls into two rather well-defined class-es. Moreover, the two groups are known todiffer in metal abundance (3) and kinematicproperties (4), which may indicate distinctorigins. Whatever the reasons for the dichot-omy, the question of whether the two groupsoriginated under fundamentally different con-ditions is of considerable interest regardingthe formation scenarios of the Galactic halo.Despite many efforts during the last decades,the origin of this phenomenon still lacks aconvincing explanation.
Figure 1 shows the Oosterhoff dichotomy.Clusters belong to groups I and II if theirvalues of Pab� fall near 0.55 and 0.65 days,respectively (5). Group I is more metal-richthan group II (3). Based on our horizontal-branch (HB) population models (6, 7), wehave found that the presence of the relativelymetal-rich (–1.9 � [Fe/H] � –1.6) clusters ingroup II (hereafter group II-a) can be under-stood by the sudden increase in Pab� at[Fe/H] � –1.6 (indicated by a blue line inFig. 1). As [Fe/H] decreases, HB stars gethotter (i.e., the HB morphology gets bluer),and there is a certain point at which thezero-age portion of the HB just crosses the
Center for Space Astrophysics, Yonsei University,Seoul 120–749, Korea.
*To whom correspondence should be addressed. E-mail: [email protected]
R E S E A R C H A R T I C L E
26 JULY 2002 VOL 297 SCIENCE www.sciencemag.org578
Print & Media
Role of Helicobacter pylori SabA Adhesin
During Persistent Infection and Chronic Inflammation
Jafar Mahdavi, Berit Sondén, Marina Hurtig, Farzad O. Olfat, Lina Forsberg,Niamh Roche, Jonas Ångström, Thomas Larsson, Susann Teneberg, Karl-Anders Karlsson, Siiri Altraja, Torkel Wadström, Dangeruta Kersulyte, DouglasE. Berg, Andre Dubois, Christoffer Petersson, Karl-Eric Magnusson, ThomasNorberg, Frank Lindh, Bertil B. Lundskog, Anna Arnqvist, LennartHammarström, Thomas Borén
Supplementary material
MATERIALS AND METHODS
Supplemental Methods 1. For Adherence In Situ (AIS) (S1), biopsies wereobtained from Drs. Ray Clouse, Washington University Medical Center, St.Louis, David Graham, VA Medical Center, Houston, and Intissar Anan, UmeåUniversity. Gastric tissue sections from Leb mice and FVB/N mice were fromDrs. Jeff Gordon and Per Falk, Washington University. To inhibit AIS (S1) H.
pylori was labeled with FITC and mixed with Leb conjugate (10 µg/mL) orsialyl-Le x/a-conjugates (20 µg/mL) (IsoSep AB, Tullinge, Sweden).Alternatively, histo sections were pre-treated with the Leb mAb or the FH6 mAb(which recognizes the sdiLex antigen), at 1:100 dilution for 3 h. Reduction inbacterial binding was estimated by the number of adherent bacteria under 200Xmagnification. Each value is the mean +- SEM of 10 different fields, withsignificance evaluated with the Student's t-Test.
__________________________________________________________
Supplemental Methods 2. H. pylori strains 26695, CCUG17875 (designated17875), the construction of the 17875babA2::cam-mutant, and the G27cagPAI-deletion mutant, were described in (S2). The construction of 17875babA1::kanbabA2::cam (abbreviated the babA1A2-mutant), sabA(JHP662)::cam and thesabB(JHP659)::cam mutants were as described in (S2) with minor modifications(see Supplemental Methods 3). J99 was kindly provided by Drs. Tim Cover,John Atherton and Martin Blaser. The 77 cag
+ and 12 cag- Swedish isolates,
including SMI65, were kindly provided by Dr. Lars Engstrand, and 6 Italian
cagA- strains were described in (S2). Strain WU12 was from a patient withgastritis (Fig 1A-M), treated at Washington University Medical Center, St.
Print & Media
Louis. Strain J166 was described in (S3). The 17875/Leb strain, a spontaneoussLex- variant was isolated by screening single colonies of strain 17875. Bacteriawere grown for 2 days on Brucella agar medium at 37°C, under 10 % CO2 and
5% O2, for optimal binding.
__________________________________________________________
Supplemental Methods 3. Construction of babA1A2-, sabA-, sabB-, and
b a b A / s a b A -deletion mutants. The 17875b a b A 1 ::kan babA2::cam,J99sabA(JHP662)::cam, J99sabB(JHP659)::cam, and the J99babA::camsabA::kan-mutant strains were constructed as described in (S2), with thefollowing modifications;
(3A) Strain CCUG17875 was used for construction of 17875babA1::kan
babA2::cam (abbreviated the babA1A2-mutant). The babA1 gene wasamplified using the F44 (forward) and R44 (reverse) primers, cloned into thepBluescript SK+/- EcoRV site (Stratagene, La Jolla CA), linearized with primersR41+F38, ligated with the KanR cassette from pILL600 (S4), and used totransform the original 17875babA2::cam mutant (S2). The transformants wereanalyzed by PCR using upstream primers F2 (babA2) or F44 (babA1) incombination with the R11 babA primer (located in the segment replaced by theKanR cassette). The 17875b a b A 1 ::kan babA2::cam (double)-mutant(abbreviated the babA1A2-mutant) could not be amplified with either primerpair.
F2: CTTAAATATCTCCCTATCCCR41: GCGAGCCTAAAGTTAATGAF38: ACGTGGCGAACTTCCAATTCF44: CAGTCAAGCCCAAAGCTATGCR11: CGATTTGATAGCCTACGCTTGTGR44: CTTAAAGGGATAGGAAGCGCT
(3B) Strain J99 was used for construction of the sabA(JHP662)::cam and
the sabB(JHP659)::cam mutants. SabA was PCR-amplified using the primersF18 and R17 and cloned in pBluescript SK+/- EcoRV site. The plasmid clonewas linearized with R20 and F21 and ligated with the camR gene, described in(S2). SabB was amplified using primers F16 and R15, and cloned in pCR2.1-TOPO vector (Invitrogen, Carlsbad, CA), linearized with HincII and ligated
with the camR gene. H. pylori transformants were analyzed for binding to 125I-labeled sLex conjugate. The location of the camR gene in sabA and sabB wasanalyzed using primers R17+F18 and F16+R15, respectively, which verified
Print & Media
that loss of binding was dependent on the cam-insertion and independent ofspontaneous OFF-(phase)-variation in sLex binding.
R15: CTATTCATGTTTACAATAF16: GGGTTTGTTGTCGCACCACTAGR17: GGTTCATTGTAAATATATF18: CGATTCTATTAGATCACCCR20: AGCGTTCAATAACCCTTACAGCGF21: GATTTAAATACTGGCTTAATTGCTCGBS22: CGCTTAAAGCATTGTTGACAGCC
(3C) Strain J99 was used for construction of the J99babA::cam sabA::kan-
(double)-mutant. For the construction of the babAsabA-mutant strain, sabA wasfirst cloned in the pBluescript vector and linearized as described above, and thenligated with the KanR cassette (see above). For the construction of the J99babA
and babAsabA-double mutants the J99 and J99sabA ::kan strains weretransformed with the babA deletion vector described in (S2). The transformantswere analyzed by PCR using primers F33 + R34. The correct J99babA::cam andthe J99babA::cam sabA::kan-(double)-mutant could not be amplified with theprimer pair.F33: ATCCTTTCATTAACTTTAGGATCGCR34: TTGAGCGCTATCAGGCACAC
__________________________________________________________
Supplemental Methods 4. Retagging (of SabA) and Identification of sabA,
the Gene which Corresponds to the Sialic acid binding Adhesin, SabA.
For identification and purification of the SabA-adhesin, the Retaggingtechnique was used, i.e., similar to the identification of the Leb antigen bindingBabA-adhesin (S2), with some modifications. Here, the babA1A2-mutant wasincubated with Sulfo-SBED (Pierce, Rockville, IL.) attached to sLex-conjugate.The crosslinker was activated by overnight UV irradiation, and biotin-(Re)tagged proteins were purified with magnetic streptavidin beads. The 66kDaband purified was digested with Trypsin (seq grade, Promega, USA) and fourpurified peptides were identified by mass spectrometry based on peptide massesand sequences. MALDI-TOF-MS on a Tof-Spec E mass spectrometer(Micromass, Manchester , England) (S 5 ) , and ProFound(www.proteometrics.com) was used for matching of the four peptide massesusing NCBI database. Peptides 1 and 2 (see below) match the JHP662(S6)/HP0725 (S 7 ) gene. Peptides 3 and 4 also matched the JHP659(S6)/HP0722) (S7) gene. Peptide identities were validated by ESI-MS/MSsequencing on a Q-Tof instrument (Micromass), using the nanospray source(S8).
Print & Media
Mascot (www.matrixscience.com) identified the peptide sequences(numbered according to the JHP662 peptide sequence:(1) aa68-QSIQNANNIELVNSSLNYLK-aa87(2) aa301-DIYAFAQNQK-aa310(3) aa500-YYGFFDYNHGYIK-aa512(4) aa620-IPTINTNYYSFLGTK-aa634
__________________________________________________________
Supplemental Methods 5. Presence of the sabA gene.The presence of sabA in clinical H. pylori isolates was analyzed by PCR usingprimer pairs F1 + 5R, and 3F + 1R.1F: CTCTAGCAATGTGTGGCAG3F: CGCTAGTGTCCAGGGTAAC1R: GCGCTGTAAGGGTTATTGAAC5R: CCGCGTATTGCGTTGGGTAG
__________________________________________________________
Supplemental Methods 6. H. pylori overlay to HPTLC separatedglycosphingolipids (GSLs).Mixtures of GSLs (20-40 µg/lane) or pure compounds (0.002-4 µg/lane) wereseparated on silica gel 60 HPTLC plates (Merck) and chemically detected withanisaldehyde. Binding of 35S-labeled H. pylori to TLC separated GSLs wasaccording to (S9). The KM-93 mAb (Seikagaku Corp., Tokyo, Japan) was
detected using 125I -rabbit anti mouse antibodies from Dako, Glostrup,Denmark.
__________________________________________________________
Supplemental Methods 7. Isolation and identification of the sdiLex GSL
The babA1A2-mutant strain was used to purify a high affinity GSL withreceptor activity from human gall bladder adenocarcinoma tissue, which is a richsource of extended a1,3-fucosylated gangliosides (S10), by chromatography onsilicic acid columns, as described (S11). Fractions were tested for babA1A2-mutant binding. Desialylation was performed by mild acid hydrolysis in 1%(v/v) acetic acid for 1 hr at 100°C. Negative ion FAB mass spectra wererecorded on a JEOL SX-102A mass spectrometer (JEOL, Tokyo, Japan). Theions were produced by 6 keV xenon atom bombardment, using triethanolamineas matrix and an accelerating voltage of -10 kV. 1H NMR spectra were obtainedon a Varian 600 MHz machine at 30°C using dimethylsulfoxide-d6/D2O (98/2)as solvent. The NMR spectrum was compared with earlier published data for
Print & Media
this structure recorded at 55°C (S12) and corresponding structural elements ofthe Y-6 glycosphingolipid recorded at 30°C (S13). The H. pylori binding GSL
was then identified by negative ion FAB mass spectrometry and 1H NMR asNeuAca2.3Galß1.4(Fuca1.3)GlcNAcß1.3Galß1.4(Fuca1.3)GlcNAcß1.3Galß1.4Glcß1Cer (Figs. 2B: lane 6, and 2C) (i.e., the sialyl-dimeric-Lewis xglycosphingolipid), abbreviated sdiLex.
__________________________________________________________
Supplemental Methods 8. RIA and Scatchard analyzes
SLea, sLex, and Leb glycoconjugates (IsoSep AB), and sdiLex-glycoconjugate synthesized according to (S14), (all based on albumin), were
labeled with 125I by the chloramine T method and mixed with H. pylori bacteriafor binding and Scatchard analyzes (S15), essentially as described in (S2). Thebinding experiments were performed in triplicates.
__________________________________________________________
Supplemental Methods 9. Gastric biopsies analyzed for H. pylori binding
activity and inflammatory scores
Gastric biopsies from 29 individuals without dysplasia were H/E-stained andevaluated for lymphocyte/ plasmacell infiltration. PMN cell infiltration wasevaluated from rare PMNs only found in the lamina propria, up to pit abscesses.SLex antigen expression was analyzed by the KM-93 mAb in the surfacemucous cells/gastric pit region. Histological gastritis was also graded. H. pylori
was Genta-stained and quantified. AIS was scored by the number ofbacteria/gastric pit region. Each value is the mean of 5 different fields. Allscores are available in Table S2A. Adherence scores were correlated withinfiltration scores of lymphocytes/plasmacells, PMNs, staining scores by sLexmAb, gastritis and also H. pylori infections, and statistically analyzed accordingto Pearson.
__________________________________________________________
Supplemental Methods 10. Gastric biopsies of monkey E6C were taken 6months post-cure (for Figs. 4A and 4B). Final biopsy samples from E6C weretaken 9 months after established re-infection (Figs. 4C-4F) (S3). Biopsies werestained with the KM-93 mAb, and AIS by the babA1A2-mutant was analyzed asdescribed in Supplemental Methods 1.
Print & Media
Supplemental Figure 1. Alignment of JHP662/ JHP659,
__________________________________________________________
10 20 30 40 50 60....|....|....|....|....|....|... .|....|....|....|....|....|
JHP662 MKKTILLSLSLSLASSLLHAEDNGF FVSAGY QIGEAVQMVK NTGELKNLNEKYEQLSQYLJHP659_ MKKTILLSLSLSLASSLLHAEDNGF FVSAGY QIGEAVQMVK NTGELKNLNDKYEQLSQSL
70 80 90 100 110 120....|....|....|....|....|....|... .|....|....|....|....|....|
JHP662 NQVASLKQSIQNANNIELVNSSLNYLKSFTNNNYNSTTQSPIFNAVQAVITSVLGFWSLYJHP659_ AQLASLKKSIQTANNIQAVNNALSDLKSFAS NNHTNKETSPIYNTAQAVITSVLAFWSLY
130 140 150 160 170 180....|....|....|....|....|....|... .|....|....|....|....|....|
JHP662 AGNYLTFFVVNKDTQKPASVQGNPPFSTIVQ --NCSGIENCAMNQTTYDKMKKLAEDLQAJHP659_ AGNALSFH-V---TGLN-D-GSNSPLGRIHR DGNCTGLQQCFMSKETYDKMKTLAENLQK
190 200 210 220 230 240....|....|....|....|....|....|... .|....|....|....|....|....|
JHP662 AQQNATTKANNLCALSGCATTQGQNPSSTVS NALNLAQQLMDLIANTKTAMMWKNIVIAGJHP659_ A-Q------GNLCALSECSSNQSNGGKTSMT TALQTAQQLMDLIEQTKVSMVWKNIVIAG
250 260 270 280 290 300....|....|....|....|....|....|... .|....|....|....|....|....|
JHP662 VSN-V--SGAIDSTGYPTQYAVFNNIKAMIP ILQQAVTLSQSNHTLSASLQAQATGSQTNJHP659_ VTNKPNGAGAITSTGHVTDYAVFNNIKAMLP ILQQALTLSQSNHTLSTQLQARAMGSQTN
310 320 330 340 350 360....|....|....|....|....|....|... .|....|....|....|....|....|
JHP662 PKFAKDIYAFAQNQKQVISYAQDIFNLFSSIPKDQYRYLEKAYLKIPN AGKTPTNPYRQEJHP659_ REFAKDIYALAQNQKQILSNASSIFNLFNSI PKDQLKYLENAYLKVPHLGKTPTNPYRQN
370 380 390 400 410 420....|....|....|....|....|....|... .|....|....|....|....|....|
JHP662 VNLNQEIQTIQNNVSYYGNRVDAALSVAKDV YNLKSNQTEIVTTYNNAKNLSQEISKLPYJHP659_ VNLNKEINAVQDNVANYGNRLDSALSVAKDV YNLKSNQTEIVTTYNDAKNLSEEISKLPY
430 440 450 460 470 480....|....|....|....|....|....|... .|....|....|....|....|....|
JHP662 NQVNTKDIITLPYDQNAPAAGQYNYQINPEQ QSNLSQALAAMSNNPFKKVGMISSQNNNGJHP659_ NQVNVTNIVMSPKD--S-TAGQ--YQINPEQ QSNLNQALAAMSNNPFKKVGMISSQNNNG
490 500 510 520 530 540....|....|....|....|....|....|... .|....|....|....|....|....|
JHP662 ALNGLGVQVGYKQFFGESKRWGLRYYGFFDYNHGYIKSSFFNSSSDIWTYGGGSDLLVNFJHP659_ ALNGLGVQVGYKQFFGESKRWGLRYYGFFDYNHGYIKSSFFNSSSDIWTYGGGSDLLVNF
550 560 570 580 590 600....|....|....|....|....|....|... .|....|....|....|....|....|
JHP662 INDSITRKNNKLSVGLFGGIQLAGTTWLNSQ YMNLTAFNNPYSAKVNASNFQFLFNLGLRJHP659_ INDSITRKNNKLSVGLFGGIQLAGTTWLNSQ YMNLTAFNNPYSAKVNASNFQFLFNLGLR
610 620 630 640 650....|....|....|....|....|....|... .|....|....|....|....|.
JHP662 TNLATAKKKDSERSAQHGVELGIKIPTINTNYYSFLGTKLEYRRLYSVYLNYVFAYJHP659_ TNLATAKKKDSERSAQHGVELGIKIPTINTNYYSFLGTKLEYRRLYSVYLNYVFAY
All four peptides were aligned with the JHP662 gene (4 peptide match),while only (2 (red) peptides matched) the JHP659 gene. Thus, theresults suggest that sabA correspond to the JHP662 gene (S6)/ HP0725(S7) gene.
Print & Media
Supplemental Figure 2. Apical localization of epithelial sLex antigens ininflamed gastric epithelium
Sialylated glycoconjugates, such as sLex are rare in healthy human stomachs,but sialylation is upregulated during chronic inflammation and gastritis. Thecellular (topical) localization of the sialylated antigens was investigated using asLex specifik mAb. In the figure (400x magn.), the sLex antigen is visualized indark staining lining the apical surfaces (single arrows). In contrast, the clue ofnucleus (which is located towards the basolateral side of the cells) stained withMayer´s hematoxylin (double arrow). The results suggest that the sLex-antigenis expressed at the apical surfaces of the epithelial cells, in response to H. pylori
adherence and stimuli.For immunohistochemistry, sections (5 mm) of inflamed gastric mucosa werestained with the sLex specific mAb ((KM-93), Seikagaku Corp., Tokyo, Japan),in 1:10 dilution, and the immuno reactivity was assessed using the VectastainElite Kit/DAB (Vector Laboratories, NY). After counterstaining in Mayer´shematoxyline, the sections were examined with Zeiss Axioscope and color videocamera (ZVS-47E, Carl Zeiss Inc.).
__________________________________________________________
Print & Media
Su
pp
lem
enta
l T
ab
le 1
. S
um
ma
ry o
f H
elic
ob
act
er p
ylo
ri b
ind
ing
to
gly
cosp
hin
go
lip
ids
No.
Nam
eS
equ
ence
12
Sou
rce
1.L
ewis
bFu
ca2G
alb3
Glc
NA
cb3G
alb4
Glc
b1C
er+
3-
Hum
an s
mal
l int
esti
ne4
Fuca
1
2.G
M3
Neu
Aca
3Gal
b4G
lcb1
Cer
--
Cal
f bra
in
3.G
M1
Gal
b3G
alN
Acb
4Gal
b4G
lcb1
Cer
--
Cal
f bra
in3
Neu
Aca
2
4.G
D1a
Neu
Aca
3Gal
b3G
alN
Acb
4Gal
b4G
lcb1
Cer
--
Cal
f bra
in3
Neu
Aca
2
5.G
D1b
Gal
b3G
alN
Acb
4Gal
b4G
lcb1
Cer
--
Cal
f bra
in3
Neu
Aca
8Neu
Aca
2
6.S
ialy
l-L
ewis
aN
euA
ca3G
alb3
Glc
NA
cb3G
alb4
Glc
b1C
er-
-H
uman
gal
l bla
dd
er c
arci
nom
a4
Fuc a
1
7.S
ialy
l-L
ewis
xN
euA
ca3G
alb4
Glc
NA
cb3G
alb4
Glc
b1C
er-
(+)
Synt
heti
c fr
om S
ymbi
com
Ltd
.
3Fu
c a1
8.S
ialy
l-d
i-L
ewis
xN
euA
ca3G
alb4
Glc
NA
cb3G
alb4
Glc
NA
cb3G
alb4
Glc
b1C
er-
+3
Hum
an g
all b
lad
der
car
cino
ma
3
3
Fuca
1Fu
ca1
1.
Bin
ding
obta
ined
wit
hst
rain
CC
UG
1787
52
.B
indi
ngob
tain
edw
ith
the
ba
bA
1A
2m
utan
tst
rain
.Bin
ding
isde
fine
das
foll
ows:
+3
deno
tes
asi
gnif
ican
tda
rken
ing
onth
eau
tora
diog
ram
whe
n1
pmol
ofsu
bsta
nce
was
app
lied
on
the
thin
-lay
er p
late
; + d
enot
es a
dar
keni
ng a
t 2
nmol
; (+
) an
occ
asio
nal
dark
enin
g at
2 n
mol
; w
hile
– d
enot
es n
o da
rken
ing
even
at
2 nm
ol.
Print & Media
Supplemental Table 2A. Gastric biopsies from 29 patients were scored forbinding by the babA1A2-mutant and the parental strain 17875 in situ, andfor several markers of inflammation.
Biopsy sample 17875-binding babA1A2-binding Lymphocyte (1) PMN cell (2) IH-total IH-gastric pit H. pylori Hist. gastritis
TB 21 5(687) 1(16) 0 0 1 1 0 0,5
TB 16 4(459) 1(14) 0 0 1 1 0 0,5
TB 15 1(38) 0(0) 0 0 1 1 0 1
TB 14 5(712) 1(15) 0 1 3 2 0 0,5
TB 12 3(289) 4(48) 2 2 5 5 3 3
TB 9 5(633) 1(17) 1 1 1 1 0 0
HS+B+A 1(38) 0(0) 1 0 1 1 0 0,5
16519 A3 5(573) 1(13) 2 3 3 2 1 2
14167 A4 5(529) 1(15) 2 1 3 4 1 2
14814 A3 5(667) 1(13) 3 2 1 1 2 3
15393 A3 4(381) 4(51) 3 2 2 3 1 3
8900 A2 5(793) 1(16) 1 1 2 2 2 1
8956 A2 5(605) 2(23) 1 2 1 1 0,5 1
9220 A4 5(593) 2(19) 2 3 4 3 1 2
12486 A4 5(558) 0(0) 1 1 4 3 1 2
15143 A3 2(111) 2(21) 2 2 5 5 1 2
15606 A4 5(629) 3(33) 1 1 3 4 0 3
15981 A3 5(803) 0(0) 2 2 3 4 2 3
16849 A4 5(587) 1(13) 2 5 4 2 1 2
17961 A3 5(839) 3(37) 3 3 4 4 1 2
15754 A5 3(223) 3(29) 2 5 5 5 0,5 2
15754 B5 4(439) 4(43) 1 3 5 5 0,5 1
15187 A5 5(750) 3(38) 3 5 4 3 1 2
15187 B6 5(713) 1(14) 1 3 4 3 1 1
14690 B5 1(39) 4(51) 3 5 4 3 1 3
14322 A4 3(267) 1(15) 2 5 3 3 3 2
14322 B4 4(419) 3(28) 2 5 4 2 2 1
14238 A3 1(38) 4(47) 2 5 5 4 1 3
14238 B6 1(41) 3(38) 1 2 5 4 1 1
Supplemental Table 2B. Gastric biopsies from 6 H. pylori non-infectedindividuals were scored for binding in situ by the babA1A2-mutant and theparental strain 17875.
Biopsy sample 17875-binding babA1A2-binding H. pylori infection
No. 61 3(160) 1(12) 0
No. 62 4(485) 1(10) 0
No. 70 2(68) 0(7) 0
No. 64 4(463) 0(8) 0
D6 5(513) 0(8) 0
F73 2(96) 0(4) 0
Print & Media
Grading scale used for Tables 2A and 2B
CCUG17875-bindingGrading scale from 0 to 5, based on number of bacterial cells/gastric pitGrade 0 = no bacteria, Grade 1 = 1-50 bacteria, Grade 2 = 50-150 bacteriaGrade 3 = 150-350 bacteria, Grade 4 = 350-500 bacteria, Grade 5 ≥500 bacteria
babA1A2-mutant bindingGrading scale from 0 to 5, based on number of bacterial cells/ gastric pitGrade 0 = no bacteria, Grade 1 = 10-15 bacteria, Grade 2 = 15-25 bacteriaGrade 3 = 25-40 bacteria, Grade 4 = 40-55 bacteria, Grade 5 ≥ 55 bacteria
Lymphocyte infiltration (1)Grading scale from 0 to 3, based on both lymphocyte and plasmacell infiltrationGrade 0= normal, Grade 1 = low inflammation, Grade 2 = moderateinflammation, Grade 3 = heavy inflammation
PMN cell infiltration (2)Grading scale from 0 to 5Grade 0 = none, Grade 1 = rare PMN, only in lamina propria (LP), Grade 2 ≤ 1intraepithelial (IE) PMN/high power field (hpf), i.e. 400X magnification, Grade3 = 1-10 IE/hpf, Grade 4 ≥ 10 IE/hpf, Grade 5 = pit abscess
Immunohisto-staining by anti-sLex-mAb (IH)Grading scale from 0 to 5IH-total: IH-staining of whole tissue sectionIH-gastric pit: IH-staining of gastric pit region
H. pylori in surface epithelium (Genta-stained)0.5 very few; 1 good colonization; 2 very good colonization; 3; heavycolonization
Histological gastritis0.5-1 minimal; 2 superficial; 3: intense
__________________________________________________________
Print & Media
References and Notes
1. T. Borén, P. Falk, K. A. Roth, G. Larson, S. Normark, Science 262, 1892(1993).
2. D. Ilver et al., Science 279, 373 (1998).3. A. Dubois et al., Gastroenterology 116, 90 (1999).4. S. Suerbaum, C. Josenhans, A. Labigne, J. Bacteriol. 175, 3278 (1993).5. T. Larsson, J. Bergström, C. Nilsson, K. A. Karlsson, FEBS Letters 469,
155 (2000).6. R. A. Alm et al., Nature 397, 176 (1999).7. J. F. Tomb et al., Nature 388, 539 (1997).8. O. Nørregaard-Jensen, M. Wilm, A. Shevchenko, M. Mann, in Methods in
Molecular Biology, A. J. Link, Ed. (Humana Press, 1999), pp. 571-588.9. J. Ångström et al., Glycobiology 8, 297 (1998)10. S. i. Hakomori, Adv. Cancer Res. 52, 257 (1989).11. K. A. Karlsson. Methods Enzymol. 138, 212 (1987).12. S. B. Levery, E. D. Nudelman, N. H. Andersen, S. i. Hakomori,
Carbohydr. Res. 151, 311 (1986).13. M. Blaszczyk-Thurin et al., J. Biol. Chem. 262, 372 (1987).14. M. Bårström, M. Bengtsson, O. Blixt, T. Norberg, Carbohydr. Res. 328,
525 (2000).15. G. Scatchard, Ann. N.Y. Acad. Sci. 51, 660 (1949).
Print & Media
III
Effects of Helicobacter pylori inoculation on host glycosylation
and H. pylori adhesion sites in rhesus monkey
Sara Lindén+¤, Jafar Mahdavi#, Cara Olsen*, Thomas Borén#, Ingemar Carlstedt+ and André Dubois¤
¤Digestive Diseases Division, Department of Medicine, and United States Military Cancer Institute, Uniformed Services University of the Health Sciences, Bethesda, MD 20814, USA +MucosalBiology Group, Department of Cell- and Molecular Biology, BMC, Lund University, SE-22184 Lund, Sweden, #Department of Odontology/Oral microbiology, Umeå University, SE-90187 Umeå, Sweden; *Department of Preventive Medicine and Biometrics, Uniformed Services University of the Health Sciences, Bethesda, MD 20814, USA.
ABSTRACT
Background & Aims: Adherence of bacteria to the host mucosal surfaces represents a pivotal step in gastrointestinal infections. Gastric mucins, carrying highly diverse carbohydrate structures, present functional binding sites for H. pylori. The aim of this study was to determine the effects of H. pylori inoculation on gastric mucin apoprotein expression/tissue localization, glycosylation and H. pylori
adherence. Methods: Gastric corpus and antral biopsies were harvested at endoscopy and studied using histological methods. Results:After H. pylori inoculation, 8 of 10 rhesus monkeys developed persistent infection, with gastritis score increasing with H. pylori density score. Expression of sialylated Lewis (Le) antigens increased 1 week after infection, and peaked before 4 months in most animals. Lea
and/or Leb expression transiently decreased in 5/7 Leb positive animals 1 week after infection, but later increased. In addition, Lea and/or Leb
expression increased in all Lea/ Leb negative monkeys. In vitro H. pylori adherence with SabA and BabA positive strains follows these glycosylation changes. muc5ac was localized to the mucous cells of the surface/foveolar epithelium and muc6 to the glands during the entire 10-month observation period. Conclusions: H. pylori infection induces time-dependent changes in host mucosal sialyl-Le and Leb expression and subsequently changes the binding sites available for H. pylori
adhesion. We propose that these alterations in
Le expression modulate host-bacterial inter-actions and thereby influence the clinical outcome of the infection.
INTRODUCTION
The stomach is lined by a protective mucus layer formed by high-molecular-mass oligo-meric glycoproteins referred to as mucins. In the healthy stomach, MUC5AC is produced by the surface/foveolar epithelium and MUC6 by the mucous glands.1 Aberrant expression of MUC6 in superficial cells may occur in H.
pylori-induced chronic gastritis, and expression of MUC2, MUC5AC, and MUC6 mucin genes can be altered in gastric precancerous lesions and cancer.2, 3 Oligosaccharides comprise most of the mucin molecular mass, and a single mucin molecule may be substituted with more than 100 different structures.4 Blood-group antigens, which differ depending on the genotype of the individual, are usually present on these oligosaccharides. The carbohydrate structures on the gastrointestinal surface vary according to cell lineage and location along the GI tract. In addition, they act as receptor molecules for microorganisms and undergo changes during malignant transformation.5 In the human stomach, the Lea and the Leb blood-group antigens are mainly expressed in the surface epithelium, whereas the Lex and the Ley
structures primarily are found in the glands, and the Leb structure appears on the MUC5AC mucin.1, 5, 6 Expression of sialylated Le antigens are common in infected and inflamed gastric
Print & Media
H. pylori induced host glycosylation changes
mucosa, and changes in mucin glycosylation occur in the rat small intestine during parasite infection and in human airway mucins fromindividuals with cystic fibrosis or chronic bronchitis. 7-10
H. pylori causes gastric and duodenal ulcers and gastric cancer in a subset of infected subjects.11 During natural infection, most H.
pylori are found in the gastric mucus layer whereas some are attached to the epithelial cells or have entered the mucosa.12, 13 Several adhesins have been implicated in H. pylori
binding, and the Leb and sialyl-Lex/sialyl-Lea
binding adhesins have been extensively investigated.7, 14
H. pylori strains that bind toLeb are associated with severe gastric disease.15
H. pylori binding to gastric mucins is dependent on bacterial strain and host blood-group.6 Rhesus monkeys can be naturally and persistently colonized by H. pylori, which leads to loss of mucus and gastritis.16, 17 Gastric ulcers and cancer also occur in rhesus monkeys.18 In addition, expression and tissue localization of mucins and Lewis antigens as well as mucin structure, density, glycoformsand H. pylori binding properties are similar in rhesus monkey and man.19
Our aim was to study the time course of the effects of H. pylori infection on the gastric mucosa of corpus and antrum with respect to:1) Mucin expression/tissue localization(muc5ac, muc6, muc2 and sulfomucin), 2) mucosal glycosylation (Lea, Leb, Lex, Ley,sialyl-Lea, sialyl-Lex, sialyl-di-Lex, sulfo-Lea)and H. pylori binding sites.
MATERIALS AND METHODS
Materials
Reagents were: polyclonal antibodies against MUC6 (LUM6-320) and MUC2 (LUM2-321),raised in rabbits using KLH-conjugated peptides; a monoclonal antibody against MUC5AC (45M1) (Sigma-Aldrich, Steinheim,Germany); monoclonal anti-Leb antibody (clone 2-25LE) and anti Lea (7LE) (kind gifts from Dr. J. Bara, INSERM, France); anti-Lea (BG5), Leb
(BG6), Lex (BG7) and Ley (BG8) (Signet,Dedham, MA); anti sialyl-Lex (clone KM93) and anti sialyl-Lea (clone IH4) (Seikagaku America, Ijamsville, MD); anti sialyl-di-Lex
(FH6) (kindly provided by Dr. H. Clausen, University of Copenhagen, Denmark); anti sulfo-Lea (Irma) (kindly provided by Dr. E. C. I. Veerman (Vrije Universiteit, The Netherlands); biotinylated goat anti-mouse and goat anti-rabbit sera and Strept AB complex-/HRP (Vector Laboratories, Burlingame, CA); and Biotin blocking system (Dako A/S, Copenhagen, Denmark).
Experimental animals
This research was conducted according to the principles enunciated in the Guide for the Care and Use of Laboratory Animals in a facilityaccredited by the Association for Assessmentand Accreditation of Laboratory Animal CareInternational.22 All procedures involving animals were reviewed and approved by our institutional animal care and use committee. 10monkeys (82A49, 85D08, 86D02, 86D06, 8PZ, 8V5, E6C, F436, F754, and T4C) were cleared of H. pylori and H. heilmanii infection at least6 months earlier and inoculated with a cocktail of 7 H. pylori strains (J258, J166, J238, J170, J178, J254 and J282) including CagA positive,sialyl-Lex and Leb binding ones.7, 23 In 4 of theanimals (82A49, 85D08, 8PZ, and E6C), the effects of inoculation on gastritis and plasmaIgG, IgA and gastrin have been reported previously.23 The monkeys were 4-13 years old (mean 7.1 years). 7 of the monkeys were Lea
and Leb positive (F754, 86D02, T4C, 8V5, F436, 82A49 and 8PZ), whereas in 3 of the monkeys (86D06, E6C, 85D08), less than 0.5% of the cells were stained with the anti-Lea andanti-Leb antibodies and these monkeys were considered Lea/Leb negative. In the resultsection, the Leb status is indicated by /Leb+ or /Leb- after the monkey code. A detailedinvestigation of the Lewis status and tissue localization of Lewis antigens in these monkeyshas been performed.19
2
Print & Media
H. pylori induced host glycosylation changes
Methods
Histochemistry
Genta-stained sections were used to determineinfection status (H. pylori and H. heilmanii)and gastritis score according to the Sydney system (0=none, 1=mild, 2=moderate and 3=marked).24, 25 Sulfo-mucins were detected according to Bravo and Correa.26
Immunohistochemistry
Immunohistochemistry was performedaccording to standard procedures as previouslydescribed.19 For quantitative histochemistry,tissue sections were stained simultaneously toensure that changes in stain intensity were not due to batch-to-batch variations. Quantificationwas carried out either with visual estimation ofintensity (un-sialylated carbohydrate structures) or with a locally developed computer program(sialylated Le antigens) (Jozsef Czege, PhD, Biomedical Instrumentation Center, USUHS). The surface/foveolar epithelium, lamina propria and glands were outlined and the percentage of the area that was stained was calculated as anaverage from 3 fields/view.
Bacterial adherence in situ
In vitro bacterial adherence to tissue sections were analyzed as previously described.27
In
vitro adherence was determined with the Leb/Htype 1-binding BabA-positive strain 176875-/Leb and the sialyl-Lex binding babA1A2
mutant 17875babA1A2, whereas the non-Leb
and non-sialyl-Lex binding strain P1 was used as a negative control.6 The area of adherent bacteria and epithelium were estimated in 10different gastric pit s on each section under200X magnification by using Leica Q-Win(standard v 2.3) and Leica DC 200 (Leica Microsystems AG, Wetzlar, Germany).
Statistics
Comparisons of means were reported from the eight persistently infected animals, and plotted with standard error bars. Changes over timewithin animals were compared using Student’s paired t-test (two-sided). In addition, Wilcoxonsigned-rank tests were calculated for the
variables that were least normally distributed, but the outcome of the test (i.e. whether changes were significantly different from pre-inoculation values or not) was similar with bothmethods. We calculated three differentcorrelation coefficients for each pair of variables using analysis of variance (ANOVA)with random effects. The within-subjects correlation (Rw) measures the extent to which an increase in one variable over time isassociated with an increase in the secondvariable over time.28 The between-subjects correlation (Rb) measures the extent to which monkeys with high values of one variable have high values of a second variable. Rb isessentially the correlation between the means of a certain parameter for each monkey.29 The overall correlation (Ro) measures the extent towhich high values for one variable are associated with high values for a second variable, across all monkeys and time points, thus Ro combines aspects of both Rw and Rb.Correlations were calculated on data from 10 monkeys at 6 time points and the significance of the correlations (p) was determined aftertaking into account the lack of independence ofmultiple observations on the same monkey. P-values less than 0.05 were considered statistically significant. Although somevariables did not appear to follow a normaldistribution, examination of residual plots showed that the assumptions of the significance test for correlation were met. Correlationsbetween variables were compared using abootstrap resampling algorithm implemented in STATA, and are reported as pRo (p value for Ro
bootstrap) and pRw (p value for Rw bootstrap). All analyses were conducted using STATA 7.0 for Windows (StatCorp, College Station, TX).
RESULTS
H. pylori infection and gastritis
After inoculation, 8 of the 10 monkeys(F754/Leb+, F436/Leb+, T4C/Leb+, 82A49/-Leb+, 8PZ/Leb+, 86D06/Leb-, E6C/Leb-,85D08/Leb-) developed persistent infection andgastritis (figure 1), 1 (8V5/Leb+) was transiently infected and 1 (86D02/Leb+) was
3
Print & Media
H. pylori induced host glycosylation changes
not infected. Correlations for gastritis score andH. pylori density are listed in Table 1A.Gastritis score correlated significantly with H.
pylori density both in antrum (p <0.001) and in corpus (p =0.001). The Rb value was higher than that of Rw for these correlations (Table 1A), indicating that individuals respond differently to the same bacterial load. The meancorpus gastritis score, antrum H. pylori scoreand corpus H. pylori score were approximatelytwice as high in Leb negative as in Leb positive persistently infected animals (data not shown).In addition, the animal that did not develop infection and the one that spontaneouslycleared the infection were both Leb positive.Thus, although the number of individuals is too
small to draw general conclusions, the Leb
positive monkeys were less affected by H.
pylori than the Leb negative ones.
Mucin tissue localization remained
unchanged
muc5ac was detected in the surface/foveolar epithelium of antrum and corpus, and muc6 in the antral glands and corpus mucous cells as well as in secreted mucus in the neck region(data not shown). The localization remainedunchanged after inoculation, and muc2 orsulfo-mucins were not detected in the stomachof any of the monkeys during the 10-monthperiod (data not shown)
Figure 1. Mean gastritis ( ) and H. pylori density ( ) scores of the 8 persistently infected monkeys. * p <0.05 vs. pre-inoculation.
4
Print & Media
H. pylori induced host glycosylation changes
Table 1. Significant correlations found in this study.
Correlated variables Ro Rb Rw pA. H. pylori density and gastritis score
gastritis score, A gastritis score, C 0.58 0.73 0.51 <0.001H. pylori density, A H. pylori density, C 0.56 0.66 0.50 <0.001H. pylori density, A gastritis score, A 0.54 0.72 0.43 <0.001H. pylori density, A gastritis score, C 0.57 0.76 0.34 <0.001H. pylori density, C gastritis score, C 0.56 0.82 0.38 0.001H. pylori density, C gastritis score, A 0.32 0.37 0.30 0.012B. Sialylated Le antigens
sialyl-Lea, A. S sialyl-Lea, A. L.P 0.28 0.06 0.43 0.002sialyl-Lea, A. S sialyl-Lea, C. S 0.60 0.45 0.51 <0.001sialyl-Lea, C. S sialyl-Lea, C. gland 0.26 0.07 0.41 0.004sialyl-Lea, C. S sialyl-Lea, C. L.P 0.41 0.27 0.47 <0.001sialyl-Lex, A. S sialyl-Lex, C. gland 0.40 0.72 0.28 0.009sialyl-Lea, A. S sialyl-Lex, A. S 0.42 0.55 0.32 0.003C. sialylated Le antigens and H. pylori density/gastritis score
sialyl-Lex, A. S H. pylori density, A 0.59 0.94 0.24 <0.001sialyl-Lea, A. S H. pylori density, A 0.49 0.55 0.46 <0.001sialyl-Lex, A. S gastritis score, A 0.42 0.78 0.18 0.014sialyl-Lea, A. S gastritis score, A 0.55 0.60 0.52 <0.001D. In vitro adherence of H. pylori strains
In.v.ad. babA1A2, A. S In v.ad. CCUG17875, A. S -0.39 -0.53 -0.22 0.038E. In vitro adherence of H. pylori strains and H. pylori density/gastritis score
In v.ad. CCUG17875, A. S H. pylori density, A -0.47 -0.77 -0.13 0.025In.v.ad. babA1A2, A. S H. pylori density, A 0.58 0.77 0.40 <0.001In v.ad. CCUG17875, A. S gastritis score, A -0.42 -0.56 -0.33 0.005In.v.ad. babA1A2, A. S gastritis score, A 0.45 0.61 0.34 0.001F. In vitro adherence of H. pylori strains and sialylated Le antigens
In.v.ad. CCUG17875, A. S sialyl-Lex, A. S -0.52 -0.78 -0.22 0.004in.v.ad. babA1A2, A. S sialyl-Lex, A. S 0.58 0.71 0.46 <0.001In.v.ad. babA1A2, A. S sialyl-Lea, A. S 0.56 0.75 0.40 <0.001In.v.ad. babA1A2, A. S sialyl-Lex +sialyl-Lea, A. S 0.66 0.84 0.50 <0.001A. =antrum, C. =corpus, L.P. =lamina propria, S.=surface/foveolar epithelium, In.v.ad. =in vitro
adherence of H. pylori strain.
Changes in expression of Lewis antigens
varied between individuals
Lea
In Lea positive monkeys, the surface/foveolar epithelium was Lea positive. One week afterinoculation, the Lea expression was lower in 4 (8V5/Leb+, F754/Leb+, F436/Leb+,86D02/Leb+) and disappeared in one (8V5/Leb+; figure 2B). 2 months afterinoculation, however, the Lea expression had returned to pre-inoculation levels or levels
higher than that (figure 2C). The decrease in Lea stain intensity was uniform in the entirespecimen and across multiple biopsies,although the same percentage of cells tended to be stained. In the Lea negative individuals, the percentage of Lea-positive surface/foveolarepithelium increased at some time points, especially in animal 85D08/Leb- (data not shown). No sulfo-Lea was detected in any of the specimens (data not shown).
5
Print & Media
Bef
ore
inoc
ulat
ion
1 w
eek
post
ino
c.2
mon
ths
post
ino
c.
A
Bef
ore
inoc
ulat
ion
BC
DE
F
GH
I
JK
L
2 m
onth
s
post
ino
c.10
mon
ths
post
ino
c.
Sia
lyl-
Lea
Lea
Leb
Sia
lyl-
Lex
Print & Media
H. pylori induced host glycosylation changes
Leb
In antrum, all surface/foveolar epithelial mucous cells and 20-70 % of the glands were Leb positive in the monkeys that were Lea
positive (data not shown). Followinginoculation, staining did not change in 4 individuals (F754/Leb+, 86D02/Leb+,T4C/Leb+ and 82A49/Leb+) and initiallydecreased in 3 (8PZ/Leb+, F436/Leb+ and8V5/Leb+). Starting at 2-4 months, the Leb
level was approximately back to pre-inoculation levels or higher (similar to the Lea
reactivity in figure 2 A-C). Leb was detected following inoculation in the 3 monkeys classified as Leb negative before inoculation with the increase being particularly strong in E6C/Leb- (figure 2D-F). No change was detected in the non-infected animal (data not shown).
Lex
Before inoculation, Lex was detected in the foveolar epithelium in 8/9 corpus and in 3/10 antral biopsies (data not shown). The corpus glands, chief and parietal cells were stained to a varying degree but the antral glands were negative (data not shown). After inoculation, staining of the mucin-producing cells remained unchanged in 6 animals (F436/Leb+,86D06/Leb-, 82A49/Leb+, 85D08/Leb-,8PZ/Leb+ and the non-infected 86D02/Leb+),increased in 2 (T4C/Leb+ and F754/Leb+),and decreased in 2 (8V5/Leb+ and E6C/Leb-;data not shown).
Ley
In antrum, Ley was expressed in the glands of all monkeys, and in the surface/foveolarepithelium of 2 animals (data not shown). In
corpus, the glands were positive, but not the surface/foveolar epithelium (data not shown). Following inoculation, Ley expression transiently decreased in the antrum of all but 1 monkey (85D08/Leb-) in early infection,whereas no change was detected in corpus or in the non-infected monkey 86D02/Leb+ (data not shown).
Sialylated Lewis antigens transiently
increased after infection
Sialyl-Lea
The percentage of sialyl-Lea positivesurface/foveolar epithelium increased in all 9infected monkeys (figure 2G-I and 3A), and the mean percentage of sialyl-Lea positive cells significantly increased in the persistentlyinfected animals (figure 3B and C). In seven of the nine infected monkeys, this increase was transient (figure 3A). In contrast, no increase was detected in the non-infectedmonkey (data not shown). Sialyl-Lea in antralsurface/foveolar epithelium significantlycorrelated with sialyl-Lea in antral laminapropria (p =0.002) and corpus surface/foveolar epithelium (p <0.001). Incorpus, sialyl-Lea in the surface/foveolar epithelium significantly correlated with sialyl-Lea in the glands (p = 0.004) and in the laminapropria (p <0.001), although the R values were low (table 1B). The correlations withinanimals (Rw) were higher than those (Rb)between them (table 1B) showing that an increase of sialyl-Lea in one tissue localizationis associated with an increase at other siteswithin the same animal at the same time point, but that the levels vary between different animals.
7
Print & Media
H. pylori induced host glycosylation changes
Fi
gure 3. Panel A shows the percentage of sialyl-Lea positive surface/foveolar epithelium in the persistently infected monkeys 85D08/Leb- ( ), 8PZ/Leb+ ( ), E6C/Leb- ( ), F754/Leb+ ( ) and the transiently infected monkey 8V5/Leb+ (+). The 82A49/Leb+ was similar to E6C/Leb- andT4C/Leb+ to 8V5/Leb+ whereas 86D06/Leb- and F436/Leb+ were similar to 85D08/Leb-. The mean percentage of sialyl-Lea positive surface/foveolar epithelium ( ) and glands ( ) for the 8 persistently infected monkeys in antrum and corpus are shown in panels B and C respectively. * indicates that the mean significantly (p <0.05) differed from the pre-inoculation value.
Sialyl-Lex
In all persistently infected monkeys, thepercentage of sialyl-Lex positivesurface/foveolar epithelium transientlyincreased in antrum (5/8 monkeys) and/or corpus (5/8 monkeys), but remained elevated only in 2 animals (fig 2J-L and 4A). The meanpercentage of sialyl-Lex positive antral surface/foveolar epithelium increased from 2%before to 18% at 1-week after inoculation(figure 4B). However, this increase did not reach a level of statistical significance (p=0.053) because the response varied between individuals (figure 4A). Interestingly, sialyl-Lex
levels did not increase in the monkey that
spontaneously cleared the infection (data not shown). The percentage of sialyl-Lex positivesurface/foveolar epithelium was significantly (p =0.011) higher in pre-inoculation biopsies of Leb positive than in Leb negative animals (8 and 76 times higher in antrum and corpus respectively). In post-inoculation biopsies, however, the differences were less pronounced, and sialyl-Lex was twice as high in Leb negative as in Leb positive animals (figure 4A). A significant correlation was found between sialyl-Lex and sialyl-Lea in surface/foveolarepithelium of antrum (p =0.003), but not in corpus (Table 1B).
8
Print & Media
H. pylori induced host glycosylation changes
Before inoculation, all monkeys were negative for sialyl-di-Lex except for a few positive cellsin the glands of the antrum and corpus in 2 animals (85D08 and 82A49). The only postinoculation change detected was a moderateincrease in 2 monkeys (86D02 and 86D06) at 10 months (data not shown). Thus, this structure does not seem to be a major adhesion factor for H. pylori in rhesus monkeys, at leastup to10 months.
H. pylori density and gastritis correlated
with sialylated Lewis antigens
H. pylori density and gastritis score significantly correlated with sialyl-Lex (p<0.001 and p =0.014) and sialyl-Lea (p <0.001 for both) expression in surface/foveolar epithelium of antrum. The correlation to sialyl-
Lex is mainly between animals (Rb), showing that individuals with high H. pylori density and gastritis also have high levels of sialyl-Lex, but not necessarily in the same tissue specimen/time point (Table 1C). In contrast, the correlations to sialyl-Lea have a higher within animal correlation (Rw) suggesting that sialyl-Lea correlates to the H. pylori density andgastritis score in the same biopsy (Table 1C). Although the correlation between sialylated antigens and gastritis score was higher than that to H. pylori density, the difference between those correlations was not significant (pRo
=0.570, pRw = 0.112). Thus, it can not be concluded that the correlation betweensialylation and H. pylori density occur because H. pylori density and gastritis score correlate.
Figure 4. Panel A shows the percentage sialyl-Lex positive surface/foveolar epithelium in the persistently infected monkeys T4C/Leb+ ( ), 85D08/Leb- ( ), 8PZ/Leb+ ( ), E6C/Leb- ( ),86D06/Leb- ( ), 82A49/Leb+ ( ), F754/Leb+ ( ), F436/Leb+ ( ). The mean sialyl-Lex positivepercentage of surface/foveolar epithelium ( ) and glands ( ) for the 8 persistently infectedmonkeys are shown in panel B (antrum) and C (corpus).
9
Print & Media
H. pylori induced host glycosylation changes
In vitro adherence of H. pylori follows the
glycosylation changes
Leb-binding strain 17875/Le
b:
In vitro adherence of strain 17875/Leb
significantly (p =0.028) decreased early during infection (figure 5A). In vitro adherence to the mucosa of the transiently infected monkey(8V5/Leb+) initially decreased as in thepersistently infected animals, but later increased as in the non-infected one with the increase preceding clearance of infection(figure 5A). No in vitro adherence to the antral glands was detected in any of the monkeys(data not shown), although Leb was detected inthe glands of most animals, supporting the theory that the topographic presentation of this structure is important for binding.6 In Leb-positive individuals, in vitro adherence to the antral surface epithelium was slightly higherbefore inoculation (mean 18 % vs 15%), and significantly higher after inoculation (1.8 times,p =0.007), than in Leb-negative individuals. However, adherence occurred in all individuals, indicating that structures other than Leb (e.g. H type 1) also mediate binding.
Sialyl-Lex binding babA1A2 mutant:
In vitro adherence to antral surface/foveolar epithelium significantly (p =0.015) increased in persistently infected monkeys (figure 5B). Themean adherence to antral glands was 20% andto corpus 7%, and it did not change significantly after inoculation (data not shown). In Leb negative animals, adherence of the babA1A2 mutant to antral surface/foveolar epithelium was twice as high as in the Leb
positive ones (data not shown). In vitro
adherence to antral surface/foveolar epithelium
progressed in opposite directions during infection compared to the Leb binding H. pylori
strain (17875/Leb) (compare figure 5A and B), and adherence of those two strains inverselycorrelated (p =0.038; table 1D). No in vitro
adherence was detected with AlpA positive strain p1 (data not shown).
In vivo H. pylori density, gastritis score and
sialylated Le antigens inversely correlated
with in vitro H. pylori adherence of the
Leb/H-type-1 binding BabA positive strain.
In vivo H. pylori density and gastritis score inversely correlated with the in vitro adherenceof strain 17875/Leb (p =0.025 vs p =0.005; table 1E). In addition, in vitro adherence of the Leb binding strain 17875/Leb correlatedinversely with sialyl-Lex (p =0.004; Table 1F). The correlation is higher between individuals (Rb; Table 1F) than within them, indicating thatanimals with high sialyl-Lex levels bind less 17875/Leb, but that this is not necessarilyreflected in individual biopsies.
In vivo H. pylori density, gastritis score and
sialylated Le antigens correlated with in vitro
H. pylori adherence of the sialyl-Lex
binding
SabA positive strain.
In vivo H. pylori density and gastritis score correlated with in vitro adherence of the babA1A2 mutant (p <0.001 vs p =0.001; table 1E) to antral surface/foveolar epithelium. In
vitro adherence of the sialyl-Lex binding babA1A2 mutant correlated significantly withsialyl-Lex (p <0.001) and sialyl-Lea (p <0.001) in surface/foveolar epithelium of antrum (Table 1F), and these correlations did not differsignificantly (PRo =0.729, pRw = 0.625).
10
Print & Media
H. pylori induced host glycosylation changes
Figure 5. Mean in vitro adherence of the BabA positive strain (17875/Leb; panel A) and of the SabA positive, BabA negative mutant (BabA1A2: panel B) to antral surface/foveolar epithelium of 8 persistently infected monkeys ( ), the transiently infected one (8V5/Leb+, ) and the non-infected one (86D02/Leb+, ).
DISCUSSION
The present study demonstrates that H. pylori
infection can induce time-dependent modifica-tions in mucin glycosylation. During the early phase of infection, Leb decreased and sialyl-Lex
as well as sialyl-Lea increased in the surface/foveolar epithelium. From 4 through 10 months after inoculation, sialyl-Lea remainedhigher than before inoculation whereas sialyl-Lex returned to pre-inoculation levels and Leb
increased. Thus, the rhesus monkey immedi-ately responds to H. pylori infection by modi-fying the expression of carbohydrate structures important in adherence of this microbe, such as Leb, sialyl-Lea and sialyl-Lex. These early alterations in glycosylation are reflected in thein vitro adherence test by an initial decrease inadherence of the Leb binding H. pylori strain
and an increase in adherence of the Lex binding strain. In subsequent months, in vitro adherence of the two H. pylori strains to antral surface/foveolar epithelium continued to progress in opposite directions, supporting the hypothesis that the BabA and SabA adhesins play complementary roles at different timepoints of infection.7
In contrast, muc5ac and muc6 tissue localiza-tion remained unchanged in all animals during the 10-month period. However, the surface/foveolar cells lost their bulginess and looked flat and depleted of mucins in somebiopsies. KATO III cell mucin synthesis haspreviously been shown to be partially inhibited when cells were incubated with H. pylori,although no effect was found on mucin secre-
11
Print & Media
H. pylori induced host glycosylation changes
tion, which could explain the depleted appear-ance.30 No muc2 was detected in the gastricspecimens. Possibly, aberrant expression of muc6 or muc2 requires a >10-month exposure to H. pylori or was present at a level that is notdetected by the present methods.
Changes in glycosylation were detected in all animals. As expected from a predominantlyantral infection, changes and correlations were more pronounced at this location. The majorityof the tested correlations were significant withp values generally less than 0.01 and thus it is highly likely that correlations are real and not occurring by chance. In all correlations with H.
pylori density, the values of Rb are higher than those of Rw, demonstrating that individuals differ in their response to this pathogen.
One week after inoculation, the Lea stain inten-sity was markedly lower in 4/7 Lea positive monkeys. Of these, only 2 became persistentlyinfected, and then only mildly (mean postinoculation gastritis score 1.4 and H. pylori
density 0.4), suggesting that this decrease could have a protective function. In total, a decrease in Lea and/or Leb was detected in 5/7 Lea/Leb
positive monkeys. In addition, an initial decrease in fucosylation affecting other struc-tures than Lea/Leb seems to occur since in vitro
adherence with H. pylori strain 17875/Leb,which binds to fucosylated antigens like Leb
and H type 1, decreased in 7/8 infected monkeys, including the Leb negative ones. Although Lea and Leb expression initially decreased in the Leb positive animals, there wasa tendency to an increase later during infection.In addition, in the three animals that were Leb
and Lea negative before inoculation, Leb and/or Lea were detected after inoculation. A similareffect is induced by some commensal bacteria (e.g. B. thetaiotaomicron) that can signal to thehost to produce more glycosidically boundfucose.31 In some animals, this increase tendedto be strongest apically (figure 2C). We haveobserved that in the non-infected stomach, the Leb binding BabA positive strains seem to bind both to Leb on secreted mucin and to H type 1
on a putative membrane associated mucin.6
(Lindén, S. Mahdavi, J. Hedenbro, J. Borén T. and Carlstedt, I., Unpublished). However, adhesion to Leb is stronger than to H type 1.6, 32
After infection, when fucosylation changes and Leb seems to increase apically on thesurface/foveolar cells, it is likely that thisproportion is changed. Competition between H.
pylori binding to membrane-bound and secreted mucins is likely to influence the outcome of the host-microbe interaction, since the membrane-bound ones have potential for signaling.33, 34
The increases in sialyl-Lea and sialyl-Lex
expression are most prominent during the first 2 months after inoculation. In humans, sialyl-Lea expression was higher before than after eradication, but only in about 50% of speci-mens,8 as explained by the change observed in monkey being time-dependent. In situ H. pylori
density and in vitro adherence of the SabA positive mutant, correlated significantly withsialyl-Lea and sialyl-Lex expression in antrum (table 1), supporting the theory that these anti-gens function as receptors for the SabA adhesin.7 In addition, SabA may also bind other sialylated structures that are not recognized by the antibodies used in this study.
H. pylori density and gastritis score tended to be lower in Leb positive than Leb negative monkeys and the animal that did not develop infection as well as the one that spontaneouslycleared the infection were both Leb positive.Although the number of individuals is too smallto draw general conclusions, attachment of H.
pylori to Leb-positive gastric mucin may reduce colonization by inhibiting attachment of the organism to epithelial cells, thus assisting removal of the bacteria by the continuous flow of mucus and decreasing colonization of the underlying mucosa. The Leb host phenotype is not associated with H. pylori infection, although patients who express Leb have a higher H. pylori density than those who do not, and H. pylori density increases with Leb
expression.35-37 These seemingly contradictory
12
Print & Media
H. pylori induced host glycosylation changes
results may be explained by our results that Leb
expression increase both in Leb positive and Leb negative monkeys later during infection, since studies in humans do not account for H.
pylori induced changes.
In conclusion, the present study demonstratesthat H. pylori infection can modify host glyco-sylation within mucus producing cells without modifying mucin apoprotein expression or tissue localization. The host glycosylation changes are time-dependent and the rapid response influences the structure of putative colonization targets. These structural changesprobably result in bacterial adaptation and selection of those bacterial strains that are bestadapted to a particular host. These host-bacte-rial interactions may determine the outcome of the infection, i.e. persistence or transience, and perhaps also whether colonization will causesubsequent overt disease.
REFERENCES
1. De Bolos C, Garrido M, Real FX. MUC6 apomucin shows a distinct normal tissue distribution that correlateswith Lewis antigen expression in thehuman stomach. Gastroenterology 1995;109:723-34.
2. Byrd JC, Yan P, Sternberg L, Yunker CK, Scheiman JM, Bresalier RS. Aberrant expression of gland-type gastric mucin in the surface epitheliumof Helicobacter pylori-infected patients. Gastroenterology 1997;113:455-64.
3. Buisine MP, Devisme L, Maunoury V, Deschodt E, Gosselin B, Copin MC, Aubert JP, Porchet N. Developmentalmucin gene expression in the gastroduodenal tract and accessorydigestive glands. I. Stomach. A relationship to gastric carcinoma. J Histochem Cytochem 2000;48:1657-66.
4. Klein A, Carnoy C, Lamblin G, Roussel P, van Kuik JA, Vliegenthart JF. Isolation and structural characterization of novel sialylated oligosaccharide-alditols from respiratory-mucus
glycoproteins of a patient suffering from bronchiectasis. Eur J Biochem1993;211:491-500.
5. Murata K, Egami H, Shibata Y, Sakamoto K, Misumi A, Ogawa M. Expression of blood group-related antigens, ABH, Lewis(a), Lewis(b), Lewis(x), Lewis(y), CA19-9, and CSLEX1 in early cancer, intestinalmetaplasia, and uninvolved mucosa of the stomach. Am J Clin Pathol 1992;98:67-75.
6. Lindén S, Nordman H, Hedenbro J, Hurtig M, Borén T, Carlstedt I. Strain- and blood group-dependent binding of Helicobacter pylori to human gastric MUC5AC glycoforms.Gastroenterology 2002;123:1923-30.
7. Mahdavi J, Sonden B, Hurtig M, OlfatFO, Forsberg L, Roche N, Angstrom J, Larsson T, Teneberg S, Karlsson KA,Altraja S, Wadstrom T, Kersulyte D, Berg DE, Dubois A, Petersson C, Magnusson KE, Norberg T, Lindh F, Lundskog BB, Arnqvist A, Hammarstrom L, Boren T. Helicobacter pylori SabA adhesin in persistent infection and chronic inflammation.Science 2002;297:573-8.
8. Ota H, Nakayama J, Momose M, Hayama M, Akamatsu T, Katsuyama T, Graham DY, Genta RM. Helicobacter pylori infection produces reversible glycosylation changes to gastric mucins.Virchows Arch 1998;433:419-26.
9. Karlsson NG, Olson FJ, Jovall PA, Andersch Y, Enerback L, Hansson GC. Identification of transient glycosylation alterations of sialylated mucin oligosaccharides during infection by the rat intestinal parasite Nippostrongylus brasiliensis. Biochem J 2000;350 Pt 3:805-14.
10. Lamblin G, Degroote S, Perini JM, Delmotte P, Scharfman A, Davril M, Lo-Guidice JM, Houdret N, Dumur V, Klein A, Rousse P. Human airway mucin glycosylation: a combinatory of
13
Print & Media
H. pylori induced host glycosylation changes
carbohydrate determinants which vary in cystic fibrosis. Glycoconj J 2001;18:661-84.
11. Uemura N, Okamoto S, Yamamoto S, Matsumura N, Yamaguchi S, YamakidoM, Taniyama K, Sasaki N, SchlemperRJ. Helicobacter pylori infection and the development of gastric cancer. N Engl J Med 2001;345:784-9.
12. Shimizu T, Akamatsu T, Sugiyama A, Ota H, Katsuyama T. Helicobacter pylori and the surface mucous gel layerof the human stomach. Helicobacter 1996;1:207-18.
13. Semino-Mora C, Doi SQ, Marty A, Simko V, Carlstedt I, Dubois A. Intracellular and interstitial expressionof Helicobacter pylori virulence genes in gastric precancerous intestinal metaplasia and adenocarcinoma. J Infect Dis 2003;187:1165-77.
14. Ilver D, Arnqvist A, Ogren J, Frick IM, Kersulyte D, Incecik ET, Berg DE, Covacci A, Engstrand L, Boren T. Helicobacter pylori adhesin binding fucosylated histo-blood group antigens revealed by retagging. Science1998;279:373-7.
15. Prinz C, Schoniger M, Rad R, Becker I, Keiditsch E, Wagenpfeil S, Classen M, Rosch T, Schepp W, Gerhard M. Key importance of the Helicobacter pylori adherence factor blood group antigen binding adhesin during chronic gastric inflammation. Cancer Res 2001;61:1903-9.
16. Dubois A, Fiala N, Weichbrod RH, Ward GS, Nix M, Mehlman PT, TaubDM, Perez-Perez GI, Blaser MJ. Seroepizootiology of Helicobacter pylori gastric infection in nonhumanprimates housed in social environments. J Clin Microbiol 1995;33:1492-5.
17. Dubois A, Fiala N, Heman-Ackah LM,Drazek ES, Tarnawski A, Fishbein WN, Perez-Perez GI, Blaser MJ. Natural gastric infection with Helicobacterpylori in monkeys: a model for spiral
bacteria infection in humans. Gastroenterology 1994;106:1405-17.
18. Parker GA, Gilmore CJ, Dubois A. Spontaneous gastric ulcers in a rhesus monkey. Brain Res Bull 1981;6:445-7.
19. Lindén S, Borén T, Dubois A, Carlstedt I. Rhesus monkey gastric mucins and their Helicobacter pylori binding properties. Biochem J;Conditionally accepted.
20. Nordman H, Davies JR, Lindell G, de Bolos C, Real F, Carlstedt I. Gastric MUC5AC and MUC6 are large oligomeric mucins that differ in size, glycosylation and tissue distribution. Biochem J 2002;364:191-200.
21. Herrmann A, Davies JR, Lindell G, Martensson S, Packer NH, Swallow DM, Carlstedt I. Studies on the "insoluble" glycoprotein complex from human colon. Identification of reduction-insensitive MUC2 oligomers and C-terminal cleavage. J Biol Chem 1999;274:15828-36.
22. Commission on Life Sciences. Guide for the Care and Use of Laboratory Animals. National Research Council, National Academy Press, Washington,DC, 1996.
23. Dubois A, Berg DE, Incecik ET, Fiala N, Heman-Ackah LM, Del Valle J, Yang M, Wirth HP, Perez-Perez GI, Blaser MJ. Host specificity of Helicobacter pylori strains and host responses in experimentally challenged nonhuman primates. Gastroenterology 1999;116:90-6.
24. Genta RM, Robason GO, Graham DY. Simultaneous visualization of Helicobacter pylori and gastric morphology: a new stain. Hum Pathol 1994;25:221-6.
25. Dixon MF, Genta RM, Yardley JH, Correa P. Classification and grading of gastritis. The updated Sydney System. International Workshop on the Histopathology of Gastritis, Houston
14
Print & Media
H. pylori induced host glycosylation changes
1994. Am J Surg Pathol 1996;20:1161-81.
26. Bravo JC, Correa P. Sulphomucinsfavour adhesion of Helicobacter pylori to metaplastic gastric mucosa. J Clin Pathol 1999;52:137-40.
27. Falk P, Roth KA, Boren T, WestblomTU, Gordon JI, Normark S. An in vitro adherence assay reveals that Helicobacter pylori exhibits cell lineage-specific tropism in the humangastric epithelium. Proc Natl Acad Sci U S A 1993;90:2035-9.
28. Bland JM, Altman DG. Calculating correlation coefficients with repeated observations: Part 1- -Correlation within subjects. Bmj 1995;310:446.
29. Bland JM, Altman DG. Calculating correlation coefficients with repeated observations: Part 2- -Correlation between subjects. Bmj 1995;310:633.
30. Byrd JC, Yunker CK, Xu QS, Sternberg LR, Bresalier RS. Inhibition of gastric mucin synthesis by Helicobacter pylori. Gastroenterology 2000;118:1072-9.
31. Bry L, Falk PG, Midtvedt T, Gordon JI. A model of host-microbial interactions in an open mammalian ecosystem.Science 1996;273:1380-3.
32. Boren T, Falk P, Roth KA, Larson G, Normark S. Attachment of Helicobacter pylori to human gastric epitheliummediated by blood group antigens. Science 1993;262:1892-5.
33. Gendler SJ, Spicer AP. Epithelial mucingenes. Annu Rev Physiol 1995;57:607-34.
34. Quin RJ, McGuckin MA. Phosphorylation of the cytoplasmicdomain of the MUC1 mucin correlates with changes in cell-cell adhesion. Int J Cancer 2000;87:499-506.
35. Heneghan MA, Moran AP, Feeley KM, Egan EL, Goulding J, Connolly CE, McCarthy CF. Effect of host Lewis and ABO blood group antigen expression on Helicobacter pylori colonisation density and the consequent inflammatory
response. FEMS Immunol Med Microbiol 1998;20:257-66.
36. de Mattos LC, Rodrigues Cintra J, Sanches FE, Alves da Silva Rde C, Ruiz MA, Moreira HW. ABO, Lewis, secretor and non-secretor phenotypes in patients infected or uninfected by the Helicobacter pylori bacillus. Sao Paulo Med J 2002;120:55-8.
37. Sheu BS, Sheu SM, Yang HB, Huang AH, Wu JJ. Host gastric Lewis expression determines the bacterialdensity of Helicobacter pylori in babA2 genopositive infection. Gut2003;52:927-32.
ACKNOWLEDGEMENTS
The work was supported by grants from Sixten Gemzeus Stiftelse; the Swedish Foundation for Strategic Research (Glycoconjugates in Biological Systems); J. C. Kempe MemorialFoundation; the Swedish Research Council (grants 7902 and 11218); the Swedish Cancer Society (Project No. 4101-B00-03XAB);Österlunds Stiftelse; Council for Medical Tobacco Research, Sweden; the County Council of Västerbotten; the MizutaniFoundation for Glycoscience; and the National Institutes of Health (grant CA82312).
The opinions and assertions contained herein are the private ones of the authors and are notto be construed as official or reflecting theviews of the Department of Defense or the Uniformed Services University of the Health Sciences.
Presented in part at the 2003 Digestive Diseases Week in Orlando, Florida and the 2003 American Society for Biochemistry and Molecular Biology meeting in San Diego, California.
Address requests for reprint to: Andre Dubois, MD, PhD, Dept of Medicine, USUHS, RoomA3075, Bethesda MD, 20814, Tel: 301-295-3607 Fax: 301-295-3676 or 301-295-3557,[email protected]
15
Print & Media
IV
INFECTION AND IMMUNITY, May 2003, p. 2876–2880 Vol. 71, No. 50019-9567/03/$08.00�0 DOI: 10.1128/IAI.71.5.2876–2880.2003Copyright © 2003, American Society for Microbiology. All Rights Reserved.
NOTES
Limited Role of Lipopolysaccharide Lewis Antigens in Adherence ofHelicobacter pylori to the Human Gastric Epithelium
Jafar Mahdavi,1 Thomas Boren,1 Christina Vandenbroucke-Grauls,2and Ben J. Appelmelk2*
Department of Odontology/Oral Microbiology, Umea University, 901 87 Umea, Sweden,1 andDepartment of Medical Microbiology, Vrije Universiteit Medical Center,
1081 BT Amsterdam, The Netherlands2
Received 3 October 2002/Returned for modification 18 November 2002/Accepted 17 January 2003
In vitro and in vivo studies from various groups have suggested that Helicobacter pylori lipopolysaccharide(LPS) Lewis x (Lex) antigens mediate bacterial adhesion. We have now reevaluated this hypothesis by studyingthe adherence in situ of H. pylori strain 11637 and its corresponding Lex-negative rfbM mutant to humangastric mucosa from patients (n � 22) with various gastric pathologies. Significant binding of the parent strainwas observed in only 8 out of 22 sections; in four out of eight patients, the Lex-negative mutant bound less well.One of these four patients displayed no gastric abnormalities, and the other three showed dysplasia, meta-plasia, and adenocarcinoma, respectively; hence, we are unable to define the circumstances under whichLPS-mediated adhesion takes place. We conclude that H. pylori LPS plays a distinct but minor role in adhesion.
More than 80% of Helicobacter pylori strain express Lewisblood group antigens on their lipopolysaccharide (LPS); mostoften, Lewis x (Lex) and/or Ley is expressed (2, 13). Thus, incontrast to the case for many other pathogens (for example,Escherichia coli), the H. pylori LPS O antigen is strongly con-served. This suggests that it may play a role in pathogenesisother than its role as a (weak) endotoxin. Despite intenseefforts, no such role has been assigned with certainty, but thefollowing three options have been investigated in some detail(2): (i) H. pylori evades the host immune response by molecularmimicry of its Lex antigen with similar blood group antigens inthe gastric mucosa, (ii) H. pylori induces a gastric autoimmuneresponse by inducing anti-Lex antibodies that bind to its LPSand also to gastric epithelium, or (iii) H. pylori LPS coulddemonstrate adhesive properties.
The data to support a role of H. pylori LPS in adhesion areas follows. Inactivation of H. pylori genes which encode glyco-syltransferases of importance for LPS glycosylation patternsyields mutants that do not express certain Lewis antigens. In aseries of studies, such Lewis antigen-negative mutants colo-nized less well in experimental mouse infection studies (9, 7,11). In a recent study it was shown that Lewis antigen-negativeLPS mutants did not adhere to the gastric mucosa; that studywas based on adherence of bacterial cells in vitro, i.e., tohistological tissue sections (5). Moreover, synthetic Lex appliedto the surface of fluorescent latex beads bound to the gastricsections in patterns similar to the adherence “blueprint” dis-played by H. pylori bacterial cells. This series of results suggests
that the Lex LPS antigens might confer adhesive properties toH. pylori (5). However, the results were based on the charac-terization of adherence properties of a single H. pylori strainand its isogenic mutants, and for the binding studies histolog-ical sections from a single patient with gastric carcinoma wereused (G. Faller, personal communication). In contrast, in sev-eral studies the H. pylori BabA adhesin has consistently beendemonstrated to mediate binding of bacterial cells to gastricepithelial cells, specifically the mucosal Leb blood groupepitope (4, 6).
The results described above raise the question of when H.pylori would use LPS-carbohydrate-based interactions for ad-herence, as alternatives to regular adhesin proteins. Could LPSbinding possibly optimize targeting of the microbe to uniquemicroniches, or could it possibly increase binding strength bythe use of multiple binding sites simultaneously (multivalentbinding)? Alternatively, would the mechanisms of phase vari-ation as described for expression of LPS antigen variation (1,3) endow the microbe with adhesive properties that would leadto escape to the immune response? One obvious answer for anLPS-dependent binding activity would be to complementBabA adhesin-mediated adherence, such as in individuals whodo not express Leb antigen (nonsecretor individuals), or formultivalent glycan-glycan interactions with highly glycosylatedstructures such as gastric mucins.
Thus, we decided to reevaluate the functional role of LPSantigens in adhesion of H. pylori. In this work we studied theinfluence of LPS Lex antigen expression among H. pyloristrains that express (positive) or do not express (negative) theBabA adhesin for its role in adherence to gastric sections frommany different patients, including an individual of the nonse-cretor phenotype.
The strains used are shown in Table 1. Most strains have
* Corresponding author. Mailing address: Department of MedicalMicrobiology, VUMC Vrije Universiteit Medical Center, van derBoechorststraat 7, 1081 BT Amsterdam, The Netherlands. Phone: 3120 4448297. Fax: 31 20 4448318. E-mail: [email protected].
2876
Print & Media
been described before (1, 3, 6, 4), apart from strains ATCC45304 KO babA and K4.1 KO babA, which were derived fromtheir respective parent strains by insertional activation of thebabA gene as described previously (4, 6). Serotyping with anti-Lewis monoclonal antibodies was done as described previously(13). Expression of a functional BabA adhesin was investigatedthrough binding studies with Leb coupled to human serumalbumin (Isosep, Tullinge, Sweden) as described previously (4,6). Briefly, the neoglycoprotein was labeled with 125I by the
chloramine T method, and approximately 20,000 cpm of la-beled material was incubated with bacterial cells Binding ofLeb to bacterial cells was expressed as the percentage of countsper minute added. Binding studies were performed on three orfour independent occasions. In situ binding of bacteria to gas-tric tissue was done essentially as described previously (6, 4).Briefly, bacteria were grown on agar plates, washed, and la-beled with fluorescein isothiocyanate, and 200 �l of bacterialsuspension with an optical density at 600 nm of 0.2 was incu-
TABLE 1. Strains used in this study and their LPS phenotypes, expression of functional BabA, and ability to bind to healthy gastric mucosa
Strain testeda LPS phenotype Leb bindingb
In situ adhesion to gastric epitheliumb
(mean � SEM)Reference(s)
Surfacemucus cells Parietal cells Deep glands
17875 No Lex or Ley 44 � 2.06 647 � 24 0 0 4, 617875 babA KO No Lex or Ley �1 � 0.03 36 � 5 0 0 4, 6ATCC 43504 Lex, Ley 25 � 1.98 545 � 13 0 0 1, 3ATCC 43504 variant 1b Ley 29.5 � 0.97 596 � 22 21 � 4 0 1, 3ATCC 43504 variant 1c Lex, Ley 32.8 � 1.0 607 � 9 0 28 � 2 1, 3ATCC 43504 variant D1.1 No Lex or Ley 29.8 � 1.22 537 � 11 25 � 5 79 � 7 1, 3ATCC 43504 variant K4.1 i-antigen, H type I 24.8 � 1.5 512 � 7 35 � 3 113 � 9 1, 3ATCC 43504 babA KO Lex, Ley �1 � 0.08 298 � 19 0 0 This studyK4.1 babA KO i-antigen, H type I �1 � 0.06 259 � 28 34 � 4 108 � 13 This study4187 Lex, Ley 3.3 � 0.25 391 � 8 0 18 � 2 34187 KO 379/651 i-antigen, H type I 1.3 � 0.12 538 � 14 0 84 � 9 3NCTC 11637 Lex, Ley �1 � 0.03 0 0 0 5NCTC 11637 galE mutant No Lex or Ley �1 � 0.07 0 0 0 5NCTC 11637 rfbM mutantc No Lex or Ley �1 � 0.03 0 0 0 5
a Compared to the NCTC 11637 parent, variant 1b phase varies in an as-yet-unidentified N-acetylglusosaminyltransferase, 1c varies in �3-fucosyltransferase HP0651,K4.1 varies in �3-fucosyltransferase HP0379, and D1.1 varies both in N-acetylglusosaminyltransferase and HP0379. In strain 4187 KO 379/651, both HP0379 andHP0379 are inactivated. The Lewis phenotypes of all strains were tested repeatedly and found to be stable.
b Based on two to four independent experiments. Bacterial adherence was defined as the number of bacteria per gastric pit.c rfbM codes for GDP mannose pyrophosphorylase, an enzyme specific for LPS biosynthesis.
TABLE 2. In situ binding of H. pylori strain NCTC 11637 and its Lewis x-negative rfbM mutant to gastric surface mucus epithelia frompatients of diverse histopathological status
Patient
Binding (mean � SEM)a of strain Patient characteristics
NCTC 11637 rfbM KO Gastric histopathologyb ABO bloodgroup
Lewis bloodgroup
1 0 0 Normal A Leb
2 0 0 Gastritis, atrophy B Leb
3 57 � 6 26 � 8 Normal A Leb
4 0 0 Normal NDc Leb
5 0 0 Gastritis, atrophy ND Leb
6 134 � 8 38 � 7** Dyplasia A Leb
7 0 0 Gastritis ND Leb
8 140 � 6 156 � 8 Normal A Leb
9 0 0 Normal ND Leb
10 6 � 2 0 Normal ND Leb
11 10 � 3 7 � 4 Normal ND Leb
12 0 20 � 3* Normal A Leb
13 306 � 21 32 � 6** Gastritis, atrophy, metaplasia ND Leb
14 0 0 Dysplasia ND Leb
15 0 0 Dysplasia ND Leb
16 119 � 18 59 � 13 Gastritis O Leb
17 336 � 8 283 � 66 Hyperplasia ND Leb
18 329 � 12 30 � 8** Normal ND Leb negative19 0 0 Adenocarcinoma ND Leb
20 331 � 12 25 � 11** Adenocarcinoma ND Leb
21 0 0 Adenocarcinoma ND ND
a Based on two to four independent experiments. Significant differences in binding between parent and rfbM mutant: *, P � 0.05, **, P � 0.01.b For an overview of gastric histopathology, see reference 10.c ND, not done.
VOL. 71, 2003 NOTES 2877
Print & Media
bated with deparaffinated gastric sections from the corpus orantrum. After FITC labeling, the bacteria are dead. Bacterialadherence was evaluated by visually counting the number ofadherent bacteria per gastric pit for 10 pits, each in a differentmicroscopic field (magnification, �200). In situ binding studieswere done on two to four independent occasions, and hence intotal 20 to 40 pits (fields) were evaluated per strain tested. Theaverage number of adherent bacteria per pit was calculated, aswell as the standard error of the mean. Significant binding wasdefined as �50 bacteria per gastric pit. Differences in bindingwere calculated with the Student’s t test. In initial studies(Table 1), three zones of binding were discriminated: surfacemucus cells, the parietal cell region, and the glandular region.In later studies (see Table 2), only the most relevant zone, i.e.,the surface mucus epithelium, was evaluated.
Table 1 shows that LPS structure had no influence on in situadherence to surface mucus cells. The isogenic strains notexpressing Lex and or Ley adhere most similarly to the parentalstrain. LPS phase variants D1.1 and K4.1 were compared withthe parental strain ATCC 43504, and the �3-fucosytransferasedouble-knockout strain 4187 KO 379/651 was compared withthe parental strain 4187. Interestingly in comparison to thewild-type strains, D1.1, K4.1, and 4187 KO 379/651 actuallydemonstrated some adherence to the deeper glandular regionalso (binding of K4.1 is shown in Fig. 1), which suggests theunmasking of additional binding properties. Strains K4.1 and4187 KO 379/651 strongly express the H type I and i antigenepitopes, and hence adhesion might be H type I (or i-antigen)mediated.
In contrast, inactivation of babA in strain 17875 stronglyaffected adherence. We reasoned that the observed lack ofeffect of LPS structure on adhesion might be due to a BabA-Leb high-affinity-mediated interaction, dominant over LPS-mediated binding, with the gastric mucosa for strains NCTC43504 and 4187 (and their phase variants and mutants, respec-tively). We therefore decided to investigate the expression of afunctional BabA adhesin activity through Leb-neoglycoproteinbinding studies for a variety of strains. Table 1 shows that
strain 17875 bound Leb very well, while its corresponding babAknockout strain was devoid of Leb binding activity. Likewise,strains ATCC 43504 and its phase variants, as well as strains4187and 4187 KO 379/651, were found to bind Leb, althoughthe latter two strains were less active. Hence, a potential effectof LPS structure on in situ adhesion might be concealed byBabA-dependent binding. For this reason, we constructedbabA knockout mutations in strain ATCC 43504 and its Lex-negative phase variant K4.1. Inactivation indeed led to a com-plete loss of Leb binding. However, again no effect of LPSstructure on adhesion was observed.
The original observation (5) of the role of LPS in adhesionwas made with strain NCTC 11637 and its Lex-negative mu-tants (galE and rfbM knockout strains), with the rfbM knockoutstrain showing the largest of decrease in adhesion. Further-more, the binding data presented in Table 1 were obtainedwith serial sections from a single patient without gastric abnor-malities. Strikingly, strain 11637 does not bind Leb and alsodoes not bind to gastric tissue of this patient; the reason for thislack of binding to Leb is not known. Recently we showed thatgastric inflammation affects H. pylori adhesion (8). We there-fore decided to test NCTC 11637 and its rfbM knockout mutantfor in situ binding to a large series of gastric tissue sectionsobtained from patients with various gastric histopathologicalabnormalities (Table 2). As this strain is negative for Leb
binding, its binding is not mediated through an active BabA.Significant binding (�50 bacteria per pit) was observed in 8 outof 21 patient sections tested. In 4 out of 21 patients, namely,patients 6, 13, 18, and 20, a statistically significant effect of LPSstructure was found, where the rfbM mutant bound less well; anexample is shown in Fig. 2. The series of sections tested wereobtained from patients with widely varying gastric inflamma-tory status, ranging from no inflammation at all through gas-tritis and atrophy, to dys-, hyper-, and metaplasia and gastricadenocarcinoma. The sections on which the rfbM knockoutstrain showed decreased binding were obtained from one pa-tient with dysplasia (patient 6), one patient with atrophy andintestinal metaplasia (patient 13), a nonsecretor patient with-out gastric inflammation (patient 18), and one patient withgastric adenocarcinoma and metaplasia (patient 20), andhence we are unable to define the circumstances under whichLPS-mediated adhesion, within the context of this adherencemodel, takes place.
Up to now a role of LPS in in situ adhesion of H. pylori hasbeen reported for tissue sections of five patients only, i.e., thefour described above in this study and the one described beforeby Edwards et al. (5). Therefore, NCTC 11637 and its rfbMmutant were tested for adhesion to sections obtained from thesame gastric carcinoma patient from the previous study (5),although we used corpus sections (patient 21), while in theother study (5) antral sections were tested. To our surprise, theparent strain did not bind, which demonstrates that the puta-tive gastric lectin is unevenly distributed over the stomach. Ina very recent study no effect of LPS structure on in situ adher-ence of H. pylori to gastric tissue sections was observed (12).
We conclude that H. pylori LPS has a limited but distinct rolein adhesion. However, the data presented here were obtainedfrom in vitro adhesion experiments, and hence we dot not yetknow how necessary H. pylori Lewis antigens are for in vivocolonization. Several studies have shown that they are crucial
FIG. 1. Binding of H. pylori 43504 (left panels) and its H typeI-expressing LPS phase variant K4.1 (right panels) to superficial mu-cosa (upper panels) or deeper glandular region (lower panels). Bothstrains bind well to superficial mucosa, but variant K4.1 displays en-hanced binding to the deeper region
2878 NOTES INFECT. IMMUN.
Print & Media
FIG. 2. Binding of H. pylori strains to gastric mucosa in situ (histological sections) of diverse pathological conditions. The strains tested wereNCTC 11637 (left column) and its galE and rfbM knockout mutants (middle and right columns, respectively). Upper row, healthy tissue, with nobinding of parent and knockout strains. Second row, metaplastic tissue; the parent strain binds well, while LPS mutants show strongly decreasedbinding. Third row, hyperplastic tissue; LPS structure has almost no effect on binding. Lower row, noninflamed gastric tissue of a nonsecretorpatient, with strongly decreased binding of the rfbM mutant.
VOL. 71, 2003 NOTES 2879
Print & Media
for colonization of mice. However, one recent study demon-strated LPS structure also to be irrelevant to mouse coloniza-tion, and �3-fucosyltransferase knockout mutants not express-ing Lex or Ley colonized mice well (14). In addition, theknockout strain adhered well to human gastric celline cells. Towhat degree the H. pylori Lewis antigens are required forcolonization of humans remains an unanswered question, asdoes the nature of Lex-binding gastric receptors.
We thank N. High (Manchester, United Kingdom) for providingstrain 11637 and its galE and rfbM knockout mutants. We thank G.Faller for providing corpus sections from a patient with gastric adeno-carcinoma.
We thank the Dutch Organization for Scientific Research (NWO)and Swedish Medical Research Council for a Research Visit Grant.
REFERENCES
1. Appelmelk, B. J., S. L. Martin, M. A. Monteiro, C. A. Clayton, A. A. McColm,P. Y. Zheng, T. Verboom, J. J. Maaskant, D. H. van den Eijnden, C. H.Hokke, M. B. Perry, C. M. J. E. Vandenbroucke-Grauls, and J. G. Kusters.1999. Phase variation in Helicobacter pylori lipopolysaccharide due tochanges in the lengths of poly(C) tracts in �3-fucosyltransferase genes. In-fect. Immun. 67:5361–5366.
2. Appelmelk, B. J., M. A. Monteiro, S. L. Martin, A. P. Moran, and C. M.Vandenbroucke-Grauls. 2000. Why Helicobacter pylori has Lewis antigens.Trends Microbiol. 8:565–570.
3. Appelmelk, B. J., B. Shiberu, C. Trinks, N. Tapsi, P. Y. Zheng, T. Verboom,J. Maaskant, C. H. Hokke, W. E. Schiphorst, D. Blanchard, I. M. Simoons-Smit, D. H. van den Eijnden, and C. M. Vandenbroucke-Grauls. 1998. Phasevariation in Helicobacter pylori lipopolysaccharide. Infect. Immun. 66:70–76.
4. Boren, T., P. Falk, K. A. Roth, G. Larson, and S. Normark. 1993. Attachmentof Helicobacter pylori to human gastric epithelium mediated by blood groupantigens. Science 262:1892–1895.
5. Edwards, N. J., M. A. Monteiro, G. Faller, E. J. Walsh, A. P. Moran, I. S.
Roberts, and N. J. High. 2000. Lewis X structures in the O antigen side-chainpromote adhesion of Helicobacter pylori to the gastric epithelium. Mol. Mi-crobiol. 35:1530–1539.
6. Ilver, D., A. Arnqvist, J. Ogren, I. M. Frick, D. Kersulyte, E. T. Incecik, D. E.Berg, A. Covacci, L. Engstrand, and T. Boren. 1998. Helicobacter pyloriadhesin binding fucosylated histo-blood group antigens revealed by retag-ging. Science 279:373–377.
7. Logan, S. M., J. W. Conlan, M. A. Monteiro, W. W. Wakarchuk, and E.Altman. 2000. Functional genomics of Helicobacter pylori: identification of abeta-1,4 galactosyltransferase and generation of mutants with altered lipo-polysaccharide. Mol. Microbiol. 35:1156–1167.
8. Mahdavi, J., B. Sonden, M. Hurtig, F. O. Olfat, L. Forsberg, N. Roche, J.Angstrom, T. Larsson, S. Teneberg, K. A. Karlsson, S. Altraja, T. Wadstrom,D. Kersulyte, D. E. Berg, A. Dubois, C. Petersson, K. E. Magnusson, T.Norberg, F. Lindh, B. B. Lundskog, A. Arnqvist, L. Hammarstrom, and T.Boren. 2002. Helicobacter pylori SabA adhesin in persistent infection andchronic inflammation. Science 297:573–578.
9. Martin, S. L., A. A. McColm, and B. J. Appelmelk. 2000. Helicobacter pyloriadhesion and Lex. Gastroenterology 119:1414–1416.
10. Meining, A., A. Morgner, S. Miehlke, E. Bayerdorffer, and M. Stolte. 2001.Atrophy-metaplasia-dysplasia-carcinoma sequence in the stomach: a realityor merely an hypothesis? Best Pract. Res. Clin. Gastroenterol. 15:983–998.
11. Moran, A. P., E. Sturegard, H. Sjunnesson, T. Wadstrom, and S. O. Hynes.2000. The relationship between O-chain expression and colonization abilityof Helicobacter pylori in a mouse model. FEMS Immunol. Med. Microbiol.29:263–270.
12. Odenbreit, S., G. Faller, and R. Haas. 2002. Role of the alpAB proteins andlipopolysaccharide in adhesion of Helicobacter pylori to human gastric tissue.Int. J. Med. Microbiol. 292:247–256.
13. Simoons-Smit, I. M., B. J. Appelmelk, T. Verboom, R. Negrini, J. L. Penner,G. O. Aspinall, A. P. Moran, S. F. Fei, B. S. Shi, W. Rudnica, A. Savio, andJ. de Graaff. 1996. Typing of Helicobacter pylori with monoclonal antibodiesagainst Lewis antigens in lipopolysaccharide. J. Clin. Microbiol. 34:2196–2200.
14. Takata, T., E. El Omar, M. Camorlinga, S. A. Thompson, Y. Minohara, P. B.Ernst, and M. J. Blaser. 2002. Helicobacter pylori does not require Lewis Xor Lewis Y expression to colonize C3H/HeJ mice. Infect. Immun. 70:3073–3079.
Editor: J. D. Clements
2880 NOTES INFECT. IMMUN.
Print & Media
Cover legend:
Sialyl-Lewis x antigen-mediated adherence of the H. pylori babA-deletion mutant in
situ, to gastric mucosa from a healthy (non-H. pylori infected) Rhesus monkey. The
binding is confined to the deeper gastric glands where sLex antigen is expressed. In
contrast, no binding was detected in surface epithelium.
Print & Media