1364

Increased Inflammatory Responses of Persons of Blood Group O

to Helicobacter pylori
Abdulhamid M. Alkout, C. Caroline Blackwell, Department of Medical Microbiology, University of Edinburgh,
and Donald M. Weir Edinburgh, Scotland

Persons of blood group O are at increased risk of peptic ulcers. Enhanced binding of
Helicobacter pylori to epithelial cells of persons of blood group O has been demonstrated.
Release of interleukin (IL)–6, IL-10, and tumor necrosis factor (TNF)–a by human leukocytes

Downloaded from https://academic.oup.com/jid/article/181/4/1364/856946 by guest on 26 September 2020

from 40 donors (10 from each ABO blood group) was measured after incubation in vitro with
outer membrane protein preparations of H. pylori. Isolates DU (from a patient with a duo-
denal ulcer), GC (from a patient with gastric cancer), NE (from a patient with normal endo-
scopic findings), and NCTC 11637 bound in significantly higher numbers to group O leu-
kocytes. Bacterial binding correlated with release of IL-6 and TNF-a but not of IL-10. Group
O cells released significantly more IL-6 in response to DU, NE, and NCTC 11637, and the
cells released more TNF-a in response to DU and NCTC 11637. Increased density of colo-
nization of epithelial cells and higher inflammatory responses to H. pylori of persons of blood
group O might contribute to increased susceptibility to peptic ulceration.

Helicobacter pylori is closely associated with type B antral
gastritis and peptic ulceration in humans [1–3]. Early epide-
miologic studies carried out before H. pylori was identified
found that nonsecretors of the glycoprotein form of their ABO
blood group antigens and persons of blood group O were over-
represented among patients with peptic ulcers [4]. Although H.
pylori is associated with development of gastric carcinoma, a
significantly higher proportion of persons of group A has been
reported among these patients [5].
The H type 2, Lewisb, and Lewisa antigens can act as recep-
tors for H. pylori on the gastric mucosa, but H type 2 appears
to be a more efficient receptor for the 61-kDa adhesin obtained
by affinity purification with H type 2 antigen (the antigen of
blood group O) than are the Lewis antigens. Biotinylated H
type 2 antigen also bound in greater quantities to outer mem-
branes, preparations of purified adhesin, and whole cells of H.
pylori isolates obtained from a series of patients with peptic
ulcer disease and patients with normal endoscopic findings [6].
The increased susceptibility of group O persons to peptic ulcer
disease [4, 7] might be due partly to higher density of coloni-
zation by H. pylori [8], compared with colonization of persons
of other blood groups. We demonstrated that epithelial cells of
Received 17 August 1999; revised 13 December 1999; electronically pub-
lished 7 April 2000.
Presented in part: Meeting of the Society for General Microbiology, Read-
ing, 1997, and workshop at the Public Health Laboratory Service, London,
United Kingdom, 1998.
Grant support: Chest Heart and Stroke Scotland.
Reprints or correspondence: Dr. C. C. Blackwell, Dept. of Medical
Microbiology, University of Edinburgh, Teviot Place, Edinburgh, Scotland
(caroline.blackwell@ed.ac.uk).
The Journal of Infectious Diseases 2000; 181:1364–9
q 2000 by the Infectious Diseases Society of America. All rights reserved.
0022-1899/2000/18104-0017$02.00
persons of group O bound significantly more H. pylori than
did cells of persons of other blood groups [6].
In most persons infected with H. pylori, there is an increase
in chronic inflammatory cells in the lamina propria of the stom-
ach, including lymphocytes, monocytes, eosinophils, and plas-
ma cells [9]. The supernatants of gastric mucosal biopsy spec-
imens from persons with H. pylori gastritis contained high levels
of tumor necrosis factor (TNF)–a and interleukin (IL)–6 [10].
Because the ABO blood group antigens are also major histo-
compatibility antigens, the aim of this study was to assess the
effect of ABO blood group on inflammatory responses of hu-
man leukocytes exposed to H. pylori antigens obtained from
isolates from patients with different clinical conditions.
Material and Methods
Preparation of bacterial antigens. Outer membrane prepara-
tions were isolated, as described elsewhere [6], from the following
strains of H. pylori: DU, isolated from a patient with duodenal
ulcer; GC, isolated from a patient with gastric cancer; NE, isolated
from a patient with normal endoscopic findings but with ischemic
heart disease; and NCTC 11637 (Public Health Laboratory Service,
London), which had been used in our previous studies on bacterial
binding [6] and inflammatory responses to H. pylori [11].
Collection of human peripheral blood leukocytes. One-day-old
human buffy coat samples (50 mL) were obtained from the Scottish
National Blood Transfusion Service and were diluted 1 : 2 in sterile
pyrogen-free PBS under aseptic conditions. Diluted blood (15 mL)
was layered carefully on Histopaque (5 mL) containing 5.7 g/dL
polysucrose and 9 g/dL sodium diatrizoate (Sigma, St. Louis) in
sterile plastic centrifuge tubes and was centrifuged for 30 min at
300 g. Leukocytes were collected from the interface and were
washed twice with pyrogen-free PBS (Life Technologies Gibco
BRL, Gaithersburg, MD), and the leukocytes were enumerated
microscopically by the trypan blue exclusion method. Cells from
JID 2000;181 (April) Blood Groups and Inflammatory Responses to H. pylori
40 donors were examined, 10 from each ABO blood group. Dif-
ferential counts for the cell preparations were within normal ranges.
Binding of H. pylori to leukocytes. A sample of each buffy
coat was stored in 1% (vol/vol) paraformaldehyde. These samples
were used in the binding assays. All cell samples were tested against
all 4 strains at the same time. Binding of fluorescein isothiocya-
nate–labeled H. pylori to leukocytes was determined by flow cytom-
etry, as described elsewhere, for binding studies with epithelial cells
[6]; 31, 62, and 125 bacteria per leukocyte were used in each assay.
The results for each sample were expressed as binding index, cal-
culated by multiplying the percentage of cells with fluorescence
greater than the control, to which no bacteria were added by the
mean fluorescence of the positive cell population [6].
Stimulation of leukocytes. The leukocytes were resuspended in
Dulbecco’s MEM (DMEM) with 1% (wt/vol) glutamine and 100
IU/mL penicillin, 200 mg/mL streptomycin, and 10% (vol/vol) hu-
man serum (each sample of leukocytes with its respective serum)
and were adjusted to 2 3 10 6/mL. Cells (500 mL/well) were placed
in 24-well tissue culture plates with 500 mL of different concen-
trations of the bacteria, adhesin, or individual outer membrane
protein (OMP) preparations obtained from the 4 H. pylori isolates,
as described by Alkout et al. [6]. The bacteria or antigens were
suspended in DMEM without antibiotics, and control wells con-
tained only medium with cells. The plates were incubated at 377C
in a humidified incubator with 5% CO2. After 18 h of incubation,
the viability of the cells was determined by the trypan blue exclusion
method. Cell-culture fluids were collected in sterile tubes and were
centrifuged at 300 g for 10 min, and the supernatants were used
to estimate the amount of TNF-a, IL-6, or IL-10.
To determine whether leukocytes from persons of blood group
O exhibited responses to other antigens, the experiments were re-
peated with sterile culture filtrates of Escherichia coli O157. The
filtrates were prepared in a category III containment facility and
were filter sterilized. No viable organisms were cultured from the
preparations. Buffy coats from blood donors (n = 32 ) were ob-
tained from the same source. There were 8 sets of A, B, O, and
AB donors; and the levels of IL-6 and TNF-a induced by the filtrate
were assessed.
ELISA for detection of IL-6 and IL-10. IL-6 and IL-10 levels
were measured with a solid-phase ELISA. Flat-bottomed 96-well
microtiter plates were coated overnight at 47C with 100 mL/well
of 0.5 mg/mL mouse monoclonal antibody specific for IL-6 or IL-
10 (R&D Systems, Abingdon, UK). The antibody was diluted in
coating buffer consisting of sodium carbonate (15 mM), sodium
bicarbonate (35 mM), and sodium azide (3 mM; pH 9.6). The plates
were washed 6 times with washing buffer consisting of 0.1% (wt/
vol) bovine serum albumin (BSA) and 0.05% (vol/vol) Tween 20
in PBS. After washing, 100 mL of blocking buffer containing 1%
(wt/vol) BSA in PBS was added to each well, and the plates were
incubated for 30 min. The blocking buffer was removed, and su-
pernatant samples (100 mL) were added to duplicate wells. Dilu-
tions of recombinant human IL-6 or IL-10 standards (R&D Sys-
tems) in blocking buffer were added to duplicate wells and were
incubated for 2 h at 377C with continuous shaking. The plates were
washed 6 times, and 100 mL of polyclonal goat anti–human IL-6
or IL-10 (R&D Systems) was added to detect IL-6 or IL-10 bound
to the wells. After incubation for 2 h at 377C with continuous
shaking, the plates were washed 6 times with washing buffer, and
1365
100 mL of horseradish peroxidase–conjugated donkey anti–goat
IgG (Scottish Antibody Production Unit, Carluke, Scotland) was
added to the wells. The plates were incubated for 1 h at 377C with
continuous shaking and then were washed 6 times with washing
buffer. The substrate (100 mL) was added to the wells; it contained
40 mg of O-phenylenediamine in 100 mL of 0.1 M phosphate citrate
buffer (0.1 M sodium hydrogen phosphate and 0.1 M citric acid;
pH 5.0) activated by 40 mL of 30% (vol/vol) H2O2 immediately
before use. The color change was stopped after 10–20 min by ad-
dition of 100 mL of 12.5% (vol/vol) H2SO4. The absorbance at 490
nm was determined by an ELISA plate reader (Dynatech Labo-
Downloaded from https://academic.oup.com/jid/article/181/4/1364/856946 by guest on 26 September 2020
ratories, Chantilly, VA) and corrected by subtracting the absorb-
ance of the corresponding blank well containing each of the com-
ponents except the cell supernatant. The amount of IL-6 or IL-10
in each sample was determined relative to the recombinant hu-
man IL-6 and IL-10 standards (R&D Systems), and results were
expressed in nanograms per milliliter.
TNF-a bioassay. TNF-a bioactivity was assessed by the L929
cell-culture method, as described by Delahooke et al. [12]. The
amount of TNF-a in each sample was determined relative to curve
for the TNF-a standard (National Institute for Biological Stan-
dards and Controls (Potters Bar, United Kingdom), and results
were expressed in international units per milliliter.
Statistical methods. Statistical analysis of the data was done
with SPSS software (SPSS, Chicago). The significance levels for
differences between groups were examined with the Kruskal-Wallis
test, and correlations between the bacterial binding and inflam-
matory responses were made with Spearman’s correlation coeffi-
cient. P ! .05 was regarded as significant.
Results
Bacterial binding to human leukocytes. Binding of different
isolates of H. pylori to human leukocytes from each ABO blood
group was assessed by flow cytometry. The mean binding in-
dices of the 4 H. pylori isolates to leukocytes from blood group
O donors were significantly higher than the binding to those
of other blood groups, for the combined data for all 4 strains
(P ! .000) and for the individual strains: DU (P = .019);
GC (P = .008); NE (P = .009); and NCTC 11637 (P = .021;
table 1).
At 31, 62, and 125 bacteria per leukocyte, there were sig-
nificantly higher binding indices for strain NE to cells from
donors of group A (P = .003) and group AB (P = .0029). There
Table 1. Mean of binding indices for 4 strains of Helicobacter pylori
to leukocytes from donors of different ABO blood groups.
Binding index
Blood
group All strains DU NE GC NCTC 11637
A (n = 10) 30541 26271 37406 25747 32739
B (n = 10) 27139 19792 32139 19456 37172
O (n = 10) 38980 36036 43319 34195 42370
AB (n = 10) 36252 34034 42806 28973 39193
P .000 .019 .008 .009 .021
NOTE. Strain DU was obtained from a patient with duodenal ulcer, NE
from a patient with normal endoscopic findings, and GC from a patient with
gastric cancer.
1366 Alkout et al. JID 2000;181 (April)

P = .046; and NCTC 11637, P = .021. The difference was not

significant with strain GC (P = .067; table 2).
For the combined data for all 4 strains, the mean TNF-a
levels released from cells of blood group O persons were higher
than levels released by cells of persons of the other blood groups
(P ! .002). For the individual strains, TNF-a levels for donors
of blood group O were significantly higher than those for per-
sons of other blood groups for 2 strains, DU (P = .049) and
NCTC 11637 (P = .041). There was no significant difference in
TNF-a released from cells in response to H. pylori antigens

Downloaded from https://academic.oup.com/jid/article/181/4/1364/856946 by guest on 26 September 2020

from isolates GC (P = .31) or NE (P = .056; table 3).
IL-10 was produced by the cells from each of the donors in
response to OMP of the 4 H. pylori isolates. Although the
highest levels were elicited from cells of group A donors in
response to strain GC, there were no significant differences
associated with ABO blood group or for the individual strains
of H. pylori (table 4).
Figure 1. Mean of tumor necrosis factor (TNF)–a (IU/mL) released
from leukocytes of blood group O stimulated with different concen-
To determine if the increased IL-6 and TNF-a responses for
trations of whole cells, outer membrane protein (OMP), or purified blood group O were antigen specific, these responses to a sterile
adhesin of Helicobacter pylori NCTC 11637 (*protein, mg/mL; **bac- culture filtrate of E. coli O157 were assessed with buffy coats
teria, 3108/mL). from 32 blood donors (4 from each ABO group). There were
no significant differences in levels of either TNF-a or IL-6
associated with ABO group.
was no significant difference in binding of any of the 4 isolates
Correlation of bacterial binding with detection of IL-6, IL-10,
of H. pylori to cells from donors of group O (P = .21; table 1).
and TNF-a. There were significant correlations between bind-
Effect of antigen concentration on release of IL-6 and TNF-a.
ing of H. pylori and detection of IL-6 (r = .210, P ! .01) or
At the end of the incubation period, 190% of the leukocytes
TNF-a (r = .292, P ! .01), and there was a significant corre-
were viable. For NCTC 11637, release of IL-6 and TNF-a was
lation between IL-6 and TNF-a levels (r = .374 , P ! .01). There
assessed for different concentrations of whole bacteria, OMP,
was no significant correlation between bacterial binding and
and the purified adhesin. Among 3 experiments with leukocytes
detection of IL-10. There was a significant inverse correlation
from blood group O donors, the responses were dose-dependent
between IL-10 and IL-6 levels (r = 2.234, P ! .01); a similar
for the bacteria and for the antigen preparations. Release of
TNF-a and IL-6 following stimulation with the purified adhesin
or OMP at protein concentrations of 1.87, 3.75, 7.5, and 15
mg/mL increased with increasing concentrations of OMP or
purified adhesin. Higher levels of TNF-a were obtained with
the purified adhesin, compared with equivalent protein con-
centrations of OMP (P = .003). Similar results were obtained
for IL-6 in response to the adhesin and the OMP (P = .03;
figures 1 and 2).
Effect of blood group on release of IL-6, TNF-a, and IL-10
from leukocytes in response to OMP of different isolates. OMP
preparations were used to compare differences in response of
leukocytes from donors of different ABO blood groups. Results
obtained with cells of 40 different persons (10 donors of each
ABO group) are summarized in tables 2–4.
IL-6 was elicited from leukocytes in response to stimulation
with 10 mg/mL OMP of the different isolates of H. pylori. For
the combined data for all 4 strains, leukocytes from persons of
group O stimulated with OMP produced the highest levels of Figure 2. Mean of interleukin (IL)–6 (ng/mL) released from leu-
kocytes of blood group O stimulated with different concentrations of
IL-6 (P ! .000). For the individual strains DU, NE, and NCTC whole cells, outer membrane protein (OMP), or purified adhesin of
11637, cells from group O produced significantly higher IL-6 Helicobacter pylori NCTC 11637 (*protein, mg/mL; **bacteria, 3108/
levels than did cells of other blood groups: DU, P = .0028; NE, mL).
JID 2000;181 (April) Blood Groups and Inflammatory Responses to H. pylori 1367

Table 2. Mean levels of interleukin (IL)–6 released from leukocytes Density of colonization is associated with inflammatory re-
of different ABO blood groups in response to outer membrane protein sponse [9]. One hypothesis is that the “virulent” strains bind
prepared from 4 different Helicobacter pylori strains.
in greater numbers to host cells. The results of the binding
IL-6, ng/mL
studies do not support this hypothesis.
Blood group All strains DU NE GC NCTC 11637 Persons of blood group O are significantly overrepresented
A (n = 10) 4.35 4.74 4.56 3.78 4.34 among patients with gastric or duodenal ulcers, compared with
B (n = 10) 4.22 4.53 4.46 3.80 4.11 patients of the other blood groups [4, 7, 21–24]. Another study
O (n = 10) 6.15 6.72 6.49 5.54 5.84
AB (n = 10) 2.95 3.44 3.31 2.46 2.57 found a correlation between ABO blood group and H. pylori
P .000 .028 .046 .067 .021 infection in 42 patients aged 50–66 years with rheumatoid ar-
thritis. Blood group O patients had H. pylori in the antral

Downloaded from https://academic.oup.com/jid/article/181/4/1364/856946 by guest on 26 September 2020

NOTE. Strain DU was obtained from a patient with duodenal ulcer, NE
from a patient with normal endoscopic findings, and GC from a patient with mucosa more frequently than did patients of other blood groups
gastric cancer.
[25], and blood group O patients had a significantly higher risk
of hospitalization [26]. Other studies failed to find an associ-
pattern was observed for IL-10 and TNF-a, but the results ation between H. pylori infection and any ABO blood group;
were not significant (r = 2.112, P 1 .05). however, some of these studies relied only on serologic evidence
of infection, and some did not distinguish infection from disease
[27–30].
Discussion This study was designed to examine the effect of ABO blood
group on cytokine release from human leukocytes in response
A previous study of cytokines released from human leuko-
to OMP from isolates from 3 patients with different endoscopic
cytes stimulated by H. pylori detected peak production of IL-1,
findings and strain NCTC 11637. Blood group O leukocytes
IL-6, IL-8, and TNF-a in response to porin proteins between
bound significantly more bacteria of each strain and, overall,
18 and 24 h [13]. In this study, the purified adhesin that binds
released significantly higher levels of IL-6 and TNF-a than did
H type 2 and Lewis antigens, OMP, and whole bacteria was
leukocytes from other blood groups. There were significant cor-
able to stimulate human leukocytes to release TNF-a, IL-10,
relations between bacterial binding and production of TNF-a
and IL-6 during the same time period. Detection of the pro-
and IL-6.
inflammatory cytokines was dose-dependent, with increasing
If the damage to the host tissue is mediated by inflammatory
TNF-a and IL-6 production observed with increasing concen-
responses to the bacteria, 2 options need to be considered: the
tration of whole bacteria, OMP, or adhesin.
levels of proinflammatory responses of the hosts who develop
It has been suggested that development of more serious
peptic ulcer disease are greater than those of patients with nor-
gastroduodenal disease might be related to infection with ul-
mal endoscopy, or the levels of responses of anti-inflammatory
cerogenic or carcinogenic strains of H. pylori [14, 15]. An in-
cytokines, such as IL-10, that regulate the proinflammatory
creased prevalence of antibodies to the cagA protein has been
responses are lower among patients who develop peptic ulcer
reported for patients with peptic ulcer disease or gastric cancer
disease. Our studies indicate that in relation to the apparent
[16–18]. Studies that make use of polymerase chain reaction
increased susceptibility of persons of blood group O to peptic
amplification and DNA hybridization found that cagA in clin-
ulcer disease, the first hypothesis is more likely to be true.
ical isolates was not significantly associated with duodenal ulcer
The mechanisms proposed for increased susceptibility of
disease [19]. A study of an Australian population reported that
group O to peptic ulceration do not appear to be relevant to
the prevalence of antibodies to the cagA protein in the duodenal
the association between group A and gastric cancer. Cells from
ulcer group was higher than that found in the control group;
however, in a Chinese population, no significant difference was Table 3. Mean concentrations of tumor necrosis factor (TNF)–a
found between the prevalence of antibodies to the cagA antigen released from leukocytes of different ABO blood groups in response
in asymptomatic subjects and that in gastric cancer patients to outer membrane protein prepared from different Helicobacter pylori
[20]. Analysis of the antibody responses to other H. pylori anti- strains.
gens, such as the 30-kDa and 45-kDa antigens, identified an TNF-a, IU/mL
association with gastric cancer, and cluster analysis of the West- Blood group All strains DU NE GC NCTC 11637
ern blot profiles showed that sera from gastric cancer patients A (n = 10) 259 258 241 239 299
reacted with almost all the major antigens of H. pylori tested B (n = 10) 231 289 164 189 281
[20]. Because the significance of the cagA antigen in relation O (n = 10) 469 499 510 430 436
AB (n = 10) 222 251 241 220 176
to disease varied with different populations tested, we tested P .002 .049 .056 .318 .41
the hypothesis that the host response might be a more consistent
NOTE. Strain DU was obtained from a patient with duodenal ulcer, NE
risk factor for disease than are some “virulence factors” of the from a patient with normal endoscopic findings, and GC from a patient with
bacteria. gastric cancer.
1368
Table 4. Mean levels of interleukin (IL)–10 released from leukocytes
of different ABO blood groups in response to outer membrane protein
prepared from 4 different Helicobacter pylori strains.
IL-10, ng/mL
Blood group All strains DU NE GC NCTC 11637
A (n = 10) 4.46 4.35 4.89 5.1 3.48
B (n = 10) 3.63 4.98 3.71 3.14 2.7
O (n = 10) 3.84 4.67 4.80 3.44 2.46
AB (n = 10) 3.37 3.54 4.14 3.21 2.59
P .146 .344 .261 .657 .478
NOTE. Strain DU was obtained from a patient with duodenal ulcer, NE
from a patient with normal endoscopic findings, and GC from a patient with
gastric cancer.
group A did not produce higher inflammatory responses to the
strain from the patient with gastric cancer or bind in greater
numbers to the cells from group A donors. One hypothesis for
the role of H. pylori in the etiology of stomach cancer is the
induction of free radicals that could damage DNA, leading to
increased numbers of mutations, increasing the probability of
cellular changes associated with cancer. Our previous work
found an inverse correlation between binding of H. pylori and
nitric oxide stimulation from human monocytes [11]. The most
intriguing observation is that cells from group A donors had
higher responses of IL-10 to H. pylori, particularly the strain
isolated from a patient with gastric cancer (table 4), and we
demonstrated a significant inverse correlation between levels of
IL-10 and IL-6. A similar pattern was observed between IL-
10 and TNF-a, but this was not significant. If the anti-inflam-
matory responses that regulate the proinflammatory responses
involved in destruction of tumor cells are greater in response
to H. pylori, similar studies comparing the inflammatory re-
sponses of persons who have peptic ulcer with those of persons
who have gastric cancer might provide a new approach to ex-
amining the host-parasite interactions involved in the etiology
of gastric cancer [11].
It has been demonstrated that density of colonization is re-
lated to inflammatory responses, epithelial damage, and duo-
denal ulcer among patients infected with H. pylori [8]. Hene-
ghan et al. [9] found that density of H. pylori and lymphocyte
infiltrates were significantly higher in blood group O non-
secretors than in secretors, and blood group O patients had
significantly higher levels of lymphocyte infiltration than did
patients of other blood groups. If the results observed in the
current study reflect the interactions in vivo, increased density
of colonization and significantly higher levels of IL-6 and TNF-
a found for group O donors could partly explain the higher
proportion of persons of group O among patients with peptic
ulcer disease. The enhanced inflammatory responses to each
strain tested indicates that the host response to the bacteria
might play a major role in pathogenesis of peptic ulcer disease
and requires as much detailed examination as do the virulence
factors for the bacteria.
Alkout et al. JID 2000;181 (April)
References
1. Armstrong JA, McGechie DB, Glancy RJ. Campylobacter infection and gas-
troduodenal disease. Med J Aust 1985; 142:436–9.
2. McNulty CAM, Watson DM. Spiral bacteria of the gastric antrum. Lancet
1984; 1:1068–9.
3. Rathbone BJ, Heatley RV. Campylobacter pylori: a new factor in peptic ulcer
disease. Gut 1986; 27:635–41.
4. Clark CA, Wyn Edwards J, Haddock DRW, Howel-Evans AW, McConnell
RB, Sheppard PM. ABO blood groups and secretor character in duodenal
ulcer. Br Med J 1956; 2:725–31.
Downloaded from https://academic.oup.com/jid/article/181/4/1364/856946 by guest on 26 September 2020
5. Aird I, Bentall HH, Roberts JAF. A relationship between cancer of stomach
and the ABO blood groups. Br Med J 1953; 1:799–801.
6. Alkout AM, Blackwell CC, Weir DM, et al. Isolation of a cell surface com-
ponent of Helicobacter pylori that binds H type 2, Lewisa, and Lewisb
antigens. Gastroenterology 1997; 112:1179–87.
7. Mentis A, Blackwell CC, Weir DM, Spiliadis C, Dailianas A, Skandalis N.
ABO blood group, secretor status, and detection of Helicobacter pylori
among patients with gastric or duodenal ulcers. Epidemiol Infect 1991;
106:221–9.
8. Atherton JC, Tham KT, Peek RM, Cover TL, Blaser MJ. Density of Heli-
cobacter pylori infection in vivo as assessed by quantitative culture and
histology. J Infect Dis 1996; 174:552–6.
9. Heneghan MA, Moran AP, Feeley KM, et al. Effect of host Lewis and ABO
blood group antigens expression on Helicobacter pylori colonisation den-
sity and the consequent inflammatory response. FEMS Immunol Med
Microbiol 1998; 20:257–66
10. Crabtree JE, Shallcross TM, Heatley RV, Wyatt JI. Mucosal tumour necrosis
factor alpha and interleukin-6 in patients with Helicobacter pylori asso-
ciated gastritis. Gut 1991; 32:1473–7.
11. Blackwell CC, Weir DM, Alkout AM, et al. Bacterial infection and blood
group phenotype. Acta Nova Leopoldina 1997; 75:67–88.
12. Delahooke DM, Barclay GR, Poxton IR. Tumor necrosis factor induction
by an aqueous phenol-extracted lipopolysaccharide complex from Bac-
teroides species. Infect Immun 1995; 63:840–6.
13. Tufano MA, Rossano F, Catalanotti P, et al. Immunological activities of
Helicobacter pylori porins. Infect Immun 1994; 62:1392–9.
14. Blaser MJ. Helicobacter pylori phenotypes associated with peptic ulceration.
Scand J Gastroenterol 1994; 205:1–5.
15. Xiang Z, Censini S, Bayeli PF, et al. Analysis of expression of CagA and
VacA virulence factors in 43 strains of Helicobacter pylori reveals that
clinical isolates can be divided into two major types and that CagA is not
necessary for expression of the vacuolating cytotoxin. Infect Immun
1995; 63:94–8.
16. Tummuru MKR, Cover TL, Blaser MJ. Cloning and expression of a high
molecular mass major antigen of Helicobacter pylori: evidence of linkage
to cytotoxin production. Infect Immun 1993; 61:1799–809.
17. Crabtree JE, Wyatt JI, Sobala GM, et al. Systemic and mucosal humoral
responses to Helicobacter pylori in gastric cancer. Gut 1993; 34:1339–43.
18. Crabtree JE, Taylor JD, Wyatt JI, et al. Mucosal IgA recognition of Heli-
cobacter pylori 120 kDa protein, peptic ulceration, and gastric pathology.
Lancet 1991; 338:332–5.
19. Owen RJ, Hurtado A, Banatvala N, et al. Conservation of the cytotoxin-
associated (cagA) gene of Helicobacter pylori and investigation of asso-
ciation with vacuolatin-cytotoxin activity and gastroduodenal disease.
FEMS Immunol Med Microbiol 1994; 9:307–15.
20. Mitchell HM, Hazell SL, Li YY, Hu PJ. Serological response to specific
Helicobacter pylori antigens: antibody against CagA antigen is not pre-
dictive of gastric cancer in a developing country. Am J Gastroenterol
1996; 91:1785–8.
21. Langman MJS, Doll R. ABO blood group and secretor status in relation to
clinical characteristics of peptic ulcers. Gut 1965; 6:270–3.
JID 2000;181 (April) Blood Groups and Inflammatory Responses to H. pylori
22. Aird I, Bentall HH, Mehigan JA, Roberts JAF. The blood groups in relation
to peptic ulceration and carcinoma of colon, rectum, breast and bronchus.
Br Med J 1954; 2:315–21.
23. McConnell RB. Peptic ulcer, early genetic evidence families, twins, and mark-
ers. In: Rotter JI, Samloff IM, eds. The genetics and heterogeneity of
common gastrointestinal disorders. New York: Academic Press, 1980:
31–41.
24. Merikas G, Christakopoulos P, Petropoulos E. Distribution of ABO blood
group in patients with ulcer disease: its relationship to gastroduodenal
bleeding. Am J Dig Dis 1966; 11:790–5.
25. Henriksson K, Uribe A, Sandstedt B, Nord C. Helicobacter pylori infection,
ABO blood group, and effect of misoprostol on gastroduodenal mucosa
in NSAID-treated patients with rheumatoid arthritis. Dig Dis Sci 1993;
38:1688–96.
26. Hein HO, Suadicani P, Gyntelberg F. Genetic markers for peptic ulcer. A
1369
study of 3387 men aged 54–74 years: the Copenhagen Male Study. Scand
J Gastroenterol 1997; 32:16–21.
27. Hook-Nikanne J, Sistonen P, Kosunen TU. Effect of ABO blood group and
secretor status on the frequency of Helicobacter pylori antibodies. Scand
J Gastroenterol 1990; 25:815–8.
28. Chesner IM, Nicholson G, Ala F, Elliott TSJ, Langman MJS. Predisposition
to gastric mucosal infection by Helicobacter pylori: an investigation of
any association with ABO or Lewis blood group and secretor status. Eur
J Gastroenterol Hepatol 1992; 4:377–9.
29. Umlauft F, Keeffe EB, Offner F, et al. Helicobacter pylori infection and blood
group antigens: lack of clinical association. Am J Gastroentrol 1996; 91:
2135–8.
Downloaded from https://academic.oup.com/jid/article/181/4/1364/856946 by guest on 26 September 2020
30. Niv Y, Fraser G, Delpre G, et al. Helicobacter pylori infection and blood
groups. Am J Gastroentrol 1996; 91:101–4.
31. Martin de Argila C, Boixeda D, Valdezate S, et al. ABO blood groups, rhesus
factor, and Helicobacter pylori. Rev Esp Enferm Dig 1998; 90:263–8.