THE AMERICAN JOURNAL OF GASTROENTEROLOGY
1, 2000
© 2000 by Am. Coll. of Gastroenterology
Published by Elsevier Science Inc.
H. pylori in the Pathogenesis of
Duodenal Ulcer: Interaction Between
Duodenal Acid Load, Bile, and H. pylori
David Y. Graham, M.D., M.A.C.G., and Michael S. Osato, Ph.D.
Department of Medicine, Veterans Affairs Medical Center and the Division of Molecular Virology,
Baylor College of Medicine, Houston, Texas
OBJECTIVE: Helicobacter pylori (H. pylori) growth is in-
hibited by bile yet it can grow in the duodenal bulb and
cause ulcer disease. The aim of this study was to test the
effect of bile on H. pylori viability and growth and to
determine whether acidification of bile reduces its inhib-
itory activity.
METHODS: Fresh human bile was collected at laparotomy
and tested for inhibitory activity of H. pylori using broth
dilution assays. Six clinical isolates of H. pylori obtained
from patients with duodenal ulcer were used for each ex-
periment. The bile was diluted from 1:3 to 1:192; its inhib-
itory effect on H. pylori was tested before and after acidi-
fication, treatment with cholestyramine, or chloroform. Bile
was acidified to a pH of 2– 6, centrifuged at 8000 rpm for 20
min to remove precipitated bile acids, and the supernatant
pH readjusted. Controls included BHI broth without bile
(positive control) and bile that was acidified to pH 2 and
neutralized without centrifugation.
RESULTS: Human bile inhibited H. pylori growth in a dose
dependent manner. Growth of all strains was supported for
all strains only at a dilution of 1:192. In contrast, after
acidification to pH ⱕ5 and centrifugation to remove precip-
itated bile acids, all strains grew at a bile dilution of 1:12.
Neither chloroform extraction of lipids, nor acidification
without centrifugation removed the inhibitory action of bile.
In contrast, cholestyramine sequestration of bile acids com-
pletely removed all inhibitory activity.
CONCLUSIONS: The duodenal acid load may be the critical
factor to explain the ability of H. pylori to colonize the
duodenal bulb by precipitating glycine-conjugated bile salts.
The combination of a high duodenal acid load and H.
pylori infection is likely the critical event in the patho-
genesis of H. pylori-related duodenal ulcer disease. (Am
J Gastroenterol 2000;95:87–91. © 2000 by Am. Coll. of
Gastroenterology)
INTRODUCTION
One of the originally described characteristics of Helico-
bacter pylori (H. pylori) was that its growth was inhibited
Vol. 95, No.
ISSN 0002-9270/00/$20.00
PII S0002-9270(99)00763-7
by ox bile (1). This led to a paradox, because the duodenum
contains bile and should be an unfavorable environment for
growth of the organism. Nevertheless, H. pylori infection of
the gastric metaplasia in the duodenal bulb occurs and may
result in the development of duodenal ulcer disease (2). To
test the hypothesis that the virulence factor promoting du-
odenal ulcer disease was the ability to survive and grow in
the presence of bile, we compared the ability of H. pylori
isolates obtained from patients with duodenal ulcer disease
and those from patients with asymptomatic gastritis to grow
in the presence of bile acids (3). Synthetic human bile
inhibited H. pylori growth in a dose-dependent manner.
There was no difference in inhibition between H. pylori
gastritis and duodenal ulcer isolates, thus disproving that
hypothesis. Glycine-conjugated bile acids were markedly
more inhibiting of H. pylori growth than were taurine con-
jugates. Glycine conjugates are precipitated from acid so-
lutions (pKa 4.3–5.2) and would not be present in solu-
tion in the acidic stomach (4). Their presence in the
duodenal bulb would depend on the duodenal acid load.
We hypothesized that the ability of H. pylori to grow in
the presence of taurine-conjugated bile acids and the
precipitation of glycine (but not taurine) bile acid conju-
gates by acid may provide the missing link between
inhibition of H. pylori by bile, acid secretion, the ability
of antisecretory therapy to accelerate ulcer healing, and
the ability of H. pylori to colonize the duodenal bulb of
ulcer patients, leading to duodenal ulcer (3, 5).
Because the previous experiments were done with mix-
tures of pure bile acids or ox bile, they may not have
accurately simulated the effects of human bile on H. pylori
growth. We therefore investigated the effect of fresh human
gallbladder bile for detrimental effects on H. pylori growth
and viability.
MATERIALS AND METHODS
Effect of Acidification
Fresh, normal-appearing human bile was collected from the
gall bladders at the time of surgery from patients with
uncomplicated chronic cholecystitis. The H. pylori status of
88 Graham and Osato
the bile donors is unknown. The bile was separated into
equal portions and portions were acidified to pH 2, 3, 4, 5,
or 6 by the addition of 1 N HCl. The volume of acid required
was recorded for each sample. To remove precipitated bile
salt, bile samples were centrifuged at 8,000 rpm for 20 min
(Sorvall RC-5B, SS-34 Rotor, DuPont Instruments, Wil-
mington, DE). The supernatant from each tube was decanted
and the pH adjusted to pH 7 by the addition of 0.1 N NaOH.
In a separate experiment bile obtained from two additional
patients was mixed, and its inhibitory ability was compared
when acidified to pH 2 and then readjusted to pH 7, with or
without centrifugation to remove precipitated bile salts. All bile
samples were paired with an identical sample of bile the vol-
ume of which was adjusted with buffer (0.1 mol/L phosphate
buffer, pH 7.2) to be equivalent to the acidified specimen.
Effect of a Bile Acid-Sequestering Agent
Cholestyramine is a bile acid-sequestering resin that is used
in the treatment of hypercholesterolemia. It is the chloride
salt of a basic anion-exchange resin and binds bile acids. A
quantity of 500 mg of cholestyramine resin (Sigma Chem-
ical, St. Louis, MO) was added to 5 ml of fresh human
gallbladder bile and vortexed. The slurry was centrifuged at
8,000 rpm for 20 min, and the supernatant was tested for the
ability to inhibit growth of H. pylori.
Effect of Lipid Extraction
A sample of bile was treated with chloroform by mixing 5
ml of chloroform (Fisher Scientific, Fair Lawn, NJ) with 5
ml of bile and inverting the tube 20 times. The mixture was
centrifuged for 20 min at 8000 rpm and the uppermost layer
collected.
Broth Dilution Assays
Each experiment of H. pylori inhibitory activity of the bile
samples was tested in broth dilution assays using six clinical
isolates of H. pylori obtained from patients with duodenal
ulcer disease; 18 different strains were tested. Each acidified
and unacidified bile solution was diluted 1:3 in BHI test
broth (Brain-Heart Infusion broth, Difco, Detroit, MI) con-
taining 0.25% yeast extract (Difco), and 10% equine serum
(HyClone, Logan, UT) and then serially diluted 1:2. The
highest dilution was 1:192. The positive control contained
BHI test broth without the addition of bile.
The inoculum was prepared by resuspending fresh growth
(48 h) in sterile saline equivalent to a McFarland 2 barium
standard (approximately 6 ⫻ 108 bacteria/ml). Twenty ␮l of
the adjusted inoculum was added to 1 ml of test solution in
wells of cluster plates (Corning Glass Works, Corning, NY)
and the cluster plates were incubated at 37°C under 12%
CO2 for 72 h. At the end of incubation, 10-␮l aliquots were
removed from each well of the cluster plate and used to
inoculate Mueller-Hinton agar plates containing 5% ovine
blood (BBL, Cockeysville, MD) to determine whether
growth was inhibited. The results were scored as growth or
no growth; data are presented as the median proportion of
strains that showed growth at each condition.
AJG – Vol. 95, No. 1, 2000
Figure 1. The growth of six H. pylori isolates from duodenal ulcer
patients broth microdilution assays is shown. Human bile was
acidified to pH 2, 3, 4, 5, or 6, centrifuged to remove precipitated
bile acids and neutralized to pH 7. H. pylori growth was assessed
in brain– heart infusion broth along with dilutions of bile (1:3 to
1:192). The 50% inhibitory dilution of previously acidified bile is
shown by the dotted line. Bile acidified to pH 6 and control (bile
with no acidification) had identical growth characteristics. There
was no inhibition of growth in controls or pH 6 acidified bile at a
dilution of 1:192 (data not shown).
RESULTS
H. pylori was remarkably inhibited by human gallbladder
bile. As part of each experiment, we tested H. pylori growth
with untreated bile. These five different experiments in-
volved 18 separate H. pylori strains and H. pylori growth
was supported by all strains only at a dilution of 1:192. To
approximate the conditions in the duodenal bulb, bile was
acidified to precipitate the bile acids. Normally, when the
bile precipitates pass distally into an area where the pH is
higher, they return to solution and are available for lipid
solubilization. Acidification of bile to pH ⱕ5, followed by
centrifugation to remove precipitated bile salts, reduced the
ability of bile to inhibit H. pylori growth (Fig. 1). Acidifi-
cation followed by neutralization without centrifugation re-
turned the inhibitory effect to the bile (e.g., it did not remove
any inhibitory factors). Bile retained an inhibitory influence
at a dilution of 1:3 despite acidification and centrifugation,
consistent with the fact that taurine-conjugated bile acids are
soluble at this pH (Fig. 1).
Both the control and the bile samples that had been
acidified to pH 6 yielded identical growth curves with the 50%
bactericidal dilution being 1:24 (Fig. 1). The 50% bactericidal
dilution for the acidified bile solutions was 1:4.5; all samples
grew at dilutions of 1:12 or greater (Fig. 1).
To test the effect of acidification on removing the H.
pylori inhibitory factor(s), we extracted bile with cholestra-
mine and with chloroform. Cholestyramine, which removes
bile acids, completely removed the ability of human gall
bladder bile to inhibit H. pylori growth (no inhibition at any
dilution) (Fig. 2). In contrast, treatment with chloroform had
no effect on the bile-associated inhibition of H. pylori
AJG – January, 2000
Figure 2. Cholestyramine treatment to remove bile acids com-
pletely removed the ability of human gall bladder bile to inhibit H.
pylori growth.
growth (growth supported for all strains only at 1:192 di-
lution).
DISCUSSION
Bile acid concentration in the duodenum ranges from 2 to 3
mmol/L fasting, to 6 to 10 mmol/L after a meal, with the
ratio of glycine to taurine conjugates being 2 or 3 to 1 (4, 6,
7). The pKa of glycine conjugates ranges between 4.3 and
5.2, and taurine conjugates range between 1.8 and 1.9 (8).
Thus, whereas both glycine and taurine conjugates are sol-
uble in neutral solutions, only taurine conjugates are soluble
in acid. This study confirmed previous observations show-
ing that growth of H. pylori is inhibited by bile in vitro (1,
3, 9 –13). The inhibitory factor(s) in bile was markedly
reduced by acidification followed by centrifugation to re-
move precipitated glycine-conjugated bile acids and was
completely eliminated by use of the bile acid-sequestering
agent cholestyramine. These results are consistent with the
notion that acidic conditions in the duodenal bulb would
serve to precipitate inhibitory bile acids (or other inhibitory
substances) and allow H. pylori to grow in an otherwise
hostile environment. The fact that, even with acidification,
H. pylori growth was inhibited by bile diluted 1:3 is con-
sistent with the fact that taurine conjugated bile acids are not
precipitated by acid conditions (pKa 1.8 and 1.9) (3). To test
whether other inhibitory factors such as fatty acids (14 –18)
were present, we extracted lipids from bile with chloroform,
and it had no effect on the inhibitory effects of bile.
The initial data on the ability of H. pylori to grow in the
presence of bile in the stomach appeared conflicting (19 –
25). For example, O’Connor et al. (26) reported an inverse
relationship between the presence of bile in the stomach and
the presence of H. pylori, whereas Karttunen and Niemela
did not (27). The difference between results can be ex-
plained by the fact that H. pylori were found when the
stomach was acidic (despite the presence of bile), in the
Pathogenesis of Duodenal Ulcer 89
study by Karttunen and Niemela (27), and are not found
with achlorhydria, as in the study by O’Connor et al. (26).
The results of these different studies, therefore, are actually
consistent when the pH of the gastric contents is taken into
account.
The factor that most characterizes duodenal ulcer disease
is the ability to generate ⬎12 mmol/h of hydrochloric acid
(28). Since the 1940s it has been recognized that the pH in
the duodenal bulb was lower in those with ulcer disease than
in those without, and that antacids or antisecretory therapy,
which reduced the duodenal acid load, accelerated ulcer
healing (29 –36). We hypothesize that, because glycine-
conjugated bile salts are precipitated at acidic pH, any event
that would lead to an increase in the duodenal acid load
would predispose to duodenal ulcer diseases in patients with
H. pylori infection (3, 5). The converse is that any condition
that reduced duodenal acid load (e.g., antisecretory therapy)
would allow bile acids to remain in solution, would inhibit
growth of H. pylori, and would promote ulcer healing. The
increase in duodenal acid load could be genetically deter-
mined or acquired (e.g., related to smoking or stress) (5, 37,
38). The duodenal acid load has at least two components:
acid secretion by the stomach, and neutralization of acid in
the duodenum (39 – 42). There are a number of factors that
are reversed after cure of H. pylori infection and that pro-
mote an increase in duodenal acid load, such as an H.
pylori-associated inhibition of the effect of antral distention
on acid secretion and impaired duodenal bicarbonate secre-
tion (43– 45). Duodenal ulcer disease is also associated with
a reduction in size and motility of the duodenal bulb, as well
as replacement of normal villous architecture by gastric
metaplasia; and areas of gastric metaplasia in the duodenal
bulb have also been shown to be capable of local acid
secretion (5, 46).
Smoking increases the duodenal acid load by increasing
acid secretion and by inhibiting duodenal bicarbonate se-
cretion (47, 48). After cure of H. pylori infection, smoking
is no longer a risk factor for ulcer recurrence (49 –52). The
role of smoking in duodenal ulcer disease can possibly be
explained via its effect on duodenal acid load and via the
precipitation of injurious components of bile, making the
duodenum a more hospitable site for H. pylori. The local-
ization of duodenal ulcer primarily to the duodenal bulb may
be a consequence of the fact that low pH is largely restricted
to this area, making the second portion of the duodenum
inhospitable for H. pylori growth because the concentration
of bile exceeds that at which H. pylori can grow well.
ACKNOWLEDGMENTS
This work was supported by the Department of Veterans
Affairs and the generous support of Hilda Schwartz.
The authors thank Drs. Tom V. Taylor and Kamal Itani
for providing the fresh human bile specimens.
90 Graham and Osato
Reprint requests and correspondence: David Y. Graham, M.D.,
Veterans Affairs Medical Center (111D), 2002 Holcombe Boule-
vard, Houston, TX 77030.
Received May 27, 1999; accepted Sep. 23, 1999.
REFERENCES
1. Marshall BJ, Royce H, Annear DI, et al. Original isolation of
Campylobacter pyloridis from human gastric mucosa. Micro-
biol Lett 1984;25:83– 8.
2. Graham DY. Campylobacter pylori and peptic ulcer disease.
Gastroenterology 1989;96:615–25.
3. Han SW, Evans DG, el-Zaatari FA, et al. The interaction of
pH, bile, and Helicobacter pylori may explain duodenal ulcer.
Am J Gastroenterol 1996;91:1135–7.
4. Hofmann AF, Mysels KJ. Bile acid solubility and precipitation
in vitro and in vivo: The role of conjugation, pH, and Ca2⫹
ions. J Lipid Res 1992;33:617–26.
5. Graham DY, Genta RM, Go MF, et al. Which is the most
important factor in duodenal ulcer pathogenesis: the strain of
Helicobacter pylori or the host? In: Hunt RH, Tytgat GNJ, eds.
Helicobacter pylori: Basic mechanisms to clinical cure.
Lancaster: Kluwer Academic Publishers, 1996: 85–91.
6. Dowling RH. The enterohepatic circulation. Gastroenterology
1972;62:122– 40.
7. Northfield TC, McColl I. Postprandial concentrations of free
and conjugated bile acids down the length of the normal
human small intestine. Gut 1973;14:513– 8.
8. Morris A, Lane M, Hamilton I, et al. Duodenal ulcer relapse
after eradication of Helicobacter pylori. New Zealand Med J
1991;104:329 –31.
9. Tompkins DS, West AP. Campylobacter pylori, acid, and bile.
J Clin Pathol 1987;40:1387 (letter).
10. Angelico M, Luzzi I, Gandin C, et al. Individual bile salts
differently modulate Helicobacter pylori growth: Implications
for human pathophysiology. Acta Gastroenterol Belg 1993;
56(suppl):107 (abstract).
11. Mathai E, Arora A, Cafferkey M, et al. The effect of bile acids
on the growth and adherence of Helicobacter pylori. Aliment
Pharmacol Ther 1991;5:653– 8.
12. Hanninen ML. Sensitivity of Helicobacter pylori to different
bile salts. Eur J Clin Microbiol Infect Dis 1991;10:515– 8.
13. Goodwin CS, Armstrong JA, Chilvers T, et al. Transfer of
Campylobacter pylori and Campylobacter mustelae to Heli-
cobacter gen. nov. as Helicobacter pylori comb. nov. and
Helicobacter mustelae comb. nov., respectively. Int J Syst
Bacteriol 1989;39:397– 405.
14. Hazell SL, Graham DY. Unsaturated fatty acids and viability
of Helicobacter (Campylobacter) pylori. J Clin Microbiol
1990;28:1060 –1.
15. Hazell SL, Markesich DC, Evans DJ, et al. Influence of media
supplements on growth and survival of Campylobacter pylori.
Eur J Clin Microbiol Infect Dis 1989;8:597– 602.
16. Thompson L, Cockayne A, Spiller RC. Inhibitory effect of
polyunsaturated fatty acids on the growth of Helicobacter
pylori: A possible explanation of the effect of diet on peptic
ulceration. Gut 1994;35:1557– 61.
17. Khulusi S, Ahmed HA, Patel P, et al. The effects of unsatur-
ated fatty acids on Helicobacter pylori in vitro. J Med Micro-
biol 1995;42:276 – 82.
18. Petschow BW, Batema RP, Ford LL. Susceptibility of Heli-
cobacter pylori to bactericidal properties of medium-chain
monoglycerides and free fatty acids. Antimicrob Agents Che-
mother 1996;40:302– 6.
19. Offerhaus GJ, Rieu PN, Jansen JB, et al. Prospective compar-
ative study of the influence of postoperative bile reflux on
AJG – Vol. 95, No. 1, 2000
gastric mucosal histology and Campylobacter pylori infection.
Gut 1989;30:1552–7.
20. O’Connor HJ, Newbold KM, Alexander-Williams J, et al.
Effect of Roux-en-Y biliary diversion on Campylobacter py-
lori. Gastroenterology 1989;97:958 – 64.
21. Sobala GM, O’Connor HJ, Dewar EP, et al. Bile reflux and
intestinal metaplasia in gastric mucosa. J Clin Pathol 1993;
46:235– 40.
22. O’Connor HJ, Wyatt JI, Dixon MF, et al. Campylobacter like
organisms and reflux gastritis. J Clin Pathol 1986;39:531– 4.
23. Niemela S, Karttunen T, Heikkila J, et al. Relationship of
Campylobacter pylori and duodenogastric reflux. Dig Dis Sci
1989;34:1021– 4.
24. Kellosalo J, Alavaikko M, Laitinen S. Effect of biliary tract
procedures on duodenogastric reflux and the gastric mucosa.
Scand J Gastroenterol 1991;26:1272– 8.
25. Robles-Campos R, Lujan-Mompean JA, Parrilla-Paricio P, et
al. Role of Helicobacter pylori infection and duodenogastric
reflux in the pathogenesis of alkaline reflux gastritis after
gastric operations. Surg Gynecol Obstet 1993;176:594 – 8.
26. O’Connor HJ, Dixon MF, Wyatt JI, et al. Effect of duodenal
ulcer surgery and enterogastric reflux on Campylobacter py-
loridis. Lancet 1986;2:1178 – 81.
27. Karttunen T, Niemela S. Campylobacter pylori and duodeno-
gastric reflux in peptic ulcer disease and gastritis. Lancet
1988;1:118 (letter).
28. Soll AH. Gastric, duodenal, and stress ulcer. In: Sleisenger M,
Fordtran J, eds. Gastrointestinal disease, 5th ed. Philadelphia:
WB Saunders, 1993:580 – 679.
29. Berk JE, Rehfuss ME, Thomas JE. Effect of ulcer on acidity
and neutralization ability in duodenal ulcer. Arch Intern Med
1942;70:959 –74.
30. Berk JE, Rehfuss ME, Thomas JE. Duodenal bulb acidity
under fasting conditions in patients with duodenal ulcer. Arch
Surg 1942;45:406 –15.
31. Lopusniak MS, Berk JE. The comparative effects of casein
hydrolysate, milk and milk-cream on gastric and duodenal
bulb acidity in duodenal ulcer patients. Gastroenterology
1948;11:891–903.
32. Berk JE, Thomas JE. The acid factor in duodenal ulcer as
evaluated by the acidity and neutralizing ability of the duode-
nal bulb. Am J Dig Dis 1947;9:371–5.
33. Berk JE, Rehfuss ME, Thomas JE. Duodenal bulb (“ulcer-
bearing area”) acidity in fasting normal people. J Lab Clin
Med 1942;27:1501–10.
34. Berk JE, Rehfuss ME, Thomas JE. The acidity of the “ulcer-
bearing area” of the duodenum in normal persons. Am J Dig
Dis 1942;9:276 – 81.
35. Wormsley KG. The pathophysiology of duodenal ulceration.
Gut 1974;15:59 – 81.
36. Archambault AP, Rovelstad RA, Carlson HC. In situ pH of
duodenal bulb contents in normal and duodenal ulcer subjects.
Gastroenterology 1967;52:940 –7.
37. Walker P, Luther J, Samloff IM, et al. Life events stress and
psychosocial factors in men with peptic ulcer disease. II.
Relationships with serum pepsinogen concentrations and be-
havioral risk factors. Gastroenterology 1988;94:323–30.
38. Feldman M, Cryer B, McArthur KE, et al. Effects of aging and
gastritis on gastric acid and pepsin secretion in humans: A
prospective study. Gastroenterology 1996;110:1043–52.
39. Hojgaard L, Mertz Nielsen A, Rune SJ. Peptic ulcer
pathophysiology: acid, bicarbonate, and mucosal function.
Scand J Gastroenterol 1996;216(suppl):10 –5.
40. Hogan DL, Ainsworth MA, Isenberg JI. Gastroduodenal bi-
carbonate secretion. Aliment Pharmacol Ther 1994;8:475– 88.
41. Allen A, Flemstrom G, Garner A, et al. Gastroduodenal mu-
cosal protection. Physiol Rev 1993;73:823–57.
AJG – January, 2000
42. Malagelada JR. Duodenal acid load. In: Mignon M, Galmiche
J-P, eds. Control of acid secretion. Paris: John Libbey
Eurotext; 1988:65–72.
43. Hamlet A, Olbe L. The influence of Helicobacter pylori in-
fection on postprandial duodenal acid load and duodenal bulb
pH in humans. Gastroenterology 1996;111:391– 400.
44. Graham DY. Helicobacter pylori and perturbations in acid
secretion: The end of the beginning. Gastroenterology 1996;
110:1647–50.
45. Calam J. Pathogenic mechanisms. Baillieres Clin Gastroen-
terol 1995;9:487–506.
46. Borgstrom S, Arborelius M Jr. Duodenal motility pattern in
duodenal ulcer disease. Scand J Gastroenterol 1978;13:349 –52.
47. Bynum TE, Solomon TE, Johnson LR, et al. Inhibition of
pancreatic secretion in man by cigarette smoking. Gut 1972;
13:361–5.
Pathogenesis of Duodenal Ulcer 91
48. Ainsworth MA, Hogan DL, Koss MA, et al. Cigarette smoking
inhibits acid-stimulated duodenal mucosal bicarbonate secre-
tion. Ann Intern Med 1993;119:882– 6.
49. Bardhan KD, Graham DY, Hunt RH, et al. Effects of smoking
on cure of Helicobacter pylori infection and duodenal ulcer
recurrence in patients treated with clarithromycin and ome-
prazole. Helicobacter 1997;2:27–31.
50. Chan FK, Sung JJ, Lee YT, et al. Does smoking predispose to
peptic ulcer relapse after eradication of Helicobacter pylori?
Am J Gastroenterol 1997;92:442–5.
51. Graham DY, Lew GM, Malaty HM, et al. Factors influencing
the eradication of Helicobacter pylori with triple therapy.
Gastroenterology 1992;102:493– 6.
52. Graham DY, Dore MP. Pertubations in gastric physiology in
Helicobacter pylori duodenal ulcer disease: Are they all epi-
phenomena? Helicobacter 1997;2:S44 –9.