GASTROENTEROLOGY 1997;113:1570 1575
Water Extract of Helicobacter pylori Inhibits Duodenal
Mucosal Alkaline Secretion in Anesthetized Rats
LARS FANDRIKS,* CHARLOTTE VON BOTHMER, BERNDT JOHANSSON, MATHIAS HOLM,
INGRID BOLIN, and ANDERS PETTERSSON
Departments of *Surgery, Physiology and Pharmacology, and Medical Microbiology and Immunology, Goteborg University, Goteborg,
Sweden
Background & Aims: The pathophysiology behind Heli-
cobacter pyloriinduced gastroduodenal dysfunction is
incompletely understood. The aim of this study was to
investigate if a water extract of H. pylori distorts acid-
induced duodenal mucosal alkaline secretion. Meth-
ods: Chloralose-anesthetized rats were prepared for du-
odenal luminal perfusion and in situ pH-stat titration of
mucosal alkaline secretion. Results: Mucosal bicarbon-
ate secretion increased approximately 55%60% after
a 5-minute exposure to 10 mmol/L HCl. This response
was absent when water extracts of three strains of H.
pylori (protein content, 0.220 mg/mL) had been
added to the perfusate. Presence of 3 mmol/L L-argi-
nine, but not the stereoisomer D-arginine, in the luminal
perfusate reversed the H. pylori extract blockade of
acid-induced mucosal alkaline secretion. High-perfor-
mance liquid chromatographybased analyses showed
that the endogenous nitric oxide synthase inhibitor
asymmetric dimethyl arginine (ADMA) increased four-
fold in duodenal perfusate and fivefold in duodenal tis-
sue after H. pylori extract exposure. In vitro proteolysis
of H. pylori extract also resulted in a substantial accu-
mulation of ADMA. Exogenously administered ADMA,
giving similar tissue concentrations, inhibited the mu-
cosal alkaline response to acid exposure. Conclusions:
Water extracts of H. pylori inhibit acid-induced mucosal
alkaline secretion via interference with mucosal NO
synthase.
H elicobacter pylori is now recognized as a key factor in
the pathogenesis of chronic gastroduodenitis and
peptic ulcer disease. A number of H. pylorirelated
pathogenic mechanisms causing mucosal injury have
been suggested, e.g., the excretion of cytotoxic substances
and indirect effects such as neutrophilic production of
oxygen radicals with effects on mucosal viability. In addi-
tion to direct damage of the gastroduodenal mucosa, H.
pylori infection is associated with functional disturbances,
e.g., defects in the feedback regulation of gastrin release
and gastric acid secretion.1 3 Another function reported
to be disturbed by the presence of H. pylori in patients
/ 5e22$$0005 10-07-97 01:42:55
with peptic ulcer is the ability of the duodenal mucosa
to increase its alkaline secretion in response to acid expo-
sure.4 The physiological mediation of acid-induced duo-
denal mucosal alkaline secretion involves several regula-
tory principles, e.g., local synthesis of prostaglandins and
neurohumoral control.5 Furthermore, recent studies
showed that the acid-induced response involves also the
6,7
L-arginine/nitric oxide pathway. The question was
raised if luminal factors, released from the H. pylori infec-
tion in patients with duodenal ulcer, disturb the mucosal
responsiveness to luminal acid, via interference with the
L-arginine/NO axis. Theoretically, this could be related
to the microbe itself by excreted toxins disturbing NO
formation or to indirect effects related to the inflamma-
tory reaction. We decided to test the former possibility
in an in vivo rat model by briefly exposing the duodenal
mucosa to a water extract of H. pylori. Presence of H.
pylori extracts (Hpx) markedly inhibited the mucosal re-
sponsiveness to acid because of local interference with
NO synthase (NOS). Some of these data have been pre-
sented in abstract form previously.8
Materials and Methods
General
The experiments were approved by the Animal Ethics
Committee of Goteborg University. Male SpragueDawley
rats (Mollegard, Denmark) were anesthetized with methohexi-
tal injected intraperitoneally (75 mg/kg body wt) and a-chlor-
alose injected intravenously (bolus, 50 mg/kg plus 25
mgrkg01rh01). The body temperature was maintained at
387C with a heating pad and lamp. A catheter was inserted
into the trachea to ensure free airways. A femoral artery and
one vein were catheterized with polyethylene 50 tubes for
subsequent blood pressure measurements and drug infusions,
respectively. Blood pressure was recorded using a Statham
P23Dc transducer (Statham, Hato Rey, Puerto Rico). To avoid
Abbreviations used in this paper: ADMA, asymmetric dimethyl argi-
nine; Hpx, Helicobacter pylori extract; NOS, nitric oxide synthase.
q 1997 by the American Gastroenterological Association
0016-5085/97/$3.00
gasas WBS-Gastro
November 1997
acidosis and dehydration caused by surgical trauma and the
long period of general anesthesia, a 2% glucose solution con-
taining 0.04 mol/L NaHCO3 was given intravenously through-
out the experiments at a rate of 1 mL/h.
Secretion
Duodenal mucosal alkaline (HCO30) secretion was
measured as described previously.9 Briefly, after a midline lapa-
rotomy, a 1.6-cm segment of the duodenum with intact vascu-
lar supply was isolated between two glass tubes connected to
a water-jacketed (387C) reservoir containing 150 mmol/L
NaCl. The saline was circulated through the segment using a
gas-lift (pure air). The common bile duct was catheterized to
avoid contamination of the segment by bile and pancreatic
juice. Alkaline secretion to the perfusate was continuously
titrated with HCl by pH-stat equipment.
Bacterial Water Extracts
H. pylori strains E50 (provided by Dr. J.-P. Butzler,
WHO Collaborating Centre for Enteric Campylobacter, Brus-
sels, Belgium); C21, a vacuolating cytotoxin negative strain
(provided by Dr. N. Figura, Siena, Italy); and Hel73, a vac-
uolating cytotoxin-positive strain (our clinical isolate from a
patient with duodenal ulcer), were used. The strains were kept
at 0707C until use and were then grown on Columbia agar
plates with 8% horse blood in microaerophilic atmosphere of
10% CO2 and 5% O2 for 4872 hours. Bacteria from agar
plates were inoculated into Brucella broth supplemented with
1% dimethyl-b-cyclodextrin (a gift from Teijin Ltd., Tokyo,
Japan), vancomycin, trimethoprim, and polymyxin. The liquid
bacterial cultures were grown for 72 hours under microaero-
philic conditions without shaking. Escherichia coli K12 strain
C 600 was grown in LuriaBertani broth at 377C overnight.
H. pylori from 1 L of liquid culture or 25 mL of E. coli culture
was pelleted by centrifugation at 6000g and resuspended in
5 mL of distilled water. The bacterial suspensions were kept
at room temperature for 20 minutes and shaken occasionally.
Proteins extracted by this procedure were separated from bacte-
ria by centrifugation at 20,000g. The protein contents were
estimated by subtracting the absorbance value at 310 nm from
the value at 280 nm, and protein concentration was expressed
in milligrams per milliliter. The water extracts were stored at
0707C until use.
Analyses of Arginine Analogues
Occurrence of monomethyl arginine, asymmetric di-
methyl arginine (ADMA), and symmetric dimethyl arginine
in luminal perfusates and in the tissue of the duodenal segment
under study was analyzed using a modified high-performance
liquid chromatographic method with fluorescence detection.10
The luminal perfusates were frozen at 0707C until analysis.
The duodenal segment, arranged for secretion study as de-
scribed above, was dissected and immediately homogenized in
4 mL of 150 mmol/L NaCl (47C) using a Polytron PT homoge-
nizer (Kinematic AG, Lucerne, Switzerland). The homogenate
/ 5e22$$0005 10-07-97 01:42:55 gasas
DUODENAL SECRETION, H. PYLORI, AND NO 1571
was centrifugated at 23,000g for 20 minutes (47C), and the
supernatant was stored at 0707C until analysis.
Drugs
For anesthesia, methohexital (Brietal; Lilly Inc., India-
napolis, IN) and a-chloralose (Sigma Chemical Co., St. Louis,
MO) were used. L- and D-arginine hydrochloride (Sigma)
and proteinase K (Boerhinger Mannheim Scandinavia AB,
Bromma, Sweden) were freshly dissolved as stock solutions.
Experimental Protocol
The animals were left undisturbed for 1 hour after
surgery. Acid exposure was performed after a 4560-minute
control period, and the circulating saline in the titration-cham-
ber was then changed to body tempered 10 mmol/L HCl (made
isotonic with the addition of NaCl) during 5 minutes. In a
separate series, without acid exposure, the acute effects of lumi-
nal administration of, e.g., bacterial extracts and L-arginine,
were studied. The animals were divided into different groups
(each 6 animals) depending on treatment. The experiments
were terminated by induction of cardiac arrest with intravenous
KCl.
Statistics
Analysis of variance (ANOVA) for repeated measure-
ments and Bonferronis post hoc test were used to evaluate
significance of changes within groups. Secretory data obtained
during the control period immediately before any interference
(e.g., acid exposure) were used as basal conditions. Responses
to stimulations were analyzed by using the induced net change
in secretion (prestimulatory value minus the value most differ-
ent from the prestimulatory value within 20 minutes after
the stimulation period). Net changes were compared between
groups by using one-way ANOVA and, when appropriate, a
t test. A P value of 0.05 was considered significant.
Results
Effects of Bacterial Extracts on Basal
Secretion
Basal mucosal alkaline secretion by the duodenal
segment in untreated controls ranged between 10 and
17 mmolrcm01rh01. Similar steady-state secretory levels
were observed in the groups with bacterial extracts added
to the perfusate resulting in a protein concentration of
20 mg/mL (Figure 1). The acute effect of adding E50
Hpx was studied in greater detail. A transient increase
(P 0.05) in secretion with a magnitude and duration
of approximately 70% and 20 minutes, respectively, was
recorded immediately after addition of E50-Hpx to the
perfusate (not shown in figure). This transient increase
in secretory values could not be attributed to either the
titration technique, because the acid buffering capacity
was similar in the control solutions, or to any significant
WBS-Gastro
1572 FANDRIKS ET AL. GASTROENTEROLOGY Vol. 113, No. 5

Figure 1. Duodenal mucosal alkaline secretion in the presence of

various bacterial extracts with a protein concentration of 20 mg/mL.
Figure 2. Effect of a 5-minute exposure to 10 mmol/L HCl on duode-
Data are means { SEM; n 6 for each group.
nal mucosal alkaline secretion in the presence or absence of bacterial
extracts (20 mg/mL protein content). Data are means { SEM; n 6
for each group. *** P 0.001, significan difference from control.
n.s., nonsignifican difference.
difference in the physicochemical properties such as tem-
perature and osmolality (data not shown).
4). A group of animals given the stereoisomer D-arginine
Effects on Acid-Induced Changes (3 mmol/L) in the Hpx-containing luminal perfusate (2
The mucosal alkaline secretion increased approxi- mg/mL; n 6) had similar steady-state basal secretion
mately 60% after 5-minute exposure to 10 mmol/L HCl. as the controls and L-arginine/Hpxtreated animals.
Interestingly, this response was absent when Hpx had
been added to the perfusate (Figure 2). Hpx was added
45 minutes before the acid exposure at a concentration
of 20 mg/mL protein content; this concentration was
maintained also during and after the acidification. The
three different H. pylori strains (E50, C21, and Hel73)
showed identical blockade of acid-induced duodenal mu-
cosal alkaline secretion. A similarly prepared extract of
a nonenterotoxin-producing E. coli strain did not block
the response, although the acid-induced response varied
considerably between animals (Figure 2). Dose-response
relations for strain E50-based Hpx showed that potent
inhibition of the acid-induced mucosal alkaline secretion
occurred at protein concentrations exceeding 0.2 mg/mL
(Figure 3).

Effects of Arginine
Short-term addition L-arginine (3 mmol/L) to the
duodenal perfusate transiently lowered the alkaline secre-
tion by approximately 50%. The secretion returned to
baseline values within 20 minutes (n 6; data not
shown). Presence of L-arginine did not influence the re- Figure 3. Dose-response relationship between the presence of strain
sponse to 10 mmol/L HCl (n 6; data not shown). E50 Hpx at various concentrations and the increase in duodenal muco-
When 3 mmol/L L-arginine had been added to the Hpx- sal alkaline secretion elicited by a 5-minute exposure to 10 mmol/L
containing (2 mg/mL) luminal perfusate, acid exposure HCl. The effect of acid exposure was tested in 6 animals for each
Hpx concentration. A plotted value represents mean { SEM from each
increased the duodenal mucosal alkaline secretion to a group. *P 0.05, *** P 0.001; significan difference from control
level similar to one observed in untreated controls (Figure (equals zero concentration).

/ 5e22$$0005 10-07-97 01:42:55 gasas WBS-Gastro

November 1997 DUODENAL SECRETION, H. PYLORI, AND NO 1573

However, acid-induced stimulatory responses were ab-

sent when D-arginine (3 mmol/L) had been added to-
gether with Hpx (Figure 4).

Direct Measurements of NOS Inhibitors

In vitro circulated Hpx perfusate contained no
detectable amounts of any of the analyzed methylated
arginines. However, significant levels of ADMA but nei-
ther monomethyl arginine nor symmetric dimethyl argi-
nine were recorded after intraduodenal circulation. Con-
trol perfusate (without Hpx) was shown to contain
approximately 0.1 mmol/L of ADMA after 45 minutes of
intraluminal circulation, whereas ADMA levels in Hpx-
containing perfusates were fourfold higher (Table 1).
Furthermore, tissue concentrations of ADMA in Hpx-
exposed duodeni were fivefold higher than in control Figure 4. Effect of a 5-minute exposure to 10 mmol/L HCl in luminal
segments treated similarly, but without Hpx exposure presence of E50 Hpx (2 mg/mL protein content) plus L-arginine (3
(Table 1). In vitro proteolytic degradation of Hpx (pro- mmol/L, ) or Hpx plus D-arginine (3 mmol/L, s). Hpx and L- or D-
arginine were administered to the luminal perfusate 45 minutes be-
tein content, 1.4 mg/mL) with proteinase K (200 mg/ fore the HCl exposure. Data shown are means { SEM; n 6 for each
mL, 377C) over 2 hours resulted in an increase in ADMA group.
concentration from 0 to 6.3 mmol/L.

Effects of Exogenously Administered

ADMA
Dose-response relationship between luminally ad-
ministered ADMA and the mucosal alkaline response to
10 mmol/L HCl showed that a perfusate concentration
of 50 and 500 mmol/L significantly (P 0.05) inhibited
the response (Figure 5). Furthermore, 45 minutes after
receiving 50 mmol/L ADMA intraluminally, duodenal
tissue concentration of the compound was 17 { 2 nmol/
g wet wt (n 6), of the same order of magnitude as
shown for the group exposed to Hpx during 45 minutes.
Discussion
This study shows that luminal administration of
H. pylori water extracts inhibits duodenal mucosal alka-
line secretion in response to acid exposure. Such an effect
was not evident when an extract of a nonpathogenic E.
coli strain was used, suggesting some specificity for the
H. pylori species. Furthermore, data indicated a point of
action for Hpx at the L-arginine/NO pathway in the
duodenal mucosa. The involvement of NO in acid-in-
duced mucosal alkaline secretion has been proposed by
Bilski and Konturek based on systemic administration
of NOS-inhibiting arginine analogues in awake dogs.6
We have shown that arginine analogues effectively in-
hibit the acid-induced mucosal alkaline response also
when administered luminally.7 The present study showed
that the Hpx inhibition of acid-induced mucosal alkaline
secretion was counteracted by luminal administration of
3 mmol/L L-arginine, but not the stereoisomer D-argi-
nine. These data indicate that Hpx acts as a false NOS
substrate that can be competed with by high concentra-
tion of the true substrate, L-arginine. Thus, the factor in
Hpx responsible for this effect is presumably an L-argi-
nine analogue. A number of such synthetically derived
arginine analogues, with specificity for NOS, are cur-

Table 1. Concentration of ADMA in Duodenal Perfusate and Tissue

Perfusate Tissue

Ex vivo After 45 min of duodenal luminal After 45 min of duodenal luminal

(mmol/L ) perfusion (mmol/L ) perfusion (nmol/g wet wt)

Control (saline) 0.03 0.11 { 0.01 4.6 { 0.8

Hpx (protein content, 20 mg/mL) 0.03 0.42 { 0.3 a 23.7 { 3.3 b

NOTE. Each value represents mean { SEM from six experiments.

a
P 0.01, bP 0.001; significan difference from control.

/ 5e22$$0005 10-07-97 01:42:55 gasas WBS-Gastro

1574 FANDRIKS ET AL.
Figure 5. Effect of ADMA administered intraluminally on the duodenal
mucosal alkaline response to a 5-minute exposure to 10 mmol/L HCl
(each value is the mean { SEM from 6 separate experiments). *P
0.05, significan difference from zero concentration.
rently used as pharmacological tools in evaluating possi-
ble NO-mediated phenomena. Such compounds have
been reported to stimulate basal duodenal mucosal alka-
line secretion, particularly when given intravenously.11,12
We also observed a transient stimulatory effect after
short-term administration of Hpx. This initial secretory
stimulation, with subsequent blockade of the mucosal
responsiveness to luminal acidity, may indicate a desensi-
tization of a common secretory pathway. Such a phenom-
enon has been described after luminal administration of
capsaicin in this model.13 On the other hand, the ob-
served effects may be caused by simultaneous influences
on different NO-dependent mechanisms. Evidence in-
creasingly shows that NO is involved at several regula-
tory levels within the gut.14 17 It may be that basal
mucosal alkaline output is determined by one NO-medi-
ated system (e.g., NO-dependent enteric nerves) and the
acid-induced increase by another (e.g., surface epithelial
NO synthesis). Further studies are needed to elucidate
these possibilities.
As mentioned, it is reasonable to assume that the
NOS-inhibitory factor in Hpx is structurally related to
L-arginine. Interestingly, some NOS-inhibiting arginine
analogues, particularly methylated arginines, occur en-
/ 5e22$$0005 10-07-97 01:42:55 gasas
GASTROENTEROLOGY Vol. 113, No. 5
dogenously during certain biological conditions.18 In an
attempt to further characterize the factor in the Hpx-
perfusate mediating the inhibition of alkaline mucosal
secretion, we screened for a number of such methylated
arginines. One of these, ADMA, was detected in the
duodenal perfusates and tissue. This arginine analogue
is a potent NOS inhibitor in various cardiovascular test
systems18 and, when administered exogenously, inhibited
the acid-induced alkaline secretory response in the pres-
ent study.
The Hpx-treated perfusate contained approximately
0.4 mmol/L ADMA when the acid-induced mucosal re-
sponse was completely blocked. The concentration was
zero in the original Hpx, which had not been circulated
in vivo. However, ADMA was detected also in duodenal
perfusates not exposed to Hpx. These values were sig-
nificantly lower than after Hpx and probably reflect the
endogenous production as reported previously.18,19 The
increased luminal ADMA concentrations after Hpx are
probably caused by intraluminal protein degradation into
free amino acid residues. It seems reasonable that mem-
brane-bound epithelial peptidases are responsible for the
occurrence of ADMA in the luminal perfusate. Such pro-
teolytic enzymes, in addition to peptide degradation,
facilitate transmembrane transport of the resulting di-
peptides or free amino acids, which, therefore, accumu-
late within the epithelial cell layer.20 When ADMA was
administered exogenously to the perfusate, a secretory
inhibition was obtained with perfusate concentration of
50 mmol/L. The duodenal tissue concentration of ADMA
at this luminal concentration was similar to those after
Hpx, supporting a causal relationship.
It has been reported that ADMA is formed endoge-
nously subsequent to nuclear lysis after cell death.18 A
substantial accumulation of ADMA was evidenced in the
present study by proteolysis of Hpx in vitro. However,
with the presently used preparation of the bacterial ex-
tracts it cannot be distinguished whether it is lysated
cell structures or specifically expressed proteins secreted
by the microbe that are degraded to ADMA. Independent
of the source for mucosal ADMA formation, the present
results explain why the NO-dependent mucosal alkalin-
ization in response to luminal acid is defective in the
presence of degraded H. pylori in our rat model. It re-
mains to be investigated if this is the case also in patients
with duodenal ulcer.4 An active H. pylori infection, by
necessity, produces numerous apoptotic or otherwise
killed bacteria that are degraded in close vicinity of the
mucosal surface. A local H. pyloriderived blockade of
gastroduodenal NOS may be an explanation to the de-
creased mucosal alkaline secretory capacity observed in
WBS-Gastro
November 1997
patients with duodenal ulcer.4 Furthermore, NO is a
regulatory factor of several principal gastrointestinal
functions (e.g., local blood perfusion, mucosa-protective
functions, epithelial permeability, motility), and NO is
used as a cytotoxic agent in inflammatory reactions.14 17
Consequently, it may be of interest to consider inhibition
of the mucosal L-arginine/NO system as a virulence factor
for gastrointestinal pathogens other than H. pylori.
In summary, the present results suggest that the water
extract of H. pylori contains certain peptides that, when
in contact with the duodenal lumen, probably by means
of the inherent digestive proteolytic systems, become
degraded into free amino acid residues, accumulating
within the mucosa. Some of these amino acid residues,
e.g., the presently analyzed ADMA, are potent inhibitors
of NOS that interfere with the mediation of the mucosal
alkaline secretory response to luminal acid.
References
1. El-Omar EM, Penman ID, Ardill JES, Chittajallu RS, Howie C,
McColl KEL. Helicobacter pylori infection and abnormalities of
acid secretion in patients with duodenal ulcer disease. Gastroen-
terology 1995; 109:681 691.
2. McGowan CC, Cover TL, Blaser MJ. Helicobacter pylori and gas-
tric acid: biological and therapeutic implications. Gastroenterol-
ogy 1996; 110:926 938.
3. Olbe L, Hamlet A, Dalenback J, Fandriks L. A mechanism by which
Helicobacter pylori infection of the antrum contributes to the
development of duodenal ulcer. Gastroenterology 1996; 110:
1386 1394.
4. Hogan DL, Rapier RC, Dreilinger A, Koss MA, Basuk PM,
Weinstein WM, Nyberg LM, Isenberg JI. Duodenal bicarbonate
secretion: eradication of Helicobacter pylori and duodenal struc-
ture and function in humans. Gastroenterology 1996; 110:705
716.
5. Allen A, Flemstrom G, Garner A, Kivilaakso E. Gastroduodenal
mucosal protection. Physiol Rev 1993; 73:823 857.
6. Bilski J, Konturek SJ. Role of nitric oxide in gastroduodenal alka-
line secretion. J Physiol Pharmacol 1994; 45:541 553.
7. Johansson B, Bothmer vC, Holm M, Pettersson A, Jonson C,
Fandriks L. Acid induced increase in duodenal mucosal bicarbon-
ate secretion in the rat involves the L-Arg/NO pathway (abstr).
Gastroenterology 1996; 110:A144.
8. Holm M, Bothmer vC, Johansson B, Bolin I, Pettersson A, Fan-
/ 5e22$$0005 10-07-97 01:42:55 gasas
DUODENAL SECRETION, H. PYLORI, AND NO 1575
driks L. A water extract of Helicobacter pylori inhibits acid induced
increase in duodenal mucosal bicarbonate secretion in the rat
(abstr). Gastroenterology 1996; 110:A136.
9. Flemstrom G, Garner A, Nylander O, Hurst BC, Heylings JR. Sur-
face epithelial HCO 30 transport by mammalian duodenum in vivo.
Am J Physiol 1982; 243:G348 G358.
10. Halawa I, Baig S, Qureshi GA. Use of high-performance liquid
chromatography in definin the abnormalities in the free amino
acid patterns in cerebrospinal flui of patients with aseptic men-
ingitis. Biomed Chromatograph 1991; 5:216 220.
11. Takeuchi K, Ohuchi T, Miyake H, Okabe S. Stimulation by nitric
oxide synthase inhibitors of gastric and duodenal HCO 30 secre-
tion in rats. J Pharmacol Exp Ther 1993; 266:1512 1519.
12. Hallgren A, Flemstrom G, Sababi M, Nylander O. Effects of nitric
oxide inhibition on duodenal function in rats: involvement of neu-
ral mechanisms. Am J Physiol 1995; 269:G246 G254.
13. Hamlet A, Jonson C, Fandriks L. The mediation of increased
duodenal alkaline secretion in response to 10 mM HCl in the
anaesthetized rat. Support for the involvement of capsaicin-sen-
sitive nerve elements. Acta Physiol Scand 1992; 146:519 525.
14. Stark ME, Szurszewski JH. Role of nitric oxide in gastrointestinal
and hepatic function and disease (review). Gastroenterology
1992; 103:1928 1949.
15. Whittle BJR. Nitric oxide in gastrointestinal physiology and pathol-
ogy. In: Johnson LR, ed. Physiology of the gastrointestinal tract.
New York: Raven, 1994:267 294.
16. Moncada S, Palmer RM, Higgs EA. Nitric oxide: physiology, patho-
physiology, and pharmacology (review). Pharmacol Rev 1991; 43:
109 142.
17. Anggard E. Nitric oxide: mediator, murderer, and medicine. Lan-
cet 1994; 343:1199 1206.
18. Vallance P, Leone A, Calver A, Collier J, Moncada S. Accumulation
of an endogenous inhibitor of nitric oxide synthesis in chronic
renal failure. Lancet 1992; 339:572 575.
19. Ueno S, Sano A, Kotani K, Kondoh K, Kakimoto Y. Distribution
of free methylarginines in rat tissues and the bovine brain. J
Neurochem 1992; 59:2012 2016.
20. Webb KJ. Intestinal absorption of protein hydrolysis products: a
review. J Anim Sci 1990; 68:3011 3022.
Received December 30, 1996. Accepted July 18, 1997.
Address requests for reprints to: Lars Fandriks, M.D., Centre for
Gastroenterological Research, Goteborg University, P.O. Box 75032,
S-400 36 Goteborg, Sweden. Fax: (46) 31-82-1866.
Supported by grant no. 8663 the Swedish Medical Research Coun-
cil, the Bank of Sweden Tercentenary Foundation, the Gothenburg
Medical Association, and the Knut and Alice Wallenberg Foundation.
The authors thank Anna Casselbrant and Soren Lundberg for excel-
lent technical assistance.
WBS-Gastro