- Email: [email protected]
Regulation of somatostatin-14 and -28 secretion by gastric acid in dogs: Differential role of cholecystokinin
1993:105:1387-1395
GASTROENTEROLOGY
Regulation of Somatostatin- 14 and -28 Secretion by Gastric Acid in Dogs: Differential Role of Cholecystokinin GORDON
R. GREENBERG,*‘*
Departments
of *Medicine
and *Physiology,
LESLIE University
FUNG,* of Toronto,
and SHERRY POKOL-DANIEL* Toronto,
13ackPround: Prosomatostatin-derived peptides include two principle bioactive molecular forms, somatostatin 28 (S-28) and somatostatin 14 (S-14). This study examined whether there is a functional relationship between gastric acid secretion and the release of S-28 and S-14 into the circulation. Methods: In conscious dogs with gastric and duodenal cannulas, S-28 and S-14 responses, measured after extraction of acidified plasma and separation by gel chromatography, were evaluated by administration of nutrients and acid-inducing secretagogues without and with omeprazole. Results: Ingestion of a solid meal caused equivalent plasma elevations of S-28 and S-14, whereas infusions of histamine and gastrin selectively increased plasma S-14. Omeprazole decreased mealstimulated S-28 (-67% + 8%; P < 0.01) and S-14 (-56 + 9%; P < 0.01) and abolished S-14 increases to histamine and gastrin. lntraduodenal perfusions of a liquid protein meal increased S-28 above S-14, comprising -7 1% of total somatostatin-like immunoreactivity released, and omeprazole suppressed S-28 (-87% + 5%; P < 0.0 1) without influencing S- 14. Similar responses occurred after exogenous cholecystokinin. Moreover, pretreatment of the intraduodenal protein meal with the cholecystokinin-A receptorantagonist MK-329 abolished increases of S-28 and S-14 and caused a further twofold increase of gastric acid (P < 0.025). Conclusions: In the fed state, gastric acid causes direct release of S-14 from the stomach, but the acid-dependent component of S-28 secretion requires cholecystokinin as a cofactor. Negative feedback regulation between somatostatin and gastric acid secretory responses to nutrients may include S-28 modulated, in part, by cholecystokinin.
I
Canada
plays a direct major sions
of HCl
crease
and humans.4 This postprandial plasma SLI response is known to be heterogenous and includes variable increases of one or both principle bioactive molecular forms of somatostatin, somatostatin 14 (S-14) and somatostatin 28 (S-2Q5-’ Although fat and protein have been regarded as the major stimuli for SLI release,8*9 recent studies in vitro suggest that gastric acid also
forms.
have
been
S-14 secretion
and mouse excess
from
stomach”
the secretion
the isolated
from
in the isolated by histamine
acidity.”
mediated Whether
gastric
pig antrum”
mouse
of food
increased
intraluminal
acid secretion
results
is, however,
have
been
only plasma
causes direct
whereas
antagonists
omeprazole,
SLI responses the suggestion
of plasma
inhibitor
Hz-receptor
plasma
observation
antagonists
SLI secretion
SLI,
of gastric
on postprandial
in dogs. l4 This latter
present
by which
study
increased
led to alter
independently
of conscious
gated without
and with
zole to exclude
inhibition
that was not related fects of acid secretion,
undertaken
acidity
nu-
of an in-
to determine
is a direct mechanism
of food elevates
SLI molecular
with
was
gastric
ingestion
the circulation
examined
increases
a more potent
that
examining
Administration
effect on acid secretion.
whether
plasma
studies
in dogs13*14 and humans15
had no effect
trient-stimulated hibitory
from
meal-stimulated
acid secretion,
The
obtained
after inges-
and conflicting
total SLI concentrations.
of Hz-receptor suppressed
unknown,
SLI
is also pre-
release of S-14 or S-28 into the circulation tion
in
pig intes-
stomach,
or gastrin
by the
in-
S-28 secretion
the isolated
secretion dominantly
perfu-
to preferentially
and stimulate
tine. ” Moreover, induced
of both
Pharmacological
shown
of S-14 secretion
gastrin, ngestion of a solid meal elevates circulating levels of somatostatin-like immunoreactivity (SLI) in dogsle3
role in regulating
SLI molecular
S-28 and S-14 in
dogs. Nutrient-stimulated
form
responses
ranitidine
were
and with
by an Hz-receptor
to suppression independent
intravenous
and cholecystokinin
of gastric
antagonist acid. Ef-
of nutrients,
infusions (CCK).
investiomepra-
were
of histamine,
A role for CCK in
modulating S-28 and S-14 responses to acid secretion was further investigated with the CCK-A receptor antagonist MK-329, because CCK has been implicated as . . a prmclpal humoral mediator of somatostatin release Abbreviations used in this paper: CCK-8, cholecystokinin octapeptide; IC,, 50% inhibitory concentration; SLI, somatostatin-like immunoreactivity; S-14 and S-28, somatostatin 14 and 28. 0 1993 by the American Gastroenterologlcal Association 0016-5085/93/$3.00
1388
GREENBERG
ET AL.
and inhibition of gastric acid,16 but without consideration of differential effects by CCK on SLI molecular forms
secretion.”
Materials and Methods Animal Preparation The studies were performed in four conscious mongrel dogs weighing 22-30 kg with surgically placed chronic gastric and duodenal cannulas. Food, but not water, was withheld for 18 hours before each study. The dogs were placed in a Pavlov harness, and two indwelling catheters were introduced into leg veins. One catheter was used for blood sampling and the second for administration of test agents. The care and use of dogs in this study were approved by the University of Toronto animal care committee.
Experimental
Protocols
Solid meal studies. In the first series of experiments, three fasting blood samples were taken during a 20-minute basal period, after which 20 g/kg of canned dog food was ingested (total caloric content of 1.5 kcal/g distributed as 31% protein, 39% carbohydrate, and 30% fat). Then blood sampling continued at 20, 30, 40, 60, 80,90, 100, 120, 130, 140, 160, and 180 minutes. bn separate occasions, the effects of oral administration of omeprazole (Astra Pharmaceuticals Inc., Mississauga, Ontario, Canada) or intravenous ranitidine (Glaxo Canada Ltd., Toronto, Canada) were studied once in each animal. Oral omeprazole was given at a dose of 40 mg daily for 4 consecutive days and 2 hours before ingestion of the meal. This dose and duration of omeprazole administered in dogs inhibits by -95% mealstimulated gastric acid secretion.‘* Intravenous infusion of ranitidine at a dose of 1.5 mg - kg-’ - h-’ was begun 40 minutes before the meal and continued for 220 minutes. This dose inhibits by -90% meal-stimulated gastric acid secretion in dogs. ‘s To allow recovery of gastric acid secretion there was a minimum 3-week interval between omeprazole studies and all other experiments. Histamine, gastrin, and CCK infusions. Secretagogue-stimulated release of SLI was studied in the basal state by examining on separate occasions intravenous infusions of histamine (Sigma Chemical Co., St. Louis, MO) at a dose of 50 pg. kg-’ - h-‘, synthetic gastrin 17 (Bachem Inc., Torrance, CA) at a dose of 150 pmol * kg-’ - h-‘, and sulfated CCK octapeptide (CCK-8) (Peninsula Laboratories Inc., Belmont, CA) at a dose of 250 pmol * kg-’ - h-’ with the gastric cannula closed. Histamine was prepared and infused in 50 mL of 0.9% saline, and gastrin and CCK-8 were administered in 50 mL 0.10/6 bovine serum albumin-0.9% saline. After a ZO-minute basal period, histamine, gastrin, or CCK-8 was infused for 90 minutes. On different occasions, the effect of acid suppression by omeprazole was studied at a dose of 40 mg orally for a minimum of 4 consecutive days before and 2 hours before the start of the infusions, or by ranitidine given at a dose of 1.5 mg * kg-’ * h-’ begun 40
GASTROENTEROLOGY
Vol. 105.
No. 5
minutes before the infusions and continued for 130 minutes. On separate days secretagogues were administered with the gastric cannula open. Gastric secretions were collected continuously for 120 minutes and divided into lo-minute aliquots, and the volume and hydrogen ion concentration of each aliquot were measured. Blood samples were obtained at 20 and 10 minutes before, at 0 minute (immediately before the start of the infusion), and at lo-minute intervals thereafter. Intestinal perfusions of a liquid protein meal and fat emulsion. The intestinal phase of SLI release was investigated on separate occasions by the intraduodenal perfusions of a defined formula liquid protein meal (Ensure Plus; Ross Laboratories, Montreal, Quebec, Canada; with a total caloric content of 1 kcal/mL distributed as 15% protein, 53% carbohydrate, and 32% fat with a pH of 7) at a rate of 2.5 mL/min and by intraduodenal perfusion of Intralipid 10% (Kabi Vitrium; Baxter Pharmaceutical, Newmarket, Ontario, Canada) at a rate of 1.7 mL/min for 90 minutes, through an indwelling catheter placed 10 cm distally into the duodenal cannula located opposite the entrance of the main pancreatic duct. The gastric cannula was left open to exclude retrograde flow of nutrients into the stomach and for quantitation of acid secretion. Three blood samples were taken during a 20-minute basal period and then at lo-minute intervals throughout the perfusions. The intraduodenal perfusion studies were performed on separate days without and with omeprazole at a dose and duration as described above. On separate days, perfusion of the intraduodenal protein meal was undertaken with the CCK receptor antagonist MK-329 (75 pg/kg) or its vehicle (2 mL of 100% dimethylsulfoxide in 20 mL of 0.9% saline containing 0.1% bovine serum albumin), administered by bolus intravenous injection 10 minutes before the start of the perfusion. MK-329, previously designated L-364,718, was a generous gift from Dr. D. Veber (Merck Sharpe & Dohme, West Point, PA). Blood sampling and gastric juice collections were performed as described above.
Laboratory
Methods
Blood for SLI determinations was collected in tubes containing 1000 kallikrein inhibitor units of aprotinin (Trasylol; Bayer AG, Germany) and 1.2 mg ethylenediaminetetraacetic acid Na/mL of blood. Samples were rapidly centrifuged, and 1 mol/L HCl was added to a known volume of plasma to adjust the pH to 3 and, thus, inactivate serine proteases and aminopeptidases.” The plasma was stored at -20°C until assayed. Plasma SLI concentrations were measured by radioimmunoassay, as described previ0us1y.~ Before radioimmunoassay, SLI was extracted from plasma with ethanol (3.6 mL 95% ethanol to 1 mL plasma). Plasma and ethanol were carefully mixed and centrifuged at 28OOg for 40 minutes; the supernatants were decanted and evaporated to dryness with compressed air overnight. After ethanol extraction of acidified plasma, the mean recoveries
November 1993
determined fmol
S-14,
in 40 sets of samples
synthetic
3%, respectively. plasma,
containing
S-14 and S-28 were In contrast,
without
the mean recovery
prior
volume.
Extracted
plasma
in assay buffer
with
an
samples
and
the S-14 molecule.
antibody
The
was collectively analysis
S-14 standards, (50% inhibitory mean
intra-assay
interassay
variation
The limit
Hydrogen (autoburette
titrator
NaOH.
on plasma
were collected then
was acidified
extracted
silyl silica Milford,
from
above
alcohol, (2.5-mL
trifluoroacetic
lo-mL
aliquot)
by
5 mL
Two plasma
aliquots air-dried,
alcohol-l%
so treated
samples and the
Associates, Acidified at
acid.
(5 mL of plasma in 2 mL column
buffer. This extraction technique provided recoveries of synthetic S-28 and S-l 4, added to acidified plasma, of 93% + 2% (n = 6) and 87% f 2% (n = 6), respectively.
hormone
the mean
responses
Statistical
comparisons
using Student’s
Reconstituted
component
95%. The mean
to the
column
was
as the mean response
basal concentration during
for each dog
as described
of means within of variance
Integrated previously.3
a group
paired t test, and differences
or the
calculated
each time period.
were calculated
by analysis
+ SEM
in plasma,
were made
between
followed
groups
by a multiple
test. P value I 0.05 was considered
significant.
Results Solid Meal SLI Responses Suppression Basal
plasma
SLI levels
to Acid
of 12 + 2 fmol/mL
were not significantly influenced by administration of omeprazole (9 * 2 fmol/mL) or ranitidine (11 k 2 fmol/mL).
The
effect
of acid
suppression
prandial
SLI responses
ingestion
of the solid meal, plasma
Administration duction
were eluted
trifluoroacetic
exceeded
SLI added
are expressed
to a peak increment
5 mL
the Sep-Pak
peptides
this
S-14 or S-28 added to
Analysis
hormone
meal
alone.
with
water.
were pooled
and reconstituted
Results incremental
and
The cartridge was then sequenvolumes of water and with 0.1%
acid, and the adsorbed
equivalent),
Waters
the
(S-l 3), the chromatographic
plasma
Statistical
gastrin,
on octadecyl
was passed through
5 mL 80% isopropyl
in extracted
of total
comparisons
forms of SLI were
was washed
1.02. Although
not differentiate plasma
omeprazole
and concentrated
followed
does
the column
of
and centrifuged,
(Sep-Pak;
used
of synthetic
were also
blood
S-14,
from S-14. The recovery
were tested
protein
volume.
the Kav for synthetic
S-14, may have been the active circulating
[Des-Ala’lsomatostatin
of the liquid
of the liquid
cartridge
a flow rate of 1 mL/min. tially washed with 5-mL with
un-
or ranitidine;
perfusions
to pH 3. The peptide
A C-18
fraction,
methodology
60 and 90
and with the vehicle
plasma
form
mean
40 and
of histamine,
with and without
(C- 18) cartridges MA).
isopropyl plasma
fasting,
at baseline;
analysis,
as indicated
S-28 was 0.68 and for synthetic smallest
was
were
forms
omeprazole
in its vehicle
conditions,
these experimental
from the concentration
during
taken
perfusion
chromatographic
and I/t the total column
Under
by subtracting
studies
the infusions
meal and fat emulsion
with MK-329
plasma
acid output
aliquot.
after the intraduodenal
and the intraduodenal
ume, I/a the void volume,
ali-
ion concentration
SLI molecular
with and without
60 minutes
For
Gastric
obtained
samples
after initiating
or CCK-8 protein
samples
were applied to a 9 X lOOO-mm Sephadex G-50 (superfine) column calibrated with synthetic cyclic S-14 (Peninsula Laboratories), synthetic S-28 (Bachem Inc.), dextran blue (void volume), cytochrome c (mol wt, 12,384), and “‘I-sodium (total volume). Elution was undertaken with 125 mmol/L ammonium bicarbonate buffer (pH 9) containing 100 mmol/L sodium chloride and 0.1% bovine serum albumin at 4°C. Fractions were collected at a flow rate of 6 mL/h and assayed. Elution positions are expressed as the coefficient of distribution Kav where Kav = (Ve Vo)/(Vt - Vo), with I/e corresponding to the elution vol-
Copenha-
each lo-minute
chromatographic
and omeprazole.
minutes
Radiometer,
from
1389
92% + 3% (n = 48).
by titration
after the meal alone, and after administration
ranitidine measured
the mean
Chromatography
Gel permeation 90 minutes
and
was determined
of each lo-minute
on plasma
was 9.5 fmol/tube.
the hydrogen
Gel Permeation dertaken
of the
AND GASTRIC ACID
samples
recovery
samples
by multiplying
by the volume
S-28 and
of the assay
was 4.3%,
and pH meter;
of 2-mL
quot to pH 7 using 0.025N calculated
from the chro-
was 7.5%.
ion concentration
gen, Denmark)
forms
of detection
[IC,,])
variation
part of
a synthetic
synthetic
and the sensitivity
concentration
antibody
S-28 on an
against
quantitated
respectively.
an
all circulating
read
were read against
assay was 0.3 fmol/tube,
in duplicate
and
also detects
basis. Total SLI, comprising
matographic
plasma
against the central
S- 14; S-28 and S-14 components
The
at 4°C
tracer
directed
S-14 or
of the initial
incubated
(1: 130,000 final dilution)
of somatostatin,
of
of 200 FL were reconsti-
[‘251-Tyr”]somatostatin
equimolar
of total plasma
was independent
tuted
acidification
for S-14 was 62% + 2% and that
for S-28 was 37% + 4%. The recovery S-28 after extraction
20, 40, or 80
78% + 2% and 76% t
S-28,
is shown
in Figure
by 44% of postprandial
After
SLI levels increased
of 17 + 2 fmol/mL
of omeprazole
on postl/l.
caused
(P < 0.01). a sustained
elevations
re-
of plasma
SLI; a similar effect occurred after ranitidine. The integrated SLI responses to the meal after omeprazole and ranitidine were not different, and both responses were significantly (both P < 0.025) value (Figure 1B).
below
the control
Gel permeation chromatography indicated that plasma SLI eluted in three peaks that coeluted with synthetic S-14 (Kav 1.02) and synthetic S-28 (Kav 0.68) and a third peak (Kav 0.15) with a molecular
1390
=
gg
GASTROENTEROLOGY Vol. 105, No. 5
GREENBERG ET AL.
25 -Meal 4
(both P < 0.001 compared
A
In contrast
20-
f
15-
: E I h
lo-
Q
histamine
were
suppressed (Figure
5o-
0
20
40
60
80
100
120
140
160
180
Time (minutes) r
-
7
Sz
acid,
2.0
plasma minutes
-
1.5
-
1.0
-
0.5
-
E E
0.0
-
Intravenous
without (Figure
Meal
min,
col-
gastric
and plasma
were not different of gastrin
of gastric
(P < 0.001) at 90
2 fmol/mL
3B). Plasma
ad-
acid increased
SLI responses
to exoge-
nous gastrin
also were composed
entirely
of S-14 and
were reduced
after omeprazole
by 90% f
5% (Figure
by 87% f 8% (both P < 0.001).
experiments
tion, exogenous
Ornep&~le
increased
infusions
collection
SLI by 17 f
In separate
Meal
histamine
with
values.
4B) and after ranitidine
P Z b
by 87% -t 5% (P < performed
(4.1 f 0.3 fmol/mL)
baseline
to
of S-14 and were
to 15.6 ? 1.4 mmol/30
S- 14 responses
ministered
5 E ,P i fn
lections
Gastrin.
T
2.5
.S E
experiments
of gastric
3A).
by 89% + 6% (P < 0.001)
and after ranitidine
acid output
(Figure
SLI responses
entirely
after omeprazole
4A)
B
E
composed
0.001). In separate
from
3.0
with controls)
to the solid meal, plasma
performed
gastrin
increased
with
acid collec-
gastric acid output
to
Meal Rarhditidine
Figure 1. (A) Mean + SE (fmol/mL) plasma SLI responses (A) after ingestion of a solid meal without (0) and with omeprazole (0) or ranitidine (A). Basal plasma SLI was subtracted from the total to obtain incremental plasma SLI. (6) Integrated release of SLI in response to ingestion of a solid meal alone and after omeprazole and ranitidine. *P < 0.025 vs. meal alone.
weight
of - 13,000
(Figure
U).
Basal levels
and S-14 and the large-molecular-weight
peak
4.1 rf: 0.6, 3.2 f 0.4, and 4.3 f 0.3 fmol/mL, tively. After ingestion plasma
S-28 from
0.8 fmol/mL S-14 from
of the meal, omeprazole
an increment
or ranitidine
2B). Ranitidine increments
(Figure
Infusions of Histamine,
6 r Solid Meal + Omeprazole
Tr,
B
respecreduced
6 r Solid Meal + Ranitidine
T
(P < 0.01)
caused similar of plasma
S-28
by 54% + 11% (P < 0.05) and S-14 by 63% + 8% (P < of the large-molecu0.01) (Figure 2C). Th e magnitude lar-weight peak was not influenced by ingesti n of the solid meal and was not changed by administration of omeprazole
-r
were
(P < 0.01) or by 67% + 8% and plasma
of postprandial
t
of 7.7 + 1.4 to 2.6 +
8.2 + 1.1 to 3.7 f 0.9 fmol/mL
or by 56% + 9% (Figure reductions
of S-28
2).
Gastrin, and CCK
Histamine. Intravenous infusions of histamine, undertaken without collection of gastric acid, caused a progressive increase in plasma SLI to a peak increment of 34 + 6 fmol/mL (P < 0.001) at 90 minutes. After omeprazole or ranitidine treatments, plasma SLI levels were reduced to values that were equivalent to baseline
20
25
30
35
40
Effluent
45
50
volume
55
60
65
70
75
(ml)
Figure 2. Characterization by gel chromatography of 5-mL plasma samples for SLI undertaken on a 9 x lOOO-mm Sephadex G-50 (superfine) column eluted with 125 mmol/L ammonium bicarbonate buffer, pH 9, and calibrated with dextran blue (Vo), cytochrome c (CC), S-28 S- 14, and ‘251-sodium as total volume marker. Entire fractions (0.6 ‘mL each) were assayed for SLI content. Points represent means + SEM of four experiments. (A) Plasma obtained before (0) and 40 minutes after (0) ingestion of a solid meal. (6) Plasma obtained 40 minutes after ingestion of a solid meal and treatment with omeprazole. (C) Plasma obtained 40 minutes after ingestion of a solid meal and treatment with ranitidine.
S- 14, S-28, AND GASTRIC ACID
November 1993
C
CCK- Of’ (250
1391
pmol ko”h”l
I
4
35 -
= 5E
30 -
zE
25 -
2
20 -
tZ
15-
Figure 3. Mean + SE (fmol/ mL) plasma SLI responses (A) without (0) and with (0) administration of omeprazole or ranitidine (A) after infusions of (A) histamine, (13) gastrin, and (C) CCK8.
a
10 -
9.2 f. 1.5 mmol/30
min,
and plasma
(3.9 +- 0.4 fmol/mL)
were not different
90
0
30
S-14 responses from baseline
omeprazole
(2.9 + 1.4 fmol/mL),
did not reach
CCK-8.
Intravenous
minutes,
and similar
Omeprazole
and
of both
respectively),
reduced
plasma
3C). In contrast
gastrin,
CCK-8
sustained.
caused
plasma
of
eleva-
(Figure
4C). After
increment
that of
omeprazole
of S-28 was reduced
1.9 fmol/mL
58% f 14%, but the incremental
(P < 0.01)
S-14 increase
was not significantly of 6.9 + 1.4 fmol/mL
different (Figure
with
a liquid
protein
by 23 f 5 fmol/mL prazole from
caused
their meal
sustained
40 to 90 minutes
plasma
S-28
meal increased
(P < 0.001)
and
S-14
reductions
the increase
(P < 0.01) (Figure increased
values after the intraduodenal (P < 0.001 and P < 0.05, respectively)
6A), although the increase s-14 (P < 0.05), comprising After
administration
influence
5B) or the plasma
the
elevations 6B).
from
ies, pretreatment meal
with
duodeSLI omeSLI from protein (Figure
the in-
crement of plasma S-28 was decreased from 10.8 f 1.3 to 1.2 f 0.4 fmol/mL (P < 0.001) or by 87% f 5%; the incremental plasma S-14 response with omeprazole (5.4 ? 2.2 fmol/mL) tended to be higher than without
of the intraduodenal
the CCK-A
abolished 0.001)
of S-28 exceeded that of 71% f 4% of the total SLI of omeprazole,
(Figure
did not
stud-
5A). Both
significantly
of omeprazole
with control
Nutrients
baseline
released.
(P < 0.01) and S-14 to (P < 0.001) (Figure 6B), although of S-14 exceeded that of S-28 (P < 0.05).
Effect of MK-329. Compared
4C).
of plasma
to
of both
treat-
of 7.8 +
plasma
increments
fat (Figure
or by
at 40 minutes;
plasma
of S-28 and S-14 after intraduodenal
from con-
of the
at 60
5B). The SLI responses
14.6 +- 2.5 fmol/mL
increases
(Figure
peak output traduodenal blockade
meal
puts during
further
with plasma
fat,
7B). After
CCK-A
increases
with controls)
from baseline; the third
period
from
a
of 3.4 -t 1.7 to 8.2 +-
however,
in-
receptor
of S-28 (1.3 +
and S-14 (2.3 + 0.1 fmol/mL)
P < 0.001 compared not different
period
(P < 0.025) (Figure
also reduced
protein MK-329
the gastric acid output
was increased
in the third min
liquid
antagonist
S-28 and S-14 (both P <
of plasma
perfusion
0.2 fmol/mL)
receptor
7A). C onversely,
the protein
2.3 mmol/30
Liquid protein meal. Perfusion
GA). a fat emul-
S-28 to 6.2 + 1.1 fmol/mL
after
SLI Responses to lntraduodenal
with
SLI by 26 + 7 fmol/mL
significant
total SLI increase
14.7 + 0.8 to 6.1 f
num
plasma
but the increase
of S-28 exceeded
perfusion
(P < 0.01) (Figure
minutes
fat comprised
SLI by 12% + to infusions
sion increased
Administration
the plasma
2.3 fmol/mL
ele-
S-28 and S-14 (P < 0.001 and P < 0.01,
S-14 (P < 0.05)
trol values
of CCK-8
(P < 0.001) at 20
levels were thereafter
marginally
3% (P = NS) (Figure histamine
infusion
SLI by 21 f 3 fmol/mL
but this difference
(P = 0.2) (Figure
significance
Fat. Intraduodenal
vated plasma
ment,
90
Time (minutes)
values.
tions
60
(both
to values that were gastric
with MK-329
acid out(0.9 z!z 0.4
mmol/30 min) and without MK-329 (0.6 f mmol/30 min) were not different from that during basal period (0.3 + 0.2 mmol/30 min).
0.3 the
Diiscussion The present study shows that in conscious dogs, gastric acid secretion stimulated by ingestion of a meal regulates the release of both S-28 and S-14 into the circulation. Administration of omeprazole or raniti-
1392
GREENBERG
GASTROENTEROLOGY
ET AL.
Histamine 1501.10kg.‘h.‘l vo cc 18
-
sion after omeprazole
4
16 -
was similar
to the
SLI reduction
after cimetidine
treat-
ment
in humans.
l5 Our
differ,
other
studies
omeprazole,
12
-
These divergent
results probably
cal differences,
because
reduced
dose intravenous 8-
a higher
6-
tion
4-
Further, bition
0t
2 B E r z
-
Gastrin (150 pmol ka”h”)
to suppression
S-14 and S-28.
6-
heterogeneous
suggestions’5
and
that inhi-
is indirect
and
acid secretion.
indicated
that in the fed secretion
SLI molecular
of both forms
are
in several species,7*12,21,22but our equiva-
with
plasma results
elevations
of other
of S-14 and S-28
studies
with
dogs21,22 in
which nutrients required adjustment to pH 3 for the measurement of circulating increases of S-14 and
O12
acidity.*’
to omeprazole
in part,
Circulating
SLI secre-
in gastric
of gastric analysis
lent postprandial contrast
stimulates
antagonists
acid mediated,
8-
2-
with
accord with previous
state gastric
4-
single-
SLI responses
Chromatographic 10
study,i4
was coadministered
dose that
of SLI by H,-receptor
SLI.
relate to methodologi-
of changes
the parallel
from but not
plasma
in the latter
omeprazole
pentagastrin
related
however,
ranitidine,
nutrient-stimulated
independently
ranitidine
2
Z
results
in dogs14 in which
-
-
No. 5
45% postprandial
14
10
.o z
or ranitidine
Vol. 105.
r
C
CCK-OP (250 pmd kg”h”)
10 -
without
detectable
rapidly
degraded
prompt
increases in plasma;
acidification
of S-28. However, recent
of samples
8-
adequate
recoveries
6-
humans,’
the total SLI increase
cular-weight 4-
This
2-
peak
ations
of gastric
20 25 30 35 40 45 50 55 60 65 70 75 Effluent
tion
is a prerequisite
for
of S-28.‘9,23 Similar
stimulates
influenced
to findings
included
in
a large-mole-
prosomatostatin.
by nutrients
or alter-
acid secretion.
The mechanisms
O-
that
peak that may represent was not
S-28 is
studies indicate
by which
release
postprandial
acid secre-
of SLI have not been clarified.
volume (ml) A
Figure 4. Characterization by gel chromatography of 5-mL plasma samples for SLI, using the column conditions described in Figure 2. Points represent means + SE of four separate experiments. (A) Plasma obtained 60 minutes after intravenous infusions of histamine without (0) and with (0) omeprazole. (6) Plasma obtained 60 minutes after intravenous infusions of gastrin without (0) and with (0) omeprazole. (C) Plasma obtained 60 minutes after infusions of CCKS without (@) and with (0) omeprazole.
dine caused equivalent suppression of postprandial plasma elevations of S-28 and S-14, but the mechanism by which gastric acidity mediated secretion of S-14 was distinct from that of S-28. Increased intraluminal gastric acidity selectively stimulated S-14 secretion, whereas the acid-dependent component of S-28 secretion required intestinal release of CCK as a cofactor. Previous studies showed increases of plasma SLI levels in the fed state without consensus on the influence of SLI suppresof gastric acidity. 14*15The magnitude
40
Intraduodmal Protein Meal i
r
0
lntraduodrnal Fat
(2.6 ml/ mid
30
60
(1.7 ml/ mid
90
0
30
50
1
90
Time (minutes) 5. Mean + SE (fmol/mL) plasma SLI responses (A) after (A) intraduodenal perfusions of liquid protein meal without (0) and with (0) omeprazole and after (13)intraduodenal perfusions of a fat emulsion without (0) and with (0) omeprazole.
Figure
November 1993
S-14,
Intraduodwd Protein Meal 12 -
VO
cc
1
I
vitro
studies
minal from
10 -
gastric
similar ach,”
4-
Stimulation
t; t” . 5 E
c i
co
18
the specificity
for stimulation The pattern
histamine,
of the
of S-14
secretion
of SLI forms
of the isolated
vagus
nerves
the fed state gastric
acid secretion,
by vagal, hormonal,
or paracrine mechanism
form
gastric
released of HCI is
perfused
also
1393
of intralu-
or the instillation
S-14 via increased
be a predominant
B
Intraduodenal Fat
r 16 -
AND GASTRIC ACID
and S-14 is the only molecular
stimulates
E
support
for both regions
6-
._
acid
the stomach.
after gastrin,
8-
further
S-28,
stom-
secreted.12
preferentially
acidity.23 Thus, whether
pathways,
in
mediated appears
to
for release of S-14 from
the stomach. In contrast
to S-14, increasing
14 -
alone could
12 -
dial S-28 release. evidence
10 -
not account Rather,
for a mechanism
lates release
intraluminal
acidity
for acid-dependent the present involving
of SLI from both
postpran-
studies
provide
CCK. CCK stimu-
fundic28
and antra12” D
86-
2O-
A
Intraduodenal Protein Meal
12 r
4=
,
81
I
I
81
I
I
I
a
1
20 25 30 35 40 45 50 55 60 65 70 75 Effluent
volume
(ml)
Figure 6. Characterization by gel chromatography of 5-mL plasma samples for SLI, using the column conditions described in Figure 2. Points represent means + SE of four separate experiments. (A) Plasma obtained 60 minutes after intraduodenal perfusions with a liquid protein meal without (0) and with (0) omeprazole. (3) Plasma obtained 60 minutes after intraduodenal perfusions with a fat emulsion without (0) and with (0) omeprazole.
.; 0 t t 5 =E i oY
VO
cc
1
I
s29
514
4
1
lo86 42-
O-
I1
I,
20 25 30 35 40 45 50 55 60 65 70 75 Effluent volume
(ml)
B
Intraduodenal Protein Meal
I
Studies
using
indicated histamine-
isolated
perfused
stomach
(2.5 mlhnin)
I T
preparations
a direct effect by acid on D cells, because and gastrin-induced luminal secretion of
SLI could be largely accounted for by the increases of intraluminal acidity. l1 Our results show that histamine- and gastrin-stimulated acid secretion potently stimulated
SLI secretion
into the circulation,
but SLI
responses were composed entirely of S-14. Pharmacological suppression of gastric acid abolished increases of S- 14 after both secretagogues, implying that intraluminal gastric acid caused direct and selective release of S-14. S-14 is present in both stomach and pancreas, but the stomach was the likely source of S-14, because the pancreas makes only a minor contribution to circulating SLI. 24 Although morphological differences2’ and SLI molecular heterogeneitya6,27 between D cells in the fundus (closed-type secreting S-28 and S-14) and antrum (open-type secreting S- 14) are well-described, in
9 Y
2-
01,
I
MK-329 (75 u9kgI or v&de
Basal
1 30 min Periods
2
3
Figure 7. (A) Characterization by gel chromatography of 5-mL plasma samples for SLI, using the column conditions described in Figure 2. Plasma was obtained 60 minutes after intraduodenal perfusions of a liquid protein meal with intravenous MK-329 (0) or its vehicle (0). (6) Mean + SE gastric acid output (mmol/30 min) from basal state and during stimulation by an intraduodenal liquid protein meal with intravenous MK-329 (0) or its vehicle (0).
GREENBERG ET AL.
1394
cells in culture
GASTROENTEROLOGY Vol. 105, No. 5
and from the same regions
3o by activation
stomach,
of specific
ceptors. 31 In conscious stimulation gastrin
dogs,
of basal gastric
in the stimulation
lated
CCK
inhibit
also causes
re-
weak
acid32 but is equipotent of acid production
to
from iso-
cells. 33 These observations lead to the that CCK in the intact dog releases SLI to
parietal
SLI released
cell function.
Our results
after CCK is composed
S-28 and that the S-28 response, gastric
acid.
The
low
after CCK compared accounted release
of S-28
close proximity lumen.
with gastrin
only closed-type are closely intact
CCK
aligned
with
instillation
separate
production releases further
tained
with
fundus
that in the in re-
S-28 in a local
fundic
by which
D
CCK
Protein
and fat both protein
CCK-A prandial
because
receptor total
recent
antagonist
SLI and causes
studies MK-329
CCK effects on acid secretion indicate function.
preferential
mediates
effects
ach, but the mechanism release
CCK,‘6,36
feedback
gastric also
S-28 release.
have shown suppresses
the post-
increase
of
begastric acid. l6 Our results extend these findings cause MK-329 abolished S-28 increases after intraduodenal protein and further increased gastric acid secretion, implying a specific reciprocal functional relationship between gastric acid and S-28 in the fed state mediated by CCK. The concept that S-28 selectively regulates CCK actions is supported by observations in vitro derived from rat pancreatic acinar plasma membranes 38, however, proof for this hypothesis in vivo awaits the development of selective S-28 antagonists. Although CCK acts via the circulation to release S-28, it is likely that S-28 regulates gastric acid by paracrine mechanisms. Inhibition of parietal cell
nisms
only
by CCK-A
on parietal
cell
that gastric
acid
gastric
and requires Models
between
humoral nega-
somatostatin
and
should
in-
the secretion
of
D cells in the fundus receptor-dependent inhibition
acid stimu-
describing
to nutrients
also S-28,
d e 1ivery to parietal
may
of S-14 from the stom-
responses
S-14 but
with consequent
a paracrine
by S-28
by which
regulation
acid secretory
The
mediating
in rats and humans
is indirect
of CCK as a cofactor.
is mediated
pathway
the secretion
lates S-28 secretion
MK-329
fat but did
in acid secretion.
this study has shown
directly
CCK was obafter intes-
increase
SLI-dependent
from closed-type
protein
a reciprocal
an
suggests
in dogs is un-
by S-14 because
after intraduodenal
a reciprocal of
which
These observations raise the issue of a role for CCKmediated S-28 secretion in the regulation of gastric acid secretion,
so without presence
S-28 secre-
stimulates
acid-dependent
S-14 release
not
release
acid.37 Our data show that intraduodenal stimulates
abolished
in
in dogs
study
effect of CCK
mediated
of gastric pathway
mechanism
‘6*44The present
the enterogastrone
likely to be entirely
suppression
an SLI-dependent
controversial.
to specific
CCK in the modula-
but a similar
clude
of SLI forms
intraduodenal
remains
gastric released
fat-induced
through
of
reflecting
responses
studies implicate
and
ingestion
only S-28,4’ perhaps
in SLI form
Recent
of basal
tive
acid environment
for acid-dependent
by endogenously
nutrients.
con-
fat in dogs caused
of S-14, whereas
releases
differences
In summary,
but not to intralu-
mechanisms
characterization
preferentially
in
and body
occurs
species
that
that
via the
cells25; also,
the fundus
increase
plasma
acid secretion.40
of intraduodenal
a predominant fat in humans
to S-14 in vitro,39 but
physiological
do not influence
Administration
rats42 and humans,43
investigation.
nutrients. only
acting
centrations
acid secretion
secreted
of S-28 and S-14 from
support
tion mediated
than
of S-28 to achieve
tion
likely
by observations
parietal from
cells,35 the intracellular
Additional
acid
by S-28 is equipotent
infusions
of
achieved
suggest
function
by
of HCl. 34 Although there are two processing pathways for the
posttranslational
requires
data
requires
acid secretion
different
selectively
these
D cells in the gastric
to stimulated
minal
acidity
or histamine16
is supported
dog, SLI release
sponse
predominantly
to D cells rather
This notion
that
of an effect by omeprazole
Together, after
indicate
in part, is mediated
intraluminal
for the absence
on S-14 secretion.
but
CCK
parietal
hypothesis
tinal
of the intact
type-A
and body mecha-
of acid secretion
by
cells.
References 1. Schusdziarra V, Rouiller D, Harris V, Unger RH. Gastric and pancreatic release of somatostatin-like immunoreactivity during the gastric phase of a meal: effects of truncal vagotomy and atropine in the anesthetized dog. Diabetes 1979;28:658-663. 2. Hollinshead JW, Debas HT, Yamada T, Elashoff J, Osadchey B, Walsh JH. Hypergastrinemia develops within 24 hours of truncal vagotomy in dogs. Gastroenterology 1985;88:35-40. 3. Greenberg GR. Influence of vagal integrity on gastrin and somatostatin release in dogs. Gastroenterology 1987;92:994100 1. 4. Lucey MR, Wass JAH, Fairclough P, Webb J, Webb S, Medbak S, Rees LH. Autonomic regulation of postprandial plasma somatostatin, gastrin and insulin. Gut 1985;26:683-688. 5. Polonsky KS, Shoelson SE, Docherty HM. Plasma somatostatin28 increases in response to feeding in man. J Clin Invest 1983;71:1514-1518. 6. O’Shaughnessy DJ, Long RG, Adrian TE, Christofides ND, Ghatei MA, Sarson DL, Bloom SR. Somatostatin-14 modulates postprandial glucose levels and release of gastrointestinal and pancreatic hormones. Digestion 1985;3 1:234-242. 7. Ensinck JW, Laschansky EC, Vogel RE, Simonowitz DA, Roos BA, Francis BH. Circulating prosomatostatin-derived peptides. Differential responses to food ingestion. J Clin Invest 1989;83: 15801589. 8. Lucey MR. Fairclough PD, Wass JAH, Kwasowski P, Medbak S,
November 1993
S-14,
Webb J, Rees LH. Response of circulating somatostatin, insulin, gastrin and GIP to intraduodenal infusion of nutrients in normal man. Clin Endocrinol (Oxf) 1984;21:209-217. 9. Schusdziarra V, Rouiller D, Pietri A. Harris V, Zyznar E, Conlon JM, Unger RH. Pancreatic and gastric release of somatostatin-like immunoreactivity during intestinal phase of a meal. Am J Physiol 1979;237:E555-E560. IO. Holst JJ, Jensen SL, Knuhtsen S, Nielsen OV, Rehfeld JF. Effect of vagus, gastric inhibitory polypeptide, and HCI on gastrin and somatostatin release from perfused pig antrum. Am J Physiol 1983;244:G5 15-G522.
28.
29.
30.
11. Schubert ML, Edwards NF, Makhlouf GM. Regulation of gastric somatostatin secretion in the mouse by luminal acidity: a local feedback mechanism. Gastroenterology 1988;94:3 17-322. 12. Baldissera FGA, Nielsen OV, Holst JJ. The intestinal mucosa preferentially releases somatostatin-28 in pigs. Regul Pept 1985;11:251-262.
31.
13. Schusdziarra V, Stepelfeldt W, Klein M, Maier V. Role of histamine H, receptor in the modulation of pancreatic and gastric function. Acta Endocrinol (Copenh) 1982;99(Suppl 246): 11 1.
32.
14. Simoens C, Woussen-Colle MC, DeGraef J. Effect of cimetidine, ranitidine and omeprazole on postprandial gastrin and somatostatin release in dogs. Regul Pept 1988;22:285-293. 15. Lucey MR, Wass JAH, Rees LH, Dawson AM, Fairclough PD. Relationship between gastric acid and elevated plasma somatostatin-like immunoreactivity after a mixed meal. Gastroenterology 1989;97:967-872. 16. Konturek SJ, Bilski J, Tasler J. Ciezkowski M. Role of cholecystokinin in the inhibition of gastric acid in dogs. J Physiol 1992;45 1:477-489. 17. Greenberg GR. Differential neural regulation ofcirculatingsomatostatinand -28 in conscious dogs. Am J Physiol 1993;264:G902-G909. 18. Ryberg B, Larsson H, Carlsson E. Gastric acid secretion and plasma gastrin levels in dogs treated with high oral doses of ranitidme and omeprazole. Stand J Gastroenterol 1986;21:9495. 19. Francis BH, Ensinck JW. Differential alterations of the circulating prosomatostatin-derived peptides during insulin-induced hypoglycemia in man. J Clin Endocrinol Metab 1987;65:880-884. 20. Shulkes A, Read M. Regulation of somatostatin secretion by gastrin- and acid-dependent mechanisms. Endocrinology 199 1; 129:2329-2334. 21. Schusdziarra V, Zynar E, Rouiller D, Harris V, Unger RH. Free somatostatin in the circulation: amounts and molecular sizes of somatostatin-like immunoreactivity in portal, aortic and vena caval plasma of fasting and meal-stimulated dogs. Endocrinology 1980;107:1572-1576. 22.
Klaff LJ, Dunmng BE, Taborsky GJ. Somatostatin-28 does not regulate islet function in the dog. Endocrinology 1988;123: 2668-2674.
23. Greenberg GR, Pokol-Daniel S, Fung L. Influence of gastric acid on circulating somatostatin-14 and -28 released after insulin-induced hypoglycemia in conscious dogs. Endocrinology 1992; 13 1: 1527- 1533. 24. Taborsky GJ, Ensinck JW. Contribution of the pancreas to circulating somatostatin-like immunoreactivity in the normal dog. J Clin Invest 1984;73:2 16-223. 25. Larsson LI, Golterman H, De Magistris L, Rehfeld JF, Schwartz TW. Somatostatin cell processes as pathways for paracrine secretion. Science 1979;205: 1393- 1395. 26. Penman E, Wass JAH. Butler MG, Penny ES, Price J, Wu P, Rees LH. Distribution and characterization of immunoreactive somatostatin in human gastrointestinal tract. Regul Pept 1983;7:5365. 27. Baldissera FGA, Host JH, Jensen SL, Krarup T. Distribution and
S-28,
AND GASTRIC ACID
1395
molecular forms of peptides containing somatostatin in immunodeterminants in extracts from the entire gastrointestinal tract of man and pig. Biochem Biophys Acta 1985;838: 132- 143. Sol1 AH, Amirian DA, Park J, Elashoff JD, Yamada T. Cholecystokinin potently releases somatostatin from canine fundic mucosal cells in short-term culture. Am J Physiol 1985;248:G569-G573. Buchan AMJ, Curtis SB, Meloche RM. Release of somatostatin immunoreactivity from human antral D-cells in culture. Gastroenterology 1990;99:690-696. Roullier D, Schusdziarra V, Harris V, Unger RH. Release of pancreatic and gastric somatostatin-like immunoreactivity in response to the octapeptide of cholecystokinin, secretin, gastric inhibitory peptide and gastrin-17 in dogs. Endocrinology 1980; 107:524529. Buchan AMJ, Meloche RM, Kwok YN, Kofed H. Effect ofcholecystokinin and secretin on somatostatin release from cultured antral cells. Gastroenterology 1993; 104: 14 14- 14 19. Mayer EA, Elashoff J, Mutt V, Walsh JH. Reassessment of gastric acid inhibition by cholecystokinin and gastric inhibitory polypep-
tide in dogs. Gastroenterology 1982;83: 1047- 1050. 33. Sol1 AH, Amiran DA, Thomas LP, Reedy JD, Elashoff JD. Gastrin receptors on isolated canine parietal cells. J Clin Invest 1984;73: 1434- 1447. 34. Schusdziarra V, Harris V, Conlon JM, Arimura A. Pancreatic and gastric somatostatin release in response to intragastric and intraduodenal nutrients and HCI in the dog. J Clin Invest 1978;62:509-5 18. 35. Chiba T, Park J, Yamada T. Biosynthesis of somatostatin in canine fundic D-cells. J Clin Invest 1988:81:282-287. 36. Lilja P, lnoue K, Wiener I, Fried GM, Greeley GH Jr, Thompson JC. Release of cholecystokinin in response to food and intraduodenal fat in pigs, and dogs and man. Surg Gynecol Obstet 1984; 15:557-56 1. 37. Konturek SJ, Konturek J, Cieszkowski M, Ebet-t R, Creutzfeldt W. Comparison of gastric inhibitory polypeptide and intraduodenal or intravenous fat on gastric acid secretion from vagally innervated and denervated canine stomach. Dig Dis Sci 1986;31:49-56. 38. Jouti-Tahiri N, Dufresne M, Viguerie N, Fourmy D, Esteve JD, Rivier J, Moroder L, Susini C, Vaysse N. Somatostatin 28 interacts with CCK receptors in brain and pancreas. Neuropeptides 199 1; 19:65-7 1. 39. Park J, Chiba T, Yamada T. Mechanisms for direct inhibition of canine gastric parietal cells by somatostatin. J Biol Chem 1987;262:14190-14196. 40. Seal A, Yamada T, Debas H, Hollinshead J, Osadchey B, Aponte G, Walsh J. Somatostatin-14 and -28: clearance and potency on gastric function in dogs. Am J Physiol 1982;243:G97-G 102. 41. Ensinck JW, Vogel RE, Laschansky EC, Francis BH. Effect of ingested carbohydrate, fat, and protein on the release of somatostatinin humans. Gastroenterology 1990;98:633-638. 42. Lloyd KCK, Raybould NE, Walsh JH. Cholecystokinin inhibits gastric acid secretion through type “A” cholecystokinin receptors and somatostatin in rats. Am J Physiol 1992;263:G287-G292. 43. Jebbink MCW, Lamers CBHW, Mooy DM, Rovati LC, Jansen JBMJ. Effect of loxiglumide on basal and gastrin- and bombesinstimulated gastric acid and serum gastrin levels. Gastroenterology 1992;103:1215-1220. 44. Lloyd KCK, Maxwell V, Ohning G, Walsh JH. Intestinal fat does not inhibit gastnc function through a hormonal somatostatin mechanism in dogs. Gastroenterology 1992; 103: 122 1- 1228.
Received March 15, 1993. Accepted May 4, 1993. Address requests for reprints to: Gordon R. Greenberg, M.D., Room 6356, Medical Sciences Building, University of Toronto, Toronto, Ontario, Canada M5S lA8. Supported by a grant from the Medical Research Council of Canada (MA6763).
GASTROENTEROLOGY
Regulation of Somatostatin- 14 and -28 Secretion by Gastric Acid in Dogs: Differential Role of Cholecystokinin GORDON
R. GREENBERG,*‘*
Departments
of *Medicine
and *Physiology,
LESLIE University
FUNG,* of Toronto,
and SHERRY POKOL-DANIEL* Toronto,
13ackPround: Prosomatostatin-derived peptides include two principle bioactive molecular forms, somatostatin 28 (S-28) and somatostatin 14 (S-14). This study examined whether there is a functional relationship between gastric acid secretion and the release of S-28 and S-14 into the circulation. Methods: In conscious dogs with gastric and duodenal cannulas, S-28 and S-14 responses, measured after extraction of acidified plasma and separation by gel chromatography, were evaluated by administration of nutrients and acid-inducing secretagogues without and with omeprazole. Results: Ingestion of a solid meal caused equivalent plasma elevations of S-28 and S-14, whereas infusions of histamine and gastrin selectively increased plasma S-14. Omeprazole decreased mealstimulated S-28 (-67% + 8%; P < 0.01) and S-14 (-56 + 9%; P < 0.01) and abolished S-14 increases to histamine and gastrin. lntraduodenal perfusions of a liquid protein meal increased S-28 above S-14, comprising -7 1% of total somatostatin-like immunoreactivity released, and omeprazole suppressed S-28 (-87% + 5%; P < 0.0 1) without influencing S- 14. Similar responses occurred after exogenous cholecystokinin. Moreover, pretreatment of the intraduodenal protein meal with the cholecystokinin-A receptorantagonist MK-329 abolished increases of S-28 and S-14 and caused a further twofold increase of gastric acid (P < 0.025). Conclusions: In the fed state, gastric acid causes direct release of S-14 from the stomach, but the acid-dependent component of S-28 secretion requires cholecystokinin as a cofactor. Negative feedback regulation between somatostatin and gastric acid secretory responses to nutrients may include S-28 modulated, in part, by cholecystokinin.
I
Canada
plays a direct major sions
of HCl
crease
and humans.4 This postprandial plasma SLI response is known to be heterogenous and includes variable increases of one or both principle bioactive molecular forms of somatostatin, somatostatin 14 (S-14) and somatostatin 28 (S-2Q5-’ Although fat and protein have been regarded as the major stimuli for SLI release,8*9 recent studies in vitro suggest that gastric acid also
forms.
have
been
S-14 secretion
and mouse excess
from
stomach”
the secretion
the isolated
from
in the isolated by histamine
acidity.”
mediated Whether
gastric
pig antrum”
mouse
of food
increased
intraluminal
acid secretion
results
is, however,
have
been
only plasma
causes direct
whereas
antagonists
omeprazole,
SLI responses the suggestion
of plasma
inhibitor
Hz-receptor
plasma
observation
antagonists
SLI secretion
SLI,
of gastric
on postprandial
in dogs. l4 This latter
present
by which
study
increased
led to alter
independently
of conscious
gated without
and with
zole to exclude
inhibition
that was not related fects of acid secretion,
undertaken
acidity
nu-
of an in-
to determine
is a direct mechanism
of food elevates
SLI molecular
with
was
gastric
ingestion
the circulation
examined
increases
a more potent
that
examining
Administration
effect on acid secretion.
whether
plasma
studies
in dogs13*14 and humans15
had no effect
trient-stimulated hibitory
from
meal-stimulated
acid secretion,
The
obtained
after inges-
and conflicting
total SLI concentrations.
of Hz-receptor suppressed
unknown,
SLI
is also pre-
release of S-14 or S-28 into the circulation tion
in
pig intes-
stomach,
or gastrin
by the
in-
S-28 secretion
the isolated
secretion dominantly
perfu-
to preferentially
and stimulate
tine. ” Moreover, induced
of both
Pharmacological
shown
of S-14 secretion
gastrin, ngestion of a solid meal elevates circulating levels of somatostatin-like immunoreactivity (SLI) in dogsle3
role in regulating
SLI molecular
S-28 and S-14 in
dogs. Nutrient-stimulated
form
responses
ranitidine
were
and with
by an Hz-receptor
to suppression independent
intravenous
and cholecystokinin
of gastric
antagonist acid. Ef-
of nutrients,
infusions (CCK).
investiomepra-
were
of histamine,
A role for CCK in
modulating S-28 and S-14 responses to acid secretion was further investigated with the CCK-A receptor antagonist MK-329, because CCK has been implicated as . . a prmclpal humoral mediator of somatostatin release Abbreviations used in this paper: CCK-8, cholecystokinin octapeptide; IC,, 50% inhibitory concentration; SLI, somatostatin-like immunoreactivity; S-14 and S-28, somatostatin 14 and 28. 0 1993 by the American Gastroenterologlcal Association 0016-5085/93/$3.00
1388
GREENBERG
ET AL.
and inhibition of gastric acid,16 but without consideration of differential effects by CCK on SLI molecular forms
secretion.”
Materials and Methods Animal Preparation The studies were performed in four conscious mongrel dogs weighing 22-30 kg with surgically placed chronic gastric and duodenal cannulas. Food, but not water, was withheld for 18 hours before each study. The dogs were placed in a Pavlov harness, and two indwelling catheters were introduced into leg veins. One catheter was used for blood sampling and the second for administration of test agents. The care and use of dogs in this study were approved by the University of Toronto animal care committee.
Experimental
Protocols
Solid meal studies. In the first series of experiments, three fasting blood samples were taken during a 20-minute basal period, after which 20 g/kg of canned dog food was ingested (total caloric content of 1.5 kcal/g distributed as 31% protein, 39% carbohydrate, and 30% fat). Then blood sampling continued at 20, 30, 40, 60, 80,90, 100, 120, 130, 140, 160, and 180 minutes. bn separate occasions, the effects of oral administration of omeprazole (Astra Pharmaceuticals Inc., Mississauga, Ontario, Canada) or intravenous ranitidine (Glaxo Canada Ltd., Toronto, Canada) were studied once in each animal. Oral omeprazole was given at a dose of 40 mg daily for 4 consecutive days and 2 hours before ingestion of the meal. This dose and duration of omeprazole administered in dogs inhibits by -95% mealstimulated gastric acid secretion.‘* Intravenous infusion of ranitidine at a dose of 1.5 mg - kg-’ - h-’ was begun 40 minutes before the meal and continued for 220 minutes. This dose inhibits by -90% meal-stimulated gastric acid secretion in dogs. ‘s To allow recovery of gastric acid secretion there was a minimum 3-week interval between omeprazole studies and all other experiments. Histamine, gastrin, and CCK infusions. Secretagogue-stimulated release of SLI was studied in the basal state by examining on separate occasions intravenous infusions of histamine (Sigma Chemical Co., St. Louis, MO) at a dose of 50 pg. kg-’ - h-‘, synthetic gastrin 17 (Bachem Inc., Torrance, CA) at a dose of 150 pmol * kg-’ - h-‘, and sulfated CCK octapeptide (CCK-8) (Peninsula Laboratories Inc., Belmont, CA) at a dose of 250 pmol * kg-’ - h-’ with the gastric cannula closed. Histamine was prepared and infused in 50 mL of 0.9% saline, and gastrin and CCK-8 were administered in 50 mL 0.10/6 bovine serum albumin-0.9% saline. After a ZO-minute basal period, histamine, gastrin, or CCK-8 was infused for 90 minutes. On different occasions, the effect of acid suppression by omeprazole was studied at a dose of 40 mg orally for a minimum of 4 consecutive days before and 2 hours before the start of the infusions, or by ranitidine given at a dose of 1.5 mg * kg-’ * h-’ begun 40
GASTROENTEROLOGY
Vol. 105.
No. 5
minutes before the infusions and continued for 130 minutes. On separate days secretagogues were administered with the gastric cannula open. Gastric secretions were collected continuously for 120 minutes and divided into lo-minute aliquots, and the volume and hydrogen ion concentration of each aliquot were measured. Blood samples were obtained at 20 and 10 minutes before, at 0 minute (immediately before the start of the infusion), and at lo-minute intervals thereafter. Intestinal perfusions of a liquid protein meal and fat emulsion. The intestinal phase of SLI release was investigated on separate occasions by the intraduodenal perfusions of a defined formula liquid protein meal (Ensure Plus; Ross Laboratories, Montreal, Quebec, Canada; with a total caloric content of 1 kcal/mL distributed as 15% protein, 53% carbohydrate, and 32% fat with a pH of 7) at a rate of 2.5 mL/min and by intraduodenal perfusion of Intralipid 10% (Kabi Vitrium; Baxter Pharmaceutical, Newmarket, Ontario, Canada) at a rate of 1.7 mL/min for 90 minutes, through an indwelling catheter placed 10 cm distally into the duodenal cannula located opposite the entrance of the main pancreatic duct. The gastric cannula was left open to exclude retrograde flow of nutrients into the stomach and for quantitation of acid secretion. Three blood samples were taken during a 20-minute basal period and then at lo-minute intervals throughout the perfusions. The intraduodenal perfusion studies were performed on separate days without and with omeprazole at a dose and duration as described above. On separate days, perfusion of the intraduodenal protein meal was undertaken with the CCK receptor antagonist MK-329 (75 pg/kg) or its vehicle (2 mL of 100% dimethylsulfoxide in 20 mL of 0.9% saline containing 0.1% bovine serum albumin), administered by bolus intravenous injection 10 minutes before the start of the perfusion. MK-329, previously designated L-364,718, was a generous gift from Dr. D. Veber (Merck Sharpe & Dohme, West Point, PA). Blood sampling and gastric juice collections were performed as described above.
Laboratory
Methods
Blood for SLI determinations was collected in tubes containing 1000 kallikrein inhibitor units of aprotinin (Trasylol; Bayer AG, Germany) and 1.2 mg ethylenediaminetetraacetic acid Na/mL of blood. Samples were rapidly centrifuged, and 1 mol/L HCl was added to a known volume of plasma to adjust the pH to 3 and, thus, inactivate serine proteases and aminopeptidases.” The plasma was stored at -20°C until assayed. Plasma SLI concentrations were measured by radioimmunoassay, as described previ0us1y.~ Before radioimmunoassay, SLI was extracted from plasma with ethanol (3.6 mL 95% ethanol to 1 mL plasma). Plasma and ethanol were carefully mixed and centrifuged at 28OOg for 40 minutes; the supernatants were decanted and evaporated to dryness with compressed air overnight. After ethanol extraction of acidified plasma, the mean recoveries
November 1993
determined fmol
S-14,
in 40 sets of samples
synthetic
3%, respectively. plasma,
containing
S-14 and S-28 were In contrast,
without
the mean recovery
prior
volume.
Extracted
plasma
in assay buffer
with
an
samples
and
the S-14 molecule.
antibody
The
was collectively analysis
S-14 standards, (50% inhibitory mean
intra-assay
interassay
variation
The limit
Hydrogen (autoburette
titrator
NaOH.
on plasma
were collected then
was acidified
extracted
silyl silica Milford,
from
above
alcohol, (2.5-mL
trifluoroacetic
lo-mL
aliquot)
by
5 mL
Two plasma
aliquots air-dried,
alcohol-l%
so treated
samples and the
Associates, Acidified at
acid.
(5 mL of plasma in 2 mL column
buffer. This extraction technique provided recoveries of synthetic S-28 and S-l 4, added to acidified plasma, of 93% + 2% (n = 6) and 87% f 2% (n = 6), respectively.
hormone
the mean
responses
Statistical
comparisons
using Student’s
Reconstituted
component
95%. The mean
to the
column
was
as the mean response
basal concentration during
for each dog
as described
of means within of variance
Integrated previously.3
a group
paired t test, and differences
or the
calculated
each time period.
were calculated
by analysis
+ SEM
in plasma,
were made
between
followed
groups
by a multiple
test. P value I 0.05 was considered
significant.
Results Solid Meal SLI Responses Suppression Basal
plasma
SLI levels
to Acid
of 12 + 2 fmol/mL
were not significantly influenced by administration of omeprazole (9 * 2 fmol/mL) or ranitidine (11 k 2 fmol/mL).
The
effect
of acid
suppression
prandial
SLI responses
ingestion
of the solid meal, plasma
Administration duction
were eluted
trifluoroacetic
exceeded
SLI added
are expressed
to a peak increment
5 mL
the Sep-Pak
peptides
this
S-14 or S-28 added to
Analysis
hormone
meal
alone.
with
water.
were pooled
and reconstituted
Results incremental
and
The cartridge was then sequenvolumes of water and with 0.1%
acid, and the adsorbed
equivalent),
Waters
the
(S-l 3), the chromatographic
plasma
Statistical
gastrin,
on octadecyl
was passed through
5 mL 80% isopropyl
in extracted
of total
comparisons
forms of SLI were
was washed
1.02. Although
not differentiate plasma
omeprazole
and concentrated
followed
does
the column
of
and centrifuged,
(Sep-Pak;
used
of synthetic
were also
blood
S-14,
from S-14. The recovery
were tested
protein
volume.
the Kav for synthetic
S-14, may have been the active circulating
[Des-Ala’lsomatostatin
of the liquid
of the liquid
cartridge
a flow rate of 1 mL/min. tially washed with 5-mL with
un-
or ranitidine;
perfusions
to pH 3. The peptide
A C-18
fraction,
methodology
60 and 90
and with the vehicle
plasma
form
mean
40 and
of histamine,
with and without
(C- 18) cartridges MA).
isopropyl plasma
fasting,
at baseline;
analysis,
as indicated
S-28 was 0.68 and for synthetic smallest
was
were
forms
omeprazole
in its vehicle
conditions,
these experimental
from the concentration
during
taken
perfusion
chromatographic
and I/t the total column
Under
by subtracting
studies
the infusions
meal and fat emulsion
with MK-329
plasma
acid output
aliquot.
after the intraduodenal
and the intraduodenal
ume, I/a the void volume,
ali-
ion concentration
SLI molecular
with and without
60 minutes
For
Gastric
obtained
samples
after initiating
or CCK-8 protein
samples
were applied to a 9 X lOOO-mm Sephadex G-50 (superfine) column calibrated with synthetic cyclic S-14 (Peninsula Laboratories), synthetic S-28 (Bachem Inc.), dextran blue (void volume), cytochrome c (mol wt, 12,384), and “‘I-sodium (total volume). Elution was undertaken with 125 mmol/L ammonium bicarbonate buffer (pH 9) containing 100 mmol/L sodium chloride and 0.1% bovine serum albumin at 4°C. Fractions were collected at a flow rate of 6 mL/h and assayed. Elution positions are expressed as the coefficient of distribution Kav where Kav = (Ve Vo)/(Vt - Vo), with I/e corresponding to the elution vol-
Copenha-
each lo-minute
chromatographic
and omeprazole.
minutes
Radiometer,
from
1389
92% + 3% (n = 48).
by titration
after the meal alone, and after administration
ranitidine measured
the mean
Chromatography
Gel permeation 90 minutes
and
was determined
of each lo-minute
on plasma
was 9.5 fmol/tube.
the hydrogen
Gel Permeation dertaken
of the
AND GASTRIC ACID
samples
recovery
samples
by multiplying
by the volume
S-28 and
of the assay
was 4.3%,
and pH meter;
of 2-mL
quot to pH 7 using 0.025N calculated
from the chro-
was 7.5%.
ion concentration
gen, Denmark)
forms
of detection
[IC,,])
variation
part of
a synthetic
synthetic
and the sensitivity
concentration
antibody
S-28 on an
against
quantitated
respectively.
an
all circulating
read
were read against
assay was 0.3 fmol/tube,
in duplicate
and
also detects
basis. Total SLI, comprising
matographic
plasma
against the central
S- 14; S-28 and S-14 components
The
at 4°C
tracer
directed
S-14 or
of the initial
incubated
(1: 130,000 final dilution)
of somatostatin,
of
of 200 FL were reconsti-
[‘251-Tyr”]somatostatin
equimolar
of total plasma
was independent
tuted
acidification
for S-14 was 62% + 2% and that
for S-28 was 37% + 4%. The recovery S-28 after extraction
20, 40, or 80
78% + 2% and 76% t
S-28,
is shown
in Figure
by 44% of postprandial
After
SLI levels increased
of 17 + 2 fmol/mL
of omeprazole
on postl/l.
caused
(P < 0.01). a sustained
elevations
re-
of plasma
SLI; a similar effect occurred after ranitidine. The integrated SLI responses to the meal after omeprazole and ranitidine were not different, and both responses were significantly (both P < 0.025) value (Figure 1B).
below
the control
Gel permeation chromatography indicated that plasma SLI eluted in three peaks that coeluted with synthetic S-14 (Kav 1.02) and synthetic S-28 (Kav 0.68) and a third peak (Kav 0.15) with a molecular
1390
=
gg
GASTROENTEROLOGY Vol. 105, No. 5
GREENBERG ET AL.
25 -Meal 4
(both P < 0.001 compared
A
In contrast
20-
f
15-
: E I h
lo-
Q
histamine
were
suppressed (Figure
5o-
0
20
40
60
80
100
120
140
160
180
Time (minutes) r
-
7
Sz
acid,
2.0
plasma minutes
-
1.5
-
1.0
-
0.5
-
E E
0.0
-
Intravenous
without (Figure
Meal
min,
col-
gastric
and plasma
were not different of gastrin
of gastric
(P < 0.001) at 90
2 fmol/mL
3B). Plasma
ad-
acid increased
SLI responses
to exoge-
nous gastrin
also were composed
entirely
of S-14 and
were reduced
after omeprazole
by 90% f
5% (Figure
by 87% f 8% (both P < 0.001).
experiments
tion, exogenous
Ornep&~le
increased
infusions
collection
SLI by 17 f
In separate
Meal
histamine
with
values.
4B) and after ranitidine
P Z b
by 87% -t 5% (P < performed
(4.1 f 0.3 fmol/mL)
baseline
to
of S-14 and were
to 15.6 ? 1.4 mmol/30
S- 14 responses
ministered
5 E ,P i fn
lections
Gastrin.
T
2.5
.S E
experiments
of gastric
3A).
by 89% + 6% (P < 0.001)
and after ranitidine
acid output
(Figure
SLI responses
entirely
after omeprazole
4A)
B
E
composed
0.001). In separate
from
3.0
with controls)
to the solid meal, plasma
performed
gastrin
increased
with
acid collec-
gastric acid output
to
Meal Rarhditidine
Figure 1. (A) Mean + SE (fmol/mL) plasma SLI responses (A) after ingestion of a solid meal without (0) and with omeprazole (0) or ranitidine (A). Basal plasma SLI was subtracted from the total to obtain incremental plasma SLI. (6) Integrated release of SLI in response to ingestion of a solid meal alone and after omeprazole and ranitidine. *P < 0.025 vs. meal alone.
weight
of - 13,000
(Figure
U).
Basal levels
and S-14 and the large-molecular-weight
peak
4.1 rf: 0.6, 3.2 f 0.4, and 4.3 f 0.3 fmol/mL, tively. After ingestion plasma
S-28 from
0.8 fmol/mL S-14 from
of the meal, omeprazole
an increment
or ranitidine
2B). Ranitidine increments
(Figure
Infusions of Histamine,
6 r Solid Meal + Omeprazole
Tr,
B
respecreduced
6 r Solid Meal + Ranitidine
T
(P < 0.01)
caused similar of plasma
S-28
by 54% + 11% (P < 0.05) and S-14 by 63% + 8% (P < of the large-molecu0.01) (Figure 2C). Th e magnitude lar-weight peak was not influenced by ingesti n of the solid meal and was not changed by administration of omeprazole
-r
were
(P < 0.01) or by 67% + 8% and plasma
of postprandial
t
of 7.7 + 1.4 to 2.6 +
8.2 + 1.1 to 3.7 f 0.9 fmol/mL
or by 56% + 9% (Figure reductions
of S-28
2).
Gastrin, and CCK
Histamine. Intravenous infusions of histamine, undertaken without collection of gastric acid, caused a progressive increase in plasma SLI to a peak increment of 34 + 6 fmol/mL (P < 0.001) at 90 minutes. After omeprazole or ranitidine treatments, plasma SLI levels were reduced to values that were equivalent to baseline
20
25
30
35
40
Effluent
45
50
volume
55
60
65
70
75
(ml)
Figure 2. Characterization by gel chromatography of 5-mL plasma samples for SLI undertaken on a 9 x lOOO-mm Sephadex G-50 (superfine) column eluted with 125 mmol/L ammonium bicarbonate buffer, pH 9, and calibrated with dextran blue (Vo), cytochrome c (CC), S-28 S- 14, and ‘251-sodium as total volume marker. Entire fractions (0.6 ‘mL each) were assayed for SLI content. Points represent means + SEM of four experiments. (A) Plasma obtained before (0) and 40 minutes after (0) ingestion of a solid meal. (6) Plasma obtained 40 minutes after ingestion of a solid meal and treatment with omeprazole. (C) Plasma obtained 40 minutes after ingestion of a solid meal and treatment with ranitidine.
S- 14, S-28, AND GASTRIC ACID
November 1993
C
CCK- Of’ (250
1391
pmol ko”h”l
I
4
35 -
= 5E
30 -
zE
25 -
2
20 -
tZ
15-
Figure 3. Mean + SE (fmol/ mL) plasma SLI responses (A) without (0) and with (0) administration of omeprazole or ranitidine (A) after infusions of (A) histamine, (13) gastrin, and (C) CCK8.
a
10 -
9.2 f. 1.5 mmol/30
min,
and plasma
(3.9 +- 0.4 fmol/mL)
were not different
90
0
30
S-14 responses from baseline
omeprazole
(2.9 + 1.4 fmol/mL),
did not reach
CCK-8.
Intravenous
minutes,
and similar
Omeprazole
and
of both
respectively),
reduced
plasma
3C). In contrast
gastrin,
CCK-8
sustained.
caused
plasma
of
eleva-
(Figure
4C). After
increment
that of
omeprazole
of S-28 was reduced
1.9 fmol/mL
58% f 14%, but the incremental
(P < 0.01)
S-14 increase
was not significantly of 6.9 + 1.4 fmol/mL
different (Figure
with
a liquid
protein
by 23 f 5 fmol/mL prazole from
caused
their meal
sustained
40 to 90 minutes
plasma
S-28
meal increased
(P < 0.001)
and
S-14
reductions
the increase
(P < 0.01) (Figure increased
values after the intraduodenal (P < 0.001 and P < 0.05, respectively)
6A), although the increase s-14 (P < 0.05), comprising After
administration
influence
5B) or the plasma
the
elevations 6B).
from
ies, pretreatment meal
with
duodeSLI omeSLI from protein (Figure
the in-
crement of plasma S-28 was decreased from 10.8 f 1.3 to 1.2 f 0.4 fmol/mL (P < 0.001) or by 87% f 5%; the incremental plasma S-14 response with omeprazole (5.4 ? 2.2 fmol/mL) tended to be higher than without
of the intraduodenal
the CCK-A
abolished 0.001)
of S-28 exceeded that of 71% f 4% of the total SLI of omeprazole,
(Figure
did not
stud-
5A). Both
significantly
of omeprazole
with control
Nutrients
baseline
released.
(P < 0.01) and S-14 to (P < 0.001) (Figure 6B), although of S-14 exceeded that of S-28 (P < 0.05).
Effect of MK-329. Compared
4C).
of plasma
to
of both
treat-
of 7.8 +
plasma
increments
fat (Figure
or by
at 40 minutes;
plasma
of S-28 and S-14 after intraduodenal
from con-
of the
at 60
5B). The SLI responses
14.6 +- 2.5 fmol/mL
increases
(Figure
peak output traduodenal blockade
meal
puts during
further
with plasma
fat,
7B). After
CCK-A
increases
with controls)
from baseline; the third
period
from
a
of 3.4 -t 1.7 to 8.2 +-
however,
in-
receptor
of S-28 (1.3 +
and S-14 (2.3 + 0.1 fmol/mL)
P < 0.001 compared not different
period
(P < 0.025) (Figure
also reduced
protein MK-329
the gastric acid output
was increased
in the third min
liquid
antagonist
S-28 and S-14 (both P <
of plasma
perfusion
0.2 fmol/mL)
receptor
7A). C onversely,
the protein
2.3 mmol/30
Liquid protein meal. Perfusion
GA). a fat emul-
S-28 to 6.2 + 1.1 fmol/mL
after
SLI Responses to lntraduodenal
with
SLI by 26 + 7 fmol/mL
significant
total SLI increase
14.7 + 0.8 to 6.1 f
num
plasma
but the increase
of S-28 exceeded
perfusion
(P < 0.01) (Figure
minutes
fat comprised
SLI by 12% + to infusions
sion increased
Administration
the plasma
2.3 fmol/mL
ele-
S-28 and S-14 (P < 0.001 and P < 0.01,
S-14 (P < 0.05)
trol values
of CCK-8
(P < 0.001) at 20
levels were thereafter
marginally
3% (P = NS) (Figure histamine
infusion
SLI by 21 f 3 fmol/mL
but this difference
(P = 0.2) (Figure
significance
Fat. Intraduodenal
vated plasma
ment,
90
Time (minutes)
values.
tions
60
(both
to values that were gastric
with MK-329
acid out(0.9 z!z 0.4
mmol/30 min) and without MK-329 (0.6 f mmol/30 min) were not different from that during basal period (0.3 + 0.2 mmol/30 min).
0.3 the
Diiscussion The present study shows that in conscious dogs, gastric acid secretion stimulated by ingestion of a meal regulates the release of both S-28 and S-14 into the circulation. Administration of omeprazole or raniti-
1392
GREENBERG
GASTROENTEROLOGY
ET AL.
Histamine 1501.10kg.‘h.‘l vo cc 18
-
sion after omeprazole
4
16 -
was similar
to the
SLI reduction
after cimetidine
treat-
ment
in humans.
l5 Our
differ,
other
studies
omeprazole,
12
-
These divergent
results probably
cal differences,
because
reduced
dose intravenous 8-
a higher
6-
tion
4-
Further, bition
0t
2 B E r z
-
Gastrin (150 pmol ka”h”)
to suppression
S-14 and S-28.
6-
heterogeneous
suggestions’5
and
that inhi-
is indirect
and
acid secretion.
indicated
that in the fed secretion
SLI molecular
of both forms
are
in several species,7*12,21,22but our equiva-
with
plasma results
elevations
of other
of S-14 and S-28
studies
with
dogs21,22 in
which nutrients required adjustment to pH 3 for the measurement of circulating increases of S-14 and
O12
acidity.*’
to omeprazole
in part,
Circulating
SLI secre-
in gastric
of gastric analysis
lent postprandial contrast
stimulates
antagonists
acid mediated,
8-
2-
with
accord with previous
state gastric
4-
single-
SLI responses
Chromatographic 10
study,i4
was coadministered
dose that
of SLI by H,-receptor
SLI.
relate to methodologi-
of changes
the parallel
from but not
plasma
in the latter
omeprazole
pentagastrin
related
however,
ranitidine,
nutrient-stimulated
independently
ranitidine
2
Z
results
in dogs14 in which
-
-
No. 5
45% postprandial
14
10
.o z
or ranitidine
Vol. 105.
r
C
CCK-OP (250 pmd kg”h”)
10 -
without
detectable
rapidly
degraded
prompt
increases in plasma;
acidification
of S-28. However, recent
of samples
8-
adequate
recoveries
6-
humans,’
the total SLI increase
cular-weight 4-
This
2-
peak
ations
of gastric
20 25 30 35 40 45 50 55 60 65 70 75 Effluent
tion
is a prerequisite
for
of S-28.‘9,23 Similar
stimulates
influenced
to findings
included
in
a large-mole-
prosomatostatin.
by nutrients
or alter-
acid secretion.
The mechanisms
O-
that
peak that may represent was not
S-28 is
studies indicate
by which
release
postprandial
acid secre-
of SLI have not been clarified.
volume (ml) A
Figure 4. Characterization by gel chromatography of 5-mL plasma samples for SLI, using the column conditions described in Figure 2. Points represent means + SE of four separate experiments. (A) Plasma obtained 60 minutes after intravenous infusions of histamine without (0) and with (0) omeprazole. (6) Plasma obtained 60 minutes after intravenous infusions of gastrin without (0) and with (0) omeprazole. (C) Plasma obtained 60 minutes after infusions of CCKS without (@) and with (0) omeprazole.
dine caused equivalent suppression of postprandial plasma elevations of S-28 and S-14, but the mechanism by which gastric acidity mediated secretion of S-14 was distinct from that of S-28. Increased intraluminal gastric acidity selectively stimulated S-14 secretion, whereas the acid-dependent component of S-28 secretion required intestinal release of CCK as a cofactor. Previous studies showed increases of plasma SLI levels in the fed state without consensus on the influence of SLI suppresof gastric acidity. 14*15The magnitude
40
Intraduodmal Protein Meal i
r
0
lntraduodrnal Fat
(2.6 ml/ mid
30
60
(1.7 ml/ mid
90
0
30
50
1
90
Time (minutes) 5. Mean + SE (fmol/mL) plasma SLI responses (A) after (A) intraduodenal perfusions of liquid protein meal without (0) and with (0) omeprazole and after (13)intraduodenal perfusions of a fat emulsion without (0) and with (0) omeprazole.
Figure
November 1993
S-14,
Intraduodwd Protein Meal 12 -
VO
cc
1
I
vitro
studies
minal from
10 -
gastric
similar ach,”
4-
Stimulation
t; t” . 5 E
c i
co
18
the specificity
for stimulation The pattern
histamine,
of the
of S-14
secretion
of SLI forms
of the isolated
vagus
nerves
the fed state gastric
acid secretion,
by vagal, hormonal,
or paracrine mechanism
form
gastric
released of HCI is
perfused
also
1393
of intralu-
or the instillation
S-14 via increased
be a predominant
B
Intraduodenal Fat
r 16 -
AND GASTRIC ACID
and S-14 is the only molecular
stimulates
E
support
for both regions
6-
._
acid
the stomach.
after gastrin,
8-
further
S-28,
stom-
secreted.12
preferentially
acidity.23 Thus, whether
pathways,
in
mediated appears
to
for release of S-14 from
the stomach. In contrast
to S-14, increasing
14 -
alone could
12 -
dial S-28 release. evidence
10 -
not account Rather,
for a mechanism
lates release
intraluminal
acidity
for acid-dependent the present involving
of SLI from both
postpran-
studies
provide
CCK. CCK stimu-
fundic28
and antra12” D
86-
2O-
A
Intraduodenal Protein Meal
12 r
4=
,
81
I
I
81
I
I
I
a
1
20 25 30 35 40 45 50 55 60 65 70 75 Effluent
volume
(ml)
Figure 6. Characterization by gel chromatography of 5-mL plasma samples for SLI, using the column conditions described in Figure 2. Points represent means + SE of four separate experiments. (A) Plasma obtained 60 minutes after intraduodenal perfusions with a liquid protein meal without (0) and with (0) omeprazole. (3) Plasma obtained 60 minutes after intraduodenal perfusions with a fat emulsion without (0) and with (0) omeprazole.
.; 0 t t 5 =E i oY
VO
cc
1
I
s29
514
4
1
lo86 42-
O-
I1
I,
20 25 30 35 40 45 50 55 60 65 70 75 Effluent volume
(ml)
B
Intraduodenal Protein Meal
I
Studies
using
indicated histamine-
isolated
perfused
stomach
(2.5 mlhnin)
I T
preparations
a direct effect by acid on D cells, because and gastrin-induced luminal secretion of
SLI could be largely accounted for by the increases of intraluminal acidity. l1 Our results show that histamine- and gastrin-stimulated acid secretion potently stimulated
SLI secretion
into the circulation,
but SLI
responses were composed entirely of S-14. Pharmacological suppression of gastric acid abolished increases of S- 14 after both secretagogues, implying that intraluminal gastric acid caused direct and selective release of S-14. S-14 is present in both stomach and pancreas, but the stomach was the likely source of S-14, because the pancreas makes only a minor contribution to circulating SLI. 24 Although morphological differences2’ and SLI molecular heterogeneitya6,27 between D cells in the fundus (closed-type secreting S-28 and S-14) and antrum (open-type secreting S- 14) are well-described, in
9 Y
2-
01,
I
MK-329 (75 u9kgI or v&de
Basal
1 30 min Periods
2
3
Figure 7. (A) Characterization by gel chromatography of 5-mL plasma samples for SLI, using the column conditions described in Figure 2. Plasma was obtained 60 minutes after intraduodenal perfusions of a liquid protein meal with intravenous MK-329 (0) or its vehicle (0). (6) Mean + SE gastric acid output (mmol/30 min) from basal state and during stimulation by an intraduodenal liquid protein meal with intravenous MK-329 (0) or its vehicle (0).
GREENBERG ET AL.
1394
cells in culture
GASTROENTEROLOGY Vol. 105, No. 5
and from the same regions
3o by activation
stomach,
of specific
ceptors. 31 In conscious stimulation gastrin
dogs,
of basal gastric
in the stimulation
lated
CCK
inhibit
also causes
re-
weak
acid32 but is equipotent of acid production
to
from iso-
cells. 33 These observations lead to the that CCK in the intact dog releases SLI to
parietal
SLI released
cell function.
Our results
after CCK is composed
S-28 and that the S-28 response, gastric
acid.
The
low
after CCK compared accounted release
of S-28
close proximity lumen.
with gastrin
only closed-type are closely intact
CCK
aligned
with
instillation
separate
production releases further
tained
with
fundus
that in the in re-
S-28 in a local
fundic
by which
D
CCK
Protein
and fat both protein
CCK-A prandial
because
receptor total
recent
antagonist
SLI and causes
studies MK-329
CCK effects on acid secretion indicate function.
preferential
mediates
effects
ach, but the mechanism release
CCK,‘6,36
feedback
gastric also
S-28 release.
have shown suppresses
the post-
increase
of
begastric acid. l6 Our results extend these findings cause MK-329 abolished S-28 increases after intraduodenal protein and further increased gastric acid secretion, implying a specific reciprocal functional relationship between gastric acid and S-28 in the fed state mediated by CCK. The concept that S-28 selectively regulates CCK actions is supported by observations in vitro derived from rat pancreatic acinar plasma membranes 38, however, proof for this hypothesis in vivo awaits the development of selective S-28 antagonists. Although CCK acts via the circulation to release S-28, it is likely that S-28 regulates gastric acid by paracrine mechanisms. Inhibition of parietal cell
nisms
only
by CCK-A
on parietal
cell
that gastric
acid
gastric
and requires Models
between
humoral nega-
somatostatin
and
should
in-
the secretion
of
D cells in the fundus receptor-dependent inhibition
acid stimu-
describing
to nutrients
also S-28,
d e 1ivery to parietal
may
of S-14 from the stom-
responses
S-14 but
with consequent
a paracrine
by S-28
by which
regulation
acid secretory
The
mediating
in rats and humans
is indirect
of CCK as a cofactor.
is mediated
pathway
the secretion
lates S-28 secretion
MK-329
fat but did
in acid secretion.
this study has shown
directly
CCK was obafter intes-
increase
SLI-dependent
from closed-type
protein
a reciprocal
an
suggests
in dogs is un-
by S-14 because
after intraduodenal
a reciprocal of
which
These observations raise the issue of a role for CCKmediated S-28 secretion in the regulation of gastric acid secretion,
so without presence
S-28 secre-
stimulates
acid-dependent
S-14 release
not
release
acid.37 Our data show that intraduodenal stimulates
abolished
in
in dogs
study
effect of CCK
mediated
of gastric pathway
mechanism
‘6*44The present
the enterogastrone
likely to be entirely
suppression
an SLI-dependent
controversial.
to specific
CCK in the modula-
but a similar
clude
of SLI forms
intraduodenal
remains
gastric released
fat-induced
through
of
reflecting
responses
studies implicate
and
ingestion
only S-28,4’ perhaps
in SLI form
Recent
of basal
tive
acid environment
for acid-dependent
by endogenously
nutrients.
con-
fat in dogs caused
of S-14, whereas
releases
differences
In summary,
but not to intralu-
mechanisms
characterization
preferentially
in
and body
occurs
species
that
that
via the
cells25; also,
the fundus
increase
plasma
acid secretion.40
of intraduodenal
a predominant fat in humans
to S-14 in vitro,39 but
physiological
do not influence
Administration
rats42 and humans,43
investigation.
nutrients. only
acting
centrations
acid secretion
secreted
of S-28 and S-14 from
support
tion mediated
than
of S-28 to achieve
tion
likely
by observations
parietal from
cells,35 the intracellular
Additional
acid
by S-28 is equipotent
infusions
of
achieved
suggest
function
by
of HCl. 34 Although there are two processing pathways for the
posttranslational
requires
data
requires
acid secretion
different
selectively
these
D cells in the gastric
to stimulated
minal
acidity
or histamine16
is supported
dog, SLI release
sponse
predominantly
to D cells rather
This notion
that
of an effect by omeprazole
Together, after
indicate
in part, is mediated
intraluminal
for the absence
on S-14 secretion.
but
CCK
parietal
hypothesis
tinal
of the intact
type-A
and body mecha-
of acid secretion
by
cells.
References 1. Schusdziarra V, Rouiller D, Harris V, Unger RH. Gastric and pancreatic release of somatostatin-like immunoreactivity during the gastric phase of a meal: effects of truncal vagotomy and atropine in the anesthetized dog. Diabetes 1979;28:658-663. 2. Hollinshead JW, Debas HT, Yamada T, Elashoff J, Osadchey B, Walsh JH. Hypergastrinemia develops within 24 hours of truncal vagotomy in dogs. Gastroenterology 1985;88:35-40. 3. Greenberg GR. Influence of vagal integrity on gastrin and somatostatin release in dogs. Gastroenterology 1987;92:994100 1. 4. Lucey MR, Wass JAH, Fairclough P, Webb J, Webb S, Medbak S, Rees LH. Autonomic regulation of postprandial plasma somatostatin, gastrin and insulin. Gut 1985;26:683-688. 5. Polonsky KS, Shoelson SE, Docherty HM. Plasma somatostatin28 increases in response to feeding in man. J Clin Invest 1983;71:1514-1518. 6. O’Shaughnessy DJ, Long RG, Adrian TE, Christofides ND, Ghatei MA, Sarson DL, Bloom SR. Somatostatin-14 modulates postprandial glucose levels and release of gastrointestinal and pancreatic hormones. Digestion 1985;3 1:234-242. 7. Ensinck JW, Laschansky EC, Vogel RE, Simonowitz DA, Roos BA, Francis BH. Circulating prosomatostatin-derived peptides. Differential responses to food ingestion. J Clin Invest 1989;83: 15801589. 8. Lucey MR. Fairclough PD, Wass JAH, Kwasowski P, Medbak S,
November 1993
S-14,
Webb J, Rees LH. Response of circulating somatostatin, insulin, gastrin and GIP to intraduodenal infusion of nutrients in normal man. Clin Endocrinol (Oxf) 1984;21:209-217. 9. Schusdziarra V, Rouiller D, Pietri A. Harris V, Zyznar E, Conlon JM, Unger RH. Pancreatic and gastric release of somatostatin-like immunoreactivity during intestinal phase of a meal. Am J Physiol 1979;237:E555-E560. IO. Holst JJ, Jensen SL, Knuhtsen S, Nielsen OV, Rehfeld JF. Effect of vagus, gastric inhibitory polypeptide, and HCI on gastrin and somatostatin release from perfused pig antrum. Am J Physiol 1983;244:G5 15-G522.
28.
29.
30.
11. Schubert ML, Edwards NF, Makhlouf GM. Regulation of gastric somatostatin secretion in the mouse by luminal acidity: a local feedback mechanism. Gastroenterology 1988;94:3 17-322. 12. Baldissera FGA, Nielsen OV, Holst JJ. The intestinal mucosa preferentially releases somatostatin-28 in pigs. Regul Pept 1985;11:251-262.
31.
13. Schusdziarra V, Stepelfeldt W, Klein M, Maier V. Role of histamine H, receptor in the modulation of pancreatic and gastric function. Acta Endocrinol (Copenh) 1982;99(Suppl 246): 11 1.
32.
14. Simoens C, Woussen-Colle MC, DeGraef J. Effect of cimetidine, ranitidine and omeprazole on postprandial gastrin and somatostatin release in dogs. Regul Pept 1988;22:285-293. 15. Lucey MR, Wass JAH, Rees LH, Dawson AM, Fairclough PD. Relationship between gastric acid and elevated plasma somatostatin-like immunoreactivity after a mixed meal. Gastroenterology 1989;97:967-872. 16. Konturek SJ, Bilski J, Tasler J. Ciezkowski M. Role of cholecystokinin in the inhibition of gastric acid in dogs. J Physiol 1992;45 1:477-489. 17. Greenberg GR. Differential neural regulation ofcirculatingsomatostatinand -28 in conscious dogs. Am J Physiol 1993;264:G902-G909. 18. Ryberg B, Larsson H, Carlsson E. Gastric acid secretion and plasma gastrin levels in dogs treated with high oral doses of ranitidme and omeprazole. Stand J Gastroenterol 1986;21:9495. 19. Francis BH, Ensinck JW. Differential alterations of the circulating prosomatostatin-derived peptides during insulin-induced hypoglycemia in man. J Clin Endocrinol Metab 1987;65:880-884. 20. Shulkes A, Read M. Regulation of somatostatin secretion by gastrin- and acid-dependent mechanisms. Endocrinology 199 1; 129:2329-2334. 21. Schusdziarra V, Zynar E, Rouiller D, Harris V, Unger RH. Free somatostatin in the circulation: amounts and molecular sizes of somatostatin-like immunoreactivity in portal, aortic and vena caval plasma of fasting and meal-stimulated dogs. Endocrinology 1980;107:1572-1576. 22.
Klaff LJ, Dunmng BE, Taborsky GJ. Somatostatin-28 does not regulate islet function in the dog. Endocrinology 1988;123: 2668-2674.
23. Greenberg GR, Pokol-Daniel S, Fung L. Influence of gastric acid on circulating somatostatin-14 and -28 released after insulin-induced hypoglycemia in conscious dogs. Endocrinology 1992; 13 1: 1527- 1533. 24. Taborsky GJ, Ensinck JW. Contribution of the pancreas to circulating somatostatin-like immunoreactivity in the normal dog. J Clin Invest 1984;73:2 16-223. 25. Larsson LI, Golterman H, De Magistris L, Rehfeld JF, Schwartz TW. Somatostatin cell processes as pathways for paracrine secretion. Science 1979;205: 1393- 1395. 26. Penman E, Wass JAH. Butler MG, Penny ES, Price J, Wu P, Rees LH. Distribution and characterization of immunoreactive somatostatin in human gastrointestinal tract. Regul Pept 1983;7:5365. 27. Baldissera FGA, Host JH, Jensen SL, Krarup T. Distribution and
S-28,
AND GASTRIC ACID
1395
molecular forms of peptides containing somatostatin in immunodeterminants in extracts from the entire gastrointestinal tract of man and pig. Biochem Biophys Acta 1985;838: 132- 143. Sol1 AH, Amirian DA, Park J, Elashoff JD, Yamada T. Cholecystokinin potently releases somatostatin from canine fundic mucosal cells in short-term culture. Am J Physiol 1985;248:G569-G573. Buchan AMJ, Curtis SB, Meloche RM. Release of somatostatin immunoreactivity from human antral D-cells in culture. Gastroenterology 1990;99:690-696. Roullier D, Schusdziarra V, Harris V, Unger RH. Release of pancreatic and gastric somatostatin-like immunoreactivity in response to the octapeptide of cholecystokinin, secretin, gastric inhibitory peptide and gastrin-17 in dogs. Endocrinology 1980; 107:524529. Buchan AMJ, Meloche RM, Kwok YN, Kofed H. Effect ofcholecystokinin and secretin on somatostatin release from cultured antral cells. Gastroenterology 1993; 104: 14 14- 14 19. Mayer EA, Elashoff J, Mutt V, Walsh JH. Reassessment of gastric acid inhibition by cholecystokinin and gastric inhibitory polypep-
tide in dogs. Gastroenterology 1982;83: 1047- 1050. 33. Sol1 AH, Amiran DA, Thomas LP, Reedy JD, Elashoff JD. Gastrin receptors on isolated canine parietal cells. J Clin Invest 1984;73: 1434- 1447. 34. Schusdziarra V, Harris V, Conlon JM, Arimura A. Pancreatic and gastric somatostatin release in response to intragastric and intraduodenal nutrients and HCI in the dog. J Clin Invest 1978;62:509-5 18. 35. Chiba T, Park J, Yamada T. Biosynthesis of somatostatin in canine fundic D-cells. J Clin Invest 1988:81:282-287. 36. Lilja P, lnoue K, Wiener I, Fried GM, Greeley GH Jr, Thompson JC. Release of cholecystokinin in response to food and intraduodenal fat in pigs, and dogs and man. Surg Gynecol Obstet 1984; 15:557-56 1. 37. Konturek SJ, Konturek J, Cieszkowski M, Ebet-t R, Creutzfeldt W. Comparison of gastric inhibitory polypeptide and intraduodenal or intravenous fat on gastric acid secretion from vagally innervated and denervated canine stomach. Dig Dis Sci 1986;31:49-56. 38. Jouti-Tahiri N, Dufresne M, Viguerie N, Fourmy D, Esteve JD, Rivier J, Moroder L, Susini C, Vaysse N. Somatostatin 28 interacts with CCK receptors in brain and pancreas. Neuropeptides 199 1; 19:65-7 1. 39. Park J, Chiba T, Yamada T. Mechanisms for direct inhibition of canine gastric parietal cells by somatostatin. J Biol Chem 1987;262:14190-14196. 40. Seal A, Yamada T, Debas H, Hollinshead J, Osadchey B, Aponte G, Walsh J. Somatostatin-14 and -28: clearance and potency on gastric function in dogs. Am J Physiol 1982;243:G97-G 102. 41. Ensinck JW, Vogel RE, Laschansky EC, Francis BH. Effect of ingested carbohydrate, fat, and protein on the release of somatostatinin humans. Gastroenterology 1990;98:633-638. 42. Lloyd KCK, Raybould NE, Walsh JH. Cholecystokinin inhibits gastric acid secretion through type “A” cholecystokinin receptors and somatostatin in rats. Am J Physiol 1992;263:G287-G292. 43. Jebbink MCW, Lamers CBHW, Mooy DM, Rovati LC, Jansen JBMJ. Effect of loxiglumide on basal and gastrin- and bombesinstimulated gastric acid and serum gastrin levels. Gastroenterology 1992;103:1215-1220. 44. Lloyd KCK, Maxwell V, Ohning G, Walsh JH. Intestinal fat does not inhibit gastnc function through a hormonal somatostatin mechanism in dogs. Gastroenterology 1992; 103: 122 1- 1228.
Received March 15, 1993. Accepted May 4, 1993. Address requests for reprints to: Gordon R. Greenberg, M.D., Room 6356, Medical Sciences Building, University of Toronto, Toronto, Ontario, Canada M5S lA8. Supported by a grant from the Medical Research Council of Canada (MA6763).