Endocrinology Vol. 140, No. 4 1687-1694
Copyright © 1999 by The Endocrine Society
Role of the Vagus Nerve in Mediating Proximal Nutrient-Induced Glucagon-Like Peptide-1 Secretion1
A. S. Rocca and
P. L. Brubaker
Departments of Physiology and Medicine, University of Toronto,
Toronto, Ontario, Canada M5S 1A8
Address all correspondence and requests for reprints to: Dr. P. L. Brubaker, Room 3366, Medical Sciences Building, University of Toronto, Toronto, Ontario, Canada M5S 1A8. E-mail: p.brubaker{at}utoronto.ca
 |
Abstract
|
|---|
Plasma levels of glucagon-like peptide-1 (GLP-1) rise
rapidly after nutrient ingestion, suggesting the existence of a
proximal gut signal regulating GLP-1 release from the L
cells of the distal small intestine. Glucose-dependent insulinotropic
peptide (GIP) has been shown to be one such proximal signal; however,
the dependence of GIP on gastrin-releasing peptide, a neuromodulator,
suggested a role for the nervous system in this proximal-distal loop.
Investigations into the nature of this proximal signal were therefore
conducted in an in situ model of the rat
gastrointestinal system. Infusions of corn oil into a 10-cm segment of
duodenum that was isolated by loose ligation (to ensure that the
luminal contents did not progress to the ileal L cell) increased the
secretion of GLP-1 in parallel with that of gut
glucagon-like immunoreactivity (gGLI; r = 0.85;
P < 0.05). Infusion of fat into a transected
segment of duodenum also significantly raised gGLI secretion compared
with saline infusion, reaching a peak value of 132 ± 37 pg/ml
above basal (P < 0.05). However, peak secretion was
significantly delayed when the gut was transected compared with that
after ligation alone (19 ± 4 vs. 6 ± 1 min,
respectively; P < 0.05). Furthermore, bilateral
subdiaphragmatic vagotomy in conjunction with gut transection
completely abolished the fat-induced rise in gGLI secretion
(P < 0.001). Consistent with a role for the vagus
in the regulation of the L cell, stimulation of the distal end of the
celiac branch of the subdiaphragmatic vagus nerve significantly
stimulated the secretion of gGLI to a level of 71 ± 14 pg/ml
above basal (P < 0.05). As found previously,
supraphysiological infusion of GIP significantly increased gGLI
secretion in control animals by 123 ± 32 pg/ml
(P < 0.05); this was not prevented by hepatic
branch vagotomy (96 ± 25 pg/ml; P < 0.05).
In contrast, although infusion of GIP at physiological levels into
sham-vagotomized animals also increased gGLI secretion, by 40 ± 6
pg/ml (P < 0.05), selective hepatic branch
vagotomy abolished GIP-induced gGLI secretion (P <
0.05). The results of these experiments therefore demonstrate that the
secretion of GLP-1 and gGLI from the ileal L cell in
response to fat is regulated by a complex neuroendocrine loop,
involving the enteric nervous system, the afferent and efferent vagus
nerves, as well as the duodenal hormone GIP.
|
Introduction
|
|---|
PROGLUCAGON is cleaved by tissue-specific
prohormone convertases to the proglucagon-derived peptides (PGDPs) (1).
The major pancreatic product of proglucagon processing is glucagon,
whereas the intestinal products include glicentin, oxyntomodulin,
glucagon-like peptide-1 (GLP-1), and GLP-2. Although
GLP-1 possesses inhibitory action in terms of gastric
emptying (2), its main effect is its role as an incretin, stimulating
the secretion of insulin in a glucose-dependent manner (2, 3, 4). Other
metabolic effects of GLP-1 include suppression of glucagon
secretion (2), enhancement of glucose disposal (5), and inhibition of
central feeding behavior (6). These pleiotropic actions of
GLP-1, therefore, offer great potential for treatment of
the insulin resistance and relative insulin insufficiency that define
type II diabetes mellitus (7).
The release of GLP-1 and the other PGDPs from ileal L
cells is under complex regulation. Agents that can directly stimulate
the secretion of GLP-1 have been analyzed in various
models of the intestinal L cell, including a primary cell culture of
fetal rat intestinal cells (8, 9), a murine intestinal endocrine cell
line (10), and a perfused model of the rat ileum (11, 12). The major
secretagogues identified using these models include glucose-dependent
insulinotropic peptide (GIP), gastrin-releasing peptide (GRP),
calcitonin gene-related peptide, and agonists of acetylcholine. Fatty
acids, when applied directly to fetal rat intestinal cell cultures,
also stimulate GLP-1 secretion (9).
In the in vivo setting, GLP-1 is rapidly
released from ileal L cells upon the ingestion of a mixed meal (13, 14)
or by the infusion of fat directly into the duodenum (15, 16). Indeed,
plasma levels of GLP-1 peak within 5–15 min of nutrient
administration despite the fact that nutrients do not reach the ileum
within this time frame (17). Furthermore, in patients with ileostomies
to divert nutrient flow from the ileum, GLP-1 levels after
ingestion of a mixed meal are similar to those observed in individuals
with an intact gastrointestinal tract (13). These observations are not
consistent with the idea that nutrients act directly on ileal L cells
to stimulate the release of GLP-1, at least during the
early phase of secretion. Therefore, it has been postulated that the
secretion of GLP-1 is under the regulation of a
proximal-distal loop, relaying information about nutrient and, in
particular, fat ingestion from the proximal duodenum to the distal site
of GLP-1-containing L cells (15, 16). Consistent with this
concept, we have demonstrated that the duodenal peptide GIP, which is
secreted in response to fat ingestion (14, 16), is an endocrine
mediator of this proximal-distal loop in rats (16). However, recent
data have demonstrated that the secretion of GLP-1 induced
by duodenal fat can be completely inhibited by infusion of an
antagonist for the neuropeptide GRP (18). These findings suggest that
physiological doses of GIP act through the nervous system (either vagal
or myenteric) to indirectly stimulate GLP-1 secretion,
rather than acting directly at the level of the L cell.
The present investigation further examined the possible existence of a
neural component of the proximal-distal loop regulating secretion of
GLP-1 in the rat. Specifically, the intrinsic and
extrinsic nervous systems were analyzed with respect to their ability
to affect duodenal fat-induced GLP-1 secretion. In
addition, the humoral mechanism by which nutrients in the proximal
duodenum signal the distal L cell was reevaluated by examining the
potential interaction of GIP with the vagus nerve.
|
Materials and Methods
|
|---|
Surgery
After fasting for 18–24 h, male Wistar rats weighing 337
± 6 g (Charles River Canada, Inc., St. Constant, Canada)
were lightly anesthetized with halothane and given an ip injection of
sodium pentobarbital (60 mg/kg) to achieve a surgical plane of
anesthesia. The carotid artery was cannulated to allow for arterial
blood sampling. A maximum of nine samples were collected from each
experimental animal. Blood was collected into 10% (vol/vol)
Trasylol (5000 kalikrein inactivating units/ml;
Bayer, Inc., Etobicoke, Canada)-EDTA (12 mg/ml)-diprotin A
(34 µg/ml; Calbiochem, La Jolla, CA). Plasma was
collected and stored at -20 C until time of assay, and the RBC were
resuspended in an equal volume of heparinized saline and reinfused via
a cannula in the jugular vein. To ensure that the experimental
procedures were not stimulating a counterregulatory response,
corticosterone levels (Diagnostic Products Corp., Los
Angeles, CA) were detected in 50 µl plasma from all animals (n =
42) upon initiation (585 ± 25 ng/ml) and completion (552 ±
18 ng/ml) of blood sampling and were not significantly different.
Model of the proximal distal loop
To separate indirect from direct effects of nutrients on the
ileal L cell, the proximal-distal loop governing GLP-1
secretion was modeled by infusing nutrients into a 10-cm segment of
proximal duodenum (15, 16, 18). The duodenum was cannulated above the
common bile duct and isolated from the remaining gut by loose ligation
10 cm distal to the site of the duodenal catheter. This ligation
ensures that nutrients do not progress to the distal portions of the
small intestine to directly stimulate GLP-1 secretion from
the ileal L cells, but permits retention of all neurovascular
connections. Blood was collected at -10 and 0 min to ascertain basal
levels of secretion, and manipulation of the gastrointestinal tract was
performed in the interval between these two time points. At 0 min, 3–4
ml fat in the form of corn oil (56% polyunsaturated fatty acids, 32%
monounsaturated fatty acids, and 12% saturated fatty acids; Sunfresh
Ltd., Toronto, Canada) were infused into the proximal duodenal segment,
and blood samples were collected every 5–10 minutes thereafter up to
60 min.
Vagotomy
To ascertain the role of vagal innervation in the regulation of
GLP-1 secretion, a series of vagotomies was performed
before the start of experimental sampling in the interval between the
first and second basal blood samplings. The procedures included 1) left
cervical vagotomy, 2) bilateral subdiaphragmatic vagotomy, and 3)
selective hepatic branch vagotomy. The bilateral subdiaphragmatic
vagotomy was performed by transection of the esophagus, including the
accompanying vagal trunks.
Vagal stimulation
The vagal stimulation protocol was based on parameters used in a
similar study of vagally induced bombesin-like immunoreactivity release
(19). Briefly, the celiac branch of the subdiaphragmatic vagus nerve
was transected 30 min before the commencement of the experimental
procedure. During this interval the distal portion of the celiac branch
was placed over a bipolar stimulating electrode connected to a Grass
stimulator (Grass Instruments, Quincey, MA). This preparation was then
immersed in paraffin oil throughout the experiment. The electrical
stimulation was begun at 0 min and lasted for 15 min at a voltage of 10
V and a frequency of 20 Hz [preliminary studies with lower frequencies
did not significantly affect gut glucagon-like immunoreactivity (gGLI)
secretion]. Blood was collected at 5, 10, and 15 min during the
stimulation period and then every 10 min throughout the remainder of
the experimental time course.
GIP infusions
Human GIP (Bachem California, Inc., Torrance, CA)
was administered as a bolus via the jugular vein followed by a
maintenance infusion through the femoral vein for 40 min and then a
20-min recovery period. Blood samples were collected every 10 min. GIP
was infused at either a high (supraphysiological; 715 ng/kg bolus plus
1070 ng/kg·h infusion) or a low (physiological; 143 ng/kg bolus plus
214 ng/kg·h infusion) dose. The physiological dose of GIP was based
on that used in a previous study, which was found to mimic levels of
GIP produced by the ingestion of a fat meal (16). The
supraphysiological dose of GIP was 5 times greater than the
physiological dose.
Assays
In correlation experiments, RIA for GLI
(glicentin/oxyntomodulin/glucagon) was carried out using 0.1 ml plasma
with antiserum K4023 (Biospacific, Emeryville, CA), whereas RIA for
immunoreactive glucagon (glucagon) used 0.2 ml plasma with antiserum
04A (Dr. R. H. Unger, Dallas, TX). The plasma levels of gGLI were
obtained by subtraction of immunoreactive glucagon from GLI (15, 16, 18). Immunoreactive GLP-1(x-36NH2) was
detected by RIA using an antiserum directed against the
carboxyl-terminus of GLP-17–36NH2
(Affinity Research Products, Nottingham, UK).
Before RIA for GLP-1(x-36NH2), 1 ml plasma was
diluted with 2 ml 1% trifluoroacetic acid (pH 2.5 with diethylamine)
and purified by passage through a C18 Sep-Pak cartridge
(Waters Corp., Milford, MA). Peptides were eluted with 3
ml 80% isopropanol/0.1% trifluoroacetic acid, and the eluate was
dried in vacuo to make a single sample for RIA (20). This
procedure resulted in a recovery rate of 53.9 ± 6% when spiked
plasma samples were analyzed. Immunoreactive GIP levels were determined
using 0.1 ml plasma in a human GIP RIA kit (Peninsula Laboratories, Inc., Belmont, CA).
Data analysis
Peptide secretion is expressed as the change from basal values.
All data are expressed as the mean ± SEM. Statistical
significance between experimental groups was assessed by ANOVA using
n-1 post-hoc custom hypothesis tests with Statistical
Analysis System Software (SAS Institute, Inc., Cary, NC).
Comparisons of plasma levels of hormones between basal levels and
subsequent time points within an experimental group were made using
repeated measures ANOVA. Significance was determined at the
P < 0.05 level in these comparisons. Where single
determinations were made, paired Student’s t test was
performed. Some data were log10 transformed before
analysis.
|
Results
|
|---|
Correlation of GLP-1 and gGLI secretion
The large amount of plasma required for assay of
GLP-1 (1 ml/determination) limited repeated sampling over
a prolonged period of time. As the proglucagon molecule is cleaved in
the intestinal L cell to produce glicentin/oxyntomodulin and
GLP-1 in a 1:1 relationship, the secretion of gGLI has
been used to monitor the secretion of GLP-1 indirectly, at
least in vitro (9). To determine the relationship between
gGLI and GLP-1 levels in the rat in vivo,
plasma levels of gGLI and GLP-1 were determined in the
same animals under basal conditions and in response to administration
of fat into the duodenum (Fig. 1
). gGLI
levels were highly correlated with those of GLP-1 (r
= 0.85; P < 0.05; n = 3). As the gGLI assay
requires only 300 µl plasma, compared with 1 ml for the
GLP-1 RIA, gGLI was used as a measure of
GLP-1 secretion in all subsequent studies.
View larger version (16K):
[in this window]
[in a new window]
|
Figure 1. Correlation of
GLP-1(x-36NH2) and gGLI RIAs. The change in
GLP-1 levels after corn oil challenge of the
proximal-distal loop model (n = 3) were corrected for
GLP-1 assay recovery rate and plotted as a function of the
change in gGLI levels in the same plasma samples. The open
circle indicates an outlier in the triplicate for
GLP-1 at that level of gGLI. This outlier was not included
in the regression analysis.
|
|
Effect of left cervical vagotomy on basal levels of gGLI
The mean basal level of gGLI was 193 ± 12 pg/ml (n =
36) in anesthetized rats. To assess whether the vagus nerve has a
global effect on basal levels of gGLI, a left cervical vagotomy was
performed in unstimulated animals. The right cervical vagus nerve was
left intact to maintain cardiorespiratory function. Left cervical
vagotomy reduced the mean basal level of gGLI from 205 ± 11 pg/ml
in control animals to 113 ± 9 pg/ml (P < 0.001;
n = 6). This finding therefore suggested that the vagus nerve
plays a role in modulating basal gGLI secretion. Levels of plasma
glucose and glucagon are shown in Table 1
and were not affected by left cervical vagotomy.
View this table:
[in this window]
[in a new window]
|
Table 1B. Changes in plasma immunoreactive glucagon over time
with left cervical vagotomy and celiac vagal stimulation
|
|
Corn oil infusion in the model of the proximal-distal loop
To confirm that the presence of fat in the proximal duodenum can
signal the GLP-1-containing L cells in the ileum, the
model of the proximal-distal loop was employed (15, 16). Infusion of
3–4 ml saline into a ligated segment of duodenum did not significantly
alter the secretion of gGLI from basal levels (Fig. 2). In contrast, infusion of 3–4 ml fat
in the form of corn oil induced a rapid rise in gGLI secretion
(P < 0.05 at 5 min), which was maintained throughout
the remainder of the experimental time course (P <
0.001 vs. saline infusion; n = 5). The role of the
enteric nervous system in mediating this proximal signal to the ileal L
cells was investigated by transection of the gastrointestinal tract at
a site immediately distal to the isolated segment of duodenum before
the infusion of corn oil. This procedure did not reduce the gGLI
secretion induced by duodenal fat, as the infusion of corn oil elicited
a significant rise in gGLI levels compared with that in saline controls
(P < 0.05; n = 6; Fig. 2). The peak change in
gGLI from basal was 132 ± 37 pg/ml (P < 0.05;
n = 6), and this was not significantly different from that
achieved by the infusion of corn oil into the ligated duodenal segment.
However, a significant difference was observed in the timing of the
corn oil-induced response of gGLI secretion in the two protocols, such
that the fat-induced rise in gGLI levels was significantly delayed by
transection of the gut compared with the animals undergoing gut
ligation (P < 0.05 for corn oil/ligation
vs. corn oil/transection at 5 min). Indeed, the peak change
in gGLI secretion occurred at 19 ± 4 min in the transected
enteric nervous system group compared with 6 ± 1 min in the
ligated group (P < 0.05).
View larger version (24K):
[in this window]
[in a new window]
|
Figure 2. Change in plasma levels of gGLI in response to
3–4 ml corn oil infused into the proximal segment of the duodenum,
isolated by loose ligation (closed triangles; n =
5) or gut transection distal to the proximal segment (open
circles; n = 6). Changes in plasma levels of gGLI in
response to 3–4 ml saline are shown by the closed
squares (n = 5). Differences between groups are
represented by the P values in the legend. Differences
from basal levels within each individual experimental group are
indicated (*, P < 0.05 vs.
basal).
|
|
To investigate the possibility that extrinsic vagal fibers mediate the
proximal signal to the ileal L cells, bilateral subdiaphragmatic
vagotomies were performed in addition to gut transection, before the
administration of fat (Fig. 3). In
contrast to the response found in rats with an intact vagus, corn oil
infusion into the proximal duodenum of vagotomized rats completely
failed to stimulate gGLI secretion (P < 0.001
vs. intact vagus; n = 5). Changes from basal levels of
plasma glucose and glucagon are shown in Table 2.
View larger version (19K):
[in this window]
[in a new window]
|
Figure 3. Effect of subdiaphragmatic vagotomy on changes in
plasma level of gGLI in response to duodenal corn oil. Changes in gGLI
induced by corn oil infusion in gut-transected control animals
(open circles; as also shown in Fig. 2 ; n = 6) and
subdiaphragmatic vagotomized animals (closed circles;
n = 5) are represented. Differences between groups are represented
by the P values in the legend. Differences from basal
levels within each individual experimental group are indicated (*,
P < 0.05 vs. basal).
|
|
Effect of vagal stimulation on gGLI secretion
Electrical stimulation of the celiac branch of the vagus nerve was
performed in unstimulated animals to determine whether the efferent
vagus nerve mediates the signal from the proximal duodenum to the
distal L cells (Fig. 4). The celiac
branch of the vagus nerve was transected, and the distal end was
stimulated for a 15-min period. Electrical stimulation using 10 V and a
frequency of 20 Hz resulted in a significant rise in gGLI levels, which
reached a peak of 71 ± 14 pg/ml above basal at the 10 min point
(P < 0.05; n = 5). gGLI levels returned to basal
during the final phase of the stimulation period, but, surprisingly,
increased again to significance during the recovery period. No
significant effect of celiac branch stimulation on plasma glucose and
glucagon levels was detected compared with those in the control group
(Table 1).
View larger version (12K):
[in this window]
[in a new window]
|
Figure 4. Responses of plasma gGLI to electrical stimulation
of the distal end of the celiac branch of the vagus nerve. Electrical
stimulation consisted of 10 V at 20 Hz for a stimulation period of 15
min. Differences from basal are indicated (*, P <
0.05 vs. basal).
|
|
Effect of high GIP infusion with or without hepatic branch
vagotomy
To determine the relative contribution of the vagus nerve to the
proximal-distal loop with respect to that of the humoral component
mediated by GIP (16), infusions of GIP at varying doses were performed
in animals with or without a hepatic branch vagotomy (Fig. 5). Infusion of GIP at a
supraphysiological dose resulted in rapidly elevated GIP levels that
were not significantly different between the control and hepatic branch
vagotomized animals. Peak immunoreactive GIP levels reached 1030
± 121 pg/ml above basal (P < 0.05, n = 7) in the
control group at 13 ± 2 min and returned toward basal during the
subsequent infusion and recovery periods. These elevations in GIP
levels resulted in significant increases in gGLI in control as well as
vagotomized animals. The peak of gGLI above basal induced by the
supraphysiological GIP infusion in the control group was 123 ± 32
pg/ml (P < 0.05; n = 7), and this was not
significantly different from the gGLI response achieved by
supraphysiological GIP infusion in the hepatic branch vagotomized
animals (96 ± 25 pg/ml; P < 0.05; n = 6).
Interestingly, GIP-induced gGLI secretion rapidly returned to baseline
within 20 min of reaching peak levels despite elevated concentrations
of GIP in both control and vagotomized groups.
View larger version (18K):
[in this window]
[in a new window]
|
Figure 5. Changes in plasma GIP (A) and gGLI (B) levels in
response to supraphysiological infusions of GIP into control (sham
hepatic branch vagotomized animals; closed circles;
n = 7) or hepatic branch vagotomized animals (open
circles; n = 6). The 40-min infusion period (solid
bar) was followed by a 20-min recovery period. Differences
between groups are represented by the P values in the
legend. Differences from basal levels within each individual
experimental group are indicated (*, P < 0.05
vs. basal).
|
|
Effect of low GIP infusion with or without hepatic branch
vagotomy
The effects of physiological increases in GIP levels on gGLI
secretion were also examined in the presence or absence of intact vagal
innervation. The infusion of the physiological dose of GIP into control
animals resulted in a peak change of 223 ± 43 pg/ml above basal
at 13 ± 3 min (P < 0.05; n = 4). The level
of GIP achieved by the physiological infusion of GIP in the vagotomized
group was not significantly different from that in the control group,
reaching a maximum of 169 ± 30 pg/ml above basal at 12 ± 2
min (P < 0.05; n = 5; Fig. 6). Furthermore, the GIP levels achieved
by the infusion were significantly decreased compared with those after
the supraphysiological dose of GIP (P < 0.001). In
control animals, the physiological dose of GIP led to a small, but
significant, elevation in gGLI secretion, reaching a peak of 40 ±
7 pg/ml above basal levels (P < 0.01; n = 4) and
rapidly returning to basal levels by the 20 min point. In contrast,
hepatic branch vagotomy completely abolished the rise in gGLI secretion
induced by the physiological infusion of GIP (P < 0.05
vs. controls). Changes in plasma glucose and glucagon
produced by the infusions of GIP are demonstrated in Table 3.
View larger version (17K):
[in this window]
[in a new window]
|
Figure 6. Changes in plasma GIP (A) and gGLI (B) levels in
response to physiological infusions of GIP into control (sham hepatic
branch vagotomized animals; closed circles; n = 4)
or hepatic branch vagotomized animals (open circles;
n = 6). The 40-min infusion period (solid bar) was
followed by a 20-min recovery period. Differences between groups are
represented by the P values in the legend. Differences
from basal levels within each individual experimental group are
indicated (*, P < 0.05 vs.
basal).
|
|
View this table:
[in this window]
[in a new window]
|
Table 3A. Changes in plasma glucose over time with
supraphysiological or physiological administration of GIP in sham- or
hepatic branch-vagotomized animals
|
|
View this table:
[in this window]
[in a new window]
|
Table 3B. Changes in plasma immunoreactive glucagon over time
with supraphysiological or physiological administration of GIP in sham-
or hepatic branch-vagotomized animals
|
|
|
Discussion
|
|---|
Due to its numerous actions on metabolic processes,
GLP-1 holds great promise as a treatment for type II
diabetes mellitus. Therefore, it is imperative that the factors that
modulate the secretion of GLP-1 are understood. The
results of the present study further this knowledge by demonstrating
that a neuroendocrine loop exists that acts to stimulate the ileal L
cells when nutrients are present in the proximal duodenum. The neural
component of this proximal-distal loop has been demonstrated to involve
the vagus nerve, whereas the humoral component has previously been
described to be mediated by the duodenal hormone GIP (16). Furthermore,
these two limbs of the proximal-distal loop have now been shown to
interact within the physiological range to induce a rapid secretory
response from the ileal L cell.
To investigate the secretion of GLP-1 in vivo,
a method to evaluate GLP-1 levels over a prolonged period
of time is necessary. The major limitation inherent to the RIA of
GLP-1 is the large plasma volume that is necessary to
accurately measure GLP-1 levels in rats. The results of
the present study demonstrate that changes in the plasma levels of
GLP-1 in rats in vivo are highly correlated
with those of gGLI, a finding that is not surprising, as
glicentin/oxyntomodulin and GLP-1 are cleaved from
proglucagon in a one to one relationship (1). Furthermore, plasma
levels of GLP-2, which is also cleaved from proglucagon in a one to one
relationship with glicentin/oxyntomodulin, has recently been shown to
be highly correlated with gGLI levels in the rat (21). Similar findings
have been made for gGLI, GLP-1, and GLP-2 using in
vitro models of the intestinal L cell (9, 22) (Brubaker, P.
L., unpublished data). The results obtained for any one of these
peptides may therefore be extrapolated to include all of the other
intestinal PGDPs, at least in the rat.
To investigate the hypothesis that a neural component of the
proximal-distal loop operates in conjunction with the previously
described endocrine-based loop (15, 16), the vagus nerve was examined
for global effects on basal GLP-1 secretion. This nerve
represents the major parasympathetic innervation of the proximal
portion of the gastrointestinal tract and is known to influence the
secretion of other gut-derived hormones, including GRP (23) and
secretin (24). Left cervical vagotomy significantly depressed the basal
secretion of gGLI to approximately 56% of basal levels in control
rats. This finding demonstrates a tonic stimulatory role for the left
cervical vagus nerve in regulating secretion of GLP-1 from
the L cell.
Nutrients such as fat can exert effects on GLP-1 secretion
directly by stimulating the L cells at their luminal surface (9) or can
have indirect effects on the L cell by exerting their influence through
an intermediate messenger such as GIP. This duodenal hormone has
previously been shown to stimulate secretion of GLP-1 in
response to fat ingestion (16). The direct and indirect effects of fat
can be separated in vivo by ensuring that the progress of
nutrients is impeded before they can reach the site of the majority of
the GLP-1-containing L cells within the ileum. This was
accomplished in our model by applying a ligature loosely around the
gastrointestinal wall at a position approximately 10 cm downstream from
the duodenal site of infusion. Saline infusion into the proximal
isolated segment did not induce secretion of GLP-1 from
the L cells, indicating that the effects of fat on GLP-1
secretion are specific to this nutrient and are not due to distension
of the duodenal wall or the mere presence of a fluid in the lumen of
the duodenum. Consistent with the results of previous studies (15, 16),
infusion of fat into the duodenum induced a rapid and prolonged rise in
gGLI secretion despite the fact that this fat did not reach the ileum
at any time during the experimental time course. Furthermore, the
secretion of GLP-1 induced by duodenal fat cannot be
explained by the presence of a few hypersensitive L cells in this
region, because the response to fat is completely abolished when all
sections of the intestine distal to the duodenum are removed (18).
However, the rapid response of the L cell to duodenal fat (
5 min) as
well as the finding that this response is dependent on the neuropeptide
GRP (18) led to the current hypothesis that in addition to GIP, a
component of the nervous system may also be involved in mediating the
proximal nutrient signal to the distal L cell.
To explore the possibility that fibers within the submucosal or
myenteric nervous plexus may play a role in the rapid L cell response
to fat ingestion, possible contributions of the enteric nervous system
were abolished by transection of the gastrointestinal tract before the
infusion of corn oil. This procedure did not prevent gGLI secretion
compared with the control group, thereby indicating that the enteric
nervous system is not responsible for relaying information about
nutrient ingestion to the ileal L cells. However, an element contained
within the walls of the gastrointestinal tract may be necessary for
maintaining the appropriate timing of the response, as transection of
the gut wall caused a marked delay in the peak of gGLI secretion. One
possible explanation for this finding is that local afferents of vagal
fibers may have been transected in this procedure. Local afferents are
extensions of vagal afferents that divide from the main afferent nerve
and supply a target organ directly without synapsing with higher
centers (25). These have been shown to contain calcitonin gene-related
peptide, which is a potent stimulator of gGLI secretion in
vitro (8). In the present scenario, such afferents could possibly
branch from vagal afferents that innervate the duodenal region and
travel to the ileum within the walls of the gastrointestinal tract to
exert effects on gGLI secretion in a reflex manner. As the main vagal
nerves are not affected by the gut wall transection, this might explain
why the gGLI response was only delayed rather than being totally
abolished in these studies.
Consistent with a major role for the vagus nerve in mediating the
proximal-distal loop, bilateral subdiaphragmatic vagotomy performed in
addition to gut transection completely abolished the fat-induced rise
in gGLI secretion. Two scenarios or a combination of the two may
provide the reason why bilateral subdiaphragmatic vagotomy totally
attenuated the fat-induced rise in gGLI secretion. The first deals with
the possibility that transection of the subdiaphragmatic vagus nerve
has rendered higher brain centers insensitive to sensory messages from
the duodenum. The vagus nerve is known to contain many afferent fibers
arising from the proximal gastrointestinal tract and has been
previously shown to mediate both nutrient and hormonal signals from the
gut in response to nutrient administration (24, 26, 27). In fact, the
vagus nerve has been shown to express receptors that are able to
differentiate between distinct classes of ingested lipid (28). This is
an important finding, considering that both fat and GLP-1
are potent inhibitors of gastric motility (2, 29). Thus, activation of
vagal afferents by ingested lipid may induce the secretion of
GLP-1, thereby augmenting the enterogastrone effects of
fat. Another possibility that may explain the reduced secretion of gGLI
in response to bilateral subdiaphragmatic vagotomy is that the efferent
mechanisms that act to stimulate GLP-1 secretion were
disrupted by the vagotomy. It has been previously suggested that the
parasympathetic nervous system is involved in the modulation of L cell
secretion (30), and indeed, the L cell is stimulated by muscarinic
agonists in vitro (8, 10). Thus, disruption of afferent
and/or efferent vagal fibers in the rat results in a loss of the rapid
L cell response to ingested nutrients. Consistent with this finding, it
has recently been reported that administration of atropine abolishes
the GLP-1 response to an oral glucose tolerance test in
humans, although the effects of cholinergic blockade to inhibit gastric
emptying were not taken into account in this study (30).
To differentiate between the afferent and efferent signals mediating
GLP-1 secretion, the celiac branch of the vagus nerve was
directly stimulated, resulting in significant stimulation of gGLI
secretion at the 10 min point. The delay observed in the secretion of
gGLI induced by electrical stimulation was very similar to that
observed in the experiments involving fat-induced secretion in
gut-transected animals. These findings lend further support to the idea
of a role for the enteric nervous system in mediating a component of
the proximal-distal loop, as this pathway was not affected by the
electrical stimulation. The delay may also occur due to involvement of
a peptidergic mediator in the efferent signaling pathway, as the
effects of these neuromodulators are known to be preceded by long
latent periods (31). Such a possibility is also supported by recent
experiments showing that infusion of an antagonist to the GRP receptor
completely abolishes proximal fat-induced gGLI secretion despite the
fact that GIP levels remained elevated (18). A role for GRP in the
signaling process at a level downstream of GIP was therefore indicated.
GRP is localized extensively in the myenteric plexus (32) and potently
stimulates the secretion of GLP-1 (8, 11, 12). Therefore,
GRP is a likely candidate neuropeptide that may be released upon
electrical stimulation of the celiac branch of the vagus nerve. It
should be noted that the delayed secretion of gGLI in response to
electrical stimulation is not dependent upon release of GIP from the
duodenum, as vagal stimulation has previously been shown to have no
effect on the GIP-containing K cell (33). The mechanism underlying the
rebound secretion of gGLI during the recovery period remains to be
elucidated. However, GLP-1 secretion is known to be
pulsatile, with a frequency of approximately one pulse every 10–15
min. This pattern of secretion has been proposed to be regulated by the
parasympathetic nervous system (30). Thus, it is possible that the
initial vagal stimulation stimulating the release of gGLI in the
present study coordinated the dispersed L cells in the distal small
intestine, leading to an integrated rebound pulse of gGLI secretion
during the recovery period.
The finding that a GRP receptor antagonist completely inhibits gGLI
secretion in the presence of elevated GIP levels suggested that the
neural and humoral arms of the proximal-distal loop may be integrated
in the regulation of the L cell. This hypothesis was examined by
experiments in which GIP was infused at different doses into animals
that possessed intact or disrupted vagal innervation at the level of
the hepatic afferent. When infused at supraphysiological doses, GIP
stimulated gGLI secretion significantly in both control and vagotomized
animals; however, a vagal dependence of GIP became evident when GIP was
infused at physiological doses. This finding indicates that in the
normal rat, the hepatic branch of the subdiaphragmatic vagus nerve
mediates the stimulatory effect of GIP on the L cell. Consistent with
this finding, both cholecystokinin (27) and secretin (24) have been
reported to act on the exocrine pancreas through vagal afferents when
administered at physiological doses, exerting direct effects only when
infused at supraphysiological concentrations. Vagal sensory endings in
the duodenum do not make contact with epithelial cells, but terminate
in close association with the lamina propria (34). Thus, GIP-containing
cells may function as taste cells by sampling the duodenal contents
and, in turn, activating vagal afferents to stimulate
GLP-1 secretion.
In conclusion, the secretion of GLP-1 and gGLI from the
rat ileal L cell is regulated by a complex proximal-distal loop that
involves both endocrine and neural factors (Fig. 7). Fat is sensed in the duodenum by
luminal K cells, which secrete GIP in response to fat and glucose. GIP
exerts its effects on the ileal L cell in two ways depending on the
concentration achieved. Physiological levels of GIP act through vagal
afferent pathways to stimulate the L cell indirectly. This stimulation
is carried to the L cell by efferent pathways located in the celiac
branch of the vagus nerve and is thought to involve GRP. GIP can also
stimulate the L cell directly at higher levels. Finally, a component of
the enteric nervous system also appears to be responsible for the early
stimulation of GLP-1 secretion within this loop. The
secretion of GLP-1 from the distal L cell is, therefore,
intimately connected with the presence of nutrients in the proximal
duodenum through an interaction of neural and endocrine pathways. As
knowledge is gained concerning the functioning of the neuroendocrine
loop, examination of the function of this axis can be made in disease
states, such as type II diabetes mellitus.
View larger version (14K):
[in this window]
[in a new window]
|
Figure 7. Overview of the proposed elements of the
proximal-distal neuroendocrine loop governing GLP-1
secretion from the rat ileal L cell. Fat in the lumen of the duodenum
stimulates the release of GIP from the K cells, which, in turn,
activates the L cell indirectly through vagal afferent pathways. The
proximal signal is then mediated by vagal efferent pathways present in
the celiac branch of the subdiaphragmatic vagus nerve that are thought
to synapse with GRPergic neurons present in the myenteric plexus. At
supraphysiological concentrations, GIP may also act directly through
receptors on the L cell (dotted line). Mechanisms that
stimulate early secretion may also be mediated through the enteric
nervous system in the walls of the gastrointestinal tract
(dashed line).
|
|
|
Footnotes
|
|---|
1 This work was supported by a Margaret A. Mollett grant from the
Canadian Diabetes Association and by a Pharmaceutical Manufacturer’s
Association of Canada-Medical Research Council of Canada
Studentship and a University of Toronto Open Studentship (to
A.S.R.). 
Received December 11, 1998.
|
References
|
|---|
-
Dhanvantari S, Seidah NG, Brubaker PL 1996 Role of prohormone convertases in the tissue-specific processing of
proglucagon. Mol Endocrinol 10:342–355[Abstract]
-
Ritzel R, Orskov C, Holst JJ, Nauck MA 1995 Pharmacokinetic, insulinotropic, and glucagonostatic properties of
GLP-1 [7–36 amide] after subcutaneous injection in healthy
volunteers. Dose-response-relationships. Diabetologia 38:720–725[Medline]
-
Mojsov S, Weir GC, Habener JF 1987 Insulinotropin:
glucagon-like peptide I(7–37) co-encoded in the glucagon
gene is a potent stimulator of insulin release in the perfused rat
pancreas. J Clin Invest 79:616–619
-
Kreymann B, Ghatei MA, Williams G, Bloom SR 1987 Glucagon-like peptide-1 7–36: a physiological incretin in man. Lancet 2:1300–1304[Medline]
-
D’Alessio DA, Prigeon RL, Ensinck JW 1995 Enteral
enhancement of glucose disposition by both insulin-dependent and
insulin-independent processes: a physiological role of glucagon-like
peptide I. Diabetes 44:1433–1437[Abstract]
-
Turton MD, O’Shea D, Gunn I, Beak SA, Edwards CMB,
Meeran K, Choi SJ, Taylor GM, Heath MM, Lambert PD, Wilding JPH, Smith
DM, Ghatei MA, Herbert J, Bloom SR 1996 A role for glucagon-like
peptide-1 in the central regulation of feeding. Nature 379:69–72[CrossRef][Medline]
-
DeFronzo RA, Banadonna RC, Ferrannini E 1992 Pathogenesis of NIDDM: a balanced overview. Diabetes Care 15:318–368[Abstract]
-
Brubaker PL 1991 Regulation of intestinal
proglucagon-derived peptide secretion by intestinal regulatory
peptides. Endocrinology 128:3175–3182[Abstract]
-
Rocca AS, Brubaker PL 1995 Stereospecific effects
of fatty acids on proglucagon-derived peptide secretion in fetal rat
intestinal cultures. Endocrinology 136:5593–5599[Abstract]
-
Abello J, Ye F, Bosshard A, Bernard C, Cuber J-C,
Chayvialle J-A 1994 Stimulation of glucagon-like peptide-1
secretion by muscarinic agonist in a murine intestinal endocrine cell
line. Endocrinology 134:2011–2017[Abstract]
-
Dumoulin V, Dakka T, Plaisancie P, Chayvialle J-A, Cuber
J-C 1995 Regulation of glucagon-like peptide-1-(7–36)amide,
peptide YY, and neurotensin secretion by neurotransmitters and gut
hormones in the isolated vascularly perfused rat ileum. Endocrinology 136:5182–5188[Abstract]
-
Hermann-Rinke C, Vöge A, Hess M, Göke B 1995 Regulation of glucagon-like peptide-1 secretion from rat ileum by
neurotransmitters and peptides. J Endocrinol 147:25–31[Abstract/Free Full Text]
-
D’Alessio D, Thirlby R, Laschansky E, Zebroski H,
Ensinck J 1993 Response of tGLP-1 to nutrients in humans.
Digestion 54:377–379
-
Elliott RM, Morgan LM, Tredger JA, Deacon S, Wright J,
Marks V 1993 Glucagon-like peptide-1(7–36)amide and
glucose-dependent insulinotropic polypeptide secretion in response to
nutrient ingestion in man: acute post-prandial and 24-h secretion
patterns. J Endocrinol 138:159–166[Abstract/Free Full Text]
-
Roberge JN, Brubaker PL 1991 Secretion of
proglucagon-derived peptides in response to intestinal luminal
nutrients. Endocrinology 128:3169–3174[Abstract]
-
Roberge JN, Brubaker PL 1993 Regulation of
intestinal proglucagon-derived peptide secretion by glucose-dependent
insulinotropic peptide in a novel enteroendocrine loop. Endocrinology 133:233–240[Abstract]
-
Borgstrom B, Dahlquist A, Lundh G, Sjovall J 1957 Studies of intestinal digestion and absorption in the human. J
Clin Invest 36:1521–1536
-
Roberge JN, Gronau KA, Brubaker PL 1996 Gastrin-releasing peptide is a novel mediator of proximal
nutrient-induced proglucagon-derived peptide secretion from the distal
gut. Endocrinology 137:2383–2388[Abstract]
-
Weigert N, Schäffler A, Reichenberger J, Madaus S,
Classen M, Schusdziarra V 1995 Effect of endogenous opioids on
vagally induced release of gastrin, somatostatin and bombesin-like
immunoreactivity from the perfused rat stomach. Regul Pept 55:207–215[Medline]
-
Brubaker PL, Crivici A, Izzo A, Ehrlich P, Tsai CH,
Drucker DJ 1997 Circulating and tissue forms of the intestinal
growth factor, glucagon-like peptide-2. Endocrinology 138:4837–4843[Abstract/Free Full Text]
-
Fischer KD, Dhanvantari S, Drucker DJ, Brubaker PL 1997 Intestinal growth is associated with elevated levels of
glucagon-like peptide-2 in diabetic rats. Am J Physiol
273:E815–E820
-
Huang THJ, Brubaker PL 1995 Synthesis and secretion
of glucagon-like peptide-1 by fetal rat intestinal cells in culture.
Endocrine 3:499–503
-
Knuhtsen S, Holst JJ, Knigge U, Olesen M, Nielsen
OV 1984 Radioimmunoassay, pharmacokinetics, and neuronal release
of gastrin-releasing peptide in anesthetized pigs. Gastroenterology 87:372–378[Medline]
-
Li P, Chang TM, Chey WY 1995 Neuronal regulation of
the release and action of secretin-releasing peptide and secretin.
Am J Physiol 269:G305–G312
-
Holzer P 1988 Local effector function of
capsaicin-sensitive sensory nerve endings: involvement of tachykinins,
calcitonin gene-related peptide and other neuropeptides. Neuroscience 3:739–768
-
Ewert WR, Wingate DL 1984 Central representation of
arrival of nutrient in the duodenum. Am J Physiol
246:G750–G756
-
Li Y, Owyang C 1993 Vagal afferent pathway mediates
physiological action of cholecystokinin on pancreatic enzyme secretion.
J Clin Invest 92:418–424
-
Melone J 1986 Vagal receptors sensitive to lipids
in the small intestine of the cat. J Auton Nerv Syst 17:231–241[CrossRef][Medline]
-
Lloyd KCK, Maxwell V, Ohning G, Walsh JH 1992 Intestinal fat does not inhibit gastric function through a hormonal
somatostatin mechanism in dogs. Gastroenterology 103:1221–1228[Medline]
-
Balks HJ, Holst JJ, Von zur Mühlen A, Brabant
G 1997 Rapid oscillations in plasma glucagon-like peptide-1
(GLP-1) in humans: cholinergic control of GLP-1 secretion via
muscarinic receptors. J Clin Endocrinol Metab 82:786–790[Abstract/Free Full Text]
-
Hirst GDS, Choate JK, Cousins HM, Edwards FR, Klemm
MF 1996 Transmission by post-ganglionic axons of the autnomic
nervous system: the importance of the specialized neuroeffector
junction. Neuroscience 73:7–23[CrossRef][Medline]
-
Sjövall M, Ekblad E, Lundell L, Sundler F 1990 Gastrin-releasing peptide: neuronal distribution and spatial
relation to endocrine cells in the human upper gut. Regul Pept 28:47–55[Medline]
-
Berthoud HR, Trimble ER, Moody AJ 1982 Lack of
gastric inhibitory polypeptide (GIP) response to vagal stimulation in
the rat. Peptides 3:907[Medline]
-
Berthoud H-R, Kressel M, Raybould HE, Neuhuber WL 1995 Vagal sensors in the rat duodenal mucosa: distribution and
structure as revealed by in vivo Dil-tracing. Anat Embryol 191:203–212[Medline]
Top
Abstract
Introduction
