Glucagon-like peptide-2 (GLP-2) reduces mouse gastric tone and small intestine transit, but its action on large intestine motility is still unknown. The purposes of the present study were 1) to examine the influence of GLP-2 on spontaneous mechanical activity and on neurally evoked responses, by recording intraluminal pressure from mouse isolated colonic segments; 2) to characterize GLP-2 mechanism of action; and 3) to determine the distribution of GLP-2 receptor (GLP-2R) in the mouse colonic muscle coat by immunohistochemistry. Exogenous GLP-2 (0.1–300 nM) induced a concentration-dependent reduction of the spontaneous mechanical activity, which was abolished by the desensitization of GLP-2 receptor or by tetrodotoxin, a voltage-dependent Na+-channel blocker. GLP-2 inhibitory effect was not affected by Nω-nitro-l-arginine methyl ester (a nitric oxide synthase inhibitor), apamin (a blocker of small conductance Ca2+-dependent K+ channels), or [Lys1,Pro2,5,Arg3,4,Tyr6]VIP7–28 (a VIP receptor antagonist), but it was prevented by atropine or pertussis toxin (PTX), a Gi/o protein inhibitor. Proximal colon responses to electrical field stimulation were characterized by nitrergic relaxation, which was followed by cholinergic contraction. GLP-2 reduced only the cholinergic evoked contractions. This effect was almost abolished by GLP-2 receptor desensitization or PTX. GLP-2 failed to affect the contractile responses to exogenous carbachol. GLP-2R immunoreactivity (IR) was detected only in the neuronal cells of both plexuses of the colonic muscle coat. More than 50% of myenteric GLP-2R-IR neurons shared the choline acetyltransferase IR. In conclusion, the activation of GLP-2R located on cholinergic neurons may modulate negatively the colonic spontaneous and electrically evoked contractions through inhibition of acetylcholine release. The effect is mediated by Gi protein.
- enteric nervous system
- gastrointestinal hormones
- colonic motility
glucagon-like peptide-2 (GLP-2) is a 33-amino acid peptide derived from proglucagon that is secreted by enteroendocrine L cells of the small and large intestine following nutrient intake, especially carbohydrate and fats (7). The gastrointestinal tract is the principal target for GLP-2 action where it affects multiple facets of physiology, including growth, absorption, and motility (28). In particular, GLP-2 has been shown to be an important intestinotrophic factor that stimulates epithelial cell proliferation and inhibits apoptosis, increases crypts and villi and enhances intestinal digestive and absorptive capacity (8, 28). Accordingly, GLP-2 is regarded with interest also in relation to its therapeutic potential in several intestinal syndromes (9, 12).
Furthermore, several studies in human and animal models have demonstrated that GLP-2 is able to affect gut motor activity. Specifically, GLP-2 reduces the vagally induced antral motility in pigs (30), it slows human gastric emptying (13, 20), and it decreases the mouse gastric fundic tone, leading to an increase of the stomach capacity (1). GLP-2 seems to act in concert with glucagon-like peptide-1 (GLP-1) to inhibit rat small bowel myoelectric activity in the fasted state (6). The peptide modulation on the gastrointestinal motility appears to be due mainly to central nervous mechanisms (30), but involvement of the enteric nervous system has been also shown (1). In addition, GLP-2 has been reported as an inhibitor of intestinal transit in the mouse (16), but the action of GLP-2 on large intestine motility has not been explored yet.
The GLP-2 effects are mediated by the interaction with a specific GLP-2 receptor (GLP-2R), belonging to the class of seven transmembrane-spanning G protein-coupled receptors. After GLP-2 binding to the transfected receptor, adenosine 3′,5′-cyclic monophosphate (cAMP) levels are increased (19, 32), consistent with findings from studies of related members of the glucagon/GLP-1R family. However, GLP-2R ability to couple to different G protein subunits and to activate multiple signaling pathways has been also demonstrated (14, 22). The presence of GLP-2R-mRNA transcripts has been demonstrated by Northern blot analysis along the murine gastrointestinal tract with high levels of expression in bowel (31) and, by in situ hybridization, in the murine enteric neurons (5). Furthermore, GLP-2R protein has been detected in rodent and human intestinal subepithelial myofibroblasts (21), in human and pig enteroendocrine cells, and in enteric neurons, suggesting that many of the GLP-2 actions may be indirect through release of not-yet-identified secondary mediators (11, 31).
The present study was undertaken to examine potential peripheral motor effects of GLP-2 on the spontaneous mechanical activity and on neurally evoked responses in mouse proximal colon and to determine the mechanism responsible for the observed effects. In addition, we determined the protein expression of GLP-2R in different regions of mouse gastrointestinal tract by Western blotting and the distribution of GLP-2R in the colonic mouse muscle coat by immunohistochemistry.
MATERIALS AND METHODS
The experiments were authorized by Ministero della Sanità (Rome, Italy), following the guidelines of the European Communities Council Directive of 24 November 1986. Adult male mice (C57BL/10SnJ; weighing 25.5 ± 1.5 g) were killed by cervical dislocation. The abdomen was immediately opened and the proximal colon was removed from a position just distal to the cecum. The content of the excised segment was cleaned with Krebs solution and segment of ∼2 cm in length was cut. Preparation was mounted in a custom-designed horizontal organ bath (volume = 5 ml), which was continuously perfused with oxygenated (95% O2 and 5% CO2) and heated (37°C) Krebs solution with the following composition (mM): 119 NaCl, 4.5 KCl, 2.5 MgSO4, 25 NaHCO3, 1.2 KH2PO4, 2.5 CaCl2, 11.1 glucose.
The mechanical activity was recorded as previously described (18). In brief, the distal end of each segment was tied around the mouth of a J tube, which was connected via a T catheter to a pressure transducer (Statham mod. P23XL; Grass Medical Instruments, Quincy, MA) and to a syringe for filling the preparation with Krebs solution. Each preparation was filled with 0.1 ml Krebs solution. The ligated proximal end was secured with a silk thread just to preload the preparations of 0.5 g. The mechanical signals were detected as changes in endoluminal pressure, which are mainly generated by circular muscle (1) and recorded on an ink writer polygraph (Grass model 7D; Grass Instruments, Quincy, MA).
To provide electrical field stimulation (EFS), we used an S88 square-wave pulse generator (Grass Medical Instruments) coupled via a stimulus isolation unit (Grass SIU5) to a pair of platinum plates, which were placed in parallel on either sides of the intestinal segment. Preparations were allowed to equilibrate for ∼60 min before the start of the experiment.
In a first set of experiments, after the equilibration time, the response of the preparation to noncumulative concentrations of GLP-2 (0.1–300 nM) was examined. The peptide was added into the bath at increasing concentrations in volumes of 50 μl and each concentration was left in contact with the tissue for 7 min. The interval between single concentrations was 40 min to avoid tachyphylaxis. In the absence of well-characterized GLP-2R antagonists, to confirm the specificity of the observed effect, the peptide-induced response was tested after 40 min of tissue pretreatment with GLP-2 (10 nM) to induce desensitization of the receptors. Moreover, to investigate the neuronal nature of inhibitory responses, a submaximal concentration of GLP-2 (30 nM) was tested in presence of tetrodotoxin (TTX) (1 μM), a voltage-dependent Na+-channel blocker. Thus the effects of increasing concentrations of GLP-2 were evaluated after pretreatment of intestinal preparations with Nω-nitro-l-arginine methyl ester (l-NAME) (300 μM), an inhibitor of nitric oxide (NO) synthase; apamin (0.1 μM), a blocker of small conductance Ca2+-dependent K+ channels; [Lys1,Pro2,5,Arg3,4,Tyr6]VIP7–28, a VIP receptor antagonist; atropine (1 μM), a muscarinic receptor antagonist; or pertussis toxin (PTX) (300 ng/ml), a Gi/o protein inhibitor. These agents were added to the perfusing solution at least 30 min before testing of the peptide, except PTX, which was left in contact with the tissue for 3 h. The concentrations of the inhibitors used were determined from previous experiments in which they have been shown to be effective in mouse colon or from literature (1, 15, 18, 26).
The influence of GLP-2 (0.1–300 nM) on the electrically evoked responses was evaluated in separate set of experiments, being the spontaneous mechanical activity abolished by the repetitive application of EFS. Trains of stimuli (duration 5 s, supramaximal voltage, 8 Hz and 0.5-ms pulse duration) were applied to colonic preparation at intervals of 70 s and stable and reproducible responses for a time period of 3 h were observed. EFS induced a biphasic response, characterized by muscular relaxation followed by contraction, which was abolished by TTX (1 μM), suggesting its neural origin. The responses evoked by EFS were analyzed in the presence of noncumulative increasing concentrations of GLP-2 (0.1–300 nM). The contact time for each concentration was 7 min. The effect of GLP-2 was evaluated after GLP-2R desensitization with GLP-2 (10 nM for 40 min) or PTX (300 ng/ml). In separate experiments, tissues were exposed to cumulative increasing concentrations of carbachol (CCh) (10 nM–30 μM) or to KCl (30 mM), and the myogenic contractions produced were evaluated in the absence or in the presence of GLP-2 (30 nM).
Data Analysis and Statistical Tests
For data analysis, the mean amplitude of the spontaneous pressure waves was determined for 7 min before and after GLP-2 administration. The inhibitory response of the colonic circular muscle to GLP-2 was taken as the percent change from the resting spontaneous activity (e.g., 100% corresponds to the abolition of spontaneous activity). The inhibitory effect of GLP-2 on evoked cholinergic contractions was expressed as a percentage of the response produced by EFS in control conditions. Contractile effects induced by CCh were expressed as a percentage of the maximal response. All data are mean values ± SE; n indicates the number of experimental animals. The concentration (EC50) with 95% confidence interval (CI) producing half-maximum response was calculated using Prism 4.0, GraphPad (San Diego, CA). Statistical analysis was performed by means of ANOVA followed by Bonferroni post hoc test. A probability value of less than 0.05 was regarded as significant.
The following drugs were used: apamin, atropine sulfate, CCh, l-NAME (Sigma Aldrich, Milan, Italy), TTX citrate (Ascent Scientific, Bristol, UK), rat glucagon-like-peptide 2 (GLP-2), PTX, and [Lys1,Pro2,5,Arg3,4,Tyr6]VIP7–28 (Tocris-Bioscience, Bristol, UK). Each compound was prepared as a stock solution in distilled water. The working solutions were prepared fresh the day of the experiments by diluting the stock solutions in Krebs.
Western Blotting and Immunohistochemistry
For Western blotting, ∼30 mg of tissue from fundus, antrum, duodenum, jejunum, and colon was incubated on ice in RIPA buffer (50 mM Tris·HCl, pH 7.4; 150 mM NaCl, 1% Nonidet P-40) containing protease inhibitors (2 mM PMSF, NaVO3) for 1 h. Subsequently, it was centrifuged for 15 min at 12,000 rpm and supernatant was isolated. Protein concentration was measured by the Bio-Rad Protein Assay (Bio-Rad Laboratories, Milan, Italy). Proteins (50 μg) were separated by 10% SDS-PAGE containing 0.1% SDS and transferred to Hybond-C nitrocellulose membranes (Amersham Life Science, Munich, Germany) by electroblotting. Loading and transfer conditions were assessed by staining of the gel with Ponceau red. The relative migration position of target protein was detected by using a coelectrophoresed prestained molecular weight protein ladder (Invitrogen, Paisley, UK). The membranes were sequentially incubated overnight with antibodies to GLP-2R (Santa Cruz Biotechnology, Santa Cruz, CA) raised against the NH2-terminal extracellular domain of rodent GLP-2R (diluted 1:200), or mouse β-tubulin (diluted 1:5,000) (Sigma Aldrich) applied as a loading control, and proteins were visualized by using an anti-goat IgG secondary antibody conjugated to horseradish peroxidase (diluted 1:10,000) (Santa Cruz Biotechnology). The target proteins were detected by enhanced chemiluminescence (Pierce, Rockford, IL).
For immunohistochemistry, proximal colon specimens (n = 3 mice) were fixed in 4% paraformaldehyde for 4 h at 4°C. The specimens were cryoprotected in 30% sucrose in phosphate-buffered saline (PBS) for ∼12 h at 4°C, frozen in Killik cryostat embedding medium (Bio-Optica, Milan, Italy). Transverse sections, 12 μm thick, were cut, collected on polylysine-coated slides, and preincubated in 0.5% Triton (Sigma Aldrich) and 1.5% bovine serum albumin (Sigma Aldrich) in PBS for 15 min at room temperature. Then they were incubated with GLP-2R polyclonal antibody (Santa Cruz Biotechnology) at final dilution of 1:50 overnight at 4°C. The immunoreaction was revealed by using the secondary antibody Alexa Fluor 488 donkey anti-goat (Invitrogen) 1:333 for 2 h at room temperature. To identify the potential localization of GLP-2R on cholinergic neurons, transverse sequential sections, 4 μm thick, were cryocut from each specimen and collected on slides (4 sections/slide, 2 slides/specimen) in two separate areas, one area containing the first and the third section, the other area containing the second and the fourth section. Each area was bordered with a pap pen, and the two sections of one area were incubated with GLP-2R antibody as described above whereas the two sections of the neighbor area were incubated with choline acetyltransferase (ChAT) polyclonal antibody (a generous gift of Dr. M. Schemann) (23), at final dilution of 1:500 overnight at 4°C. The two immunoreactions were revealed by using the Alexa Fluor 488 secondary donkey anti-goat 1:333, and the Alexa Fluor 568 secondary goat anti-rabbit 1:333 respectively, both incubated for 2 h, at room temperature and then observed under an epifluorescence Zeiss Axioskop microscope. Negative controls were performed by omitting the primary antibodies, and all of them had no labeling (data not shown).
By using a ×40 objective, GLP-2R-immunoreactivity (IR) neurons were counted along the entire perimeter of each section (4 sections each animal) taking as starting and ending point the insertion of the mesentery. Quantification of the neurons sharing GLP-2R- and ChAT-IR was done on the sequential sections collected as above described. Digitized images of the entire perimeter of the muscle wall were acquired via a ×40 objective and transformed into TIFF files by use of Scion Image. Field edges were defined on the basis of structural details within the tissue section to ensure that the fields did not overlap. Comparison between pictures taken from GLP-2R-IR and ChAT-IR sections at the same level was done to identify and quantify those neurons that shared the two markers. Only the labeled neuronal bodies were considered for quantification. The count was done by two of us blind to each other's results on a total of 12 slices for each antibody. The results were expressed as means ± SE.
Influence of GLP-2 on spontaneous mechanical activity.
Mouse proximal colon exhibited spontaneous mechanical activity consisting of phasic contractions at a frequency of 1.8 ± 0.3 (n = 15) and an amplitude of 10.6 ± 2.5 cmH2O (n = 15). GLP-2 (0.1 nM 300 nM) produced inhibitory effects on the pressure waves, characterized by a decrease in the mean amplitude of spontaneous contractions, without affecting the frequency and the resting tone (Fig. 1A). The effect occurred within 1 min after addition of the peptide and was completely reversible after washout with normal Krebs solution. Figure 1B shows the concentration-response curve for the inhibitory effects induced by GLP-2 on the spontaneous contractions of isolated mouse proximal colon. GLP-2 produced a maximal effect corresponding to ∼75% of reduction of amplitude of spontaneous contractions with an EC50 = 4.0 nM (CI = 2.3–7 nM; n = 6). To assess the specificity of the effect, the preparations were pretreated for 40 min with GLP-2 (10 nM) to desensitize the receptors. This treatment produced an early and transient reduction of the spontaneous contractions, which was followed by recovery to the initial amplitude. In these conditions, the inhibition of mechanical activity induced by the peptide was significantly reduced over the full concentration used (Fig. 1B). The desensitization effect was not reversible after 2-h washout.
The response to GLP-2 (0.1–300 nM) was abolished by TTX (1 μM), which per se reduced the spontaneous contraction amplitude (∼4 cmH2O), suggesting its neural origin. However, it was not affected by l-NAME (300 μM), a blocker of the NO synthase that per se increased amplitude of spontaneous activity; apamin (100 nM), a blocker of small conductance Ca2+-dependent K+ channels; or [Lys1,Pro2,5,Arg3,4,Tyr6]VIP7–28 (300 nM), a VIP receptor antagonist (Fig. 2). In contrast, the GLP-2 inhibitory effect was significantly reduced by atropine (1 μM), a muscarinic receptor blocker, that per se decreased the amplitude of the spontaneous contractions (∼7 cmH2O), or by PTX (300 ng/ml), which per se was without any effect on spontaneous mechanical activity (Fig. 3).
Influence of GLP-2 on the neurally evoked responses.
EFS (trains of 8 Hz for 5 s) induced a nitrergic small relaxation (0.4 ± 0.2 cmH2O; n = 12) (abolished by l-NAME) which was always followed by cholinergic contraction (14.1 ± 2.3 cmH2O; n = 12) (abolished by atropine). GLP-2 (0.1–300 nM) caused a concentration-dependent reduction of the electrically evoked cholinergic contractions, without affecting the neural inhibitory response (Fig. 4, A and B). GLP-2 (100 nM) produced ∼60% of reduction of the evoked contractile response amplitude and this effect was reversible after washing out. The inhibitory action of GLP-2 was significantly reduced by GLP-2R desensitization or by PTX (300 ng/ml) (Fig. 4B). GLP-2R desensitization caused an transient reduction of the evoked contractions, which was followed by recovery to the initial amplitude, whereas PTX did not affect the evoked responses.
In the tissue, GLP-2 (30 nM) did not affect the concentration-dependent contractile response induced by CCh (10 nM–30 μM), a muscarinic agonist, (Fig. 5) or the contraction induced by KCl (30 mM) (data not shown).
Western blot analyses of mouse fundus, antrum, duodenum, jejunum, and colon yielded a single 72-kDa band corresponding to the molecular weight of the GLP-2R. Tissue-specific differences in protein expression were observed given that GLP-2R levels were relatively higher in gastric fundus and colon than in the small bowel (Fig. 6).
GLP-2R-IR in the proximal colon was detected only in the neuronal cells of myenteric and submucous plexuses. The IR neurons were round or oval in shape; the labeling was detected in the perikaryon and in numerous nerve varicosities inside the myenteric ganglia. The labeling had a granular aspect. IR fibers were seen outside the ganglia, few in the thickness of the circular muscle coat, many at the submucosal border (Fig. 7) in the region containing the so-called submucosal interstitial cells of Cajal. Some of the IR myenteric neurons were also ChAT-IR (Fig. 8).
ChAT-IR neurons were present in both plexuses (myenteric and submucous plexuses). They had a round or oval perikaryon and the labeling was homogenously distributed in the cytoplasm. Some IR nerve fibers were detected in the muscle coat and numerous at the submucosal border of the circular muscle layer (data not shown). Some of the IR neurons were also GLP-2R-IR (Fig. 8).
The mean number of GLP-2R-IR and ChAT-IR neurons per slice in the myenteric plexus is reported in Table 1. The mean number of neurons per slice that, by comparison, shared the two markers was 18.44 ± 1.32 at the myenteric plexus.
The present study demonstrates that, in mouse proximal colon, GLP-2 is able to modulate negatively the spontaneous mechanical activity and the electrically evoked cholinergic contractions through inhibition of acetylcholine release from enteric neurons. These conclusions are also supported by immunohistochemistry, showing the presence of the GLP-2R on myenteric neurons, half of which shared ChAT-IR.
It is well accepted that GLP-2 affects multiple facets of gastrointestinal physiology, concerning mainly the control of mucosal growth and function (e.g., epithelial integrity, secretion, local blood flow, nutrient uptake, and utilization) (8, 28). However, little is known about its role in the control of the gastrointestinal motility. In anesthetized pigs, GLP-2 acts as an inhibitor of gastric motility, because it abolishes the motor response induced by hypoglycemia (30). In humans, results about the ability of GLP-2 to suppress gastric motility are conflicting and could depend on the methodology used to assess antral emptying or on the type of administered test meal (low-calorie liquid meal or high-calorie solid meal) (13, 17, 20, 24). In mouse, GLP-2 decreases the gastric tone leading to an increase of the stomach capacity (1) and inhibits the intestinal transit (16).
We initially used Western blotting to examine the distribution and relative abundance of GLP-2R along the mouse gastrointestinal tract, because to date this has not been investigated and the only available information is about GLP-2R gene expression (31). GLP-2R was observed in all gastrointestinal regions examined, with relatively higher levels of expression in gastric fundus and colon than in small intestine. The GLP-2R high level in gastric fundus supports our recent evidence for GLP-2 ability to induce gastric relaxation (1).
Therefore, our working hypothesis was that GLP-2R activation may induce motor effects also in mouse colon. Actually, our data show that exogenous GLP-2 reduced in a concentration-dependent manner spontaneous mechanical activity suggesting a potential inhibitory role of the peptide on mouse colonic circular muscle. The effect is mediated by GLP-2R because the agonist-induced desensitization almost abolished the GLP-2 response. In fact, in the absence of GLP-2R well-characterized antagonists, desensitization can represent a useful pharmacological tool in the study of receptors, and GLP-2R has been shown to undergo rapid and sustained homologous desensitization induced by the agonist (10, 29). Indeed, GLP-23–33 has been uses as a GLP-2R antagonist (3, 25), but it works as well as a weak partial agonist (27). The potency (EC50) of GLP-2 in inducing colonic inhibitory effects was 4.0 nM, which is in agreement with that reported for the human and/or rat GLP-2 ability to stimulate GLP-2 receptor in various cell types (EC50 ranging from 0.04 to 14 nM) (10, 19, 27, 29) or for inducing mouse gastric relaxation (1). Preliminary analyses in our laboratory have indicated that the GLP-2 plasma concentration is ∼0.7 ng/ml in ad libitum-fed mice, in good accord with the range of effective concentrations used in this study. Moreover, a dipeptidyl peptidase IV inhibitor (DPPI 1c) caused reduction of evoked contractions suggesting the occurrence of a system for metabolizing GLP-2 in the colon (F. Mulè, unpublished observations). Therefore, we are encouraged to consider the GLP-2 effect as a physiological action. We can only speculate about hormonal role on colonic motility because it could be released in the blood stream from L-type enteroendocrine cells of the small and large intestine after nutrient ingestion or it could act with a paracrine mechanism near its site of production (colonic enteroendocrine L-cells). Reduced colonic propulsion caused by GLP-2 could lead to greater absorption of water and electrolytes, in agreement with its nutrient absorptive function (8, 28).
Moreover, we investigated whether GLP-2 induces inhibition of colonic motility via a direct action on the smooth muscle cells and/or via an indirect action, mediated by enteric nervous system, because expression in enteric neurons of both submucosal and myenteric plexus has been reported (3, 5, 11). The observation that TTX, a blocker of neuronal voltage-dependent Na+ channels, abolished the GLP-2 effects suggests that neurons within the intramural plexuses are responsible for the action of the peptide in mouse colonic circular muscle. Indeed, our immunohistochemical study showed that in the proximal colon the GLP-2R was expressed only by the neuronal cells. IR was detected both in the soma of myenteric and submucosal neurons and in the nerve fibers, especially at the border between the circular muscle layer and the submucosa.
The inhibitory effects induced by GLP-2 were not mediated by the release of any inhibitory neurotransmitters because the NO synthase inhibitor l-NAME, the VIP receptor antagonist [Lys1,Pro2,5,Arg3,4,Tyr6]VIP7–28, or the blocker of small conductance Ca2+-dependent K+ channels apamin, which antagonizes the inhibitory purinergic transmission in this preparation (26), failed to affect the mechanical response to GLP-2.
Therefore, we addressed the possibility that GLP-2 inhibitory effects could be due to a reduction of ongoing release of acetylcholine. Cholinergic motor neurons have been reported to be spontaneously active in mouse ileum (4), and we observed a reduction of the amplitude of spontaneous contractions by atropine suggesting that, in our preparation, muscarinic receptors are tonically activated, as reported in a previous investigation (18). Atropine significantly attenuated GLP-2 inhibitory effects, indicating that muscarinic receptors were involved on the hormone mechanism of action. Therefore, our data can be interpreted as suggesting that circular smooth muscle of mouse proximal colon is under a tonic influence by neural acetylcholine and GLP-2R activation would reduce the release of the excitatory transmitter from the cholinergic nerves. To further support our hypothesis we tested GLP-2 on neurally mediated responses evoked by EFS. The observations that the peptide reduced, in a concentration-dependent manner, the electrically evoked cholinergic responses, without affecting the nitrergic relaxation, and this effect was almost completely abolished after GLP-2R desensitization, provide further evidence that GLP-2 receptor activation is able to modulate the release of acetylcholine. On the other hand, GLP-2 failed to affect the contractions induced by CCh or KCl, confirming that the peptide does not directly interfere with smooth muscle muscarinic receptors and the inhibitory action is achieved primarily by acting on prejunctional receptors. Once more, immunohistochemical data support our hypothesis because more than 50% of myenteric GLP-2R-IR neurons shared ChAT-IR. Our proposed mechanism is in agreement with a previous report in guinea pig ileum, showing that activation of GLP-2R, present on cholinergic secretomotor neurons, decreases epithelial chloride secretion by suppressing acetylcholine release (3). Consistent with our functional and morphological results, recent findings have pointed out the importance of the enteric excitatory motoneurons in the downstream signaling of the glucagon-like peptides to inhibit mouse intestinal motility (16). In fact, they showed that in a murine animal model with a partial enteric nervous system deficit, characterized by a dramatic decrease of cholinergic neurons number (GFRα2-deficient animals), GLP-2 was not able to induce reduction of intestinal transit, as did in wild-type animals (16). The reduction of acetylcholine release could represents an explanation for the GLP-2 inability to inhibit motility in these animals.
Because studies characterizing GLP-2R-regulated intracellular signaling pathways in transfected cell lines and in intestinal mucosa have reported increases in cAMP (19, 29, 32), GLP-2 should be thought as facilitating synaptic transmission. The observation that, in our experimental model, the inhibitory effects of GLP-2 on spontaneous mechanical activity and on evoked cholinergic contractions were significantly inhibited by pretreatment with PTX, a Gi/o protein inhibitor, supports the hypothesis that, in mouse colonic cholinergic myenteric neurons, GLP-2R may be coupled to inhibitory G protein, leading in turn to reduction of the cAMP level. GLP-2R ability to couple to different G protein subunits and activate multiple signaling pathways has been demonstrated (14, 22). In particular, GLP-2R may be coupled in a dose-dependent manner to alternate G protein and cAMP accumulation occurs at moderate concentrations (0.1–1 nM), but there is reduction with higher levels of GLP-2 (25, 29).
In conclusion, in mouse proximal colon, GLP-2 is able to inhibit the spontaneous and electrically evoked contractions acting peripherally through inhibition of prejunctional acetylcholine release. The slowing of colonic motility could lead indirectly to an increase of water and electrolyte absorption, in according with GLP-2 role as nutrient absorption stimulator.
This work was supported by a grant from Ministero dell'Istruzione, dell'Università e della Ricerca (PRIN 2007), Italy.
No conflicts of interest, financial or otherwise, are declared by the author(s).
↵* M. G. Vannucchi is the coordinator of the immunohistochemical studies.
- Copyright © 2010 the American Physiological Society