Address for Correspondence Takafumi Sakai, Area of Regulatory Biology, Division of Life Science, Graduate School of Science and Engineering, Saitama University, 255 Shimo-ohkubo, Sakura-ku, Saitama 338-8570, Japan. Tel: +81-48-858-3869; fax: +81-48-858-3422; e-mail: email@example.com
Background It has been shown in human and canine studies that motilin, a gastroprokinetic hormone, induces gastric phase III contractions via the enteric nervous; however, the center of motilin action in the stomach has not been clearly revealed. In the present study, we investigated the neural pathway of motilin-induced gastric contraction by using Suncus murinus, a new animal model for motilin study.
Methods An isolated suncus stomach was used in vitro to determine the mechanism of motilin action through the myenteric plexus. Synthetic suncus motilin (10−11–10−7 mol L−1) was added to an organ bath, and the spontaneous contraction response was expressed as a percent of ACh (10−5 mol L−1) responses. Motilin-induced contractions were also studied by a pharmacological method using several receptor antagonists and enzyme inhibitor.
Key Results Suncus motilin induced a concentration-dependent gastric contraction at concentrations from 10−9 to 10−7 mol L−1. The responses to suncus motilin in the stomach were completely abolished by atropine and tetrodotoxin treatment and significantly suppressed by administration of hexamethonium, verapamil, phentolamine, yohimbine, ondansetron, and naloxone, whereas ritanserin, prazosin, timolol, and FK888 did not affect the action of motilin. Additionally, N-nitro l-arginine methylester slightly potentiated the contractions induced by motilin.
Conclusions & Inferences The results indicate that motilin directly stimulates and modulates suncus gastric contraction through cholinergic, adrenergic, serotonergic, opioidergic, and NO neurons in the myenteric plexus.
Motilin, a fascinating hormone for the physiologist, is a gastrointestinal (GI) hormone with 22 amino acid residues1 and is thought to be an important peptide involved in the control of interdigestive contractile activity in the gastrointestine of several animals including humans and dogs. Many researchers have shown that an increase in plasma motilin concentration caused simultaneous contractile activity in the stomach,2,3 that phase III contractions in the stomach were induced by exogenous motilin administration,4 and that spontaneous gastric phase III contractions were completely eliminated by neutralizing circulating motilin with motilin antiserum or motilin antagonist,5–7 indicating endogenous motilin stimulates gastric phase III contraction in the interdigestive motor complex.
The mechanisms of motilin-induced gastric contractions have been studied, and it has been reported that the effect of motilin is species-specific and motilin evokes contractions in the GI tract through direct and indirect pathways, i.e., neural pathway8,9 or smooth muscles cells.10–14 Similarly, the effect of motilin on the antrum and duodenum is exerted through different stimulatory pathways in humans.15 For example, in human studies, high doses of motilin-induced gastric contraction through direct stimulation of smooth muscle, whereas motilin at a low dose exerted its effect via a neural pathway.16,17 Moreover, two subtypes of motilin receptors were found in the nerve- or smooth muscle-enriched preparations of the human or rabbit antrum and duodenum by using a receptor binding assay.18–21 In the dog study, the vagus nerve may not be important for exogenous motilin-induced gastric contractions,22 but rather the complete muscle and myenteric neural connection is important, because motilin induced strong phasic contractions that were mediated through the intrinsic nerves in both ex vivo and in vivo study23,24 but strip preparation did not react to motilin.23 Immunohistochemical findings in dogs also indicate the presence of motilin receptors on neuronal cell bodies and fibers in the myenteric plexus of the stomach and duodenum of dogs but not on longitudinal or circular smooth muscle cells,25 suggesting that the pharmacological effect and the physiological effect of motilin is elicited through the myenteric plexus in the GI tract. Taken together, these results suggest that at least one of the target sites of motilin is the myenteric plexus, and motilin neural receptors seem to have a more imperative contribution than receptors located on muscle cells for regulation of the physiological action of motilin (induction of interdigestive phase III). Therefore, it is important to clarify the neural networking system involved in motilin-induced gastric contraction. Nevertheless, the mechanism of motilin action through the myenteric plexus in the stomach has still not been completely elucidated.
The lack of a suitable animal for motilin study may be the reason for the slow progress in this field. To date, dogs have been used for motilin study because many physiological properties of migrating motor complex (MMC) in the fasting state and the relationship between motilin and MMC were successfully studied by using dogs. However, the dog may not always be a good model animal in an in vitro system because dogs are too large for studies to clarify the comprehensive motilin system, e.g., studies on the distribution of motilin receptors and the neuronal signal pathway of motilin stimulation. One the other hand, rodents, the most classical and widely used laboratory animals, cannot be used for motilin experiments because the motilin gene has been inactivated in the common ancestor of mouse and rat.26
To address this problem, we have been screening suitable small laboratory animals for motilin study and focused on Suncus murinus, a small mammal laboratory animal. Suncus belongs to the order Insectivora, family Soricidae, genus Suncus, and this order of animals has traditionally been considered as one of the key groups for understanding the origin of mammals.27,28 Suncus has been used for studying the mechanism of vomiting and development of antiemetic drugs.29–32 We have determined cDNA sequences of suncus motilin and ghrelin by using PCR cloning.33,34 We also studied the contractile properties of the suncus stomach, in both the conscious free-moving suncus and an organ bath experiment and found that the suncus has almost the same GI motility and motilin response as that found in humans and dogs33,35 indicating that this animal can be used as a suitable small model animal for motilin study.
In this study, we investigated the myenteric nervous system involved in motilin-induced gastric contraction by using suncus whole stomach in vitro.
Materials and methods
The experiments were conducted by using more than 5-week-old female suncus of an outbred KAT strain established from a wild population in Kathmandu, Nepal,29 weighting 45–75 g. Animals were housed individually in plastic cages equipped with an empty can for a nest box and were allowed food (trout pellets; Nippon Formula Feed Manufacturing Co., Ltd., Yokohama, Japan) and water ad libitum. The animal room was maintained at 21–24 °C and the light and dark cycle was controlled to change every 12 h (lights on from 8.00 to 20.00 hours). All procedures were approved and performed in accordance with the Saitama University Committee on Animal Research. All efforts were made to minimize animal suffering and to reduce the number of animals used in the experiment.
Preparation of suncus isolated stomach
After having been deeply anesthetized with diethyl ether, the animals were killed by decapitation, and the stomachs were immediately placed into freshly prepared Krebs’ solution (composition in mmol L−1: NaCl 118, KCl 4.75, CaCl2 2.5, MgSO4 1.2, NaH2PO4 1.8, NaHCO3 25, and glucose 11.5; pH 7.2) after laparotomy. The mesentery attachments and fatty tissues were removed, and the inside of each stomach was washed with Krebs’ solution through a small incision in the gastric fundus. The stomachs were then mounted in 10 mL water-jacketed organ baths and initially loaded with approximately 1.0 g weights. The temperature of Krebs’ solution was maintained at 37 ± 0.5 °C and the solution was aerated continuously with a mixture of 95% O2 and 5% CO2.
Gastric contractility study
Contractile activities of the stomach with motilin treatment were monitored by using an isometric force transducer (UM-203; Iwashiya Kishimoto Medical Instruments, Kyoto, Japan) and software (PicoLog for Windows; Pico Technology Ltd., St Neots, UK). To normalize the contractions of this experiment, ACh (10−5 mol L−1) was given twice into the organ bath before the cumulative administration of motilin with the absence and presence of an antagonist, and at the end of the experiment, ACh (10−5 mol L−1) was given once again into the organ bath. Thereafter, the percentage of maximal contractions was calculated by averaging the tonic response induced by these three administrations. Note that in each case the ACh administration evoked almost the same tonic gastric contraction. The effects of suncus motilin in the absence or presence of antagonists were expressed as a percentage of the control contractions. Concentration–response curves were obtained by cumulative addition of suncus motilin with or without antagonists or an inhibitor at appropriate intervals to the organ bath.
The concentrations of agonists and antagonists added did not exceed 6% of bath volume. Acetylcholine chloride (Sigma, St. Louis, MO, USA) was dissolved in DW and synthetic suncus motilin (Bex, Tokyo, Japan) was dissolved in 0.1% BSA/PBS. In antagonist or inhibitor experiments, the stomachs were equilibrated before the application of suncus motilin with the antagonists: atropine sulfate (10−6 mol L−1; Merck, San Diego, CA, USA),36,37 hexamethonium bromide (10−4 mol L−1; Wako, Osaka, Japan),38 verapamil hydrochloride (10−6 mol L−1; Sigma),36 tetrodotoxin (TTX) (10−6 mol L−1; Wako),36,37 prazosin hydrochloride (10−6 mol L−1; Wako),39 timolol maleate (10−6 mol L−1; Wako),40 naloxone (10−6 mol L−1; Wako),41 FK888 (10−6 mol L−1; Tocris Bioscience, Ellisville, USA),42 ondansetron (10−5 mol L−1; Hikari Pharmaceutical, Imado, Japan)43 and phentolamine mesylate (10−5 mol L−1; MP Biomedicals, Illkirch, France)39 for 30 min, yohimbine hydrochloride (10−6 mol L−1; Tocris Bioscience)40 for 25 min, ritanserin (10−7 mol L−1; Tocris Bioscience)44 for 1 h, or N-nitro l-arginine methylester (l-NAME; 10−4 mol L−1; Sigma)37 for 15 min. Concentrations of drugs were expressed as final molar concentrations in the bath solution. Ritanserin and FK888 were dissolved in ethanol, and other drugs were dissolved in DW before use. All reagents were prepared for each experiment according to the manufacturer’s instructions.
The results of experiments are expressed as mean ± SEM of more than four separate experiments using whole stomachs. One-way analysis of variance followed by Student’s t-test was used for statistical analysis of data. P < 0.05 was considered significant.
Spontaneous stomach contractile activity in the isolated suncus stomach
The isolated stomach of S. murinus showed spontaneous contraction under a basal 1.0 g resting tension. The rate of spontaneous activity of the stomach was approximately 12–15 cycles per minute in the organ bath, which is almost the same as physiologically phasic contraction in vivo.35 The maximum tension produced by ACh (10−5 mol L−1) was about 5 g in the suncus stomach (Fig. 1A). Cumulative addition of suncus motilin (10−11–10−7 mol L−1) to the organ bath induced contraction of the isolated suncus stomachs in a concentration-dependent manner (Fig. 1B). Suncus motilin-induced maximum contractions were about 90% of that of ACh (10−5 mol L−1), and these contractions were completely abolished by TTX (10−6 mol L−1) treatment (Fig. 1C,D).
We have tried to address the motilin-induced contraction by strip preparation of the suncus stomach, but the strip did not respond to motilin treatment. Considering that the motilin receptor is expressed throughout the suncus stomach (manuscript in preparation) and motilin-induced contractions are neurally mediated in the myenteric plexus, in strip preparation some unintentional breakdown of the neural network may have occurred, causing the contractile response of motilin to disappear. Therefore, it is necessary to study the whole stomach to understand the effect of motilin on suncus gastric contractions, which is also true for the canine stomach.23 Also, considering the rabbit gastric sensitivity to motilin in vitro,8,13 the suncus stomach requires a relatively low concentration of motilin to evoke gastric contractions (as demonstrated by the ability of 10−9 mol L−1 motilin to evoke maximum contractions).
Atropine, a muscarinic receptor antagonist, completely abolished the response to suncus motilin in the stomach and suppressed the spontaneous contractile activity (Fig. 2A). On the other hand, hexamethonium, a nicotinic receptor antagonist, significantly decreased the contractions induced by 10−9, 10−8, and 10−7 mol L−1 motilin but did not affect the spontaneous contractile activity.
We also examined the involvement of adrenergic neurons. Phentolamine, an α receptor antagonist, markedly inhibited the motilin-induced contractions (Fig. 3A) and also decreased spontaneous contractions (data not shown). The motilin-induced spontaneous and tonic contractions were not affected by pretreatment with prazosin, α1 receptor antagonist, or timolol, β receptor antagonist (Fig. 3B,D). However, yohimbine, an α2 receptor antagonist, suppressed the spontaneous contractions (data not shown) and almost completely inhibited the motilin-induced contractions (Fig. 3C).
According to previous studies, 5-hydroxytryptamine (5-HT) is involved in motilin action in the dog GI tract.20 In the present study, ritanserin, a 5-HT2 receptor antagonist (Fig. 4A), did not decrease the motilin-induced contraction at each concentration but markedly suppressed the spontaneous contractile activity (data not shown). In contrast to ritanserin, ondansetron, a 5-HT3 receptor antagonist, significantly suppressed the contraction induced by motilin at 10−9 and 10−8 mol L−1 (Fig. 4B).
Other receptor antagonists and nitric oxide synthase inhibitor
The effects of other receptor antagonists and nitric oxide synthase inhibitors were also investigated for further characterization of the response to motilin. Naloxone, an opiate receptor antagonist, significantly suppressed the contraction induced by a low dose of motilin (10−9 mol L−1; Fig. 4C) but did not affect the spontaneously occurring phasic contractions (data not shown). Neither spontaneously occurring contractions nor motilin-induced contractions were decreased by FK888, a neurokinin 1 (NK1) receptor antagonist (Fig. 4D). Pretreatment with verapamil, an inhibitor of calcium inflow, significantly decreased the motilin-induced contraction (Fig. 4E).
l-NAME, an inhibitor of nitric oxide synthase, potentiated the contraction induced by 10−9.7 mol L−1 motilin, but did not significantly change contractions induced by other concentrations of motilin (Fig. 4F).
The percentages of the maximum contraction induced by 10−7 mol L−1 motilin did not significantly differ for prazosin, ondansetron, ritanserin, naloxone, or l-NAME treated motilin-induced contractions (Table 1). We did not calculate the EC50 value for atropine- and yohimbine-treated contractions, as atropine completely and yohimbine almost completely inhibited such contractions. However, the EC50 value of using these antagonists and inhibitors showed significant inhibition or potentiation on specific doses of motilin-induced suncus gastric contractions (Table 1). Moreover, in the case of hexamethonium, phentolamine, and verapmil pretreatment, both the maximum contraction (%) and EC50 values differed significantly. Therefore, from these EC50 values, it is clear that the pretreatment of hexamethonium, prazosin, phentolamine, ondansetron, ritanserin, naloxone, and verapmil effectively suppressed motilin-induced contractions and that l-NAME induced a significant shift in motilin-induced contractions.
Table 1. The EC50 values and % max (mean values ± SEM) of the concentration–response curves to motilin in the absence and presence of different pharmacological blockers
Motilin (10−11–10−7 mol L−1)
% of Max (mean ± SE)
EC50 (nmol L−1; mean ± SE)
Anta, antagonist; inhib, inhibitor; NS = non significant.
*P < 0.05; **P < 0.01.
74.4 ± 1.5
42.9 ± 3.7
0.29 ± 0.01
3.14 ± 0.58
83.3 ± 2.8
35.8 ± 2.9
0.27 ± 0.00
0.56 ± 0.05
75.4 ± 3.6
78.7 ± 3.6
0.35 ± 0.03
0.34 ± 0.05
89.9 ± 8.5
79.4 ± 6.9
0.44 ± 0.03
2.22 ± 0.59
77.0 ± 9.6
84.1 ± 6.7
0.38 ± 0.07
0.37 ± 0.09
87.8 ± 3.7
77.7 ± 1.9
0.37 ± 0.03
1.04 ± 0.24
70.5 ± 1.1
79.43 ± 1.1
0.49 ± 0.07
0.24 ± 0.04
81.8 ± 5.5
21.6 ± 4.6
0.52 ± 0.15
1.70 ± 0.37
76.6 ± 6.2
82.1 ± 8.8
0.89 ± 0.31
0.87 ± 0.24
In the present study, we used the stomach of S. murinus, a new model animal for motilin study, and revealed the mechanisms of motilin-induced contraction by a pharmacological in vitro method. To investigate the neural network in the enteric nervous system of motilin’s action, we examined the effects of various receptor antagonists and a nitric oxide synthase inhibitor on motilin-induced contraction and characterized the pharmacological properties in the suncus stomach in vitro.
The present study clearly showed a concentration-dependent (10−9 mol L−1 to 10−7 mol L−1) contractile effect of motilin on the isolated suncus whole stomach. Although various studies have revealed the importance of vagal nerves in GI motility in vivo,45–47 an ex vivo study showed that motilin-evoked contraction in the perfused dog stomach in a vagus-independent manner.23 Therefore, the relationship between the vagus nerve and MMC in the gastric body remains obscure. In a recent in vivo study using suncus, we observed that vagal innervation is not essential for either the occurrence of MMC or exogenous motilin-induced gastric contraction (manuscript in preparation), indicating that the vagus nerve is not essential for the occurrence of motilin-induced contraction in the suncus stomach. Our results also showed that the response of isolated whole suncus stomach in the organ bath to motilin treatment was completely blocked by TTX and atropine pretreatment, suggesting that motilin-induced suncus gastric contraction is mediated in the myenteric cholinergic neural network as found in dogs.23 It is now well established that one of the major stimulatory pathways of GI motility is through myenteric neurons,8,9,16,25 and it has been reported that the motilin receptor was found in neural structures of the human GI tract.20 We previously demonstrated that in vivo motilin-induced suncus gastric contractions were completely abolished by atropine pretreatment.35 Taken together, the results indicate that motilin-induced suncus gastric contraction is mediated through the myenteric plexus and that the final mediator to evoke contraction in smooth muscle is cholinergic neurons. Moreover, in the case of independent administration of different doses of motilin, the tonic response induced by motilin administration and the concentration–response curve showed that the motilin-induced contraction most likely has a flip-flop shape, as found in experiments testing its cumulative administration (Figure S1). Therefore, it appears that the motilin-induced contractions in the present study does not progress directly to the smooth muscle, but rather are neurally mediated.
It has been demonstrated in an ex vivo experiment that motilin-induced GI contraction is mediated through muscarinic and nicotinic cholinergic receptors in dog.23 We found that atropine completely inhibited the motilin-induced suncus gastric contraction and that hexamethonium, a ganglion blocking agent, significantly suppressed motilin’s action, suggesting that myenteric preganglionic cholinergic neurons and postganglionic cholinergic neurons are important for motilin-induced contraction in the suncus stomach. However, in contrast to the complete inhibition achieved by atropine treatment, hexamethonium showed a relatively weak inhibition of muscle contraction, especially at a high dose of motilin, suggesting that the stimulatory effect of motilin on postganglionic cholinergic neurons is mediated not only by preganglionic cholinergic neurons but by other neurons as well. However, we recently found that hexamethonium did not completely abolish the responses of the nicotinic stimulant (data not shown), indicating that hexamethonium treatment resulted in only partial inhibition. Therefore, it is difficult to draw a final conclusion regarding the involvement of the preganglionic non-cholinergic pathway in motilin-induced suncus gastric contractions. Conversely, we examined the combinatory pretreatment of hexamethonium, ondansetron, and naloxone and found that the motilin-induced contraction was almost completely inhibited by pretreatment with these antagonists (Figure S2), which suggests that hexamethonium-insensitive contractions in response to motilin were blocked by ondansetron and naloxone. These results indicate that, in the motilin-induced suncus gastric contraction, the existence of the preganglionic cholinergic, serotonergic, and opioidergic neuron is important.
Consequently, the notable inhibitory effect of phentolamine and yohimbine indicate that α2 receptors are a possible pathway for motilin-induced contractions in the suncus stomach. However, other subtypes of adrenergic receptors, α1 and β receptors, may not be involved in the motilin-induced suncus gastric contraction. It has also been demonstrated that serotonergic neurons are important for the regulation of phase III contraction in the dog stomach.48,49 In the present study, two different serotonergic receptor antagonists (5-HT2 and 5HT3 receptors) were used. These findings suggest that the 5-HT3 receptor is mainly involved in the motilin-induced suncus gastric contraction, as ondansetron treatment exerted an inhibitory effect on motilin-induced gastric contractions. A similar inhibitory effect of a 5-HT3 receptor antagonist on motilin-induced contraction has also been reported in dogs,23,50 On the other hand, a 5-HT2 receptor antagonist did not affect the motilin-induced contraction, are considered to be independent of motiln’s action.
The significant suppression of motilin-induced suncus gastric contraction by naloxone in our study suggests the involvement of opioid receptors in the myenteric plexus, although Mizumoto et al.23 reported that opioid receptors are not involved in motilin-induced contraction in the dog. However, Fox-Threlkeld et al.51 found that opiate receptors are involved in the occurrence of motilin-induced contractions in the canine isolated ileum. This difference in results regarding opioid receptor involvement in motilin-induced gastric contraction may be due to the species and organ used. We observed that NK1 receptors do not mediate motilin-induced suncus gastric contraction; however, Shiba et al.52 reported that NK1 receptors are involved in the stimulator pathway of motilin, and probably the variation of this participation of NK1 receptors in motilin-induced contraction might be due to the difference between in vivo and in vitro experiments or species, and it would be interesting to clarify this by in vivo study of suncus.
It has been reported that external Ca2+ influx through L-type Ca2+ channels mediate motilin-induced contraction in the rabbit.10,53 Kitazawa et al.9 also showed that extracellular Ca2+ is important for motilin-induced contractile response in the chicken. Our results indicate that motilin-induced gastric contraction in suncus is also mediated by the influx of external Ca2+. On the other hand, Kitazawa et al.37 observed that l-NAME potentiated the contractile response to motilin through reduction of NO-mediated presynaptic inhibition of neural ACh release in the chicken proventriculus. In our study, l-NAME also significantly potentiated contraction induced by motilin treatment. Functional and morphological evidence has also suggested that NO neurons exist in the myenteric plexus54,55 and dorsal motor nucleus,56,57 and both of these are important sites for the regulation of GI motility, together suggesting the involvement of NO neurons in motilin-induced contraction through the myenteric plexus of the suncus gastric body.
In view of physiology or pharmacology, it is an upcoming issue to make a clear description of motilin action in the stomach and its mechanisms of sensitization to each receptor agonist or antagonist for exploiting the motilin as a potential drug target. In this respect, showing the involvement of postganglionic neurons and preganglionic cholinergic neurons and serotonergic neurons, adrenergic neurons, opioidergic neurons, and NO neurons in the myenteric plexus would provide useful information for elucidating the mechanism of motilin-induced contraction. Moreover, this motilin and ghrelin-producing animal has been proved to be a useful laboratory animal for motilin study from receptor level to whole body physiology.
Acknowledgements and disclosures
This work was supported by a Grant-in-Aid for Scientific Research (C) from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) KAKENHI (21590785). We thank Mr. Zuoyun Xie and Mr. Yoshiaki Shimada for their technical assistance. Part of this work has been presented at Joint International Meeting of Neurogastroenterology and Motility, 2009.58
AM performed the research, analyzed the data and wrote the paper; YK performed the research, analyzed the data; TY & CT performed the research; TS & IS designed research study; SO & TT contributed the animals; TS approved the final version to be published.