Characterization of the gastric motility response to human motilin and erythromycin in human motilin receptor-expressing transgenic mice

Motilin is a gastrointestinal peptide hormone that stimulates gastrointestinal motility. Motilin is produced primarily in the duodenum and jejunum. Motilin receptors (MTLRs) are G protein-coupled receptors that may represent a clinically useful pharmacological target as they can be activated by erythromycin. The functions of motilin are highly species-dependent and remain poorly understood. As a functional motilin system is absent in rodents such as rats and mice, these species are not commonly used for basic studies. In this study, we examine the usefulness of human MTLR-overexpressing transgenic (hMTLR-Tg) mice by identifying the mechanisms of the gastric motor response to human motilin and erythromycin. The distribution of hMTLR was examined immunohistochemically in male wild-type (WT) and hMTLR-Tg mice. The contractile response of gastric strips was measured isometrically in an organ bath, while gastric emptying was determined using phenol red. hMTLR expression was abundant in the gastric smooth muscle layer. Interestingly, higher levels of hMTLR expression were observed in the myenteric plexus of hMTLR-Tg mice but not WT mice. hMTLR was not co-localized with vesicular acetylcholine transporter, a marker of cholinergic neurons in the myenteric plexus. Treatment with human motilin and erythromycin caused concentration-dependent contraction of gastric strips obtained from hMTLR-Tg mice but not from WT mice. The contractile response to human motilin and erythromycin in hMTLR-Tg mice was affected by neither atropine nor tetrodotoxin and was totally absent in Ca2+-free conditions. Furthermore, intraperitoneal injection of erythromycin significantly promoted gastric emptying in hMTLR-Tg mice but not in WT mice. Human motilin and erythromycin stimulate gastric smooth muscle contraction in hMTLR-Tg mice. This action is mediated by direct contraction of smooth muscle via the influx of extracellular Ca2+. Thus, hMTLR-Tg mice may be useful for the evaluation of MTLR agonists as gastric prokinetic agents.

Motilin is a gastrointestinal peptide hormone that stimulates gastrointestinal motility. Motilin is produced primarily in the duodenum and jejunum. Motilin receptors (MTLRs) are G protein-coupled receptors that may represent a clinically useful pharmacological target as they can be activated by erythromycin. The functions of motilin are highly speciesdependent and remain poorly understood. As a functional motilin system is absent in rodents such as rats and mice, these species are not commonly used for basic studies. In this study, we examine the usefulness of human MTLR-overexpressing transgenic (hMTLR-Tg) mice by identifying the mechanisms of the gastric motor response to human motilin and erythromycin.
The distribution of hMTLR was examined immunohistochemically in male wild-type (WT) and hMTLR-Tg mice. The contractile response of gastric strips was measured isometrically in an organ bath, while gastric emptying was determined using phenol red.
hMTLR expression was abundant in the gastric smooth muscle layer. Interestingly, higher levels of hMTLR expression were observed in the myenteric plexus of hMTLR-Tg mice but not WT mice. hMTLR was not co-localized with vesicular acetylcholine transporter, a marker of cholinergic neurons in the myenteric plexus. Treatment with human motilin and erythromycin caused concentration-dependent contraction of gastric strips obtained from hMTLR-Tg mice but not from WT mice. The contractile response to human motilin and erythromycin in hMTLR-Tg mice was affected by neither atropine nor tetrodotoxin and was totally absent in Ca 2+ -free conditions. Furthermore, intraperitoneal injection of erythromycin significantly promoted gastric emptying in hMTLR-Tg mice but not in WT mice.
Human motilin and erythromycin stimulate gastric smooth muscle contraction in hMTLR-Tg mice. This action is mediated by direct contraction of smooth muscle via the influx of extracellular Ca 2+ . Thus, hMTLR-Tg mice may be useful for the evaluation of MTLR agonists as gastric prokinetic agents. PLOS

Introduction
Motilin, a 22-amino acid peptide, was firstly identified in 1973 from porcine intestine [1]. The most important physiological function of this hormone is the stimulation of gastrointestinal motility. Motilin is thought to be primarily involved in gastric phase III activity of the migrating motor contraction (MMC) observed during the fasting state. Indeed, several studies have shown that plasma motilin concentrations are increased in parallel with this contractile response in the stomach [2,3]. In addition, exogenous administration of motilin caused MMC phase III-like contractions, while gastric phase III contractions were inhibited by anti-motilin serum and administration of a motilin antagonist [4][5][6].
The receptor for motilin (MTLR) was identified in 1999 from human stomachs and was previously known as GPR38 [7]. MTLR is a G protein-coupled receptor and has 52% sequence homology with ghrelin receptors [8]. Previous studies have demonstrated that MTLR is expressed in gastrointestinal muscle cells and the myenteric plexus, but not in the mucosa and submucosa of the gastrointestinal tract in humans [9,10]. These findings are consistent with results that suggest that the motilin-induced motor response is mediated through direct action on smooth muscle cells and the activation of cholinergic pathways [11][12][13]. Furthermore, the antibiotic drug erythromycin has been shown to stimulate MTLRs and mimic the motor response observed following treatment of motilin [14,15].
The actions of motilin are known to be highly species-dependent. In addition to humans, pigs, and dogs, motilin has been identified in rabbits, cows, cats, sheep, horses and shrews; in contrast, a functional motilin system is absent in rodents such as mice, rats, and guinea pigs [16][17][18][19][20][21][22]. A pseudogene of MTLR has been identified in rodents that fails to respond to motilin, which may be a result of differences in the anatomy and physiology of rodents; this is corroborated by the inability of these organisms to vomit [20,21]. Further, since these rodents are commonly used for basic science studies of many pathways, the lack of a motilin system prevents their use for elucidating the basic physiological and pathophysiological functions of motilin. It also means these models are not useful for the screening of MTLR agonists.
We have previously produced human MTLR-overexpressing transgenic (hMTLR-Tg) mice. hMTLR-Tg mice had an increased intake of food and water during human motilin and erythromycin treatment when compared to wild-type (WT) mice [23]. However, the effect of human motilin and erythromycin treatment on gastric motor activity in hMTLR-Tg mice is yet to be examined.
In order to identify the usefulness of hMTLR-Tg mice, in this study, we examined and characterized the gastric motor response following human motilin and erythromycin treatment in hMTLR-Tg mice.

Animals and ethics statement
This study was carried out in strict accordance with the recommendations found in the Guide for Care and Use of Laboratory Animals (8 th Edition, National Institutes of Health). The protocols were approved by the committee on the Ethics of Animal Research of the Kyoto Pharmaceutical University (Permit Number: 16-014). Male C57BL/6 mice (8-10 weeks) weighing 22-26 g were purchased from Japan SLC Inc. (Shizuoka, Japan). Human motilin receptoroverexpressing transgenic (hMTLR-Tg) mice were generated on an C57BL/6 mice background as described previously [23]. All mice were maintained in plastic cages with free access to food and water. All animals were housed at 22 ± 1˚C with a 12-hour light/dark cycle.

Drugs
Human motilin was purchased from Anygen Co., Ltd (GwangJu, Korea). Erythromycin was purchased from Mylan (Tokyo, Japan). Acetylcholine (ACh), atropine and tetrodotoxin (TTX) were purchased from Wako Pure Chemical (Osaka, Japan). For in vivo experiments (gastric emptying), erythromycin was diluted in physiological salt solution. The dose of erythromycin was expressed as mg/kg body weight. For in vitro experiments (contraction), erythromycin was diluted in, and other drugs were dissolved in, physiological salt solution. Drug concentrations were expressed as final molar concentrations in the organ bath solution. All drugs were prepared immediately before use.

Immunohistochemistry
Tissue preparation and immunohistochemical procedures were performed as described previously [24,25]. After 18 hours of fasting, the stomach was removed. For glass-mounted sections, tissues were fixed by immersion in 4% paraformaldehyde (PFA; in 0.1 M phosphate buffer) for 2 hours at 4˚C, cryoprotected in 0.1 M 20% sucrose (in phosphate buffer) overnight, and subsequently frozen in OCT compound (Sakura Finetec, Tokyo, Japan). The preparation was then cut into 30 μm sections using a cryostat (CM3050, Leica Biosystems, Nussloch, Germany) and mounted onto Superfrost Plus slides (Matsunami Glass, Osaka, Japan). For wholemounted sections, the mucosal and submucosal layers were gently removed under a dissecting microscope and the muscularis preparations were fixed in 4% PFA (in 0.1 M phosphate buffer) for 30 min at 4˚C, and cryoprotected in 0.1 M 20% sucrose in phosphate buffer overnight. Glass-mounted and whole mounted sections were treated with 10% donkey serum containing 0.2% Triton X-100, in phosphate buffered saline (PBS), for 1 hour. Subsequently, the preparations were probed for 40 hours at room temperature with rabbit anti-human motilin antibody (MBL, Aichi, Japan), goat anti-vesicular acetylcholine transporter (VAChT) (Phoenix Pharmaceuticals, Burlingame, CA, USA), and Alexa Flour 488-conjugated mouse anti-alpha smooth muscle actin (α-SMA) antibody (Abcam, Cambridge, MA, USA). After washing in PBS, preparations were labeled for 4 h at room temperature with donkey anti-rabbit IgG (conjugated to Alexa Fluor 488 or 594) and anti-goat IgG (conjugated to Alexa Fluor 594) (1:400, Thermo Fisher Scientific, Waltham, MA, USA). Preparations were imaged on a confocal microscope (A1R+, Nikon, Tokyo, Japan) using NIS-Elements AR 4.20,00 software (Nikon) to capture images, project and reconstruct multiple images in Z-stacks.

Determination of gastric contractile response in vitro
After 18 hours of fasting, the stomach was removed and immediately placed in a freshly prepared normal physiological salt solution (136.9 mM NaCl, 5 mM KCl, 1.5 mM CaCl 2 , 1 mM MgCl 2 , 23.8 mM NaHCO 3 , 5.5 mM glucose, and 0.01 mM ethylene-diamine tetraacetic acid (EDTA); pH 7.4). Ca 2+ -free solutions were prepared by removing CaCl 2 and adding 0.5 mM EGTA in a normal physiological salt solution. Muscle strips (3×5 mm) were obtained from the fundus, corpus, and antrum regions along the circular axis, and then suspended in an organ bath filled with physiological salt solution at 37˚C in an atmosphere of 95% O 2 :5% CO 2 . Contractile responses were measured isometrically under a resting tension of 10 mN and this data recorded (Pantos Co., Ltd, Kyoto, Japan).
Tesque, Kyoto, Japan) containing 0.05% phenol red (Wako Pure Chemical) orally at a volume of 0.5 mL/mouse. Ten minutes later, mice were sacrificed by cervical dislocation, their abdomen was opened, and the gastroesophageal junction and pylorus were clamped. Then, the stomach was removed, placed in 0.1 M NaOH and homogenized. The suspension was allowed to settle for 1 hour at room temperature, after which 2% trichloroacetic acid (final concentration) was added and samples were centrifuged at 2,500 g for 10 min. The supernatant was mixed with 0.05 M NaOH, and the amount of phenol red was determined colorimetrically at 560 nm using a microplate reader (Multiskan GO, Thermo Fisher). The standard sample (0%) was determined by the amount of phenol red recovered from mice sacrificed immediately after oral administration of 1.5% CMC containing 0.05% phenol red. Gastric emptying rate was calculated according to the following equation; Gastric emptying rate ð%Þ ¼ ð1À amount of test sample=amount of standard sampleÞ �100:

Statistical analysis
Data are presented as mean ± SEM. Statistical analysis was performed with GraphPad Prism 6.0h (GraphPad Software, La Jolla, CA, USA) using a Student t-test and one-way or two-way ANOVA followed by Bonferroni's test. We considered p values less than 0.05 to represent statistically significant results.

Distribution of hMTLR in the stomach
To confirm the presence of hMTLR in the stomach of hMTLR-Tg mice, immunohistochemical analysis was performed on hMTLR-Tg and WT mice using an anti-hMTLR antibody. In glass-mounted sections, hMTLR-positive immunostaining was not detected in the stomach obtained from WT mice (Fig 1A and 1B). However, in contrast, hMTLR-positive immunostaining was observed in the muscle layer, with higher levels of hMTLR-positive immunostaining observed in the myenteric plexus of stomachs obtained from hMTLR-Tg mice. High levels of hMTLR expression were also found in gastric epithelial cells of hMTLR-Tg mice. To characterize the localization of hMTLR, double immunostaining using an anti-hMTLR antibody with either an anti-VAChT antibody (a marker of cholinergic neurons) or an anti-α-SMA antibody (a marker of smooth muscle) was performed in the stomachs of WT and hMTLR-Tg mice. VAChT staining was detected in neuronal fibers of the smooth muscle layer and myenteric plexus of both WT and hMTLR-Tg mice ( Fig 1A). The expression of VAChT in the myenteric plexus mostly colocalized with hMTLR-positive staining. Positive immunostaining for α-SMA was also identified in the smooth muscle layer of both WT and hMTLR-Tg mice ( Fig 1B). Double immunostaining revealed the colocalization of α-SMA and hMTLR in muscle layer of hMTLR-Tg mice.
To further investigate the expression of hMTLR in the myenteric plexuses of hMTLR-Tg mice, double immunostaining of anti-hMTLR antibody was performed. In these experiments, alongside an anti-MTLR antibody, anti-VAChT antibody and anti-α-SMA antibodies were used in whole mounted sections of the stomach from hMTLR-Tg mice. Although the colocalization of hMTLR and VAChT was observed in the myenteric plexus, hMTLR-positive staining was not consistent with VAChT-positive staining in the myenteric plexus.

Gastric contractile response to human motilin and erythromycin
To examine the gastric contractile response to human motilin, human motilin (10 −10 -10 −6 M) and erythromycin (10 −9 -10 −5 M) were introduced into an organ bath and the contractile amplitude of the gastric fundus, corpus and antrum was obtained from WT and hMTLR-Tg mice. Although human motilin had no influence on smooth muscle tension in WT mice, motilin caused contractile responses at all sites examined in a concentration-dependent manner (Fig 2A and 2B). Statistically significant differences in the contractile response to human motilin were observed between WT and hMTLR-Tg mice at concentrations of >10 −8 in the fundus, and >10 −7 in the corpus and antrum. In terms of efficacy, the order of the contractile response to human motilin was fundus > corpus > antrum. Likewise, erythromycin caused a concentration-dependent contractile response at all sites obtained from hMTLR-Tg mice but not from WT mice (Fig 3A and 3B). Statistically significant differences in the contractile response to erythromycin between WT and hMTLR-Tg mice were observed at concentrations of >10 −6 in the fundus and antrum, and >10 −7 in the corpus. In terms of efficacy, the order of the contractile response to erythromycin was also fundus > corpus > antrum. The contractile response of gastric strips was more potent to human motilin than erythromycin. Gastric contractile response in human motilin receptor-transgenic mice

Gastric contractile response to ACh
The contractile responses to ACh in the gastric fundus, corpus, and antrum were examined in WT and hMTLR-Tg mice. ACh (10 −5 M) caused visible contractile responses in the gastric fundus, corpus, and antrum that were observed in both, the WT and hMTLR-Tg mice (Fig 4). The gastric contractile responses in WT and hMTLR-Tg mice were almost similar. The order of contractility was as follows: fundus > corpus > antrum.

Characterization of gastric contractile response to human motilin and erythromycin in hMTLR-Tg mice
We next sought to characterize the gastric contractile response to human motilin and erythromycin in hMTLR-Tg mice. The effect of atropine, tetrodotoxin and Ca 2+ -free conditions on motilin and erythromycin-dependent contractile responses in the gastric corpus obtained from hMTLR-Tg mice were examined. Human motilin (10 −6 M) caused visible contractile response in the gastric fundus of hMTLR-Tg mice (Fig 5A and 5B). This contractile response was not diminished by co-treatment with atropine (10 −6 M) or tetrodotoxin (10 −6 M). However, the contractile response was completely abrogated in Ca 2+ -free conditions. Likewise, the erythromycin-induced contractile response was not affected by atropine or tetrodotoxin, but totally abolished in Ca 2+ -free conditions.

Effect of erythromycin on gastric emptying in WT and hMTLR-Tg mice
Finally, we examined the effect of erythromycin on gastric emptying in hMTLR-Tg mice in vivo. Gastric emptying rates in WT and hMTLR-Tg mice were 53.5 ± 7.8% and 54.4 ± 5.1% at 10 min after oral administration of 1.5% CMC containing 0.05% phenol red, respectively (Fig 6). There were no significant differences in gastric emptying between WT and hMTLR-Tg mice. Intraperitoneal injection of erythromycin (10 mg/kg) had no effect on gastric emptying rate in WT mice (53.2 ± 5.9%) but significantly promoted gastric emptying rate to 70.3 ± 4.4% in hMTLR-Tg mice.

Discussion
Motilin is a peptide hormone that stimulates gastrointestinal motor activity via actions at MTLR receptors. Rodents, such as rats and mice, fail to respond to motilin due to lack of functional a signaling pathway and MTLR receptors [20,21]. This fact impedes progress of motilin Gastric contractile response in human motilin receptor-transgenic mice research because these rodents are commonly used for other basic scientific studies. We have previously produced an hMTLR-Tg mouse that has hMTLR cDNA incorporated into the transgene; hMTLR mRNA was expressed in the gastrointestinal tract, liver, and skin of hMTLR-Tg mice [23]. In this study, we examined the expression of hMTLR and the observed motor activity in response to human motilin and erythromycin in hMTLR-Tg mice.
Immunohistochemical analyses identified enhanced expression of hMTLR in the smooth muscle layer, myenteric plexus, and stomach mucosa of hMTLR-Tg mice but not in WT mice. Takeshita et al. [10] reported that mRNA expression of MTLR were detectable in the human gastrointestinal tract. Further, they showed MTLR-immunoreactivity in the smooth muscle layer and myenteric plexus, but not the mucosa and submucosa. A similar result was reported by Ter Beek et al. [27], although in this instance, they found expression of MTLR even in the mucosa. Although the expression of MTLR in the mucosa in basic studies is controversial, the expression of MTLR in the smooth muscle layer and myenteric plexus of hMTLR-Tg mice is similar to that in humans. We further observed that hMTLR in the myenteric plexus was not colocalized with VAChT, a maker of cholinergic neurons. Several studies have demonstrated that the gastrointestinal motor response to motilin is caused by the direct action of this agent on smooth muscle cells and the activation of cholinergic pathways as evidenced in rabbits and shrew, as well as humans [11-13, 28, 29]. These findings may suggest that a lack of motilininduced motor response in hMTLR-Tg mice via activation of the cholinergic pathway occurs in contrast to the mechanism observed in humans. However, the types of neurons positive for hMTLR in the myenteric plexus of hMTLR-Tg mice are not known. Non-cholinergic neurons Gastric contractile response in human motilin receptor-transgenic mice (without VAChT expression) may be certain inhibitory motor neurons. Further studies are needed to elucidate this point.
The most important finding in this study was that human motilin produced concentrationdependent contractile responses in gastric smooth muscles strips obtained from hMTLR-Tg mice. Of course, human motilin failed to induce contraction in gastric smooth muscles obtained from WT mice. In addition, we show that the efficacy of the contractile response to motilin was gastric fundus < corpus < antrum. An especially apparent contractile response was observed even at a low concentration (10 −10 M) of human motilin in the fundus but not in the corpus and antrum. Broad et al. [29] showed that the efficacy of motilin-induced contractile response was the highest in the gastric antrum compared with that in other regions of the human stomach. In contrast, Dudani et al. [30] reported that the contractile response to motilin in house shrew was more prominent in the gastric fundus than in the antrum. In general, the contractile response in mice to various stimulation such as acetylcholine, serotonin, and electrical field stimulation was more prominent in the gastric fundus than in the antrum [31]. In the present study, we observed that ACh induced contractile responses in hMTLR-Tg and WT mice. Further, the order of the contractile response to ACh was similar to that observed for response to human motilin and erythromycin, that is, gastric fundus > corpus > antrum. Thus, it is likely that the differences in contractile responses according to the stomach region may result from the fundamental differences in the contractile apparatus, depending on the animal species. However, the possibility that these regional differences may have occurred because of differential expression levels of hMTLR could not be excluded. Further detailed studies must be conducted to elucidate this point.
The previous study showed that EC 50 value of motilin to produce gastric contraction in humans is around 10 −9 M [32]. In contrast, the present results revealed lower contractile response to human motilin because human motilin at concentration of 10 −9 M induces minimal contraction in hMTLR-Tg mice. It is likely that a lower contractile response to human motilin in hMTLR-Tg mice compared to that in humans might be because coupling of the contractile apparatus is reduced in hMTLR-Tg mice compared with WT mice.
Similar to human motilin, erythromycin produced a concentration-dependent gastric contractile response, and this response was more pronounced in fundi obtained from hMTLR-Tg mice. However, the contractile response to erythromycin was about one tenth the potency of that to human motilin, as consistent with a previous report in the human stomach [33]. These findings strongly suggest that the gastric contractile response to human motilin and erythromycin in hMTLR-Tg mice closely resembles the response observed in humans.
Interestingly, the acetylcholine muscarinic receptor antagonist atropine, and the neurotoxin tetrodotoxin, failed to attenuate the contractile responses to human motilin and erythromycin in hMTLR-Tg mice. Several studies have demonstrated that motilin-induced motor responses are mediated through direct action on smooth muscles and the activation of cholinergic pathways in dogs, rabbits, shrew and humans [11][12][13]34]. In contrast, Satoh et al. [33] showed that the contractile response induced by an erythromycin-derivative in human gastric antrum was not influenced by atropine and tetrodotoxin. Further, in human and rabbit isolated stomachs, facilitation of cholinergic contraction was caused by low concentrations of motilin and erythromycin, while the higher concentrations directly contracted the muscle [12,[35][36][37]. Thus, it is likely that the gastric contractile response to motilin and erythromycin may be mediated by the different pathways (i.e. either cholinergically or directly stimulated) depending on the concentration of each agent. As mentioned above, while the present study revealed that hMTLR was expressed in the myenteric plexus in addition to smooth muscles, the expression of hMTLR in the myenteric plexus was not colocalized with cholinergic neurons in hMTLR-Tg mice. Thus, the observations of contractile response to human motilin and erythromycin independent of cholinergic pathways fit well with the results obtained from immunohistochemical analyses. These findings suggest that hMTLR-Tg mice reproduce motilin-induced motor activity by direct action on smooth muscle cells and not via the operation of cholinergic pathways.
In humans and rabbits, electrically stimulated gastric contraction was increased by motilin and erythromycin [12,29,37]. Thus, motilin may facilitate cholinergic pathways to produce gastric contraction. However, the immunohistochemical study revealed that hMTLR was not expressed in cholinergic neurons of the stomach obtained from hMTLR-Tg mice. Since the influence of motilin on cholinergic pathways could not be excluded, further studies are needed to clarify this point.
Several studies have demonstrated that the contractile effect of motilin and erythromycin is mediated via intracellular Ca 2+ signaling in smooth muscle of rabbits, dogs and humans [38][39][40][41]. Further, Van Assche et al. [41] showed that motilin-induced increases in intracellular Ca 2+ levels are dependent on Ca 2+ influx; this is in contrast to acetylcholine, which triggers the release from intracellular Ca 2+ stores. In this study, we observed that the gastric contractile response to human motilin and erythromycin in hMTLR-Tg mice was totally abolished in Ca 2+ -free conditions. Thus, it is likely that human motilin and erythromycin can produce direct contraction of gastric smooth muscle cells in hMTLR-Tg mice through activation of Ca 2+ influx via MTLR.
Finally, we examined the effect of erythromycin on gastric motor activity in vivo by evaluating gastric emptying. Intraperitoneal injection of erythromycin significantly promoted gastric emptying in hMTLR-Tg mice without any effect in WT mice. A previous study has shown that an erythromycin-derivative promoted gastric emptying in dogs [42]. Thus, our evidence suggests that hMTLR stimulation can promote gastric motor activity via a contractile response of smooth muscles in hMTLR-Tg mice. For future studies, we plan to investigate the prokinetic effect of human motilin and erythromycin in pathological conditions. Going forward, the hMTLR-Tg model will be a useful tool to study the interactions between the motilin and ghrelin pathways.

Conclusions
In the stomach of hMTLR-Tg mice, hMTLR was primarily expressed in smooth muscle and the myenteric plexus, but in the latter structure, not colocalized with cholinergic neurons. Human motilin and erythromycin promoted a gastric motor response both in vivo and in vitro conditions. This action is mediated by the direct contraction of smooth muscle via influx of extracellular Ca 2+ . Thus, hMTLR-Tg mice may be a useful model for the evaluation of MTLR agonists as gastric prokinetic agents.