Motilin stimulates food intake linked to gastric motility in Suncus murinus: Simultaneous recordings of food intake and gastric motility in the conscious state

Significance

Motilin, a peptide hormone produced in the proximal small intestine, peaks during gastric phase III contractions in the interdigestive state and is considered a potential hunger hormone. However, the mechanism whereby motilin modulates food intake during gastric contractions remains unclear. In Suncus murinus, a small motilin-producing mammal, motilin stimulates food intake via the vagus nerve during phase I contractions. Furthermore, motilin activates feeding-related neurons, including tyrosine hydroxylase (TH) neurons in the hindbrain area postrema (AP) and nucleus of the solitary tract (NTS), as well as neuropeptide Y (NPY) and TH neurons in the hypothalamic arcuate nucleus (ARC). These physiological findings indicate that motilin regulates feeding behavior in response to gastrointestinal conditions.

Abstract

The intestinal-derived hormone motilin, a peptide belonging to the ghrelin family, induces strong gastric contractions and hunger sensations. However, whether motilin regulates food intake and its association with gastric motility remains unclear. Rodents are unsuitable for animal studies of motilin function, as both the motilin and motilin-receptor genes exist only as pseudogenes in their genomes. In this study, we investigated the role of motilin on food intake by simultaneously monitoring gastric contractions and their central mechanisms in the house musk shrew (Suncus murinus), a small mammal that produces motilin. In the interdigestive state, plasma motilin concentrations were elevated during spontaneous phase III contractions of the migrating motor complex (MMC) than during phase I contractions. Food intake during spontaneous phase III contractions was higher than that during phase I contractions. Intravenous motilin administration stimulated food intake during phase I, although its effect was weaker than that of ghrelin. Motilin-induced food intake was abolished in vagotomized suncus. Additionally, motilin increased c-Fos expression in tyrosine hydroxylase (TH) neurons in the area postrema and nucleus of the solitary tract of the brain stem, as well as in activated neuropeptide Y and TH neurons in the arcuate nucleus of the hypothalamus. These results revealed that motilin stimulated feeding linked to gastric motility through the vagus nerve and activated brain regions associated with food intake. Our findings provide evidence that motilin regulates food intake, highlighting its potential as a therapeutic strategy for appetite disorders.

The migrating motor complex (MMC) is a cyclic motility pattern that occurs during the interdigestive state and is commonly referred to as ‘hunger contractions’ (1). The MMC begins when postprandial contractions (PPCs) end during the feeding period. The MMC comprises three phases. Phase I represents a period of motor quiescence, whereas phase II is characterized by irregular and moderate activity. Notably, phase III contractions—cluster contractions that originate in the antrum or small intestine and migrate distally—are the most critical and essential for clearing the stomach in preparation for the next meal (2). Motilin, a peptide hormone produced in the proximal small intestine, peaks during gastric phase III contractions (3). Recently, motilin has been considered as a hunger hormone in humans, as it elevates hunger ratings during phase III contractions (4). Furthermore, erythromycin (a motilin receptor agonist) induced gastric phase III contractions, increased hunger scores, and stimulated food intake (soup-based meals) in healthy individuals during discontinuous monitoring of contractions (5). However, in another study where gastric motility was not monitored, erythromycin did not increase total caloric intake in humans (excess free-choice buffet) (6). Additionally, few mechanisms and action sites through which motilin stimulates hunger stimulation are known because related animal studies have not been performed. Hence, a need exists for investigating the effects of motilin on food intake and the potential mechanisms associated with gastric contractions.

Gut–brain communication plays a crucial role in regulating food intake. Importantly, gastric vagal afferent nerves transmit sensory information to the nucleus of the solitary tract (NTS) and the area postrema (AP) of the brainstem by sensing gastrointestinal (GI) nutritional hormones, such as cholecystokinin (CCK) and glucagon-like peptide-1 (GLP-1) to suppress feeding (7, 8). Furthermore, NTS epinephrine neurons that coexpress neuropeptide Y (NPY) can receive vagal afferents and promote feeding (9). The AP is located adjacent to the NTS, which is mainly exposed to circulating factors and projected to other brain regions, such as the paraventricular nucleus (PVN), dorsomedial hypothalamic nucleus, and dorsal motor nucleus of the vagus (10–12). In turn, hypothalamic agouti-related protein (AgRP) in the arcuate nucleus (ARC) receives ascending projections from NTS norepinephrine neurons to mediate feeding behavior (13). Thus, cooperation between the brainstem and hypothalamus modulates feeding in response to GI tract hormones. The orexigenic hormone ghrelin is mainly released from the stomach during fasting, activates its receptor in the nodose ganglion, and is transmitted to the AP and NTS through the vagal afferent nerve (14). Central administration of ghrelin stimulates feeding through appetite-regulating neurons that express AgRP and NPY (15, 16). However, few animal studies have been conducted to explore the mechanism whereby motilin regulates feeding related to gastric motility because the motilin and its receptor genes exist as pseudogenes in rodents (17).

In this study, the house musk shrew (Suncus murinus; order, Insectivora; suncus named laboratory strain) (18, 19), which has almost identical GI motility and motilin responses to those of humans and dogs, was used as an animal model. We first measured food intake during spontaneous phases I and III of the MMC and identified the role of exogenous motilin on food intake while monitoring gastric motility in the interdigestive state. Next, we investigated whether vagal nerves were involved in motilin-induced feeding. Finally, the motilin-induced neuronal targets in the hindbrain and hypothalamus were examined.

Results

Endogenous Motilin–Induced Spontaneous Phase III Contractions Stimulated Food Intake.

Endogenous motilin levels increase during spontaneous gastric phase III contractions, and hunger scores are strongly associated with these contractions in humans (4, 5). As GI motility and feeding behavior are closely related, simultaneous recordings of food intake and gastric motility in conscious animals are important for understanding their physiological relevance. We measured plasma motilin levels to validate the relationship between motilin fluctuations and different MMC phases in suncus. We found that motilin secretion was higher during phase III contractions than during phase I contractions [F (1.386, 4.157) = 24.16, P = 0.0061; phase III vs. phase I, P = 0.0095; phase III vs. phase II, P = 0.0646; Fig. 1 B and C]. Thus, we monitored gastric motility to investigate whether food intake is regulated by endogenous motilin release. During spontaneous phase I contractions, a small amount of food was consumed by suncus within 30 min; however, food intake increased during spontaneous phase III contractions in the same suncus (P = 0.0366, Fig. 1 D and E). To observe whether the motilin receptor (GPR38) could mediate endogenous motilin release to promote food intake during spontaneous phase III contractions, we administered the motilin receptor antagonist (MA-2029) via intravenous (IV) infusion during spontaneous phase III, which lowered feeding in suncus (P = 0.0080, Fig. 1 F and G).

Fig. 1.

Motilin-induced food intake is associated with phase III contractions in suncus. (A) Representative traces indicating MMC phase I, II, and III contractions during the fasting state, as well as PPCs and PPGCs in the postprandial period. (B) Plasma was collected during phase I–III contractions to measure plasma motilin levels. (C) The plasma motilin concentrations during different MMC phases are shown (n = 4). (D) Typical traces demonstrated that food intake was performed during spontaneous gastric phase I and phase III contractions in the same suncus. (E) Food intake was analyzed within 30 min between spontaneous phase I and phase III (n = 5). (F) Responses to IV infusion of vehicle or MA-2029 (MA, 1 mg·kg⁻¹·h⁻¹, 30 min) during spontaneous phase III contractions and accompanied by performing feeding to suncus. (G) Amount of food intake within 30 min after MA infusion during the phase III contractions (n = 6). (H) Representative traces of gastric contractions after vehicle (0.1% BSA/PBS), motilin (10 μg·kg⁻¹), or ghrelin (10 μg·kg⁻¹) administration, combined with feeding during the fasting state in the same suncus. (I) Effects of exogenous motilin administration on food intake within 30 min (n = 6). *PPGCs or phase III contractions. The data shown are expressed as the mean ± SEM. ^#P < 0.05, ^##P < 0.01.

Motilin Administration Promoted Food Intake during Phase I Contractions of MMC.

We then explored whether exogenous motilin administration affected food intake during phase I to the same extent as it did in spontaneous phase III. First, we observed the effects of different doses of motilin and ghrelin on food intake without monitoring the gastric motility of suncus (SI Appendix, Fig. S1). Therefore, motilin and ghrelin were subsequently administered at a dose of 10 μg·kg⁻¹ to study the effect on food intake during phase I contractions. As shown in Fig. 1H, vehicle administration did not induce contractions, but suncus consumed food, and phase I contractions changed to PPCs, which have characteristics similar to phase II contractions. Ghrelin then stimulated phase II-like contractions and initiated feeding. Food intake significantly increased within 30 min, which is consistent with the results of previous studies on other mammals (20–22). In the same suncus, motilin administration induced phase III-like contractions and promoted food intake compared with those after vehicle administration. However, the effect of motilin on food intake was weaker than that of ghrelin [F (1.532, 7.658) = 20.47, P = 0.0012. Motilin vs. Veh, P = 0.0356; Ghrelin vs. Veh, P = 0.0014; Motilin vs. Ghrelin, P = 0.0989. Fig. 1 H and I].

The Vagus Nerve Mediated Motilin-Induced Food Intake During Phase I Contractions of MMC.

The vagus nerve plays an important role in regulating metabolism through gut–brain crosstalk (9, 23). Therefore, we examined whether the vagal pathway is involved in motilin-induced food intake during phase I contractions, using vagotomized suncus. In sham-operated suncus, motilin administration induced phase III-like contractions and stimulated greater feeding than vehicle administration (P = 0.0210, Fig. 2 A and B). Conversely, gastric contractions in vagotomized suncus showed a longer phase I period and a shorter phase II period, followed by phase III contractions. Although vagotomized suncus exhibited motilin-induced phase III-like contractions, food intake decreased (P = 0.0004, Fig. 2 C and D).

Fig. 2.

Involvement of the vagus nerve in motilin-induced food intake during phase I contractions. (A) In sham-operated suncus, food intake was observed after vehicle or motilin administration while monitoring gastric contractions. (B) Comparison of the effects of motilin or vehicle treatment on feeding within 30 min in sham-operated suncus (n = 6). (C) Typical traces indicating the effects of vehicle and motilin administration on feeding during phase I contractions in vagotomized suncus. (D) Quantified results showing the amount of food consumed within 30 min after vehicle or motilin treatment during phase I contractions in vagotomized suncus (n = 8). *PPGCs or phase III contractions. The data shown are expressed as the mean ± SEM. ^#P < 0.05, ^###P < 0.001.

Motilin Drives Hindbrain Tyrosine Hydroxylase (TH) Neuron Activation during Phase I Contractions.

We hypothesized that motilin regulates neuronal activity in the hindbrain and hypothalamus, the main brain regions responsible for controlling feeding behaviors. While monitoring gastric motility, brains were collected 90 min after vehicle, motilin, or ghrelin administration during phase I contractions, in the absence of feeding (SI Appendix, Fig. S2). First, we examined motilin-activated TH neurons in the hindbrain. The central endings of GI vagal afferents mainly terminate at the NTS, with a small subset terminating at the AP. Within the NTS, catecholamine (CA) neurons can regulate feeding via the vagal pathway (9). As demonstrated in Fig. 3A and quantified in Fig. 3B, 47.39 ± 1.78% of c-Fos-positive neurons coexpressed TH after motilin administration. This proportion was higher than that observed following vehicle (27.35 ± 3.00%) or ghrelin (31.35 ± 5.54%) treatment in the AP [F (2, 12) = 7.877, P = 0.0065. motilin vs. Veh, P = 0.0072; ghrelin vs. Veh, P = 0.7403; motilin vs. ghrelin, P = 0.0276]. Additionally, 16.27 ± 1.76% of double positive neurons for TH and c-Fos were observed in the NTS under motilin administration, representing an increase compared with that following vehicle administration (8.33 ± 1.67%). However, the effect of ghrelin in exciting the TH neurons (AP: 31.35 ± 5.54%, NTS: 11.83 ± 1.03%) did not significantly increase from that observed after vehicle treatment in the AP and NTS [F (2, 12) = 6.862, P = 0.0103; motilin vs. Veh, P = 0.0080; ghrelin vs. Veh, P = 0.2723; motilin vs. ghrelin, P = 0.1385; Fig. 3 C and D].

Fig. 3.

TH neuron activation in hindbrain AP and NTS following motilin administration during phase I contractions. (A) Representative images demonstrating colocalization between TH⁺ neurons (red) and c-Fos⁺ (green) in the area postrema (AP) along the mid axis (distance to bregma: −7.2 to −7.76 mm) following vehicle, motilin, or ghrelin administration. The arrows indicate double-labeled neurons. (B) Quantified results showing the percentage of TH⁺ neurons among AP c-Fos⁺ neurons (n = 5). (C) Representative images revealing colocalization between TH+ neurons (red) and c-Fos⁺ neurons (green) in the nucleus of the solitary tract (NTS) along the mid axis (distance to bregma: −7.2 to −7.76 mm) following vehicle, motilin, or ghrelin administration. The arrows indicate double-labeled neurons. (D) Quantified results showing the percentage of TH⁺ neurons among NTS c-Fos⁺ neurons (n = 5). (Scale bar, 100 μm.) The data shown are expressed as the mean ± SEM. ^#P < 0.05, ^##P < 0.01.

Arcuate NPY and TH Neuron Responses to Motilin.

The hypothalamic ARC is an important center for regulating feeding and energy balance. It receives ascending gut-related projections from the NTS and responds directly to circulating food intake-related factors (24–26). Therefore, we examined whether motilin activated NPY neurons during phase I contractions. Double in situ hybridization chain reaction (isHCR)-based analysis of NPY and immunofluorescence for c-Fos were performed to confirm the motilin-induced activity of NPY neurons. As shown in Fig. 4A, the number of c-Fos-immunopositive cells in the ARC was higher after motilin or ghrelin administration than that following vehicle administration. Furthermore, the percentage of NPY neurons colocalized with c-Fos-immunopositive cells among all c-Fos-positive neurons was increased after motilin administration (56.71 ± 5.45%) than that after vehicle administration (30.74 ± 6.84%). This effect was comparable to that observed after ghrelin administration (67.68 ± 6.79%) [F (2, 11) = 8.873, P = 0.0051; motilin vs. Veh, P = 0.0477; ghrelin vs. Veh, P = 0.0045; motilin vs. ghrelin, P = 0.5037; Fig. 4B].

Fig. 4.

Activation of NPY and TH neurons in the hypothalamic ARC under motilin administration during phase I contractions. (A) Representative images displaying colocalization between NPY⁺ neurons (red) and c-Fos⁺ neurons (green) in the ARC (distance to bregma: −1.32 to −2.8 mm) after vehicle, motilin, or ghrelin administration. The arrows indicate double-labeled neurons. (B) Quantified results showing the percentage of NPY⁺ neurons among ARC c-Fos⁺ neurons (n = 4 to 5). (C) Representative images showing the distribution between NPY⁺ neurons (red) and TH⁺ (green) in the ARC (distance to bregma: −1.32 to −2.8 mm) in suncus (n = 5). (D) Representative images showing colocalization between TH⁺ neurons (red) and c-Fos⁺ neurons (green) in the ARC (distance to bregma: −1.32 to −2.8 mm) after vehicle, motilin, or ghrelin administration. The arrows indicate double-labeled neurons. (E) Quantified results showing the percentage of c-Fos⁺ neurons among ARC TH⁺ neurons (n = 5 to 6). (Scale bar, 100 μm.) The data shown are expressed as the mean ± SEM. ^#P < 0.05, ^##P < 0.01, ^###P < 0.001.

TH neurons in the ARC have been shown to promote food intake, and their distribution was independent of NPY neurons in mice (27). In suncus, we found that TH neurons were mainly expressed in the dorsomedial ARC and that NPY neurons were distributed in the mediobasal area and lateral or ventral borders of the medial ARC, indicating that they were not colocalized in the ARC (Fig. 4C). Therefore, we examined whether TH neurons in the ARC would respond to motilin. Both motilin and ghrelin administration induced the activation of TH-positive neurons (Fig. 4D). The percentage of TH⁺ c-Fos⁺ cells versus TH⁺ neurons was elevated in suncus-administered motilin (48.27 ± 3.03%) or ghrelin (50.66 ± 3.26%) than in those given vehicle-control treatment (26.44 ± 2.58%) [F (2, 13) = 19.09, P = 0.0001; motilin vs. Veh, P = 0.0005; ghrelin vs. Veh, P = 0.0003; motilin vs. ghrelin, P = 0.8372; Fig. 4E].

Discussion

Fluctuations in GI hormone release are associated with gastric motility. Our previous findings suggested that ghrelin receptor antagonists inhibit motilin-induced phase III-like contractions and that ghrelin stimulates the motilin pathway by inhibiting GABAergic neurons (28, 29). Therefore, ghrelin is important for phase II contractions, and the coordination of motilin and ghrelin is necessary to initiate phase III contractions in the MMC. Humans often feel stomach rumblings when they experience hunger, which is attributed to strong gastric phase III contractions. Therefore, studying the involvement of motilin in regulating hunger is important. This study investigated the effects of motilin on food intake with simultaneous measurement of gastric contractions using suncus, a motilin-producing mammal. To the best of our knowledge, the motilin plasma peak occurs exclusively prior to the onset of gastric phase III contractions in humans and is not associated with a duodenal origin. Thus, gastric phase III is closely linked to motilin release, which was the focus of this study. First, we demonstrated that motilin secretion increased during phase III contractions in suncus, consistent with findings in humans (30, 31). Comparing the level of feeding during spontaneous phase I (low motilin levels) and phase III (high motilin levels) contractions showed that food intake was significantly higher during phase III than it was during phase I, indicating that feeding was regulated by endogenous motilin during fasting.

We found that exogenous motilin administration evoked phase III-like contractions and promoted food intake, confirming that the effects of motilin on feeding are linked to gastric contractions. As expected, the orexigenic effect of ghrelin, used as a positive control, was confirmed in suncus, aligning with observations in other mammals. However, this effect was stronger than that observed with motilin treatment, suggesting that differences occur in their mechanisms of action. Although ghrelin was essential for motilin-induced contractions dependent on the gastric ghrelin receptor (GHSR) and motilin receptor (GPR38) (29), our results suggest that motilin and ghrelin receptors (expressed in the central or vagus nerve) involve in regulating food intake. These results indicate that distinct pathways link motilin-induced gastric contractions and food intake. Abnormal motilin levels and the absence of phase III contractions are associated with food intake and gastric-dysmotility disorders. Unlike healthy controls, patients with an unexplained loss of appetite do not exhibit gastric phase III symptoms (4). In severely obese patients, gastric phase III activity and plasma motilin levels were decreased (32). The origin of phase III activity shifts from the antrum to the duodenum, and motilin secretion does not peak before phase III despite its high levels, resulting in no elevated hunger scores during phase III (33). Thus, our findings provide evidence illustrating the importance of peak motilin levels during phase III contractions in stimulating food intake. However, whether motilin administration can alleviate diseases associated with food intake and gastric contractions requires further investigation.

The gastric vagal afferent nerve transmits gut information regarding nutritional status by acting as a physical and chemical sensor from the gut to the brain, and it regulates energy metabolism (34). The dorsal vagal complex of the hindbrain, which receives neural signals from vagal sensory neurons, contains several nuclei crucial for feeding control. In most related studies, the AP and NTS are thought to regulate food intake. GLP-1 agonists activated NTS neurons to suppress feeding, and the stimulation of NTS CCK neuron projections leading to the parabrachial nucleus and PVN of the hypothalamus resulted in reduced food intake (35–38). Ghrelin is the only peptide shown to increase feeding via the GHSR, and it is expressed in the vagal afferent nerve (14, 39), which highlights the importance of understanding how motilin stimulates feeding in suncus during fasting. Previously, we found that GPR38 was expressed in the nodose ganglion (40) and that the vagal afferent nerve is involved in the regulation of gastric motility in suncus (41). Thus, we examined the role of the vagus nerve in motilin-induced food intake, observing that motilin administration significantly inhibited food intake in vagotomized suncus, although gastric phase III contractions were not affected. Our previous data also showed that vagotomy did not affect spontaneous gastric phase III contractions (41), suggesting that motilin-induced phase III contractions directly affect receptors in the stomach and are independent of the vagus nerve. Therefore, motilin regulates gastric contractions and food intake via an independent pathway. Motilin-induced gastric phase III contractions are thought to empty the stomach and prepare it for the next meal, and elevated motilin levels drive food intake.

NTS TH neurons (markers of catecholamine neurons) have been considered to control feeding. TH neuron activation can promote feeding through vagal afferent nerves in mice (9). Additionally, the receptor for the GI tract hormone GLP-1 is expressed in TH neurons of the AP (42), which innervates the caudal NTS (43, 44). Thus, we explored TH neuronal activity in the AP and NTS following peripheral motilin administration and showed that motilin strongly activated TH neurons. Furthermore, vehicle-treated suncus showed activated TH neurons in the NTS, consistent with previous findings showing that fasting excited TH neurons and stimulated feeding in mice (9). The NTS receives most vagal afferents, and the AP receives only a small proportion. However, our results suggest that the NTS and AP help mediate the effects of motilin-activated TH neurons through the vagal pathway. Furthermore, TH neurons comprise two subpopulations with opposing functions: NPY/E^NTS neurons that stimulate feeding and NE^NTS neurons that inhibit feeding (9). Moreover, the expression and neuronal activity of AgRP in the hindbrain increased under fasting conditions (45). Thus, it is necessary to confirm the response of NPY/AgRP neurons in the NTS to motilin and its effect on food intake in suncus.

Similar to the hindbrain, the hypothalamus receives circulatory and neural signals. The arcuate nucleus receives ascending gut-related projections from the NTS [NTS^NE–ARC^AgRP] (13) and responds directly to circulating feeding-related factors such as ghrelin (46, 47). NPY and AgRP colocalize in ARC neurons and are effective orexigenic peptides (48, 49). In this study, NPY neurons in the ARC were stimulated with motilin in the same manner as with ghrelin in suncus. Similar to rodents, ghrelin directly interacts with NPY-containing neurons to activate calcium signaling through protein kinase A and N-type channel-dependent mechanisms in the rat ARC (50). In contrast, ghrelin excites NPY neurons via vagal afferent nerves (14). Ghrelin is the only known peripheral peptide orexigenic hormone. In this study, we demonstrated that motilin is a peripheral orexigenic peptide that targets brain regions following peripheral motilin administration. To determine whether motilin and ghrelin result in different behavioral outcomes, histologically detailed mapping of GPR38 and GHSR is necessary to understand whether both hormones act synergistically, additively, or independently. Elucidating their precise localization will advance the understanding of the neural mechanisms underlying motilin-mediated feeding regulation.

The ARC contains a distinct population of TH neurons, which can be triggered by ghrelin, mediating its direct effects on feeding (27). Similar to ghrelin, motilin increased c-Fos expression in TH neurons of the suncus ARC. However, TH neuron activation also occurred in vehicle-administered suncus during the fasting state, consistent with the finding that fasting activated TH neurons in the ARC of mice. Although TH neurons can release dopamine, they did not stimulate NPY neurons because of the lack of synaptic connections between NPY and TH neurons (27). Blood–brain barrier leaking in the ARC increased the potential that blood-derived signals from the GI tract reached the ARC (51); thus, motilin might act directly on NPY and TH cells in the ARC. However, peripheral motilin administration does not stimulate food intake in the absence of the vagal pathway. Determining whether the central administration of motilin can directly stimulate food intake will help to clarify its central effects.

In conclusion, motilin, a hunger-promoting hormone released from the GI tract during phase III contractions in Suncus murinus, stimulates food intake through the vagal pathway. This finding may relate to motilin’s stimulation of brainstem AP^TH and NTS^TH neurons and hypothalamic ARC^NPY and ARC^TH neurons. The orexigenic role of motilin provides crucial insights into how gut hormones coordinate food intake and gastric motility.

Materials and Methods

Animals.

Adult female Suncus murinus (8 to 24 wk old) of an outbred KAT strain established from a wild population in Kathmandu, Nepal. The weight of the animals ranged from 45 to 70 g. They were housed individually under controlled conditions (23 ± 2 °C; lights on from 8:00 AM to 8:00 PM) in plastic cages equipped with an empty can as a nest box, with free access to water and commercial trout pellets (No. 5P; Nippon Formula Feed Manufacturing). The metabolizable energy content of the pellets was 344 kcal/100 g, and the pellets consisted of 54.1% protein, 30.1% carbohydrates, and 15.8% fat. All procedures were approved and performed in accordance with the guidelines of the Saitama University Committee on Animal Research (R5-A-1-12).

Animal Surgery.

After fasting for 2 to 3 h, the animals were anesthetized via intraperitoneal (IP) injection of a mixture containing the psychotropic drug midazolam (4 mg·kg⁻¹; Sandoz K.K., Yamagata, Japan), the pain relieving and suppressing drug domitor (0.3 mg·kg⁻¹; Nippon Zenyaku Kogyo Co., Fukushima, Japan), and vetorphale, a drug for postoperative pain relief (5 mg·kg⁻¹; Meiji Seika Pharma Co., Tokyo, Japan).

Transducer implantation and internal jugular vein catheterization were performed as previously described (29). Briefly, to monitor gastric contractions, strain-gauge force transducers were developed in our laboratory using a small strain gauge (KFG-02-120-C1-11N30C2; Kyowa Electronic Instruments). Before implantation, all transducers were evaluated for waterproof integrity and their response properties. A strain-gauge force transducer was implanted on the serosal surface of the gastric body via a midline laparotomy. The wires from the transducers were exteriorized through the abdominal wall and subsequently routed subcutaneously toward the back of the neck, where they were secured.

Jugular vein cannulation was performed for IV drug delivery. A silicone tube (1.0 mm outside diameter × 0.5 mm inside diameter; Kaneka Medics, Osaka, Japan) was inserted into the left or right jugular vein, passed under the skin on the left or right side of neck, and secured at the back of the neck. Heparinized saline (100 U/mL) was administered via catheter to prevent coagulation.

In vagotomy experiments, we performed truncal vagotomy in the suncus, which had undergone both transducer implantation and internal jugular vein catheterization, to investigate the role of the vagus nerve in food intake following vehicle or motilin administration, as described (52). The stomach and lower esophagus were exposed; the liver, pancreas, and blood vessels remained unharmed; and the dorsal and ventral vagus nerves were isolated. Both branches of the vagus nerve were cut, and all neural connections in the resected area were completely peeled away by wiping the tissues (Kimwipes; Nippon Paper Crecia, Tokyo, Japan). In sham-operated suncus, the vagus nerve was exposed without incision or dissection. The animals were given ad libitum access to food pellets after surgery, and gastric motility was recorded after at least 2 d.

Simultaneous Monitoring of Food Intake and Gastric Motility.

The gastric motility of free-moving suncus was recorded during fasting. The wire from the transducer to the amplifier and the amplified analog signals were converted using an analog-to-digital converter (model ADC-24; Pico Technology Ltd.), after which the digital signals were recorded with PicoLog software (Pico Technology, Ltd.) using a sampling interval of 100 ms. The gastric-contractions data were processed with a frequency cutoff of 0.8 Hz using LabChart 8 Reader software (ADInstruments, Dunedin, New Zealand). Spontaneous GI motility was recorded for 8 to 10 h (10:00 to 18:00 or 20:00). The definition of phase III contractions of the MMC in conscious animals was based on that used in dogs and humans: clustered strong contractions with an amplitude of > 8 g, lasting > 5 min. Phase I of the MMC was described as a period of motor quiescence, and phase II contractions were identified based on irregular and moderate characteristics. Immediately after feeding disrupts the MMC, the gastric-motility pattern shifts to PPCs, which are irregular contractions accompanied by strong postprandial giant contractions (PPGCs; Fig. 1A).

During the initial 1 to 3 h of monitoring, traces showed PPCs in the feeding state due to free feeding before the experiment began. The MMC then emerged after the end of the PPGCs. Considering that differences occurred in the physiological state of GI motility in the animals during monitoring, we standardized the timing of administration and feeding as follows.

First, to investigate the effects of spontaneous phase I and III contractions on food intake in suncus, we monitored gastric motility during the fasting state. Food intake within 30 min was observed during the second phase I contractions that started after 10 to 30 min or during the first peak of phase III contractions in the same animal (n = 5). Next, to determine whether spontaneous phase III would promote food intake through the motilin receptor (GPR38), we administered either a vehicle control [hydrogen chloride (HCl)/saline, 80 μL of 1 N HCl in 50 mL saline] or the motilin receptor antagonist, MA-2029 (kindly donated by Chugai Pharmaceutical Company, Tokyo, Japan, dissolved in HCl/saline). MA-2029 was delivered via IV infusion (1 mg·kg⁻¹·h⁻¹) for 30 min during the first occurrence of phase III. Simultaneously, we measured feeding for 30 min (n = 6).

To explore whether exogenous motilin administration regulates food intake, vehicle (0.1% bovine serum albumin/phosphate-buffered saline [BSA/PBS]), motilin (10 μg·kg⁻¹, synthetic suncus motilin, Scrum Inc., dissolved in 0.1% BSA/PBS), or ghrelin (10 μg·kg⁻¹, active human ghrelin, Asubio Pharma, Hyogo, Japan, dissolved in 0.1% BSA/PBS) was administered via IV injection during the second phase I (at 10 to 30 min), and the animal was immediately fed 30 min posttreatment (n = 6).

To study the involvement of the vagus nerve in motilin-induced food intake during the interdigestive state, we administered either vehicle or motilin 10 to 30 min after initiation of the second phase I contractions and subsequently measured the food intake within 30 min using sham-operated suncus (n = 6) and vagotomized suncus (n = 8). Above all, gastric contractions were monitored throughout the observation period. Food intake within 30 min was calculated as follows: amount of food (g)/body weight (g) × 100.

Suncus brains were sampled to observe brain regions with neuronal activity linked to feeding after the animals were treated with the vehicle (0.1% BSA/PBS), motilin (10 μg·kg⁻¹), or ghrelin (10 μg·kg⁻¹) via IV injection, without feeding during the phase I contractions. Brain samples were collected 90 to 110 min thereafter (n = 5 to 6). The brain samples were obtained as outlined in the “Tissue Processing”.

Measuring Plasma Motilin Concentrations.

Heparinized blood was collected during MMC phases I–III from a catheter placed in the jugular vein of suncus (n = 4) and centrifuged (12,000 rpm, 3 min, 4 °C) to obtain plasma. Motilin concentrations were measured via liquid chromatography-tandem mass spectrometry (LC–MS/MS), as previously reported (53). Briefly, LC separation was performed using an LC-20A HPLC system (Shimadzu) equipped with an XBridge BEH300 C18 column (2.1 mm inside diameter × 50 mm, 3.5 μm particle size; Waters). The mobile phases consisted of water with 20 mM ammonium bicarbonate (MPA) and acetonitrile (MPB), and gradient elution was performed. The flow rate was 0.5 mL/min, and the column temperature was maintained at 50 °C. Detection was conducted using a 5500 QTRAP® system (AB SCIEX) in positive mode via multiple-reaction monitoring. Suncus motilin was quantified based on mass/charge (m/z) transitions of 669.7 → 600.1 and m/z 669.7 → 799.7, with human motilin (m/z 675.7 → 614.1 and m/z 675.7 → 818.5) serving as an internal standard. The data were analyzed using SCIEX Analyst software. Calibration samples (20 to 5,000 pg/mL) were prepared in 0.5% BSA. Test samples were processed using OASIS HLB μElution plates (Waters), followed by injection into the LC–MS/MS instrument.

Tissue Processing.

To assess whether neuronal activity in specific brain regions was related to food intake after motilin administration, suncus (n = 5 to 6) were anesthetized using a mixture of midazolam, domitor, and vetorphale, as described for the surgical procedure. They were then deeply anesthetized with an overdose of sodium pentobarbital (100 mg·kg⁻¹, IP) (54) and transcardially perfused with 75 mL of PBS, followed by 75 mL of 4% paraformaldehyde (PFA). Suncus brains were removed and postfixed for 24 h in 4% PFA. Thereafter, the brains were immersed for 1 d in 20% and 30% sucrose/PBS (pH 7.4) at 4 °C. The postfixed brains were coronally sectioned using a freezing microtome to a thickness of 40 μm, extending from the cervical spinal cord to the olfactory bulb. Subsequently, they were immersed in an antifreeze solution, separated into four equal series, and stored at −20 °C until further analysis.

Immunofluorescence.

To study the activity of TH neurons in the AP, NTS, and ARC, we used free-floating hindbrain sections from one series containing the AP and NTS. Sections containing the AP and NTS extended from −7.2 to −7.76 mm relative to the bregma, whereas sections containing the ARC ranged from −1.32 to −2.8 mm from the bregma. All sections were immersed thrice in PBS containing 0.1% Tween 20 at pH 7.6 (PBST) and then treated with 1% normal fetal bovine serum in PBS/0.5% Triton X-100 (TNBS) for 1 h at room temperature. Then, the sections were incubated overnight at 4 °C in a mixture of rabbit polyclonal anti-TH antibody (GTX102424, 1:4,000, Genetex) and guinea pig monoclonal recombinant IgG anti-c-Fos antibody (Synaptic Systems, 226308, 1:40,000). After washing them in PBST, the sections were incubated for 2 h with donkey anti-rabbit IgG (Alexa Fluor® 555, ab150074, 1:1,000; Abcam) and donkey anti-guinea pig IgG (Alexa Fluor® 488, 706-545-148, 1:1,000; Jackson ImmunoResearch), as secondary antibodies. Then, the sections were washed and mounted on glass slides with DAPI Fluoromount-G (Southern Biotech). The sections were observed using a confocal fluorescence microscope (FV1000-D; Olympus), and the contrast and brightness of the sections were corrected and merged using ImageJ software. We quantified the percentage of TH and c-Fos double-positive neurons among c-Fos-positive cells in the AP and NTS in two or three sections of each suncus. Similarly, the percentages of TH and c-Fos double-positive neurons relative to TH-positive cells in the ARC were calculated using three sections of each suncus.

In Situ Hybridization Chain Reaction (isHCR) with Immunofluorescence.

The isHCR protocol was based on modifications to previous methods involving short-hairpin DNA (55, 56). For NPY targets, three sets of split probes were designed and synthesized as standard oligos (SI Appendix, Table. S1). These sets were mixed and diluted to a concentration of 2 μM. The hypothalamus sections in one series were mounted, which contained the ARC and extended from −1.32 to −2.8 mm relative to the bregma. The slides were washed once with PBST, treated with methanol for 10 min, and then washed twice with PBST. Next, the slides were prehybridized in a moist chamber for 5 min at 37 °C in hybridization buffer (5× saline sodium citrate solution [SSC, Nacalai, 10% dextran sulfate (500,000; Fujifilm Wako Pure Chemicals), 30% formamide (Nacalai), 0.1% Tween 20 (Fujifilm Wako Pure Chemicals), 11 U/mL heparin sodium (Mochida pharmaceutical Co., LTD), 1× Denhardt’s solution (Fujifilm Wako Pure Chemicals), and distilled water]. The NPY probe was denatured at 95 °C for 5 min and then diluted to a final concentration of 20 nM in hybridization solution, which was then added to the sections, spread evenly by covering each section with parafilm, and incubated at 37 °C overnight in a moist chamber.

After hybridization, the slides were washed three times with probe-washing buffer (5× SSC, 30% formamide, and 0.1% Tween 20; 10 min/wash step) and once with 5× SSCT (5× SSC, and 0.1% Tween 20) for 10 min at 37 °C. The sections were equilibrated for 5 min with amplification buffer [8× SSC, 10% dextran sulfate (500,000; Fujifilm Wako Pure Chemicals), 0.2% Triton X-100 (Fujifilm Wako Pure Chemicals), and 100 mM MgCl₂]. A fluorophore-conjugated in situ hybridization (ISH) short-hairpin amplifier (ATTO550-S41, Nepagene) was heated to 95 °C for 1 min and gradually cooled to 65 °C in the span of 15 min and to 25 °C in the span of 40 min to allow formation of the hairpin structure. Hairpin DNA was diluted to a final concentration of 60 nM in amplification buffer and applied to the sections. The sections were then incubated at 25 °C for 2 h for chain-reaction development. Next, the sections were washed with PBST and PBS for 5 min at room temperature. NPY signaling was first confirmed via confocal fluorescence microscopy (FV1000-D; Olympus), followed by costaining with primary antibodies (against c-Fos and TH) and incubation with second antibodies, namely donkey anti-guinea pig IgG (Alexa Fluor® 488, 706-545-148, 1:1,000; Jackson ImmunoResearch) or donkey anti-rabbit IgG (Alexa Fluor™ 488, A21206, 1:500, Thermo Fisher Scientific). Immunofluorescence analysis was performed as described above. The percentages of NPY and c-Fos double-positive neurons to c-Fos-positive cells in the ARC were analyzed using three sections from each suncus.

Statistical Analysis.

At least three individual experiments were conducted to record the data. The experimental results are expressed as standard error of the mean (SEM). GraphPad Prism 9 software (GraphPad Software Inc., La Jolla, CA) was used for data analysis. Repeated one-way analysis of variance (ANOVA) (Fig. 1 C and I) or one-way ANOVA (Figs. 3 B and D, and 4 B and E) followed by Tukey’s multiple-comparison test and Student’s paired t-test (Figs. 1 E and G and 2 B and D) were to analyze the data. Statistical significance was set at P < 0.05.

Supplementary Material

Appendix 01 (PDF)

Data, Materials, and Software Availability

All study data are included in the article and/or SI Appendix.