Abstract
EM574, an erythromycin derivative and potent motilin receptor agonist, is now undergoing clinical trials as a gastroprokinetic drug. The aim of this study was to compare the effect of EM574 with that of cisapride on gastric motility and emptying in normal and gastroparesis dogs. Six dogs were each implanted with two duodenal cannulas for infusion of phenolsulfonphthalein into the proximal duodenum and for aspiration of luminal samples from the distal duodenum. Both solid and liquid gastric emptying were determined by a novel freeze-drying method developed in our laboratory. A freeze-dried standard meal (100 g, 400 kcal) was given with 100 ml normal saline containing 15 g of polyethylene glycol as a liquid marker. Gastric muscle contractility was measured by means of a force transducer implanted on the gastric antrum. EM574 (3–30 μg/kg) and cisapride (0.3–3.0 mg/kg) were administered intraduodenally at the start of feeding. Clonidine (3–30 μg/kg) was injected subcutaneously 15 min before feeding to induce gastroparesis. EM574 and cisapride both enhanced gastric muscle contractility in a dose-dependent manner. EM574 (30 μg/kg and 10 μg/kg) significantly accelerated gastric emptying of solids and liquids, respectively. Cisapride (1 mg/kg) significantly accelerated solid gastric emptying, but 3.0 mg/kg significantly delayed liquid gastric emptying. Clonidine (10 and 30 μg/kg) significantly delayed solid and liquid gastric emptying and reduced gastric muscle contractility. EM574, at a dose of 30 μg/kg, completely restored solid and liquid gastric emptying and muscle contractility to the normal range in dogs with clonidine-induced gastroparesis. Cisapride (1 mg/kg) restored liquid gastric emptying in dogs with gastroparesis to the normal range and partially restored solid emptying. EM574 accelerated gastric muscle contractility and emptying of solids and liquids in normal dogs. The stimulating activity of EM574 on gastric muscle contractility and emptying was comparable to that of cisapride, but EM574 was as effective as cisapride in normalizing gastric muscle contractility and emptying in dogs with clonidine-induced gastroparesis.
Ever since erythromycin was observed to mimic motilin (Itoh et al., 1984), it has attracted the attention of many investigators because it has been found to improve the delayed gastric emptying seen in diabetic patients with gastroparesis (Janssens et al., 1990). The mechanism of action of erythromycin seems to be different from that of other gastroprokinetic drugs developed in the past. It has now been demonstrated that the erythromycin derivatives known as motilides are nonpeptide motilin agonists in rabbits (Peeters et al., 1989) and humans (Satoh et al., 1994). The activity of the motilide EM523, which induces phase III contractions in the interdigestive state, has been reported in dogs (Ohtawa et al., 1993) and humans (Kawamura et al., 1993). More recently, Shiba et al. (1995) reported that EM523 also stimulates postprandial contractile activity in a dose-dependent manner by enhancing the amplitude of existing contractions in the gastric antrum of dogs. EM523 has also been shown to accelerate gastric emptying in dogs (Tanaka et al., 1997) and humans (Okanoet al., 1996). The motilides EM574 (Satoh et al., 1994) and ABT-229 (Maes et al., 1996) have been developed as orally ingestible formulations and are now undergoing clinical trials as promising prokinetics. EM574 is 248 times more potent than erythromycin in inducing phase III contractions in dogs and has no antibacterial activity (Tsuzuki et al., 1989). This compound has been confirmed as a potent motilin agonist in vitro in human stomach preparations (Satoh et al., 1994). However, its effect on gastric emptying has not been investigated. Cisapride, on the other hand, is a 5-HT4 receptor agonist (Taniyama et al., 1991) and is the most widely used prokinetic drug in the clinical field. This drug is highly effective in accelerating gastric emptying in normal subjects (Duan et al., 1995) and improving delayed gastric emptying in patients with various diseases (Stacher et al., 1987; Fraser et al., 1994). Most of the studies on the effect of cisapride on gastric emptying, to date, have been performed in healthy volunteers and patients, and few studies have been undertaken in normal dogs or those with gastroparesis. In the present study, therefore, we set out to determine the effect of EM574 on gastric motor activity and emptying in normal dogs and dogs with clonidine-induced gastroparesis. The activity of EM574 was compared with that cisapride.
Methods
The procedures used in the present animal experiments were approved by the Review Committee on Animal Use at Gunma University, Maebashi, Japan.
Preparation of animals.
Six healthy mongrel dogs, weighing 8 to 13 kg, were used in the present study. The dogs were anesthetized with a single intravenous (i.v.) dose of thiopental sodium (20 mg/kg), and general anesthesia was maintained by inhalation of halothane and oxygen. After laparotomy, a silicone tube (SH No. 1, Silascon Medical Tube, Kaneka Medix, Osaka, Japan) was inserted into the proximal duodenum, 6 cm distal to the pyloric ring, and used as a route to infuse the nonabsorbable marker PSP. Another tube (X2–50, Top Extension Tube; Meditop, Tokyo, Japan) was also inserted into the distal duodenum, 20 cm distal to the proximal duodenal tube; this tube was used as a route for intraduodenal administration of test materials and for aspiration of luminal contents. The distal duodenal tube was led out through a stainless steel cannula implanted on the right lateral abdominal wall. A force transducer (Itoh et al., 1977) was implanted on the gastric antrum, 3 cm proximal to the pyloric ring, to measure circular muscle contractions. The lead wires of the transducer and the proximal duodenal tube were led out through a skin incision made on the back via a subcutaneous tunnel in the costal flank. After surgery, the lead wires and proximal duodenal tube were protected with a jacket protector. The dogs were fasted and maintained with an i.v. drip-infusion of Solita T3G (Shimizu Pharmaceutical Co., Shizuoka, Japan) containing antibioticsvia a silicone tube (SH No. 1, Silascon Medical Tube, Kaneka Medix, Osaka, Japan) placed in the superior vena cava through a branch of the external jugular vein. Antibiotics were given for the first 3 days postsurgery as prophylaxis against postoperative infections, and the dogs were then gradually returned to normal dog food. The dogs were housed in individual cages and fed once daily at 10:00 a.m. Drinking water was given ad libitum.
Test meals and collection of samples.
Approximately 2 weeks were required before the dogs fully recovered from laparotomy. During that period, the dogs were trained to stand in a Pavlov sling for 7 hr per day and to eat a test meal in the Pavlov sling.
The dogs were placed in the Pavlov sling at 10:00 a.m. after being fasted for 24 hr, and the lead wires of the transducer were connected to the cable leads of the amplifiers (UG-6, Nihon Kohden Kohgyo, Tokyo, Japan). PSP solution, at a concentration of 375 μg/ml (pH 7.0), was infused at a constant rate of 3.0 ml/min through the proximal duodenal tube by means of a peristaltic pump (PST-100, Iwaki Glass, Tokyo, Japan) after being warmed to 37°C using a hot stirrer (SR-350, Advantec Toyo, Tokyo, Japan). The PSP was used for measuring duodenal volume. After a 30-min equilibration period, 1.5-ml samples of duodenal contents were collected through the distal duodenal tube at 5-min intervals. Three consecutive samples were combined to make a 15-min sample and the gastric emptying rate was determined at 15-min intervals. After two more successive samples, all dogs ate their test meal within 5 min and the collection of samples at 5-min intervals was continued for 6 hr after feeding.
The test meal consisted of 100 g freeze-dried dog food (4 kcal/g; 26% protein, 41% carbohydrate, 12% fat; Adult Dog Formula, Natural Life Petproducts, Frontenac, KS) mixed with 100 ml normal saline containing 15 g polyethylene glycol (PEG 3350; Sigma Chemical, St. Louis, MO). The food was freeze-dried using a freeze dryer (VD-41, Taitec, Saitama, Japan) and vacuum pump (DW-120, Sato Vacuum Machinery Industrial, Saitama, Japan). PEG was used for measuring liquid gastric emptying.
Administration of EM574, cisapride and clonidine.
EM574 and cisapride were administered into the duodenum through the indwelling distal duodenal tube at the start of feeding. EM574, at doses of 3, 10 and 30 μg/kg, and cisapride, at doses of 0.3, 1.0 and 3.0 mg/kg, were used. The doses of EM574 and cisapride were determined from previous studies (Shiba et al., 1995; Inatomi et al., 1989). To provide a model of gastroparesis, clonidine, analpha-2 adrenergic agonist, at doses of 3, 10 and 30 μg/kg, was injected subcutaneously into the dogs 15 min before the test meal was given (Gullikson et al., 1991a, 1991b, 1993).
Measurement of gastric emptying and gastric antral motility.
Measurements of gastric emptying of solids and liquids were performed by the freeze-drying method as reported previously (Tanaka et al., 1997).
Briefly, the lead wires of the transducer within the small pocket on the jacket protector were connected to the cable leads of the amplifier, and the changes in contractile activity in the gastric antrum were recorded continuously with a chart recorder (WT685G, Nihon Kohden Kohgyo, Tokyo, Japan) through amplifiers and a computer diskette. Data accumulated in the computer diskette were used to calculate quantitative activity (motor index, MI) of the gastric antrum in response to a test meal by measuring the area under the curve.
Drugs.
EM574 was a gift from Takeda Chemical Industries, Osaka, Japan. In figure 1, the chemical structure of EM574 is shown, together with those of erythromycin A (EMA, antibiotic agent) and EM523. This compound was dissolved in ethanol to a concentration of 20 mg/ml and the solution was diluted with 2.7% lactobionic acid. Cisapride was kindly supplied by Janssen-Kyowa, Tokyo, Japan, and was dissolved in 2.0% lactic acid. Clonidine (Wako Pure Chemical, Osaka, Japan) was dissolved in distilled water.
Statistics.
All values were expressed as mean ± S.E. of the mean values obtained from two experiments on each of the dogs. Gastric emptying of solids and liquids was expressed as the half-emptying time (t1/2) or percent remaining in the stomach at 1, 3 and 6 hr after feeding. Gastric antral motility was estimated by the integrated MI at 1, 3 and 6 hr after feeding. Statistical significance between groups was determined by Fisher’s protected least-squares difference method after analysis of variance. P values of <.05 were considered to represent a significant difference between paired data.
Results
Effects of EM574 and cisapride on postprandial contractile activity in the gastric antrum of normal dogs.
EM574 (A), at a dose of 30 μg/kg, and cisapride (B), at doses of ≥1.0 mg/kg, enhanced the contractile activity in the gastric antrum as shown in figure2. However, the effect of EM574 on the mean integrated MI was not significant. Cisapride, on the other hand, increased gastric antral motor activity in a dose-dependent manner and enhanced it significantly at a dose of 3.0 mg/kg (fig.3).
Effect of EM574 on gastric emptying of solids and liquids in normal dogs.
Figure 4 shows the effect of EM574 on solid (A) and liquid (B) gastric emptying in normal dogs. EM574, at a dose of 3 μg/kg, did not affect solid gastric emptying, and doses of ≥10 μg/kg shifted the emptying curve to the left. EM574 (30 μg/kg) did not significantly shorten the meant1/2 of the solid emptying but the mean amount of solid remaining in the stomach decreased significantly at 6 hr after feeding as shown in table 1. EM574, at a dose of 3 μg/kg, did not affect liquid gastric emptying, whereas EM574, at 10 μg/kg, caused a marked acceleration in the liquid gastric emptying. This effect on liquid emptying was less remarkable with EM574 at 30 μg/kg as shown in figure 4B. The meant1/2 of the liquid emptying was found to be significantly shorter with EM574 at 10 μg/kg than in the controls, and the mean amount of liquid remaining in the stomach was significantly smaller at 1, 3 and 6 hr after feeding as shown in table1.
Effect of cisapride on gastric emptying of solids and liquids in normal dogs.
Figure 5 shows the effect of cisapride on solid (A) and liquid (B) gastric emptying in normal dogs. Cisapride, at a dose of 1.0 mg/kg, significantly shortened the mean t1/2 of the solid emptying and the mean amount of solid remaining in the stomach was decreased significantly at 1, 3 and 6 hr after feeding. However, 0.3 and 3.0 mg/kg cisapride did not affect the solid gastric emptying as shown in figure 5 and table 1. The effect of cisapride on liquid emptying was unexpected. Cisapride, at a dose of 3.0 mg/kg, markedly prolonged the meant1/2 of the liquid emptying and the mean amount of liquid remaining in the stomach was increased significantly at 1 and 3 hr after feeding as shown in table 1. Cisapride, at doses of 0.3 and 1.0 mg/kg, did not affect the liquid gastric emptying when compared with the controls.
Effect of clonidine on postprandial contractile activity in the gastric antrum of normal dogs.
When clonidine was given 15 min before the start of feeding, the postprandial contractile activity was found to be suppressed in a dose-dependent manner as shown in figure6. The mean integrated MI was significantly smaller with clonidine (10 μg/kg) than in the controls at 1 hr after feeding. Clonidine (30 μg/kg) significantly inhibited gastric antral motor activity at 1 and 3 hr as shown in figure7A.
Effect of clonidine on gastric emptying of solids and liquids in normal dogs.
Treatment with clonidine, at concentrations of 10 and 30 μg/kg, significantly delayed gastric emptying of solids (A) and liquids (B) as shown in figure 8. The mean solid emptying curves were shifted upward at 10 and 30 μg/kg (fig. 8A). The mean t1/2 of the solid emptying was prolonged significantly by clonidine at a dose of 30 μg/kg. The mean amount of solid remaining in the stomach with 10 μg/kg clonidine was increased significantly at 1 and 3 hr and that with 30 μg/kg clonidine was increased significantly at 1, 3 and 6 hr (table2). The mean liquid emptying curves, on the other hand, were greatly shifted to the right and were delayed by 10 and 30 μg/kg of clonidine as shown in figure 8B. Clonidine, at doses of 10 and 30 μg/kg, markedly prolonged the meant1/2 of the liquid emptying as shown in table 2. The mean amount of liquid remaining in the stomach was increased significantly by clonidine (10 and 30 μg/kg) at 1 hr after feeding. Therefore, the submaximum dose of 10 μg/kg was used to induce a state of gastroparesis in subsequent experiments.
Effects of EM574 and cisapride on contractile activity in the gastric antrum of dogs with clonidine-induced gastroparesis.
As EM574, at doses of 10 and 30 μg/kg, and cisapride, at a dose of 1.0 mg/kg, were found to accelerate gastric emptying in normal dogs, the same doses of the two drugs were used to investigate gastric motor activity and emptying in dogs with gastroparesis. When EM574 or cisapride was given to dogs with gastroparesis, the contractile activity in the gastric antrum was restored as shown in figure 6. Figure 7B shows the effects of EM574 (10 and 30 μg/kg) and cisapride (1.0 mg/kg) on the mean integrated MI at 1, 3 and 6 hr after feeding. Treatment with EM574 and cisapride tended to improve the decrease in gastric motor activity seen in clonidine-treated dogs, and this improvement was significant when EM574 was administered at a dose of 30 μg/kg (fig. 7B).
Effects of EM574 and cisapride on gastric emptying of solids and liquids in dogs with clonidine-induced gastroparesis.
The delayed solid gastric emptying induced by clonidine was restored completely by EM574 at a dose of 30 μg/kg but not 10 μg/kg. Indeed, the solid emptying curve obtained with 30 μg/kg EM574 was even accelerated compared with the controls as shown in figure9A. After an EM574 dose of 30 μg/kg, the mean t1/2 of the solid emptying was improved significantly, and the mean amount of solid remaining in the stomach was significantly smaller at 1, 3 and 6 hr after feeding than with clonidine (10 μg/kg) (table 2). As to the effect of EM574 on liquid gastric emptying, the delayed emptying induced by clonidine was returned to the normal range by 30 μg/kg EM574 as shown in figure 9B. The mean t1/2 of the liquid emptying was shortened significantly by this dose of EM574, as shown in table 2.
Cisapride (1.0 mg/kg) accelerated the delayed gastric emptying of both solids (fig. 10A) and liquids (fig.10B) in clonidine-treated dogs. The mean t1/2 of the liquid emptying, but not solid emptying, was improved significantly compared with dogs given clonidine only, as shown in table 2.
Discussion
Ever since Bass and Wiley (1972) and Itoh et al. (1977)described how to record contractile activity in the gastrointestinal tract in conscious animals, gastroprokinetic functions have been investigated worldwide using force transducers. Dye dilution techniques were used formerly to measure gastric emptying of liquids until the isotope technique was introduced by Griffith et al. (1966). Since then, the question of surface adsorption of isotopes in the mixed solid-liquid meal has arisen. Meyer et al. (1976)established a method for measuring solid gastric emptying by incorporating 99 mTc into chicken liver by injecting the isotope into a live chicken in 1976. The isotope method reported by Meyer et al. is ideal for measuring solid gastric emptying, but there are still some problems concerning the dissociation of tagged solid from the rest of a test meal. At present, most methods for measuring gastric emptying of solid-liquid meals use foods tagged with radionuclide markers. However, the numbers of test meals to which radionuclide markers can be tagged are limited, and it is not always possible to tag all solid foodstuffs. To solve these problems several new methods for measuring gastric emptying without exposure to radionuclides have recently been reported (Ghoos et al., 1993; Braden et al., 1995; Hveem et al., 1996; Demedts et al., 1997). In our laboratory, we have developed a freeze-drying method to measure gastric emptying of solids in conscious dogs. The advantage to our method is that it is unnecessary to tag foods with a radioactive nuclide, and there is no problem about the possible dissociation of the radioactivity from the solid phase after digestion of a liquid meal. Another advantage of our technique is that any type of foods can be used as a test meal, as long as the absolute weight of the food before feeding and the weight of the samples collected from the duodenum can be determined by freeze-drying. We have confirmed (Tanaka et al., 1997) the validity of this method for measuring gastric emptying of solids using the nitric oxide synthase inhibitor L-NNA and an erythromycin derivative, EM523.
Another advantage to this method is that it is possible to measure both gastric motility and emptying at the same time. By measuring gastric motor activity and emptying simultaneously, we confirmed that increasing gastric motor activity is not necessary for an acceleration in gastric emptying. Gastric antral motor activity was increased by cisapride (3.0 mg/kg), whereas gastric emptying was delayed. Wulschkeet al. (1986) reported that an increase in the force of antral contractions by stimulating substances did not speed the emptying rate of the nutrient meal. In a previous study (Tanakaet al., 1997), we also showed that EM523 and L-NNA had opposite effects on gastric emptying, although both of these compounds enhanced the contractile activity in the gastric antrum. With cisapride (3.0 mg/kg), the delayed gastric emptying may have occurred because of disruption of antropyloroduodenal coordination or simultaneous increase in gastrointestinal contractile activity, although we have no data to indicate such an occurrence. In the present study, however, a force transducer was chronically implanted on the duodenum in an attempt to examine antroduodenal coordination, but strong adhesion developed around the transducer and prevented precise recording of the contractile waves due to additional implantation of the two tubes used as routes for infusion and collection in the duodenum. Therefore, if the true reason for delayed emptying due to the high dose of cisapride is to be investigated, a new technique should be applied that enables us to monitor contractile changes precisely not only in the duodenum but also in the pyloric ring.
Models of gastroparesis have been set up using morphine, dopamine, CCK and oil components, and the gastroprokinetic activity of various drugs has been studied using these models (Müller-Lissner et al., 1986; Rowbotham and Nimmo, 1987; Hatanaka et al., 1996; Jones et al., 1996). Sninsky et al. (1986)suggested that the alpha-2 adrenergic antidiarrhea agent lidamidine delayed gastric emptying in humans, and Allescher et al. (1989) reported that adrenergic receptors inhibit pyloric motor function through presynaptic alpha receptors. Furthermore, Gullikson et al. (1991b) demonstrated that SC-39585A, an alpha-2 adrenergic agonist, inhibited antroduodenal motility and stimulated gastroparesis in dogs. In this study, we have shown that clonidine, an alpha-2 adrenergic agonist, decreased postcibal gastric antral motor activity in a dose-dependent manner and delayed solid and liquid gastric emptying. As a result, we were also able to induce gastroparesis in dogs in a similar way to Gullikson et al. (1991b).
Alpha-2 adrenergic agonists inhibit acetylcholine release because of activation of presynaptic alpha-2 receptors, and as a result, contractile activity of the gastrointestinal smooth muscle is suppressed. Erythromycin (a motilin receptor agonist) and cisapride have been reported to be beneficial to patients with diabetes (Horowitzet al., 1987; Desautels et al., 1995), progressive systemic sclerosis (Dull et al., 1990) and many other conditions where gastroparesis is common. We demonstrated that EM574 restored postprandial gastric antral motor activity and solid and liquid gastric emptying in dogs with clonidine-induced gastroparesis to the normal range and that cisapride restored liquid gastric emptying.
The mechanism of action of EM574 is largely unknown, although recent studies have shown that erythromycin can mimic the effect of motilin on gastrointestinal motility (Itoh et al., 1984) and improve gastroduodenal coordination (Annese et al., 1992). Furthermore, it is well known that erythromycin, as well as cisapride, increases gastric antral motor activity and accelerates solid and liquid gastric emptying in humans (Annese et al., 1992) and dogs (Lin et al., 1994). Moreover, erythromycin enhances postprandial proximal gastric tone in humans (Bruley des Varanneset al., 1995). In this study, we demonstrated that when EM574 was administered intraduodenally, postprandial gastric antral motor activity increased and solid and liquid gastric emptying accelerated in normal dogs. It is proposed that erythromycin derivatives act at the motilin receptors in the area postrema through 5-HT3 receptors to stimulate vagal preganglionics, which finally releases acetylcholine via the cholinergic pathway in the interdigestive state (Ohtawa et al., 1993). After food, however, erythromycin derivatives mediate gastric contractions through both the cholinergic and noncholinergic pathways, which involve 5-HT3 and neurokinin 1 receptors, but not alpha adrenergic receptors (Shibaet al., 1995). On the other hand, cisapride, a 5-HT4 receptor agonist (Taniyama et al., 1991), increased postprandial gastric antral motor activity in a dose-dependent manner and accelerated solid gastric emptying in normal dogs in this study. The effect of cisapride is thought to be due to the augmentation of acetylcholine release from cholinergic neuronsvia 5-HT4 receptors. Both EM574 and cisapride have different modes of action as gastroprokinetic drugs and actvia different routes.
In conclusion, our method for measuring gastric emptying after ingestion of a mixed solid-liquid meal is reliable and reproducible in conscious dogs and is superior to the isotope method. The method has the advantages that there is no exposure to radioactive nuclides, it can easily be used in different laboratories, and any foodstuff can be used as long as the absolute weight of that food before feeding, and the weight of the samples collected from the duodenum, can be determined by freeze-drying. We have shown that the two gastroprokinetic drugs acting via a different route, EM574 and cisapride, increased postprandial gastric antral motor activity and accelerated gastric emptying in normal dogs, although increasing the gastric motor activity does not always accelerate gastric emptying. Moreover, we have shown that clonidine decreased motor activity and delayed gastric emptying when used to induce gastroparesis in dogs. EM574 completely restored gastric motor activity and gastric emptying to the normal range in dogs with clonidine-induced gastroparesis, but cisapride only partially restored these functions. EM574 can be considered to be an efficacious gastroprokinetic agent.
Acknowledgments
The authors thank Miss Fumie Mizusawa for technical assistance.
Footnotes
-
Send reprint requests to: Toshiyuki Tanaka, M.D., The Second Department of Surgery, Gunma University School of Medicine, Maebashi, Japan.
-
↵1 T.T. is a research fellow from the Second Department of Surgery, Gunma University School of Medicine, Maebashi, Japan.
- Abbreviations:
- EM574
- De(N-methyl)-N-isopropyl-8,9-anhydroerythromycin A-6,9-hemiacetal
- PSP
- phenolsulfonphthalein
- PEG
- polyethylene glycol
- EM523
- De(N-methyl)-N-ethyl-8,9-anhydroerythromycin A-6,9-hemiacetal
- 5-HT
- 5-hydroxytryptamine
- EMA
- erythromycin A
- t1/2
- half-emptying time
- MI
- motor index
- Tc
- technetium
- L-NNA
- Nω-nitro-l-arginine
- Received February 19, 1998.
- Accepted June 8, 1998.
- The American Society for Pharmacology and Experimental Therapeutics