Rikkunshito, an Herbal Medicine, Suppresses Cisplatin-Induced Anorexia in Rats Via 5-HT2 Receptor Antagonism

Transcription

1 GASTROENTEROLOGY 2008;134: Rikkunshito, an Herbal Medicine, Suppresses Cisplatin-Induced Anorexia in Rats Via 5-HT2 Receptor Antagonism HIROSHI TAKEDA,* CHIHARU SADAKANE, TOMOHISA HATTORI, TAKEHIKO KATSURADA,* TATSUYA OHKAWARA,*, KOICHI NAGAI, and MASAHIRO ASAKA* *Department of Gastroenterology and Hematology, Hokkaido University Graduate School of Medicine, Kita-ku, Sapporo, Japan; Tsumura & Co. Tsumura Research Laboratories, Inashiki-gun, Ibaraki, Japan; Department of Medical Information, Hokkaido Information University, Ebetsu, Japan; and the Department of Oral Pathological Science, Graduate School of Dental Medicine, Hokkaido University, Kita-ku, Sapporo, Japan Background & Aims: Chemotherapy with an anticancer agent generally causes gastrointestinal tract disorders such as vomiting and anorexia, but the mechanism remains unclear. Rikkunshito, a kampo preparation, is known to alleviate such adverse reactions. In this study, we attempted to clarify the mechanism. Methods: We investigated the decreases of plasma acylated-ghrelin level and food intake caused by cisplatin, serotonin (5-HT), 5-HT agonists, and vagotomy as well as the decrease-suppressing effects of rikkunshito and 5-HT antagonists. In addition, binding affinities of rikkunshito components were determined in receptor-binding assays using 5-HT2B and 5-HT2C receptors. Results: Cisplatin, 5-HT, BW723C86 (5-HT2B receptor agonist), and m-chlorophenylpiperazine HCl (5-HT2C agonist) markedly decreased plasma acylated-ghrelin levels, although 5-HT3 and 5-HT4 agonists had no effect. In contrast, 5-HT2B and 5-HT2C antagonists suppressed the cisplatin-induced decrease of plasma acylated-ghrelin level and food intake. Administration of rat ghrelin improved the cisplatin-induced decrease in food intake. Vagotomy decreased the plasma acylated-ghrelin level, which was decreased further by cisplatin. Rikkunshito suppressed such cisplatin-induced decreases of plasma acylated-ghrelin level and food intake. The suppressive effect of rikkunshito was blocked by a ghrelin antagonist. Components of rikkunshito, 3,3=,4=,5,6,7,8-heptamethoxyflavone, hesperidin, and isoliquiritigenin showed a 5-HT2B-antagonistic effect in vitro, and oral administration of rikkunshito suppressed the cisplatin-induced decrease in the plasma acylated-ghrelin level. Conclusions: The cisplatininduced decreases of the plasma acylated-ghrelin level and food intake are mediated by 5-HT2B/2C receptors and suppressed by flavonoids in rikkunshito. Chemotherapy is used to treat various types of cancer in the expectation of extending the life of the patient. However, gastrointestinal tract disorders caused by chemotherapy, such as nausea, vomiting, and anorexia, markedly decrease the patient s quality of life and make continuation of chemotherapy difficult. Cisplatin, being representative of the anticancer agents used in chemotherapy, causes acute gastrointestinal tract disorders within 24 hours after administration and delayed gastrointestinal tract disorders occurring more than 24 hours postdose. It has been suggested that the acute gastrointestinal tract disorders involve serotonin (5-HT) secretion from the enterochromaffin cells. 1 Furthermore, it is known that the co-administration of a 5-HT3 receptor antagonist with cisplatin suppresses nausea and vomiting in the acute stage, further suggesting that 5-HT is at least partially responsible for the gastrointestinal tract disorders caused by cisplatin. 5-HT has been implicated in a variety of physiologic effects and plays particularly important roles in gastrointestinal functions. The accumulated findings suggest that 5-HT2 receptor family is involved in the control of appetite. 2 The 5-HT2B receptor subtype is known to exist in smooth muscle such as those in the gastrointestinal tract, stomach fundus, 3 blood vessels, and uterus. It is known that stimulation of the 5-HT2B receptor by administration of selective agonist induces a marked decrease of food intake. 2 On the other hand, the 5-HT2C receptor subtype is localized mainly in the brain, 4 and knockout of this receptor in mice caused leptin-independent hyperphagia and hypoactivity leading to obesity. 5 Some studies using a 5-HT2C receptor agonist have provided evidence supporting the hypothesis that stimulation of the 5-HT2C receptor induces anorexia. 2,6 The involvement of 5-HT2B/2C receptors in the control of food intake has been suggested by previous findings. However, the relationship between the gastrointestinal tract disorders caused by cisplatin and the peripheral/ intracerebral 5-HT2B/2C receptors has not been clarified sufficiently. Abbreviations used in this paper: CPB, 1-(3-chlorophenyl) biguanide HCl; GHRP-6, [D-Lys 3 ] ghrelin receptor antagonist; HMF, 3,3=,4=,5,6,7,8- heptamethoxyflavone; 5-HT, serotonin; Ki, inhibition constant 5-HT2B/ 2C receptor binding-inhibitory activities; mcpp, m-chlorophenyl piperazine HCl; CD, hydroxypropyl- -cyclodextrin by the AGA Institute /08/$34.00 doi: /j.gastro

2 June 2008 RIKKUNSHITO INCREASES PLASMA GHRELIN LEVEL 2005 Ghrelin is an endogenous ligand of the growth-hormone secretagogue receptor consisting of 28 amino acids, and it is secreted mainly from the stomach. Ghrelin is known to have an intense appetite-enhancing effect in addition to the growth hormone secretion-promoting effect. 7 Ghrelin and its receptor also are produced and expressed in the hypothalamus arcuate nucleus neurons as well as in the stomach, and ghrelin administered to rodents peripherally or intracerebrally causes enhancement of gastrointestinal motility and increases food intake. 8 In addition, changes in the plasma ghrelin level were reported in functional dyspepsia, 9 chronic gastritis, and gastric ulcer. 10 Thus, the association of ghrelin with gastrointestinal diseases has been variously reported, but the control mechanism for ghrelin secretion has not been clarified. Our hypotheses are as follows: (1) the administration of cisplatin would cause a decrease in plasma ghrelin concentration, and (2) a decreased ghrelin level was attributable to a stimulation of 5-HT receptor, especially 5-HT2B/2C receptors. Rikkunshito is a kampo preparation used in Japan to treat various gastrointestinal tract disorders Some studies in rats administered rikkunshito have shown an enhancing effect on gastric emptying. 15 Recently, it was reported that rikkunshito administered in combination with an anti-emetic drug proved useful against the anorexia and vomiting occurring as adverse reactions to chemotherapy in patients with advanced breast cancer. 13 But, again, the mechanism by which rikkunshito alleviates the gastrointestinal disorders induced by anticancer agents has not been clarified. Thus, in this study, we also investigated the mechanism by which co-administration of rikkunshito with anticancer agents proves useful in clinical practice by attempting to show that rikkunshito prevents the decreases of appetite and plasma ghrelin level caused by cisplatin treatment in rats. In addition, an attempt was made to identify the active components of rikkunshito. Materials and Methods Experimental Animals Male SD rats aged 7 weeks (Charles River Laboratories Japan, Inc, Tokyo, Japan) were used. During the study period, the rats were kept in an animal room under the following conditions: controlled temperature and humidity, 4 5 rats in each cage, with a 12-hour light/ darkness cycle (7:00 AM to 7:00 PM); and food and water given ad libitum. All experiments were performed between 9:00 AM and 6:00 PM under compliance with the guidelines established by the Animal Ethics Committee of Tsumura & Co. Test Substances Cisplatin (Wako, Osaka, Japan), 5-HT, BW723C86 (5-HT2B receptor agonist), m-chlorophenyl piperazine HCl (mcpp, 5-HT2C receptor agonist), 1-(3-chlorophenyl) biguanide HCl (CPB), 2-methyl serotonin (5-HT3 receptor agonist) (Sigma Aldrich Chemical Co, St Louis, MO), and cisapride (5-HT4 receptor agonist) (Tocris Bioscience, Glasgow, UK) were dissolved in sterile 0.9% physiologic saline (Otsuka, Tokyo, Japan) immediately before use. Similarly, SB (5-HT2B receptor antagonist) and SB242084HCl (5-HT2C receptor antagonist) (Sigma Aldrich Chemical Co) were suspended or dissolved in saline or 40% hydroxypropyl- -cyclodextrin solution (Tocris Bioscience, Ellisville, MO) immediately before use. Rat acylated-ghrelin (Peptide Institute Inc, Osaka, Japan) and [D-Lys 3 ] ghrelin receptor antagonist (GHRP-6) (Bachem, Bubendorf, Switzerland) were dissolved in physiologic saline before use. Rikkunshito was used in the form of a powdered extract obtained by spraydrying the hot-water extract of the mixture of 8 crude drugs (ie, Atractylodis lanceae rhizoma, Ginseng radix, Pinelliae tuber, Hoelen, Zizyphi fructus, Aurantii nobilis pericarpium, Glycyrrhizae radix, and Zingiberis rhizoma). Hesperetin, hesperidin, naringenin, 3,3=,4=,5,6,7,8-heptamethoxyflavone (HMF), nobiletin, tangeretin (Aurantii nobilis pericarpium), isoliquiritigenin, liquiritigenin, glycycoumarin (Glycyrrhizae radix), 8-shogaol, 10-shogaol, 10-dehydrogingerdin, 10-gingerdion, and 8-gingerol (Zingiberis rhizoma), typical components of rikkunshito, were used as samples for 5-HT receptor binding assays. Rikkunshito and these ingredients were supplied by Tsumura & Co. Effects of Cisplatin and Various 5-HT Receptor Agonists on Plasma Ghrelin Level Rats were administered cisplatin (2, 6, 10, or 14 mg/kg) and 5-HT (1, 2, 4, or 8 mg/kg) and subjected to blood collection by decapitation at 120 minutes (cisplatin) or 30 minutes (5-HT) after administration. Other groups of rats were fasted for 24 hours and administered BW723C86 (4 or 16 mg/kg), mcpp (3 or 9 mg/kg), CPB (1.25 or 6 mg/kg), 2-methylserotonin (1 or 4 mg/kg), or cisapride (1.25 or 5 mg/kg) and subjected to blood collection by decapitation at 60 minutes after administration. The stomach was removed from animals of the treated groups in which the plasma ghrelin level was decreased. All drugs were given as solutions in physiologic saline. Effect of 5-HT2 Receptor Antagonist and Vagotomy on Plasma Acylated-Ghrelin Level in Rats Administered Cisplatin Rats fasted for 24 hours were administered SB (6 mg/kg) or SB242084HCl (10 mg/kg) intraperitoneally. After 30 minutes, cisplatin was administered intraperitoneally. Two hours after administration of cisplatin, the rats were subjected to blood collection by decapitation, and plasma samples were obtained by the method detailed later. Vagotomy was performed as described previously. 16 The vagotomized rats were fasted

3 2006 TAKEDA ET AL GASTROENTEROLOGY Vol. 134, No. 7 Figure 1. Plasma acylated- (A and C) or desacylated (B and D) ghrelin levels in cisplatin- or 5-HT treated rats. Each bar represents the mean SE of 8 rats. P values were obtained with Dunnett s test. for 24 hours, administered cisplatin, and subjected to blood collection by decapitation at 2 hours after administration. Effects of Rikkunshito and Active Herbal Components on Plasma Acylated-Ghrelin Level in Rats Administered Cisplatin The experiments to verify the effect of rikkunshito on plasma ghrelin level in cisplatin-treated rats were performed both on rats fasted for 24 hours and on the rats given food ad libitum. The rats were orally administered rikkunshito (500 or 1000 mg/kg) dissolved in 1 ml of distilled water per 100 g of body weight. After 16 hours, the rats were administered cisplatin intraperitoneally and at the same time a second dose of rikkunshito (which was the same size as the first dose) was administered orally. A single dose of HMF (0.8, 4.0, or 20 mg/kg), hesperidin (4.0 or 20 mg/kg), or isoliquiritigenin (4.0 mg/kg) was given orally simultaneously with the administration of cisplatin. At 2 hours after the last dose, the rats were subjected to blood collection by decapitation, and plasma samples were obtained by the method detailed later. Determination of Ghrelin Level The plasma samples promptly were centrifuged at 4 C, and the supernatant was acidified with 1 mol/l HCl (1/10 volume). The tissue samples were boiled in water for 7 minutes and acidified with 1 mol/l HCl after cooling. The samples then were homogenized and centrifuged at 3,000 rpm for 15 minutes. The supernatants were stored at 80 C until use. The ghrelin level was determined using the Active Ghrelin or Desacyl Ghrelin Enzyme-Linked Immunoassay Kit (Mitsubishi Kagaku Iatron, Inc, Tokyo, Japan). Effect of 5-HT2B or 2C Receptor Antagonist on Food Intake in Rats Administered Cisplatin The rats were given water and food ad libitum, and SB or SB242084HCl was administered intraperitoneally. After 30 minutes, cisplatin was administered intraperitoneally and the food intake for 24 hours thereafter was measured. Effects of Rat Ghrelin and Rikkunshito, Alone or in Combination With Ghrelin-Receptor Antagonist, on Food Intake in Rats Administered Cisplatin Cisplatin was administered intraperitoneally, and after 2 hours rat acylated-ghrelin (1 or 5 nmol/rat) was administered intravenously via the tail vein, and thereafter the food intake in 4 hours was measured. Rikkunshito (500 or 1000 mg/kg) was administered orally, and after 16 hours cisplatin was administered intraperitoneally and at the same time the second dose of rikkunshito was administered orally (which was the same size as the first dose). The food intake and body weight gain in the subsequent 24 hours were recorded. In the second experiment, after following the procedures de-

4 June 2008 RIKKUNSHITO INCREASES PLASMA GHRELIN LEVEL 2007 Figure 2. Plasma acylated-ghrelin levels (A), 24-hour food intakes in rats treated with SB or SB HCl in combination with cisplatin (B), and effects on plasma acylatedghrelin level in sham-operation rats, vagotomized rats, and cisplatin-treated vagotomized rats (C). (A and B) SB and SB HCl were administered intraperitoneally at dose levels of 6 and 10 mg/kg, respectively, 30 minutes before administration of cisplatin. Each bar represents the mean SE of 8 rats. (C) Each bar represents the mean SE of 8 rats. P values were obtained with Student s t-test or Dunnett s test. scribed earlier, physiologic saline or a solution of [D-Lys 3 ] GHRP-6 (0.4 mol/rat) was administered intraperitoneally simultaneously with cisplatin and the second dose of rikkunshito (1000 mg/kg); the food intake in the subsequent 6 hours was recorded. Radioligand Binding Assays CHO-K1 cells expressing human recombinant 5-HT receptor were used to prepare membranes in modified Tris- HCl buffer. A membrane protein was incubated with 1.2 nmol/l [ 3 H]LSD for 60 minutes at 37 C. Nonspecific binding was estimated in the presence of 10 mol/l 5-HT. The filters then were counted for radioactivity to determine the amount of specifically bound [ 3 H]-LSD. 17 CHO-K1 cells stably transfected with a plasmid encoding the human 5-HT2C receptor were used to prepare membranes in modified Tris-HCl buffer. A membrane protein was incubated with 1.0 nmol/ml [ 3 H]-Mesulergine for 60 minutes at 25 C. Nonspecific binding was estimated in the presence of 1 mol/l Mianserin (Nippon Organon K.K., Osaka, Japan). The filters then were counted for radioactivity to determine the amount of specifically bound [ 3 H]-Mesulergine. 18 Likewise, membranes prepared from cells stably transfected with human recombinant 5-HT3, 5-HT6 receptors, and guinea pig striatum (5-HT4) were incubated with radiolabelled ligands with a high affinity for the given receptor, that is, [ 3 H]LSD, [ 3 H]-GR-65630, and [ 3 H]-GR , respectively. Nonspecific radioligand binding was defined by 5-HT or MDL Statistical Analysis Data are given as the mean SEM of 8 rats. The Student t test and the Dunnett method were used for statistical analysis as appropriate; a P value of less than.05 was considered statistically significant. Results Effects of Cisplatin and 5-HT on Plasma Ghrelin Level, Effects of 5-HT2B and 2C Receptor Antagonists on Plasma Acylated-Ghrelin Level and Food Intake in Rats Administered Cisplatin, and Effect of Vagotomy on Plasma Acylated-Ghrelin Level The plasma acylated-ghrelin level decreased significantly when cisplatin was administered intraperitoneally at all dose levels, as compared with the control group administered saline intraperitoneally (Figure 1A). The plasma desacylated-ghrelin level also significantly decreased in the cisplatin-treated rats at all doses as compared with the control group (Figure 1B). When 5-HT was administered intraperitoneally to rats, the plasma acylated-ghrelin level decreased in a dose-dependent manner and was significantly lower in the groups receiving 5-HT at 4 mg/kg and 8 mg/kg than in the control group (Figure 1C). The plasma desacylated-ghrelin also decreased in the 5-HT treated rats (Figure 1D). When cisplatin was administered to rats that had been fasted for 24 hours, the plasma acylated-ghrelin level decreased significantly as compared with the control rats ad-

5 2008 TAKEDA ET AL GASTROENTEROLOGY Vol. 134, No. 7 Figure 3. Plasma acylated- (A) or desacylated (B) ghrelin levels in BW723C86- and mcpptreated rats, and gastric acylated- or desacylated-ghrelin levels in BW723C86- (C) or mcpp (D) treated rats. The rats fasted for 24 hours, administered BW723C86 (4 or 16 mg/ kg) or mcpp (3 or 9 mg/kg), and subjected to blood collection by decapitation at 60 minutes after administration (A and B), and the and removal of the stomach at 60 minutes after administration (D and E). Each bar represents the mean SE of 8 rats. P values were obtained with Dunnett s test. ministered 40% hydroxypropyl- -cyclodextrin solution. The decrease caused by administration of cisplatin was suppressed significantly by intraperitoneal administration of SB or SB242084HCl as shown in Figure 2A. When cisplatin was administered to rats, the food intake decreased significantly as compared with the control rats administered only 40% hydroxypropyl- -cyclodextrin solution. The decrease in food intake caused by administration of cisplatin was suppressed significantly by intraperitoneal administration of SB Similarly, the food intake decreased significantly in rats administered cisplatin as compared with rats administered physiologic saline, but in the rats administered SB242084HCl, the decrease of food intake caused by the administration of cisplatin was suppressed significantly (Figure 2B). As indicated in Figure 2C, the plasma acylated-ghrelin level showed no marked change in the sham-operation group as compared with the untreated control group. On the other hand, the plasma acylated-ghrelin level decreased significantly in the vagotomized group as compared with the sham-operation group. In the vagotomized rats administered cisplatin, the plasma acylated-ghrelin level was markedly decreased further. Effects of Various 5-HT Receptor Agonists on Plasma Ghrelin Level and Effects of BW723C86 and mcpp on Intragastric Ghrelin Level As the next step, to obtain the direct evidence by which 5-HT2B/2C receptor activation causes a decrease in ghrelin secretion, we investigated the effects of 5-HT receptor agonists on plasma and gastric ghrelin levels. When BW723C86 was administered intraperitoneally at a dose level of 16 mg/kg, the plasma acylated-ghrelin (Figure 3A) and desacylated-ghrelin (Figure 3B) levels were significantly lower, as compared with the control group at 60 minutes postdose. When mcpp was administered intraperitoneally at dose levels of 3 mg/kg and 9 mg/kg, the plasma acylated-ghrelin level was significantly lower, as compared with the control group, at 60 minutes postdose (Figure 3A). The plasma desacylated-ghrelin level in the group treated with 9 mg/kg mcpp showed a tendency to decrease at 60 minutes postdose (Figure 3B). When BW723C86 was administered intraperitoneally to rats at a dose level of 16 mg/kg, the intragastric acylated- and desacylated-ghrelin levels were significantly higher as compared with the control group (Figure 3C), and when mcpp was administered at 9 mg/kg, the intragastric acylated- and desacylated-ghrelin level were significantly higher as compared with the control group (Figure 3D). To confirm the participation of 5-HT3/4 receptor activation on ghrelin secretion, we evaluated the effects of 5-HT3 and 5-HT4 receptor agonists on plasma acylated-ghrelin level. As shown in Table 1, when cisapride was administered intraperitoneally, the plasma acylated-ghrelin level was not affected significantly as compared with the saline-treated group. When treated with 2-methylserotonin or CPB, the plasma acylated-ghrelin level tended to be higher than that of the saline-treated group.

6 June 2008 RIKKUNSHITO INCREASES PLASMA GHRELIN LEVEL 2009 Table 1. Effects of Several 5-HT Receptor Agonists on Plasma Acylated-Ghrelin Level in Rats Group Plasma ghrelin level (fmol/ml) P values Saline, ip Methylserotonin 1.0 mg/kg, ip NS 2-Methylserotonin 4.0 mg/kg, ip NS Saline, ip CPB 1.25 mg/kg, ip NS CPB 6.0 mg/kg, ip NS Saline, ip Cisapride 1.25 mg/kg, ip NS Cisapride 5.0 mg/kg, ip NS NOTE. Each datum represents the mean SE of 8 rats. The rats were fasted for 24 hours, administered a test drug, and subjected to blood collection by decapitation at 60 minutes after administration. ip, intraperitoneally. Effects of Rikkunshito on Plasma Acylated- Ghrelin Level, Food Intake, and Body Weight in Rats Administered Cisplatin, and the Effects of Rat Ghrelin, Rikkunshito in Combination With [D-Lys 3 ] GHRP-6 When cisplatin was administered to rats fasted for 24 hours, the plasma acylated-ghrelin level decreased significantly as compared with that of control rats administered physiologic saline (Figure 4). The decrease caused by administration of cisplatin was suppressed significantly by oral administration of rikkunshito (1000 mg/kg) (Figure 4A). Similarly, when cisplatin was administered to the rats given food ad libitum, the plasma acylated-ghrelin level decreased significantly as compared with the control rats administered physiologic saline. The decrease caused by administration of cisplatin was suppressed significantly when rikkunshito was administered orally at dose levels of 500 and 1000 mg/kg (Figure 4B). Furthermore, as compared with the group administered physiologic saline, both the 24-hour food intake and the body weight gain decreased significantly in the group administered cisplatin (24-hour food intake: g/24 h vs g/24 h, P.05 by Student t test; body weight gain: g/24 h vs g/24 h, P.05 by Student t test). When rikkunshito (1000 mg/kg) was co-administered with cisplatin, the decreases of food intake and body weight gain were suppressed significantly as compared with the group administered cisplatin alone ( g/24 h and g/24 h, respectively, P.05 vs cisplatin-treated rats by the Dunnett test) (data not shown in Tables and Figures). When cisplatin was administered to rats, the 6-hour food intake decreased significantly as compared with the control rats administered physiologic saline. The decrease of food intake caused by administration of cisplatin was not suppressed in the group administered rat ghrelin at a dose level of 1 nmol/rat, but it was suppressed significantly in the group administered ghrelin at a dose level of 5 nmol/rat (Figure 4C). In the rats administered cisplatin alone, the 6-hour food intake decreased significantly compared with that of saline-treated rats. When cisplatin was combined with rikkunshito, the decrease of food intake was suppressed, and when [D-Lys 3 ] GHRP-6 also was added, the effect of rikkunshito was suppressed significantly by the [D-Lys 3 ] GHRP-6. The food intake did not differ between rats administered cisplatin alone and rats administered cisplatin in combination with [D-Lys 3 ] GHRP-6 (Figure 4D). Binding Assays of Rikkunshito Components Using 5-HT Receptors Table 2 shows the 5-HT2B/2C receptor bindinginhibitory activities of the crude drug components contained in rikkunshito. Thirteen components showed 5-HT2B receptor binding-inhibitory activity, and, among these, HMF showed an inhibition constant (Ki) value of 0.21 mol/l, nobiletin of 0.31 mol/l, tangeretin of 0.59 mol/l, 8-shogaol of 1.8 mol/l, and glycycoumarin of 2.4 mol/l. Hesperetin, an aglycon form of hesperidin, had a Ki value of 5.3 mol/l against the 5-HT2B receptor and 20.9 mol/l against the 5-HT2C receptor. Isoliquiritigenin showed a Ki value of 3.3 mol/l on binding between 5-HT and 5-HT2C receptor. The binding assay also was performed using 5-HT3, 4, and 6 receptors, but only isoliquiritigenin and glycycoumarin showed binding activity with the 5-HT4 receptor, with Ki values of 13.0 mol/l and 4.1 mol/l, respectively. The other components showed no inhibitory activities. On the other hand, although not shown in the Table 2, ketanserin as the positive control showed a Ki value of 0.25 mol/l for binding of the 5-HT2B receptor and SB242084HCl showed a Ki value of 0.49 nmol/l for binding of the 5-HT2C receptor. Effects of HMF, Hesperidin, and Isoliquiritigenin on the Decrease of Plasma Acylated-Ghrelin Level in Rats Administered Cisplatin Figure 5 shows the structures of HMF (Figure 5A), hesperidin (Figure 5B), and isoliquiritigenin (Figure 5C). In rats administered cisplatin, the plasma acylated-ghrelin level decreased significantly as compared with the rats administered physiologic saline. When HMF (4 or 20 mg/kg) was administered orally in combination with cisplatin, the decrease of plasma acylated-ghrelin level was prevented (Figure 5D). The decrease of plasma acylated-ghrelin level was suppressed significantly by administration of hesperidin as a glycoside of hesperetin (4.0 or 20 mg/kg) (Figure 5E), and administration of isoliquiritigenin at a dose of 4.0 mg/kg to cisplatin-treated rats also inhibited the decrease of plasma acylated-ghrelin compared with control (Figure 5E). Discussion Gastrointestinal tract disorders are considered to be the most common adverse drug reactions to anticancer agents in severity and frequency. The vomiting and anorexia

7 2010 TAKEDA ET AL GASTROENTEROLOGY Vol. 134, No. 7 Figure 4. Plasma acylated-ghrelin levels in 24-hour fasting rats (A) or freely fed rats (B) treated with rikkunshito, 6-hour food intakes in rats treated with rat ghrelin (C) and rats treated with rikkunshito in combination with [D-Lys 3 ] GHRP-6 (D) in cisplatin-treated rats. (A) Rikkunshito was administered orally at dose levels of 500 or 1000 mg/kg at 16 hours before administration of cisplatin and again simultaneously with administration of cisplatin to 24-hour fasting (A) or free feeding (B) rats. Each bar represents the mean SE of 8 rats. (C) Cisplatin was administered intraperitoneally via the tail vein, and after 2 hours, rat acylated-ghrelin was administered intravenously via the tail vein, and food intake was recorded for 4 hours. Each bar represents the mean SE of 8 rats. (D) Rikkunshito was administered orally at a dose of 1000 mg/kg at 16 hours before administration of cisplatin and again simultaneously with administration of cisplatin, alone or in combination with [D-Lys 3 ] GHRP-6. [D-Lys 3 ] GHRP-6 was administered intraperitoneally simultaneously with cisplatin and the second dose of rikkunshito and the food intake in the subsequent 6 hours was determined. Each bar represents the mean SE of 8 rats. P values were obtained with Student s t-test or Dunnett s test. caused by cisplatin involve neurotransmitters such as dopamine, catecholamine, histamine, and acetylcholine in a complex manner in both peripheral and central regions. It has been suggested that these gastrointestinal tract disorders involve 5-HT because the nausea and vomiting occurring as adverse reactions to cisplatin are markedly alleviated when a 5-HT3 receptor antagonist is used in combination with cisplatin. In other words, these gastrointestinal tract disorders are thought to occur when a large amount of 5-HT released from enterochromaffin cells binds to the 5-HT receptors. 1 Thus, the relationship between cisplatininduced emesis and activation of 5-HT3 receptors has been fully recognized. However, contribution of 5-HT receptors to anorexia still remains to be defined. Our hypotheses are as follows: (1) cisplatin causes a decrease in plasma ghrelin concentration, and (2) 5-HT receptors, especially 5-HT2B/2C receptors, are involved in the mechanism of the decrease in plasma ghrelin concentration. In this study, intraperitoneal administration of cisplatin or 5-HT decreased the plasma acylated-ghrelin level in a dose-dependent manner and decreased the 24-hour food intake significantly. In addition, supplementation of rat ghrelin to the cisplatin-treated rats by intravenous administration apparently suppressed the decrease of food intake. These results suggested that cisplatin or 5-HT induced anorexia is mediated by the decrease in plasma ghrelin level. The present study confirmed that administration of 5-HT2B (SB215505) or 5-HT2C (SB242084HCl) receptor antagonists attenuated the decrease in food intake and plasma acylated-ghrelin levels in cisplatin-treated rats. These results strongly imply that activation of 5-HT2B or 2C receptor plays an important role in the decrease in plasma ghrelin level in cisplatin-induced anorexia. More recently, the involvement of 5-HT2B or 2C receptor in ingestive behavior in rats has been characterized using the selective agonists. 2,6 Therefore, to further confirm the results obtained by administration of receptor antagonists, selective agonists 2,4 6 that have abundant evidence in hypophagic effect were administered to untreated rats. Administration of BW723C86 (5-HTB receptor agonist) or mcpp (5-HT2C receptor agonist) clearly caused a decrease in plasma acylated-ghrelin level, and an increase in gastric ghrelin level, but, interestingly, administration of 5-HT3 or 5-HT4 agonists did not.

8 June 2008 RIKKUNSHITO INCREASES PLASMA GHRELIN LEVEL 2011 Table 2. Binding Assays Using 5-HT Receptor Ki values 5-HT receptor subtype and conc. ( mol/l) 2B 2C Aurantii nobilis pericarpium HMF (flavonoids) Nobiletin (flavonoids) Tangeretin (flavonoids) Hesperetin (flavonoids) Naringenin (flavonoids) Glycyrrhizae radix Isoliquiritigenin (flavonoids) Liquiritigenin (flavonoids) Glycycoumarin (coumarins) Zingiberis rhizoma 8-Shogaol (phenols) Gingerol (phenols) Dehydrogingerdion (phenols) Gingerdion (phenols) Gingerol (phenols) ( ) indicates more than 100 mol/l as Ki value. Each value indicates the mean SE of 3 samples as mol/l. A 5-HT3 receptor antagonist is the first-line drug for the prevention or treatment of vomiting, so 5-HT3 receptors also are considered to have a role in gastrointestinal tract disorders caused by cisplatin. 19 Although not shown in the detailed results of this study, the subcutaneous administration of ondansetron was not effective in preventing the decrease of food intake caused by cisplatin (cisplatin-treated group: g/24 h; ondansetron-treated group: g/24 h) or the decrease of plasma acylated-ghrelin (cisplatin-treated group: fmol/ml; ondansetrontreated group: fmol/ml). In contrast to the previous studies concerning vomiting, our results indicated that the administration of ondansetron failed to improve the anorexia in cisplatin-treated rats. The results of this study suggested that cisplatin-induced emesis and anorexia may develop by different mechanisms (ie, 5-HT3 receptors Figure 5. Chemical structures of HMF (A), hesperidin (B), isoliquiritigenin (C), and the effect of HMF (D) and hesperidin (E) or isoliquiritigenin (E) on plasma acylated-ghrelin level in cisplatin-treated rats. A single dose of HMF, hesperidin, or isoliquiritigenin was given orally simultaneously with administration of cisplatin. Two hours after the last dose, the rats were subjected to blood collection by decapitation. Each bar represents the mean SE of 8 rats. P values were obtained with Student s t-test or Dunnett s test.

9 2012 TAKEDA ET AL GASTROENTEROLOGY Vol. 134, No. 7 may be involved in cisplatin-induced emesis and 5-HT2B/2C receptors may be involved in anorexia). To prove this presumption, it is necessary to show that 5-HT2B/2C receptors are not involved in emesis. Because cisplatin-induced emesis does not occur in rats, we could not prove this presumption in this study, and additional investigation using other species such as ferrets is necessary. To assess the release of ghrelin protein into the blood from the stomach, it is necessary to assess changes in the total pool of ghrelin by measuring not only acylatedghrelin but also desacylated-ghrelin, a metabolite of acylated-ghrelin. Total ghrelin concentration is comparable with the sum of acylated- and desacylated-ghrelin concentration and changes in gastric ghrelin levels may be comparable with those in total plasma ghrelin levels. It is presumed from this study that 5-HT2B/2C receptor stimulation inhibits the release of gastric ghrelin protein into the blood, resulting in decreased plasma ghrelin levels and increased gastric ghrelin levels. The ratio of acylated- to desacylated-ghrelin in gastric granules has not been clarified fully. There was a discrepancy between changes in plasma acylated-ghrelin levels and changes in gastric levels after administration of 5-HT2C receptor agonists at low doses. A difference in the activity of the enzyme that degrades acylated-ghrelin might cause a time difference in changes between plasma and gastric levels. It is unlikely that stimulation of the 5-HT2B or 2C receptors affected the synthesis of ghrelin protein because the decrease of plasma acylated-ghrelin level occurred within a very short period of time after the receptor stimulation. In addition, although not shown as the results of this study, when the expression of prepro-ghrelin messenger RNA (mrna) in the gastric fundus was determined by reverse-transcription polymerase chain reaction at 60 minutes after administration of BW723C86 or mcpp, no marked difference from the mrna expression level in the control rats was seen (control: ; BW723C86 16 mg/kg: ; and mcpp 9 mg/kg: ). The 5-HT2B receptors mainly are distributed peripherally (eg, in regions of the gastrointestinal tract, such as the gastric corpus), and 5-HT2C receptors mainly are distributed centrally (eg, in the hypothalamic area). In this study, when the connection between the peripheral and central regions was blocked by vagotomy, the plasma acylated-ghrelin level decreased significantly, although the decrease was small. On the other hand, when cisplatin was administered to vagotomized rats, the plasma acylated-ghrelin level decreased further. This result suggested that the secretion of acylated-ghrelin from stomach to plasma is controlled by the peripheral nerves, but that in the rats administered cisplatin the control of the plasma acylated-ghrelin level may occur in the peripheral tissues. Rikkunshito is prescribed widely in syndromes with a decreased gastric function such as decreased gastric motility after surgery, 11 chronic idiopathic dyspepsia, 12 and abdominal symptoms of gastritis, 14 and the beneficial effects have been shown. Recently, rikkunshito was used in combination with granisetron to alleviate the side effects of anticancer drugs without affecting their efficacy. 13 The common symptoms among all these diseases are decreased gastric motility and decreased appetite, which may be caused by a decreased plasma ghrelin level 9,10 or ghrelin function. In the clinical application, rikkunshito is particularly effective against these symptoms, 12,14 and its efficacy also is supported by basic research in rats with delayed gastric emptying function. 15 These findings suggest the possibility that rikkunshito affects the secretion or function of ghrelin. On the other hand, the results obtained in this study showing that rikkunshito ameliorated the reduced food intake and plasma acylated-ghrelin level suggest that the increased plasma acylated-ghrelin enhanced motility of the gastrointestinal tact, including the stomach, 15 improving gastric emptying and stimulating the appetite center in the brain via the vagus nerve, leading to increased feeding behaviors. We screened 33 compounds out of many components contained in rikkunshito (all data not shown). As a result, 13 compounds showed antagonistic activity against binding to any of 5-HT 1A, 1B/D, 2A, 2B, 2C, 3, 4, 6, and 7 receptors. Among the 13 compounds that are active for receptor binding inhibition, the HMF, nobiletin, and tangeretin contained in Aurantii nobilis pericarpium had potent 5-HT2B receptor antagonistic activity with Ki values of 0.21 to 0.59 mol/l. In addition, 8-shogaol contained in Zingiberis rhizoma has a relatively high 5-HT2B receptor antagonistic activity with a Ki value of 1.8 mol/l. The inhibition activity of hesperetin against the 5-HT2B receptor is weaker than that of the first 3 compounds, but the concentration of hesperidin in rikkunshito is the highest among the ingredients tested in the binding assay, 15 and hesperidin is converted quickly to the aglycon hesperetin after oral administration. In addition, isoliquiritigenin, which is an ingredient of Glycyrrhizae radix, has the most potent antagonistic activity against the 5-HT2C receptor of all the components, besides having 5-HT2B receptor antagonistic activity. Therefore, we conducted an in vivo assessment using components of high selectivity and potency against 5-HT2B receptors, high concentration in the rikkunshito s formulation, and activity against 5-HT2C receptors as the index. Our study indicated that the administration of HMF, isoliquiritigenin, and hesperidin attenuated the decrease in plasma ghrelin level, although tangeretin, nobiletin, and 8-shogaol did not (in vivo data not shown). Contrary to the efficacy of HMF, nobiletin, tangeretin, and 8-shogaol showed contradictory results in vivo compared with the results of the binding assay (in vivo data not shown). We did not investigate the in vivo effects of these components on plasma ghrelin level at various times intervals after administration. It was speculated that the blood pharmacokinetics of these 3 components might be different from those of HMF because these 3 components have a different chemical structure. In addition, the amount of nobiletin, tangeretin, and 8-shogaol contained in herbal medicine is very small, indicating that the involvement of these components

10 June 2008 RIKKUNSHITO INCREASES PLASMA GHRELIN LEVEL 2013 among rikkunshito on ghrelin secretion may be lessened. The efficacy of hesperetin and isoliquiritigenin in plasma ghrelin suggested that the ingredients that inhibit 5-HT2C receptor binding in addition to 5-HT2B receptor are likely to be effective in vivo. Hesperetin and isoliquiritigenin have very weak antagonistic action for 5-HT4 and 5-HT6 receptor in addition to the 5-HT2B receptor antagonism. The relationship between 5-HT4/5-HT6 receptor and plasma ghrelin is not currently known. In this study, because administration of 5-HT4 agonist did not affect plasma ghrelin levels, the involvement of 5-HT4 in the in vivo evaluation results also seemed to be small. In addition, because the Ki values for each receptor are high, it is supposed that the activity is negligible at the dose used in the in vivo evaluation. The induction of ghrelin secretion by rikkunshito is supposed to be based on the 5-HT2B/2C receptor antagonism owing to multiple active ingredients. Further exploration and clarification of active ingredients with 5-HT2B/2C receptor antagonist activity is necessary. We speculate that in this study, candidates for the representative active ingredients, HMF, hesperidin, and isoliquiritigenin only were identified. In summary, anorexia caused by cisplatin therapy mainly is attributable to a decrease in plasma acylatedghrelin levels caused by stimulation of 5-HT2B/2C receptors via 5-HT secretion. This study suggested that rikkunshito, a herbal medicine, may improve cisplatininduced gastrointestinal dysfunction through improvement of ghrelin release owing to the 5-HT2B/2C antagonistic action of flavonoids contained in rikkunshito. References 1. Cubeddu LX, Hoffmann IS. Participation of serotonin on early and delayed emesis induced by initial and subsequent cycles of cisplatin-based chemotherapy: effects of antiemetics. J Clin Pharmacol 1993;33: De Vry J, Schreiber R. Effects of selected serotonin 5-HT1 and 5-HT2 receptor agonists on feeding behavior: possible mechanisms of action. Neurosci Biobehav Rev 2000;24: Kursar JD, Nelson DL, Wainscott DB, et al. Molecular cloning, functional expression, and pharmacological characterization of a novel serotonin receptor (5-hydroxytryptamine 2F) from rat stomach fundus. Mol Pharmacol 1992;42: Wright DE, Seroogy KB, Lundgren KH, et al. Comparative localization of serotonin 1A, 1C and 2 receptor subtype mrnas in rat brain. J Comp Neurol 1995;351: Nonogaki K, Strack AM, Dallman MF, et al. Leptin-independent hyperphagia and type 2 diabetes in mice with a mutated serotonin 5-HT2C receptor gene. Nat Med 1998;4: Hayashi A, Suzuki M, Sasamata M, et al. Agonist diversity in 5-HT2C receptor-mediated weight control in rats. Psychopharmacology 2005;178: Kojima M, Hosoda H, Date Y, et al. Ghrelin is a growth-hormonereleasing acylated peptide from stomach. Nature 1999;402: Asakawa A, Inui A, Kaga T, et al. Antagonism of ghrelin receptor reduces food intake and body weight gain in mice. Gut 2003;52: Shinomiya T, Fukunaga M, Akamizu T, et al. Plasma acylated ghrelin levels correlate with subjective symptoms of functional dyspepsia in female patients. Scand J Gastroenterol 2005;40: Isomoto H, Ueno H, Nishi Y, et al. Circulating ghrelin levels in patients with various upper gastrointestinal disease. Dig Dis Sci 2005;50: Yagi M, Homma S, Kubota M, et al. The herbal medicine Rikkunshi-to stimulates and coordinates the gastric myoelectric activity in post-operative dyspeptic children after gastrointestinal surgery. Pediatr Surg Int 2004;19: Tatsuta M, Iishi H. Effect of treatment with liu-jun-zi-tang(tj-43) on gastric emptying and gastrointestinal symptoms in dyspeptic patients. Aliment Pharmacol Ther 1993;7: Tomono H, Ito Y, Watanabe T. Successful antiemetic treatment of Tsumura rikkunshi-to extract granules for ethical use in addition to other antiemetic agents in neoadjuvant chemotherapy for an advanced breast cancer patient. Jpn J Cancer Chemother 2006;33: Miyoshi A, Yachi A, Masamune O, et al Clinical evaluation of the efficacy of TJ-43 Tsumura Rikkunshi-to on chronic gastritis with abdominal symptom-multicenter groups study in comparison with cisapride. Prog Med 1991;11: Kido T, Nakai Y, Kase Y, et al. Effects of rikkunshi-to, a traditional Japanese medicine, on the delay of gastric emptying induced by N(G)-nitro-L-arginine. J Pharmacol Sci 2005;98: Masuda Y, Tanaka T, Inomata N, et al. Ghrelin stimulates gastric acid secretion and motility in rats. Biochem Biophys Res Commun 2000;276: Wolf WA, Schuts JS. The serotonin 5-HT2C receptor is a prominent serotonin receptor in basal ganglia; evidence from functional studies on serotonin-mediated phosphoinositide hydrolysis. J Neurochem 1997;69: Butcher ME. Global experience with ondansetron and future potential. Oncology 1993;50: Ozaki A, Sukamoto T. Improvement of cisplatin-induced emesis and delayed gastric emptying by KB-R6933, a novel 5-HT3 receptor antagonist. Gen Pharmacol 1999;33: Received August 8, Accepted February 28, Address reprint requests to: Hiroshi Takeda, MD, PhD, Department of Gastroenterology and Hematology, Hokkaido University Graduate School of Medicine, N15 W7, Kita-ku, Sapporo , Japan. h_takeda@med.hokudai.ac.jp; fax: (81) Hiroshi Takeda has received grant support from Tsumura & Co. Chiharu Sadakane and Tomohisa Hattori are employees of Tsumura & Co.