Antidepressant-Like Behavioral and Neurochemical Effects of the Citrus

Available online at www.sciencedirect.com

Life Sciences 82 (2008) 741–751 www.elsevier.com/locate/lifescie

Antidepressant-like behavioral and neurochemical effects of the citrus-associated chemical apigenin ⁎ Li-Tao Yi, Jian-Mei Li, Yu-Cheng Li, Ying Pan, Qun Xu, Ling-Dong Kong

State Key Laboratory of Pharmaceutical Biotechnology, School of Life Sciences, Nanjing University, Nanjing 210093, PR China Received 14 July 2007; accepted 16 January 2008

Abstract

Apigenin is one type of bioflavonoid widely found in citrus fruits, which possesses a variety of pharmacological actions on the central nervous system. A previous study showed that acute intraperitoneal administration of apigenin had antidepressant-like effects in the forced swimming test (FST) in ddY mice. To better understand its pharmacological activity, we investigated the behavioral effects of chronic oral apigenin treatment in the FST in male ICR mice and male Wistar rats exposed to chronic mild stress (CMS). The effects of apigenin on central monoaminergic neurotransmitter systems, the hypothalamic–pituitary–adrenal (HPA) axis and platelet adenylyl cyclase activity were simultaneously examined in the CMS rats. Apigenin reduced immobility time in the mouse FST and reversed CMS-induced decrease in sucrose intake of rats. Apigenin also attenuated CMS-induced alterations in serotonin (5-HT), its metabolite 5-hydroxyindoleacetic acid (5-HIAA), dopamine (DA) levels and 5-HIAA/ 5-HT ratio in distinct rat brain regions. Moreover, apigenin reversed CMS-induced elevation in serum corticosterone concentrations and reduction in platelet adenylyl cyclase activity in rats. These results suggest that the antidepressant-like actions of oral apigenin treatment could be related to a combination of multiple biochemical effects, and might help to elucidate its mechanisms of action that are involved in normalization of stress- induced changes in brain monoamine levels, the HPA axis, and the platelet adenylyl cyclase activity. © 2008 Elsevier Inc. All rights reserved.

Keywords: Apigenin; Antidepressant; Serotonin; 5-Hydroxyindoleacetic acid; Dopamine; Corticosterone; Adenylyl cyclase

Introduction establishing normal mood by antidepressants (Duman et al., 1997; Nestler et al., 2002a; Coyle and Duman, 2003). Recent Depression is a serious emotional disorder with estimated research in the field has had the goal of discovering new targets lifetime prevalence as high as 21% of the general population in and developing novel therapeutics that act faster with higher some developed countries (Gainotti et al., 2001; Wong and efficacy and fewer side effects. Licinio, 2001; Nestler et al., 2002a,b). The neurobiology of Dysregulation of the hypothalamic–pituitary–adrenal (HPA) depression and its response to antidepressant treatment are axis is one of the most prominent neurobiological findings in not well understood. Some of the research on depression has major depressive disorder, and is considered as another im- focused on the interactions between the monoamine neuro- portant mechanism in the investigation of new antidepressant transmitters and their reuptake and receptor proteins. However, agents (Heuser et al., 1996; Nickel et al., 2003; Young et al., pharmacotherapy for depression often requires week- or month- 2004; Aihara et al., 2007). Adenylyl cyclase is an enzyme long treatments despite the fact that antidepressants immedi- that regulates the physiological effects of numerous drugs and ately affect the brain monoamine neurotransmission (Nestler, hormones through the production of cyclic adenosine-3′,5′- 1998), suggesting that other mechanisms may be involved in re- monophosphate (cAMP). Clinical and epidemiologic research has provided suggestive evidence regarding the association between adenylyl cyclase activity and major depression (Cowburn ⁎ Corresponding author. Tel.: +86 25 8359 4691; fax: +86 25 8359 4691. et al., 1994; Reiach et al., 1999). Patients with major depression E-mail address: [email protected] (L.-D. Kong). were observed to have lower platelet adenylyl cyclase activity

(Cowburn et al., 1994; Menninger and Tabakoff, 1997). No- Furthermore, other reports demonstrated that apigenin inhibited ticeably, antidepressants could dramatically attenuate the reduc- γ-aminobutyric acid (GABA) receptor function and reduced tion of platelet adenylyl cyclase activity in depressed patients N-methyl-D-aspartate (NMDA) receptor function. Since dif- (Hines and Tabakoff, 2005). Thus, platelet adenylyl cyclase ferent GABA and NMDA receptor antagonists were effective activity might serve as a biological marker for major depression in many animal models of depression, these could account for and the therapeutic effect of antidepressants (Reiach et al., apigenin's antidepressant activity (Skolnick, 1999; Nakazawa 1999; Donati and Rasenick, 2003; Hines and Tabakoff, 2005; et al., 2003). These findings indicated that the antidepressant- Abdel-Razaq et al., 2007). like effects of apigenin might not be explained by only one Citrus fruits may be of potential interest to pharmaceuti- mechanism. To better understand the pharmacological activity of cal and food industries since they contain bioactive bioflavo- apigenin, we investigated the behavioral effects of oral apigenin noids with health-related properties. These components act as treatment in the FST in male ICR mice and in the chronic mild anti-oxidants in various biological systems (Morel et al., 1993; stress (CMS) model in male Wistar rats. The effects of apigenin Salah et al., 1995). They prevented pregnancy (Garg et al., on monoaminergic function in various brain regions, serum 2001), inhibited cancer cell proliferation (Manthey and Guthrie, corticosterone concentrations (an index of the HPA axis status), 2002), and displayed anti-allergic and anti-inflammatory ac- and platelet adenylyl cyclase activity were simultaneously studied tivities (Struckmann and Nicolaides, 1994). Early animal stud- in the CMS rats. ies confirmed that citrus fragrance reduced immobility time in the forced swimming test (FST) (Komori et al., 1995a,b) and Materials and methods accelerated the metabolic turnover of dopamine (DA) in hippocampus and of serotonin (5-HT) in prefrontal cortex and Materials striatum. Moreover, these effects were significantly blocked by pre-treatment with apomorphine, a nonselective DA receptor Apigenin was obtained from Shanxi Huike Botanical agonist, but not by agonists or antagonists to 5-HT receptor and Development Co., Ltd. (purityN98% by HPLC). Fluoxetine α-2 adrenaline receptor, indicating that citrus exerted anti- hydrochloride was purchased from Changzhou Siyao Pharma- depressant-like effects via modulating the 5-HT and DA func- ceuticals Co., Ltd. (P. R. China). All other chemicals used tions (Komiya et al., 2006). It was reported that citrus fragrance were of high-purity analytical grade obtained from commercial markedly reduced the doses of antidepressant treatment needed sources. for depressed patients in the clinic (Komori et al., 1995c). These observations have aroused our interest in searching for new Animals agents with antidepressant-like action from citrus. Apigenin (Fig. 1), one type of bioflavonoid widely found in Male ICR mice (Laboratory Animal Center, Nanjing Uni- citrus fruit, has been demonstrated to have anti-oxidation, anti- versity of Traditional Chinese Medicine, Jiangsu Province, P. R. inflammatory, and anti-tumor activities (Chen et al., 2005; Czyz China), weighing 23–25 g, were used. Animals were housed 5 et al., 2005; Fang et al., 2005; Hougee et al., 2005). Apigenin per cage (320×180×160 cm) under a 12-h/12-h light/dark was found to exert a variety of pharmacological actions on the schedule with the lights on at 07:00 a.m. and had free access to central nervous system, such as anxiolytic and sedative pro- tap water and food pellets. Ambient temperature and relative perties (Avallone et al., 2000; Zanoli et al., 2000). It was humidity were maintained at 22±2 °C and 55±5%, respectively. reported that acute intraperitoneal (i.p.) administration They were allowed at least 1 week to adapt to the laboratory of apigenin in ddY mice decreased immobility time in the environment before experiments. Experiments, performed by an FST and attenuated swim stress-induced decrease in DA turn- observer that was unaware of the treatment each mouse had over in amygdala and increase in DA turnover in hypothalamus, received, were carried out between 1:00 p.m. and 3:00 p.m. indicating that apigenin possessed antidepressant-like ef- Male Wistar rats (220–250 g), purchased from the Lab- fects, which might be mediated by dopaminergic mechanisms oratory Animal Center, Nanjing University of Traditional (Nakazawa et al., 2003). In addition, apigenin inhibited mono- Chinese Medicine, Jiangsu Province, P. R. China, were used in amine oxidase (MAO) activity (Lorenzo et al., 1996; Han et al., the experiments. Except as described below, the rats were sin- 2007). MAO inhibitors increase the levels of brain mono- gly housed and were kept on a 12-h light/dark cycle under con- amines, such as 5-HT, that have been related to the allevia- trolled temperature at 22±2 °C and humidity at 55±5%. They tion of clinical depression (Kanazawa, 1994; Wouters, 1998). were allowed free access to a laboratory chow diet and water. All studies were conducted in accordance with the Institu- tional Animal Care Committee at the Nanjing University and the China council on Animal Care at Nanjing University.

FST in mice

The FST used was the same as described in detail elsewhere (Porsolt et al., 1977, 1978). Briefly, mice were individually Fig. 1. Structure of apigenin. placed in a glass cylinder (20 cm in height, 14 cm in diameter) L.-T. Yi et al. / Life Sciences 82 (2008) 741–751 743 with 10-cm deep water (25±2 °C). All animals were forced to fluctuations in hormone levels due to circadian rhythms, rats swim for 6 min, and the duration of immobility was observed were bled at 09:00 a.m. and 10:00 a.m. Some blood samples and measured during the final 4 min interval of the test. The were immediately collected on ice and separated in a immobility period was regarded as the time spent by the mouse refrigerated centrifuge at 4 °C. Serum was stored at −80 °C floating in the water without struggling and making only those until corticosterone assay was performed. The regions of the rat movements necessary to keep its head above the water. The brain including frontal cortex, hippocampus, striatum, hypotha- animals were used only once in this test. lamus and nucleus accumbens were carefully dissected and Different groups of mice, 10 animals per group, were used stored in liquid nitrogen for determination of monoamine for drug treatment and for each test. Drugs were dispersed in neurotransmitter levels. water. All doses are expressed as milligrams per kilogram body weight of the respective drugs. Food, but not water, was Determination of monoamine neurotransmitter levels withdrawn from the animals 1 h prior to drug administration. Apigenin (10 and 20 mg/kg), fluoxetine hydrochloride 5-HT, 5-HIAA, noradrenaline (NA) and DA levels in various (15 mg/kg) and water (vehicle) were administered in a vol- brain regions of the control and CMS rats were determined by ume of 15 ml/kg by gastric gavage once daily at 11:00 a.m.– modification of methods described in the literature (Welch and 12:00 p.m. to different groups of mice in the FST. Animals were Welch, 1969; Curzon and Green, 1970). Briefly, each brain subjected to the FST 1 h after drug treatment for 1 week and region was homogenized in acidified n-butanol and shaken for 2 weeks, respectively. 5 s. It was then centrifuged at 3000 ×g for 5 min at 4 °C. The supernatant was mixed with n-heptane and 0.1 M HCl, shaken CMS procedure for 30 s and centrifuged at 3000 ×g for 5 min at 4 °C. The aqueous phase was retained for determination of 5-HT, NA, DA According to a modification of the method of Willner et al. and the organic phase for determination of 5-HIAA. The ratio of (1987), animals were first trained to consume 1% (w/v) sucrose 5-HIAA/5-HT was used as an index of 5-HT turnover. solution before the start of the CMS protocol. The baseline tests of sucrose solution intake were performed 9 times (once every Assay of corticosterone levels three days), in which rats could select between two pre-weighed bottles, one with 1% (w/v) sucrose solution and one with tap Serum corticosterone levels in the control and CMS rats water, after 14 h food and water deprivation (Bekris et al., 2005; were measured using an enzyme immunoassay kit (Adlitteram Grippo et al., 2006). On the basis of sucrose preference fol- Diagnostic Laboratories Inc.). The minimum detectable con- lowing the last training test, the animals with varying sucrose centration of corticosterone in this assay was estimated to be preference scores were more or less equally distributed into two 0.1 nmol/l. matched groups, control and stressed rats. Stressed rats were subjected to the CMS procedure for an 8-week period. The Adenylyl cyclase activity assay weekly stress regime consisted of two periods of food or water deprivation, 45 degree cage tilt, intermittent illumination (lights The platelet adenylyl cyclase activity in the control and CMS on and off every 2 h), soiled cage (250 ml water in sawdust rats was determined by the method of Hines and Tabakoff bedding), paired housing, and low intensity stroboscopic illu- (2005). Blood samples were centrifuged at 700 ×g for 10 min at mination (150 flashes/min), and two periods of no stress. All room temperature. The platelet-rich layer was transferred to a stressors were 10–14 h of duration and were applied in- fresh centrifuge tube, and this procedure was repeated. The dividually and continuously, day and night. Control animals upper platelet-rich layer was transferred again to a fresh tube were housed in separate room and had no contact with the for 15-min centrifugation (2800 ×g) at room temperature. The stressed groups. They were deprived of food and water for the platelet pellet was thawed and washed twice at 4 °C by sus- 14 h preceding each sucrose test, but otherwise food and water pension in 1.5 ml of 50 mmol/l Tris–HCl (pH 7.5) containing were freely available in the home cage. 20 mmol/l EDTA, followed by centrifugation at 17,000 ×g for On the basis of their sucrose intake scores following 4 weeks 10 min. The final pellet was resuspended in 1.5 ml of 5 mmol/l of CMS, both stressed and control rats were divided into Tris–HCl (pH 7.5) containing 5 mmol/l EDTA, with a hand- matched subgroups (n=8). These groups received water (ve- held Teflon homogenizer. The homogenate was diluted with hicle, 5 ml/kg), apigenin (7 and 14 mg/kg) or fluoxetine 5 mmol/l Tris–HCl (pH 7.5) containing 1 mmol/l EDTA to hydrochloride (7 mg/kg) by daily oral administration beginning attain a protein concentration of 200–1000 μg/ml and im- at Week 4. All drug administrations were given at 10:00 a.m. for mediately assayed for platelet adenylyl cyclase activity. Protein the subsequent 4 weeks and the weekly sucrose tests were determinations were performed according to the method of carried out 24 h after the last drug treatment. Lowry et al. (1951). Adenylyl cyclase activity was determined by the calcula- Blood and tissue sample collection tion of cAMP quantity, resulting after incubation (15 min) of membrane preparations (10–15 μg of protein) in standard me- After the last sucrose intake test, all animals were left dium including 25 mmol/l Tris-maleate at pH 7.5, 10 mmol/l without any treatment until the following morning. To avoid theophylline, 4 mmol/l MgCl2, 25 mmol/l ATP, and [α-32P] 744 L.-T. Yi et al. / Life Sciences 82 (2008) 741–751

ATP at 1.2–2.0 ×10 cpm/assay. The incubation was terminated on the FST behavior in this study [F(1,18)=1.37, P=0.255]. by heating (96 °C) for 3 min. Determination of cAMP con- Administration for 2 weeks of apigenin also modified im- tent was performed with a commercially available RIA Kit mobility time [F(2,27)=8.69, Pb0.01]. Apigenin noticeably re- (Shanghai University of Traditional Chinese Medicine). duced immobility time at two doses [10 mg/kg: F(1,18)=13.47, Pb0.01; 20 mg/kg: [F(1,18)=12.64, Pb0.01]. Fluoxetine hy- Statistical analyses drochloride treatment at 15 mg/kg reduced immobility in the mouse FST [F(1,18)=28.30, Pb0.001]. Data are expressed as the mean±S.E.M. Behavioral data from the FST were analyzed using a one-way analysis of Effects of apigenin on 1% sucrose and water intake in the CMS variance (ANOVA). Data on body weight, water and sucrose rats intakes from the CMS experiments were analyzed using a repeated ANOVA with treatment and stress as between factors Fig. 3 presents water (Panel A) and sucrose (Panel B) intake and time (weeks) as within factor. For the estimation of bio- in the control and CMS rats at baseline and following 8 weeks chemical results in the CMS experiments, a two-way ANOVA of CMS. Compared with the control rats, neither CMS nor (stress×treatment) was performed. According to the statistically drug treatment significantly affected water intake of animals significant interactions revealed by the aforementioned multi- throughout the study. factor analysis, a separate analysis was performed in order to A one-way ANOVA revealed that baseline sucrose intake reveal specific differences between groups. Following ANOVA did not differ between the control and CMS groups (Week 0). analyses, LSD post hoc tests were used. A probability level of From Week 0 to Week 4, a repeated ANOVA with stress as Pb0.05 was taken to be statistically significant in the analyses. independent factor and week as repeated factor, revealed statistically significant effects of stress [F(1,62)=17.10, Pb0.001], Results week [F(4,248)=2.57, P b0.05] and stress×week interaction [F(4,248)=5.21, P b0.001] on sucrose intake. Further separate Effects of apigenin on immobility time in the mouse FST repeated ANOVA, with week as repeated factor, revealed a gradual sucrose intake increase in control rats [F(4,124)=5.12, The effects of apigenin and fluoxetine hydrochloride on P b0.01], and a significant sucrose intake decline in the CMS immobility time in the mouse FST are shown in Fig. 2. rats [F(4,124)=3.02, P b0.05] (Fig. 3B). In addition, sucrose Administration for 1 week of apigenin modified immobility intake in the CMS groups was significantly lower than that time [F(2,27)=3.57, Pb0.05] in the mouse FST. Post-hoc incontrolgroupsatWeek2[F(1,63)=9.02, P b0.01], Week 3 analysis revealed that apigenin at 20 mg/kg significantly [F(1,63)=14.72, P b0.001] and Week 4 [F(1,63)=15.54, decreased immobility time [F(1,18)=5.52, Pb0.05]. However, P b0.001]. Such a difference between the control and CMS apigenin at 10 mg/kg exhibited a slight but insignificant re- animals treated with vehicle, persisted for the drug-treatment duction in immobility time [F(1,18)=4.28, P=0.053]. Fluox- period [Week 5: F(1,14)=5.36, Pb0.05; Week 6: F(1,14)=5.18, etine hydrochloride treatment at 15 mg/kg was devoid of effect Pb0.05; Week 7: F(1,14)=11.14, Pb0.01; Week 8: F(1,14)= 6.09, Pb0.05]. Furthermore, two-way ANOVA performed on sucrose intake yielded significant stress effect [F(1,14)=11.73, Pb0.01], and no week effect [F(3,42)=0.11, P=0.950] and stress×week interaction [F(3,42)=0.17, P=0.917]fromWeek5 to Week 8 (Fig. 3B). In rats treated with apigenin, repeated ANOVA indicated significant week effect [F(4,84)=2.65, Pb0.05], treatment effect [F(2,21)=5.43, Pb0.05] and no week×treatment interaction [F(8,84)=0.86, P=0.548]. Apigenin treatment reversed CMS-induced deficit in sucrose intake and the onset of ame- lioration, i.e. a significant increase in sucrose intake was seen at Week 5 and remained in subsequent weeks in the CMS rats treated with 14 mg/kg apigenin [Week 5: F(1,14)=5.57, Pb0.05; Week 6: F(1,14)=8.10, Pb0.05; Week 7: F(1,14)= 11.99, P b0.01; Week 8: F(1,14)=13.19, P b0.01] compared with the sucrose intake at Week 4. Significant increases in sucrose intake were observed in the CMS rats treated with Fig. 2. Effects of apigenin and fluoxetine hydrochloride on immobility time in 7 mg/kg apigenin at Week 7 [F(1,14)=4.61, Pb0.05] and Week the mouse FST. The duration of immobility was counted in mice orally 8[F(1,14)=4.77; Pb0.05] (Fig. 3B). administered with water (vehicle), fluoxetine hydrochloride (15 mg/kg) and In the CMS rats treated with fluoxetine hydrochloride, re- apigenin (10 and 20 mg/kg) for 1 week or 2 weeks. Data are expressed as the mean±S.E.M. of 10 animals. A one-way ANOVA showed an effect of apigenin peated ANOVA indicated significant treatment effect [F(1,14)= treatment (one-week: Pb0.05; two-week: Pb0.01). For statistical significance, 9.93, Pb0.01], and no week effect [F(4,56)=1.27, P=0.290] ⁎Pb0.05, ⁎⁎Pb0.01 and ⁎⁎⁎Pb0.001 compared to the vehicle. and week×treatment interaction [F(4,56)=1.18, P=0.329]. L.-T. Yi et al. / Life Sciences 82 (2008) 741–751 745

apigenin or fluoxetine hydrochloride (Fig. 3). The body weight of animals in the whole experiment showed a continual increase, but there was no significant difference between any two groups (data not shown).

Effects of apigenin on 5-HT, 5-HIAA, NA and DA levels and 5-HIAA/5-HT ratio in various brain regions in the CMS rats

Table 1 shows the effects of apigenin and fluoxetine hydrochloride on 5-HT, 5-HIAA, NA and DA levels and 5-HIAA/ 5-HT ratio in various brain regions in the control and CMS rats.

Prefrontal cortex In the CMS rats treated with apigenin, two-way ANOVA indicated significant stress effect [F(1,36)=6.20, Pb0.05], treatment effect [F(2,36)=5.89, Pb0.01] and no stress×treatment interaction [F(2,36)=1.32, P=0.278] on 5-HT levels; there was no significant effect of stress, treatment or stress×treatment interaction both on 5-HIAA levels and 5-HIAA/5-HT ratio. One-way ANOVA revealed that CMS induced significant decreases in 5-HT levels [F(1,12)=7.58, P b0.05] and 5-HIAA [F(1,12)=4.87, P b0.05] without change in 5-HIAA/5-HT ratio [F(1,12)=0.00, P =0.985]. Compared with vehicle-treated CMS animals, apigenin treatment at 14 mg/kg significantly increased 5-HT levels [F(1,12)=6.65, P b0.05], 5-HIAA [F(1,12)=4.82, P b0.05] and did not affect 5-HIAA/5-HT ratio. The CMS rats receiving 7 mg/kg apigenin showed un- altered 5-HT, 5-HIAA levels and 5-HIAA/5-HT ratio. Two- way ANOVA indicated stress×treatment interaction on DA levels [F(2,36)=3.90, P b0.05]. CMS caused a significant increase in DA levels [F(1,12)=7.48, P b0.05]. Apigenin treatment only at 14 mg/kg noticeably decreased DA levels compared with vehicle-treated CMS animals [F(1,12)=7.05, P b0.05]. In the CMS rats treated with fluoxetine hydrochloride, two- way ANOVA indicated significant stress effect [F(1,24)=5.97, Fig. 3. Effects of apigenin and fluoxetine hydrochloride on 1% sucrose and b b water intake in the control and CMS rats. Sucrose and water intake was P 0.05], treatment effect [F(1,24)=12.88, P 0.01] and no measured in the control and CMS rats receiving water (vehicle), 7 mg/kg stress×treatment interaction [F(1,24)=0.57, P=0.456] on 5- fluoxetine hydrochloride, 7 and 14 mg/kg apigenin. Data are expressed as the HT levels. However, there was no significant effect of stress, mean±S.E.M. (n=8). A one-way ANOVA performed on baseline sucrose and treatment and stress×treatment interaction on 5-HIAA levels water intake showed no significant difference between the control and CMS and 5-HIAA/5-HT ratio. Fluoxetine hydrochloride treatment groups (Week 0). From Week 0 to Week 4, a repeated ANOVA showed effects of stress (Pb0.001), week (Pb0.05) and stress×week interaction (Pb0.001) on significantly exhibited elevations in 5-HT [F(1,12)=9.45, sucrose intake but not on water intake. A two-way ANOVA performed on Pb0.01] and 5-HIAA [F(1,12)=8.81, Pb0.05] levels without sucrose intake yielded an effect of stress (Pb0.01) from Week 5 to Week 8. A altering 5-HIAA/5-HT ratio [F(1,12)=0.00, P=0.972]. In ad- repeated ANOVA showed effects of week (Pb0.05) and apigenin treatment dition, there was a significant stress effect [F(1,24)=5.49, b + b (P 0.05) in the CMS rats. For statistical significance, P 0.05 compared to the Pb0.05], treatment effect [F(1,24)=9.13, Pb0.01] and no baseline level (Week 0); ##Pb0.01 and ###Pb0.001 compared to control vehicle, and ⁎Pb0.05, ⁎⁎Pb0.01 and ⁎⁎⁎Pb0.001 compared to drug-treated CMS stress×treatment interaction [F(1,24)=3.24, P =0.084] on animals at Week 4. DA levels. Fluoxetine hydrochloride treatment significantly reversed CMS-induced increase in DA levels [F(1,12)=17.82, Pb0.01]. Fluoxetine hydrochloride gradually reversed CMS-induced decrease in sucrose consumption from Week 6 compared with the Hippocampus sucrose intake at Week 4 [Week 6: F(1,14)=5.67, Pb0.05; In the CMS rats treated with apigenin, two-way ANOVA Week 7: F(1,14)=9.19, Pb0.01; Week 8: F(1,14)=9.62, indicated no significant effect of stress, treatment and stres- Pb0.01]. s×treatment interaction on 5-HTand 5-HIAA levels and 5-HIAA/ In all the control groups, neither water intake nor sucrose 5-HT ratio. Separate one-way ANOVA revealed that CMS pro- intake was significantly increased after a 4-week treatment with duced a significant decrease in 5-HT levels [F(1,12)=7.55, 746 L.-T. Yi et al. / Life Sciences 82 (2008) 741–751

Table 1 Effects of apigenin and fluoxetine hydrochloride on 5-HT, 5-HIAA, NA and DA levels and 5-HIAA/5-HT ratio in various brain regions in the control and CMS rats Group Does 5-HT 5-HIAA NA DA 5-HIAA/5-HT ratio (mg/kg) (μg/g wet tissue) Prefrontal cortex Control Vehicle – 0.515±0.0598 0.264±0.0350 0.460±0.0497 0.408±0.0416 0.50±0.03 Apigenin 7 0.574±0.0256 0.250±0.0516 0.420±0.0345 0.415±0.0919 0.44±0.07 14 0.633±0.0537 0.262±0.0277 0.468±0.0475 0.435±0.0662 0.43±0.03 Fluoxetine 7 0.672±0.0502 0.327±0.0767 0.523±0.0399 0.346±0.0517 0.48±0.06 CMS Vehicle – 0.338±0.0233# 0.173±0.0216# 0.472±0.0321 0.618±0.0647# 0.52±0.04 Apigenin 7 0.391±0.0514 0.169±0.0184 0.403±0.0467 0.624±0.0522 0.42±0.04 14 0.623±0.108⁎⁎ 0.270±0.0383⁎ 0.444±0.0542 0.315±0.0800⁎⁎ 0.56±0.07 Fluoxetine 7 0.579±0.0746⁎ 0.257±0.0179⁎ 0.497±0.0485 0.374±0.0413⁎ 0.45±0.04

Hippocampus Control Vehicle – 0.649±0.0619 0.309±0.0400 0.519±0.0848 0.357±0.0329 0.50±0.05 Apigenin 7 0.666±0.0784 0.240±0.0545 0.511±0.0511 0.380±0.0504 0.39±0.04 14 0.635±0.0547 0.280±0.0471 0.528±0.0429 0.368±0.0580 0.44±0.07 Fluoxetine 7 0.715±0.0541 0.308±0.0492 0.509±0.0553 0.369±0.0516 0.43±0.07 CMS Vehicle – 0.411±0.0607# 0.296±0.0511 0.493±0.0642 0.359±0.0476 0.82±0.06## Apigenin 7 0.664±0.0744⁎ 0.295±0.0586 0.504±0.0634 0.345±0.0532 0.44±0.07⁎ 14 0.631±0.0501⁎ 0.231±0.0249 0.502±0.0538 0.339±0.0594 0.39±0.07⁎⁎ Fluoxetine 7 0.698±0.0521⁎⁎ 0.294±0.0688 0.513±0.0500 0.369±0.0493 0.43±0.06⁎

Striatum Control Vehicle – 0.613±0.0592 0.400±0.0283 0.644±0.0323 4.39±0.272 0.68±0.06 Apigenin 7 0.661±0.0783 0.421±0.0583 0.623±0.0702 4.20±0.524 0.60±0.05 14 0.666±0.0404 0.426±0.0313 0.635±0.0758 4.10±0.400 0.64±0.04 Fluoxetine 7 0.641±0.0600 0.501±0.0668 0.641±0.0587 4.59±0.402 0.66±0.05 CMS Vehicle – 0.517±0.0332 0.305±0.0111## 0.653±0.0532 4.56±0.379 0.60±0.04 Apigenin 7 0.538±0.0666 0.378±0.0531 0.635±0.0542 4.30±0.630 0.70±0.04 14 0.604±0.0557 0.416±0.0379⁎ 0.626±0.0467 4.17±0.401 0.73±0.05 Fluoxetine 7 0.612±0.0665 0.455±0.0579⁎ 0.640±0.0488 4.52±0.320 0.65±0.05

Hypothalamus Control Vehicle – 0.952±0.0452 0.488±0.0450 2.51±0.177 0.453±0.0565 0.52±0.05 Apigenin 7 0.936±0.0533 0.479±0.0616 2.46±0.186 0.451±0.0322 0.51±0.07 14 1.03±0.0861 0.455±0.0202 2.78±0.114 0.425±0.0452 0.44±0.02 Fluoxetine 7 1.08±0.0431 0.550±0.0621 2.70±0.169 0.466±0.0302 0.50±0.06 CMS Vehicle – 0.757±0.0608# 0.476±0.0669 2.51±0.246 0.452±0.0197 0.65±0.06 Apigenin 7 0.795±0.0540 0.469±0.0219 2.45±0.346 0.407±0.0123 0.60±0.04 14 1.01±0.0722⁎ 0.447±0.0468 2.65±0.136 0.436±0.0390 0.47±0.06 Fluoxetine 7 1.05±0.0532⁎⁎ 0.519±0.0467 2.72±0.142 0.493±0.0166 0.50±0.05

Nucleus accumbens Control Vehicle – 0.716±0.0687 0.711±0.0677 1.76±0.272 4.54±0.667 0.93±0.09 Apigenin 7 0.748±0.0778 0.643±0.103 1.89±0.440 4.45±0.558 0.86±0.07 14 0.881±0.117 0.669±0.0523 2.10±0.197 4.68±0.390 0.76±0.06 Fluoxetine 7 0.756±0.0525 0.648±0.0919 1.73±0.311 4.34±0.152 0.85±0.08 CMS Vehicle – 0.450±0.0796# 0.507±0.0424# 1.71±0.228 7.83±0.640## 1.30±0.17# Apigenin 7 0.756±0.0595⁎ 0.522±0.0258 1.75±0.452 5.99±0.945 0.75±0.08⁎⁎ 14 0.820±0.116⁎⁎ 0.659±0.0495⁎ 1.64±0.150 4.94±0.876⁎ 0.76±0.05⁎⁎ Fluoxetine 7 0.824±0.0605⁎⁎ 0.695±0.0838 1.70±0.129 4.54±0.986⁎⁎ 0.82±0.11⁎ The 5-HT, 5-HIAA, NA and DA levels and 5-HIAA/5-HT ratio were measured in the control and CMS rats receiving water (vehicle), 7 mg/kg fluoxetine L.-T. Yi et al. / Life Sciences 82 (2008) 741–751 747

Pb0.05] and no change in 5-HIAA levels [F(1,12)=0.04, P= stress×treatment interaction on 5-HIAA levels and HIAA/5-HT 0.842] resulting in a significant increase of 5-HIAA/5-HT ratio. Separate one-way ANOVA revealed that CMS caused a ratio [F(1,12)=9.59, P b0.01]. Apigenin treatment signifi- significant decrease in 5-HT levels [F(1,12)=6.64, Pb0.05,] cantly elevated 5-HT levels [7 mg/kg: F(1,12)=6.94, P b0.05; but no changes in 5-HIAA levels and 5-HIAA/5-HT ratio. 14 mg/kg: F(1,12)=7.79, Pb0.05] and reduced 5-HIAA/5-HT Apigenin did not influence 5-HIAA levels and 5-HIAA/5-HT ratio [7 mg/kg: F(1,12)=6.89, Pb0.05; 14 mg/kg: F(1,12)= ratio. However, only the 14 mg/kg apigenin treatment sig- 10.11, Pb0.01], but it did not alter 5-HIAA levels in the CMS nificantly increased 5-HT levels [F(1,12)=7.15, Pb0.05] in the rats. In addition, there was no significant effect of stress, treatment CMS rats. In addition, there was no significant effect of stress, or stress×treatment interaction on DA levels. DA levels did not treatment and stress×treatment interaction on DA levels, and show any differences between the control and CMS rats treated apigenin did not alter the levels. with vehicle. Apigenin also exhibited no effect. In the CMS rats treated with fluoxetine hydrochloride, two- In the CMS rats treated with fluoxetine hydrochloride, two- way ANOVA indicated significant stress effect [F(1,24)=4.81, way ANOVA indicated significant stress effect [F(1,24)= Pb0.05], treatment effect [F(1,24)=16.8, Pb0.001] and no 4.96, Pb0.05], treatment effect [F(1,24)=9.49, Pb0.01] and stress×treatment interaction [F(1,24)=2.65, P=0.116] on 5- no stress×treatment interaction [F(1,24)=3.70, P=0.066] on HT levels. However, there was no significant effect of stress, 5-HT levels. However, there was no significant effect of stress, treatment and stress×treatment interaction on 5-HIAA levels treatment and stress×treatment interaction on 5-HIAA and 5-HIAA/5-HT ratio. Fluoxetine hydrochloride treatment levels and 5-HIAA/5-HT ratio, as well as DA levels. Fluoxe- significantly increased 5-HT levels [F(1,12)=13.10, Pb0.01], tine hydrochloride treatment significantly increased 5-HT lev- but did not alter 5-HIAA levels and 5-HIAA/5-HT ratio com- els [F(1,12)=12.88, Pb0.01], but did not alter 5-HIAA levels pared with vehicle-treated CMS animals. There was no signif- [F(1,12)=0.00, P=0.98], resulting in a significant decrease of icant effect of stress, treatment and stress×treatment interaction 5-HIAA/5-HT ratio [F(1,12)=8.25, Pb0.05]. However, it did on DA levels. Also, fluoxetine hydrochloride did not change not affect DA levels [F(1,12)=0.01, P=0.890]. DA levels.

Striatum Nucleus accumbens In the CMS rats treated with apigenin, two-way ANOVA In the CMS rats treated with apigenin, two-way ANOVA indicated no significant effect of stress, treatment and indicated significant treatment effect [F(2,36)=5.36, Pb0.01], stress×treatment interaction on 5-HT, 5-HIAA levels and 5- but no stress effect [F(1,36)=2.51, P=0.122] and stress×treat- HIAA/5-HT ratio, as well as DA levels. Separate one-way ment interaction [F(2,36)=1.48, P=0.239] on 5-HT levels; ANOVA revealed that CMS induced a significant decrease only there was no effect of stress, treatment and stress×treatment in- in 5-HIAA levels [F(1,12)=9.79, Pb0.01] and no changes in 5- teraction on 5-HIAA levels. Furthermore, only significant treat- HT, DA levels and 5-HIAA/5-HT ratio. Apigenin treatment at ment effect on 5-HIAA/5-HT ratio was observed [F(2,36)= two doses did not alter 5-HT, DA levels and 5-HIAA/5-HT ratio 6.16, Pb0.01]. Separate one-way ANOVA revealed that CMS compared with vehicle-treated CMS rats. However, only the induced significant decreases in 5-HT [F(1,12)=6.41, Pb0.05] 14 mg/kg dose significantly increased 5-HIAA levels [F(1,12)= and 5-HIAA [F(1,12)=6.56, Pb0.05] levels, resulting in a 7.87, Pb0.05]. significant increase in 5-HIAA/5-HT ratio [F(1,12)=6.74, In the CMS rats treated with fluoxetine hydrochloride, there Pb0.05]. Apigenin treatment significantly increased 5-HT lev- was no significant effect of stress, treatment and stress×treat- els [7 mg/kg: F(1,12)=9.46, Pb0.05; 14 mg/kg: F(1,12)= ment interaction on 5-HT, 5-HIAA, DA levels and 5-HIAA/5- 10.61, P b0.01] and decreased 5-HIAA/5-HT ratio [7 mg/kg: HT ratio. Fluoxetine hydrochloride treatment revealed a F(1,12)=13.16, P b0.01; 14 mg/kg: F(1,12)=14.24, P b0.01] significant increase only in 5-HIAA levels [F(1,12)=6.50, compared with vehicle-treated CMS rats; only the dose of Pb0.05]. 14 mg/kg noticeably elevated 5-HIAA levels [F(1,12)=5.44, P b0.05]. Two-way ANOVA indicated significant stress effect Hypothalamus [F(1,36)=8.73, P b0.01] without treatment effect [F(2,36)= In the CMS rats treated with apigenin, two-way ANOVA in- 1.99, P =0.150] and stress×treatment interaction [F(2,36)= dicated significant stress effect [F(1,36)=5.36, Pb0.05], treat- 2.31, P =0.113] on DA levels. CMS displayed a significant ment effect [F(2,36)=4.31, Pb0.05] and no stress×treatment increase in DA levels [F(1,12)=12.65, Pb0.01]. Only 14 mg/kg interaction [F(2,36)=0.96, P=0.390] on 5-HT levels. How- apigenin treatment noticeably decreased DA levels compared ever, there was no significant effect of stress, treatment and vehicle-treated CMS animals [F(1,12)=7.05, Pb0.05].

Notes to Table 1 The 5-HT, 5-HIAA, NA and DA levels and 5-HIAA/5-HT ratio were measured in the control and CMS rats receiving water (vehicle), 7 mg/kg fluoxetine hydrochloride, 7 and 14 mg/kg apigenin. Data are expressed as mean value±S.E.M (n=7). In prefrontal cortex, a two-way ANOVA showed effects of stress (Pb0.05) and of apigenin treatment (Pb0.01) on 5-HT levels, and of stress×apigenin treatment interaction on DA levels (Pb0.05). In hypothalamus, there were effects of stress (Pb0.05) and apigenin treatment (Pb0.05) on 5-HT levels. In nucleus accumbens, there was an effect of apigenin treatment (Pb0.01) on 5-HT levels, and of stress (Pb0.01) on DA levels. In hippocampus and striatum, there were no effects of stress, apigenin treatment and stress×apigenin treatment interaction on 5-HT, 5-HIAA, DA levels and 5-HIAA/5-HT ratio. For statistical significance, #Pb0.05 and ##Pb0.01 compared to control vehicle, ⁎Pb0.05 and ⁎⁎Pb0.01 compared to CMS- treated vehicle. 748 L.-T. Yi et al. / Life Sciences 82 (2008) 741–751

In the CMS rats treated with fluoxetine hydrochloride, two- way ANOVA indicated significant treatment effect [F(1,24)= 9.77, Pb0.01], stress×treatment interaction [F(1,24)=6.43, Pb0.05] and no stress effect [F(1,24)=2.22, P=0.149] on 5- HT levels. However, there was no significant effect of stress, treatment and stress×treatment interaction on 5-HIAA levels and 5-HIAA/5-HT ratio. Fluoxetine hydrochloride treatment significantly increased 5-HT levels [F(1,12)=14.01, Pb0.01] with a slight but insignificant decrease in 5-HIAA levels [F(1,12)=4.03, P=0.067], resulting a significant decrease in 5-HIAA/5-HT ratio [F(1,12)=9.19, Pb0.05] compared with vehicle-treated CMS rats. In addition, there were significant effects of stress [F(1,24)=6.59, P b0.05], treatment [F(1,24)=6.58, P b0.05] and stress×treatment interaction [F(1,24)=5.14, Pb0.05] on DA levels. Fluoxetine also significantly attenuated CMS-induced Fig. 5. The effects of apigenin and fluoxetine hydrochloride on platelet adenylyl DA increase [F(1,12)=17.82, Pb0.01] in this study. cyclase activity in the control and CMS rats. Platelet adenylyl cyclase activities No significant change of NA levels was observed both in were measured in the control and CMS rats receiving water (vehicle), 7 mg/kg vehicle-treated CMS rats and drug-treated CMS rats. There was fluoxetine hydrochloride, 7 and 14 mg/kg apigenin. Data are expressed as the mean±S.E.M. (n=8). A two-way ANOVA showed effects of stress (Pb0.01) no significant alteration of 5-HT, 5-HIAA, NA and DA levels and apigenin treatment (Pb0.05). For statistical significance, ##Pb0.01 and 5-HIAA/5-HT ratio after drug treatments in any of the brain compared to control vehicle and ⁎Pb0.05 compared to CMS-treated vehicle. regions in the control rats not exposed to CMS. treated CMS animals, chronic treatment with 14 mg/kg apigenin Effects of apigenin on serum corticosterone levels in the CMS significantly reversed CMS-induced elevation in serum corti- rats costerone concentrations [F(1,14)=13.67, Pb0.01]. However, apigenin at 7 mg/kg had no effect. The effects of apigenin and fluoxetine hydrochloride on In the CMS animals treated with fluoxetine hydrochloride, serum corticosterone levels in the control and CMS rats are two-way ANOVA indicated significant stress effect [F(1,28)= shown in Fig. 4. In the CMS rats treated with apigenin, two-way 7.56, Pb0.05], treatment effect [F(1,28)=11.80, Pb0.01] and ANOVA indicated significant stress effect [F(1,42)=15.76, stress×treatment interaction [F(1,28)=7.63, Pb0.05]. Flu- P b0.001], treatment effect [F(2,42)=4.20, P b0.05] and oxetine hydrochloride significantly reversed the elevated cor- stress×treatment interaction [F(2,42)=3.97, P b0.05] on ticosterone levels caused by CMS compared with vehicle- serum corticosterone levels. One-way ANOVA revealed that treated CMS animals [F(1,14)=12.72, Pb0.01]. CMS caused a significant increase in serum corticosterone lev- None of the treatments with apigenin and fluoxetine hy- els of rats [F(1,14)=15.35, Pb0.01]. Compared with vehicle- drochloride resulted in significant alterations of serum corticosterone levels in the control rats.

Effects of apigenin on platelet adenylyl cyclase activity in the CMS rats

The effects of apigenin and fluoxetine hydrochloride on platelet adenylyl cyclase activity in the control and CMS rats are shown in Fig. 5. In the CMS rats treated with apigenin, two-way ANOVA indicated significant stress effect [F(1,42)=10.51, Pb0.01], treatment effect [F(2,42)=4.98, Pb0.05] and no stress×treatment interaction [F(2,42)=1.10, P =0.340] on platelet adenylyl cyclase activity. One-way ANOVA revealed that CMS displayed a significant reduction in platelet adenylyl cyclase activity of rats [F(1,14)=22.25, Pb0.001]. Compared with vehicle-treated CMS animals, chronic treatment with 14 mg/kg apigenin significantly increased platelet adenylyl Fig. 4. Effects of apigenin and fluoxetine hydrochloride on serum corticosterone cyclase activity in the CMS rats [F(1,14)=6.38, Pb0.05]. levels in the control and CMS rats. Serum corticosterone levels were measured in However, insignificant elevation in platelet adenylyl cyclase the control and CMS rats receiving water (vehicle), 7 mg/kg fluoxetine activity was observed in the CMS rats receiving 7 mg/kg hydrochloride, 7 and 14 mg/kg apigenin. Data are expressed as the mean±S.E.M. apigenin [F(1,14)=0.13, P=0.716]. (n=8). A two-way ANOVA showed effects of stress (Pb0.001), apigenin treatment (Pb0.05) and stress×apigenin treatment interaction (Pb0.05). For In the CMS rats treated with fluoxetine hydrochloride, two- statistical significance, ###Pb0.001 compared to control vehicle and ⁎⁎Pb0.01 way ANOVA indicated significant stress effect [F(1,28)=37.71, and ⁎⁎⁎Pb0.001 compared to CMS-treated vehicle. Pb0.001], but no treatment effect [F(1,28)=2.13, P=0.155] L.-T. Yi et al. / Life Sciences 82 (2008) 741–751 749 and stress×treatment interaction [F(1,28)=1.06, P=0.311] served in frontal cortex and nucleus accumbens in this study. on platelet adenylyl cyclase activity. Fluoxetine hydrochloride Due to the limitation of the method used to determine mono- failed to alter the activity in the CMS rats [F(1,14)=0.110, amine neurotransmitters and their metabolites, it was difficult to P=0.744]. get a stable status of 4-dihydroxyphenylacetic acid in animals None of the treatments with apigenin and fluoxetine hy- in this study. These data indicated that CMS might alter drochloride resulted in significant changes in platelet adenylyl serotonergic and dopaminergic functions in some of the rat cyclase activity in the control groups. brain regions. Interestingly, apigenin administrated i.p. restored DA turnover changes in amygdala and hypothalamus induced Discussion by FST (Nakazawa et al., 2003). Apigenin was reported to be a MAO inhibitor in vitro (Lorenzo et al., 1996; Han et al., 2007). The search for new antidepressants with novel strategies may MAO inhibitors are known to increase the levels of brain mono- help to develop faster and more effective antidepressant agents. amines (Kanazawa, 1994). Therefore, we hypothesized that oral Recently, researchers paid more attention to natural products for apigenin treatment might regulate the brain monoaminergic seeking safe and effective antidepressant agents (Ito et al., 2006; neurotransmitter systems in the CMS rats. In the present study, Pan et al., 2005, 2007; Xu et al., 2006, 2007). Apigenin has a apigenin enhanced 5-HT levels and decreased 5-HT turnover variety of pharmacological actions on the central nervous in some of the rat brain regions as well as attenuated DA al- system. In the present study, we first confirmed antidepressant- terations in frontal cortex and nucleus accumbens in the CMS like effects of oral apigenin treatment in the mouse FST and in rats. A dose–response tendency of apigenin's action on 5-HT the CMS rats. We also explored potential mechanisms involved levels in frontal cortex and hypothalamus and DA levels in in its antidepressant actions. Here we showed that apigenin frontal cortex was observed in the CMS rats. In addition, the resulted in the normalization of changes in central mono- brain regional changes of monoamine levels were partly re- aminergic neurotransmitter, the HPA axis, and adenylyl cyclase versed by chronic treatment with fluoxetine hydrochloride, activity systems in the CMS rats. which was consistent with a previous study (Li et al., 2003). A previous study showed the antidepressant-like effects of These results suggested that the antidepressant-like effects of apigenin at the acute doses of 12.5 and 25 mg/kg i.p. in the FST apigenin might be mediated by actions on serotonergic and in ddY mice (Nakazawa et al., 2003). Contrary to the conclusion dopaminergic systems in rats exposed to CMS. by Nakazawa et al. (2003), in our pilot study, acute oral Basic and clinical evidence suggested that there was an treatment with apigenin did not show any significant effect in HPA axis hyperactivity in depression, which was character- the FST in male ICR mice. Therefore, subchronic and chronic ized by elevated serum corticosterone concentrations (Muscat treatments with apigenin were used in the present study. In fact, and Willner, 1992; Ayensu et al., 1995; Barden et al., 1995; under an oral two-week treatment condition, apigenin, at the Risbrough and Stein, 2006; Aihara et al., 2007). Successful daily doses of 10 and 20 mg/kg, exerted marked antidepressant- treatment with antidepressants normalized the HPA axis like effects in the mouse FST, suggesting that chronic treatment hyperactivity in depressed patients (Heuser et al., 1996; Nickel of antidepressants might be necessary for full efficacy in the et al., 2003; Young et al., 2004). In addition, there was a close mouse FST. relationship between monoamine neurotransmission and the As reported with conventional antidepressants (Di Chiara HPA axis systems in major depression (Price and Lucki, 2001; et al., 1999; Li et al., 2003; Bekris et al., 2005; Grippo et al., Valentino and Commons, 2005; Keeney et al., 2006). A recent 2006), chronic treatment with 7 and 14 mg/kg apigenin caused a report showed that concurrent manipulation of the HPA axis reversal of sucrose intake in rats exposed to CMS without sig- system might improve the efficacy of fluoxetine in the treatment nificantly affecting water intake and body weight in the present of patients with major depression (Young et al., 2004). The study. These results further confirmed the antidepressant-like present study exhibited that chronic apigenin administration actions of oral apigenin administration in animal models of alleviated serum corticosterone elevation induced by CMS with depression. However, there was no clear dose–response trend a significant dose–response tendency. These findings suggested observed in the CMS rats treated with the two doses of apigenin. that the antidepressant actions of oral apigenin treatment might Indeed, a U-shaped dose–response pattern induced by anti- be partly related to its concurrent modulation of the HPA axis depressants has been always observed in some animal ex- activity and the brain monoamine neurotransmitter levels. periments (MacSweeney et al., 1998; Muñoz and Papp, 1999; It is well known that adenylyl cyclase activity is regulated by Kulkarni and Dhir, 2007). Therefore, the dose-dependent rela- serotonergic receptors, which are coupled to adenylyl cyclase tionship of apigenin in behavioral studies needs to be further affecting cAMP production. The hippocampus comprises a key investigated based on our findings. structure as a neurobiological substrate of stress and depression, As reported by others (Torres et al., 2002; Li et al., 2003; due to its involvement in the regulation of the HPA axis (Bekris Bekris et al., 2005), CMS induced significant decreases in 5-HT et al., 2005). Antidepressants might induce up-regulation in levels in prefrontal cortex, hippocampus, hypothalamus and cAMP signaling and, consequently, increase adenylyl cyclase nucleus accumbens, in 5-HIAA levels in prefrontal cortex, stri- activity in hippocampus (Duman, 1998; Donati and Rasenick, atum and nucleus accumbens, resulting in significant increases 2003; Hines and Tabakoff, 2005; Abdel-Razaq et al., 2007). in 5-HIAA/5-HT ratio in hippocampus and nucleus accumbens. In the present study, the dosages of apigenin were validated A simultaneous significant increase of DA levels was also ob- by several approaches, including measurements of behaviors, 750 L.-T. Yi et al. / Life Sciences 82 (2008) 741–751 brain 5-HT and serum corticosterone levels in the CMS rats. prefrontal cortex and limbic/paralimbic regions. Psychiatry Research 155 (3), – Therefore we hypothesized that oral apigenin treatment might 245 256. Avallone, R., Zanoli, P., Puia, G., Kleinschnitz, M., Schreier, P., Baraldi, M., change platelet adenylyl cyclase activity. In the present study, 2000. Pharmacological profile of apigenin, a flavonoid isolated from Ma- CMS caused a significant decrease in platelet adenylyl cyclase tricaria chamomilla. Biochemical Pharmacology 59 (11), 1387–1394. activity reflected as a reduction of cAMP concentrations, in- Ayensu, W.K., Pucilowski, O., Mason, G.A., Overstreet, D.H., Rezvani, A.H., dicating that platelet adenylyl cyclase activity could serve as a Janowsky, D.S., 1995. Effects of chronic mild stress on serum complement biological marker for the CMS model in animals. Further- activity, saccharin preference, and corticosterone levels in Flinders lines of rats. Physiology & Behavior 57 (1), 165–169. more, apigenin reversed the reduction in cAMP contents to Barden, N., Reul, J.M., Holsboer, F., 1995. Do antidepressants stabilize mood control levels with a clear dose–response tendency in the CMS through actions on the hypothalamic–pituitary–adrenocortical system? rats. However, fluoxetine hydrochloride failed to affect platelet Trends in Neurosciences 18 (1), 6–11. adenylyl cyclase activity in this study. Indeed, a previous study Bekris, S., Antoniou, K., Daskas, S., Papadopoulou-Daifoti, Z., 2005. Behavioural had reported that chronic treatment with fluoxetine did not and neurochemical effects induced by chronic mild stress applied to two different rat strains. Behavioural Brain Research 161 (1), 45–59. produce any change in the 5-HT1A-induced inhibition of the Chen, D., Daniel, K.G., Chen, M.S., Kuhn, D.J., Landis-Piwowar, K.R., Dou, adenylyl cyclase activity (Varrault et al., 1991). These results Q.P., 2005. Dietary flavonoids as proteasome inhibitors and apoptosis indicated that the antidepressant-like effects of apigenin might inducers in human leukemia cells. Biochemical Pharmacology 69 (10), be related to an up-regulation in cAMP signaling associated 1421–1432. with elevation in platelet adenylyl cyclase activity. More im- Cowburn, R.F., Marcusson, J.O., Eriksson, A., Wiehager, B., O'Neill, C., 1994. Adenylyl cyclase activity and G-protein subunit levels in postmortem frontal portantly, the mechanisms of the antidepressant-like actions by cortex of suicide victims. Brain Research 633 (1–2), 297–304. apigenin might be partly different with that by fluoxetine Coyle, J.T., Duman, R.S., 2003. Finding the intracellular signaling pathways hydrochloride. This issue needs to be further studied. affected by mood disorder treatments. Neuron 38 (2), 157–160. In summary, this investigation confirmed antidepressant-like Curzon, G., Green, A.R., 1970. Rapid method for the determination of 5- properties of oral apigenin administration in the mouse FST and hydroxytryptamine and 5-hydroxyindoleacetic acid in small regions of rat brain. British Journal of Pharmacology 39 (3), 653–655. in the CMS rats. Apigenin attenuated CMS-induced serotoner- Czyz, J., Madeja, Z., Irmer, U., Korohoda, W., Hulser, D.F., 2005. Flavonoid gic and dopaminergic alterations in some of the rat brain apigenin inhibits motility and invasiveness of carcinoma cells in vitro. regions. Moreover, apigenin reversed CMS-induced elevation International Journal of Cancer 114 (1), 12–18. in serum corticosterone concentrations as well as reduction in Di Chiara, G., Loddo, P., Tanda, G., 1999. Reciprocal changes in prefrontal and platelet adenylyl cyclase activity. These results suggested that limbic dopamine responsiveness to aversive and rewarding stimuli after chronic mild stress: implications for the psychobiology of depression. Biological the antidepressant-like actions of oral apigenin treatment could Psychiatry 46 (12), 1624–1633. be related to a combination of multiple biochemical effects, and Donati, R.J., Rasenick, M.M., 2003. G protein signaling and the molecular basis might help to elucidate its mechanisms, that are involved in of antidepressant action. Life Sciences 73 (1), 1–17. normalization of central monoamine neurotransmitter systems, Duman, R.S., 1998. Novel therapeutic approaches beyond the serotonin – HPA axis alterations, and down-regulation of the cAMP path- receptor. Biological Psychiatry 44 (5), 324 335. Duman, R.S., Heninger, G.R., Nestler, E.J., 1997. A molecular and cellular way in animal models of depression. This study may provide theory of depression. Archives of General Psychiatry 54 (7), 597–606. new information and a novel target for the development of Fang, J., Xia, C., Cao, Z., Zheng, J.Z., Reed, E., Jiang, B.H., 2005. Apigenin therapeutic agents from natural products. In addition, no ap- inhibits VEGF and HIF-1 expression via PI3K/AKT/p70S6K1 and HDM2/ parent toxicity was observed for apigenin at the doses used in p53 pathways. FASEB Journal 19 (3), 342–353. the tested rats. This feature of apigenin makes it an attractive Gainotti, G., Antonucci, G., Marra, C., Paolucci, S., 2001. Relation between depression after stroke, antidepressant therapy, and functional recovery. candidate for the treatment of depression, particularly if it is to Journal of Neurology Neurosurgery and Psychiatry 71 (2), 258–261. be taken on a long term basis. Obviously, in view of the in vivo Garg, A., Garg, S., Zaneveld, L.J., Singla, A.K., 2001. Chemistry and pharma- antidepressant actions of apigenin, further studies on its bio- cology of the citrus bioflavonoid hesperidin. Phytotherapy Research 15 (8), logical properties compared with currently available oral anti- 655–669. depressant agents in humans are warranted. Grippo, A.J., Beltz, T.G., Weiss, R.M., Johnson, A.K., 2006. The effects of chronic fluoxetine treatment on chronic mild stress-induced cardiovascular changes and anhedonia. Biological Psychiatry 59 (4), 309–316. Acknowledgements Han, X.H., Hong, S.S., Hwang, J.S., Lee, M.K., Hwang, B.Y., Ro, J.S., 2007. Monoamine oxidase inhibitory components from Cayratia japonica. Archives The work was co-financed by grants from NSFC (No. of Pharmacal Research 30 (1), 13–17. 90409009), JSNSF (BK2007143), NCET-06-0442 and 07-C- Heuser, I.J., Schweiger, U., Gotthardt, U., Schmider, J., Lammers, C.H., Dettling, M., Yassouridis, A., Holsboer, F., 1996. Pituitary–adrenal-system 016 to Ling-Dong Kong. regulation and psychopathology during amitriptyline treatment in elderly depressed patients and normal comparison subjects. The American Journal References of Psychiatry 153 (1), 93–99. Hines, L.M., Tabakoff, B., 2005. Platelet adenylyl cyclase activity: a biologi- Abdel-Razaq, W., Bates, T.E., Kendall, D.A., 2007. The effects of antidepres- cal marker for major depression and recent drug use. Biological Psychiatry sants on cyclic AMP-response element-driven gene transcription in a model 58 (12), 955–962. cell system. Biochemical Pharmacology 73 (12), 1995–2003. Hougee, S., Sanders, A., Faber, J., Graus, Y.M., van den Berg, W.B., Garssen, J., Aihara, M., Ida, I., Yuuki, N., Oshima, A., Kumano, H., Takahashi, K., Fukuda, Smit, H.F., Hoijer, M.A., 2005. Decreased pro-inflammatory cytokine pro- M., Oriuchi, N., Endo, K., Matsuda, H., Mikuni, M., 2007. HPA axis duction by LPS-stimulated PBMC upon in vitro incubation with the dysfunction in unmedicated major depressive disorder and its normaliza- flavonoids apigenin, luteolin or chrysin, due to selective elimination of tion by pharmacotherapy correlates with alteration of neural activity in monocytes/macrophages. Biochemical Pharmacology 69 (2), 241–248. L.-T. Yi et al. / Life Sciences 82 (2008) 741–751 751

Ito, N., Nagai, T., Yabe, T., Nunome, S., Hanawa, T., Yamada, H., 2006. Clinical and neurobiological effects of tianeptine and paroxetine in major Antidepressant-like activity of a Kampo (Japanese herbal) medicine, Koso- depression. Journal of Clinical Psychopharmacology 23 (2), 155–168. san (Xiang-Su-San), and its mode of action via the hypothalamic–pituitary– Pan, Y., Kong, L., Xia, X., Zhang, W., Xia, Z., Jiang, F., 2005. Antidepressant- adrenal axis. Phytomedicine 13 (9–10), 658–667. like effect of icariin and its possible mechanism in mice. Pharmacology Kanazawa, I., 1994. Short review on monoamine oxidase and its inhibitors. Biochemistry and Behavior 82 (4), 686–694. European Neurology 34 (Suppl 3), 36–39. Pan, Y., Kong, L.D., Li, Y.C., Xia, X., Kung, H.F., Jiang, F.X., 2007. Icariin Keeney, A., Jessop, D.S., Harbuz, M.S., Marsden, C.A., Hogg, S., Blackburn- from Epimedium brevicornum attenuates chronic mild stress-induced Munro, R.E., 2006. Differential effects of acute and chronic social defeat behavioral and neuroendocrinological alterations in male Wistar rats. stress on hypothalamic–pituitary–adrenal axis function and hippocampal Pharmacology Biochemistry and Behavior 87 (1), 130–140. serotonin release in mice. Journal of Neuroendocrinology 18, 330–338. Porsolt, R.D., Le Pichon, M., Jalfre, M., 1977. Depression: a new animal model Komiya, M., Takeuchi, T., Harada, E., 2006. Lemon oil vapor causes an anti- sensitive to antidepressant treatments. Nature 266 (5604), 730–732. stress effect via modulating the 5-HT and DA activities in mice. Behavioural Porsolt, R.D., Anton, G., Blavet, N., Jalfre, M., 1978. Behavioural despair in Brain Research 172 (2), 240–249. rats: a new model sensitive to antidepressant treatments. European Journal of Komori, T., Fujiwara, R., Tanida, M., Nomura, J., 1995a. Potential antide- Pharmacology 47 (4), 379–391. pressant effects of lemon odor in rats. European Neuropsychopharmacology Price, M.L., Lucki, I., 2001. Regulation of serotonin release in the lateral sep- 5 (4), 477–480. tum and striatum by corticotropin-releasing factor. Journal of Neuroscience Komori, T., Fujiwara, R., Tanida, M., Nomura, J., 1995b. Application of 21 (8), 2833–2841. fragrances to treatments for depression. Nihon Shinkei Seishin Yakurigaku Reiach, J.S., Li, P.P., Warsh, J.J., Kish, S.J., Young, L.T., 1999. Reduced Zasshi 15 (1), 39–42. adenylyl cyclase immunolabeling and activity in postmortem temporal Komori, T., Fujiwara, R., Tanida, M., Nomura, J., Yokovama, M.M., 1995c. cortex of depressed suicide victims. Journal of Affective Disorders 56 (2–3), Effects of citrus fragrance on immune function and depressive states. 141–151. Neuroimmunomodulation 2 (3), 174–180. Risbrough, V.B., Stein, M.B., 2006. Role of corticotropin releasing factor in Kulkarni, S.K., Dhir, A., 2007. Effect of various classes of antidepressants in anxiety disorders: a translational research perspective. Hormones and Behavior behavioral paradigms of despair. Progress in Neuro-Psychopharmacology & 50 (4), 550–561. Biological Psychiatry 31 (6), 1248–1254. Salah, N., Miller, N.J., Paganga, G., Tijburg, L., Bolwell, G.P., Riceevans, C., Li, J.M., Kong, L.D., Wang, Y.M., Cheng, C.H., Zhang, W.Y., Tan, W.Z., 2003. 1995. Polyphenolic flavanols as scavengers of aqueous phase radicals and Behavioral and biochemical studies on chronic mild stress models in rats as chain-breaking antioxidants. Archives of Biochemistry and Biophysics treated with a Chinese traditional prescription Banxia-houpu decoction. Life 322 (2), 339–346. Sciences 74 (1), 55–73. Skolnick, P., 1999. Antidepressants for the new millennium. European Journal Lorenzo, P.S., Rubio, M.C., Medina, J.H., Adler-Graschinsky, E., 1996. In- of Pharmacology 375 (1–3), 31–40. volvement of monoamine oxidase and noradrenaline uptake in the posi- Struckmann, J.R., Nicolaides, A.N., 1994. Flavonoids. A review of the phar- tive chronotropic effects of apigenin in rat atria. European Journal of macology and therapeutic efficacy of Daflon 500 mg in patients with chronic Pharmacology 312 (2), 203–207. venous insufficiency and related disorders. Angiology 45 (6), 419–428. Lowry, O.H., Rosebrough, N.J., Farr, A.L., Randall, R.J., 1951. Protein Torres, I.L., Gamaro, G.D., Vasconcellos, A.P., Silveira, R., Dalmaz, C., 2002. measurement with the Folin phenol reagent. Journal of Biological Chemistry Effects of chronic restraint stress on feeding behavior and on monoamine 193 (1), 265–275. levels in different brain structures in rats. Neurochemical Research 27 (6), MacSweeney, C.P., Lesourd, M., Gandon, J.M., 1998. Antidepressant-like 519–525. effects of alnespirone (S 20499) in the learned helplessness test in rats. Valentino, R.J., Commons, K.G., 2005. Peptides that fine-tune the serotonin European Journal of Pharmacology 345 (2), 133–137. system. Neuropeptides 39 (1), 1–8.

Manthey, J.A., Guthrie, N., 2002. Antiproliferative activities of citrus flavo- Varrault, A., Leviel, V., Bockaert, J., 1991. 5-HT1A-sensitive adenylyl cyclase of noids against six human cancer cell lines. Journal of Agricultural and Food rodent hippocampal neurons: effects of antidepressant treatments and Chemistry 50 (21), 5837–5843. chronic stimulation with agonists. Journal of Pharmacology and Experi- Menninger, J.A., Tabakoff, B., 1997. Forskolin-stimulated platelet adenylyl cy- mental Therapeutics 257 (1), 433–438. clase activity is lower in persons with major depression. Biological Psychiatry Welch, A.S., Welch, B.L., 1969. Solvent extraction method for simultaneous 42 (1), 30–38. determination of norepinephrine, dopamine, serotonin, and 5-hydroxyindo- Morel, I., Lescoat, G., Cogrel, P., Sergent, O., Pasdeloup, N., Brissot, P., Cillard, leacetic acid in a single mouse brain. Analytical Biochemistry 30 (2), 161–179. P., Cillard, J., 1993. Antioxidant and iron-chelating activities of the fla- Willner, P., Towell, A., Sampson, D., Sophokleous, S., Muscat, R., 1987. vonoids catechin, quercetin and diosmetin on iron-loaded rat hepatocyte Reduction of sucrose preference by chronic unpredictable mild stress, and its cultures. Biochemical Pharmacology 45 (1), 13–19. restoration by a tricyclic antidepressant. Psychopharmacology (Berl) 93 (3), Muñoz, C., Papp, M., 1999. Alnespirone (S 20499), an agonist of 5-HT1A 358–364. receptors, and imipramine have similar activity in a chronic mild stress model Wong, M.L., Licinio, J., 2001. Research and treatment approaches to depres- of depression. Pharmacology Biochemistry and Behavior 63 (4), 647–653. sion. Nature Reviews Neuroscience 2 (5), 343–351. Muscat, R., Willner, P., 1992. Suppression of sucrose drinking by chronic mild Wouters, J., 1998. Structural aspects of monoamine oxidase and its reversible unpredictable stress: a methodological analysis. Neuroscience and Biobe- inhibition. Current Medicinal Chemistry 5 (2), 137–162. havioral Reviews 16 (4), 507–517. Xu, Y., Ku, B.S., Tie, L., Yao, H.Y., Jiang, W.G., Ma, X., Li, X.J., 2006. Nakazawa, T., Yasuda, T., Ueda, J., Ohsawa, K., 2003. Antidepressant-like Curcumin reverses the effects of chronic stress on behavior, the HPA axis, effects of apigenin and 2,4,5-trimethoxycinnamic acid from Perilla BDNF expression and phosphorylation of CREB. Brain Research 1122 (1), frutescens in the forced swimming test. Biological & Pharmaceutical 56–64. Bulletin 26 (4), 474–480. Xu, Y., Ku, B.S., Cui, L., Li, X.J., Barish, P.A., Foster, T.C., Ogle, W.O., 2007. Nestler, E.J., 1998. Antidepressant treatments in the 21st century. Biological Curcumin reverses impaired hippocampal neurogenesis and increases sero- Psychiatry 44 (7), 526–533. tonin receptor 1A mRNA and brain-derived neurotrophic factor expression Nestler, E.J., Barrot, M., DiLeone, R.J., Eisch, A.J., Gold, S.J., Monteggia, L.M., in chronically stressed rats. Brain Research 1162, 9–18. 2002a. Neurobiology of depression. Neuron 34 (1), 13–25. Young, E.A., Altemus, M., Lopez, J.F., Kocsis, J.H., Schatzberg, A.F., Nestler, E.J., Gould, E., Manji, H., Buncan, M., Duman, R.S., Greshenfeld, H., DeBattista, C., Zubieta, J.K., 2004. HPA axis activation in major depression 2002b. Preclinical models: status of basic research in depression. Biological and response to fluoxetine: a pilot study. Psychoneuroendocrinology 29 (9), Psychiatry 52 (6), 503–528. 1198–1204. Nickel, T., Sonntag, A., Schill, J., Zobel, A.W., Ackl, N., Brunnauer, A., Murck, Zanoli, P., Avallone, R., Baraldi, M., 2000. Behavioral characterization of the H., Ising, M., Yassouridis, A., Steiger, A., Zihl, J., Holsboer, F., 2003. flavonoids apigenin and chrysin. Fitoterapia 71 (Suppl. 1), S117–S123.