Psychopharmacology DOI 10.1007/s00213-007-1037-z ORIGINAL INVESTIGATION Anticonvulsant and anxiolytic-like effects of compounds isolated from Polygala sabulosa (Polygalaceae) in rodents: in vitro and in vivo interactions with benzodiazepine binding sites Filipe Silveira Duarte & Mariel Marder & Alexandre Ademar Hoeller & Marcelo Duzzioni & Beatriz Garcia Mendes & Moacir Geraldo Pizzolatti & Thereza Christina Monteiro De Lima Received: 5 April 2007 / Accepted: 28 November 2007 # Springer-Verlag 2007 Abstract Rationale Polygala sabulosa, a folk medicine, presents dihydrostyryl-2-pyrones (DST) and styryl-2-pyrones (STY), compounds structurally similar to kavalactones. Our previous study showed that the ethyl acetate fraction (EA) and these constituents present anxiolytic-like, hypnosedative, and anticonvulsant effects in mice. F. S. Duarte : M. Duzzioni : T. C. M. De Lima Laboratory of Neuropharmacology - Department of Pharmacology, Federal University of Santa Catarina, Florianópolis, SC, Brazil M. Marder Instituto de Química y Fisicoquímica Biológicas - Faculdad de Farmacia y Bioquímica, University of Buenos Aires, Buenos Aires, Argentina A. A. Hoeller Laboratory of Physiology—Department of Physiological Sciences, Federal University of Santa Catarina, Florianópolis, SC, Brazil B. G. Mendes : M. G. Pizzolatti Laboratory of Organic Chemistry - Department of Chemistry, Federal University of Santa Catarina, Florianópolis, SC, Brazil T. C. M. De Lima (*) Department of Pharmacology, Center of Biological Sciences, Federal University of Santa Catarina, Campus Universitário, Trindade, Florianópolis, SC 88049–900, Brazil e-mail: thereza@farmaco.ufsc.br Objectives This study investigated the role of benzodiazepine binding site (BDZ-bs) in the central effects of either EA or three DST (1, 2, and 3) and three STY (4, 5, and 7), using in vivo and in vitro assays. Methods and results In the elevated plus-maze (EPM), flumazenil (FMZ), a BDZ antagonist, partially blocked the anxiolytic-like effect of DST-3 or STY-4 and STY-7, but not DST-1. Using electroencephalogram (EEG), EA protected against pentylenetetrazole (PTZ)-induced convulsion in rats, an effect partially blocked by FMZ, suggesting the participation of the BDZ-bs in this action. EA also protected against the maximal electroshock (MES)-induced convulsions in mice, a profile distinct from diazepam (DZP). DST and STY compounds inhibited the [3H]flunitrazepam ([3H]-FNZ) binding to BDZ-bs in rat cortical synaptosomes with Ki higher than 100 μM (DST-1), 41.7 μM (DST-2), 35.8 μM (DST-3), 90.3 μM (STY-4), 31.0 μM (STY-5) and 70.0 μM (STY-7). In the saturation assay, DST-3 and STY-7 competitively inhibited the binding of [3H]-FNZ to BDZ-bs with a significant decrease in apparent affinity (Kd) and no change in maximal binding (Bmax). Conclusions The present data support a partial BDZ-bs mediation of the anxiolytic-like and anticonvulsant effects of EA of P. sabulosa and its main isolated constituents, DST and STY. Keywords Phytomedicines . Polygala sabulosa . Benzodiazepine site . Anxiolytic . Anticonvulsant Psychopharmacology Introduction γ-Aminobutyric acid (GABA) plays a key role in the overall balance between neuronal excitation and inhibition, regulating convulsions, anxiety and sleep (Nutt and Malizia 2001; Möhler 2006). Some drugs acting at GABAA receptors, such as benzodiazepines (BDZ) and barbiturates, act as positive allosteric modulators thereby increasing the chloride conductance and promoting anxiolytic, hypnosedative and anticonvulsant effects. BDZ can present various side effects including drowsiness, mental dullness, amnesia, muscle relaxation, physiological dependence and withdrawal syndrome (Ashton 1994). Partial agonists of BDZ receptors have been investigated to diminish the side effects which might eventually overcome the limitations of full agonists such as diazepam. On this regard, great efforts have been made in search of new anxiolytic drugs, including studies of active constituents obtained from medicinal plants (Hui et al. 2002; Leung et al. 2003; Dhawan et al. 2004; Zhang 2004). The plant Polygala sabulosa A.W. Bennett (Polygalaceae) and its ethyl acetate fraction (EA) show anticonvulsant, hypno-sedative and anxiolytic-like effects while its main constituents dihydrostyryl-2-pyrones (DST) and styryl-2-pyrones (STY) exhibit an anxiolytic-like action in mice (Duarte et al. 2007), suggesting a BDZ-like action. However, so far, the underlying mechanisms of these central effects have not been studied. Phytochemical studies of P. sabulosa revealed that their DST and STY isolated compounds present a structural skeleton similar to kavalactones found in kava-kava, a plant widely used as anxiolytic but with alleged toxicity (Escher et al. 2001; Russmann et al. 2001; Blumenthal 2002; Brauer et al. 2003; Singh 2005). Thus, the present work investigated, to the best of our knowledge for the first time, the underlying mechanisms involved in the central actions of EA and DST or STY compounds isolated from P. sabulosa, studying the interaction with the BDZ binding site (BDZ-bs) of the GABAA receptor with in vivo and in vitro approaches. Materials and methods Animals Adult male Wistar rats (250–300 g) were used for biochemical assays and electroencephalogram (EEG) recordings. Adult male Swiss mice (30–35 g) were used for behavioral evaluation. The animals were maintained on a 12-h light–dark cycle (lights on at 7:00 A.M.) at constant room temperature (23±2 °C). The rats and mice were housed in groups of 5 or 20 per cage, respectively, and had free access to food and water, except during the experi- ments. All the animals were allowed to adapt to the laboratory conditions for at least 1 week before the beginning of the experiments. Each behavioral test was conducted during the light phase of the cycle (between 1:00 and 6:00 P.M.), and the animals were used just once. All experiments were conducted in accordance with international standards of animal welfare recommended by the Brazilian Society of Neuroscience and Behavior (Act 1992) and approved by the local Committee for Animal Use and Care in Research (no. 23080.027554/2004–49 and 23080.08007244/2006–70/CEUA/UFSC). Stereotaxic surgery and intracerebroventricular injections Rats were anesthetized with ketamine (100 mg/kg) and xylazine (20 mg/kg; Vetbrands, SP, Brazil) and implanted stereotaxically (Stoelting®, USA) with bipolar twisted NiCr wire electrodes (tip diameter 150 μm) in the left CA3 region (AP=−3.3; ML=+2.5; DV=−3.8) for deep recordings. Surface recordings were obtained from electrodes positioned in the right A3 region (AP=−1.5; ML=−3.0). An additional screw placed in the frontal sinus served as a reference electrode. The electrodes were fixed to the skull with dental acrylic. The correct location of implanted electrodes was verified in sections using a cryostat microtome (Leica® CM1850, Germany), stained with Cresyl violet and examined under light microscope (Nikon®, Japan). Intracerebroventricular (i.c.v.) microinjections were performed with a hand-driven Hamilton syringe according to the procedure originally described by Haley and McCormick (1957) and modified by Maurice et al. (1996). In brief, the mice were slightly anesthetized with ethyl ether until the loss of their postural reflex. Thereafter, the animals were gently grasped by the loose skin above the head, pulled tightly back, and a 27 gauge needle attached to a 5 μl Hamilton syringe was inserted perpendicularly through the skull in to the brain. The site of injection was 1 mm lateral to the midline and 1 mm posterior to a line draw through the anterior base of the eyes (used as external reference); the depth of insertion of the needle was 3 mm. Upon termination of the experiment, each mouse was decapitated and the accurate placement of the injection site (needle track) was verified a fresco (without any fixation procedure). The mice with cannula misplacement or any sign of cerebral hemorrhage were discarded from the statistical analysis (less than 5% overall). Procedures Electroencephalographic recording Electroencephalogram (EEG) was recorded through a polygraph digital system (BIOPAC System, MP-100/WSW). Psychopharmacology Signals were amplified 20,000× and filtered (0.5–30 Hz), at a sampling rate of 256 Hz and recorded using the acquisition software ACQ-Knowledge version 3.2. EEGs were registered in a glass tank (0.3×0.5×0.4 m), the floor covered with sawdust, placed inside a Faraday cage (1.0×0.6×0.7 m). Five days after surgery, the rats were individually placed in the glass tank and the acquisition cable was connected to the micro-connector on the animal head, with signals of electrical activity captured by a BIOPAC® system. EEG recordings were analyzed by direct visual inspection. The animals’ behavior was recorded simultaneously through a webcam located inside the Faraday cage. Pentylenetetrazole-induced convulsions After a 30-min period of habituation, a 20-min baseline recording was obtained followed by treatments (Treatments section). The rats received an i.p. injection of pentylenetetrazole (PTZ) and the behavioral and electrographic changes were simultaneously recorded for up 1 h. Latency to and duration of the first myoclonic jerk/clonus (MJ/C) or generalized clonic seizures (GCS) as well as the lethality ratio were recorded. Convulsion severity was registered using a modified Racine scale (Racine 1972) as follows: stage 0, no change in behavior; stage 1, ear and facial twitching; stage 2, isolated myoclonic jerks; stage 3, clonus of the forelimbs, neck and/or head; stage 4, clonus of the forelimbs, neck and/or head with rearing and falling; stage 5, GCS (without the tonic phase) beginning with running, followed by loss of righting reflex. Each animal was assigned the score of the most severe seizure presented. Maximal electroshock-induced convulsions Electro-convulsions in mice were evoked using alternating current produced by a shocker generator (PLAT-2), and delivered via ear-stainless steel electrodes. The suprathreshold stimulus of maximal electroshock (MES) was adjusted with rectangular pulses of 50 mA intensity delivered at 60 Hz for 200 ms of stimulus duration. This stimulus is necessary to produce tonic hind-limb extension in 95% of the animals tested in our laboratory condition. Full tonic extension of both hind-limbs was taken as the endpoint. The tonic extension time (s) and the percentage of animals that presented the tonic extension were used as an index of convulsion severity (Swinyard et al. 1952). Elevated plus-maze The anxiolytic-like activity of compounds isolated from P. sabulosa was assessed using the elevated plus-maze (EPM), as proposed by Lister (1987) and earlier described (Duarte et al. 2007). After treatments, each mouse was placed on the central platform, facing an enclosed arm, and observed for a 5-min period. Frequencies of entries into either open or enclosed arms, as well as the time spent in each arm type were recorded (in seconds). The number of entries into the enclosed arms was used as an index of general activity (Rodgers and Dalvi 1997). The incidence of ethological parameters such as unprotected head-dipping (uHD), protected stretch-attend postures (pSAP), and open-arms end activity (OAEA) was also recorded (Cole and Rodgers 1994; Duarte et al. 2007). Biochemical assay ( 3H-flunitrazepam binding assay) A radioligand binding assay was used to evaluate the putative action of DST (1, 2, or 3) and STY (4, 5, or 7) compounds on the BDZ-bs of the GABAA receptor complex. The binding of [3H]-FNZ (81.8 Ci/mmol; obtained from PerkinElmer Life and Analytical Sciences, Boston, MA, USA) to the BDZ-bs in washed crude synaptosomal membranes from rat cerebral cortex was determined as previously described (Medina et al., 1990; Viola et al. 2000; Marder et al. 2003). In the competition assays, the incubations were done with [3H]-FNZ 0.42 nM in the presence of 0.1 to 600 μM concentrations of DST (1, 2 or 3) and STY (4, 5 or 7). Diazepam (DZP) was used as positive control in concentrations between 1 to 100 nM. In saturation assays, increasing concentrations of [3H]-FNZ (0.3–9 nM) were incubated in the presence of vehicle, DST3 40 μM or STY-7 70 μM. Non-specific binding was measured in the presence of FNZ 10 μM and represented 5–15% of the total binding. After incubation, the assays were terminated by filtration under vacuum through Whatman GF/B glass-fiber filters followed by washing three times with 3 ml each of incubation medium. Individual filters were incubated overnight with scintillation cocktail (OptiPhase ‘HiSafe’ 3) before measuring radioactivity in a Wallac Rackbeta 1214 liquid scintillation counter. For saturation binding, membrane homogenates were incubated with vehicle or test drug and increasing concentrations of [3H]-FNZ (0.3–9 nM). Drugs and solvents EA and its isolated compounds 4-methoxy-6-(11,12methylenedioxydihydrostyryl)-2-pyrone-1 (DST-1), 4methoxy-6-(11,12-methylenedioxy-14-methoxy-dihydrostyryl)-2-pyrone-2 (DST-2), 4-methoxy-6-(11,12-methylenedioxy-10,14-dimethoxy-dihydrostyryl)-2-pyrone-3 (DST3), 4-methoxy-6-(11,12-methylenedioxystyryl)-2-pyrone-4 (STY-4), 4-methoxy-6-(11,12-methylenedioxy-14methoxystyryl)-2-pyrone-5 (STY-5), and 4-methoxy-6(11,12-dimethoxystyryl)-2-pyrone-7 (STY-7) were obtained from the whole plant P. sabulosa, as previously described (Duarte et al. 2007). The drugs used were ethyl ether (F. Psychopharmacology Maia Ind. & Com. Ltda, Cotia, SP, Brazil), pentylenetetrazole hydrochloride (PTZ; Sigma Chemical Co., St. Louis, MO, USA), diazepam (DZP; Dienpax®, SanofiWinthrop Lab., São Paulo, SP, Brazil, and Hoffman-La Roche Ltd, Basel, Switzerland), and flumazenil (FMZ; Tocris, Ellisville, MO, USA). PTZ, DZP, and FMZ were dissolved in saline (0.9% NaCl) immediately before i.p. injection. EA was freshly suspended in 10% Tween-80 and tap water, before PTZ or electroshock-assays. DST (1, 2, or 3) and STY (4, 5, or 7) compounds were initially dissolved in 100% dimethylsulfoxide (DMSO; CAQ Ltda, SP, Brazil), then subsequently diluted in sterile phosphatebuffered saline (PBS; pH 7.4; Sigma Chemical Co., St Louis, MO, USA) to a final concentration of 0.8% DMSO, and injected into the brain lateral ventricle (i.c.v.). Phenytoin, phenobarbital (Cristália, SP, Brazil), carbamazepine and valproic acid (Sigma Co., St Louis, MO, USA) were dissolved in 10% Tween-80 and injected by i.p. route. Treatments Experiment 1: Effects of ethyl acetate fraction (EA) and GABA-acting drugs on the PTZ-induced electroencephalographic (EEG) activity and behavioral convulsions The rats were divided into six groups: three groups received saline subcutaneously (s.c.) followed by oral administration (per os route, p.o.) of vehicle (10% Tween-80; group 1), EA of P. sabulosa (750 mg/kg; group 2) or DZP (5 mg/kg; group 3) through an intragastric cannula and, 1 h later, received PTZ (80 mg/kg; i.p.); the other three groups received FMZ (5 mg/kg s.c.) followed by p.o. administration of vehicle (10% Tween-80; group 4), EA (750 mg/kg; group 5) or DZP (5 mg/kg; group 6), and 40 min later received an additional dose of FMZ (5 mg/kg s.c.). After 20 min, convulsions were induced by PTZ (80 mg/kg i.p.). Experiment 2: Effects of ethyl acetate fraction (EA) and anticonvulsant drugs on maximal electroshock-induced convulsions (MES) in mice Mice received vehicle (10% Tween-80, p.o.) or EA (250, 500 or 1,000 mg/kg, p.o.) and were submitted to the MESinduced convulsions. Positive control groups received phenytoin (7.5–30 mg/kg), phenobarbital (10–30 mg/kg), carbamazepine (5–30 mg/kg) or valproic acid (100– 300 mg/kg), by i.p. route, 1 h before MES. Vehicle (10% Tween-80) and DZP (1–10 mg/kg) were administered, by the same route, 30 min before MES. Experiment 3: Effects of pretreatment with flumazenil (FMZ) on the elevated plus-maze test performance of mice treated with i.c.v. injections of DST (1 or 3), STY (4 or 7) or DZP DST (1 or 3) or STY (4 or 7) compounds were injected (0.04–25 pmol) into the lateral ventricles (i.c.v.) of mice in a volume of 2 μl, 15 min after i.p. injection of either saline or FMZ (10 mg/kg). Afterwards, the mice were submitted to the EPM test. Control animals were similarly treated with vehicle (0.8% DMSO in PBS) and the standard anxiolytic DZP (7,000 pmol, effective dose chosen from previous studies) was used as the positive control drug. Data analysis All values are expressed as means±SEM. Data of experiments 1 and 2 were analyzed by two-way ANOVA (pretreatment×treatment) followed by the post-hoc Student Newman-Keuls’ test for multiple comparisons. Non-parametric data were compared by the Kruskal-Wallis ANOVA test followed by the Mann-Whitney U test (experiment 1). In the MES protocol (experiment 3), the statistical analysis was carried out using one-way ANOVA followed by the post-hoc Dunnett’s test for comparisons between treated and control groups. Lethality and MES-induced tonic extension ratios were compared by Fisher’s exact test. For the competition binding, data were analyzed by nonlinear regression of specific bound vs radioligand concentration. Regressions for two (full model) and one (reduced model) binding site(s) were further compared by the extra sum-of-squares F test. According to the best fit model, the affinity constant (Kd) and number of binding sites (Bmax) were estimated and compared (GraphPad Prism version 4.0® software). Ki values were calculated using the Cheng-Prusoff/Chou equation: Ki ¼IC50 =½1þðL=Kd ފ, where Ki refers to the inhibition constant of the unlabeled ligand, IC50 is the concentration of unlabeled ligand required to reach half-maximal binding, Kd refers to the equilibrium dissociation constant of the radioactive ligand and L refers to the concentration of radioactive ligand. For saturation binding, data were fitted by non-linear regression using the equation y ¼ Bmax :x=ðKd þ xÞ, where y is specifically bound [3H]-ligand in dpm, Bmax is maximal binding, and x is the concentration of [3H]-ligand. Differences were considered significant at P≤0.05. All tests were performed using the Statistica version 6.0® software package and the graphics were drawn with the GraphPad Prism version 4.0® Software. Results Effects of ethyl acetate fraction and GABA-acting drugs on the PTZ-induced electroencephalographic activity and behavioral convulsions EEG tracings of PTZ effects in both control and treated rats are shown in Fig. 1. After PTZ administration, control Psychopharmacology Fig. 1 Flumazenil (FMZ, 10 mg/kg, s.c.) effects on EEG recordings of PTZ-induced convulsions (80 mg/kg, i.p.) of rats (n=6) treated with EA (750 mg/kg, p.o.) of P. sabulosa or diazepam (DZP, 5 mg/kg, p.o.). Representative EEG recordings (8 s epochs) from A3 (Cx, upper) and CA3 (Hp, bottom) cortical and hippocampal areas, respectively, are shown animals (saline+ vehicle) presented a 2-min transition period characterized by a high-voltage fast epileptiform activity in cortex (A3) and hippocampus (CA3), followed by isolated spikes indicating the first MJ/C. In the third minute, bursts of high-voltage fast polyspiking activity denoted the first GCS, with a second GCS 12 min later. EA (saline + EA) and DZP (saline + DZP) blocked the PTZinduced seizures. Pretreatment with FMZ (FMZ + saline) had no effects per se on the seizures, but partially blocked the anticonvulsant effects of both EA (FMZ + EA) and DZP (FMZ + DZP). PTZ-induced convulsions in control rats (saline + vehicle) begun with a brief period of MJ of head and neck followed by clonus spreading to forelimbs. Afterwards, rats presented GCS with running fits followed by loss of posture and generalized clonus. Table 1 shows the duration, severity and lethality of the first MJ/C or GCS. The latency and duration of the first MJ/C were significantly affected by FMZ pretreatment [F(5,30)=7.90, P<0.01 and F(5,30)=7.74, P< 0.01, respectively] and EA or DZP treatments [F(5,30)=4.49, P<0.05 and F(5,30)=5.75, P<0.01, respectively], with an interaction of pretreatment and treatment [F(5,30)=3.73, P< 0.05 and F(5,30)=4.17, P<0.05, respectively]. Treatment with EA (saline + EA) increased the latency to first MJ/C (P<0.01) and reduced its duration (P<0.01), similar to DZP (saline + DZP; P<0.05 and P<0.01, respectively). There was no significant effect of FMZ (FMZ + saline) per se but this treatment was able to partially block the anticonvulsant effect of EA as well as of DZP (P<0.05). Latency and duration of the first GCS were significantly affected by FMZ pretreat- Table 1 Effects of oral administration of the EA (750 mg/kg) of P. sabulosa on the latency to the first myoclonic jerk/clonus (MJ/C) and generalized clonic seizure (GCS), as well as on the duration of GCS and the mortality, evaluated in the PTZ-induced convulsions, in rats Treatment s.c SAL SAL FMZ FMZ SAL FMZ Variables p.o VEH EA VEH EA DZP DZP MJ/C GCS Severity Latency (s) Duration (s) Latency (s) Duration (s) 59.0±6.2 93.5±10.5** 64.7±4.7 68.0±5.3*** 89.2±6.7* 63.5±4.1*** 22.8±2.7 14.5±1.2** 21.0±1.0 22.0±1.9*** 12.8±1.8** 19.2±1.4*** 120.2±7.8 600.0±0.0**a 141.3±11.3 600.0±0.0**a 600.0±0.0**a 136.7±23.8**** 66.5±3.1 0.0±0.0**a 63.7±5.7 0.0±0.0**a 0.0±0.0**a 69.8±1.6**** 5 3 5 3 3 5 (5–5) (3–3) (5–5) (3–3) (3–3) (5–5) Lethality (%) ** ** ** **** 100 0** 100 0** 0** 100**** Values of latency and duration represent mean+SEM for six animals each group being analyzed by a two-way ANOVA followed by Student Newman-Keuls’ test. The severity values represent median+interquartile range (Q1–Q3) being analyzed by non-parametric Kruskal-Wallis ANOVA test followed by Mann-Whitney U test. The percentage of death in 24 h was assessed by Fischer’s exact test. SAL, saline; VEH, vehicle a Animals that not presented GCS, being the latency and duration time considered as 600 s and 0 s, respectively. *P<0.05 or **P<0.01 as compared to control group (SAL/VEH) and ***P<0.05 or ****P<0.01 as compared to SAL/EA or SAL/DZP. Psychopharmacology the seizure severity (P<0.01), without modifying the protective effects of EA on this same behavioral parameter (P>0.05). Moreover, EA (saline + EA) as well as DZP (saline + DZP), fully protected the rats from death after PTZ convulsion (P<0.01). Regarding lethality, FMZ blocked the protective effect of DZP (P<0.01,), without modifying the EA effect (P>0.05). ment [F(5,30)=258.14, P<0.0001 and F(5,30)=100.37, P< 0.0001, respectively], EA and DZP treatments [F (5,30)=872.24, P<0.0001 and F(5,30)=284.65, P<0.0001, respectively] with interactions [F(5,30)=296.99, P<0.0001 and F(5,30)=113.65, P<0.0001, respectively]. EA treatment delayed the latency to first GCS (P<0.0001) and reduced its duration (P<0.0001), similar to DZP effects (P<0.0001). FMZ totally blocked the anticonvulsant effect of DZP (P< 0.0001), without affecting the anticonvulsant effect of EA. Statistical analysis revealed significant between-group differences for seizure severity [H(5) = 36.00; P<0.0001]. EA treatment (saline + EA) significantly reduced the seizure severity similar to DZP (saline + DZP)-treated group (P< 0.01). FMZ (FMZ + vehicle) had no effect per se (P>0.05), but completely blocked the anticonvulsant effect of DZP on Fig. 2 a and b Effects of EA (250–1,000 mg/kg; p.o.) of P. sabulosa and standard anticonvulsant drugs c and d diazepam (DZP), phenytoin (PHT), e and f phenobarbital (PHB), carbamazepine (CBZ) and valproic acid (VALP) on duration of tonic hind-limb extension and percentage of animals that presented tonic extension in the MES-induced convulsions in mice. Each value represents the mean±SEM or percentage of 10–12 animals. **P≤0.01 as compared to control (one-way ANOVA followed by Dunnett’s test or Fisher’s exact probability test, respectively) Effects of ethyl acetate fraction and anticonvulsant drugs on maximal electroshock-induced convulsions in mice EA (500–1,000 mg/kg) protected mice against MES-induced seizures, showing a significant decrease in the duration of tonic hind-limb extension (Fig. 2a) and in the percentage of animals presenting tonic extension (P<0.01; Fig. 2b). These effects % Tonic extension Time of tonic hindlimb extension (s) a b 18 100 15 75 12 9 50 ** 6 ** 3 0 ** 25 ** 0 - 250 500 EA - 1000 c 250 500 EA 1000 (mg/Kg) d 18 100 15 75 12 9 50 ** ** 6 ** 3 0 5 7.5 10 DZP 1 ** 25 ** - ** ** 0 7.5 15 30 PHT e - ** 7.5 15 30 (mg/Kg) PHT 5 7.5 10 DZP 1 f 100 18 15 75 12 9 6 ** ** ** 3 0 ** ** - 10 15 20 30 PHB 50 ** ** ** ** ** 25 ** 0 5 10 20 30 100 200 300 CBZ VALP ** * - 10 15 20 *30 PHB ** ** ** 5 10 20 30 100 200 300 (mg/Kg) CBZ VALP Psychopharmacology were similar to those obtained with the standard anticonvulsant drugs (phenytoin, phenobarbital, carbamazepine or valproic acid), except DZP (Fig. 2c,d,e, and f). Effects of flumazenil on the elevated plus-maze performance of mice treated with i.c.v. injections of DST (1 or 3), STY (4 or 7) or DZP The time spent into the open arms of the EPM was significantly affected by i.c.v. treatment with DST-1 [F (5,56)=16.58, P < 0.0001], DST-3 [F(7,63)=26.47, P < 0.0001], STY-4 [F(5,64)=16.05, P<0.0001], STY-7 [F (5,45)=14.95, P < 0.001] or DZP [F(3,23)=23.71, P < 0.0001]. DST-1 (5 pmol) increased the exploration of open arms (P<0.001; Table 2). This compound also promoted a reduction in the number of pSAP as well as an increase in the uHD and OAEA (P<0.001) whereas FMZ, which had no effect per se, did not block these effects. Moreover, FMZ (1 or 5 mg/kg) did not affect the reduction in pSAP nor the increase in uHD and OAEA induced by DST-1. DST-3 (0.04 pmol) increased the exploration of open arms (P< 0.001) and reduced pSAP (P<0.001).and also increased uHD and OAEA (P<0.0001) whereas FMZ (1–10 mg/kg) partially blocked these effects. STY-4 (25 pmol) and STY-7 (5 pmol) promoted a similar anxiolytic-like profile of action which was partially blocked by FMZ. Those findings were also similar to those observed after DZP (7,000 pmol) and FMZ completely blocked this anxiolytic effect, except for the parameters entries into the open arms and pSAP (P>0.05). All data obtained in the EPM test are presented in Table 2. The other behavioral parameters were not affected by any treatment (data not shown). Table 2 Effects of i.c.v. injection of DST (1 or 3), STY (4 or 7) isolated from P. sabulosa or DZP, 15 min after the i.p. injection of vehicle (saline) or FMZ (1–10 mg/Kg), on the parameters evaluated in EPM test in mice Treatments Variables i.p. (mg/kg) i.c.v. (pmol) N % TOA % OE uHD pSAP OAEA SAL FMZ FMZ SAL FMZ FMZ SAL FMZ FMZ FMZ SAL FMZ FMZ FMZ SAL FMZ FMZ SAL FMZ FMZ SAL FMZ FMZ SAL FMZ FMZ SAL FMZ SAL FMZ PBS PBS PBS DST-1 5 DST-1 5 DST-1 5 PBS PBS PBS PBS DST-3 0.04 DST-3 0.04 DST-3 0.04 DST-3 0.04 PBS PBS PBS STY-4 25 STY-4 25 STY-4 25 PBS PBS PBS STY-7 5 STY-7 5 STY-7 5 PBS PBS DZP 7,000 DZP 7,000 12 10 8 16 7 9 12 10 8 8 8 8 7 10 17 8 8 17 11 9 12 9 8 7 7 8 7 7 6 7 16.4±3.8 9.1±3.1 19.8±7.7 46.8±7.9** 34.3±9.2 36.6±7.1 15.5±4.0 9.6±3.0 19.8±7.7 18.8±5.5 46.2±10.7* 43.4±7.6* 42.1±8.4* 28.0±6.0 12.8±2.9 9.59±3.8 19.8±7.7 49.8±7.9** 39.8±10.2 27.8±8.0 14.4±3.7 10.1±3.3 19.8±7.7 45.0±7.1* 40.0±13.2 27.3±8.2 14.3±4.8 19.3±7.2 58.2±7.8** 35.3±4.6**** 28.6±5.0 27.8±4.6 26.6±6.5 52.5±5.9** 47.9±5.6 43.4±5.2 26.3±4.3 29.8±3.6 26.6±6.5 26.5±5.8 50.5±5.6* 48.8±3.6* 47.5±4.6 37.8±6.8 29.9±3.6 26.4±5.7 26.6±6.5 54.1±5.4* 43.9±7.6 42.8±7.2 26.7±4.2 30.9±3.8 26.6±6.5 55.5±6.9* 41.2±10.9 37.7±8.0 11.0±1.0 9.0±1.2 13.0±2.2 11.0±0.6 8.7±2.9 4.8±2.9 4.5±2.2 25.0±4.2** 22.7±2.7* 19.2±2.9* 7.6±2.5 4.8±1.6 4.5±2.2 8.2±3.3 27.8±5.6** 26.9±6.0** 20.4±3.9 14.1±4.0 6.1±1.7 5.0±1.9 4.5±2.2 26.9±4.2** 24.1±6.5** 10.7±3.5 6.3±2.00 5.3±1.6 4.5±2.2 18.9±5.3* 14.3±6.2 11.8±4.0 7.3±2.3 7.6±3.1 25.3±4.3** 13.3±1.7*** 19.4±2.6 22.2±4.0 18.1±1.6 7.2±1.4** 10.4±1.4* 8.4±2.1* 19.2±1.9 21.1±3.7 18.1±1.6 19.1±2.3 9.6±2.3** 12.6±3.9 12.1±1.7 16.3±2.3 17.8±1.5 24.5±4.7 18.1±1.6 8.8±2.1* 15.4±4.3 13.3±1.9 21.3±2.6 20.7±4.2 18.1±1.6 6.1±1.6** 10.6±3.9 13.8±2.2 21.4±1.2 21.1±1. 7 10.5±1.0** 15.3±1.2 1.3±0.6 0.7±0.4 1.2±0.7 6.8±1.2** 6.9±1.9** 5.4±1.6* 0.9±0.4 0.7±0.4 1.2±0.7 1.4±0.5 5.6±1.4** 3.9±1.2 4.0±1.0 3.1±0.8 0.6±0.3 0.9±0.5 1.2±0.7 4.6±1.1* 4.3±1.4 1.1±0.7 0.8±0.4 0.8±0.5 1.2±0.7 5.4±1.1** 2.6±1.2 2.4±1.2 2.4±0.8 2.9±0.8 8.8±0.6** 6.1±0.6 1 5 1 5 1 5 10 1 5 10 1 5 1 5 1 5 1 5 10 10 Each value represents the mean+SEM. Data were analyzed by a two-way ANOVA followed by Student Newman-Keuls’ test N, number of animals, % TOA, time spent into open-arms, % EOA, frequencies of entries into open-arms. *P<0.05 or **P<0.01 as compared to respective control group (SAL/PBS); ***P<0.05 or ****P<0.01 as compared to SAL/DZP group Psychopharmacology b a 125 125 DST (3) % specific bound of 3[H]-FNZ % specific bound of 3[H]-FNZ Fig. 3 Competition curves of a DST-3 and b STY-7 for [3H]FNZ binding to washed crude synaptosomal membranes of the rat cerebral cortex. Data represent the mean of four to five individual experiments done in duplicate 100 75 50 25 0 -7 -6 -5 Effects of the DST (1, 3 or 4), STY (5, 6 or 7) or DZP on [3H]-FNZ binding DST-3 (Fig. 3a) and STY-7 (Fig. 3b) inhibited the binding of [3H]-FNZ to the BDZ-bs with Ki (CI95%) values of 35.8 (28–45) μM and 70.0 (60–79) μM, respectively. The other DST compounds (1 or 2) showed Ki (CI95%) values above 100 μM and 41.7 (20–86) μM, respectively, while STY (4 or 5) inhibited the [3H]-FNZ binding with Ki (CI95%) values of 90.3 (67–122) μM and 31.0 (22–41) μM, respectively. DZP, the positive control, caused inhibition of [3H]-FNZ binding with Ki (CI95%) value of 0.008 (0.006–0.01) μM. Effects of the DST-3 and STY-7 on the affinity and density of [3H]-FNZ binding sites In all cases, data obtained were best fitted to one site binding hyperbola (Fig. 4). Scatchard plot analysis of saturation curves showed that both the compound DST-3 40 μM and STY-7 70 μM competitively inhibited the binding of [3H]-FNZ with a significant decrease in [3H]FNZ binding apparent affinity (Kd) [F(2,9)= 5.8, P<0.05] and no change in maximal binding (Bmax) [F(2,9)= 1.4, P> 0.05]. Kd (95% CI) and Bmax (95% CI) values for DST-3 were 3.3 (0–8) nM and 0.56 (0.18–0.94) nM, respectively, 0.12 0.2 0.10 Bound/Free Specific binding (pmol/mg) vehicle DST (3) STY (7) 0.3 0.1 vehicle DST (3) STY (7) 0.08 0.06 0.04 0.02 0.00 0.0 0.1 0.2 0.3 0.4 Bound (pmol/mg) 0.0 0 1 2 3 75 50 25 0 -8 0.4 STY (7) 100 4 5 [3H]-FNZ (nM) Fig. 4 Effects of DST-3 (40 μM) or STY-7 (70 μM) on [3H]-FNZ binding to washed crude synaptosomal membranes of the rat cerebral cortex. Saturation isotherms and Scatchard plot of [3H]-FNZ (0.3– 9 nM) are shown. Data were fitted to one site binding hyperbolas and the apparent affinity constant (Kd) and the number of binding sites (Bmax) were estimated. Data represent the mean of three individual experiments done in duplicate -4 -3 log[M] -7 -6 -5 -4 -3 log[M] while the values for STY-7 were 2.7 (1.2–4.2) nM and 0.48 (0.35–0.60) nM. Discussion In our previous study, we have demonstrated the anticonvulsant, hypno-sedative, and anxiolytic-like properties of the EA obtained from P. sabulosa in mice, suggesting a BDZ-like action. The active principle(s) responsible for these central actions were suggested to be DST and STY compounds, the major constituents of EA (Duarte et al. 2007). However, the underlying mechanism of action of this fraction and its isolated compounds was not investigated. Thus, the present work was carried out to study the participation of the BDZ-bs of the GABAA receptor in these actions in in vivo and in vitro studies. The rats pretreated with EA (a fraction containing all DST and STY compounds) exhibited only MJ/C and did not show any GCS, indicating a protective effect of EA against PTZ-induced convulsions. As MJ/C and GCS indicate initiation and spread of the seizures, respectively, our results suggest that EA blocked the spread of seizures, distinct from DZP, which blocked both epileptic parameters (Crestani et al. 2000). The anticonvulsant effect of EA was also observed in the EEG recording, through a dramatic decrease in the epileptiform activity induced by PTZ, a DZP-like profile. Moreover, FMZ blocked the protective effect of EA on the MJ/C, but did not alter the protective effect against the GCS as well as the mortality rate, in contrast to DZP, suggesting that FMZ only partially counteracted the anticonvulsant effect of EA. Furthermore, the MES test was included in this study to open up the pharmacological characterization of P. sabulosa since this test was not previously used. EA significantly reduced the time of tonic hind limb extension as well as the percentage of mice that displayed tonic extension in the MES-induced convulsions, an effect similar to several anticonvulsant drugs. MES-induced tonic extension can be prevented either by drugs that present GABAergic action, such as phenobarbital or valproic acid (Rogawski and Porter 1990; Löscher 2002), drugs that inhibit voltage-dependent sodium channels, such as phenytoin, carbamazepine, valproic acid, Psychopharmacology felbamate, and lamotrigine (Rogawski and Porter 1990; MacDonald and Kelly 1995), or drugs that block glutamatergic excitation mediated by the NMDA receptor, such as felbamate (Subramaniam et al. 1995). Our results suggest that the EA presents an anticonvulsant profile of action which differs from BDZ compounds, maybe with the participation of other underlying mechanisms, ion channels and/or neurotransmitters, since DZP, in contrast to EA, phenytoin, phenobarbital, carbamazepine and valproic acid, only promoted a weak reduction in the hind limb tonic extension in the MES test, and at doses considered ataxic and hypnotic. These findings are in agreement with the weak ability of BDZ drugs to prevent MES-induced seizures at doses lower than the neurotoxic TD50 (Kupferberg 1992; Mandhane et al. 2007). Taken together, the present results extend those of our previous study and indicate the involvement of BDZ-s in the anticonvulsant actions of EA, besides indicating other FMZ-insensitive mechanisms in its central effects, which will deserve further investigation. In the EPM test, DST (1 and 3) and STY (4 and 7) elicited several behavioral alterations which characterize an anxiolytic-like action and the absence of motor effects supports a specific action in accordance with our recent data (Duarte et al. 2007). In the present work, FMZ promoted a dose-dependent partial blockade of the anxiolytic-like actions of DST-3 and STY (4 and 7) but not of DST-1, suggesting that other FMZ-insensitive sites of the GABAA receptor complex, or another neurotransmitter system such as the 5-HT receptors (Clenet et al. 2005; Starr et al. 2007), could be involved in the DST-1 action, a finding that deserves further investigation. Furthermore, FMZ was not able to completely block all the behavioral alterations elicited by DZP in the EPM test which could be explained by the potential involvement of both populations of FMZ-sensitive and insensitive GABAA receptors or even of non-GABAergic mechanisms in the effects observed for DZP, DST, and STY compounds. Some studies support this explanation since it is well known that BDZ drugs influence a range of non-GABAergic mechanisms (Polc 1991; Ishizawa et al. 1997), any one of which could potentially play a role in the anxiolytic effects of DZP, as well as of the DST and STY compounds here studied. To characterize the pharmacological profile of compounds isolated from P. sabulosa and to substantiate the extrapolation made in vivo, three DST and three STY were studied in the 3H-FNZ binding assay to test their putative action as modulators of the BDZ-bs. All compounds were able to displace 3H-FNZ binding to a different extent, suggesting a central BDZ-like activity. However, DST-1 inhibited only 49% of the BDZ binding, showing a very low affinity for the BDZ-bs, at concentrations above 100 μM. The other compounds, DST (2 and 3) and STY (4, 5 and 7), exhibited a moderate affinity for the BDZ-bs (in the μM range), in contrast to DZP (nM). In the saturation experiments, the ability of DST-3 or STY-7 to modify the Kd without any significant change in the Bmax suggests that these compounds interact competitively at only one recognition site, probably the BDZ-bs. Our in vivo results revealed that DST and STY are very effective compounds, with anxiolytic-like effects in the range of doses from fmol to pmol, an order of magnitude lower than those found for DZP (nmol). Our in vitro study suggest the involvement of the BDZ-bs of the GABAA receptor in underlying the anticonvulsant and anxiolyticlike actions of the EA as well as of DST and STY compounds isolated from P. sabulosa. However, DST and STY compounds inhibited 3H-FNZ binding with an affinity in the μM range which suggests that besides an action on the GABAA receptor, other mechanisms of action may contribute to the central effects of this plant species. Therefore, these compounds could be useful as new tools to open up the search for new therapeutic agents to treat anxiety and related disorders. Acknowledgment F. S. Duarte is a recipient of a post-doctoral fellowship from the Brazilian National Research Council (CNPq), M. Duzzioni and B. G. Mendes are recipients of a Ph.D. scholarship from CAPES. CNPq also provided research grants to T. C. M. De Lima and M.G. Pizzolatti. The authors would like to thank Dr. José Marino-Neto for providing laboratory facilities to perform EEG studies, and Dr. Gareth Cuttle for final English revision of the text. References Ashton H (1994) Guidelines for the rational use of benzodiazepines. When and what to use. Drugs 48(1):25–40 Blumenthal M (2002) Kava safety questioned due to case reports of liver toxicity. HerbalGram 55:26–32 Brauer RB, Stangl M, Stewart JR, Pfab R, Becker K (2003) Acute liver failure after administration of herbal tranquilizer kava-kava (Piper methysticum). J Clin Psychiatry 64(2):216–218 Clenet F, Hascoet M, Fillion G, Galons H, Bourin M (2005) Role of GABA-ergic and serotonergic systems in the anxiolytic-like mechanism of action of a 5-HT-moduline antagonist in the mouse elevated plus maze. Behav Brain Res 158(2):339–348 Cole JC, Rodgers RJ (1994) Ethological comparison of the effects of diazepam and acute/chronic imipramine on the behaviour of mice in the elevated plus-maze. Pharmacol Biochem Behav 52(3):473–478 Crestani F, Martin JR, Mohler H, Rudolph U (2000) Mechanism of action of the hypnotic zolpidem in vivo. Br J Pharmacol 131 (7):1251–1254 Dhawan K, Dhawan S, Sharma A (2004) Anti-anxiety studies on extracts of Passiflora incarnata Linneaus. J Ethnopharmacol 78 (2–3):165–170 Duarte FS, Duzzioni M, Mendes BG, Pizzolatti MG, De Lima TC (2007) Participation of dihydrostyryl-2-pyrones and styryl-2pyrones in the central effects of Polygala sabulosa (Polygalaceae), a folk medicine topical anesthetic. Pharmacol Biochem Behav 86(1):150–161 Psychopharmacology Escher M, Desmeules J, Giostra E, Mentha G (2001) Hepatitis associated with Kava, a herbal remedy for anxiety. BMJ 322 (7279):139 Haley TJ, McCormick WG (1957) Pharmacological effects produced by intracerebral injection of drugs in the conscious mouse. Br J Pharmacol 12(1):12–15 Hui KM, Huen MSY, Wang HY, Zheng H, Sigel E, Baur R, Ren H, Li ZW, Wong JT-F, Xue H (2002) Anxiolytic effect of wongonin, a benzodiazepine receptor ligand isolated from Scutellaria baicalensis Georgi. Biochem Pharmacol 64(9):1415–1424 Ishizawa Y, Furuya K, Yamagishi S, Dohi S (1997) Non-GABAergic effects of midazolam, diazepam and flumazenil on voltagedependent ion currents in NG108–15 cells. Neuroreport 8 (11):2635–2638 Kupferberg HJ (1992) Strategies for identifying and developing new anticonvulsant drugs. Pharm Weekbl Sci 14(3A):132–138 Leung WC, Zheng H, Huen M, Law SK, Xue H (2003) Anxiolyticlike action of orally administered dl-tetrahydropalmatine in elevated plus-maze. Prog Neuro-Psychopharmacol. Biol Psychiatry 27(5):775–779 Lister RG (1987) The use of a plus-maze to measure anxiety in the mouse. Psychopharmacology (Berl) 92(2):180–185 Löscher W (2002) Basic pharmacology of valproate: a review after 35 years of clinical use for the treatment of epilepsy. CNS Drugs 16(10):669–694 Luddens H, Korpi ER, Seeburg PH (1995) GABAA/benzodiazepine receptor heterogeneity: neurophysiological implications. Neuropharmacology 34(3):245–254 MacDonald RL, Kelly KM (1995) Antiepileptic drug mechanisms of action. Epilepsia 36(2):S2–12 Mandhane SN, Aavula K, Rajamannar T (2007) Timed pentylenetetrazol infusion test: A comparative analysis with s.c. PTZ and MES models of anticonvulsant screening in mice. Seizure 16 (7):636–644 Marder M, Viola H, Wasowski C, Fernandez S, Medina JH, Paladini AC (2003) 6-methylapigenin and hesperidin: new valeriana flavonoids with activity on the CNS. Pharmacol Biochem Behav 75(3):537–545 Maurice T, Lockhart BP, Privat A (1996) Amnesia induced in mice by centrally administered beta-amyloid peptides involves cholinergic dysfunction. Brain Res 706(2):181–193 Medina JH, Paladin AC, Wolfman C, Levi de Stein M, Calvo D, Diaz LE, Pena C (1990) Chrysin (5,7-di-OH-flavone), a naturally- occurring ligand for benzodiazepine receptors, with anticonvulsant properties. Biochem Pharmacol 40(10):2227–2231 Möhler H (2006) GABAA receptors in central nervous system disease: anxiety, epilepsy, and insomnia. J Recept Signal Transduct Res 26(5–6):731–740 Nutt DJ, Malizia AL (2001) New insights into the role of the GABA (A)-benzodiazepine receptor in psychiatric disorder. Br J Psychiatry 179:390–396 Polc P (1991) GABA-independent mechanisms of benzodiazepine action. In: File SE, Briley M (eds) New concepts in anxiety. London, MacMillan Publishers, pp 211–236 Racine RJ (1972) Modification of seizure activity by electrical stimulation. II. Motor seizure. Electroencephalogr Clin Neurophysiol 32(3):281–294 Rodgers RJ, Dalvi A (1997) Anxiety, defence and the elevated plusmaze. Neurosci Biobehav Rev 21(6):801–810 Rogawski MA, Porter RJ (1990) Antiepileptic drugs: pharmacological mechanisms and clinical efficacy with consideration of promising developmental stage compounds. Pharmacol Rev 42(3):223–286 Russmann S, Lauterburg BH, Helbling A (2001) Kava hepatotoxicity. Ann Intern Med 135(1):68–9 Singh YN (2005) Potential for interaction of kava and St. John’s wort with drugs. J Ethnopharmacol 100(1–2):108–13 Starr KR, Price GW, Watson JM, Atkinson PJ, Arban R, Melotto S, Dawson LA, Hagan JJ, Upton N, Duxon MS (2007) SB-649915B, a Novel 5-HT(1A/B) Autoreceptor Antagonist and Serotonin Reuptake Inhibitor, is Anxiolytic and Displays Fast Onset Activity in the Rat High Light Social Interaction Test. Neuropsychopharmacology 32(10):2163–2172 Subramaniam S, Rho JM, Penix L, Donevan SD, Fielding RP, Rogawski MA (1995) Felbamate block of the N-methyl-Daspartate receptor. J Pharmacol Exp Ther 273(2):878–886 Swinyard EA, Brown WC, Goodman LS (1952) Comparative assays of antiepileptic drugs in mice and rats. J Pharmacol Exp Ther 106 (3):319–330 Viola H, Marder M, Wasowski C, Giorgi O, Paladini AC, Medina JH (2000) 6,3′-dibromoflavone and 6-nitro-3′-bromoflavone: new additions to the 6,3′-disubstituted flavone family of high-affinity ligands of the brain benzodiazepine binding site with agonistic properties. Biochem Biophys Res Commun 273(2):694–698 Zhang ZJ (2004) Therapeutic effects of herbal extracts and constituents in animal models of psychiatric disorders. Life Sci 75 (14):1659–1699