Phytomedicine , Vol. 7(6), pp . 463-469© Urban & Fischer Verlag 2000http: //www.urba nfischer.de/journal s/ph ytom ed

Anxiolytic-antidepressant activity of Withania somniferaglycowithanolides: an experimental study

S. K. Bhattacharya1, A. Bhattacharya/ , K. Sairam 1 and S. GhosaF

IDepar tment of Pharm acology, Institute of Medical Sciences, Banaras Hindu University, Varanasi, India2Drug Research & Development Centre, Jessore Road, Calcutta, India

Summary

The roots of Withania somnifera (WS) are used extensively in Ayurveda, the classical Indian system ofmedicine, and WS is categorized as a rasayana, which are used to promote physical and mental health,to provide defence against disease and adverse environmental factors and to arrest the aging process. WShas been used to stabilize mood in patients with behavioural disturbances. The present study investigat­ed the anxiolytic and antidepressant actions of the bioactive glycowithanolides (WSG), isolated from WSroots, in rats. WSG (20 and 50 mg/kg) was administered orally once daily for 5 days and the results werecompared by those elicited by the benzodiazepine lorazepam (0.5 mg/kg, i.p.] for anxiolytic studies, andby the tricyclic anti-depressant, imipramine (10 mg/kg, i.p.), for th e antidepressant investigations. Boththese standard drugs were administered once, 30 min prior to the tests. WSG induced an anxiolytic ef­fect, comparable to that produced by lorazepam, in the elevated plus-maze, social interaction and feed­ing latency in an unfamiliar environment, tests. Further, both WSG and lorazepam, reduced rat brainlevels of tribulin, an endocoid marker of clinical anxiety, when the levels were increased following ad­ministration of the an xiogenic agent, penrylenetetrazole. WSG also exhibited an antidepressant effect,comparable with that induced by imipramine, in the forced swim-induced 'behavioural despair' and'learn ed helpl essness ' tests. The investigations support the use of WS as a mood stabilizer in clinical con­ditions of anxiety and depression in Ayurveda.

Key words: Withania somnifera, glycowithanolides, anxiolytic-antidepressant acti on s, tribulin.

Introduction

Withania somnifera Dunal (family: Solanaceae ),known as Ashwagandha in Ayurvcda, the classical In­dian system of med icine, has been categorized as arasayana drug. Rasayanas are used to promote physicaland mental health by increasing defence against dis­eases and adverse environmenta l factors, for arrestingthe aging process, revitalizing the body in debil itatedconditions and as a nervine tonic for mood sta biliza­tion (Rege et aI., 1999). Extracts of the roots of W.somnifera (WS) are extensively used to reverse memorydeficits and to normalize behavioural perturbations(Weiner and Weiner, 1994 ). Limited clinical tri als sug­gest th at WS may exert an anxiolytic-antidepressant ef­fect in psychiatric patients (Udupa and Singh, 1993).Th e bioacti ve glycowithanolides of WS (WSG), com -

pnsmg of sitoindosides VII to X, and withaferin A(Bha ttacharya et aI., 198 7; Ghosal et al., 198 9), haveearl ier been found to pre vent acute stress-ind uced anx­iety (Bha ttacha rya et aI., 1987) and chronic stress-in­duced depression (Bhattacharya, 1998 ) in rats . WSGalso inhibited the cognitive deficits seen in animal mod­els of Alzheimer's disease (Bhattacharya et aI., 1995).The present investigation was undertaken to assess thepossible anxiolytic and antidepressant actions of WSG,using well valida ted rat model s of anx iety and dep res­sion,

0944-7113199107/06-463 $ 15.00/0

464 S. K. Bhattacharya et al.

Materials and Methods

AnimalsThe study was conducted on inbred Charles Fosterstrain male rats (160-180 g), housed in colony cages atambient temperature of 25 ± 2 °C and 45-55% rela­tive humidity, with a 12 h light/12 h dark cycle. Unlessotherwise mentioned, the rats had free access to stan­dard pellet chow and drinking water. Experiments wereconducted between 9.00 and 14.00 hours. 'Principlesof laboratory animal care' (NIH publication no. 85-23,revised 1985) guidelines were followed.

Drugs and vehicle treatmentAn aqueous concentrate of WS roots (Dabur, India)was exhaustively extracted with chloroform to re­move fatty materials and free withanolides. The aque­ous solution was then spray-dried. The compositionof sitoindosides VII-X and withaferin (collectively re­ferred to as glycowithanolides, WSG) in the spray­dried residue was determined by HPTLC (CAMAGassembly, software (C) 1990; II, V. 3.14/Pc/CATSversion; 3.05 1 amp, deuterium, wavelength 254 nm,reflection mode; developer, (1) n-BuOH-AcOH-H20(4:1:2); (2) CHClrMeOH (90:10), using authenticmarkers (Bhattacharya et al., 1987; Ghosal et al.,1989). The equimolar composition of the constituentswas made in an aqueous stock solution by adding ap­propriate amounts of the deficient compounds (Ghos­al et al., 1989). The combined formulation of WSGwas freely soluble in water and normal saline. Totalsteroid content of WS roots was found to be 1.13mg% of the dry powder. The method adopted for de­termining was modified Liebermann-Burchard reac­tion (Akbarsha et al., in press).

WSG was dissolved in distilled water and adminis­tered orally (p.o.) in the doses of 20 and 50 mg/kg,once daily for 5 days, the last administration being onday 5, one hour prior to the test procedures, unlessmentioned otherwise. Lorazepam (LZP, Cipla, India;0.5 mg/kg, i.p.) and imipramine (IPM, Sun Pharma, In­dia, 10 mg/kg, i.p.) were used as the standard drugs forcomparison and were administered acutely once, 30min before the test procedures.

Experimental methodsThe following methods were used:

Anxiolytic activitya. Elevated plus-maze test: The maze consisted of twoopposite open arms (50 em x 10 em) crossed with twoclosed arms of the same dimension but with walls50 em high. The maze had a central square (10 em x 10em), which gave the maze the shape of a plus sign, and

was kept elevated 50 em above the floor in a dimly-litroom. Naive rats were placed individually on the mazecentre and the number of entries and time spent by therat on the closed and open arms was recorded duringthe next 5 min (Pellow et al., 1985; Bhattacharya andSatyan, 1997; Vogel and Vogel, 1997).b. Social Interaction test: Rats were housed individual­ly 5 days prior to the test. The social interaction arenawas a wooden box (60 cmx60 cmx35 em), placed in adimly-lit room. On day 6, the rats were placed individ­ually in the box and given two familiarization sessionsof 7.5 min at 2 h interval. On day 7, the rats werepaired on the basis of weight and placed in the box for7.5 min. During this time, the total time spent by therat pair in active social interaction (sniffing, grooming,following, boxing, kicking, biting, crawling over or un­der the partner) was recorded (File and Hyde, 1978;Bhattacharya and Satyan, 1997; Vogel and Vogel,1997) WSG or the vehicle was administered from day3 to day 7, whereas LZP was administered on day 7,30 min prior to the pairing of the rats. Immediately af­ter recording social interaction, the rats were placed in­dividually in an eight-beam photocell motor activitychamber (Actophotometer, Blue Line Inc). Locomotoractivity was recorded for 7.5 min, as indicated by pho­tobeam break counts.c. Novelty-suppressed feeding latency test: The test ap­paratus was the same used for the social interactiontest. The box floor was covered with a 2.5 em layer ofwooden chips on which 15 laboratory chow pelletswere evenly placed. A similar arrangement was made inthe home cage of the rats. Food was withdrawn fromthe home cage 48 h prior to the test but water was giv­en ad libitum. The rats were placed individually in thetest cage and the latency to begin eating, defined aschewing the pellet and not merely sniffing or playingwith it, was recorded, with a cut off period of 300 s.The results were compared with that of another groupof rats where latency to feed was recorded in the homecage under identical conditions (Bodnoff et al., 1988;Bhattacharya and Satyan, 1997).d. Rat brain tribulin activity: Tribulin activity was as­sessed in terms of endogenous monoamine oxidase(MAO) A and MAO B inhibiting activity. The extrac­tion procedure and the assay technique were similar toearlier described procedures (Medvedev et al., 1992).The enzyme sources and substrates for estimatingMAO A and MAO B inhibiting activity were humanplacental homogenate and human platelets, and (He)5-hydroxytryptamine and (14e) phenylethylamine, re­spectively (Medvedev et al., 1982).

Antidepressant activitya. Swim stress-induced 'behavioural despair' test: Theapparatus used was a modified version of that described

Results

Rat brain tribulin activity

Although WSG (20 and 50 mg/kg, p.o.) and LZP tend­ed to lower rat brain tribulin activity, assessed in termsof endogenous MAO A and MAO B inhibitory act ivity,the effects remained sta tis tica lly insignificant (Table 4).Pent ylenetetrazole (PT Z, 20 mg/kg, i.p. ), a well-estab­lished anxiogenic agent (Bhattacharya and Sat yan ,1997), markedly increased rat brain M AO A and B in­hib itory activities by 109.1 % and 70.6 %, respectively.Thi s increase was inhibited by WSG (20 and 50 mg/kg,p.o. ) and by LZP (Table 4).

Novelty-suppressed feeding latency test

WSG (20 and 50 mg/kg, p.o.l produced a decrease inthe latency to feed in the test chamber in 48 -h food-de­priv ed rats, as did LZP. H owever, neither WSG or LZPhad an y significant effect on feeding latency in th ehome cage (Table 3).

Social Interaction test

WSG (20 and 50 mg/kg, p.o.) produced an increase inthe total time spent by the rat pairs in acti ve social in­teraction, as did LZP. Ne ith er WSG or LZP had anysignificant effect on locomot or acti vity in th e dos esused (Table 2).

Elevated plus-maze test

WSG (20 and 50 mg/kg, p.o.) -produced, an increase inthe number of entries an d time spent on the open armsof the maze, with concomitant decrease in these indic eson the closed arms, as did the standard anxiolyticagent, LZP. There was no statistical difference in the to­tal time spent and number of entries on both the openand closed arms of the maze between the vehicle-treat­ed control group and any of the drug-treated groups(Table 1).

Rotarod testThis test was used to assess th e pre sence of drug-in­duc ed motor defic it, if any, which can vitiate the re­sults. Rats were placed indiv idua lly on a rotarod appa­ratus (Techno), rotating at a speed of 16 rpm. The testcriteri on was the ability of the ra t to rema in on the ro ­tating rod for 30 s. Each rat was allo wed up to 3 trialsto reach criterion and th e percent age of rats not reach­ing criterion in each expe rimental gro up wa s recorded(Bennett et al. , 1985).

Anxiol ytic-antidepressant activity of Withania somnifera glycowithanolides: an experimental study 465

by Porsolt et al. (1978 ). Rats were forced to swim indi- treated groups. The Chi squa re test wa s utilized for thevidually in a pol ypropylene vessel (45 ern x 40 cm x If quantal data in the rotarod test.ern), with a water level of 20 em, which ensured that therat 's feet did not touch the vessel floor and that it couldnot climb out. The rat was allowed to swim for 10 min ,thereafter during the next 5 min the total period of im­mobility, characterized by complete cessation of swim­ming with the head floating just above the water level,wa s noted. The immobility period , after initial frenziedattempts to escape, is postulated to represent 'behav­ioural despair' as an experiment al model of endogenousdepression (Porsolt et al., 1978; Vogel and Vogel, 1997;Bhattacharya et al., 1999).b. 'Learned helplessness' test: The rats were subjectedto footshock (60 scrambled sho cks, 15 s duration, 0.8rnA, every min) in a two-compartment jumping box(Techno) with the escape door to the adjoining unelec­trified 'safe' compartment closed. The exercise contin­ued for 1 h. 48 h later, the rats were subjected to avoid­ance training, using the same ap paratus, but keepingthe escape route open. During this avoidance trainingthe rats were placed in the electr ified chamber and al­lowed to acclimatize for 5 min before being subjectedto 30 avoidance trials, with an inter-trial interval of30 s. During the first 3 s of the trial, a buzzer stimulus(conditioned stimulus, e S) was sounded followed byelectroshock (unconditional stimulus DeS; 0.8 rnA) de­livered via the grid floor for the next 3 s. The avoidanceresponse was characterized by esca pe to the adjoining'safe' chamber during DeS within 15 s and was as­sessed as ' escape failure' , which is postulated to repre­sent despair or depression (Thiebot et al., 1992; Vogeland Vogel, 1997; Bhattacharya et al., 1999). WSG, orthe vehicle, was administered for 5 days before testingfor avoidance and escape, whereas IPM was adminis­tered once 30 min prior to the test.

Statistical analysis

The one -way analysis of var iance (ANOVA), followedposthoc by the Tuke y test , was used. A probability val­ue of < 0.05 was accepted as statistically significant dif­ference between the vehicl e-treated control and drug-

Forced swim-induced 'behavioural despair' test

WSG (20 and 50 mg/kg, p.o. ) induced a decrease in theduration of immobility by 30.4% and 44 .7%, similarto the reduction (54.5 %) produced by IPM (Table 5).

466 S. K. Bhattacharya et al.

Table 1. Effects of Withania somnifera glycowithanolides (WSG) and lorazepam (LZP) on the elevated plus-maze test in rats(data are means ± SEM).

Treatment Total time spent Total entries % Time spentGroups (mg/kg) n on both open and on both open on open

closed arms (s) and closed arms (s)arms (N)

Vehicle 16 219.6 ± 14.2 29.6 ± 3.8 19.6 ± 3.9WSG (20, p.o.) 10 229.2 ± 11.9 32.6 ± 4.4 38.2 ± 4.6*WSG (50, p.o.) 10 233.6 ± 15.4 38.4 ± 4.9 46.8 ± 3.2 *LZP (0.5, i.p.) 8 239.4 ± 16.2 36.4 ± 3.6 50.3 ± 3.8*

* P < 0.05 different from vehicle-treated group

% Entries onopen arms (N)

29.8 ± 3.640.4 ± 2.9*49.3 ± 3.8*54.4 ± 4.9*

Table 2. Effects of Withania somnifera glycowithanolides (WSG) and lorazepam (LZP) on the social interaction test in pairedrats (data are means ± SEM ).

Treatment groups(mg/kg)

VehicleWSG (20, p.o.)WSG (50, p.o.)LZP (0.5, i.p.)

n

10886

% Time spent insocial interaction

23.4 ± 4.238.9 ± 5.2"44.6 ± 4.9"49.2 ± 4.4*

n

20161612

Locomotor activity(counts)

185.2 ± 26.4204.0 ± 22.2174.2 ± 19.8138.4 ± 22.6

* P < 0.05 different from vehicle-treated group

Table 3. Effects of Withania somnifera glycowithanolides (WSG) and lorazepam (LZP) on the feeding latency test in rats in fa­miliar and unfamiliar environments (data are means ±SEM).

Treatmentgroups (mg/kg)

VehicleWSG (20, p.o.)WSG (50, p.o.)LZP (0.5, i.p.)

n

8666

Latency to feedin home cage (s)

46.4 ± 5.240.2 ± 6.344.9 ± 8.238.4 ± 6.2

n

10888

Latency to feed intest cage (s)

178.4 ± 19.8129.4 ± 12.2"108.4 ± 14.2*

94.4 ± 9.8"

* P <0.05 different from vehicle-treated group

Table 4. Effects of Withania somnifera glycowithanolides (WSG) and lorazepam (LZP) on rat brain tribulin activity (data aremeans ± SEM).

Treatment groups (mg/kg) n Brain weight (g) Tribulin activity (% inhibitionlg wet weight)

MAO A MAOBVehicle 8 1.72 ± 0.09 22 ± 6 34 ± 6WSG (20, p.o.) 6 1.82 ± 0.08 20 ± 4 29 ± 5WSG (50, p.o.) 6 1.74 ± 0.08 16 ± 6 26 ± 6LZP (0.5, i.p.) 6 1.64 ± 0.09 16 ± 4 24 ± 6Pentylenetetrazole (20, i.p.; PTZ) 8 1.77±0.1 46 ± 8" 58 ± 6*WSG (20)+ PTZ 6 1.80 ± 0.8 30 ± 6"" 38 ± 6"*WSG (50)+ PTZ 6 1.68 ± 0.7 24 ± 6** 32 ± 6""LZP + PTZ 6 1.77 ± 0.8 20 ± 4** 28 ± 5**

" P < 0.05 different from vehicle-treated group; "" P < 0.05 different from PTZ-treated group.

Anxiolytic-antidepress ant activity of Withania somnifera glycowith anolides: an experimental study 467

'Learned helplessness' test

WSG (20 and 50 mg/kg) produced a decrea se in escapefailures with concomitant incre ase in avoidance re­sponse, similar to the response elicited by IPM(Table 6).

Rotarod test

WSG (20 and 50 mg/kg, p.o. ), LZP and IPM had mar­ginal and statistically non-significant effects in thi s test.The performance deficit (n =10 rat s in each gro up)with WSG (20 and %0 mg/kg), LZP and IPM was 0,10,20 and 0, respectively.

Discussion

Experimental evaluation of psychotropic agents, in­clud ing putative anxiolytics and antidepressa nts, hasits limit ations since it is extremely difficult to developsuitable animal models of psychiat ric disorders. Thecriteria proposed by McKinney (1979) are virtually im­possible to achieve in the laboratory. Later guidelines(Barrett and Miczek, 1995) suggest that the animalmodels should have face and predictive validity so thatthey co uld correctly identify putat ive psychotropicagents . The use of more than a single test and the use ofsta nda rd agents for comparison; adds to the reliab ilityof the test procedure. The animal models of anxietyand depression used in this study have been subjectedto critical appraisal and extensively validated (Bhat­tach ar ya and Saryan , 1997 ; Vogel and Vogel, 1997;Bhattacharya et aI., 1999 ). Lorazepam, a widely usedanxiolytic agent, and imipramine, a well establishedantidepressant agent, were used as the standard drugsfor comparison.

WSG exhibited significant anxiolytic act ivity in allthe test procedures used, wh ich was comparabl e to thatshown by LZP. Thus, both the dru gs increased th enumber of entries and time spent on the open arms ofthe elevated plus-maze, increa sed active social interac­tion in previously isolat ed and then paired rats, and de-

creased the latency to feed in an unknown enviro n­ment. Likewise, both WSG and LZP decreased ratbrain tr ibulin activity when th e latter wa s increased byan anxiogenic agent, PTZ. Extensive experimental andclinical investigations have shown that tribulin mayfuncti on as an endocoid marker of anxiety and stress(Glover and Sandl er, 1993 ). Several anxiogenic agents,including PTZ, have been show n to increase rat braintribulin activity, assessed in terms of endogenous MAOA and B inhibito r activity (Bhattacha rya et al., 1996 ).WSG also exhibited significant an tidepressan t activity,compa rable to that sho wn by IPM, in the swim-in­duced behav ioural desp air and the learned helplessnesstest.

Neurochemical investigations on WS root extract in­dicate that it has significant GABA-mimetic act ivity inrodent brain. Thus, WS extract pr oduced inh ibit ion of(3H) GABA and (355) r-butylbicyclophosphorothionate(TBPS) binding, with concomitant increase in (3H ) flu­nitrazepam binding to their respective receptor sites inrat brain membranes. The extract also increased 36c1influx in mouse spina l cord neuron e preparation in theabsence of GABA. Th e increased influx was attenuatedby th e GABA receptor antagonists bicuculline and pi­crotox in (Mehta et al., 1991). This in vitro study wasconfirmed by the finding that the anticonvulsant effectof WS root extract in a lithium-pilocarpine model ofsta tus epilepticus invo lves a GABAergic mechanism(Kulka rni et al., 1998 ). However, receptor autoradi­ographic studies failed to demonstr ate any effect ofWSG on GABA receptor subtypes (Schliebs et al.,1997).

Anxiety and depression show important overlaps.Certain forms of chro nic anxiety disorders respondbest to treatment with antidepressants and anxiolyticsha ve proved useful as adjuncts in many depressives(Haefely, 1992). GABAergic mech an isms, wh ich formthe basis of action of several anxiolytics, including ben­zodi azepines, may be involved in ant idepressant activi­ty as well (Lloyd er aI., 19 89). The GABA-mim eticagent, pr ogabide, has been claimed to be as effective anantidepressa nt as imipramine. Likewise, its congener,

Table 5. Effects of Withania somnifera glycowithanolides(WSG) and imipramine (IPM) on the forced swimming-in­duced immobility ('behavioural despair' ) test in rats (data aremeans e SEM)

Table 6. Effects of Withania somnifera glycowithanolides(WSG) and imipramine (IPM) on the 'learned helplessness'test in rats (data are means: SEM)

Treatment groups (mg/kg) n Duration of immobility (s)Treatment groups(mg/kg) n

Escape failures(N)

Avoidanceresponse (N)

VehicleWSG (20, p.o.)WSG (50, p.o.)IPM (10, i.p.)

14 122:!: 16.810 78 : 8.4*10 62 : 6.9*

8 52: 5.4 "

VehicleWSG (20, p.o.)WSG (50, p.o.)IPM (10, i.p.)

12886

12.0 :!: 0.79.6 : 0.9 <'7.4 :!: 0.6 *8.4 :!: 0.7"

4.4:!: 0.46.4 :!: 0.5 *7.6 :!: 0.6*7.0 :!: 0.6*

" p < 0.05 different from vehicle-treated group " P < 0.05 different from vehicle-treated group

468 S. K. Bhattacharya et al.

fengabine, also a GABA-mimetic agent, has significantclinical antidepressant action (Lloyd et al., 1989). Lowplasma and CSF GABA levels have been reported in de­pressed patients with an ap parent correlation betweenGABA-CSF concentrations and clinical status (Lloyd etal., 1989). These clinical findings together with experi­mental evidence of perturbed GABAergic transmissionand the antidepressant activity of GABA-mimeticagents in animal models of depression, have led to theformulation of a GABA hypothesis of depression(Lloyd et al., 1989 ; Morselli and Garreau, 1991). Inview of these observat ions, it is possible that the anx i­olytic - antidepressant action of WSG is because of itsGABA-mimetic effect.

The present investigations supports the clinical ob­servations made with WS extracts and the use of theseextracts in Ayurveda for stabilization of disturbedmood.

References

Akbarsha M. A., Vijendrakumar, S., Kadalman i, B., Girija,R., Faridha, A.: Curative property of Withania somniferaDunal root in the context of carbendazim-inducedhistopathological changes in the liver and kidne y of ra t,Phytomedicine in press.

Barrett , ]. E., Miczek, K. A.: Behavio ural techniques in pre­clinical neuropsychophar maco logy research. In: Psy­chopharmacology: the fourth generation of progress(Bloom, F. E., Kupfer, D. ]. eds.), Raven Press, New York,1995, pp. 65-73 .

Bennett, D. A., Amrick, C. L., Wilson, D. E., Bernard, P. S.,Yokoyama, N., Liebman, j .: Behavioural profile of CGS9895, a novel anxiomodul ator with selective benzodi­azepine agonist and ant agonist properties. Drug Dev. Res.6: 313-325, 1985.

Bhatt acharya S. K.: Behavioural and physiological perturba­tions induced by chronic unpredictable footshock stress inrats . Effect of a standar dized ext ract of Withania somni fera(Ashwagandha). Proc. Int. Conf. Stress Adaptation, BoseInstitute, Calcutta, 1998, p. 39.

Bhattacharya, S. K., Satyan , K. S.: Experimental methods forevaluation of psychotropic agents in rodents. Anti-anxie tyagents. Indian J. Ex p. BioI. 35: 565-575, 1997.

Bhattacharya, S. K., Chakra barti, A., Sandler, M., Glover, V.:Effects of anx iogenic agents on rat brain monoamine oxi­dase (MAO) A and B inhibitory [tribulin) activity. Ind ian J.Exp . BioI. 34: 1190-1193, 1996.

Bhattacharya, S. K., Goel, R. K., Kaur, R., Ghosal, S.: Anti­stress activity of sito-indosides VII and VIII, new acyl­sterylglucosides from Witha nia somnifera. Phytother. Res.1: 32-37, 1987.

Bhattacharya, S. K., Kumar, A., Ghosal, S.: Effects of gly­cowithanolides fro m Withania somn ifera on an animalmodel of Alzheimer's disease and perturbed central cholin­ergic markers of cognition in ra ts. Phytother. Res. 9:110-113,1995.

Bhattacharya, S. K., Satyan, K. S., Rarnnathan, M.: Experi­mental methods for evaluation of psychotropic agents inrodents. II. Antidepressants. Indian j. Exp, Bioi. 37:117-123, 1999.

Bodnoff, S. R., Suranyi-Cadotte, B., Aitken, D. H. , Quirian,R., Meaney, M. ]. : The effects of chro nic antidepressanttreatment on an animal model of anxiety. Psychopharma­cology 95: 298-302, 1988.

File, S. E., Wyde, ] . R. G.: Can social interact ion be used tomeasure anxiety? Br.]. Pharmacol. 62: 19-24, 1978.

Ghosal, S., Lal, ] ., Srivastava, R., Bhattacharya, S. K., Upad­hyay, S. N., ]aiswal, A. K., Chatto padhyay, U.: Im­munomodulatory and CNS effects of sitoindosides IX andX, two new glycowithanolides from Withania somnifera.Phytother. Res. 3: 201-206,1 989.

Glover, V., Sandler, M.: Tribulin and isatin: an upd ate. In:Monoamine oxidase: basic and clinical aspects. (Yasuhara,H. , Parvez, S. M., Oguchi, K., Sandler, M. eds). VSP,Utrecht, 1993, pp. 61-71.

Haefely, W.: The role of GABA in anxiolytic/antidepressantdrug action. In: Experimental approaches to anx iety anddepression. (Elliot, ]. M. , Heal, D. ]., Marsden, C. A., eds.)John Wiley, London, 1992, pp. 151-1 68.

Kulkarni, S. K., George, B., Mathur, R.: Protective effect ofWithan ia somnifera root extract on electrographic act ivityin a lithium-pilocarpine model of sta tus epilepticus. Phy­tother Res. 12: 451-453, 1998.

Lloyd, K. G., Zivkovic, B., Scatton, B., Morselli, P. L.,Bartholoni , G.: The GABAergic hypoth esis of depression.Prog. Neuropsychopharmacol. Bioi. Psychiat. 13:341-351, 1989.

McKinney, W. T.: Behaviou ral models of depression in mon­keys. In: Animal models in psychiatry and neurology(Hanin, E., Usdin, E., eds.). Pergamon Press, Ox ford,1979, pp. 117-1 26.

Medvedev, A. E., Gork in, V. Z., Fedotova, I. B., Semiokhina,A. E, Glover, V., Sandler, M.: Increase of brain endogen ousmonoamine oxidase inhibitory activity (tribulin) in experi­mental audiogenic seizures in rats: evidence for amonoamine oxida se A inhibiting component of tribulin.Biochem. Pharmacal. 44: 1209-1210, 1992.

Mehta, A. K., Binkley, P., Gandhi, S. S., Ticku , M. K.: Phar­macological effects of Witha nia somnifera root extract onGABA receptor complex. Indian J. Med. Res. 94: 312-315,1991.

Morse lli, P. L., Garreau, M.: GABAergic drugs as putative an­tidepressants: a review of available data. In: Biological Psy­chiatry (Racagni, G., ed.). Elsevier, New York, 1991, pp .325-3 27.

Pellow, S., Chopin, P., File, S. E., Briley, M .: Validatio n ofopen closed arm entries in the elevated plus-maze measureof anxiety in the rat .]. Neu rosci. Meth. 14: 149-1 67, 185.

Porsolt , R. D., Anton, G., Blavet, N. , Jalfre, M.: Behaviouraldespair in rats: a new model sensitive to ant idepressivetreatments, Eur.]. Pharmacol47: 379-391, 1978.

Rege, N. N. , Thatte, U. M., Dahanukar, S. A.: Adapt ogenicpropert ies of six Rasayana herbs used in Ayur vedic medi­cine. Phytother. Res. 13: 275- 291, 1999.

Schliebs, R., Liebmann, A., Bhattacharya, S. K., Kumar, A.,Ghosal, S., Bigl, V.: Systemic administration of defined ex-

Anxiolytic-antidepressant activity of Withania somnifera glycowithanolides: an experimental stud y 469

tracts from Withania somn ifera (Indian Ginseng) and Shi­lajit differentiall y affects cholinergic but not glutamatergicand GABAergic markers in rat brain . Neurochem. Int . 30:181-190,1997.

Th iebot , M . H. , Martin, P., Puech, A. ]. : Animal behaviourstudies in the evaluation of antidepressant drugs . Br.]. Psy­chiat. 160 (Suppl. 15): 44-50,1 992.

Udup a, K. N., Singh, R. H.: Clinical and Experimental Stud­ies on Rasayana Drugs and Panchkarma Therapy. CentralCouncil for Research in Ayurveda and Siddha, New Delhi,1993.

Vogel, H. G., Vogel, N. H.: Drugs Discovery and Evaluation,Springer-Verlag, Berlin, Heidelberg, 1997, pp. 292-312.

Weiner, M. A., Weiner,].: Ashwagandha (Indian Ginseng). In:Herbs that heal, Quantum Books, Mill Valley, 1994, pp.70-72 .

Address

S. K. Bhattacharya, Department of Pharmacology, In­stitute of Medical Sciences, Banaras Hindu University,Varanasi - 221005, India.Tel: (05 42) 31 1099; Fax: (05 42) 31 6483;e-mail: salil@banaras.ernet.in