Pharmacological activities of Vitex agnus-castus extracts in vitro

Phytomedicine , Vol. 7(5), pp. 373-381 C Urban & Fischer Verlag 2000 http://www.urbanfischer.de!journalslphytomed

Pharmacological activities of Vitex agnus-castus extracts in vitro B. Meier-, D. Berger', E. Hoberg.', O. Sticher' and W. Schaffner! IDepartment of Pharmaceutical Biology, Institute of Pharmacy, University of Basel, Basel, Switzerland 2Zeller AG, Herbal Medicinal Products, Romanshorn, Switzerland 3Department of Applied BioSciences, Institute of Pharmaceutical Sciences, Swiss Federal Institute of Technology (ETH) Zurich, Zurich, Switzerland

Summary The pharmacological effects of ethanolic Vitex agnus-castus fruit-extracts (especially Ze 440) and various extract fractions of different polarities were evaluated both by radioligand binding studies and by superfusion experiments. A relative potent binding inhibition was observed for dopamine D z and opioid (11 and K subtype) receptors with IC so values of the native extract between 20 and 70 mg/mL. Binding, neither to the histamine HI' benzodiazepine and OFQ receptor, nor to the binding-site of the serotonin (S-HT) transporter, was significantly inhibited. The lipophilic fractions contained the diterpenes rotundifuran and 6~,7~-diacetoxy-13-hydroxy-Iabda-8,14-dien. They exhibited inhibitory actions on dopamine Dz receptor binding. While binding inhibition to 11 and K opioid receptors was most pronounced in lipophilic fractions, binding to () opioid receptors was inhibited mainly by a aqueous fraction. Standardised Ze 440 extracts of different batches were of constant pharmacological quality according to their potential to inhibit the binding to D z receptors. In superfusion experiments, the aqueous fraction of a methanolic extract inhibited the release of acetylcholine in a concentration-dependent manner. In addition, the potent D z receptor antagonist spiperone antagonised the effect of the extract suggesting a dopaminergic action mediated by D z receptor activation. Our results indicate a dopaminergic effect of Vitex agnus-castus extracts and suggest additional pharmacological actions via opioid receptors. Key words: Vitex agnus-castus, diterpenes, dopaminergic activity, opioid receptor binding inhibition

•

Introduction

Several clinical studies have confirmed the beneficial effect of extracts of the fruits of Vitex agnus-castus 1. in the treatment of premenstrual symptoms (PMS) (Coeugniet et al., 1986; Liebl, 1992; Lauritzen et al., 1997, Berger 1998, Schrader, 2000). It has previously been shown that a hypersecretion of prolactin resulting in a latent hyperprolactinemia may be responsible for PMS-symptoms (Halbreich et al., 1976; Carroll and Steiner, 1978; Jarry et al., 1991; Merz et al., 1996). The neurotransmitter dopamine was identified as a key factor that regulates prolactin synthesis via dopamine D 2 receptors (MacLeod, 1969). Extracts of Vitex agnus-castus have recently been shown to exert dopaminergic actions (Jarry et al.,

1994; Wuttke et al., 1995; Winterhoff, 1998). Radioligand binding studies with aqueous extracts indicated that this action is restricted to D 2 receptors (larry et al., 1994). Fractionation of such an aqueous extract into different lipophilic subfractions revealed the more lipophilic fractions to be the active ones (Wuttke et al., 1995). In this study, the interest was focussed to a ethanolic extract according to the proposal of the German Commission E and to various extract fractions of distinct lipophilicities of Vitex agnus-castus and their potential pharmacological actions. Both, radioligand binding and functional studies were performed in order to evaluate the proposed dopaminergic action of Vitex agnus0944-7113/00/07/05-373 $ 15.00/0

374

B. Meier et al.

mAu 2IXXl

1750

15:Xl

l2!iO

llXXl

10

20

so

30

eo

70

Fig. 1. HPLC-fingerprint-chromatogram of several batches tested of extract Ze 440. The fingerprints correspond from each batch to the other. Blue trace: V 7003; red: V 7002; green: V 7001; pink: V 6024; yellow V 23/95.

castus and to establ ish a pharmacological profile in vitro. The following receptors wer e selected for radio ligand binding studies: dopamine D 2, histamine H 1, benzodi azepine, opi oid and the binding-site of th e sero tonin (5-H T) transporter. Dop aminergic action s of Vitex agnus-castus extracts wer e studied by superfusio n exp eriments determining their inhibitory potential on acet ylcholine release in rat striatal slices. Beside dopamine D 2, binding to th e histamine 1-1 1 receptor was studied, since histamine is known to con trol neuroendocrine mechanisms including prolactin secre-

tion (Schwartz et al., 1991). Since PMS is frequently accompanied by sleep disorders, anxiety and depression, the effects on benzodiazepine and opioid receptor s as well as on the 5-HT transporter may be of interest. Benzodiazepines are kn own to overcome PM S related sympto ms, but th eir use is restricted due to the ir pot ential to ca use dep endency (Severino and Moline, 1995 ). Opioids, as well as a deficit of 5-HT, is further th ou ght to play a role in the pathophysiology of PMS (Sondheimer, 1994; Rapkin et al., 19 96 ).

Table 1. Chemical characterisation of the tested extracts (diterpenes and casticin in % of the dry extract). Batch No

Vitexilactone

6 ~, 7~-diacetoxy-

Rotundifuran

Casticin

DER native

13-hydroxylabda-8,14-diene V 2 3/95

0.4 78

0 .4 22

1. 042

0.677

22: 1

V 6024 V 7001 V 7002 V 7003

1.009 0.653 0.595 0.338

0.801 0.221 0.176 00408

2.225 1.584 10467 1.226

1.223 0.665 0.714 0.543

21:1 10:1 7:1 9:1

Pharmacological activities of Vitex agnus-castus extra cts in vitro

375

120 100

2

80

60 Fig. 2. Competition binding curve of 3H-spiperone binding to the dopam ine Dz receptor in the presence of various concentrations of the concentra ted extr act V23/95 derived fro m fruits of Vitex agnus-castus. Values repr esent mean and standard error of th ree tests perfor med in tripl icates.

40 20

o 0.0

Materials and Methods

10.0

100.0

1000.0

Extract Concentration (ug/ml)

concentrated extract V23 /95 . All ICso values are given in mg native extract per mL.

Plant material and extracts

The dry extract Ze 440 is the result of a standard ised pr ocedure. It includ es a mace ration of milled, dried fruits of Vitex agnus-castus with 60 % (VN ) eth an ol as a solvent in a rati o of 1:8 (first macerat ion ) and 1:5, respectively (second maceration). Th e solvent is evaporated to a concentrated extract with a dr y mass of appr ox. 40 % (V23 /95 with 29 %). Th e concent rated extr act is adsor bed to colloida l silica to produce a dry extract pr eparation with a content of approx. 50 % of the native extract (V23/95 with 25 % ). The dry extract preparations V23/95; V6024; V7001; V7002; V7003 ha ve been tested. The fingerprint chro matogram of the extracts are show n in Figure 1, th e qu antitative characterisation of the dr y extract prepa ra tion and the drug/extract-ratio (DER) of th e native extract is given in Table 1. For V23/95 and V6024 fru its with a relatively low amount of extractives have been used. Subsequently, fru its with a higher amo unt of extractives were selected. T herefore, the DER native decreased. No relevant influence of DER to th e chemical composition of the extracts has been observed. The ext ract is produced of dri ed fruit s with at least 0.08 % of casticin . The fru its were pu rchased fro m Germ an wholesalers Paul Miigge n burg (Alveslohe}, Ambro (Ham burg) and Chri stof Peter (Schebheim). Th e inte rna l batch-numbers have been PR-00 34 8-95, 13 7466 and 032867. Th e qu ality of the fruit s was in acco rdance to the mon ograph in HAB. The extract V23 /95 was tested as a con centrated extract (29 % dry weight ) and as a dry extract preparation. The screening was done with the

Bioguided fractionation

Dried frui ts of Vitex agnus-castus were purchased fro m Paul Mu ggenburg AG (Alveslohe, Germany. Internal batch-number PR-00 348-95 ). About 80 g of the powdered fruits were extrac ted for four times with 600 mL 80 % meth anol by turbo-extr action. The meth an ol extr act s were combined and evap or ated to dryness (10 mb ar, 30° C). Th e residue was solved in 200 mL 90 % methanol. Th e solution obta ined wa s partiti oned against 300 mL hexane three tim es. The hexane solutions were combined and evap orated to dr yness (300 mbar, 30° C). Th e 90 % methanol fraction wa s diluted with wat er to a 60 % solution (300 mL) and was partition ed aga inst 300 mL chloro form three times. Th e chlor oform solutions were combined and evaporated to dryness (450 mbar, 30° C). Th e 60 % meth an ol solut ion was evaporated so th at a aq ueous solutio n was o btai ned onl y. The wa ter solutio n was filled up with water to a volume of 300 mL. Th is was partition ed against 300 mL butyl alcohol three times. The butyl alcoh ol solutions were combined and evaporated to dryness (10 mbar, 30 ° C). The water fraction was evaporat ed to dryness as well (10 mb ar, 30° C). From this procedure, four frac tions, nam ely a hexane, a chlor oform, a butyl alcoho l and a wa ter fraction, were obtained. To increa se the yield and to det ect active prin ciples powdered frui ts (80 g, batch as before) were extra cted directly with 600 mL pure hex ane by turbo extraction. The obta ined extract was evaporated to dryness to yield 7.25 g of a residue (hexane extract), which was frac-

376

B. Meier et al.

120 Vl

____

a.>

100

0

0

e "ii-;;:

80

1:: ...... :.:::: 0

60

"'~

a.>

~ 0 a.> u

o -I:: ..c: u a.> ~

u

-a.>

a.> 0.

>., ....

40

Fig. 3. Inhibition of acetylcholine release by an aqueous Vitex agnus-castus extract. Acetylcholine release was electrically evoked in rat striatal slices preloaded with 3H-choline and superfused with Krebs buffer. Values represent mean and standard error of two independently performed tests.

o '-'

«

20 0 0.10

0.00

1.00

10.00

100.00

Concentration Aqueous Extract (ug/ml)

tionated by VLC on a silica gel column (6.5 x 22 em) and eluted with hexane followed by a gradient of EtOAc up to 100 %. Fraction H7 and H8 (hexane-erhylacetate 80:20) showed activity. For details and for further steps to identify diterpenes consult Hoberg (1999). The aqueous extract for the superfusion experiments was prepared from the concentrated extract V23/95. 3 g have been suspended in 60 mL of water. After centrifugation, filtration (0.2Ilm) and lyophilization, the residue was 5.5 % of the dry mass. Up to 10.5 mg have been dissolved in Krebs bicarbonate buffer for experiments.

tex agnus-castus as described elsewhere (Hoberg 1999, Hoberg et al. 1999).

Preparation of receptors for radioligand binding studies The sources of receptor material were: calf striatum (dopaminee D 2 receptor), male guinea pig cerebellum (histamine HI receptor), rat cortex (5-HT transporter and benzodiazepine receptor), stable CHO cell lines (opioid receptors). Brain tissues were homogenized in different buffers using a Polytron homogenizer: 50 mM Tris-HCI buffer pH 7.6, 0.1 % ascorbic acid, 120 mM NaCI, 5 mM KCI, 2 mM CaCI2 and 1 mM MgCI 2 (calf striatum); 50 mM phosphate buffer, pH 7.5 (guinea pig Compounds cerebellum); 50 mM Tris/HCI buffer, pH 7.4, 120 mM Aucubin and vitexin were obtained from Karl Roth NaCI, 5 mM KCI (rat cortex, serotonin transporter); GmbH&Co., Karlsruhe (Germany). Homoorientin, 15 mM Tris/HCI buffer pH 7.4, 118 mM NaCI, 4.8 isovitexin, luteolin-7-glucoside, and orientin were obtained from Extrasynthese, Genay (France). Casticin, mM KCI, 1.2 mM CaCI 2 , 1.2 mM MgCl 2 (rat cortex, vitexilactone, rotundifuran, 6~,7~-diacetoxy-13-hy­ benzodiazepine receptor). The homogenates were then droxy-Iabda-8,14-dien were isolated from fruits of Vi- centrifuged at 31000g for 10 min at 4°C. After a wash

Table 2. Receptors, Radioligands and Conditions used for Radioligand Binding Studies.

Receptor

Receptor Cone. (ug Protein)

Radioligand

Ligand Cone. (nM)

Displacer

Incubation (min; Temp.)

Dopamine D2 Histamine HI 5-HT transporter Benzodiazepine Opioid OFQ

120 630 600 188

3H-Spiperone 3H-Mepyramine 3H-Paroxetine 3H-Flumazenil

0.2 4.0 2.0 1.0

(+)-Butaclamole Promethazine Clomipramine Diazepam

90; RT 45;RT 60; RT 60; 0 °C

39 54 39 54

3H-Nociceptin 3H-Naloxone 3H-Naloxone 3H-Naltrindole

0.4 3.0 5.0 1.0

Orphanin-FQ Naloxone Naloxone Naloxone

60; RT 60; RT 60;RT 120; 37°C

Jl K

8

RT - Room Temperature

Pharmacological activities of Vitex agnus-castus extracts in vitro

377

Table 3. Inhibition of receptor binding by various extracts of Vitex agnus-castus expressed as ICso-values (pg/ml)", Dopamine D2

Histamine H 1

5-HT transporter

Benzodiazepine

Opioid OFQ

Concentrated extract (V23/95) Subfractions of a methanolic extract (80% VN) Water Butanol Chloroform Hexane Hexane extract Subfraction H7 Subfraction H8 1

I..l

K

0

200

36

22

194

300

200

> 300 > 300

> 200 > 200 > 200

70 24 36 22

52 22 35 12

58 > 200 > 200 195

52

>

370

500

>

1000

32

>

15 33 17

ICso-values represent triplicates of at least one experiment.

step (repeated homogenization and centrifugation) using the same buffer as described above, the pellets were stored at -80°C. Opioid receptors were expressed in transfected CHO cells and harvested as previously described (Reinscheid et al., 1995). Protein concentrations were determined by the Bradford method with bovine serum albumin as standard (Bradford, 1976). Radioligand binding assays

Receptor binding was conducted in triplicates in a total volume of 1 mL under the conditions summarised in Table 2. Extracts and compounds were added as 50 % ethanol solutions in a volume of 100 ul., Both total binding and non-specific binding also contained a final ethanol concentration of 5 %. Binding was terminated by rapid filtration with GF/C filters preincubated in 0.2 % polyethyleneimine under reduced pressure using a Brandel MR 48 cell harvester. The filters were washed 4 times with 5 mL cold Tris/HCI buffer pH 7.4 and air dried. Radioactivity on filters was determined by liquid scintillation counting (Betamatic I, Kontron). Specific binding of various extract concentrations was calculated by determining the difference between total binding and non-specific binding in the presence of the corresponding extract concentration. Extract concentrations that inhibited 50 % of the binding (IC50 ) were determined by curve fittings using the Sigma Plot program.

For acetylcholine release experiments rats were pretreated with reserpine (2.5 mg/kg s.c.) 12 and 17 hours before sacrifice in order to deplete endogenous dopamine. Striatal slices were incubated for 30 min at 30 )lC in Krebs-bicarbonate buffer containing 0.16 u.M 3H-choline (80 Ci/mMol) and superfused with Krebs buffer containing 10 u.M hemicholine in order to prevent reuptake of choline formed by hydrolysis of released acetylcholine. For noradrenaline release experiments rat occipital cortex slices were labelled by preincubation for 30 min at 30 °C in Krebs buffer containing 0.1 }..IM 3H-noradrenaline (54 Ci/mMol) and 500 llM ascorbic acid. Collection of 5 min fractions (6 mL) began after 60 min of superfusion. The slices were stimulated by electric pulses (2 ms, 12 rnA) twice for 2 min after 75 min (51) and 150 min (52) of superfusion. Stimulation frequencies were 2 and 5 Hz in the experiments with 3H-choline preloaded rat striatal slices and 3H-noradrenaline preloaded rat cerebral cortex slices, respectively. The overflow of tritium evoked by 51 was approximately 10 and 2 % of total tissue tritium in 3H_ choline and 3H-noradrenaline preloaded slices, respectively. Test extracts were added 30 min before 52 and

Table 4. Comparison of different batches of Vitex agnus-castus extracts on dopamine D2 receptor binding expressed as ICso-values (ug/ml) of the native extract (n = 3).

Superfusion experiments in brain slices

Superfusion experiments were performed as described earlier (Markstein et al., 1987). Briefly, male rats (OFA strain) were sacrified by decapitation, the brains rapidly removed and dissected. Disk-shaped slices with a diameter of 3 mm and a thickness of 0.3 mm were prepared from rat striatum or rat occipital cortex and loaded with tritium labelled precursor or transmitter.

Dopamine D2 Batch-Number V23/95 V6024 V7001 V7002 V7003

40 ± 61 ± 65 ± 66 ± 69 ±

8

10 7

10 12

378

B. Meier et a1.

Table 5. Inhibition of dopamine D2 receptor binding by commercially available and purified compounds expressed as ICso-values (pg/ml)I.

Table 6. Antagonism by spiperone or atropine of the acetylcholine release induced by V. agnus-castus ' : Release of acetylcholine

Dopamine D2 pg/ml

Aucubin Casticin Homoorientin Isovitexin Linoleic acid Luteolin-7-glucoside

release e SE

(%, relative to control)

nmo l/ml

> 300

Control Vitex agnus-castus (70 pg/ml) " + Spiperone (3.10-9 M) " + Spiperone (3.10- 8 M ) " + Spiperone (3.10-7 M )

100 40:!: 1.4 44:!: 1.1 60:!: 7.2 69 :!: 2.5

6 ~, 7~-diacetoxy- 13 -hydroxy-

300 > 300 > 300 40:!: 12 > 300 79 :!: 12

labda-8,14-dien Orientin Rotundifuran Vitexilactone Vitexin

> 300 45:!: 7 > 300 > 300

Vitex agnu s-castus (70 ug/ml) " + Atropine (3.10- 9 M ) " + Atropine (3-10- 8 M ) " + Atropine (3-10- 7 M )

31 :!: 5.0 26:!: 8.6 26:!: 4.7 56:!: 9.2

1

>

143 194

124

ICso-values represent triplicates of at least one experiment.

were pre sent in the medium between 120 and 170 min of superfusion. Tritium overflow is expressed as fractional rate per min. Th e effects of extracts on tritium overflow are expressed as the ratio of tritium overflow evoked by 52 to that by 51. Analyses

Casticin, diterpenes and agnuside have been analysed according to earlier published methods (Hoberg et ai, 2000 a-c, Hoberg 1999). The HPLC-fingerprint in Figure 1 shows the result of a RP-C 18 -analysis after a linear gradient elution with changing steepness (solvent A methanol, solvent B 0,5 % ph osphoric acid in water. Initi al: 5 % A; 0-15 min: 5-27 % A; 15-20min: 2735% A; 20-26 min: 35 - 52% Aj 26-33min: 52-80 % A; 33 -37min:80- 90% A. Post run: rinsing with 100% B). Detection wa velength 254 nm . The extracts ha ve been dissolved therefore in 80 % (VN) methanol (30 mg/mL).

Results Radioligand binding studies

Based on the positive screening results obtained with the concentrated extract V2 3/95 (see Table 3) and the findings of earlier research, the current experiments have been mainly focused onto the binding on the dopamine D z receptor sites. The comparison of five, chemically well characterised, batches of the dry extr act preparation Ze 440 on the inhibition of the binding of 3H-spiperone to th e dopamine D z receptor

1 Striatal slices were preincubated with 3H-choline superfused with Krebs buffer and stimulated twice by electric pulses in the absence (SI) and presence of extract (S2). The extract was added 30 min before S2 alone or together with various concentrations of spiperone or atropine. Relative inhibition rates were calculated by comparing the S1IS2 ratios.

showed similar inh ibition potencies with IC so values between 40 and 70 ug/ml. , The results are given in Tabl e 4. Chemical (H PLC-Fingerprint, Figure 1) and ph armacological results support the standardisation con cept of the extract. A typical dos e-/response curve is show n in Figure 2. N o loss of activity ha s been observed after the dr ying pr ocess: The con centrated extr act and the dr y extract prep aration of batch V23/95 showed the same activity. To find some mole cular principles responsible for the pharmacological effect, a bioguided fractionation has been undertaken. First, subfractions of a methanolic (80 % ) extract were prepared. Hexane, chloroform, butanol and water subfractions were obtained by partition and studied for their inhibitory effect on the binding to D z receptors. Onl y the hexane extract inhibited D z binding with an IC so of 32 ug/rnl., whereas the others did not show an y inhibition up to 300 ug/ml, (Table 3). Chromatographic analysis of the fractions revealed th at the hexane fraction contained fat ty acid s and diterpenes. The lipophilic flavonoids casticin and penduletin were present in th e chloroform extract, whil e th e butanol extract contained the iridoids agnuside and aucubin as well as p-hydr oxyb enzoic acid. In order to identify th e diterp enes exhibiting inhibitory binding effects to D z receptors, dried fru its were directly extracted with hexane and studied. Th is extract potently inhibited D z binding with an IC so value of 1511g/mL (Ta ble 3 ). Binding studies with subfractions of this extract obtained by subsequent vacuum liquid chromatography on a NP silica gel column resulted in two active fractions (H7 and H8) exhibiting

Pharmacological activities of Vitex agnus-castus extracts in vitro similar inhibitory potencies as the hexane extract (Table 3). Chemical and chromatographic analysis revealed that these fractions contained the known diterpene rotundifuran and the unsaturated fatty acid linoleic acid. In addition, vitexilactone and a new diterpene 6p,7p-diacetoxy-13-hydroxy-labda-8,14-dien were isolated from two further fractions without activity but characteristic spots in TLC-control. All compounds have been identified by lD-, 2D-NMR and MS-spectroscopy as described earlier (Hoberg et al., 1999, Hoberg 1999). When subjected to radioligand binding studies two of these diterpenes (vitexilactone was inactive) exhibited inhibitory actions with IC so values of 79 (6P,7p-diacetoxy-13-hydroxy-Iabda-8,14-dien) and 45 (Rotundifuran) ug/ml., In addition, linoleic acid showed a similar potent binding inhibition as the diterpenes, whereas other characteristic constituents of Vitex agnus-castus such as agnusid, aucubin, casticin and other flavonoids did not show any inhibitory effect at all (Table 5). Inhibitions to the histamine H 1, benzodiazepine and opioid receptors as well as to the 5-HT transporter binding are summarized in Table 3. The concentrated extract V23/95 did not significantly inhibit binding to histamine H 1, benzodiazepine and OFQ receptor nor to the 5-HT transporter. However, inhibition of the binding to the u, K and 8 opioid receptor with IC so values of 36, 22 and 194 ug/ml., respectively was observed. While all extract fractions showed similar effects on II and K subtypes, the aqueous extract was most effective on the 0 subtype. Investigations with pure compounds on the opioid receptors are planned. In order to determine the relative stability of the extract, the same experiments on II and K receptors were repeated one year later. Similar results (15 and 20 ug/rnl, for II and K receptor) were obtained, confirming the pharmacological stability of the extract. Superfusion experiments

Inhibition of acetylcholine release The effect of the aqueous fraction of the concentrated extract Ze 440 (V23/95) on functional D 2 receptors was studied by measuring its effect on acetylcholine release from rat striatal slices. In this test, the aqueous extract inhibited electrically evoked tritium overflow from 3H-choline preloaded rat striatal slices in a concentration-dependent manner with an IC so value of approximately 30 ug/ml, (Figure 3). In addition, the potent D 2 receptor antagonist spiperone antagonized the effect of the extract in a concentration-dependent manner, whereas the muscarinic receptor antagonist atropine failed to exert an antagonistic action up to 0.1

379

u.M (Table 6). Atropine antagonized the effect of the extract only at the highest concentration tested (l11M).

Stimulation of noradrenaline release The effect of the same extract on functional u 2-adrenoceptors was investigated by measuring its stimulating effect on noradrenaline release from rat occipital cortex slices. The aqueous extract failed to enhance electrically evoked tritium overflow from 3H-noradrenaline preloaded rat cerebral cortex slices up to a concentration of 100 llg/mL.

•

Discussion

Radioligand binding studies with various extracts of Vitex agnus-castus on dopamine D 2 receptors confirmed the previously described effects of an aqueous extract on dopamine receptors (larry et al., 1994; Wuttke et al., 1995). Binding inhibition of the ethanolic extract Ze 440 and a special hexane extract were relatively potent with IC so values of 52 and 15 ug/ml., respectively. Subsequent fractionation of the hexane extract by vacuum liquid chromatography resulted in only few fractions exhibiting inhibitory actions. Rotundifuran, 6p,7p-diacetoxy-13-hydroxy-Iabda-8,14-dien and the unsaturated fatty acid linoleic acid have been isolated and showed similar ICso-values like the extracts. Linoleic acid has recently been proven to inhibit binding to various receptors in a non-competitive manner (Ingkaninan et al., 1999). This compound has therefore most probably no pharmacological relevance onto the doparnine-Dj-effect. The effect of the diterpenes on doparnine-Dj-receptors has been confirmed at the same time by another working group (Christoffel et al., 1999; Jarry et al., 1999) with similar results. The observed low inhibitory potential of the concentrated extract V23/95 to histamine Hj, benzodiazepine and OFQ receptors as well as to the 5-HT transporter indicated that none of these receptors are direct mediators of the pharmacological action of Vitex agnus-castus. Bindings to opioid receptors were relatively potently inhibited by the extract and the fractions of different polarity showing the most potent inhibition to II and K subtypes with even lower IC so values than observed for D 2 receptors. Interestingly, active compounds were not exclusively concentrated in the hexane fraction as it was the case for D 2 receptors, but were present in all subfractions (aqueous, butanol, chloroform and hexane fractions), Electrically evoked tritium overflow from rat striatal slices pre loaded with 3H-choline has been shown to reflect neuronal release of acetylcholine and to be regu-

380

B. Meier et al.

lated in an inhibitory manner by Oz receptors (Stoof et al., 1979). Extracts of Vitex agnus-castus inhibited acetylcholine release in rat striatal slices in a concentration-dependent manner indicating a dopamine Oz agonistic action. In addition, the potent Oz receptor agonist spiperone antagonized the effect on acetylcholine release of the extract confirming a dopaminergic action mediated by Oz receptor activation. To what extent the observed weak cholinergic action may be valid for Vitex agnus-castus, needs to be studied in more detail, since the muscarinic receptor antagonist atropine antagonised the effect on acetylcholine release only at a high concentration of 111M. This is in agreement with the suggested dopaminergic inhibition of prolactin secretion in rat pituitary cell cultures (larry et al., 1991) The functional effect of Vitex agnus-castus on (Xzadrenoceptors was assessed by measuring its effect on stimulated noradrenaline release in rat occipital cortex slices. Noradrenaline release in rat cerebral cortex slices is regulated in an inhibitory manner by autoreceptors with the pharmacological properties of (Xzadrenoceptors (Starke, 1980). Blockade of these receptors prevents feedback inhibition resulting in enhanced transmitter release. Electrically evoked tritium outflow from rat occipital cortex slices preincubated with 3H_ noradrenaline has been shown to reflect action potential-induced release of the transmitter (Taube et al., 1977). The aqueous extract failed to enhance electrically evoked tritium overflow from preloaded rat cerebral cortex slices up to a concentration of 100 ug/rnl, indicating the absence of an antagonistic action at the autoreceptor. In conclusion, both radioligand binding studies and functional experiments indicated that extracts of Vitex agnus-castus exhibited dopaminergic actions mediated by Oz receptors. Fractionation of the extract revealed lipophilic constituents such as diterpenes being responsible for the observed effect to Oz receptors. Additional pharmacological actions may be mediated by opioid receptor 11 and K subtypes according to their relative potent inhibitory binding effects in vitro. Acknowledgement

We thank Novartis Consumer Health S.A., CH-3007 Berne and Zeller AG, Herbal Medicinal Products, CH-8590 Romanshorn for the financial and scientific, Roy Upton, American Herbal Pharmacopeia, Santa Cruz CA and Urs Simmen, University of Basel for editorial support. Receptor binding studies on opioid receptors were carried out in cooperation with Dr. R. Reinscheid, Hoffmann-La Roche AG, Basel, Switzerland. Superfusion experiments were performed in collaboration with C. Kohler and Dr. R. Markstein, Novartis AG, Basel, Switzerland.

References Berger, D.: Vitex agnus-castus: Unbedenklichkeit und Wirksamkeit beim pramenstruellen Syndrom, Wirkprinzipien und Wirkmechanismen eines neu entwickelten Extraktes. Thesis University of Basel, 1998. Bradford, M. M.: A rapid and sensitive method for the quantification of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72: 248-250, 1976. Carroll, B. l. and Steiner, M.: The psychobiology of premenstrual dysphoria: The role of prolactin. Psychoneuroendocrinol. 3: 171-180, 1978. Christoffel v., Spengler B., larry H., Wuttke W: Prolactin inhibiting dopaminergic activity of diterpens from Vitex agnus-castus in Leow D., Blume H., Dingermann Th.: Phytopharmaka V, Forschung und klinische Anwendung. Steinkopf-Verlag, Darmstadt, pp. 209-214, 1999. Coeugniet E., Elek E., Kuhnast R.: Das pramenstruelle Syndrom (PMS) und seine Behandlung. Arztezeitschrift fur Naturheilverfahren 27: 619-622, 1986. Halbreich, U., Assad, M., Ben-David, M. and Bornstein, R.: Serum-prolactin in women with premenstrual syndrome. Lancet 2: 654-656, 1976. Hoberg E.: Phytochemical and analytical investigations on Vitex agnus-castus L. Thesis No. 13285, Swiss Federal Institute of Technology, Zurich, 1999. Hoberg E., Meier B., Sticher 0.: Quantitative high performance liquid chromatography analysis of diterpenoids in Agni-casti fructus. Planta med. 66: 352-355, 2000a. Hoberg E., Meier B., Sticher 0.: A new analytical HPLC method for the determination of agnuside and p-hydroxybenzoic acid contents in Agni casti fructus. Phytochemical Analysis, in press, 2000b. Hoberg E., Meier B., Sticher 0.: Quantitative High Performance Liquid Chromatographic Analysis of Casticin in the Fruits of Vitex agnus-castus. Pharmaceutical Biology, submitted, 2000c. Hoberg, E., Orjala, J. E., Meier, B. and Sticher, 0.: Diterpenoids from the fruits of Vitex agnus-castus, Phytochemistry 52: 1555-1558, 1999. Ingkaninan, K., von Frijrag Drappe Kiinzel, ]. K., Ijzerman, A. P. and Verpoorte, R.: Interference of linoleic acid fraction in some receptor binding assays. ]. Nat. Prod. 62: 912-914, 1999. larry, H., Leonhardt, S., Wuttke, W., Behr, B. and Gorkow, c.: Agnus-castus als dopaminerges Wirkprinzip in Mastodynon N. Z. Phytother. 3: 77-82, 1991. larry H., Leonhardt S., Wuttke W., Spengler B., Christoffel v.: Auf der Suche nach dopaminergen Substanzen in Agnicasti-fructus-Praparaten: Warum eigentlich? in Meier B, Hoberg E (Editors): Agni-casti-fructus: Neue Erkenntnisse zur Qualitat und Wirksamkeit. Z. Phytother. 20: 150-153, 1999. larry, H., Leonhardt, S., Gorkow, C. and Wuttke, W.: In vitro prolactin but no LH and FSH release is inhibited by compounds in extracts of Agnus-castus: direct evidence for a dopaminergic principle by the dopamine receptor assay. Exp. Clin. Endocrinol. 102: 448-454,1994.

Pharmacological activities of Vitex agnus-castus extracts in vitro Lauritzen CH., Reuter H.D., Repges R., Bohnert K.]., Schmidt D.: Treatment of premenstrual tension syndrom with Vitex agnus-castus - controlled double blind study versus pyridoxine. Phytomedicine 4: 183-189, 1997. Liebl A.: Behandlung des priimenstruellen Syndroms - Agnus-castus-haltiges Kombinationsarzneimittel im Test. TW Gynakologie 5: 147-154, 1992. Markstein, R., Enz, A., Vigouret,]. M., laton, A., Closse, A., Briner, U. and Gull, P.: Biochemical, behavioural, and endocrine effects of CK 204-933, a novel 8-~-ergolene. ]. Neural. Transm. 69: 179-199, 1987. MacLeod, R. M.: Influence of norepinephrine and catecholamine depleting agents on the synthesis and release of prolactin and growth hormone. Endocrinol. 85: 916-923, 1969. Merz, P. G., Gorkow, C, Schrodter, A., Rietbrock, S., Sieder, C, Loew, D., Dericks-Tan,]. S. E. and Taubert, H. D.: The effects of a special Agnus-castus extract (BP1095E1) on prolactin secretion in healthy male subjects. Exp, Clin. Endocrinal. Diab. 104: 447-453, 1996. Rapkin, A. ]., Shoupe, D., Reading, A., Daneshgar, K. K., Goldman, L., Bohn, Y, Brann, D. W. and Wahesh, V. B.: Decreased central opioid activity in premenstrual syndrome: luteinizing hormone response to naloxone. ]. Soc. Gynecol. Invest. 3: 93-98, 1996. Reinscheid, R. K., Nothacker, H. P., Bourson, A., Ardati, A., Henningsen, R. A., Bunzow,]. R., Grandy, D. K., Langen, H., Monsma, F.]. and Civelli, 0.: Orphan FQ: A neuropetide that activates an opioidlike G-protein coupled receptor. Science 270: 792-794,1995. Schwartz, ]. C, Arrang, ]. M., Garbarg, M., Pollard, H. and Ruat, M.: Histaminergic transmission in the mammalian brain. Physiol. Rev. 71: 1-51, 1991. Schrader E.: Pre-menstrual syndrome treatment with Agnuscastus extract: A randomised, placebo-controlled study. In preparation.

381

Severino, S. K. and Moline, M. L.: Premenstrual syndrome identification and management. Drugs 49: 71-82,1995. Sondheimer, S. ].: Etiology of premenstrual syndrome. In: Smith S. and Schiff 1. (eds) Modern management of premenstrual syndrome. Norton Medical Books New York, pp 46-53, 1994. Starke, K.: Presynaptic receptors and the control of noradrenaline release. Trends. Pharmacal. Sci. 13: 268-271, 1980. Stoof,]. C, Thieme, R. E., Vrijmoe-de Vries, M. C and Mulder, A. H.: In vitro acetylcholine release from rat caudate nuclues as a new model for testing drugs with dopamine receptor activity. Naunyn Schmiedebergs Arch. Pharmacal. 323:298-306,1979. Taube, H. D., Starke, K. and Borowski, E.: Presynaptic receptor systems on the noradrenergic neurons of rat brain. Naunyn Schmiedebergs Arch. Pharmacal. 299: 123-141, 1977. Wuttke, W., Gorkow, C H. and larry, H.: Dopaminergic compounds in Vitex agnus-castus. In: Loew D. and Rietbrock, D.: Phytopharmaka: Forschung, Klinik und Anwendung, Steinkopf-Verlag Darmstadt, Vol 1, pp: 81-90, 1995. Winterhoff, H.: Vitex agnus-castus (Chaste Tree): Pharmacological and clinical data. In: Lawson L. D. and Bauer R. (eds): Phytomedicines of Europe, Chemistry and Biological Activity. American Symposium Series 691, Amer. Chern. Soc., Washington, DC, pp 299-308, 1998.

Address B. Meier, Zeller AG, Herbal Medicinal Products, CH8590 Romanshorn 1. Tel.: 0041714660571; Fax: 0041 71 4660550; e-mail: [email protected]