Ginsenoside RG1 and Corticosteroid Receptors in Rat Brain Rudolf

Endocrinol. Japon. 1987, 34 (2), 213-220

Ginsenoside RG1 and Corticosteroid Receptors in Rat Brain

E. R. DE KLOET, J. M. H. M. REUL, F. R. VAN DEN BOSCH, J. A. D. M. TONNAER1, AND H. SAITO2

Rudolf Magnus Institute for Pharmacology, Medical Faculty, University of Utrecht, Vondellaan 6, NL-3521 GD Utrecht, The Netherlands and 1) Scientific Development Group, ORGANON International, Oss, The Netherlands and 2) Experimental Station for Medicinal Plant Studies, Faculty of Phar- maceutical Sciences, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo, 113 Japan

Abstract

Old (28 months) male Wistar rats were treated chronically for two weeks with ginsenoside Rgi or with vehicle delivered via sc implanted Alzet minipumps (rate of ginsenoside release 2.4 ƒÊg/0.5 pl/h). The number of Type 1 corticosterone-preferring receptor sites (CR) and Type 2 glucocorticoid receptors (GR) was measured in the cytosol of hippocampus tissue of rat brain with an in vitro binding assay. In old rats the B. of Type 1 CR and Type 2 GR was reduced by 51.5% and 28.3% respectively. Following the two week treatment with Rgi the B. of Type 1 CR increased by 60% and a receptor concentration was reached which was 21% lower than that observed in the young control animals. Minor differences in affinity of steroid binding to both receptor systems were observed in the groups of rats. The possible binding of ginsenosides to brain corticosteroid receptors in vitro was investigated as well. The inclusion of a 500 fold molar excess of Rgi in hippocampus cytosol did not displace 3H-corticosterone from its soluble receptor sites. The affinity of Rgi with these sites in vitro is therefore negligible. In conclusion, the binding capacity of Type 1 CR and Type 2 GR is reduced in the hippocampal brain region of aged rats. Upon chronic infusion of ginsenoside Rgi, only Type 1 CR capacity is restored towards the level observed in young control animals. This finding suggests that in old rats the ginsenoside enhances the CORT signal via Type 1 CR on the function of the hippocampus, which is a limbic brain structure involved in cognition, mood and affect.

Two distinct corticosteroid receptor sys- Veldhuis et al., 1982; Beaumont and Fane- tems have been identified in rat brain i.e. stil, 1983; Reul and de Kloet, 1985; Fun- Type 1 (mineralocorticoid-like) and Type 2 der, 1986). The Type 1 receptors can be (glucocorticoid) receptors (de Kloet et al., divided into two subtypes (de Kloet and 1975; Moguilevski and Raynaud, 1980; Veldhuis, 1984; Fun der, 1986; McEwen et al., 1986a). One subtype is the Type 1 Received July 2, 1986 CORT preferring receptor (CR), which re- Endocrinol. Japon. 214 DE KLOET et al. April 1987 sponds with stringent specificity to CORT. component of the ginseng root (Panax gin- The other subtype is the Type 1 mineralo- seng C. A. Meyer) (Takagi, 1974), selectively corticoid receptor (MR), which responds to increases the binding capacity of this Type ALDO in the regulation of salt appetite 1 CR towards the level observed in young (McEwen et al., 1986b). Type 1 CR is control animals. almost exclusively localized in hippocampus neurons (McEwen et al., 1968; Reul and de Kloet 1985; 1986). Type 1 CR is in- Materials and Methods volved in the regulation of aspects of adaptive behaviour (Bohus and de Kloet, 1981) Animals and surgery and of hippocampus-associated neurotrans- Old (28 months) and young (3 months) male mission (de Kloet et al., 1986a; McEwen et Wistar rats were supplied by TNO, Zeist, The al., 1986a). Type 1 CR is 80% or more Netherlands. The animals were housed singly occupied by endogenous CORT under all in cages under standard lighting (lights on from conditions of intact adrenocortical function. 6.00-20.00 h) and temperature conditions (23•Ž). Type 1 CR mediates an activating and Food and water were given ad libitum. Six old rats were treated for two weeks with the gin- synchronizing influence of CORT on cir- senoside Rgi and a group of six old rats and cadian-driven brain processes and contributes one of six young rats were treated with vehicle. to proper expression of the stress response Alzet minipumps (model 2002) were used to (de Kloet and Reul, in press). deliver the ginsenosides. The minipumps were implanted sc under anesthesia with Nembutal

The Type 2 glucocorticoid receptors (40 mg/kg body weight). The rate of release of (GR) have a widespread localization in ginsenoside Rgi was 2.4ƒÊg/0.5ƒÊl/h for 14 days. neurons and glial cells as shown by auto- The vehicle was 0.9% saline containing 1.7% radiography of in vitro labelled tissue sec- (v/v) ethanol. tions (Reul and de Kloet, 1985; 1986) and Twenty-four hours prior to killing, the animals by immunocytochemistry (Fuxe et al., 1985). were bilaterally adrenalectomized (ADX) via the dorsal approach between 9.00 and 10.00 am un- Type 2 GR is occupied in parallel with der ether anesthesia in order to deplete endo- rising plasma CORT levels during circadian genous corticosteroids. The rats were anes- variation and after stress. Type 2 GR thetized with Nembutal before killing and were mediates the glucocorticoid feedback on then perfused with ice-cold saline through the stress-activated brain processes (de Kloet heart (about 25 ml per rat). A blood sample and Reul, in press; Reul and de Kloet, was drawn for measurement of the plasma CORT 1985). level by a radioimmunoassay procedure (Veld- huis et al., 1982). All rats had a plasma CORT Previous studies have demonstrated that level below 0.2ƒÊg/100 ml. The antiserum to the senescent rat has significantly fewer CORT was kindly provided by Dr. Th. J. Ben- CORT binding receptors in the hippocampus raad, University of Nijmegen, The Netherlands. The hippocampus was dissected on ice (Gispen (Rigter et al., 1984; Sapolski et al., 1983; et al., 1972) and stored until receptor assay at -80℃ 1986) than the young animals. The total . number of CORT receptors (Type 1 CR+ Type 2 GR) was determined in these pre- Receptor binding assays vious studies. We showed in the present The hippocampal tissue of six rats was pooled study that the reduction in CORT binding and homogenized in 5 mM Tris buffbr(500μl/ capacity of the senescent rat occurs in Type lobe) containing 1 mM EDTA, 1 mM 2-mercap- toethanol, 10 mM sodium molybdate and 5% 1 CR as well as in Type 2 GR. In addi- glycerol adjusted to pH 7.4 with hydrochloric tion, we show that chronic infusion of the acid (Reul and de Kloet, 1985). The homogenate ginsenoside Rgi, which is a stimulatory was centrifuged for 1 h at 2•Ž at 100 000 gay for Vol.34, No.2 GINSENG AND BRAIN CORTICOSTEROID RECEPTORS 215

preparation of cytosol. Cytosol was added to corrected for remaining traces of CBG. Type 2 previously evaporated 3H-steroids. Type 1 GR was measured with 3H-RU 28362. B. CORT-preferring sites (CR) were measured with and KD were determined from Scatchard and 3H-corticosterone (spec . act 105.0 Ci/mmol, New Woolf analysis of the binding data (Keightley et England Nuclear, Boston, USA) in a concentra- al., 1983). tion range from 0.1 to 20 nM labelled steroid in the presence of a 500 fold molar excess of RU 28362, which completely prevents the bind- Results ing of 3H-CORT to the Type 2 glucocorticoid receptors (GR) (Philibert and Moguilevski, 1983). Figure 1 depicts the Scatchard analysis Non-specific binding was determined with a 500 fold molar excess of unlabelled CORT and of 3H-CORT binding to Type 1 CR. The specific binding was corrected for traces of con- binding to Type 2 GR was blocked by ex- taminating blood transcortin (CBG), that might cess RU 28362. The Bmax of Type 1 CR possibly have remained after perfusion. Correc- was 51.5% lower in cytosol of hippocampus tion for CBG was done after the determination tissue of the old than that of the young of non-specific binding in the presence of a 500 rats. The old rats responded to treatment fold molar excess of dexamethasone (DEX). with Rgi with a 60% increase in Type 1 Type 2 GR were measured by 3H-RU28362 CR number, reaching a Bmax 21.2% below (spec. act 78 Ci/mmol, Roussel-Uclaf, Romain- ville, France) binding using a concentration range that observed in the young animals (Table from 0.5 to 20 nM in the absence or presence 1). Figure 1 also shows the Scatchard of a 500 fold molar excess of unlabeled RU analysis of 3H-RU 28362 binding to Type 2 28362 to correct for the non-specific binding. GR. The binding capacity of Type 2 GR Incubation was carried out for 3 h at 0•Ž (de in the old rats was 28.3% lower than in Kloet et al., 1975; Veldhuis et al., 1982). Separa- the young rats and was not significantly tion of bound from unbound 3H-steroids was changed (p<0.05) upon Rg1 infusion. The performed by Sephadex LH20 gel filtration and the binding data were expressed as fmoles/mg changes in the binding affinity were small. cytosol protein. A separate series of experi- Only in the case of the Type 1 CR, the ments served to measure the displacement of old animals had a slightly higher KD for 3H-CORT binding to receptors in the hippo- 3H-CORT binding , which was restored to campus cytosol by various saponins purified that observed in the young animals after from the ginseng root (Panax-ginseng C. A. the Rg1 treatment (Table 1). Meyer). Incubation of 2 nM 3H-CORT was Table 2 summarizes the results of the performed in the presence of increasing concentrations of CORT in a range from 1 to 1000 nM experiments designed to analyze the affinity as well as with 1000 nM concentrations of the of various ginsenosides with the CORT ginsenoside components Rg1, Rb1, Rd, Re, and binding receptor types. Displacement of a C3 (Saito et al., 1984). The non-specific binding 2 nM 3H-CORT concentration by the gin- of 3H-CORT was determined after inclusion of senosides was studied. Thus Table 2 shows a five hundred fold excess of unlabelled CORT. that a 500 fold molar excess (1 pMol) of CORT, DEX and ALDO were kindly donated Rg1, Rb1, Rd, Re and C3 was not able to by Organon, Oss, The Netherlands and RU 28362 was a gift of Roussel-Uclaf, Romainville, compete for 3H-CORT binding to the mixed France. population of Type 1 CR and Type 2 GR. The 2 nM 3H-CORT concentration could Calculation of Type 1 CR and Type 2 GR binding saturate the Type 1 CR sites and half- constants maximally saturate the Type 2 GR sites. The number of binding sites for Type 1 CR Note that a 5 fold molar excess of CORT was estimated from the apparent maximum bind- already showed 75% displacement of 3H- ing capacity (Bmax) of 3H-CORT in the presence CORT. of a 500 fold molar excess of RU 28362 and was Endocrinol. Japon. 216 DE KLOET et al. April 1987

Fig.1.•@ Scatchard plot of 3H-CORT (Type 1 CR, left) and of 3H-RU 28362 (Type 2 GR, right) in hippocampal cytosol of young and old rats and in old rats treated chronically with the ginsenoside Rgi. △-△Young rats ○-○ Old rats ●-● Old rats+Rg1. Type 1 CR was determined by 3H-CORT binding assay in the presence of a 500 fold excess of RU 28362, which completely prevented the binding of 3H-CORT to Type 2 GR. Type 2 GR was determined directly by 3H-RU 28362 binding. See Materials and Methods for details of Rgi treatment protocol and binding assays.

Table 1. Effect of chronic ginsenoside R-Gi treatment on age-induced changes in Type 1 corticosterone-preferring receptors (CR) and Type 2 glucocorticoid receptors (GR) in cytosol of rat hippocampus.

KD and Bmax SEM were determined from Scatchard and Woolf analyses respectively (Keightly et al., 1983). Analysis was performed with 9 determinations per group. rcorr was 0.99. Data from Figure 1. 1) nM 2) in fmoles/mg protein 3) old vs young rats; p<0.001; Student's t-test 4) old+Rgi vs old rats; p<0.001; Student's t-test

(Rigter et al., 1984; Sapolski et al., 1983). Discussion The present study showed that this decrease in CORT binding occurs in both CORT Previous studies had revealed that senes- binding receptor types, Type 1 CR and cent rats have a reduced maximal binding Type 2 GR. The Bmax of Type 1 CR from capacity for CORT in cytosol of the cortex old animals was about 50% of that observed (Roth, 1980) and of the hippocampus in young animals while Type 2 GR was Vol.34, No.2 GINSENG AND BRAIN CORTICOSTEROID RECEPTORS 217

Table2.•@ Displacement of3H-CORT from re- level and is subject to downregulation by ceptors in hippocampus cytosol by the homologous hormone, when the plasma various ginsenosides. CORT level is elevated. In this connec- tion, the reduced receptor concentration can be viewed as an adaptive response (Reul and de Kloet, 1985; Reul et al., 1987; de Kloet et al., 1986b). Recent experiments have shown that enhanced downregulation due to stress and aging may lead to irre- versible Type2GR loss and consequent further loss of CORT feedback (Sapolski et al., 1986). Type1CR appears insensitive to homologous downregulation, but instead 1) in fmoles/mg protein. The results are ex- responds in plasticity to neurotrophic agents pressed as means (n=4). SEM does not such as certain neuropeptides (Veldhuis and exceed5% of the given values shown. de Kloet, 1982; de Kloet and Veldhuis, 2) Incubations were performed with2nM3H- 1984; de Kloet et al., 1986b). In fact, a CORT (see Materials and Methods). behaviourally active ACTH analog (Org 2766), devoid of corticotrophic activity (de Wied and Jones, 1982), appeared to be decreased by28%. We also showed that effective in selectively increasing Type1CR treatment of old rats for two weeks with capacity in the hippocampus of aged rats the ginsenoside Rg1increased the number as was observed in the present study with of hippocampal Type1CR by60%, to the ginsenoside (Rigter et al., 1984; Reul near the receptor number observed in the et al. submitted manuscript). The ACTH hippocampus of the young control animals, analog also delays age-associated changes while the number of Type2GR was not in the hippocampus morphology of rats significantly changed. Fulder (1980; 1981) (Landfield et al., 1981). reported a stimulation of cell nuclear uptake of3H-CORT in the hippocampus of animals The ginsenoside RG1 which was selected treated for one week with0.5mg per day for this study is thought to exert a stimu- of a ginseng extract. As indicated by this latory effect on the central nervous system author, the in vivo uptake of3H-CORT may (Shibato and Saito, 1977; Takagi, 1974; also be affected by the altered metabolic Saito et al., 1984). How the Rg1exerts its clearance of the steroid induced by chronic action in the brain is not known; when the treatment with the ginseng extract. The compound is administered as a radio-labelled present study shows, however, that the en- tracer to rats only a very low concentration hanced uptake and retention of3H-CORT of intact ginsenoside can be recovered from primarily seemed to be due to the ginseno- brain tissue (Strombom et al., 1985). An side-induced increase in Type1CR capacity. interaction of RG1with Type1CR and Although Type1CR and Type2GR Type2GR seems unlikely in spite of a probably reside in the same hippocampal claim that a44% enriched saponin extract cells (Fuxe et al., 1985), the downregulatory from Panax Ginseng had demonstrable mechanism leading to a reduced number of affinity for mineralocorticoid receptors (MR) receptors is different for each receptor type. and Type2GR of kidney. The relative Our previous studies have shown that Type potency of the extract when compared with 2GR responds to changing plasma CORT ALDO and DEX was, however, only1: Endocrinol. Japon. 218 DE KLOET et al. April1987 10000and1: 2000respectively (Pearce et al., 1982). The present study showed that References RG1in a500fold molar excess did not displace3H-CORT from the hippocampal Beaumont, K. and D. D. Fanestil (1983). Char- receptor sites. Moreover, if RG1would acterization of rat brain aldosterone receptors bind to corticosteroid receptors in vivo one reveals high affinity for corticosterone. Endo- would not expect an increase in cortico- crinology113, 2043-2051. steroid receptor concentration as was ob- Brekhman, I. I. and I. V. Dardymov (1969). New served for Type1CR. substances of plant origin which increase non- specific resistance. Ann. Rev. Pharmacol. 9, An important question is whether the 419-430. glucocorticoid-like effects of ginseng (Fulder, Bohus, B. and E. R. de Kloet (1981). Adrenal 1986; 1987; Brekhman and Dardymov, steroids and extinction behaviour: antagonism by progesterone, deoxycorticosterone and dex- 1969; Hiai et al., 1979) can be ascribed to amethasone of a specific effect of corticoste- the RG1-induced increase in Type1CR rone. Life Sci. 28, 433-440. capacity. This receptor system is involved De Kloet, E. R., G. Wallach and B. S. McEwen in synchronization and coordination of cir- (1975). Differences in corticosterone and dex- cadian-driven brain processes underlying amethasone binding to rat brain and pituitary. daily activities and sleep related events. Endocrinology96, 598-609. Superimposed on this basal control exerted De Kloet, E. R. and H. D. Veldhuis (1984). Adrenocortical hormone action on the brain. by CORT via Type1CR is the activation In: Handbook of Neurochemistry Vol8 (A. of pituitary-adrenal activity by stress and Lajtha ed.) Pergamon Press, New York. pp. the subsequent feedback of CORT on stress- 47-91. activated neural circuitry and metabolism De Kloet, E. R., H. Sybesma and J. M. H. M. via Type2GR. According to this line of Reul (1986a). Selective control by cortico- reasoning, the selective increase in Type1 sterone of serotonin1 receptor capacity in the CR after RG1treatment measured in the raphehippocampal system. Neuroendocrinology 42, 513-522. present study and suggested by Fulder De Kloet, E. R. and J. M. H. M. Reul (1987). (1980: 1981) may explain why the saponin Feedback action and tonic influence of cortico- might affect the responsiveness of the or- steroids on brain function: a concept arising ganism to the environment. from heterogeneity of brain receptor systems. Psychoneuroendocrinology (in press). Taken together, the results of the present De Kloet, E. R., J. M. H. M. Reul, F. S. W. study have shown that the number of both de Ronde, M. Bloemers, and A. Ratka (1986b). Type1CR and Type2GR is reduced in Function and plasticity of brain corticosteroid the old rat, but that only Type1CR can receptor systems: action of neuropeptides. J. be restored towards the level observed in Ster. Biochem.(in press). De Wied, D. and J. Jolles (1982). Neuropep- young control animals after chronic infusion tides derived from pro-opiocortin: behavioural, of the ginsenoside Rg1. This effect is seen psychological and neurochemical effects. Phys- in the hippocampus which is a brain struc- iol. Rev. 62, 976-1056. ture involved in cognition, mood and affect. Fulder, S. J.(1980). The Root of Being: Gin- seng and the pharmacology of harmony. Hutchinson, London, UK. Acknowledgement Fulder, S. J.(1981). Ginseng and the hypo- thalamic-pituitary control of stress. Am. J. Chin. Med. 9, 112-118. The editorial assistance of Dr. Schonbaum- Desbarats is gratefully acknowledged. Funder, J. W.(1986). Adrenocorticoid receptors in the brain, In: Frontiers in Neuroendo- crinology Vol.9 (L. Martini and W. F. Ga- Vol.34, No.2 GINSENG AND BRAIN CORTICOSTEROID RECEPTORS 219

nong eds.) Raven Press, New York, USA. pp. 335. 169-189. Reul, J. M. H. M. and E. R. de Kloet (1985). Fuxe, K., A. C. Wikstrom, S. Okret, L. F. Two receptor systems for corticosterone in rat Agnati, F. Harfstrand, Z. Y. Yu, L. Granholm, brain: microdistribution and differential oc- M. Zoli, W. Vale and J. A. Gustaksson (1985). cupation. Endocrinology117, 2505-2511. Mapping of glucocorticoid receptor immuno- Reul, J. M. H. M. and E. R. de Kloet (1986). reactive neurons in the rat tel-and diencepha- Anatomical resolution of two types of cortico- ion using a monclonal antibody against rat sterone receptor sites in rat brain with in liver glucocorticoid receptor. Endocrinology vitro autoradiography and computerized image 117, 1803-1812. analysis. J. Ster. Biochem. 24, 269-272. Gispen, W. H., P. Schotman and E. R. de Kloet Reul, J. M. H. M., F. R. van den Bosch and (1972). Brain RNA and hypophysectomy: a E. R. de Kloet (1987). Differential response topographical study. Neuroendocrinology9, of type1and type2corticosteroid receptor to 285-296. changes in plasma steroid level and circadian Hiai, S. H., H. Yokoyama, H. Oura and S. Yano rhythmicity. Neuroendocrinology (in press). (1979). Stimulation of pituitary-adrenocortical Reul, J. M. H. M., J. A. D. M. Tonnaer and system by ginseng saponin. Endocrinol. Japon. E. R. de Kloet (submitted manuscript). Neuro- 26, 661-665. trophic ACTH analog promotes plasticity of Keightley, D. D., R. J. Fisher and N. A. C. corticosteroid subtype in brain of senescent Cressie (1983). Properties and interpretation rats. of the Woolf and Scatchard plots in analysing Rigter, H., H. D. Veldhuis and E. R. de Kloet steroid receptor assays. J. Steroid Biochem. (1984). Spatial orientation and the hippo- 19, 1407-1412. campal corticosterone receptor system of old Landfield, P. W., R. W. Baskin and T. A. Pitler rats: effect of ACTH4-9analog Org2766. (1981). Brain-aging correlates: retardation by Brain Res. 309, 393-398. hormonal-pharmacological treatments. Science Roth, G. S.(1980). Interaction of hormones 214, 581-584. with receptors and alterations of these pro- McEwen, B. S., J. M. Weiss and L. S. Schwartz cesses with age. Adv. Exptl. Med. Biol. 129, (1968). Selective retention of corticosterone 157-169. by limbic structures in rat brain. Nature220, Saito, H., N. Nishiyama, H. Fujimori, K. Hi- 911-912. nata, T. Kamegya, Y. Kato and T. Bao (1984). McEwen, B. S., E. R. de Kloet and W. H. Effects of drugs on sex and learning behavior Rostene (1986a). Adrenal steroid receptors and tyroxin hydroxylase activities of adrenal and actions in the nervous system. Physiol. gland and hypthalamic regions in chronic Rev. 66, 1121-1188. stressed mice. In: Stress the role of catechola- McEwen, B. S., L. T. Lambdin, T. C. Rainbow mines and other neurotransmitters (E. Usdin, and A. deNicola (1986b). Aldosterone effects R. Kvetnansky, J. Axelrod eds) Gordon and on salt appetite in adrenalectomized rats. Breach Science Publishers, New York, USA. Neuroendocrinology43, 38-43. pp.467-480. Moguilevski, M. and J. P. Raynaud (1980). Sapolski, R., L. Krey and B. S. McEwen (1983). Evidence for a specific mineralocorticoid recep- Corticosterone receptors decline in a site tor in rat pituitary and brain. J. Ster. Bio- specific manner in the aged rat brain. Brain chem. 12, 309-314. Res. 289, 235-240. Pearce, P. T., I. Zois, K. N. Wynne and J. W. Sapolski, R., L. C. Krey, B. S. McEwen and T. Funder (1982). Panax ginseng and eleuthro- C. Rainbow (1984). Do vasopressin-related coccus senticosus extracts- in vitro studies on peptides induce hippocampal corticosterone binding to steroid receptors. Endocrinol. Japon. receptors? Implications for aging. J. Neurosci. 29, 567-573. 4, 1479-1485. Philibert, D. and M. Moguilevski (1983). RU Sapolski, R., L. C. Krey and B. S. McEwen 28362, a useful tool for the characterization (1986). The Neuroendocrinology of Stress and of the glucocorticoid and mineralocorticoid Aging: The Glucocorticoid Cascade Hypo- receptors. 65th Annual meeting of the Endo- thesis. Endocrine Rev. 7, 284-301. crine Society, San Antonio, Abstr nr1018, p Shibata, S. and H. Saito (1977). Some recent Endocrinol. Japon. 220 DE KLOET et al. April1987

studies on ginseng saponins. In: Proc. 2nd Vasopressin related peptides increase the hip- Int. Geront. Symp. Singapore: Pharmaton Ltd, pocampal corticosterone receptor capacity of Switzerland. Bioggio, Lugano, Switzerland. diabetes insipidus (Brattleboro) rats. Endo- pp.183-197. crinology110, 153-157. Strombom, J., F. Sandberg and L. Dencker Veldhuis, H. D., C. van Koppen, M. van Itter- (1985). Studies on absorbtion and distribution sum and E. R. de Kloet (1982). Specificity of ginsenoside Rgi by whole body autoradio- of adrenal steroid receptor system in rat hip- graphy and chromatography. Acta Pharmaceu- pocampus. Endocrinology110, 2044-2051. tica Suecia22, 113-122. Veldhuis, H. D. and E. R. Kloet (1983). An- Takagi, K.(1974). Pharmacological studies on tagonistic effects of aldosterone on cortico- ginseng. In: Proc. 1st Int. Ginseng Sympo- sterone-mediated changes in exploratory be- sium. Seoul, Korea. Central Res. Inst., Office havior of adrenalectomized rats. Hormones of Monopoly. pp.119-128. and Behavior17, 225-232. Veldhuis, H. D. and E. R. de Kloet (1982).