t. maurice, t.p. su and a. privat- sigma1 (sigma1) receptor agonists and neurosteroids attenuate...

8/3/2019 T. Maurice, T.P. Su and A. Privat- Sigma1 (sigma1) Receptor Agonists and Neurosteroids Attenuate B25-35-Amyloid Peptide-Induced Amnesia in Mice Through

1/16

SIGMA1 (1) RECEPTOR AGONISTS AND NEUROSTEROIDS

ATTENUATE 2535-AMYLOID PEPTIDE-INDUCED

AMNESIA IN M ICE THROUGH A COMMON MECHANISM

T. MAURICE,* T.-P. SU and A. PRIVAT*

*I.N .S.E.R .M . U nite 336, D eveloppement, Plast icite et Vieillissement du Systeme Ner veux,8 rue de lEcole Normale, 34296 Montpellier Cedex 5, France

Unit of Pathobiology, Molecular Neuropsychiatry Section, Division of Intramural Research,N.I.D.A., N.I.H., Baltimore, MD 21224, U.S.A.

AbstractThe sigma1 (1) receptor agonists exert potent anti-amnesic effects, as they appa rently block t helearning impairments either induced by t he muscarinic receptor antagonist scopolamine, the N -methyl--aspartate receptor antagonist dizocilpine or inherently due to the age-related deficits in senescence-accelerated mice. We recently described the amnesia induced by the -amyloid-related peptide 2535,administered centrally in an aggregated form, in mice. The deficits were sensitive to cholinomimetics or to

N-methyl--aspartate/glycine modulatory site agonists. Herein, we examined the effects of 1 receptorligands on t he 2535 peptide-induced amnesia. The effects of neuro(active) steroids, which interact in vitroan d in vivo with 1 receptors were examined in parallel. Mnesic capacity was evaluated seven days afteradministration of aggregated 2535 peptide (3 nmol), using spontaneous alternation in the Y-maze forspatial short-term memory, or after 14 days, using the step-down type passive avoidance test for long-termmemory. The 1 receptor agonists (+ )-pentazocine, PR E-084, or SA4503 attenuated, in a dose-dependentand bell-shaped manner, the 2535 peptide-induced deficits o n both tests. T hese effects were antagonizedby haloperidol or BMY-14802, confirming the 1 receptor pharma cology. Pregnenolone, dehydroepian-drosterone, and their sulphate esters, but not progesterone, also dose-dependently attenuated the 2535peptide-induced deficits. Progesterone blocked the beneficial effects of each other neurosteroid, behavingas an antagonist. Furthermore, haloperidol blocked the effects induced by neurosteroids, whereasprogesterone antagonized the effects of the non-steroidal 1 receptor agonists, showing a clear crossedpharmacology of different drug classes.

These results demonstrate that: (i) the anti-amnesic effect o f1 receptor agonists may b e of therapeuticrelevance in pathological states affecting the cholinergic and/or glutamatergic systems, such as inpathological aging; (ii) neurosteroids play an important role in learning processes and may collectivelyconstitute a therapeutic tar get; (iii) t he interaction between 1 systems and neurosteroids appears indeedof behavioural relevance. 1997 IBRO. Published by Elsevier Science Ltd.

Key words: -amyloid peptides, sigma1 (1) receptor ligands, neurosteroids, alternation behavior, passiveavoidance, amnesia model.

Until its recent cloning by Hanner et al.,14 the sigma

() receptors appeared as particularly enigmatic

molecular targets. Initially introduced by Martin

et al.28 and considered as one subtype of opioid

receptors, the receptors were th en considered as

distinct entities, not only from the other classical

opioid receptors, but also from the high-affinity

phencyclidine binding site, located within the ion

channel associated with the N-methyl--aspartate

(NMDA)-type of glutamate receptor.56 Both phen-

cyclidine sites and receptors share moderate to

h igh affinities for different chemical classes of

drugs, leading to a long-lasting confusion between

their respective effects. At least, two different site

subtypes have been proposed, on the basis of

their pharmacological profiles, termed 1 an d 2.55

Hanner et al.14 recently reported the p urification and

cloning of the 1 binding site from guinea-pig liver.

The 25,000 mol. wt protein revealed no homology

to known mammalian proteins and presented one

putative transmembrane domain. Interestingly the

amino acid sequence shared some homology with

fungal proteins involved in the steroid synthesis and

north ern blot analysis showed high densities of 1site mR N A in steroid-producing tissues, par ticularly

in the adrenal gland, ovary and the 20-day gestation

fetus.14

Numerous physiological functions have been

proposed for 1 receptors.40,62 Among the most

important ones, they mediate a potent neuro-

modulatory role on two excitatory systems, the

cholinergic neurotransmission and the NMDA-type

To whom correspondence should be addressed.Abbreviations: DH EA, dehydroepiandrosterone; DH EAs,

dehydroepiandrosterone sulphate; D MSO, dimethylsul-phoxide; KW, KruskalWallis statistics; NMDA,

N-methyl--aspartate; N.S., not significant; SAM,senescence-accelerated mice; SDL, step-down latency.

Pergamon

Neuroscience Vol. 83, No. 2, pp. 413428, 1998Copyright 1997 I BRO. Published by Elsevier Science Ltd

Printed in Great Britain. All rights reserved03064522/98 $19.00+0.00PII: S0306-4522(97)00405-3

413


2/16

of glutamatergic system. The 1 receptor agonists

induced increases in extracellular acetylcholine levels

measured by in vivo microdialysis in the rat frontal

cortex and hippocampus,30,31,33 in [3H]acetylcholine

release from hippocampal slices,18 and a selective

potentiation of several NM DA-evoked responses

in vitro an d in vivo, such as the electric activity

of CA3 dorsal hippocampal neurons,50,51 an d

t h e [3H]norepinephrine release from hippocampal

slices.49,58 With regard to the behavioural relevance

for these effects, 1 receptor agonists were demon-

strated to have anti-amnesic properties, as recently

reviewed by Maurice and Lockhart,40 in amnesia

models induced by cholinergic antagonists,6,3436 or

by the non-competitive NM DA antagonist dizo-

cilpine.7,37,38,45,53 More recently, we reported that

these compounds were also effective in alleviating the

learn ing deficits exhibited by senescence-acceleratedmice (SAM), an original rodent model of age-related

cognitive deficits with early onset.44

Neuro(active)steroids, such as progesterone, preg-

nenolone, or dehydroepiandrosterone (DHE A), and

their sulphate esters (pregnenolone sulphate, DHEA

sulphate), are involved in regulating the balance

between excitation and inhibition in the brain,54,65

and also in modulating learning and memory mecha-

nisms.10,11,13,47,57 Steroid hormones are known to

bind to specific cytoplasmic receptors, which trans-

locate into the nucleus and regulate gene expression,affecting translational efficiency and protein stability.

These processes may be involved in long-lasting

learning.11,57 H owever, they also affect more directly

the activity of neurotransmitter systems involved in

learning. F irstly, they act as a llosteric mod ulators of

the GABAA receptor.2527 Secondly, they modulate

several NMD A-evoked responses, such as the

NMDA-gated currents in cerebellar or spinal cord

neurons,61,65 the convulsant potency of NMDA in

mice,24 the NM DA-mediated increases in intra-

cellular Ca

2+

in cultured rat or chick hippocampalneurons,2,9,15,16 the NMD A-evoked [3H]norepine-

phrine release from rat hippocampal slices,52 and the

NM DA-induced electrical activity in CA3 dorsal

hippocampal neurons.1 Furthermore, pregnenolone

sulphate or DHEA sulphate were shown to attenuate

the learning impairments induced by competitive or

non-competitive NMDA receptor antagonists.4,29,39

On the other hand, progesterone and, to a lesser

extent pregnenolone sulphate, testosterone, and 17-

estradiol, inhibited the in vitro binding of1 radio-

ligands in rat brain, spleen and liver preparations,

and, conversely, 1 receptor ligands inhibited the

b in din g o f [3H]progesterone.19,48,59,64,66 Together

with the homology shared by the recently cloned 1receptor with fungal steroidal synthesizing enzyme,14

several functional studies confirmed the interaction

between neurosteroids and 1 receptors.1,52 We

also reported that DH EA sulphate attenuated the

dizocilpine-induced learning impairment in mice by a

mechanism involving 1 receptors.39 These observa-

tions suggested a crossed pharmacology between the

functional effects of neurosteroids and 1 receptor

ligands regarding, at least, their effect s o n t he

NMDA receptor activation.

Central administration of the aggregated form of

2535 peptidergic fragment of the -amyloid protein

into the mouse or rat brain induced histological and

biochemical changes and memory deficits.3,5,21,41

These changes appeared reminiscent of the deposi-

tion of amyloid plaques, endogenously constituted

from 140 an d 142-amyloid proteins, but not

2535, which extent correlates with the progressive

cognitive deficits and memory impairment observed

in Alzheimers disease patients.60 The mechanism of

the neurotoxicity induced by aggregated 2535 pep-

tide involved, similarly for 140 or 142-amyloid

proteins, its ability to self-aggregate and form Ca2+-

permeable channels in membranes, resulting inexcessive Ca2+ influx and induction of neurotoxic

cascades.2022

In this study, we checked the ability of the refer-

ence 1 receptor agonist (+)-pentazocine and of two

new and selective 1 receptor agonists, PR E-08463

and SA4503,32 to attenuate th e learning impairments

induced by central administration of 2535-amyloid

peptide, by using the spontaneous alternation

and passive avoidance tests in mice. The antagonist

effects of haloperidol and BMY-14802 were also

checked. Then, we examined the effects of severalneurosteroids.

EXPERIMENTAL PROCEDURES

Animals

Male Swiss mice (Breeding centre of the Faculty ofPharmacy, M ontpellier, F rance), a ged five to six weeks an dweighing 3035 g, at the beginning of the experiments, wereused throughout the study. Animals were housed in plasticcages, with free access to laboratory chow and water, exceptduring behavioural experiments, and kept in a regulated

environment (231C, 4060% humidity), under a 12 hlight/dark cycle (light on at 8:00 a.m.). Experiments werecarried out between 10.00 a.m. and 6.00 p.m., in a sound-proof and air-regulated experimental room, to which micewere habituated at least 30 min before each experiment.Animal care followed the protocols and guidelines approvedby I.N.S.E.R.M., in particular all efforts were made tominimize animal suffering and to reduce the number ofanimals used.

Drugs and administration procedures

(+)-pentazocine was donated by Dr F. J. Roman (Institutde Recherche Jouveinal, Fresnes, France); PRE-084 was

donated by Dr D. W. Parish (S.R.I. International, MenloPark, CA); SA4503 was donated by D r K . Ma tsuno (SantenPharmaceuticals, Osaka, Japan); BMY-14802 was fromBrystolMyers (U.S.A.); haloperidol (Haldol) was fromJanssen (BoulogneBillancourt, France); pregnenolone(5-pregnen-3-ol-20-one), pregnenolone sulphate, dehy-droepiandrosterone (5-androsten-3-ol-17-one, DH EA),dehydroepiandrosterone sulphate (DHEAs), and progester-one (4-pregnene-3,20-dione) were from Sigma (Saint-Quentin F allavier, Fr ance). BMY-14802 was dissolved in aminimal volume of 0.1 N HCl and then in saline solution,

414 T. M aurice et al


3/16

pH being adjusted to 4 with 1 N NaOH. Pregnenolone,DH EA, a nd progesterone were dissolved in pur e sesame oil(Sigma). Pregnenolone sulphate and DHEAs were dissolvedin d imethylsulfoxide (DM SO), t hen in saline solution, finalvehicle being DMSO 5% in saline. Other compounds weredissolved in saline. Drugs were injected subcutaneously(s.c.) or intraperitoneally (i.p.), in a volume of 100 l/20 g of

body weight. The 2535 peptide and scrambled 2535 pep-tide were from Neosystems (Strasbourg, France). Peptideswere dissolved in sterile bidistilled water, at a concentrationof 1 mg/ml, and stored at 20C until use. The peptideswere aggregated by incubation, at 1 mg/ml in sterile bi-distilled water, at 37C for four days. Light microscopicobservation indicated that incubating the 2535 peptide, butnot the scrambled 2535 peptide, led the presence of twotypes of insoluble precipitates, birefringent fibril-like struc-tures and amorphous globular aggregates (data not shown).Th e -amyloid peptides were then administered intra-cerebroventricularly (i.c.v.), as previously described.41,42 Inbrief, each mou se was ana esthetized lightly with ether, anda 28-gauge stainless-steel needle (Exmire, Ito Corpor ation,Fuji, Japan), 3 mm long, was inserted unilaterally 1 mm tothe right of the midline point equidistant from each eye, atan equal distance between the eyes and the ears andperpendicular to the plane of the skull. Peptides or vehicle(3 l) were delivered gradually within appro ximately 3 s.Mice exhibited normal behaviour within 1 min after injec-tion. The administration site was checked by injectingIndian ink in preliminary experiments. Neither insertion ofthe needle, nor injection of the vehicle had a significantinfluence on survival, behavioural responses or cognitivefunctions.41

Spontaneous alternation performances

Spatial working memory performance was assessed sevendays after the -amyloid peptide a dministration, by record-ing spontaneous alternation behaviour in a Y-maze.37

39,4146 The maze was made of black painted wood. Eacharm was 40 cm long, 13 cm high, 3 cm wide at the bottom,10 cm wide at the top, and converged at an equal angle.Each mouse was placed at the end of one arm and allowedto move freely through the maze during an 8 min session.The series of arm entries, including po ssible returns into thesame arm, was recorded using an Apple IIe computer. Analternation was defined as entries into all three arms onconsecutive occasions. The number of maximum alterna-tions was therefore the tota l number of arm entries minustwo and the percentage of alternation was calculated as(actual a lternations/maximum a lternations)100. The 1receptor ligands or neurosteroids were a dministered 30 minbefore t he session.

S tep-down t ype passive avoidance test

Delayed amnesia was examined u sing t he step-down typeof passive avoidance task, 14 days after the -amyloidpeptide administration.37,39,4146 The apparatus consistedin a transparent acrylic cage (303040 cm high) witha gridfloor, inserted in a semi-soundproof outer box(353590 cm high). The cage was illuminated with a15 W lamp during the experimental period. A woodenplatform (444 cm) was fixed at the centre of the grid-

floor. Electric shocks (1 Hz, 500 ms, 45 V DC) were deliv-ered to the gridfloor using an isolated pulse stimulator(Mo del 2100, AM Systems, Everett, WA). Th e test consistedin two training sessions, at 90 min time interval, and in aretention session, carried out 24 h after the first training.During training sessions, each mouse was placed on theplatform. When it stepped down and placed its four pawson the gridfloor, shocks were delivered for 15 s. Step-downlatency (SDL) and the nu mbers of vocalizations and flinch-ing reactions were measured. Shock sensitivity was evalu-ated by summ ing these two num bers. N one of the

treatments used in this study affected significantly the SDLor shock sensitivity showed by the animals during the firsttraining session as compared to control animals (e.g.,SDL= 5[313] s, shock sensitivity=121, n=12), indicatingthat the behaviours observed during the retention sessionmay directly reflect learning and memory abilities. Animalswhich did not step down within 60 s during the second

training session were considered as remembering the taskand taken off, without receiving electric shocks any more.This rout inely-used pro cedure allows to minimize th e intra-group variability without affecting the relevance of thebehavioural measure. The retention test was performed in asimilar manner as training, except that the shocks were notapplied to the gridfloor. Each mouse was placed again onthe platform, and the SDL was recorded, with an uppercut-off time of 300 s. Two parametric measures of retentionwere analysed: the SDL and the number of animals reachingthe avoidance criterion, defined as reached if the SDLmeasured during the retention session was greater thanthree-fold the SDL showed b y the an imal during the secondtraining session and, at least, greater than 60 s. Basically,median SDL could be considered as a qualitative index ofmnesic capacities, whereas the percentage of animals tocriterion could be considered as a quant itative index.39,4144

Th e 1 receptor ligands or neurosteroids were routinelyadministered 30 min before the first training, and once,injections not being repeated before the second training, orthe retention test. In some experiments, PRE-084 wasadministered immediately after the first training or 30 minbefore retention.

Experimental series

Half of the mice were examined seven days after the i.c.v.

administration of-amyloid peptide for spontaneous alter-nation behavior in the Y-maze. 14 days after the peptideadministration, the second half were used for the passiveavoidance task. This schedule was chosen for convenience,since previous studies showed that marked learning impair-ments appear six days after peptide administration41 an dcould still be observed after at least 30 days (unpublishedobservation). I n a preliminary series of experiments, t he 1receptor ligands and neurosteroids were examined onvehicle-treated mice, in order to confirm that none of themaffect the mnesic capa cities, measured using either spon-taneous alternation o r passive avoidance, b y themselves, asobserved in untreated animals for (+)-pentazocine,37 PRE-08445 and SA450334,43, and for the neurosteroid DH EAs.39

Then, the doseresponse effects of the 1 ligands wereexamined in 2535-treated animals at doses ranging from0.1 to 3 mg/kg, and the antagonist effects of haloperidoland BMY-14802 as tested. In the third series, the doseresponse effects of systemic injections of neurosteroids wascharacterized at doses ranging from 5 to 20 mg/kg. Theantagonist effect of pr ogesterone was tested at the 20 mg/kgdose. Finally, the crossed pharmacology between the 1receptor ligands and the neurosteroids was investigated byexamining (i) the antagonist effect of progesterone on thepeak effect of each 1 receptor agonists, and (ii) the block-ade by haloperidol of the highest efficient dosage of eachneurosteroid.

S tatistical analysis

Results are expressed a s meansS.E.M., excepting SDL,which are expressed in terms of medians and interquartileranges. Data did not show a normal distribution, sincecut-off times were set. They were analysed using theKruskalWallis nonparametric ANOVA (KW values),group comparisons being made with the Dunns non-parametric multiple comparisons test. The levels forstatistical significance were P


4/16

RESULTS

As shown in Table 1, i.c.v administration of the

vehicle, sterile bidistilled water, or of scrambled 2535-amyloid peptide (3 nmol), did not affect the m nesic

performances on each test, as compared to control

untreated animals. A decrease in locomotion was,

however, observed with the group administered with

scrambled 2535 peptide. Mice treated with aggre-

gated 2535 peptide (3 nmol) exhibited marked

behavioral deficits, that affected spontaneous

alternation, but not locomotion, and both passive

avoidance parameters (Table 1). These deficitsappear thus due to the selective effect of aggregated

2535-amyloid peptide.

Effects of the 1

receptor agonists ( +) -pentazocine,

PRE-084 and SA4503

The preliminary experiments showed that none of

th e 1 receptor agonists affected the mnesic capacities

by itself (data not shown).

All compounds significantly attenuated the deficitsobserved in animals treated with aggregated 2535peptide. Administration of increasing doses of (+)-

pentazocine led to a bell-shaped attenuation of the

deficits in alternation behavior (KW=29.02, P


5/16

The beneficial effects of 1 receptor agonists

against the learning deficits induced by the 2535peptide were also observed during the retention

session of the step-down passive avoidance test, aspresented in Fig. 2. First, (+)-pentazocine allowed a

bell-shaped attenuat ion of the decreases in b oth SDL

(KW=19.78, P


6/16

showed that 1 receptor agonists appeared more

effective in the learning and consolidation phases

than on retention.

Effects of haloperidol and BM Y -14802

The effects of the putative 1 receptor antagonists

haloperidol and BMY-14802 were investigated, in

order to confirm the implication of the 1 system. A s

shown in Fig. 3A, haloperidol, administered i.p. at

0.030.1 mg/kg (80266 nmol/kg), did not aff

ect the2535-induced decrease in spontaneous alternation.

Furthermore, the simultaneous administration of

haloperidol, 0.1 mg/kg, with the most active dose of

each 1 receptor agonist led to a complete and highly

significant blockade of its effect (Fig. 3A). It must be

noted that haloperidol, due to its dopaminergic D2receptor antagonist properties, affected the loco-

motor activity of the animals: at 0.1 mg/kg, animals

completed 171 arm entries, vs 301 for controls

(P < 0 .01) a nd 343 for 2535-treated animals

(P


7/16

pregnenolone sulphate: KW=24.81, P


8/16

A complementary experiment was performed to

examine the effect of DHEAs or pregnenolone sul-

phate on the consolidation and retention phases ofthe learning processes. DHEAs or pregnenolone sul-

phate (20 mg/kg s.c. each) was administered either

immediately after the first training, or 30 min before

the retention session. The control group showed

SDL= 156 [85234] s, with a percentage of an imals to

criterion of 68.8% (n=16). The 2535-treated group

showed SDL=24 [1673] s (n=21, P< 0.01 vs con-

trols), with 19.0% of animals to criterion (P


9/16

the passive avoidance task, haloperidol blocked

the eff

ects of pregnenolone and DHEA, in term ofSDL (F ig. 9C ) a nd in p ercen ta ge o f a nim als

to criterion (Fig. 9D), although the difference

between the pregnenolone-treated group and the

(haloperidol+pregnenolone)-treated group did not

appear significant. Similarly, haloperidol blocked the

effects of pregnenolone sulphate and DHEAs

(Fig. 9E, F ). R egarding th e percentages of animals to

criterion, the difference b etween the pregnenolone

sulphate-treated group and the (haloperidol+

pregnenolone sulphate)-treated grou p remained non-

significant (Fig. 9F ). T hese d ata indicated, however,

that haloperidol suppressed the attenuating effects

of neurosteroids on the 2535-induced learning

impairment.

Antagonism by progesterone of the effects of non-

steroidal 1

ligands

The beneficial effects of (+)-pentazocine, PRE-084,

or SA4503 on the spontaneous alternation deficits

induced by 2535 could be significantly blocked by

progesterone (Fig. 10A). The eff

ects on passiveavoidance could also be blocked in terms of SDL

(Fig. 10B) and in terms of animals to criterion (Fig.

10C), although some differences did not reach signifi-

cant levels between the 1 receptor agonist-treated

group and the (progesterone+1 receptor agonist)-

treated group.

DISCUSSION

The present study describes the beneficial effects of

selective 1 receptor agonists and neurosteroids on

the learning impairment induced by central admin-

istration of 2535-amyloid peptide in mice. These

results showed that the anti-amnesic effects of 1receptor agonists, previously evidenced on pharma-

cological models of learning impairment induced

by scopolamine, mecamylamine, dizocilpine, nimo-

dipine, or ageing, in SAM, an age-related amnesia

model,44 could be extended to a model of pathologi-

cal aging related to the Alzheimers disease aetiology,

Fig. 6. Effect of the neurosteroids on the step-down passive avoidance deficits in 2535-treated mice.Passive avoidance tra ining was performed 14 days after the i.c.v. administration of distilled water (Veh),or aggregated 2535-amyloid p eptide (3 nmol/mouse), and retention was examined 24 h after. R esults arepresented as median SDL and interquartile range (A, C, E, G ) and p ercentages of animals that reached theavoidance criterion (B, D, F, H). Pregnenolone (PREG) (A, B), PREGs (C, D), DHEA (E, F) andDHEAs (G, H), 520 mg/kg each, were administered s.c. 30 min before the first training. Vehicle (Veh)was sesame oil for PR EG or D HEA, a nd D MSO 5% in saline for PREG s or D HEAs. The numb er of miceper group is indicated below the columns. *P


10/16

the amyloid-type amnesia. Furthermore, the inter-

action between neurosteroids and 1 receptors,

already observed in functional studies, appeared here

to be of behavioural relevance, the anti-amnesic

effects induced by neurosteroids and 1 receptor

agonists sharing a clear crossed pharmacology. The

therapeutic impact of t hese effects ar e of importance.

In vivo rodent models of-amyloid peptide depo-

sition and related amnesia, which are essential for

discovering new therapeutical outcomes against the

Alzheimers type p ath ogenesis, were repor ted. M icro-injections o f-amyloid peptides into the rat cortex or

hippocampus produces neuronal loss and damaged

neurites.8,12,21,22 Among the different models, we

reported t hat the acute i.c.v. administration o f aggre-

gated 2535-amyloid peptide (39 nmol) in mice in-

duced after six to 13 days spontaneous alternation

deficits in the Y-maze, and impairments of step-down

type passive avoidance and place learning in a water-

maze.41 Similarly, acute administration of aged 2535peptide (15 nmol) into the rat brain ventricle led after

14 days to significant learning deficits in the Morris

water-maze.5 A m odera te cell loss was observed u sing

Cresyl Violet-stained brain sections, together with

Congo Red-stained amyloid deposits, in the fronto-

parietal cortex and h ippocampal formation of2535-

treated mice.41 The learning impairments could be

alleviated using the cholinomimetics tacrine o r ()-

nicotine,41 or using the NMDA/glycine site agonists

-cycloserine or milacemide,42 indicating that

both cholinergic and glutamatergic dysfunctions are

involved in the -amyloid type amnesia. Chen et al.3

also reported that repeated intra-hippocampal ad-

ministrations of 2535 peptide induced after one

week increases in norepinephrine, serotonin and

metabolite levels in the contralateral side. These

observations suggested an enhanced synthesis of

these monoamines in order to compensate for a loss

of tyrosine hydroxylase and recurrent inhibition in

the ipsilateral side.3 In summary, central admin-

istration of2535-amyloid peptide induced a delayed

amnesia, that appeared closely related to the 140-induced amnesia, involving perturbations of several

neurotransmitter systems. It must be noted that if

these rodent peptidergic models a ppeared of interest

regarding their pharmacological profiles, they could

not encompass the evident complexicity of the

myriad of subsequent mechanisms implicated in the

pat hological pro gression of Alzheimers d isease, such

as apolipoprotein E genotype, neuronal degenera-

tion, inflammation and oxidative stress, nor the

progression over years of amyloid deposits in

Alzheimers disease patients. T he r esults o bserved on

such rodent peptidergic models would thus allow a

relevant preliminary screening of new therapeutic

approaches, but extension to human therapy must be

put forward cautiously.

As recently reviewed,40 th e 1 receptor agonists

play an important neuromodulatory role in learning

and memory processes. None of them have been

shown to affect by themselves th e learning ability.

Ho wever, they were report ed to improve the amn esia

Fig. 7. Effect of p rogesterone (PROG ) and o f the simultaneous administrations of PROG with each ot herneurosteroid on the spontaneous alternation deficits in 2535-treated mice: PR OG alone (A), and incombination with pregnenolone (PREG) or DHEA (B), or PREGs or DHEAs (C). Spontaneousalternation was examined seven days after the i.c.v. administration of distilled water (Veh), or aggregated2535-amyloid peptide (3 nmol/mouse). PROG, 20 mg/kg, was administered s.c. simultaneously with eachother neurosteroid, 20 mg/kg each, which were given s.c. 30 min before the session. Results are expressedas meanS.E.M. of the number of mice indicated below the columns. **P


11/16

induced by the muscarinic cholinergic receptor antag-

onist scopolamine,6,3436 and by th e nicotinic cholin-

ergic antagonist mecamylamine.37,45 On the other

hand, they also exert a potent beneficial effect on

the learning impairment induced by blockade of the

NM DA receptor activation.7,37,38,45,53 Among the

ligands tested, SA4503 appeared particularly interest-

ing, since it reversed the dizocilpine-induced deficits

with an equal po tency as the scopolamine-induced

deficits, at sub-mg/kg doses systemically.34,43 This

apparent non-selectivity of the anti-amnesic effects o f

1 receptor agonists on several different neurotrans-

mitter systems, together with their effectiveness in

attenuating the learning impairment induced by the

L-type voltage d ependent calcium channel blocker

nimodipine,46 strongly suggested that their effect

involved intracellular modulation of calcium homeo-

stasis rather t han a direct modulation of t he function-

ing of extracellular receptor complexes.40,46 Th e

present results indicated that 1 receptor agonists

appeared particularly efficient in attenuating the

-amyloid type amnesia. This effect is likely to

involve the combined effects of1 receptor agonists

on cholinergic as well as glutamatergic neurons,

which both appeared to be a putative target of

therapeutic interest. Interestingly, the compounds

appeared efficient on the short-term as well as on the

long-term memory tests, indicating that they may

induce a general facilitat ion o f learning, involving no t

only a particular type of learning. Furthermore,

PRE-084 appeared efficient on the learning phase of

the memory process as well as on consolidation,

indicating that the drug facilitated not only the

NMDA-dependent learning, but also affected

independently the cholinergic dependent memory

processes, confirming the non-selectivity of the 1anti-amnesic effect.

The impact of endogenous neurosteroid levels and

of exogenously applied neurosteroids on the learning

capacities have been report ed in several stu dies. F irst,

Fig. 8. Effect of p rogesterone (PROG ) and of the simultaneous administrations of PROG with each ot herneurosteroid on the step-down p assive avoidance deficits in 2535-treated mice: PROG alone (A, B), andin combination with pregnenolone (PREG) or DHEA (C, D), or PREGs or DHEAs (E, F). Passiveavoidance training was performed 14 days after the i.c.v. administration of distilled water (Veh), oraggregated 2535-amyloid peptide (3 nmol/mouse), and retention was examined 24 h after. Results arepresented a s median SDL a nd interquartile range (A, C, E) and percentages of animals to criterion (B, D,F). PROG , 20 mg/kg, was administered s.c. simultaneously with each other neurosteroid, 20 mg/kg each,which were given s.c. 30 min before the first training. The number of mice per group is indicated below thecolumns. *P


12/16

administration of DHEA or its sulphate was shown

to ameliorate the learning ability of mice submitted

to a footshock active avoidance training test in the

T-maze, or of aged mice, the 18-month-old animals

treated centrally or systemically with DHEAs being

able to learn as well as two-month-old animals.11,57

The memory enhancing effects of neurosteroids could

be observed on several other compounds, including

pregnenolone and its sulphate, or testosterone, but

not progesterone.10 The authors thus suggested that

active neurosteroids, such as DH EA or pregne-

nolone, could be of clinical utility in pathologies

F ig. 9. Effect of the simultaneous administrations of haloperidol with each neurosteroid on thespontaneous alternation deficits (A, B), and on the step-down passive avoidance deficits (C, D, E, F) in2535-treated mice: haloperidol in combination with pregnenolone (PREG) or DHEA (A, C, D), orPREGs or DHEAs (B, E, F). Haloperidol, 0.1 mg/kg, was administered i.p. simultaneously with eachneurosteroid, 20 mg/kg each, which were given s.c. 30 min before th e Y-maze session or the first trainingin t he p assive avoidance test. Y-maze test results a re expressed a s meanS.E.M. of the number of miceindicated below the columns (A, B). Passive avoidance results are presented as median SDL andinterquartile range (C, E) and percentages of animals to criterion (D, F). **P


13/16

characterized by lower serum levels of these neuro-

steroids, such as normal or pathological aging, the

evolution of cancers or disorders in individual receiv-

ing drugs that block the synthesis of cholesterol,

the p recursor of pr egnenolone.10 A d irect implication

of neurosteroids in Alzheimers pathology was

suggested by Mayo et al.,47 reporting the beneficial

effect of pregnenolone sulphate, and conversely the

disrupting effect of tetrahydroprogesterone, in rats

submitted to a two-trial forced alternation task in a

Y-maze, when the drugs were locally administered

into the nucleus basalis magnocellularis. On the

other hand, neurosteroids, such as DHEAs or preg-

nenolone sulphate, have been reported to attenuate

the scopolamine-induced amnesia in rodents.23,47

More recent studies described anti-amnesic effects

of neurosteroids on the learning impairments

induced by blockade of the N MD A receptor activa-tion. First, Mathis et al.29 reported th at i.c.v. admin-

istration of pregnenolone sulphate in rats attenuated

the deficits induced by the competitive NMD A

receptor antagonist 3-(()-2-carboxypiperazin-4-

yl)-propyl-1-phosphonic acid in the step-through

passive avoidance test. These observations were

confirmed by Cheney et al.,4 who also reported

a similar effect on the dizocilpine-induced learning

deficits. In the present study, we confirmed that

neurosteroids administered systemically exert a

potent modulation of learning processes, as evi-d en ced o n t he 2535-amyloid peptide-induced

amnesia model in mice. Furthermore, we described

that pregnenolone, DHEA and their sulphate esters

behaved as anti-amnesic drugs, whereas progesterone

behaved as an antagonist for this anti-amnesic

effect.

Most of the studies cited suggested that the anti-

amnesic effect of neurosteroids was likely to involve

their interaction with the GABAA systems, as pre-

viously reported.2527 Indeed, p regnenolone sulphate

or DHEAs bind to modulatory sites associated withthe GABAA /benzodiazepine receptor complex and

are negative allosteric modulators of the GABAAreceptor-mediated Cl conductance, whereas proges-

terone and several of its metabolites act as positive

modulators. However, the effects o f n eurosteroids on

NMDA systems could also be involved, since similar

pharmacology was evidenced, i.e. pregnenolone,

DH EA and their sulphate esters, potentiated the

responses induced by the excitatory neurotrans-

mitter, whereas progesterone attenuated it or blocked

the potentiating effects of the oth er neur o-

steroids.1,2,9,15,16,24,52,61,65 Indeed, a similar pharma-

cology could be observed in behavioural studies

examining the neurosteroidal effects on learning

impairment induced by competitive or non-

competitive NMD A antagonists, pregnenolone

sulphate and DHEAs attenuating the deficits,4,29,39

and progesterone behaving as an antagonist.43

Some discrepancies still remain between the

different functional studies describing the crossed

Fig. 10. Effect of the simultaneous administrations ofprogesterone (PROG) with 1 receptor agonists on thespontaneous alternation deficits (A), and on the step-downpassive avoidance deficits (B, C) in 2535-treated mice.

PROG, 20 mg/kg, was administered s.c. simultaneouslywith (+)-pentazocine, 0.1 mg/kg, or PRE-084, 0.3 mg/kg, orSA4503, 0.3 mg/kg, which were given s.c. 30 min before theY-maze session or the first training in the passive avoidancetest. Y-maze test results are expressed as meanS.E.M. ofthe number of mice indicated below the columns (A).Passive avoidance results are presented as median SDL andinterquartile range (B) and percentages of animals to cri-terion (C). *P


14/16

pharmacology between neurosteroids and 1 receptor

agonists. First, Monnet et al.,52 using the NMDA-

induced [3H]norepinephrine release from rat hippo-

campal slices, reported that DH EAs potentiated,

whereas pregnenolone sulphate inhibited, the

NMD A response, behaving thus as agonist and

inverse agonist, respectively. Then, Bergeron et al.,1

using the NM DA-evoked excitatory response of

rat hippocampal pyramidal neurons, reported that

DHEA potentiated the NMDA response, but that

pregnenolone and pregnenolone sulphate were

inactive. In the present study, all neurosteroids

appeared effective and behaved similarly, attenuating

th e 2535-induced deficits. In a convergent manner,

however, progesterone, which is the neurosteroid

that inhibited the binding to 1 sites most effi-

ciently,19,48,64 behaved as a potent antagonist in all

three studies.In light of the present results, Jansen et al.17

previously reported that binding sites are signifi-

cantly decreased in the hippocampal CA 1 area in the

brains of Alzheimers disease patients. The sites

were however labelled with the non-selective 1/2site radiotracer [3H]1,3-di(2-tolyl)guanidine, and the

exact proportion of 1 sites affected is unknown.

Neverth eless, some correlation was ob served b etween

losses of CA1 pyramidal cells, which were reported

to be preferentially impaired during Alzheimers

disease, and the reduction in ligand binding to sites, suggesting a link between these sites and the

disease.17 The impact of this alteration upon the

putat ive therapeutic effectiveness of ligand s against

pathological ageing-related behavioural deficits

remains to be examined.

CONCLUSIONS

The selective 1 receptor agonists (+)-pentazocine,

PRE-084 and SA4503 alleviated the learning

and memory deficits observed after the in vivo

central administration of 2535-amyloid related

peptide in mice. These effects were blocked byco-administration of the 1 receptor antagonist

haloperidol or BMY-14802. The neurosteroids pr eg-

nenolone, DHEA, and their sulphate esters reduced

the learning deficits induced by the a dministration of

2535-amyloid peptide. Progesterone behaved as

an antagonist, since it blocked the neurosteroidal

anti-amnesic effects, without having any effect by

itself. F urthermor e a crossed pha rmacology could be

evidenced, since haloperidol blocked the neuroster-

oidal effects, whereas progesterone blocked the anti-

amnesic effects induced by the 1 receptor agonists. Itappears from these observations that selective 1receptor agonists may be of therapeutic interest

against the behavioural deficits observed in patho-

logical ageing. Furthermore, 1 receptor ligands an d

neurosteroids may act at a related receptor or, at

least, through a common mechanism.

AcknowledgementsThe authors acknowledge Dr F. P.Monnet (Le Kremlin-Bicetre, France), and Dr B. P.Lockhart (Croissy-sur-Seine, France) for helpful discussionsand support throughout the study; D rs F . J . R oman

(Fresnes, France), D. W. Parish (Sunnyvale, CA), and K.Matsuno (Osaka, Japan) for their gift of drugs. We alsothank D. Petite (Montpellier, France) for preparing theincubation of the peptide, and J. Bayle (Montpellier,France) for building the apparatus used for behaviouraltesting. This work was supported by INSERM and SantenPharmaceutical Co., Ltd.

REFERENCES

1. Bergeron R., De Montigny C. and Debonnel G . (1996) Potentiation of neuronal N MDA response induced bydehydroepiandrosterone and its suppression by progesterone: effects mediated via sigma receptors. J. Neurosci. 16,11931202.

2. Bowlby M. R. (1993) Pregnenolone sulphate potentiation ofN-methyl--aspartate receptor channels in hippocampalneurons. M olec. P harmac. 43, 813819.

3. Chen S.-Y., Wright J. W. and Barnes C. D . (1996) The neurochemical and behavioral effects of-amyloid peptide(2535). Brain Res. 720, 5460.

4. Cheney D . L., U zunov D. and Guidotti A. (1995) Pregnenolone sulfate anta gonizes dizocilpine amnesia: role forallopregnanolone. NeuroReport6, 16971700.

5. Delobette S., Privat A. and Ma urice T. (1996) In vitro aggregation facilitates the -amyloid peptide-[2535]-inducedamnesia in the rat. Eur. J. Pharmac. 319, 14.

6. Earley B., Burke M., Leonard B. E., Gouret C. J. and Junien J.-L. (1991) Evidence for an anti-amnesic effect ofJO-1784 in the rat: a potent and selective ligand for the sigma receptor. Brain Res. 546, 282286.

7. Earley B., Glennon M., Can e D., Leonard B. E. and Junien J.-L. (1995) Behavioural evidences for a modulatory ro leof the ligands JO-1784 and 2-DTG in memory processes in the rat. Soc. Neurosci. Abstr. 21, 1233.

8. Erme M., Geula C., Ransil B. J. and M esulam M. M . (1992) The acute neurotoxicity and effects upon cholinergic

axons of intracerebrally injected beta-amyloid in the rat brain. Neurobiol. Aging 13, 553559.9. Fahey J. M., Lindquist D . G., Pritchard G. A. and Miller L. G . (1995) Pregnenolone sulfate potentiation of

NM DA-mediated increases in intracellular calcium in cultured chick cortical neurons. Brain Res. 669, 183188.10. Flood J. F ., Morley J. E. and R oberts E. (1992) Memory-enhancing effects in male mice of pregnenolone an d steroids

metabolically derived from it. Proc. natn. Acad. Sci. U.S.A. 89, 15671571.11. Flood J . F. an d Ro berts E. (1988) Dehydroepiandrosterone sulfate improves memory in aging mice. Brain Res. 448,

178181.12. Fr autschy S. A., Baird A. and Cole G. M. (1991) Effects of injected Alzheimer -amyloid cores in rat brain. Proc. natn.

Acad. Sci. U.S.A. 88, 83628366.

13. Frye C. A. and Sturgis J. D. (1995) Neurosteroids affect spatial/reference, working, and long-term memory of femalerats. Neurobiol. L earning M em. 64, 8396.



15/16

14. Hanner M., M oebius F. F., F landorfer A., Knaus H . G., Striessnig J., Kempner E. and Glossman H. (1996)Purification, molecular cloning, an d expression o f th e mamma lian sigma1-binding site. Proc. natn. Acad. Sci. U.S.A.93, 80728077.

15. Irwin R . P., Lin S. Z., Rogawski M. A., Purdy R. H. a nd Paul S. M. (1994) Steroid potentiation an d inhibition ofN-methyl--aspartate receptor-mediated intracellular Ca2+ responses: structur e-activity studies. J. Pharmac. exp. Ther.271, 677682.

16. Irwin R . P., Maragakis N. J., Rogawski M. A., Purdy R. H., F arb D . H. and Paul S. M. (1992) Pregnenolone sulfate

augments NMDA receptor mediated increases in intracellular Ca 2+ in cultured rat hippocampal neurons. Neurosci.Lett. 141, 3034.

17. Jansen K. L. R ., Faull R. L. M ., Storey P. and Leslie R. A. (1993) Loss of sigma binding sites in the CA1 area of theanterior hippocampus in Alzheimers d isease correlates with CA1 pyramidal cell loss. Brain Res. 623, 299302.

18. Junien J.-L., Roman F. J., Brunelle G. and Pascaud X. (1991) JO-1784, a novel ligand, p otentiates [3H]acetylcholinerelease from rat hippocampal slices. Eur. J. Pharmac. 200, 343345.

19. Klein M. and Musacchio J. M. (1994) Effects of cytochrome P-450 ligands on the binding of [3H]dextromethorphanand sigma ligands to guinea-pig brain. In Sigma Receptors (ed. Itzhak Y.), pp. 243262. Academic, San Diego.

20. Koh J., Yang L. L. and Cotman C. W. (1990) -amyloid protein increases the vulnerability of cultured corticalneurons to excitotoxic damage. Brain Res. 533, 315320.

21. Kowall N. W., Beal M . F., Busciglio J., Duffy L . K . and Y anker B. A . (1991) A n in vivo model for theneurodegenerative effects of-amyloid and protection by substance P. Proc. natn. Acad. Sci. U.S.A. 88, 72477251.

22. Kowall N. W., McKee A. C., Yankner B. A. and Beal M. F . (1992) In vivo neurotoxicity of-amyloid (140) and the(2535) fragment. Neurobiol. Aging 13, 537542.

23. Li P.-K., R hodes M. E., Jagannathan S. and Johnson D. A. (1995) Reversal of scopolamine induced amnesia in ratsby the steroid sulfatase inhibitor estrone-3-O-sulfamate. Cognit. Brain Res. 2, 251254.

24. Maione S., Berrino L., Vitagliano S., Leyva J. and Rossi F. (1992) Pregnenolone sulfate increases the convulsantpotency of N-methyl--aspartate in mice. Eur. J. Pharmac. 219, 477479.

25. Majewska M . D., D emirgor en S., Spivak C. E. and L ondon E. D. (1990) The neurosteroid dehydroepiandrosteronesulfate is an allosteric antagonist of the GABAA receptor. Brain Res. 526, 143146.

26. Majewska M. D., M ienville J. M. and Vicini S. (1988) Neurosteroid pr egnenolone sulfate an tagonizes electrophysi-ological responses to GABA in neurons. Neurosci. Lett. 90, 279284.

27. Majewska M . D. and Schwartz R. D. (1987) Pregnenolone-sulfate: an endogenous antagonist of the -aminobutyricacid receptor complex in brain?. Brain Res. 404, 355360.

28. Martin W. R., Eades C. G., Thompson J. A., Huppler R. E. and Gilbert P. E. (1976) The effects of morphine andnalorphine-like drugs in th e non-dependent an d morphine-dependent chron ic spinal do g. J. Pharmac. exp. Ther. 197,517532.

29. Mathis C., Paul S. M. and Crawley J. N. (1994) The neurosteroid pregnenolone sulfate blocks NMD A an tagonist-induced deficits in a passive avoidance memory task. Psychopharmacology 116, 201206.30. Matsuno K., Matsunaga K . a nd Mita S. (1992) Increase of extracellular acetylcholine level in rat frontal cortex

induced by (+)N-allylnormetazocine as measured by brain microdialysis. Brain Res. 575, 315319.31. Mat suno K ., Mat sunaga K., Senda T. a nd M ita S. (1993) Increase in extracellular a cetylcholine level by sigma ligands

in rat frontal cortex. J. Pharmac. exp. Ther. 265, 851859.32. Mat suno K., Na kazawa M., Okamoto K ., Kawashima Y. and M ita S. (1996) Binding properties of SA4503, a novel

and selective 1 receptor agonist. Eur. J. Pharmac. 306, 271279.33. Matsuno K ., Senda T., Kobayashi T. and M ita S. (1995) Involvement of1 receptor in (+)-N-allylnormetazocine-

stimulated hippocampal cholinergic functions in rats. Brain Res. 690, 200206.34. Matsuno K., Senda T., Kobayashi T., Okamoto K., N akata K. and Mita S. (1997) SA4503, a novel cognitive

enhancer, with 1 receptor agonistic properties. Behav. Brain Res. 83, 221224.35. Mat suno K., Senda T., Ma tsunaga K. and Mita S. (1994) Ameliorating effects of receptor ligands on the impairment

of passive avoidance tasks in mice: involvement in the central acetylcholinergic system. Eur. J. Pharmac. 261, 4351.

36. Matsuno K ., Senda T., Matsunaga K., Mita S. and K aneto H. (1993) Similar ameliorating effects of benzomorphansand 5-HT2 antagonists and drug-induced impairment of passive avoidance response in mice: Comparison withacetylcholinesterase inhibitors. Psychopharmacology 112, 134141.

37. Mau rice T., Hiramatsu M ., Itoh J., K ameyama T., Hasegawa T. and N abeshima T. (1994) Behavioral evidence for amodulating role of ligands in memory processes. I. Attenuation of dizocilpine (MK-801)-induced amnesia. Brain

Res. 647, 4456.38. Maurice T., Hiramatsu M., Itoh J., Kameyama T., Hasegawa T. and Nabeshima T. (1994) Low dose of

1,3-di(2-tolyl)guanidine (DTG) attenuates MK-801-induced spatial working memory impairment in mice. Psycho-pharmacology 114, 520522.

39. Mau rice T., Junien J.-L. and Privat A. (1997) Dehydroepiandro sterone sulfate attenuates dizocilpine-induced learningimpairment in mice via 1 receptors. Behav. Brain Res. 83, 159164.

40. Maurice T. and Lockhart B. P. (1997) Neuroprotective and anti-amnesic potentials of sigma () receptor ligands.Prog. Neuro-Psychopharmac. biol. Psychiat. 21, 69102.

41. Maurice T., Lockhart B. P. and Privat A. (1996) Amnesia induced in mice by centrally administered -amyloid

peptides involves cholinergic dysfunction. Brain Res. 706, 181193.42. Maurice T., Lockhart B. P., Su T.-P. and Privat A. (1996) Reversion of2535 amyloid peptide-induced amnesia by

NM DA receptor-associated glycine site a gonists. Brain Res. 731, 249253.43. Mau rice T. and Privat A. (1997) SA4503, a novel cognitive enhancer with 1 receptor agonist properties, facilitates

NMDA receptor-dependent learning in mice. Eur. J. Pharmac. 328, 918.44. Mau rice T., Roman F. J., Su T.-P. and Privat A. (1996) Beneficial effects of sigma agonists on the age-related learning

impairment in the senescence-accelerated mouse (SAM). Brain Res. 733, 219230.

45. Maurice T., Su T.-P., Parish D. W., Nabeshima T. and Privat A. (1994) PRE-084, a selective PCP derivative,attenuates MK-801-induced impairment of learning in mice. Pharmac. Biochem. Behav. 49, 859869.

46. Mau rice T., Su T.-P., Parish D. W. an d Privat A. (1995) Prevention of nimodipine-induced impairment of learning bythe selective ligand PRE-084. J. neural Transm. 102, 118.

1 ligands and neurosteroids attenuate 2535-amyloid amnesia 427


16/16

47. Ma yo W., Dellu F., Robel P., Cherkaou i J., Le Moal M., Baulieu E.-E. and Simon H. (1993) Infusion of neurosteroidsinto the nucleus b asalis ma gnocellularis affects cognitive processes in the rat. Brain Res. 607, 324328.

48. McCann D. J. and Su T.-P. (1991) Solubilization an d chara cterization of haloperidol-sensitive (+)-[3H]SKF-10047binding sites (sigma sites) from rat liver membranes. J. Pharmac. exp. Ther. 257, 547554.

49. Mon net F. P., Blier P., Debonnel G. and D e Montigny C. (1991) The modulation by sigma ligands of NMD A-evoked[3H]noradrenaline release involve G i/o proteins. Soc. Neurosci. Abstr. 17, 1340.

50. Monnet F. P., Debonnel G . and De Montigny C. (1992) In vivo electroph ysiological evidence for a selective

modulation of N-methyl--aspartate-induced neuronal activation in rat CA3 dorsal hippocampus by sigma ligands.J. Pharmac. exp. Ther. 261, 123130.

51. Monnet F. P., Debonnel G., Junien J.-L. and De Montigny C. (1990) N-methyl--aspartate-induced neurona lactivation is selectively modulated by sigma receptors. Eur. J. Pharmac. 179, 441445.

52. Monnet F. P., Mahe V., Robel P. and Baulieu E.-E. (1995) Neurosteroids, via receptors, modulate the[3H]norepinephrine release evoked by N-methyl--aspartate in the rat hippocampus. Proc. natn. A cad. S ci. U.S .A. 92,37743778.

53. Ohno M. and Watanabe S. (1995) Intrahippocampal administration of (+)-SKF 10047, a sigma ligand, reversesMK-801-induced impairment of working memory in rats. Brain Res. 684, 237242.

54. Paul S. M. and Purdy R. H. (1992) Neuroactive steroids. Fedn Am. Socs exp. Biol. J. 6, 23112322.55. Quirion R., Bowen W. D., Itzhak Y., Junien J.-L., Musacchio J. M., Rothma n R. B., Su T.-P., Tam S. W. and Taylor

D. P. (1992) A proposal for the classification of sigma binding sites. Trends pharmac. Sci. 13, 8586.56. Quirion R ., Chicheportiche R., Contreras P. C., Johnson K. M ., Lodge D., Tam S. W., Woods J. H. and Zukin S. R.

(1987) Classification and nomenclature of phencyclidine and sigma receptor sites. Trends Neurosci. 10, 444446.57. Ro berts E., Bologa L., Flood J. F. and Smith G. E. (1987) Effect of dehydroepiandrosterone and its sulphate on brain

tissue in culture and memory in mice. Brain Res. 406, 357362.58. Roman F. J., Pascaud X., Duffy O. and Ju nien J.-L. (1991) Mod ulation by neuropeptide Y and peptide YY of NM DA

effects in hippocampal slices: R ole for sigma receptors. In NM DA R elated A gents: Biochemistry, Pharmacology, andBehavior(eds Kameyama T., N abeshima T. an d Domino E. F.), pp. 211218. NPP Books, Ann Arbor.

59. Ro ss S. B. (1991) Heterogeneous binding of radioligands in the ra t b rain a nd liver: possible relationships to subformsof cytochrome P-450. Pharmac. Toxicol. 468, 293301.

60. Selkoe D. J. (1991) The molecular p athology of Alzheimers disease. Neuron 6, 487498.61. Smith S. S. (1991) Progesteron e administration a ttenu ates excitator y amino acid respon ses of cerebellar Purkinje cells.

Neuroscience 42, 309320.62. Su T.-P. and Junien J.-L. (1994) Sigma receptors in the central nervous system and the periphery. In Sigma Receptors,

(ed. Itzhak Y.), pp. 2144. Academic Press, San Diego.63. Su T.-P., Wu X.-Z., Cone E. J., Shukla K ., Gund T. M., D odge A. L. and Parish D. W. (1991) Sigma compounds

derived from phencyclidine: identification of PRE-084, a new, selective sigma ligand. J. Pharmac. exp. Ther. 259,543548.64. Su T.-P., London E. D. and Jaffe J. H. (1988) Steroid binding at receptors suggest a link between endocrine, nervous,

and immune systems. Science 240, 219221.65. Wu F.-S., G ibbs T. T. and Farb D. H. (1991) Pregnenolone sulphate: a positive allosteric modulator at the

N-methyl--aspartate receptor. M olec. Pharmac. 40, 333336.66. Yamada M., N ishigami T., Nak asho K., N ishimoto Y. and Miyaji H. (1994) Relationship between sigma-like site and

progesterone-binding site o f a dult male rat liver microsomes. Hepatology 20, 12711280.

(Accepted30 July 1997)


t. maurice, t.p. su and a. privat- sigma1 (sigma1) receptor agonists and neurosteroids attenuate...

Documents