118 DRUG DEVELOPMENT RESEARCH 47:118–126 (1999)

GHELARDINI ET AL.

Research Article
Improvement of Cognitive Functions by the
Acetylcholine Releaser SM 21
Carla Ghelardini,1* Nicoletta Galeotti,1 Fulvio Gualtieri,2 Serena Scapecchi,2 and
Alessandro Bartolini1
1
Department of Pharmacology, Viale G. Pieraccini 6, Florence, Italy
2
Department of Pharmaceutical Sciences, Via G. Capponi 9, Florence, Italy

Strategy, Management and Health Policy

Venture Capital Preclinical Development Clinical Development

Enabling Preclinical Toxicology, Formulation Phases I-III Postmarketing
Technology Research Drug Delivery, Regulatory, Quality, Phase IV
Pharmacokinetics Manufacturing

ABSTRACT The effect of administration of SM 21 on memory processes was evaluated in the mouse
passive avoidance and in the rat social learning tests. SM 21 (10–20 mg kg–1 i.p.) prevented amnesia in-
duced by scopolamine and dicyclomine as tested by the mouse passive avoidance test and prevented
memory disruption by AF-64A and benehexol ascertained by the rat passive avoidance test. Both SM 21
enantiomers were able to abolish dicyclomine-induced amnesia in mice. SM 21, starting from the dose of
10 mg kg–1 i.p., antagonized the memory impairment produced by mecamylamine, baclofen, and diphen-
hydramine in mice, as well as amnesia induced by diazepam in rats. SM 21, at doses ranging between 10
and 30 mg kg–1 i.p., prevented memory reduction in mice by hypoxia in the passive avoidance test. In the
social learning test, SM 21 (10 mg kg–1 i.p.) injected in adult rats reduced the duration of active exploration
of a familiar partner in the second session of the test. SM 21 prevented amnesia in both mice and rats
comparable to that of the cholinesterase inhibitor physostigmine (0.2 mg kg–1 i.p.), the M1 selective agonist
AF-102B (10 mg kg–1 i.p.), and the nootropic drug piracetam (30 mg kg–1 i.p.). These results demonstrated
the ability of SM 21 to modulate memory functions and suggests that SM 21 could be useful in the treat-
ment of cognitive deficits. Drug Dev. Res. 47:118–126, 1999. © 1999 Wiley-Liss, Inc.

Key words: SM 21; ACh releaser; memory; amnesia; cholinergic system

INTRODUCTION and in Alzheimer patients, has been reported [Deutsch,

Cholinergic activity has long been associated with 1971; Bartus and Johnson, 1976]. Ghelardini et al. [1997b]
memory processes. Morphological and neurochemical reported that the antimuscarinic compound atropine, at
studies of Alzheimer’s disease, the major type of demen- very low doses, was able to prevent amnesia in mice and
tia, have revealed marked decreases in the cholinergic in particular this beneficial effect was attributable to the
innervation of the cortex and hippocampus [Bartus et al., R-(+)-enantiomer of atropine, R-(+)-hyoscyamine, since
1982; Mash et al., 1985; Whitehouse, 1986]. Drugs in- S-(–)-hyoscyamine was ineffective. It is interesting to note
volving cholinergic stimulation alleviate cognitive dys- that this antiamnesiac activity, different from that pro-
functions in Alzheimer’s disease [Bartus et al., 1982] and, duced by direct muscarinic agonists and cholinesterase
in particular, M1-selective agonists have been proposed inhibitors, was not accompanied by typical cholinergic
as a promising treatment strategy in this pathology [Mash symptomatology, such as tremors, lacrimation, sialorrhea,
et al., 1985; Whitehouse, 1986; Fisher et al., 1989;
Gualtieri et al., 1995]. On the other hand, cholinergic Contract grant sponsor: MURST.
blockade produces significant impairments of cognitive *Correspondence to: Dr. Carla Ghelardini, Dept. of Pharma-
functions. A delay-dependent disruption following treat- cology, Viale G. Pieraccini, 6, I-50139 Florence, Italy. E-mail:
ment with scopolamine and atropine, which appeared to ghelard@server1.pharm.unifi.it
resemble that occurring spontaneously in aged subjects Received 17 January 1999; Accepted 24 February 1999

© 1999 Wiley-Liss, Inc.

 SM 21 IMPROVEMENT OF COGNITION 119

diarrhea etc., Bartolini et al. [1994], investigating the 1:10 India ink and their brains examined macroscopically
paradoxical effect of atropine, using microdialysis tech- after sectioning. The accuracy of the injection technique
niques demonstrated that R-(+)-hyoscyamine, at was evaluated and 95% correct.
antiamnesic doses, produced an increase in acetylcho-
line (ACh) release from the rat cerebral cortex in vivo. Passive-Avoidance Test
On these bases, a synthetic program to modify the chemi- The test was performed according to the step-
cal structure of atropine was started, aimed at develop- through method described by Jarvik and Kopp [1967].
ing cholinergic amplifiers as endowed with ameliorated The apparatus consisted of a two-compartment acrylic
antiamnesic activity as atropine, but lacking the cholin- box with a lighted compartment connected to a darkened
ergic side effects of atropine. These compounds would, one by a guillotine door. As soon as they entered the dark
therefore, be potentially useful in pathological conditions compartment, mice received a punishing electrical shock
such as Alzheimer’s disease, that are characterized by (0.5 mA, 1 sec). The latency times for entering the dark
cholinergic deficit. Of the many compounds synthesized compartment were measured in the training test and af-
and studied, the racemate [Gualtieri et al., 1994] and the ter 24 h in the retention test. For memory disruption,
enantiomers [Romanelli et al., 1996] of the compound animals were either exposed to a hypoxic environment
labeled SM 21 (3-α-tropanyl 2-[4-(Cl-phenoxy)] buta- (5% O2 in water-saturated nitrogen) for 8 min up to 30
noate) showed the best pharmacological profile. sec before passive avoidance training or injected with
SM 21 antiamnesic properties were investigated in amnesic drugs. Scopolamine, dicyclomine, diazepam,
mice and rats using the passive avoidance and social learn- benzhexol, baclofen, mecamylamine, and diphenhy-
ing tests, whereas the incidence of behavioral side ef- dramine were i.p. injected immediately after the train-
fects was detected by the rota-rod test and Animex ing session, whereas AF-64A was i.c.v. injected 4 h before
apparatus. the training session. To improve memory, animals were
treated 20 min before the training session with SM 21,
MATERIALS AND METHODS piracetam, AF-102B, or physostigmine. The drug admin-
Animals istration schedule was chosen on the basis of prelimi-
nary experiments in which the time-course for every
Male Swiss albino mice (23–30 g) and Wistar rats compound was determined. The maximum entry latency
(90–110 g, 200–300 g, 350–450 g) from Morini (San Polo allowed in the training session was 30 sec for mice and
d’Enza, Italy) breeding farms were used. Fifteen mice 150 sec for rats, whereas in the retention session the en-
and four rats were housed per cage. The cages were trance latency allowed was 120 sec and 20 min, respec-
placed in the experimental room 24 h before the test for tively, for mice and rats. The memory degree of received
acclimatization. The animals were kept at 23 ± 1°C with punishment was expressed as latencies recorded in the
a 12-h light/dark cycle, lights on at 7 AM, with food and retention and training sessions.
water ad libitum. All experiments were carried out ac-
cording to the guidelines of the European Community Social Learning Test
Council for experimental animal care. The social learning test was performed according
to Mondadori et al. [1992]. Male Wistar rats (350–450 g)
Intracerebroventricular Injection Technique
were used throughout the experiments and juvenile males
Intracerebroventricular (i.c.v.) administration was (90–110 g) were used as social stimuli. All the adult ani-
performed under ether anesthesia using isotonic saline mals were housed individually and placed in the testing
as solvent, according to the method described by Haley room at least 24 h before the experiment. On the day
and McCormick [1957] for mice and which we adapted preceding the experiment, adult rats were handled to
for rats. Briefly, during anesthesia, mice and rats were become familiar with the operator. Juvenile rats were
grasped firmly by the loose skin behind the head. A 0.4 housed four per cage and brought into the testing room
mm external diameter, hypodermic needle attached to a on the day of the experiment. Experimental sessions were
10 µl syringe was inserted perpendicularly through the always conducted between 10 AM and 2 PM. Each mature
skull at a depth of no more than 2 mm into the brain of male rat was tested in its home cage. The first day of the
the mouse and 4 mm into the brain of the rat, where 5 µl experiment, a juvenile rat was introduced into the adult
(mice) or 10 µl (rats) were then administered. The injec- male’s cage and the time spent in social investigatory
tion site was 1.5 mm (mice) or 2.5 mm (rats) from either behavior by the adult male within a 5-min fixed interval
side of the midline on a line drawn through to the ante- was recorded. Social investigatory behavior was defined
rior base of the ears. To ascertain that the drugs were as being proximally oriented to the juvenile or in direct
administered exactly into the cerebral ventricle, some contact while sniffing, following, nosing, grooming, or
mice and rats were i.c.v. injected with 5–10 µl of diluted generally inspecting any body surface of the juvenile.
120 GHELARDINI ET AL.

After 24 h, either the same juvenile or an unfamiliar one pared from 10 mM acetylethylcholine mustard hydro-
was placed again into the mature male’s cage and social chloride. It was dissolved in distilled water (pH 11.5–
investigatory behavior was recorded in a 5-min interval. 11.7), and the solution was vigorously stirred for 20 min.
SM 21 and piracetam were i.p. injected 20 min before This solution was diluted with 0.9% NaCl and finally ti-
the first session of the experiment. trated with HCl to adjust the pH to 7.3–7.4. The cere-
bral ventricles were then bilaterally infused with 2.0 nmol
Spontaneous Activity Meter (Animex) of AF-64A. A total volume of 10 µl was delivered to each
Locomotor activity in rats was quantified using an side. Drug concentrations were prepared in such a way
Animex activity meter Type S (LKB, Farad, Sweden) set that the necessary dose could be administered in a vol-
to maximum sensitivity. Every movement of rats, which ume of 5 µl per mouse by i.c.v. injection and 10 ml kg–1
were placed on the top of the Animex activity meter, pro- by i.p. injection.
duced a signal due to variation in inductance and capac-
ity of the apparatus resonance circuit. Signals were then Statistical Analysis
automatically converted to numbers. On the day of the All experimental results are given as the mean ±
experiment, the rats were treated and the cage, contain- SEM. Analysis of variance (ANOVA), followed by Fisher’s
ing three rats, was put on the measuring platform. Activ- Protected Least Significant Difference (PLSD) procedure
ity counts were made for 5 min at 15-min intervals for 45 for post-hoc comparison, was used to verify significance
min (total of three sessions) starting immediately after between two means. Data were analyzed with the
injection of the drug. Because of the arbitrary scale StatView software for the Macintosh. P values less than
adopted to quantify movements, drug-treated rats were 0.05 were considered significant.
always compared with saline-treated ones.
Rota-Rod Test RESULTS

The apparatus consisted of a base platform and a Prevention by SM 21 of Amnesia Induced by

rotating rod of 3 cm diameter with a nonslippery surface. Antimuscarinic Drugs
This rod was placed at a height of 15 cm from the base. Pretreatment of ACh releaser SM 21 prevented
The rod, 30 cm in length, was divided into five equal amnesia induced by the administration of antimuscarinic
sections by six disks. Thus, up to five mice were tested drugs in both mice and rats (Figs. 1, 2).
simultaneously on the apparatus, with a rod-rotating SM 21 dose-dependently prevented scopolamine
speed of 16 r.p.m. The integrity of motor coordination (1 mg kg–1 i.p.; Fig. 1A) and dicyclomine (10 mg kg–1 i.p.;
was assessed on the basis of endurance time of the ani- Fig. 1B) in the mouse passive avoidance test. SM 21, at
mals on the rotating rod. One day before the test, the the dose of 1 and 5 mg kg–1 i.p. was completely ineffec-
animals were trained twice. On the day of the test, only tive, whereas starting from a dose of 10 mg kg–1 i.p. it
the mice that were able to stay balanced on the rotating prevented antimuscarinic amnesia, reaching entrance
rod between 70 and 120 sec (cut-off time) were selected latency values comparable to those produced by saline-
for testing. The performance time was measured before treated mice. SM 21 was also able to prevent amnesia
and at various times after treatment. induced by AF-64A (2 nmol per rat i.c.v.; Fig. 2A) and
benzhexol (10 mg kg–1 i.p.; Fig. 2B), as demonstrated by
Reagents and Compounds the rat passive avoidance test. Similar to the profile ob-
The following drugs were used: SM 21 racemate served in mice, the antiamnesic effect of SM 21 reached
was prepared according to Gualtieri et al. [1994]; R-(+)- statistical significance at 10 mg kg–1 i.p..
SM 21 and S-(–)-SM 21 were prepared according to SM 21 prevented antimuscarinic amnesia in both
Romanelli et al. [1996]; dicyclomine hydrochloride (Le mice and rats comparably to that exerted by the cholinest-
Petit, Italy); benzhexol (Cyanamid); diazepam (Valium, erase inhibitor physostigmine (0.2 mg kg–1 i.p.; Figs. 1A,B,
Roche); mecamylamine hydrochloride; acetylethylcholine 2B) and the M1 selective agonist AF-102B (10 mg kg–1
mustard hydrochloride (RBI; Natick, MA); AF-102B (Inst. i.p.; Fig. 2A). The maximum antiamnesic effect of SM 21
for Neurobiol. Res., Bruxelles, Belgium); diphenhy- (20 mg kg–1 i.p.) was also equal to that produced by the
dramine hydrochloride (De Angeli); scopolamine well-known nootropic drug piracetam (30 mg kg–1 i.p.)
hydrobromide; baclofen, piracetam; and physostigmine (Fig. 1A,B). However, at active doses SM 21 did not en-
hemisulphate (all Sigma Chemical Co., St. Louis, MO). hance the entrance latency in unamnesic mice in com-
Drugs were dissolved in isotonic (NaCl 0.9%) sa- parison with the control group (Figs. 1A, 2). There were
line solution, with the exception of diazepam, which was no differences observed in the various entrance laten-
dissolved in a water and dimethyl sulphoxide (DMSO) cies of every group in the training session of the passive
(3:1) vehicle immediately before use. AF-64A was pre- avoidance test (Figs. 1, 2).
 SM 21 IMPROVEMENT OF COGNITION 121

Fig. 1. Dose–response curves of SM 21 in comparison with piracetam Fig. 2. Prevention by SM 21 of amnesia induced by AF-64A (2 nmol icv)
and physostigmine on amnesia induced by scopolamine (1.5 mg kg i.p.) in comparison with AF-102B (A) and benzhexol (10 mg kg–1 i.p.) in com-
(A) and dicyclomine (10 mg kg–1 i.p.) (B) in mouse passive avoidance test. parison with physostigmine (B) in rat passive-avoidance test. AF-64A was
SM 21, piracetam, and physostigmine were administered 20 min before injected 4 h before training session, while SM 21, AF-102B, and physostig-
training session, while scopolamine and dicyclomine were injected im- mine 20 min before training session. Benzhexol was administered imme-
mediately after. The number of mice is inside the column. *P < 0.01 in diately after training session. The number of rats is inside the column. *P <
comparison with antimuscarinic-treated mice. 0.01 in comparison with AF-64A or benzhexol-treated rats.

The antiamnesic effect of the two enantiomers of

SM 21, R-(+)-SM 21 and S-(–)-SM 21 was evaluated in
mecamylamine (20 mg kg–1 i.p.), baclofen (2 mg kg–1 i.p.),
the mouse passive avoidance test (Fig. 3). Both enanti-
and diphenhydramine (20 mg kg–1 i.p.) in the mouse pas-
omers dose-dependently were able to prevent dicyclo-
sive avoidance test (Fig. 4A) as well as the amnesia in-
mine-induced amnesia, even if R-(+)-SM 21 was slightly
duced by diazepam (1 mg kg–1 i.p.) in the rat passive
more effective than S-(–)-SM 21. As a matter of fact, the
avoidance test (Fig. 4B).
lowest active doses of R-(+)-SM 21 and S-(–)-SM 21
SM 21, at 10 mg kg–1 i.p., partially prevented diaz-
were, respectively, 10 and 20 mg kg–1 i.p. (Fig. 3). For
epam-induced amnesia, even if statistical significance was
both enantiomers, the maximum antiamnesic effect was
not reached. The dose of 20 mg kg–1 i.p. increased the
reached at 20 mg kg–1 i.p., obtaining an entrance latency
entrance latency in the retention session up to a value
value comparable to that in saline-treated mice.
comparable to that produced by control animals (Fig. 4B).
Conversely, the amnesia induced by mecamylamine,
Prevention by SM 21 of Amnesia Induced by
baclofen, and diphenhydramine was prevented by lower
Mecamylamine, Baclofen, Diphenhydramine,
doses of SM 21, with 10 mg kg–1 i.p. being the first active
and Diazepam
dose (Fig. 4A). The maximum antiamnesic effect was
The administration of SM 21 dose-dependently reached at the dose of 20 mg kg–1 i.p.
antagonized the memory disruption produced by SM 21 was active in facilitating memory similar to
122 GHELARDINI ET AL.

Fig. 3. Dose–response curves of R-(+)- SM 21 and S-(–)- SM 21 on

amnesia induced by dicyclomine in the mouse passive avoidance test.
Both enantiomers were administered 20 min before training session while
dicyclomine was injected immediately after. Each column represents the
mean of at least 18 mice. ^P < 0.05; *P < 0.01 in comparison with
dicyclomine-treated mice.

the nootropic drug piracetam, regardless of the amnesic

drug used (Fig. 4A,B).
Prevention by SM 21 of the Amnesia
Induced by Hypoxia
A strong reduction in memory was obtained in mice Fig. 4. Dose–response curves of SM 21 in comparison with piracetam
on amnesia induced in mouse by mecamylamine, baclofen, and diphen-
exposed for 8 min to 5% O2 in water-saturated nitrogen hydramine (A) and in rat by diazepam (B) in passive avoidance test. SM
in the passive avoidance test (Fig. 5). The intensity of 21 and piracetam were administered 20 min before training session, while
amnesia was comparable to that produced by all the am- the other drugs were injected immediately after. Each column represents
nesic drugs used in this study. At doses ranging between the mean of at least 18 mice or 8 rats. ^P < 0.05; *P < 0.01 in compari-
son with amnesic drug-treated animals.
10–30 mg kg–1 i.p., SM 21 was able to prevent the amne-
sic effect of this low oxygen concentration with an effi-
cacy comparable to that produced by piracetam, used as
a nootropic reference drug (Fig. 5). On the other hand,
SM 21 was inactive in preventing hypoxia-induced am-
nesia at lower doses (Fig. 5). At the highest active dose
SM 21 did not increase the entrance latency in
nonhypoxic mice in comparison with the saline group
(Fig. 5).
Effect of SM 21 on the rat social learning test
Adult rats were treated with SM 21 (5–10 mg kg–1
i.p.) and piracetam (30 mg kg–1 i.p.), or saline 20 min
before the first session of the test. In the adult animals
treated with SM 21 (10 mg kg–1 i.p.), the duration of ac-
tive exploration of the familiar partner at the second pair-
ing at 24 h was shortened in comparison with Fig. 5. Comparison of prevention by SM 21 and by piracetam of
hypoxia-induced amnesia in the mouse passive avoidance test. Mice
saline-treated rats (Fig. 6). No curtailment was observed
were exposed to hypoxic environment (5% O2, 95% N2) for 8 min.
if an unknown partner was presented (Fig. 6). At the dose SM 21 and piracetam were injected 20 min before the training test.
of 5 mg kg–1 i.p., SM 21 was ineffective. In the same The number of mice is inside the column. *P < 0.01 in comparison
experimental conditions, a nootropic drug such as with hypoxic-treated mice.
 SM 21 IMPROVEMENT OF COGNITION 123

dicyclomine (Fig. 7). By contrast, in the same experimen-

tal conditions piracetam (30 mg kg–1 i.p.), used as refer-
ence drug, still exerted antiamnesic activity (Fig. 7).
Subacute treatment with SM 21 (10 mg kg–1 i.p.) did not
evidence any loss of body weight nor the typical symp-
tomatology of withdrawal syndrome.
Effect of SM 21 on Mouse Rota-Rod Test and
Rat Animex Apparatus
It should be noted that SM 21 elicited its modu-
latory effect on cognitive processes without changing ei-
ther gross behavior or motor coordination and spontaneous
motility as revealed, respectively, by the mouse rota-rod
test (Table 1) and the rat Animex apparatus (Fig. 8). SM 21,
administered at the highest active doses, did not reduce
the endurance time on the rotating rod in comparison
with saline-treated mice (Table 1). By contrast, in the
same experimental conditions physostigmine (0.2 mg
Fig. 6. Comparison of the effect of SM 21 and of piracetam in the rat kg–1 i.p.) reduced time spent by the animals on the rotat-
social learning test. SM 21 and piracetam were administered 20 min be- ing rod (Table 1).
fore the first session. The number of rats is inside the column. *P < 0.01 The spontaneous motility of rats was unmodified
in comparison with saline-treated rats. by SM 21 administration (30 mg kg–1 i.p.) as revealed by
the Animex apparatus in comparison with saline-treated
rats (Fig. 8).
piracetam reduced the time spent on exploratory behav-
ior (Fig. 6). All drugs used did not modify the duration of DISCUSSION
active exploration in comparison with saline-treated rats
The present results describe acute and subacute
(data not shown).
effects observed with SM 21 on experimentally impaired
Effect of Subacute Treatment With SM 21 memory in mice and unimpaired learning in rats. SM 21
has been demonstrated to ameliorate cognitive processes
SM 21 induced tolerance after repeated adminis-
not only by preventing amnesia induced by pharmaco-
tration. SM 21, injected twice daily for 2 weeks at a dose
logical treatments or exposure to hypoxic environment
at which it demonstrates full antiamnesic activity (10 mg
in the passive avoidance test, but also by producing a
kg–1 i.p.), lost the ability to prevent amnesia induced by
procognitive activity in a social learning task.
That stimulation of the cholinergic system improves
cognitive processes has long been observed [Coyle, 1995].
On the other hand, a blockade of the cholinergic system
produces a disruption of memory functions. The admin-
istration of scopolamine, an unselective muscarinic ACh
receptor antagonist, results in impaired learning and
memory in humans [Frumier et al., 1976] and animals
[Dilts and Berry, 1967; Levin and Bowman, 1986]. Ani-
mals treated with the M1 selective antagonist pirenzepine
[Hammer et al., 1980] had impaired passive avoidance
learning in mice [Caufield, 1993] and impaired spatial
learning [Hagan et al., 1987; Hunter and Roberts, 1988],
radial arm maze performance [Sala et al., 1991], and ac-
tive avoidance acquisition [Sen and Bhattacharya, 1991]
in rats. Moreover, the M1 selective antagonists dicyclo-
Fig. 7. Comparison of the effect of subacute treatment with SM 21 and mine [Nilvebrant and Sparf, 1986] and S-(–)-ET-126
piracetam on dicyclomine-induced amnesia in mouse passive-avoidance
[Ghelardini et al., 1996] were able to induce amnesia in a
test. SM 21 (10 mg kg–1 i.p.) and piracetam (30 mg kg–1 i.p.) were in-
jected twice daily for 14 days; on the day of the experiment, both were
mouse passive avoidance task [Ghelardini et al., 1997b;
administered 20 min before training. The number of mice is inside the Matucci et al., 1997]. Disruption of the cholinergic sys-
column. *P < 0.01 in comparison with dicyclomine-treated mice. tem can also be obtained by the use of cholinotoxins. The
124 GHELARDINI ET AL.

TABLE I. Effect of SM 21 in Comparison With Physostigmine in the Rota-Rod Test

Endurance time on rota-rod(s)
After treatment
Before treatment 15 min 30 min 45 min
Saline ip 103.5 ± 5.6 (11) 99.7 ± 7.2 (11) 102.9 ± 5.9 (11) 96.3 ± 4.9 (11)
SM 21 30 mg kg–1 ip 98.5 ± 6.9 (10) 103.6 ± 7.7 (10) 96.2 ± 8.3 (10) 101.2 ± 8.6 (10)
Physostigmine 0.2 mg kg–1 ip 102.4 ± 4.5 (10) 65.3 ± 8.1* (10) 65.7 ± 7.3* (10) 87.4 ± 8.4* (10)
*P < 0.05 in comparison with the respective pretest value. The number of mice is shown in parentheses.

i.c.v. injection of the ethylcholine aziridinium ion (AF- of GABAB receptors by baclofen disrupts memory after
64A), a selective cholinotoxin, produced a deficient per- systemic, intraamygdala or intraseptal administration
formance in a rat passive avoidance task and a delayed [Swartzwelder et al., 1987; Castellano et al., 1989;
alteration task in the T-maze test [Nakahara et al., 1988]. Stackman and Walsh, 1994]. Furthermore, benzodiaz-
Furthermore, the administration of nicotinic ACh recep- epines have long been known to impair many forms of
tor antagonists, such as mecamylamine, produces a dose- learning in many species, including humans [Izquierdo
dependent impairment of performance in the passive and Medina, 1991]. The antihistaminics are known to
avoidance test [Elrod and Buccafusco, 1991]. exert a variety of effects on the central nervous system.
It has been demonstrated by microdialysis studies Central depression usually accompanies therapeutic
that SM 21 is able to increase ACh release from rat cere- doses of the H1 antagonists, which appears to be related
bral cortex [Ghelardini et al., 1997a] and hippocampus to occupancy of cerebral H1 receptors; impairment of
(data not shown), two cerebral structures highly involved cognitive functions is a common manifestation [Simons
in the modulation of cognitive processes [Bartus et al., and Simons, 1994]. Furthermore, the administration of
1982]. Therefore, it is not unexpected that SM 21 is able the cerebral H1 antagonist diphenhydramine induces
to prevent amnesia induced by selective and unselective amnesia in animals also [Kamei et al., 1990; Galeotti et
antimuscarinic drugs, by disruption of the cholinergic al., 1998].
neurones by the use of the antimuscarinic toxin AF-64A, Cerebral hypoxia is associated with a series of mo-
as well as by the administration of a nicotinic antagonist. lecular events which can culminate in neuronal cell death.
Amnesia can also be obtained by modulating neu- Exposure of cerebral structure to an environment with a
rotransmitter systems different from the cholinergic. low concentration of oxygen is well known to produce am-
GABA is the main inhibitory neurotransmitter in the brain nesia. The administration of cholinomimetics or nootropic
and it plays an important role in learning and memory. drugs can prevent the amnesic effect produced by hypoxia
The activation of GABAA receptors impairs memory per- in both laboratory animals [DeNoble et al., 1986; Coyle,
formance [Jerusalinsky et al., 1994] and the stimulation 1995] and humans [Saletu and Grünberger, 1984].
SM 21 was able to prevent amnesia induced by the
administration of baclofen, diazepam, diphenhydramine,
and hypoxia. Thus, SM 21 counteracts amnesia not only
induced by antimuscarinic drugs, but also that obtained
independently from a cholinergic blockade.
SM 21 is also endowed with antinociceptive prop-
erties [Ghelardini et al., 1997a], but the analgesia and
antiamnesia seem to be unrelated, since the first
antiamnesic dose was lower than that able to enhance
the pain threshold. A complete prevention of amnesia
was, in fact, obtained at a dose (10 mg kg–1) that was
weakly analgesic only in the hot-plate test [Ghelardini et
al., 1997a]. The time-course of the antiamnesic activity
of SM 21 was equal to that observed for its antinociceptive
action, reaching its maximum between 15–30 min after
injection (data not shown). Therefore, in the learning and
memory experiments SM 21 was administered 20 min
before the training session.
Fig. 8. Lack of effect of SM 21 on rat spontaneous motility. Vertical In the passive avoidance test, an improvement of
lines show SEM. Each column represents the mean of eight rats. cognition in animals which have no memory impairment
 SM 21 IMPROVEMENT OF COGNITION 125

is difficult to demonstrate. As a matter of fact, not only DeNoble VJ, Repetti SJ, Gelpke LW, Wood LM, Keim KL. 1986.
SM 21 but also well-known nootropic drugs, such as Vinpocetine: nootropic effects on scopolamine-induced and hypoxia-
induced retrieval deficits of a step-through passive avoidance re-
piracetam and aniracetam, or cholinomimetics, such as sponse in rats. Pharmacol Biochem Behav 24:1123–1128.
physostigmine and oxotremorine, do not show any Dilts SL, Berry CA. 1967. Effect of cholinergic drugs on passive avoid-
memory facilitation in unamnesic animals [Gouliaev and ance in the mouse. J Pharmacol Exp Ther 158:279–285.
Senning, 1994; Coyle, 1995]. However, a procognitive Deutsch JA. 1971. The cholinergic synapse and the site of memory.
activity of SM 21 was unmasked by using a social learn- Science 174:788–794.
ing test in which adult rats with unimpaired memory were Elrod K, Buccafusco JJ. 1991. Correlation of the amnestic effects of
used. SM 21, as well as piracetam, improved cognitive nicotinic antagonists with inhibition of regional brain acetylcholine
performance by prolonging the time spent by rats delet- synthesis in rats. J Pharmacol Exp Ther 258:403–409.
ing mnemonic information. Fisher A, Brandeis R, Pittel Z, Karton I, Sapir M, Dachir S, Levy A,
In the first session, the latency to enter the dark Heldman E. 1989. (±)-cis-2-methyl-spiro(1,3-oxathiolane-5,3′)
compartment of the light–dark box in the passive avoid- quinuclidine (AF102B): a new M1 agonist attenuates cognitive dys-
functions in AF64A-treated rats. Neurosci Lett 102:325–331.
ance test, as well as the duration of exploratory activity
Frumier MJ, Herckar VR, Jarvik ME. 1976. Amnesic actions of diaz-
in the social learning test, were not modified by the ad- epam and scopolamine in man. Anesthesiology 45:406–410.
ministration of SM 21. This observation was confirmed
Galeotti N, Ghelardini C, Bartolini A. 1998. Effect of pertussis toxin
by evaluation of the motor coordination in mice and spon- on baclofen and diphenhydramine induced amnesia. Psycophar-
taneous motility in rats. SM 21, at the highest doses used, macology 136:328–334.
did not impair motor coordination as revealed by the rota- Ghelardini C, Galeotti N, Gualtieri F, Romanelli MN, Bartolini A. 1996.
rod test or modify spontaneous motility as indicated by S-(–)-ET126: a potent and selective M1 antagonist in vivo and in
the Animex apparatus. Furthermore, SM 21 did not elicit vitro. Life Sci 58:991–1000.
the typical cholinergic symptoms (tremors, sialorrhea, Ghelardini C, Galeotti N, Gualtieri F, Bellucci C, Manetti D, Giotti A,
diarrhea, rhinorrhea, lacrimation, etc.) produced by in- Malmberg-Aiello P, Galli A, Bartolini A. 1997a. Antinociceptive pro-
jection of direct postsynaptic muscarinic agonists. In file of SM 21: a novel analgesic with a presynaptic cholinergic mecha-
nism of action. J Pharmacol Exp Ther 82:430–439.
other words, SM 21 is able to counteract amnesia and to
Ghelardini C, Gualtieri F, Romanelli MN, Angeli P, Pepeu G,
exert a procognitive activity in a more physiological man- Giovannini MG, Casamenti F, Malmberg-Aiello P, Giotti A, Bartolini
ner than the cholinergic activators, such as physostigmine A. 1997b. Stereoselective increase in cholinergic transmission by
or AF-102B, used as reference drugs. R-(+)-hyoscyamine. Neuropharmacology 36:281–294.
In conclusion, these results indicate the ability of Gouliaev AH, Senning A. 1994. Piracetam and other structurally re-
SM 21 to modulate memory processes. On these bases, lated nootropics. Brain Res Rev 19:180–222.
SM 21 could be considered a new potential antiamnesic Gualtieri F, Bottalico C, Calandrella A, Dei S, Giovannoni P, Mealli S,
drug useful in the treatment of cognitive disorders. Romanelli MN, Scapecchi S, Teodori E, Galeotti N, Ghelardini C,
Bartolini A, Giotti A. 1994. Presynaptic cholinergic modulators as
ACKNOWLEDGMENTS potent nootropic and analgesic drugs. II. 2-Phenoxy, 2-phenylthio
and 2-phenylamino alkanoic acid esters. J Med Chem 37:1712–1719.
The authors thank Mary Forrest for linguistic revi-
Gualtieri F, Dei S, Manetti D, Romanelli MN, Scapecchi S, Teodori E.
sion of the manuscript.
1995. The medicinal chemistry of Alzheimer’s and Alzheimer-like
REFERENCES diseases with emphasis on the cholinergic hypothesis. Il Farmaco
50:489–503.
Bartolini A, Ghelardini C, Giovannini MG, Casamenti F, Malmberg-
Aiello P, Pepeu G, Giotti A. 1994. Modulators of ACh release as Hagan JJ, Jansen JHM, Broekkamp CLE. 1987. Blockade of spatial
potent cognition enhancers and analgesics: pharmacodynamic stud- learning by the M1 muscarinic antagonist pirenzepine. Psychophar-
ies. Proceeding of XXVII Symposium of the Italian Pharmacologi- macology 93:470–476.
cal Society, Turin, September 25–29, p 29. Haley TJ, McCormick WG. 1957. Pharmacological effects produced
Bartus RT, Johnson HR. 1976. Short-term memory in the rhesus mon- by intracerebral injection of drugs in the conscious mouse. Br J
key: disruption from the anticholinergic scopolamine. Pharmacol Pharmacol Chemother 12:12–15.
Biochem Behav 5:39–46. Hammer R, Berrie CP, Birdsall NJM, Burgen AS, Hulme EC. 1980.
Bartus RT, Dean RL, Beer B, Lippa AS. 1982. The cholinergic hy- Pirenzepine distinguishes between different subclasses of muscar-
pothesis of geriatric memory dysfunctions. Science 217:408–417. inic receptors. Nature 283:90–91.
Castellano C, Brioni JD, Nagahara AH, McGaugh JL. 1989. Post-train- Hunter AJ, Roberts FF. 1988. The effect of pirenzepine on spatial
ing systemic and intra-amygdala administration of the GABAB ago- learning in the Morris water maze. Pharmacol Biochem Behav
nist baclofen impairs retention. Behav Neural Biol 52:170–179. 30:519–523.
Caulfield MP. 1993. Muscarinic receptors—characterization, coupling Izquierdo I, Medina JH. 1991. GABAA receptor modulation of memory:
and function. Pharmacol Ther 58:319–379. the role of endogenous benzodiazepines. TIPS 12:260–265.
Coyle MJ. 1995. A cholinergic hypothesis for Alzheimer’s disease. In: Jarvik ME, Kopp R. 1967. An improved one-trial passive avoidance
Meyer L, Nordeberg GH, editors. Learning and memory molecu- learning situation. Psychol Rep 21:221–224.
lar bases. London: Pergamon Press. p 11–32. Jerusalinsky D, Quillfeldt JA, Walz R, Da Silva RC, Silva MB, Bianchin
126 GHELARDINI ET AL.
M, Schmitz P, Zanatta MS, Ruschel AC, Paczko N, Medina JH, Romanelli MN, Bartolini A, Bertucci C, Dei S, Ghelardini C,
Izquierdo I. 1994. Effect of the infusion of the GABAA receptor ago- Giovannini MG, Giotti A, Pepeu G, Scapecchi S, Teodori E. 1996.
nist, muscimol, on the role of the entorhinal cortex, amygdala, and Synthesis and enantioselectivity of the enantiomers of PG-9 and
hippocampus in memory processes. Behav Neural Biol 61:132–138. SM 21, new potent analgesic drugs. Chirality 8:225–233.
Kamei C, Chung YH, Tasaka K. 1990. Influence of certain H1-blockers, Sala M, Braida D, Calcaterra P, Leone MP, Comotti FA, Gianola S,
on the step-through active avoidance in rats. Psychopharmacology Gori E. 1991. Effect of centrally administered atropine and
102:312–318. pirenzepine on radial arm maze performance in the rat. Eur J
Levin ED, Bowman RE. 1986. Scopolamine effects on Hamilton Pharmacol 194;45–49.
search task performance in monkeys. Pharmacol Biochem Behav Saletu B, Grünberger J. 1984. The hypoxia model in human psycho-
24:819–821. pharmacology: neurophysiological and psychometric studies with
Mash DC, Flynn DD, Potter LT. 1985. Loss of M2 muscarine recep- aniracetam i.v. Hum Neurobiol 3:171–181.
tors in the cerebral cortex in Alzheimer’s disease and experimental Sen AP, Bhattacharya SK. 1991. Effect of selective muscarinic recep-
cholinergic denervation. Science 228:115–117. tor agonists and antagonists on active-avoidance learning acquisi-
Matucci R, Ghelardini C, Galeotti N, Teodori E, Gualtieri F, Bartolini tion in rats. Ind J Exp Biol 29:136–139.
A. 1997. Amnesic properties of the selective M1 antagonist S-(–)- Simons FER, Simons KJ. 1994. The pharmacology and use of H1-re-
ET-126. 27th Annual Meeting Soc. for Neurosci., New Orleans, ceptor-antagonist drugs. N Engl J Med 330:1663–1670.
Louisiana, October 25–30, 23:213. Stackman RW, Walsh TJ. 1994. Baclofen produces dose-related work-
Mondadori C, Preiswerk G, Jaekel J. 1992. Treatment with a GABAB ing memory impairments after intraseptal injection. Behav Neural
receptor blocker improves the cognitive performance of mice, rats Biol 61:181–185.
and rhesus monkeys. Pharmacol Commun 2:93–97. Swartzwelder HS, Tilson HA, Mclamb RL, Wilson WA. 1987. Baclofen
Nakahara N, Iga Y, Mizobe F, Kawanishi G. 1988. Effect of the disrupts passive avoidance retention in rats. Psychopharmacology
intracerebroventricular injection of AF64A on learning behaviors 92:398–401.
in rats. Jpn J Pharmacol 48:121–130. Whitehouse PJ. 1986. Neuronal loss and neurotransmitter receptor
Nilvebrant L, Sparf B. 1986. Dicyclomine, benzhexol and oxybutynine alterations in Alzheimer’s disease. In: Fisher A, Hanin I, Lachman
distinguish between subclasses of muscarinic binding sites. Eur J C, editors. Alzheimer’s and Parkinson’s diseases: strategies for re-
Pharmacol 123:133–143. search and development. New York: Plenum. p 85–94.