- Email: [email protected]
dark transition test in mice: Evaluation of classic and putative anxiolytic and anxiogenic drugs
Gen. Pharmac. Vol. 26, No.
1,pp.205-210, 1995
Copyright 0 1995Elsevier Science Ltd Printed in Great Britain. All rights reserved 0306-3623/95 $9.50 + 0.00
The Modified Light/Dark Transition Test in Mice: Evaluation of Classic and Putative Anxiolytic and Anxiogenic Drugs TAKASHI SHIMADA,’ KINZO MATSUMOTO,’ MIYUKI OSANAI,2 HARUMI MATSUDA,’ KATSUTOSHI TERASAWA’ and HIROSHI WATANABE’* ‘Department of Japanese Oriental (Kampo) Medicine, Faculty of Medicine and 2Division of Pharmacology, Research Institute for Wakan - Yaku (Oriental Medicines), Toyama Medical and Pharmaceutical University, 2630 Sugitani, Toyama-shi, 930-01 Toyama, Japan /Tel: 0764-34-2281; Fax: 0764-34-50561 (Received
I2 April 1994)
Abstract-l. We devised a new light/dark transition apparatus, recorded transitions, % time animals spent outside the dark chambers (% time) and locomotor activity, and evaluated this apparatus by testing anxiolytics, non-anxiolytic drugs and putative anxiogenic drugs in mice. 2. Diazepam and alprazolam significantly increased transitions, % time and locomotor activity. The effects of 1 mg/kg (i.p.) diazepam on these parameters in this modified test were blocked by flumazenil, a selective benzodiazepine antagonist. 3. Anxiogenic drugs such as B-carboline-3-carboxylic acid ethyl ester (J-CCE) and picrotoxin significantly decreased all three parameters. Another anxiogenic drug, yohimbine, also significantly decreased transitions and locomotor activity, but it significantly increased % time at 5 mg/kg (i.p.). 4. Imipramine (5-10 mg/kg, i.p.), an antidepressant, sulpiride (10-25 mg/kg, i.p.), an antipsychotic drug, and scopolamine (0.1-l mg/kg, i.p.), an anticholinergic drug, had no effect. 5. Buspirone, a partial 5-HT,, receptor agonist, produced parameter changes similar to those induced by anxiolytic benzodiazepines. 8-OH-DPAT, a full 5-HT,, receptor agonist, significantly increased transitions and locomotor activity but not % time. 5-HT, receptor antagonists, ICS205-930 and MDL72222, did not have any effect on these parameters. 6. Methamphetamine (l-2 mg/kg, i.p.) increased all parameters, while caffeine increased only locomotor activity. 7. The present findings indicate that the modified light/dark transition test is very simple and easy to perform for testing the anxiolytic and anxiogenic effects of drugs. Key Words: Modified light/dark transition test, anxiety, diazepam, benzodiazepine, DPAT, ICS205-930, MDL72222, /3-CCE, anxiogenic, anxiolytic
INTRODUCTION A number of animal models of anxiety have been reported for evaluating new compounds with potential anxiolytic action, and for experimentally reproducing the effects of drugs with clinical efficacy for treatment of generalized or panic anxiety (for a review, see File, 1990). They also have been used to mechanisms investigate the pathophysiological underlying such emotional disorders. The two-compartment light/dark transition test is one of the most widely used experimental models. It *To whom all correspondence
should be addressed.
buspirone, S-OH-
is based on the ethological exploratory behavior of animals in a two-compartment chamber, where one compartment is lit and the other dark (for a review, see Crawley, 1985; Crawley and Goodwin, 1980; Costa11 et al., 1987). In such a model, anxiolytic benzodiazepines reportedly increased the transitions between the two compartments, the time rodents spent in the lit compartment and the locomotor activity in the lit area (Crawley et al., 1984; Costa11 et al., 1988; Young and Johnson, 199la, b). However, the effects of anxiolytic drugs on these parameters in this model are still controversial, as Young and Johnson (1991 b) demonstrated that anxiolytic benzodiazepines and putative anxiolytic agents increase the 205
206
Takashi
time spent in the lit area but between the two compartments.
Shimada
not the transitions
In the present study, a modified light/dark transition apparatus for mice was devised, and its pharmacological selectivity was evaluated by examining the effects of drugs with known or putative anxiolytic activity. The effects of the anxiogenic drugs pcarboline-3-carboxylic acid ethyl ester (P-CCE) and yohimbine, the psychomotor stimulants caffeine and methamphetamine, the antidepressant imipramine, and the antipsychotic drug sulpiride, were also examined using this apparatus. MATERIALS
AND METHODS
Animals Male ddY mice (Nippon SLC, Shizuoka, Japan) were obtained at the age of 4 weeks. Twenty mice were housed in each cage (17 x 30 x 35 cm) in the laboratory animal room for at least 1 week before the tests. Housing conditions were thermostatically maintained at 25 Ifr 1“C, with a 12 hr light-dark cycle. Food
and water were given ad libitum.
Apparatus The apparatus was a corridor-type runway, with outer (40 x 12 cm) and inner walls (20 x 12 cm) made of black and transparent Plexiglas, respectively (Fig. 1). The floor of the apparatus (10 cm width) was made of black Plexiglas and lined into 10 cm squares. Two rectangular chambers (10 x 10 x 12 cm) made of black Plexiglas (the dark chambers) were diagonally placed on the corners of the corridor. Each dark chamber arranged
had two small passageways (3.5 x 3.5 cm) next to each other. The apparatus was
enclosed in a dark box (100 x 100 x 100 cm) and illuminated by a fluorescent lamp (10 W) 90 cm above the center of the apparatus (about 100 Lux).
et al.
Through a CCD camera installed above the apparatus, animal behavior was monitored on a TV screen outside the dark box and video-recorded for subsequent analysis. Experimental procedure Each mouse was gently placed in one of the corners away from the dark chambers. The behavioral events were recorded over a 3 min observation period using a keyboard interfaced with a computer (NEC 9801LV). Three behavioral parameters were analyzed: the number of exploratory transitions between the lit aisles and the dark chambers; % time mice spent outside the dark chambers (X time), and the number of line passages (locomotor activity). The entry into a chamber was defined as the placement of all four paws in the chamber. The experiments were carried out between 11 a.m. and 5 p.m. Animals were completely
unacquainted
with the apparatus.
Drugs The following
drugs were used for the experiments:
diazepam (Cercine” injection) and alprazolam (Takeda Chemical Industries Ltd., Osaka), flumazenil” (Roche Co. Ltd., Basel), methamphetamine HCl (Dainippon Pharmaceutical Co. Ltd., Osaka), scopolamine hydrobromide, picrotoxin, buspirone HCl (+)-&hydroxy-di-propylaminotetralin HBr @-OHDPAT) (Sigma, St Louis, MO), (f)sulpiride (Dogmatyl% injection, Fujisawa Pharmaceutical Co. Ltd., Osaka), caffeine monohydrate and yohimbine HCl (Wako Pure Chem., Osaka), (Research Biochemicals Inc., 930 (Sandoz Pharmaceutical solutions except alprazolam
p-CCE and MDL72222 Natick, MA), ICS205Co. Ltd., Base]). Drug were prepared just be-
fore the experiments, and were administered 30 min before tests except for specifically stated cases. Alprazolam was suspended in saline containing 0.5% (w/v)
dark chamber 1
400
(mm)
Fig. 1. The structure of the modified light/dark transition apparatus. The inner walls (A) and the dark chambers (B) with two passageways were made of black and transparent Plexiglas, respectively. The floor (C) was made of black Plexiglas and lined into 10 cm squares. The recording was started by placing a mouse in the corners of the corridor.
Modified light/dark transition test in mice carboxymethylcellulose. All drugs were intraperitoneally injected in a constant volume of 0.01 ml/g
207
Table 2. Effect of flumazenil on diazepam-induced changes of behavioral parameters in the modified light/dark transition test in mice
body weight.
% Time spent outside chambers
Locomotor activity (counts)
Transitions (counts)
Statistics
Drug @g/kg)
Data were analyzed with one-way analysis of variance (ANOVA) followed by the Dunnett’s test for comparing drug treatment with control or with twoway ANOVA followed by Tukey’s test for multiple
Control Diazepam Flumazenil Diazepam + flumazenil
comparison among groups. A difference was considered statistically significant at P < 0.05.
Diazepam or vehicle was injected (i.p.) 30 min before the experiments. Flumazenil or vehicle was injected (i.p.) 20 min after diazepam. Each datum represents the mean f SEM (n = IO). *P < 0.05 and **P < 0.01 compared with controls; tP < 0.05 and $P < 0.01 compared with diazepam alone (Tukey’s test). Interaction: F(l,36) = 8.236, P i 0.01 (% time spent outside the chambers), F(l,36) = 16.656, P < 0.001 (transitions), F(l,36) = 29.115, P < 0.001 (locomotor activity).
RESULTS Effects of GABA -benzodiazepine receptor related compounds on behavioral parameters in the modified light/dark transition test In control ddY locomotor activity
mice, transitions, % time and were 11.2 f 0.4 counts, 50.9 f
1.2% and 41.3 f 1.9 counts (mean + SEM; n = 90), respectively. As summarized in Table 1, acute administration of diazepam (0.25-l mg/kg) significantly increased transitions [F(3,36) = 14.782, P < O.OOl], % time [F(3,36) = 5.275, P < O.Ol] and locomotor activity [F(3,36) = 16.993, P < O.OOl]. Alprazolam (0.05-0.20 mg/kg) also significantly and dosedependently increased transitions [F(3,36) = 6.639, P < O.OOl], % time [F(3,36) = 7.733, P < O.OOl] and locomotor activity [F(3,36) = 10.160, P < O.OOl]. On the other hand, benzodiazepine
/3-CCE agonist,
(0.5-2 mgikg), significantly
transitions [F(3,36) = 5.903, P < 0.011, [F(3,36) = 5.317, P < 0.011 and locomotor P < O.Ol]. Picrotoxin [F(3,36) = 5.317, 2 mg/kg), blocker,
an inverse decreased % time activity (1 and
a GABA, receptor-coupled Cl- channel also significantly decreased transitions
Table I. Effects of GABA-benzodiazepine receptor-related drugs on behavioral parameters in the modified light/dark transition test in mice Dose (mg/kg) Diazepam
Alprazolam
a-CCE
Picrotoxin
% Time spent outside chambers
Locomotor activity (counts)
10.4 14.6 15.2: 21.7 z
I.1 1.2’ 0.x* l.6**
37.1 52.9 67.3 87.7
i i f i
4.5 6.0 2.3.’ 6.9*’
I
47.2 48.7 58.0 62.8
f f + +
0 0.05 0.1 0.2
45.0 50.3 56. I 66.4
k 2.8 i 2.9 f 3.0* If- 4.2**
II.7 I3 6 16.5 18.9
I.2 0.7 l.4* l.S**
40.1 49.1 62.7 80.7
f f f f
5.0 3.3 6.l** 7.0**
0 0.5 I ^
56.0 59.2 42.6 44.6
f ? + f
ll.O+O.8 ll.4tO.X 7.2 f 0.9** 7.x f 1.0*
38.4 40.6 21.4 27.0
k k + f
4.1 3.8 3.7** 4.3
0 I 2
52.9 f 3.6 46.7 + 5.4 35.4 + 5.0’
0 0.25 0.5
3.1 4.0 3.6* 2.2’*
Transitions (counts)
2.5 2.7 4.5’ 4. I *
‘. + f f
13.0 + 1.3 45.6 f 4.4 37.7 + 8.1 10.7 + 2.0 5.0 + 1.3** 16.0 + _ 4.2”
Test drugs were intraperitoneally injected 30 min before the experiments. Each datum represents the mean I SEM (n = IO). *P < 0.05 and **f < 0.01 compared with respective controls.
I I
40.9 f 3.4 60.1 ? 2.9’ 43.8 f 2.9
10.4 i I.6 20.0 f l.6** 13.1 + I.2
35.9 + 5.8 82.4 f 6.2** 48.4 f 4. I
45.7 k 2.97
II.5 f I.11
37.3 * 5.l$
I I
[F(2,27) = 7.178, P < 0.011, % time [F(2,27) = 3.528, P < 0.051 and locomotor activity [F(2,27) = 6.903, P < 0.011. As shown in Table 2, a two-way ANOVA revealed a significant interaction between diazepam (1 mg/kg) and flumazenil (1 mg/kg) treatment (% time spent, outside the dark chambers: F(l,36) = 8.236, P < 0.01; transitions: F(l,36) = 16.656, P < 0.001; locomotor activity: F( 1,36) = 29.115, P < 0.001). Moreover, post hoc test showed
that flumazenil
(1 mg/kg),
a selective benzodiazepine receptor antagonist, significantly blocked the effect’s of 1 mg/kg diazepam on transitions, locomotor activity and % time, but had no direct effect on these parameters
when used alone.
Efsects of serotonergic drugs on behavioral parameters in the modified light/dark As shown
in Table
and 8-OH-DPAT
transition test
3, buspirone
(0.03-0.3 mg/kg),
(0.25-I 5-HT,,
Table 3. Effects of 5-HT receptor related drugs parameters in the modified light/dark transition
Drug
Dose @g/kg)
% Time spent outside chambers
mg/kg) agonists,
on behavioral test in mice
Transitions (counts)
Locomotor activity (counts)
I
45.9 52.1 61.1 62.7
& 2.9 * 3.3 ir 2.2” + 2.2**
9.3 II.3 13.2 12.8
f 0.7 * 1.0 + 0.7** +0.X**
35.2 40.1 51.2 54.4
& 4.1 * 4.1 i 3.2** + 4.0**
0 0.03 0.1 0.3
53.7 55.9 61.2 56.4
+ + f f
3.8 2.9 2.9 2.4
9.5 * 0.8 ll.6+ I.0 13.8 k 0.5** 14.7 f 0.8**
35.3 40.3 52.3 53.4
* i f +
4.4 3.6 4.7** 3.6**
rcs205-930t
0 I 2.5 5
54.0 50.8 43.3 52.1
+ f + *
4.2 3.8 2.5 2.2
12.8 13.1 IO.1 12.0
49.9 53.3 36.4 45.7
f f f *
5.4 6.3 5.0 5.9
MDL72222
0 0.125 0.5
49.4 + 4.2 56.8 f 3.7 49.7 * 5.1
Buspirone
0 0.25 0.5
8-OH-DPAT
Test
f f f f
I.1 1.4 0.9 I.8
12.1 + I.5 10.5 * 0.9 9.8 + 1.4
42.9 i 6.6 38.1 15.3 28.6 i 3.8
drugs except buspirone were injected (i.p.) 30min before experiments. Buspirone was injected (i.p.) IO min before experiments. Each datum represents the mean k SEM (n = 8-10). *P < 0.05 and **P < 0.01 compared with respective controls. tlrglkg.
Takashi Shimada et al.
208
Table 4. Effects of various drugs on behavioral parameters modified light/dark transition test in mice Dose (mg/kg)
Drug Yohimbine
0 2.5 5
% Time spent outside Transitions chambers (counts) 58.3 + 4.2 61.0 rt 4.3 74.0 * 4.1’
13.9 + 1.7 9.5 f 0.6; 5.2 + _ 0.8**
in the
Locomotor activitv (counts) 56. I f 8.2 38.5 f 4.6’ 26.6 + - 3.7**
Yohimbine
(2.5 and
5mg/kg),
an CQ adrenocep-
tor antagonist, significantly decreased transitions [F(2,27) = 14.232, P < O.OOl] and locomotor activity [F(2,27) = 6.487, P < 0.011, but it significantly increased % time [F(2,27) = 3.977, P < 0.051 at 5 mg/kg. Methamphetamine
(1 and
2 mg/kg)
lmipramine
0 5 IO
38.3 f 2.9 43.7 * 3.3 46.8 i 4.2
8.2 f 1.1 9.2 + I .o 9.6 + 1.3
22.3 f 4.5 30.9 f 5.4 32.6 f 5.4
Sulpiride
0 IO 25
57.7 + 3.6 43.5k3.0 52.4 It 5.7
13.5 f I.1 ll.O? 1.0 11.5* I.5
54.7 * 5.3 38.7 f 5.6 44.6 + 5.0
activity [F(2,21) = 11.129, P < O.OOl]. Caffeine (15-60 mg/kg) did not
0 0. I I
50.0 f 3.7 50.8 f 5.0 58.5 i 3.3
12.7 f I .3 12.9 f I.3 10.8 f I.6
48.4 k 6.2 47.6 f 6.8 41.0 + 8.0
change
0 I 2
55.9 i 4.0 12.5 i 1.3 48.9 i 6.0 62.8 i 3.4 17.0 + 0.9’ 66.3 i 5.4 70.3 _ + 2.7** 20.9 + I .O** 91 3 _ + 7.6**
Scopolamine
Methamphetamine
Caffeine
0 I5 30 60
52.8 66.6 60.9 63.6
f 4.0 * 3. I + 3.9 & 3.7
10.6 13.9 12.0 12.3
+ i + f
0.8 I.5 0.9 I.1
38.0 62.2 57.5 53.7
f ?r + f
3.6 6.3** 3.5* 5.9.
Each datum represents the mean f SEM (n = 10-13). *P < 0.05 and ?? *f < 0.01 compared with respective controls.
significantly increased transitions [F(3,36) = 4.658, P < 0.01 for buspirone; F(3,36) = 8.709, P < 0.001 for 8-OH-DPAT]
and locomotor
activity
[F(3,36) =
5.509, P -C0.01 for buspirone; F(3,36) = 4.686, P < 0.01 for 8-OH-DPAT]. Buspirone but not 8-OHDPAT significantly increased % time [F(3,36) = 8.645, P < 0.001 for buspirone; F(3,36) = 1.076, P > 0.05 for 8-OH-DPAT]. MDL72222 (0.125-0.5 mg/kg), a 5-HT, antagonist, did not significantly alter transitions [F(2,27) = 1.840, P > 0.051, % time [F(2,27) = 0.925, P > 0.051 or locomotor activity [F(2,27) = 1.847, P > 0.051. Neither did ICS205-930 (l-5 pg/kg) change the behavioral parameters recorded [transitions: F(3,27) = 1.114, P > 0.05; % time: F(3,27) = 2.005, P > 0.05; locomotor
activity:
F(3,27) = 1.708, P > 0.051.
Eflect of various drug administrations on behavior in the modified light /dark transition test A one-way ANOVA revealed significant effect of imipramine (5 and 10 mg/kg) on % time [F(2,34) = 6.253, P < O.Ol] but a post hoc test did not show significant difference between saline and imipramine treated animals. None of the other parameters recorded were significantly changed by imipramine (5 and 10 mg/kg) [transitions: F(2,34) = 2.129, P > 0.05; locomotor activity: F(2,34) = 1.192, P > 0.051, sulpiride (10-25 mg/kg) [transitions: F(2,27) = 2.328, P > 0.05; % time: F(2,27) = 2.852, P > 0.05; locomotor activity: F(2,27) = 1.208, P > 0.051 or scopolamine (0.1 and 1 mg/kg) [transitions: F(2,27) = 0.769, P > 0.05; % time: F(2,27) = 0.406, P > 0.05; locomotor activity: F(2,27) = 1.429, P > 0.051.
significantly
increased transitions [F(2,21) = 14.840, P < O.OOl], % time [F(2,21) = 4.477, P c 0.051 and locomotor
transitions
[F(3,36) = 1.516,
significantly P > 0.051, or
% time [F(3,36) = 2.541, P > 0.051, while it significantly increased locomotor activity [F(3,36) = 4.441, P < 0.011.
DISCUSSION The light/dark Crawley
transition
and Goodwin
test initially
described
(1980) is based
by
on both the
aversive responses of mice or rats to a brightly lit area and the spontaneous exploratory behavior in such an environment. In those tests, classic anxiolytics such as diazepam and sitions between
chlordiazepoxide increase the tranthe two compartments (Crawley and
Goodwin, 1980; Crawley, however, have demonstrated
1981). Recent reports, using a computer con-
trolled two-compartment light/dark apparatus that anxiolytic drugs significantly increase exploratory activity in the lit area (Onaivi and Martin, 1989) or time spent in the lit area (Young and Johnson, 1991a, b) without changing the transitions between two compartments. These controversial prompted us to devise a new light/dark
findings transition
apparatus. In this modified apparatus, we measured three parameters: % time, the transitions between lit area and dark chambers and locomotor activity in the lit area. Although the total area of the dark chambers was one-fifth of the total lit area, the time naive control mice spent in the dark chambers was about 50% of the total recording period (3 min). This clear-cut preference of mice for the dark chambers of this apparatus agrees well with the data reported by other groups using the two-compartment light/dark transition apparatus (Costa11 et al., 1987) and indicates possible aversive properties of the lit area. Benzodiazepine anxiolytics, diazepam and alprazolam, dose-dependently increased all the parameters recorded during the 3 min test period, whereas the antipsychotic drug sulpiride, the antidepressant imipramine and the anticholinergic drug scopolamine did not significantly change them. These results are consistent with previous reports (Crawley and Goodwin, 1980; Crawley et al., 1984) and give rise to the possibility that the parameters measured in the
Modified light/dark transition test in mice modified light/dark transition apparatus are sensitively changed by drugs affecting anxiety. Moreover, the fact that flumazenil, a selective benzodiazepine receptor antagonist, blocked the diazepam-induced change in these parameters indicates the involvement of benzodiazepine receptor in the anxiolytic-like effect of diazepam. b-CCE, an inverse benzodiazepine agonist, and picrotoxin, a GABA, receptor-chloride channel blocker, reportedly produce anxiogenic action in animal models of anxiety (Prado de Carvalho et al., 1983; File and Lister, 1984). In the present study, the changes in parameters (% time, transitions and locomotor activity) caused by these two drugs contrasted markedly with those induced by anxiolytic benzodiazepines. Thus, these results suggest that the decreases and increases of all parameters represent the anxiogenic and anxiolytic states, respectively, of mice placed in the modified light/dark transition apparatus, and that the parameters are useful for assessing anxiolytic or anxiogenic action of test agents. Yohimbine, which is known to have anxiogenic properties in man and rodents (Charney and De Redmond, 1983; Handley and Mithani, 1984; Pellow et al., 1985), produced parameter changes similar to those induced by /I-CCE and picrotoxin, except for the significant increase by yohimbine (5 mg/kg) of % time spent outside the chambers. Since yohimbine reportedly decreases locomotor activity in mice at a high dose (>4mg/kg, i.p.; Njung’e and Handley, 1991) the apparent increase of % time seems to be due to the sedative effect of yohimbine. The present findings disagree with the data of Venault et al. (1993) that yohimbine did not show any effect at 0.125-4 mg/kg (i.p.) in the light/dark choice situation. The reason for this discrepancy remains unclear, but it may be due to differences in the apparatus and/or the strains of mice used for the experiments. Of the serotonergic drugs tested, only buspirone, a partial 5HT,, receptor agonist, significantly increased all parameters measured in the modified light/dark transition test. Moreover, buspirone-induced parameter changes were quite similar to those induced by anxiolytic benzodiazepines, suggesting an anxiolyticlike effect of buspirone in this animal model of anxiety. On the other hand, 8-OH-DPAT, a full 5HT,, receptor agonist, also significantly increased transitions and locomotor activity, but it failed to increase the % time mice spent outside the dark chambers. These results disagree with the data reported by other groups that 8-OH-DPAT produces parameter changes similar to those induced by diazepam in the light/dark test (Young and Johnson, 199la, b). The reason for this discrepancy remains unresolved. However, as the selectivity of parameters
209
recorded in the light/dark test to anxiolytic drugs is still controversial (Crawley and Goodwin, 1980; Costa11 et al., 1988; Onaivi and Martin, 1989; Young and Johnson, 199la, b; Lopez-Rubalcava et al., 1992), the failure of 8-OH-DPAT to increase the % time mice spent outside the chambers does not seem to completely exclude the putative anxiolytic action of this drug. On the other hand, neither ICS205-930 nor MDL72222, 5-HT, receptor antagonists, showed an anxiolytic-like effect in our modified light/dark transition test. These observations disagree with other reports using behavioral models such as the twocompartment light/dark transition test (Costa11 et al., 1988, 1989; Onaivi and Martin, 1989) and the elevated plus-maze (Wright et al., 1992) that 5-HT, receptor antagonists exhibit anxiolytic-like effects. The question, however, has been raised repeatedly as to whether 5-HT, receptor antagonists show anxiolytic-like effects in ordinary axiety models used for testing benzodiazepine anxiolytics (for reviews, see Barrett and Vanover, 1993; Handley and McBlane, 1993). Therefore, the failure of these compounds to produce the parameter changes under the same experimental conditions as those used for busipirone may caution against the conclusion that ICS205-930 and MDL72222 have anxiolytic-like effects. Amphetamine and caffeine both have been reported to have anxiogenic-like actions (Pellow et al., 1985) and to increase exploratory activity in the dark chamber in the light/dark transition test (Charney et al., 1984; Lister, 1987). In the modified light/dark transition test, however, methamphetamine produced the same parameter changes as anxiolytic drugs did, whereas caffeine significantly increased only locomotor activity without affecting the other two parameters. Our and other groups’ data have demonstrated that caffeine dose-dependently stimulates locomotor activity in mice at the dose range of lO-30mg/kg (Fujii et al., 1989; Matsumoto et al., 1990). In the present study, however, the caffeinestimulation of locomotor activity was not dosedependent. The reason for this discrepancy remains unclear, but it may be due to the difference in the systems used for measuring. Nevertheless, these profiles of methamphetamine and caffeine were different from those of anxiogenic drugs such as P-CCE and picrotoxin, and the parameter changes induced by methamphetamine were similar to those by the anxiolytic benzodiazepines. The reason for the apparent anxiolytic-like effect of methamphetamine is not clear. However, it is possible that the effect of methamphetamine on motor activity in the dark chambers may be underestimated in our modified light/dark apparatus, since the area ratio of the dark chambers
210
Takashi
to the lit aisle (115) was smaller light/dark
that
in the usual
(> l/3)
apparatus
1989; Young
than
Shimada
and Johnson,
(Onaivi and Martin, 1991b). In fact, the sensi-
tivity of the light/dark transition test to anxiogeniclike effects of drugs appears to be increased by employing two compartments of equal size (File, 1990). In conclusion, the present findings suggest that the increse of all three parameters in this test are useful measure
for
assessing
anxiolytic-like
activity,
and
that the procedure described is very simple and can quickly test anxiolytic- and anxiogenic-like actions of compounds. Nevertheless, whether the increases of all three parameters are crucial requirements for a compound to be judged as a putative remains to be further investigated.
anxiolytic
drug
REFERENCES Barrett J. E. and Vanover K. E. (1993) 5-HT receptors
as targets for the development of novel anxiolytic drugs: models, mechanisms and future directions. Psychopharmacology 112, l-12. Charney D. S. and De Redmond. (1993) Neurobiological mechanisms in human anxiety: evidence supporting central noradrenergic hyperactivity. Neuropharmacology 22, 1531-1536. Charney D. S., Galloway M. P. and Heninger G. R. (1984) The effects of caffeine on plasma MHPG, subjective anxiety, autonomic symptoms and blood pressure in healthy humans. I@ Sci. 35, 1355144. Costa11 B., Hendric C. A., Kelly M. E. and Naylor R. J. (1987) Actions of sulpiride and tiapride in a simple model of anxiety in mice. Neuropharmacology 26, 195-200. Costall B., Kelly M. E., Naylor R. J. and Onaivi E. S. (1988) Actions of buspirone in a putative model of anxiety in the mouse. J. Pharm. Pharmac. 40, 494-500. Costall B., Jones B. J., Kelly E., Naylor R. J. and Tomkins D. M. (1989) Exploration of mice in a black and white test box: validation as a model of anxiety. Pharmac. Biochem. Behac. 32, 177-785. Crawley J. N. (1981) Neuropharmacologic specificity of a simple animal model for the behavioral actions of benzodiazepines. Pharmac. Biochem. Behav. 15, 695-699. Crawley J. N. (1985) Exploratory behavior models of anxiety in mice. Neurosci. Biobehav. Rev. 9, 3744. Crawley J. N. and Goodwin F. K. (1980) Preliminary report of a simple animal behavior model for the anxiolytic effects of benzodiazepines. Pharmac. Biochem. Behac. 13, 167-170. Crawley J. N., Skolnick P. and Paul S. M. (1984) Absence of intrinsic antagonist actions of benzodiazepine antagon-
er al.
ists on an exploratory model of anxiety in the mouse. Neuropharmacology 2j, 531-537. File S. E. (1990) New strategies in the search for anxiolytics. Drug Design and Delivery 5, 195-201. File S. E. and Lister R. G. (1984) Do the reductions in social
interaction produced by picrotoxin and pentylenetetrazole indicate anxiogenic action? Neuropharmacology 23, 793-796. Fujii W., Kuribara
H. and Tadokoro
S. (1989) Interaction
between caffeine and methamphetamine by means of ambulatory activity in mice. Jpn. J. Psychopharmac. 11, 225-231.
Handley S. L. and Mithani S. (1984) Effects of alphaadrenoceptor agonists and antagonists in a mazeexploration model of ‘fear’-motivated behaviour. Naunvn-Schmiedebere’s Arch. Pharmac. 321. 1-5. Handle; S. L. and McBlane J. W. (1993) 5HT drugs in animal models of anxiety. Psychopharmacology 112, 13-20. Lister R. G. (1987) The use of a plus-maze to measure anxiety in the mouse. Psychopharmacology 92, 180-185. Lopez-Rubalcava C., Saldivar A. and Fernandez-Guasti A. (1992) Interaction of GABA and serotonin in the anxiolytic action of diazepam and serotonergic anxiolytics. Pharmac. Biochem. Behav. 43, 433440. Matsumoto K., Cai B., Nakamura S. and Watanabe H. (1990) A new system for the measurement and analysis of motor activity in mice: effect of several central stimulants. Folk Pharmacol. japon. 96, 31-39. Njung’e K. and Handley S. L. (1991) Evaluation of marbleburying behaviour as model of anxiety. Pharmac. Biothem. Behav. 38, 63-67. Onaivi E. S. and Martin B. R. (1989) Neuropharmacological and physiological validation of a computer-controlled two-compartment black and white box for the assessment of anxiety. Prog. Neuropsychopharmac. Biol. Psychiat. 13, 963-976. Pellow S., Chopin P. and File S. E. (1985) Are the anxiogenie effects of yohimbine mediated by its action at benzodiazepine receptors? Neurosci. Left. 55, 5-9. Prado de Carvalho L., Venault P., Rossier J. and Chapouthier G. (1983) Anxiogenic properties of convulsive agents. Sot. Neurosci. Abstr. 9, 128. Venault P., Jaquot F., Save E., Sara S. and Chapouthier G. (1993) Anxiogenic-like effects of yohimbine and idazoxan in two behavioral situations in mice. Life Sci. 52,639-5. Wright I. K., Heaton M., Upton N. and Marsden C. A. (1992) Comparison of acute and chronic treatment of various serotonergic agents with those of diazepam and idazoxan in the rat elevated X-maze. P:ychopharmacology 107, 405414. Young R. and Johnson D. N. (1991a) Comparison of routes of administration and time course effects of zaclopride and buspirone in mice using an automated light/dark test. Pharmac. Biochem. Behav. 40, 733-737. Young R. and Johnson D. N. (1991b) A fully automated light/dark apparatus useful for comparing anxiolytic agents. Pharmac. Biochem. Behav. 40, 739-743.
1,pp.205-210, 1995
Copyright 0 1995Elsevier Science Ltd Printed in Great Britain. All rights reserved 0306-3623/95 $9.50 + 0.00
The Modified Light/Dark Transition Test in Mice: Evaluation of Classic and Putative Anxiolytic and Anxiogenic Drugs TAKASHI SHIMADA,’ KINZO MATSUMOTO,’ MIYUKI OSANAI,2 HARUMI MATSUDA,’ KATSUTOSHI TERASAWA’ and HIROSHI WATANABE’* ‘Department of Japanese Oriental (Kampo) Medicine, Faculty of Medicine and 2Division of Pharmacology, Research Institute for Wakan - Yaku (Oriental Medicines), Toyama Medical and Pharmaceutical University, 2630 Sugitani, Toyama-shi, 930-01 Toyama, Japan /Tel: 0764-34-2281; Fax: 0764-34-50561 (Received
I2 April 1994)
Abstract-l. We devised a new light/dark transition apparatus, recorded transitions, % time animals spent outside the dark chambers (% time) and locomotor activity, and evaluated this apparatus by testing anxiolytics, non-anxiolytic drugs and putative anxiogenic drugs in mice. 2. Diazepam and alprazolam significantly increased transitions, % time and locomotor activity. The effects of 1 mg/kg (i.p.) diazepam on these parameters in this modified test were blocked by flumazenil, a selective benzodiazepine antagonist. 3. Anxiogenic drugs such as B-carboline-3-carboxylic acid ethyl ester (J-CCE) and picrotoxin significantly decreased all three parameters. Another anxiogenic drug, yohimbine, also significantly decreased transitions and locomotor activity, but it significantly increased % time at 5 mg/kg (i.p.). 4. Imipramine (5-10 mg/kg, i.p.), an antidepressant, sulpiride (10-25 mg/kg, i.p.), an antipsychotic drug, and scopolamine (0.1-l mg/kg, i.p.), an anticholinergic drug, had no effect. 5. Buspirone, a partial 5-HT,, receptor agonist, produced parameter changes similar to those induced by anxiolytic benzodiazepines. 8-OH-DPAT, a full 5-HT,, receptor agonist, significantly increased transitions and locomotor activity but not % time. 5-HT, receptor antagonists, ICS205-930 and MDL72222, did not have any effect on these parameters. 6. Methamphetamine (l-2 mg/kg, i.p.) increased all parameters, while caffeine increased only locomotor activity. 7. The present findings indicate that the modified light/dark transition test is very simple and easy to perform for testing the anxiolytic and anxiogenic effects of drugs. Key Words: Modified light/dark transition test, anxiety, diazepam, benzodiazepine, DPAT, ICS205-930, MDL72222, /3-CCE, anxiogenic, anxiolytic
INTRODUCTION A number of animal models of anxiety have been reported for evaluating new compounds with potential anxiolytic action, and for experimentally reproducing the effects of drugs with clinical efficacy for treatment of generalized or panic anxiety (for a review, see File, 1990). They also have been used to mechanisms investigate the pathophysiological underlying such emotional disorders. The two-compartment light/dark transition test is one of the most widely used experimental models. It *To whom all correspondence
should be addressed.
buspirone, S-OH-
is based on the ethological exploratory behavior of animals in a two-compartment chamber, where one compartment is lit and the other dark (for a review, see Crawley, 1985; Crawley and Goodwin, 1980; Costa11 et al., 1987). In such a model, anxiolytic benzodiazepines reportedly increased the transitions between the two compartments, the time rodents spent in the lit compartment and the locomotor activity in the lit area (Crawley et al., 1984; Costa11 et al., 1988; Young and Johnson, 199la, b). However, the effects of anxiolytic drugs on these parameters in this model are still controversial, as Young and Johnson (1991 b) demonstrated that anxiolytic benzodiazepines and putative anxiolytic agents increase the 205
206
Takashi
time spent in the lit area but between the two compartments.
Shimada
not the transitions
In the present study, a modified light/dark transition apparatus for mice was devised, and its pharmacological selectivity was evaluated by examining the effects of drugs with known or putative anxiolytic activity. The effects of the anxiogenic drugs pcarboline-3-carboxylic acid ethyl ester (P-CCE) and yohimbine, the psychomotor stimulants caffeine and methamphetamine, the antidepressant imipramine, and the antipsychotic drug sulpiride, were also examined using this apparatus. MATERIALS
AND METHODS
Animals Male ddY mice (Nippon SLC, Shizuoka, Japan) were obtained at the age of 4 weeks. Twenty mice were housed in each cage (17 x 30 x 35 cm) in the laboratory animal room for at least 1 week before the tests. Housing conditions were thermostatically maintained at 25 Ifr 1“C, with a 12 hr light-dark cycle. Food
and water were given ad libitum.
Apparatus The apparatus was a corridor-type runway, with outer (40 x 12 cm) and inner walls (20 x 12 cm) made of black and transparent Plexiglas, respectively (Fig. 1). The floor of the apparatus (10 cm width) was made of black Plexiglas and lined into 10 cm squares. Two rectangular chambers (10 x 10 x 12 cm) made of black Plexiglas (the dark chambers) were diagonally placed on the corners of the corridor. Each dark chamber arranged
had two small passageways (3.5 x 3.5 cm) next to each other. The apparatus was
enclosed in a dark box (100 x 100 x 100 cm) and illuminated by a fluorescent lamp (10 W) 90 cm above the center of the apparatus (about 100 Lux).
et al.
Through a CCD camera installed above the apparatus, animal behavior was monitored on a TV screen outside the dark box and video-recorded for subsequent analysis. Experimental procedure Each mouse was gently placed in one of the corners away from the dark chambers. The behavioral events were recorded over a 3 min observation period using a keyboard interfaced with a computer (NEC 9801LV). Three behavioral parameters were analyzed: the number of exploratory transitions between the lit aisles and the dark chambers; % time mice spent outside the dark chambers (X time), and the number of line passages (locomotor activity). The entry into a chamber was defined as the placement of all four paws in the chamber. The experiments were carried out between 11 a.m. and 5 p.m. Animals were completely
unacquainted
with the apparatus.
Drugs The following
drugs were used for the experiments:
diazepam (Cercine” injection) and alprazolam (Takeda Chemical Industries Ltd., Osaka), flumazenil” (Roche Co. Ltd., Basel), methamphetamine HCl (Dainippon Pharmaceutical Co. Ltd., Osaka), scopolamine hydrobromide, picrotoxin, buspirone HCl (+)-&hydroxy-di-propylaminotetralin HBr @-OHDPAT) (Sigma, St Louis, MO), (f)sulpiride (Dogmatyl% injection, Fujisawa Pharmaceutical Co. Ltd., Osaka), caffeine monohydrate and yohimbine HCl (Wako Pure Chem., Osaka), (Research Biochemicals Inc., 930 (Sandoz Pharmaceutical solutions except alprazolam
p-CCE and MDL72222 Natick, MA), ICS205Co. Ltd., Base]). Drug were prepared just be-
fore the experiments, and were administered 30 min before tests except for specifically stated cases. Alprazolam was suspended in saline containing 0.5% (w/v)
dark chamber 1
400
(mm)
Fig. 1. The structure of the modified light/dark transition apparatus. The inner walls (A) and the dark chambers (B) with two passageways were made of black and transparent Plexiglas, respectively. The floor (C) was made of black Plexiglas and lined into 10 cm squares. The recording was started by placing a mouse in the corners of the corridor.
Modified light/dark transition test in mice carboxymethylcellulose. All drugs were intraperitoneally injected in a constant volume of 0.01 ml/g
207
Table 2. Effect of flumazenil on diazepam-induced changes of behavioral parameters in the modified light/dark transition test in mice
body weight.
% Time spent outside chambers
Locomotor activity (counts)
Transitions (counts)
Statistics
Drug @g/kg)
Data were analyzed with one-way analysis of variance (ANOVA) followed by the Dunnett’s test for comparing drug treatment with control or with twoway ANOVA followed by Tukey’s test for multiple
Control Diazepam Flumazenil Diazepam + flumazenil
comparison among groups. A difference was considered statistically significant at P < 0.05.
Diazepam or vehicle was injected (i.p.) 30 min before the experiments. Flumazenil or vehicle was injected (i.p.) 20 min after diazepam. Each datum represents the mean f SEM (n = IO). *P < 0.05 and **P < 0.01 compared with controls; tP < 0.05 and $P < 0.01 compared with diazepam alone (Tukey’s test). Interaction: F(l,36) = 8.236, P i 0.01 (% time spent outside the chambers), F(l,36) = 16.656, P < 0.001 (transitions), F(l,36) = 29.115, P < 0.001 (locomotor activity).
RESULTS Effects of GABA -benzodiazepine receptor related compounds on behavioral parameters in the modified light/dark transition test In control ddY locomotor activity
mice, transitions, % time and were 11.2 f 0.4 counts, 50.9 f
1.2% and 41.3 f 1.9 counts (mean + SEM; n = 90), respectively. As summarized in Table 1, acute administration of diazepam (0.25-l mg/kg) significantly increased transitions [F(3,36) = 14.782, P < O.OOl], % time [F(3,36) = 5.275, P < O.Ol] and locomotor activity [F(3,36) = 16.993, P < O.OOl]. Alprazolam (0.05-0.20 mg/kg) also significantly and dosedependently increased transitions [F(3,36) = 6.639, P < O.OOl], % time [F(3,36) = 7.733, P < O.OOl] and locomotor activity [F(3,36) = 10.160, P < O.OOl]. On the other hand, benzodiazepine
/3-CCE agonist,
(0.5-2 mgikg), significantly
transitions [F(3,36) = 5.903, P < 0.011, [F(3,36) = 5.317, P < 0.011 and locomotor P < O.Ol]. Picrotoxin [F(3,36) = 5.317, 2 mg/kg), blocker,
an inverse decreased % time activity (1 and
a GABA, receptor-coupled Cl- channel also significantly decreased transitions
Table I. Effects of GABA-benzodiazepine receptor-related drugs on behavioral parameters in the modified light/dark transition test in mice Dose (mg/kg) Diazepam
Alprazolam
a-CCE
Picrotoxin
% Time spent outside chambers
Locomotor activity (counts)
10.4 14.6 15.2: 21.7 z
I.1 1.2’ 0.x* l.6**
37.1 52.9 67.3 87.7
i i f i
4.5 6.0 2.3.’ 6.9*’
I
47.2 48.7 58.0 62.8
f f + +
0 0.05 0.1 0.2
45.0 50.3 56. I 66.4
k 2.8 i 2.9 f 3.0* If- 4.2**
II.7 I3 6 16.5 18.9
I.2 0.7 l.4* l.S**
40.1 49.1 62.7 80.7
f f f f
5.0 3.3 6.l** 7.0**
0 0.5 I ^
56.0 59.2 42.6 44.6
f ? + f
ll.O+O.8 ll.4tO.X 7.2 f 0.9** 7.x f 1.0*
38.4 40.6 21.4 27.0
k k + f
4.1 3.8 3.7** 4.3
0 I 2
52.9 f 3.6 46.7 + 5.4 35.4 + 5.0’
0 0.25 0.5
3.1 4.0 3.6* 2.2’*
Transitions (counts)
2.5 2.7 4.5’ 4. I *
‘. + f f
13.0 + 1.3 45.6 f 4.4 37.7 + 8.1 10.7 + 2.0 5.0 + 1.3** 16.0 + _ 4.2”
Test drugs were intraperitoneally injected 30 min before the experiments. Each datum represents the mean I SEM (n = IO). *P < 0.05 and **f < 0.01 compared with respective controls.
I I
40.9 f 3.4 60.1 ? 2.9’ 43.8 f 2.9
10.4 i I.6 20.0 f l.6** 13.1 + I.2
35.9 + 5.8 82.4 f 6.2** 48.4 f 4. I
45.7 k 2.97
II.5 f I.11
37.3 * 5.l$
I I
[F(2,27) = 7.178, P < 0.011, % time [F(2,27) = 3.528, P < 0.051 and locomotor activity [F(2,27) = 6.903, P < 0.011. As shown in Table 2, a two-way ANOVA revealed a significant interaction between diazepam (1 mg/kg) and flumazenil (1 mg/kg) treatment (% time spent, outside the dark chambers: F(l,36) = 8.236, P < 0.01; transitions: F(l,36) = 16.656, P < 0.001; locomotor activity: F( 1,36) = 29.115, P < 0.001). Moreover, post hoc test showed
that flumazenil
(1 mg/kg),
a selective benzodiazepine receptor antagonist, significantly blocked the effect’s of 1 mg/kg diazepam on transitions, locomotor activity and % time, but had no direct effect on these parameters
when used alone.
Efsects of serotonergic drugs on behavioral parameters in the modified light/dark As shown
in Table
and 8-OH-DPAT
transition test
3, buspirone
(0.03-0.3 mg/kg),
(0.25-I 5-HT,,
Table 3. Effects of 5-HT receptor related drugs parameters in the modified light/dark transition
Drug
Dose @g/kg)
% Time spent outside chambers
mg/kg) agonists,
on behavioral test in mice
Transitions (counts)
Locomotor activity (counts)
I
45.9 52.1 61.1 62.7
& 2.9 * 3.3 ir 2.2” + 2.2**
9.3 II.3 13.2 12.8
f 0.7 * 1.0 + 0.7** +0.X**
35.2 40.1 51.2 54.4
& 4.1 * 4.1 i 3.2** + 4.0**
0 0.03 0.1 0.3
53.7 55.9 61.2 56.4
+ + f f
3.8 2.9 2.9 2.4
9.5 * 0.8 ll.6+ I.0 13.8 k 0.5** 14.7 f 0.8**
35.3 40.3 52.3 53.4
* i f +
4.4 3.6 4.7** 3.6**
rcs205-930t
0 I 2.5 5
54.0 50.8 43.3 52.1
+ f + *
4.2 3.8 2.5 2.2
12.8 13.1 IO.1 12.0
49.9 53.3 36.4 45.7
f f f *
5.4 6.3 5.0 5.9
MDL72222
0 0.125 0.5
49.4 + 4.2 56.8 f 3.7 49.7 * 5.1
Buspirone
0 0.25 0.5
8-OH-DPAT
Test
f f f f
I.1 1.4 0.9 I.8
12.1 + I.5 10.5 * 0.9 9.8 + 1.4
42.9 i 6.6 38.1 15.3 28.6 i 3.8
drugs except buspirone were injected (i.p.) 30min before experiments. Buspirone was injected (i.p.) IO min before experiments. Each datum represents the mean k SEM (n = 8-10). *P < 0.05 and **P < 0.01 compared with respective controls. tlrglkg.
Takashi Shimada et al.
208
Table 4. Effects of various drugs on behavioral parameters modified light/dark transition test in mice Dose (mg/kg)
Drug Yohimbine
0 2.5 5
% Time spent outside Transitions chambers (counts) 58.3 + 4.2 61.0 rt 4.3 74.0 * 4.1’
13.9 + 1.7 9.5 f 0.6; 5.2 + _ 0.8**
in the
Locomotor activitv (counts) 56. I f 8.2 38.5 f 4.6’ 26.6 + - 3.7**
Yohimbine
(2.5 and
5mg/kg),
an CQ adrenocep-
tor antagonist, significantly decreased transitions [F(2,27) = 14.232, P < O.OOl] and locomotor activity [F(2,27) = 6.487, P < 0.011, but it significantly increased % time [F(2,27) = 3.977, P < 0.051 at 5 mg/kg. Methamphetamine
(1 and
2 mg/kg)
lmipramine
0 5 IO
38.3 f 2.9 43.7 * 3.3 46.8 i 4.2
8.2 f 1.1 9.2 + I .o 9.6 + 1.3
22.3 f 4.5 30.9 f 5.4 32.6 f 5.4
Sulpiride
0 IO 25
57.7 + 3.6 43.5k3.0 52.4 It 5.7
13.5 f I.1 ll.O? 1.0 11.5* I.5
54.7 * 5.3 38.7 f 5.6 44.6 + 5.0
activity [F(2,21) = 11.129, P < O.OOl]. Caffeine (15-60 mg/kg) did not
0 0. I I
50.0 f 3.7 50.8 f 5.0 58.5 i 3.3
12.7 f I .3 12.9 f I.3 10.8 f I.6
48.4 k 6.2 47.6 f 6.8 41.0 + 8.0
change
0 I 2
55.9 i 4.0 12.5 i 1.3 48.9 i 6.0 62.8 i 3.4 17.0 + 0.9’ 66.3 i 5.4 70.3 _ + 2.7** 20.9 + I .O** 91 3 _ + 7.6**
Scopolamine
Methamphetamine
Caffeine
0 I5 30 60
52.8 66.6 60.9 63.6
f 4.0 * 3. I + 3.9 & 3.7
10.6 13.9 12.0 12.3
+ i + f
0.8 I.5 0.9 I.1
38.0 62.2 57.5 53.7
f ?r + f
3.6 6.3** 3.5* 5.9.
Each datum represents the mean f SEM (n = 10-13). *P < 0.05 and ?? *f < 0.01 compared with respective controls.
significantly increased transitions [F(3,36) = 4.658, P < 0.01 for buspirone; F(3,36) = 8.709, P < 0.001 for 8-OH-DPAT]
and locomotor
activity
[F(3,36) =
5.509, P -C0.01 for buspirone; F(3,36) = 4.686, P < 0.01 for 8-OH-DPAT]. Buspirone but not 8-OHDPAT significantly increased % time [F(3,36) = 8.645, P < 0.001 for buspirone; F(3,36) = 1.076, P > 0.05 for 8-OH-DPAT]. MDL72222 (0.125-0.5 mg/kg), a 5-HT, antagonist, did not significantly alter transitions [F(2,27) = 1.840, P > 0.051, % time [F(2,27) = 0.925, P > 0.051 or locomotor activity [F(2,27) = 1.847, P > 0.051. Neither did ICS205-930 (l-5 pg/kg) change the behavioral parameters recorded [transitions: F(3,27) = 1.114, P > 0.05; % time: F(3,27) = 2.005, P > 0.05; locomotor
activity:
F(3,27) = 1.708, P > 0.051.
Eflect of various drug administrations on behavior in the modified light /dark transition test A one-way ANOVA revealed significant effect of imipramine (5 and 10 mg/kg) on % time [F(2,34) = 6.253, P < O.Ol] but a post hoc test did not show significant difference between saline and imipramine treated animals. None of the other parameters recorded were significantly changed by imipramine (5 and 10 mg/kg) [transitions: F(2,34) = 2.129, P > 0.05; locomotor activity: F(2,34) = 1.192, P > 0.051, sulpiride (10-25 mg/kg) [transitions: F(2,27) = 2.328, P > 0.05; % time: F(2,27) = 2.852, P > 0.05; locomotor activity: F(2,27) = 1.208, P > 0.051 or scopolamine (0.1 and 1 mg/kg) [transitions: F(2,27) = 0.769, P > 0.05; % time: F(2,27) = 0.406, P > 0.05; locomotor activity: F(2,27) = 1.429, P > 0.051.
significantly
increased transitions [F(2,21) = 14.840, P < O.OOl], % time [F(2,21) = 4.477, P c 0.051 and locomotor
transitions
[F(3,36) = 1.516,
significantly P > 0.051, or
% time [F(3,36) = 2.541, P > 0.051, while it significantly increased locomotor activity [F(3,36) = 4.441, P < 0.011.
DISCUSSION The light/dark Crawley
transition
and Goodwin
test initially
described
(1980) is based
by
on both the
aversive responses of mice or rats to a brightly lit area and the spontaneous exploratory behavior in such an environment. In those tests, classic anxiolytics such as diazepam and sitions between
chlordiazepoxide increase the tranthe two compartments (Crawley and
Goodwin, 1980; Crawley, however, have demonstrated
1981). Recent reports, using a computer con-
trolled two-compartment light/dark apparatus that anxiolytic drugs significantly increase exploratory activity in the lit area (Onaivi and Martin, 1989) or time spent in the lit area (Young and Johnson, 1991a, b) without changing the transitions between two compartments. These controversial prompted us to devise a new light/dark
findings transition
apparatus. In this modified apparatus, we measured three parameters: % time, the transitions between lit area and dark chambers and locomotor activity in the lit area. Although the total area of the dark chambers was one-fifth of the total lit area, the time naive control mice spent in the dark chambers was about 50% of the total recording period (3 min). This clear-cut preference of mice for the dark chambers of this apparatus agrees well with the data reported by other groups using the two-compartment light/dark transition apparatus (Costa11 et al., 1987) and indicates possible aversive properties of the lit area. Benzodiazepine anxiolytics, diazepam and alprazolam, dose-dependently increased all the parameters recorded during the 3 min test period, whereas the antipsychotic drug sulpiride, the antidepressant imipramine and the anticholinergic drug scopolamine did not significantly change them. These results are consistent with previous reports (Crawley and Goodwin, 1980; Crawley et al., 1984) and give rise to the possibility that the parameters measured in the
Modified light/dark transition test in mice modified light/dark transition apparatus are sensitively changed by drugs affecting anxiety. Moreover, the fact that flumazenil, a selective benzodiazepine receptor antagonist, blocked the diazepam-induced change in these parameters indicates the involvement of benzodiazepine receptor in the anxiolytic-like effect of diazepam. b-CCE, an inverse benzodiazepine agonist, and picrotoxin, a GABA, receptor-chloride channel blocker, reportedly produce anxiogenic action in animal models of anxiety (Prado de Carvalho et al., 1983; File and Lister, 1984). In the present study, the changes in parameters (% time, transitions and locomotor activity) caused by these two drugs contrasted markedly with those induced by anxiolytic benzodiazepines. Thus, these results suggest that the decreases and increases of all parameters represent the anxiogenic and anxiolytic states, respectively, of mice placed in the modified light/dark transition apparatus, and that the parameters are useful for assessing anxiolytic or anxiogenic action of test agents. Yohimbine, which is known to have anxiogenic properties in man and rodents (Charney and De Redmond, 1983; Handley and Mithani, 1984; Pellow et al., 1985), produced parameter changes similar to those induced by /I-CCE and picrotoxin, except for the significant increase by yohimbine (5 mg/kg) of % time spent outside the chambers. Since yohimbine reportedly decreases locomotor activity in mice at a high dose (>4mg/kg, i.p.; Njung’e and Handley, 1991) the apparent increase of % time seems to be due to the sedative effect of yohimbine. The present findings disagree with the data of Venault et al. (1993) that yohimbine did not show any effect at 0.125-4 mg/kg (i.p.) in the light/dark choice situation. The reason for this discrepancy remains unclear, but it may be due to differences in the apparatus and/or the strains of mice used for the experiments. Of the serotonergic drugs tested, only buspirone, a partial 5HT,, receptor agonist, significantly increased all parameters measured in the modified light/dark transition test. Moreover, buspirone-induced parameter changes were quite similar to those induced by anxiolytic benzodiazepines, suggesting an anxiolyticlike effect of buspirone in this animal model of anxiety. On the other hand, 8-OH-DPAT, a full 5HT,, receptor agonist, also significantly increased transitions and locomotor activity, but it failed to increase the % time mice spent outside the dark chambers. These results disagree with the data reported by other groups that 8-OH-DPAT produces parameter changes similar to those induced by diazepam in the light/dark test (Young and Johnson, 199la, b). The reason for this discrepancy remains unresolved. However, as the selectivity of parameters
209
recorded in the light/dark test to anxiolytic drugs is still controversial (Crawley and Goodwin, 1980; Costa11 et al., 1988; Onaivi and Martin, 1989; Young and Johnson, 199la, b; Lopez-Rubalcava et al., 1992), the failure of 8-OH-DPAT to increase the % time mice spent outside the chambers does not seem to completely exclude the putative anxiolytic action of this drug. On the other hand, neither ICS205-930 nor MDL72222, 5-HT, receptor antagonists, showed an anxiolytic-like effect in our modified light/dark transition test. These observations disagree with other reports using behavioral models such as the twocompartment light/dark transition test (Costa11 et al., 1988, 1989; Onaivi and Martin, 1989) and the elevated plus-maze (Wright et al., 1992) that 5-HT, receptor antagonists exhibit anxiolytic-like effects. The question, however, has been raised repeatedly as to whether 5-HT, receptor antagonists show anxiolytic-like effects in ordinary axiety models used for testing benzodiazepine anxiolytics (for reviews, see Barrett and Vanover, 1993; Handley and McBlane, 1993). Therefore, the failure of these compounds to produce the parameter changes under the same experimental conditions as those used for busipirone may caution against the conclusion that ICS205-930 and MDL72222 have anxiolytic-like effects. Amphetamine and caffeine both have been reported to have anxiogenic-like actions (Pellow et al., 1985) and to increase exploratory activity in the dark chamber in the light/dark transition test (Charney et al., 1984; Lister, 1987). In the modified light/dark transition test, however, methamphetamine produced the same parameter changes as anxiolytic drugs did, whereas caffeine significantly increased only locomotor activity without affecting the other two parameters. Our and other groups’ data have demonstrated that caffeine dose-dependently stimulates locomotor activity in mice at the dose range of lO-30mg/kg (Fujii et al., 1989; Matsumoto et al., 1990). In the present study, however, the caffeinestimulation of locomotor activity was not dosedependent. The reason for this discrepancy remains unclear, but it may be due to the difference in the systems used for measuring. Nevertheless, these profiles of methamphetamine and caffeine were different from those of anxiogenic drugs such as P-CCE and picrotoxin, and the parameter changes induced by methamphetamine were similar to those by the anxiolytic benzodiazepines. The reason for the apparent anxiolytic-like effect of methamphetamine is not clear. However, it is possible that the effect of methamphetamine on motor activity in the dark chambers may be underestimated in our modified light/dark apparatus, since the area ratio of the dark chambers
210
Takashi
to the lit aisle (115) was smaller light/dark
that
in the usual
(> l/3)
apparatus
1989; Young
than
Shimada
and Johnson,
(Onaivi and Martin, 1991b). In fact, the sensi-
tivity of the light/dark transition test to anxiogeniclike effects of drugs appears to be increased by employing two compartments of equal size (File, 1990). In conclusion, the present findings suggest that the increse of all three parameters in this test are useful measure
for
assessing
anxiolytic-like
activity,
and
that the procedure described is very simple and can quickly test anxiolytic- and anxiogenic-like actions of compounds. Nevertheless, whether the increases of all three parameters are crucial requirements for a compound to be judged as a putative remains to be further investigated.
anxiolytic
drug
REFERENCES Barrett J. E. and Vanover K. E. (1993) 5-HT receptors
as targets for the development of novel anxiolytic drugs: models, mechanisms and future directions. Psychopharmacology 112, l-12. Charney D. S. and De Redmond. (1993) Neurobiological mechanisms in human anxiety: evidence supporting central noradrenergic hyperactivity. Neuropharmacology 22, 1531-1536. Charney D. S., Galloway M. P. and Heninger G. R. (1984) The effects of caffeine on plasma MHPG, subjective anxiety, autonomic symptoms and blood pressure in healthy humans. I@ Sci. 35, 1355144. Costa11 B., Hendric C. A., Kelly M. E. and Naylor R. J. (1987) Actions of sulpiride and tiapride in a simple model of anxiety in mice. Neuropharmacology 26, 195-200. Costall B., Kelly M. E., Naylor R. J. and Onaivi E. S. (1988) Actions of buspirone in a putative model of anxiety in the mouse. J. Pharm. Pharmac. 40, 494-500. Costall B., Jones B. J., Kelly E., Naylor R. J. and Tomkins D. M. (1989) Exploration of mice in a black and white test box: validation as a model of anxiety. Pharmac. Biochem. Behac. 32, 177-785. Crawley J. N. (1981) Neuropharmacologic specificity of a simple animal model for the behavioral actions of benzodiazepines. Pharmac. Biochem. Behav. 15, 695-699. Crawley J. N. (1985) Exploratory behavior models of anxiety in mice. Neurosci. Biobehav. Rev. 9, 3744. Crawley J. N. and Goodwin F. K. (1980) Preliminary report of a simple animal behavior model for the anxiolytic effects of benzodiazepines. Pharmac. Biochem. Behac. 13, 167-170. Crawley J. N., Skolnick P. and Paul S. M. (1984) Absence of intrinsic antagonist actions of benzodiazepine antagon-
er al.
ists on an exploratory model of anxiety in the mouse. Neuropharmacology 2j, 531-537. File S. E. (1990) New strategies in the search for anxiolytics. Drug Design and Delivery 5, 195-201. File S. E. and Lister R. G. (1984) Do the reductions in social
interaction produced by picrotoxin and pentylenetetrazole indicate anxiogenic action? Neuropharmacology 23, 793-796. Fujii W., Kuribara
H. and Tadokoro
S. (1989) Interaction
between caffeine and methamphetamine by means of ambulatory activity in mice. Jpn. J. Psychopharmac. 11, 225-231.
Handley S. L. and Mithani S. (1984) Effects of alphaadrenoceptor agonists and antagonists in a mazeexploration model of ‘fear’-motivated behaviour. Naunvn-Schmiedebere’s Arch. Pharmac. 321. 1-5. Handle; S. L. and McBlane J. W. (1993) 5HT drugs in animal models of anxiety. Psychopharmacology 112, 13-20. Lister R. G. (1987) The use of a plus-maze to measure anxiety in the mouse. Psychopharmacology 92, 180-185. Lopez-Rubalcava C., Saldivar A. and Fernandez-Guasti A. (1992) Interaction of GABA and serotonin in the anxiolytic action of diazepam and serotonergic anxiolytics. Pharmac. Biochem. Behav. 43, 433440. Matsumoto K., Cai B., Nakamura S. and Watanabe H. (1990) A new system for the measurement and analysis of motor activity in mice: effect of several central stimulants. Folk Pharmacol. japon. 96, 31-39. Njung’e K. and Handley S. L. (1991) Evaluation of marbleburying behaviour as model of anxiety. Pharmac. Biothem. Behav. 38, 63-67. Onaivi E. S. and Martin B. R. (1989) Neuropharmacological and physiological validation of a computer-controlled two-compartment black and white box for the assessment of anxiety. Prog. Neuropsychopharmac. Biol. Psychiat. 13, 963-976. Pellow S., Chopin P. and File S. E. (1985) Are the anxiogenie effects of yohimbine mediated by its action at benzodiazepine receptors? Neurosci. Left. 55, 5-9. Prado de Carvalho L., Venault P., Rossier J. and Chapouthier G. (1983) Anxiogenic properties of convulsive agents. Sot. Neurosci. Abstr. 9, 128. Venault P., Jaquot F., Save E., Sara S. and Chapouthier G. (1993) Anxiogenic-like effects of yohimbine and idazoxan in two behavioral situations in mice. Life Sci. 52,639-5. Wright I. K., Heaton M., Upton N. and Marsden C. A. (1992) Comparison of acute and chronic treatment of various serotonergic agents with those of diazepam and idazoxan in the rat elevated X-maze. P:ychopharmacology 107, 405414. Young R. and Johnson D. N. (1991a) Comparison of routes of administration and time course effects of zaclopride and buspirone in mice using an automated light/dark test. Pharmac. Biochem. Behav. 40, 733-737. Young R. and Johnson D. N. (1991b) A fully automated light/dark apparatus useful for comparing anxiolytic agents. Pharmac. Biochem. Behav. 40, 739-743.