Further evidence that the mouse defense test battery is useful for screening anxiolytic and panicolytic drugs: Effects of acute and chronic treatment with alprazolam

Neuropharmacology,

Pergamon

002th3908(95)00121-2

Vol. 34, No. 12, pp. 1625-1633, 1995 Copyright 8 1995 Elsevier Science Ud Printed in Great Britain. All rights reserved 002S-3908/95 $9.50 + 0.00

Further Evidence that the Mouse Defense Test Battery is Useful for Screening Anxiolytic and Panicolytic Drugs: Effects of Acute and Chronic Treatment with Alprazolam G. GRIEBEL,l* D. C. BLANCHARD,1’2 A. JUNG,l J. C. LEE,r C. K. MASUDA’ and R. J. BLANCHARD1,3 ‘Bt%%y Laborator) of Neurobiology, John A. Burns School of Medicine, University of Hawaii 1993 East-West Road, Honolulu, HI 96822, U.S.A., ‘Department of Anatomy and Reproductive Biology, John A. Burns School of Medicine, University of Hawaii, 1993 East-West Road, Honolulu, HI 96822, U.S.A. and 3Departinent of Psychology, University of Hawaii, 1993 East-West Road Honolulu, HI 96822, U.SA. (Accepted 24 July 1995) Summary-The Mouse Defense Test Battery (MDTB) has been designed to investigate defensive responses of Swiss-Webster mice confronted with a natural predator, a rat. These behaviors include flight, avoidance, defensive threat/attack responses, and risk assessment activities. Previous studies with the MDTB have suggested that this model may have some utility for the investigation of panicogenic and antipanic compounds. In the present stud:y the MDTB was used to investigate the effects of acute (0.05-l mg/kg, i.p., 30 min) or chronic (OS-2 mg/kg, one daily i.p. injection during 10 days) treatment with the benzodiazepine receptor (BZPR) full agonist and panicolytic agent alprazolam. At non motor-impairing doses (0.0545 mg/kg), acute alprazolam failed to alter the avoidance distance between the subject and the predator, the number of avoidances when the rat is approaching, predator assessment activities, defensive threat/attack responses when contact is forced bsetween the subject and the predator or contextual escape attempts after the predator was removed. This was in contrast to chronic treatment which decreased both avoidance variables at 0.5 and 1 mg/ kg, defensive threat/attack responses at all doses, and predator assessment responses at 0.5 mg/kg. In addition, the latter treatment reduced post-predator potentiation of escape attempts at 2 mg/kg. These results (1) confirm previous findings with the BZPR full agonist chlordiazepoxide, indicating that these compounds generally

attenuate antipredator defensive responses in Swiss-Webster mice; (2) support recent data indicating that panic-altering drug,s modulate flight/escape reactions, and suggest that the primary mechanism of action of drugs with efficacy against panic disorder may involve neural systems controlling flight; (3) confirm that the MDTB may be useiful for the investigation of panicolytic as well as auxiolytic agents. Keywords-Benzodiazepine,

alprazolam, antipredator defense, flight, anxiety, panic.

The clinical evidence for a dissociation of generalized anxiety disorder (GAD) and panic disorder (PD), on the basis of drug response, is controversial (Lister, 1991). However, there is general agreement that a range of tricyclic antidepressants, 5HT reuptake inhibitors (SSRIs), and also the triazolobenzodiazepine, alprazolam (e.g. Chouinard et al., 1982; Sheehan et al., 1984; Shehi and Patterson, 1984), administered on a chronic basis, are effective against panic. Some studies of the classic

*To whom correspondence should be addressed, at: CNS Pharmacology Group, CNS Research Department, SynthClabo Recherche, 31 avenue Paul-Vaillant Couturier, 92220 Bagneux, France.

benzodiazepine receptor (BZPR) full agonists such as diazepam and chlordiazepoxide indicate that they are as effective as alprazolam (Dunner et al., 1986). However, most studies suggest that classic BZPR are somewhat less effective (e.g. Wilcox, 1986; Adami et al., 1987) than the tricyclics, SSRIs or alprazolam, or, that they are as effective, but require considerably higher doses than these compounds (Dunner et al., 1986; Pollack and Rosenbaum, 1988). Alprazolam has a unique clinical and neurochemical spectrum among BZPR full agonists. Unlike these latter, it has been successfully used in the clinical management of depression (Kravitz et al., 1993). Some authors have claimed that the behavioral effects of this compound may be due to unusual interactions with BZPR (Lopez et al.,

1626

G. Griebel et al.

1988) or to the involvement of other mechanisms, like an interaction with adrenergic (Sethy and Hodges, 1982; Charney et al., 1986), adenosinergic (Kaplan et al., 1990) and serotonergic systems (Pandey et al., 1993) or an antagonistic action of the platelet activating factor (Kornecki et al., 1986). Preclinical investigations with alprazolam in experimental models of anxiety disorders revealed variable effect of the drug when administered acutely. For example, alprazolam has been reported to elicit anxiolytic-like responses in rodents confronted in conditioned as well as with ethologically-based procedures (e.g. Soderpalm et al., 1989; Jenck et al., 1992; Jones et al., 1994; Prune11 et al., 1994; Molewijk et al., 1995; Shimada et al., 1995). However, a few studies failed to reveal such actions of the drug after acute drug challenge (File, 1985; File and Pellow, 1985; Johnston and File, 1988). These latter used the rat social interaction paradigm, a method that generally yields anxiolytic effects with classic BZPR agonists like chlordiazepoxide or diazepam. By contrast, when alprazolam was administered on a repeated basis, it invariably produced anxiolytic-like effects (Johnston and File, 1988; Ellis et al., 1990). There are few well accepted animal models of psychiatric disorders. However, a nu, .ber of animal models of anxiety have been proposed, most of which involve exposure of animals tti external (e.g. cues previously paired with footshock) or internal (e.g. drug states) stimuli which are assumed to be capable of inducing anxiety in humans. The actual measures taken include suppression of previously punished activities (Howard and Pollard, 1991), conditioned emotional responses such as the fear-potentiated startle (Davis, 1991) a range of sonic and ultrasonic vocalizations (Newman, 1991) and social, exploratory, and defensive behaviors (Lister, 1991). Many similar tests involving both conditioned (e.g. Martin, 1993) or unconditioned (e.g. Graeff, 1991) behaviors have also been used to model panic, and the major differentiation of whether a particular measure is interpreted as ‘anxiety-like’ or ‘panic-like’ often reflects the drugs to which these measures respond. Several authors have suggested that the spontaneous activation of neuronal systems mediating the flight component of defense reactions may underlie human panic disorder (Graeff, 1990; Deakin and Graeff, 1991; Deakin et al., 1991). We recently adressed this issue by showing that yohimbine potentiated flight behavior in Swiss-Webster mice confronted with a natural predator in an oval runway, while chronic treatment with the panicolytic agents imipramine and fluoxetine markedly reduced flight responses (Blanchard et al., 1993a; Griebel et al., 1995a). Moreover, the specific anti-GAD compounds chlordiazepoxide and gepirone, a 5-HTrA receptor ligand, were found to be devoid of any flightmodulating action in this test (Griebel et al., 1995~).

Taken together, these findings suggested that drugs effected changes in defense reactions by potentiating or inhibiting neural mechanisms mediating flight, and that the oval runway may have some utility for the investigation of panic-modulating compounds. This Mouse Defense Test Battery (MDTB) has also provided specific profiles of drug effects on several other defensive behaviors, for chlordiazepoxide and gepirone. For example, gepirone reduced contextual escape attempts to the situation associated with the predator, and, it also inhibited predator assessment behaviors, while chlordiazepoxide only altered the latter. Finally, both drugs strongly reduced defensive threat and attack responses when contact was forced between the subject and the predator. These results suggest that when a wide variety of defensive behaviors are measured, differential patterns of drug response may be associated with anxiolytic action (e.g. defensive threat/attack responses and risk assessment measures) or with panicolytic action (flight/avoidance). The objective of the present study was to further investigate the validity of the MDTB as an animal model of anxiety and PD by assessing the behavioral effects of acute as well as chronic treatment with the anxiolytici panicolytic compound alprazolam in this experimental procedure.

METHODS Animals Subjects were 139 naive male Swiss-Webster mice obtained from Simonsen Laboratories (CA), 60-75 days old at the beginning of the experiment, and 2 male Long Evans rats (400-500 g) bred in the laboratory. Prior to experimental testing, they were housed singly in 30 x 20 x 14 cm; standard cages (mice: rats: 44 x 30 x 20 cm) containing a constant supply of food pellets and water. All animals were maintained under standard laboratory conditions (21-23°C) and kept on a 12 hr light/dark cycle with light onset at 6 a.m. Drug and treatment groups Alprazolam (The Upjohn Co., Kalamazoo, MI) was suspended in an isotonic saline vehicle with a drop of tween 80 to various concentrations such that intraperitoneal injections were always at a constant volume of 10.0 ml/kg. Mice were randomly assigned to the following three conditions: (a) acute alprazolam: control group (n = 15) and drug treatment groups (0.05 mg/kg, n = 17; 0.5 mg/kg, n = 19; 1 mg/kg, n = 18); (b) chronic alprazolam treatment: control group (n = 15) and drug treatment groups (0.5, 1 and 2 mg/kg, n = 15). In this last case, mice received 10 daily injections of either saline or alprazolam. Since some evidence of tolerance to the behavioral effects of alprazolam has emerged from previous literature (see Discussion), higher doses were used in the chronic experiment; (c) Non exposed group:

Alprazolam and antipredator defense to evaluate the impact of predator exposure on animals’ performance during the post-test, we measured the behavioral parameters in a non-exposed group of mice (n = 10, saline). The last injection for each subject was given 30 min before the experiment was carried out. The mice were tested in an order randomized for drug treatment. Apparatus The test was conducted in an oval runway, 0.40 m wide, 0.30 m high, and 6.0 m in total length, consisting of two 2 m straight segments joined by two 0.4 m curved and separalted by a median wall segments (2.0 x 0.30 x 0.06) (Fig. 1). The apparatus was elevated to a height of 0.80 m from the floor to enable the experimenter to easily hold the rat, while minimizing the mouse’s visual contact with him. All parts of the apparatus were made of black Plexiglas. The floor was marked every 20 cm to facilitate distance measurement. Activity was recorded with videocameras mounted above the apparatus. Behavioral assessments were made from the recordings with the observer unaware of the original pretreatment. Experiments were performed under red light between 1 p.m. and 5 p.m. After removal of each animal, the runway field was carefully mopped using hot soapy water to remove any residual odour due to urine, faeces or to the predator. Procedure Pre-test: motor activity before exposure to the predator. Thirty minuteis after injection of alprazolam or vehicle, subjects were placed into the runway for a 3min familiarization period, in which line crossings, wall rears, wall climbs, and ju.mp escapes were recorded (min 1-3). Wall rear, wall climb and jump escape measures provide an index of contextual escape attempts. Consequently, these measures were combined in one single measure called ‘escape attempts’. In the non-exposed condition, the pre-test was followed by a 10 min period without predator exposure. After this period, the 3-min post-test was given. Reactions to the predatolr Predator avoidance test (min 4-6). Immediately after the 3-min familiarization period, a hand-held conscious rat was introduced into tlhe runway and brought up to the subject at a speed of approx 0.5 m/set. Approach was

Fig. 1. The oval runway test.

1627

terminated when contact with the subject was made or the subject ran away from the approaching rat. If the subject fled, avoidance distance (the distance from the rat to the subject at the point of flight) was recorded. This was repeated five times. Chase/flight test (min 7-S). The hand-held conscious rat was brought up to the subject at a speed of approx 2.0 m/set. The time it took to chase the subject a distance of 18 m was recorded. Overall flight speed (rn/sec) and maximum flight speed (measured when the subject is running straight) were subsequently calculated from these measures. In addition, the following parameters were recorded: number of stops (pause in movement), orientations (subject stops, then orients the head toward the rat) and reversals (subject stops, then runs in the opposite direction). Straight alley (min 9-11). The runway was then converted to a straight alley by the closing of a door at each end (Fig. 1). Three approaches, 15 set each, respectively at 1.20, 0.80 and 0.40 m were made by a hand-held terminally anaesthetized rat toward the subject in this inescapable runway. Measures taken included immobility time, closest distance between the subject and the rat and the number of approaches/withdrawals (subject must move more than 0.2 m forward from the closed door, then return to it). Finally, the experimenter brought the rat up to contact the subject. For each such contact, bites, vocalizations, upright postures and jump attacks by the subjects were noted. This was repeated three times. Post-test: contextual defense Immediately after the straight alley test, the predator was removed and doors were opened. Line crossings, wall rears, wall climbs, and jump escapes were recorded during a 3-min session. As was the case in the pre-test, the three latter responses were combined in an “escape attempts” measure (min 12-14). Statistics Data were analyzed by one-way analysis of variance (ANOVA) (avoidance distance, flight speeds, immobility time and closest distance between animals) or the nonparametric Kruskal-Wallis ANOVA for some infrequently occurring or highly variable behaviors (number of avoidances, reversals, orientations, approaches/withdrawals, bites, vocalizations, upright postures and jump attacks). Subsequent comparisons between treatment groups and control were carried out using NewmanKeuls procedures or the nonparametric Mann-Whitney U-test. In the contextuel defense test, differences were evaluated by a combined repeated measures ANOVA followed by a Newman-Keuls post-hoc comparison (line crossings) or by the Mann-Whitney U-test and Wilcoxon matched pairs test if the behavior occurred infrequently (escape attempts). Non parametric data are displayed as

1628

G. Griebel et

mean + standard error in order to illustrate the group variation. RESULTS

Contextual escape attempts: motor activity before and after exposure to the predator (Fig. 2) Non-exposed mice. ANOVA revealed that line crossings were significantly decreased in the post-test = 16.55, P < 0.0007). By contrast, the Wilcoxon Fl,lS matched pairs test failed to indicate any reliable effect of the 10 min free running session on escape attempts made during the first, as opposed to the last, 3-min period of the session. Alprazolam (acute treatment). ANOVA revealed a reliable treatment effect for line crossing (F3,64= 3.94, P < 0.012) and the frequency of escape attempts (&$s = 28.54, P < 0.0001). Post-hoc analyses indicated a reliable decrease in both measures at 1 mg/kg (P c 0.001 vs control). Furthermore, both behavioral measures increased significantly in the post-test period following presentation and removal of the predator (line crossing: F1,e4 = 34.4, P < 0.0001; escape attempts: Wilcoxon pair test: P < 0.0001). Finally, 4 x 2 (dose x pre/post-test) two-way ANOVA revealed a reliable interaction effect for line crossing (F3,a = 11.58, attempts (Friedman: escape P < 0.0001) and N1,6s = 21.06, P c 0.0001). Subsequent Newman-Keuls analyses indicated that predator exposure significantly increased post-test line crossings in mice treated with the two highest doses (0.5 and 1 mgkg) (P < 0.004 and

0-l

I

I

NON EXPOSED

I

I

0.00 0.05

I

I

0 50

1m

ACUTE

I

I

I

,

I

al.

P < 0.0001, respectively, vs pre-test) and Wilcoxon matched pairs test analyses indicated that escape attempts increased in the saline-treated group and for the 0.05 and 0.5 mgikg groups. Alprazolam (chronic treatment). Alprazolam produced a reliable effect on frequency of line crossing = 3.94, P< 0.01) and escape attempts @3,56 (H3,6o = 12.72, P c 0.003). Post-hoc analyses indicated reliably more line crossings (Newman-Keuls: P < 0.03 vs control) and fewer escape attempts at 2 mg/kg (MannWhitney: P < 0.0001 vs control). In addition, both behavioral measures increased significantly in the posttest period following presentation and removal of the predator (line crossings: F1,56 = 14.59, P < 0.003; escape attempts: Wilcoxon pair test: P < 0.0001). Finally, 4 x 2 ANOVA failed to indicate a reliable interaction effect for line crossing (F3,56 = 1.4) but this interaction was reliable for escape attempts (j!r1,60= 31.34, P < 0.0001). Subsequent analysis with Wilcoxon pair tests indicated that post-test escape attempts increased for ail groups. Reactions (Fig. 3)

to the predator: predator -

avoidance

test

Alprazolam (acute treatment). ANOVA revealed a reliable drug effect on frequency of avoidance = 7.57, P < 0.0002), and subsequent Newman(F3,65 Keuls comparisons indicated a reliable decrease in this measure at 1 mg/kg. Alprazolam (chronic treatment). ANOVA indicated a reliable drug effect on the avoidance frequency

I

0.00 0.50 tea 2.00 lpi CHRONIC

Fig. 2. Effects of acute and chronic (10 days, once a day) treatments with alprazolam on two response measures before @e-test) and after (post-test) the exposure to the predator. Non Exposed group was not confronted with the predator and injected with saline. Data represent Mean + SEM. *P < 0.05, **P < 0.01 and ***P e 0.001 (vs pre-test); (d) (vs vehicle control).

0.09 0.05 0.50 1.M ACUTE

0.W

050

1.00

CHRONIC

Z.Wmg/kg ”

Fig. 3. Runway measures of avoidance to an approaching predator for mice administered alprazolam. Data represent Mean + SEM. *PC 0.05, **PC 0.01 and ***PC 0.001 (vs vehicle control).

1629

Alprazolam and antipredator defense Table 1. Effects of alprazolam on behavioral responses of mice chased by a predator Maximum speed (m/set)

Overall speed (m’sec)

Frequency of stops

Frequency of orientations

Frequency of reversals

Acute 0 0.05 mg/kg 0.5 1

0.60 0.60 0.60 0.50

f * f f

0.05 0.04 0.02

0.80 1.00 0.90 0.80

+ + * f

0.06 0.07 0.06 0.10

6.40 8.60 4.60 3.70

f + k +

1.71 1.73 0.98 0.65

1.60 1.20 0.70 0.10

* * * +

0.43 0.41 0.39 0.1ot

0.60 0.30 0.10 0.10

f f f f

0.19 0.14 0.07 0.10*

Chronic 0 0.5 mgkg 1 2

0.50 0.50 0.50 0.40

* f * *

0.04 0.03 0.05 0.02

1.00 0.90 0.90 0.70

+ k f f

0.08 0.06 0.05 0.05t

8.50 6.90 6.10 6.20

+ + f f

1.53 4.50 2.14 1.15

1.20 0.40 0.60 0.60

k + + +

0.28 0.16* 0.24 0.16

0.90 0.10 0.20 0.20

+ * f f

0.26 0.09 0.11 0.11

0.04

Data represent Mean + SEM. “P c 0.05 and tP < 0.001 (vs vehicle control).

Table 2. Effects of alprazolam in the straight alley on behavioral reactions to a predator which remained at a constant distance from the subject Frequency of approaches/withdrawals

Closest distance between animals (cm)

Immobility time (set)

Acute 0 0.05 mg/kg 0.5 1

3.5 4.1 4.6 2.7

f * + +

0.5 0.5 0.5 0.4

112.1 103.8 114.9 110.8

f + * +

13.8 8.7 11.7 12.9

5.5 5.5 6.6 14.2

Chronic 0 0.5 mg/kg 1 2

3.2 3.3 4.4 3.5

+ k * +

0.2 0.6 0.4 0.5

122.1 106.9 90.9 97.3

k * f +

8.1 15 15.7 18.5

2.9 4.1 6.0 6.1

f 1.8 + 1.8 f 1.6 &- 3.3* + f + f

1.2 1.7 2.2 1.9

Data represent Mean + SEM. *P c 0.05 (vs vehicle control).

3.14, P < 0.03) and the predator-subject distance at which avoidance occurred (F3,56 = 6.94, P < 0.0005). The latter measure was reliably decreased at all doses tested, but the former meaLsurewas reduced only at 2 mg/ kg. Taken together, Predator Avoidance findings demonstrate higher efficacy of the drug after repeated administration. (F3,56 =

Flight/predator orientation test (Table 1) Alprazolam (acute treatment). ANOVA revealed a reliable main effect for the frequency of orientations (Kruskal-Wallis: H3,e9 :=16.33, P c 0.001) and the number of reversals (Kruskal-Wallis: Hs,eg = 10.39, P < 0.015). Alprazolam dlecreased the frequency of both behavioral parameters, but this was statistically significant only in the 1 mg/kg group. Alprazolam (chronic treatment). ANOVA indicated a reliable effect of drug trea.tment on maximum flight speed of orientations cF3,56 = 4.97, P < 0.004) and frequency (Kruskal-Wallis: H3,60 :=9.21, P < 0.02). Subsequent analyses revealed that the 2 mg/kg dose significantly reduced flight speed while frequency of orientations was lower at 0.5 mg/kg.

Alprazolam (chronic treatment). ANOVA failed to indicate any reliable effects of chronic alprazolam treatment. Forced contact with the predator (Fig. 4) Alprazolam (acute treatment). ANOVA indicated a reliable effect for frequency of vocalization (KruskalWallis: H3,69 = 10.03, P < 0.01) and frequency of biting (Kruskal-Wallis: H3,69 = 10.20, P < 0.0002). MannWhitney comparisons indicated that both measures were reliably decreased at 1 mg/kg. Alprazolam (chronic treatment). ANOVA indicated a reliable effect for frequency of vocalization (KruskalWallis: H3,60 = 21.52, P c O.OOOl), biting (KruskalWallis: H3,60 = 29.83, P < 0.0001) and upright posture (Kruskal-Wallis: H3,60 = 8.18, P < 0.04). Mann-Whitney U-tests indicated a reliable decrease in vocalization and biting at all doses and in upright posture at 1 mg/kg. As was the case in the Predator Avoidance test, chronic treatment was required to induce a clear-cut decrease in these defensive behaviors. DISCUSSION

Predator approach: straight alley (Table 2)

Myorelaxation and sedation measures

Alprazolam (acute treatment). ANOVA failed to indicate reliable effects of treatment on any measure other than immobility time (F3,65 = 2.74, P < 0.05), which reflected an increase at 1 mglkg.

Two measures, line crossings and escape attempts, provide data relevant to an evaluation of the sedative or myorelaxant effects of alprazolam. The line crossing measure in the pre-test suggested that the highest acute

G. Griebel et al.

1630

Effects preceding and following predator “contextual escape attempts”

4 VOCALIZATIONS

1

3 ,UPRlGHT

POSTURES

1

3 JUMP ATTACKS TOWARD RAT

0.00

0 05

0.50

ACUTE

1 00

0.00

0 50

1.00

2.00 ;g/kg

CHRONIC

Fig. 4. Frequencies of biting, defensive threat vocalization, upright posture and jump attacks to forced contact with a deeply anaesthetized rat for subjects under varying doses of alprazolam. Data represent Mean f SEM. *P c 0.05, **P < 0.01 and ***P c 0.001 (vs vehicle control).

dose of alprazolam may have produced such effects. This view is strongly supported by findings of escape attempts, indicating that the higher doses of acute and chronic alprazolam produced a decrease in these responses. However, because anxiolysis may also involve reduced contextual defensiveness, such that escape attempts might reflect anxiety reduction as well as myorelaxant or sedative effects, we considered only those dose levels that produced both a reduction in line crossings and a decrease in escape attempts to be sedative/myorelaxant. Consequently, only doses lower than 1 mg/kg can be considered to produce a more specific effect on the defensive behaviors measured in the acute alprazolam experiment. Also, on this basis, all three doses of chronic alprazolam can be considered to have specific effects in the present study. The sedative/myorelaxant effect of acute alprazolam is consistent with preclinical and clinical studies reporting evidence of muscle relaxant and sedative actions of the drug (e.g. Smith and Kroboth, 1987; Tesar et al., 1991; Bourin et al., 1992; Jenck et al., 1992). For instance, Jenck et al. (1992) recently showed that a single administration of alprazolam in mice and rats impaired their motor performances in the horizontal wire test and the rotarod procedures with ED5e values ranging from 0.4 to 4.4 mg/kg. Furthermore, the lack of a clear-cut effect of chronic alprazolam on sedative/ myorelaxant measures also agrees with several reports demonstrating that tolerance develops to the sedative and myorelaxant effects of the drug (Seidel et al., 1985; Soderpalm et al., 1989; Bourin et al., 1992).

exposure:

In the present oval runway situation there was a dramatic rise in escape attempts against the wall in the post-predator period, compared to an equivalent period prior to the introduction of the predator. These pre-post predator differences were not seen in control animals given similar handling and placement in the test apparatus but not exposed to the predator. They thus provide an index of contextual defensiveness, i.e. defensiveness to the situation following its association with the predator. Blanchard and collaborators (Blanchard et al., 1993a) also obtained contextual defense changes after predator exposure in a closed chamber, although the actual behaviors involved were different, as is consonant with previous analyses (Blanchard et al., 1990a) indicating that reactions to the predator itself are strikingly different in situations that permit, or do not permit, escape. Both treatments with alprazolam prevented the potentiation of these responses at the two highest doses tested. However, on the basis of the findings of decreased locomotor activity in the pre-test, this effect with acute alprazolam may be nonspecific. The lack of specific action of acute alprazolam on these responses closely resemble the results we recently obtained with acute chlordiazepoxide (5-25 mg/kg) (Griebel et al., 1995b), and hence confirm a general lack of specific effect of acutely administered BZPR full agonists on the post-encounter contextual escape behaviors. In addition, the 2 mg/kg dose of chronic alprazolam produced a profile on escape attempts which is akin to the effect we recently obtained after repeated administration of the panicolytic agents imipramine and fluoxetine on these responses (Griebel et al., 1995a). Similarly, acute administration of the 5-HTtA receptor agonists 8-OH-DPAT (0.5-10 mg/kg) and gepirone (510 mg/kg) was also found to reduce escape responses (Griebel et al., 1995~). However, in this latter case, postrat inhibition of escape attempts appeared to be nonspecific as line crossings were markedly reduced at a similar dosage. This suggested that some components of the mouse 5-I-IT motor syndrome were involved in this effect (Griebel et al., 1995~). Taken together, these findings demonstrate that post-rat escape attempts are reduced by chronic antipanic drug challenge only and thus suggest that this response may be of particular relevance in the screening of panicolytic drugs. Effects during exposure to the predator Flight. Saline-treated mice almost invariably showed active flight to the approaching predator, with a consistent (prey-predator) avoidance distance of about 0.8 m in all control groups. Except at sedative/myorelaxant level, acute alprazolam did not reduce avoidance distance or frequency. This was in contrast to chronic alprazolam which dramatically reduced the prey predator distance at which flight occurred, at all doses tested. On

1631

Alprazolam and antipredator defense the basis of recent findings in the MDTB that the panicpromoting drug yohimbine potentiates flight reactions (Blanchard et al., 1993a), while the panicolytic agents imipramine and fluoxetine, given on a repeated basis, markedly reduced the prey-predator distance at which flight occurred (Griebel ei! al., 1995a), the present data further suggest that flight behavior in the MDTB provides measures that serve as an effective experimental model of PD. Moreover, this view is supported by recent findings in the MDTB with the specific anti-GAD compounds chlordiazepoxide and gepirone, administered acutely, which were found inactive on flight-related responses (Griebel et al., 1995b; Griebel et al., 1995~). Predator assessment in the chasel’ight and the straight alley test. During the chase/flight test, saline-treated mice

frequently (about 8 times per test session) exhibited a salient behavior pattern consisting of an abrupt movement arrest often followed by orientation to the oncoming predator and sometimes a reversal of movement to approach the predator. Furthermore, when the subject was constrained in one part of the runway by closing doors at opposite ends, mouse subjects again showed a pattern of behavioral responses such as frequent movements, which often consisted of approaches to the predator, followed by witlhdrawals, a pattern apparently related to predator assessment. None of the nonsedative/ myorelaxant doses of acute alprazolam produced a reliable reduction of any of these behaviors. A single predator assessment measure, orientation, was reduced with chronic alprazolam and this effect was reliable at 0.5 mg/kg only. A low baseline level might account for the lack of dose-related effect on this measure. In any case, these findings sugg,est only a minimal effect of acute and chronic alprazolam on predator or risk assessment behaviors, and thus contrast with the chlordiazepoxide data in the MDTB (Griebel et al., 1995b) as well as with pr’evious findings with diazepam in the rat (Blanchard et al., 1990; Blanchard et al., 1990b) showing that BZPR full agonists reduce risk assessment when measured from a control baseline that includes risk assessment behavior. Interestingly, similar weak effects on predator assessment :responses have recently been observed after long-term treatments with two other panicolytic agents, namely imipramine and fluoxetine (Griebel et al., 1995a). Defensive threat and attack. When contact was forced between the predator and1 the subject, the latter almost invariably showed a pattern of defensive threat and attack toward the rat. This included vocalization, upright posture, jump attack toward the rat, and biting to it. Chronic treatment with alprazolam reduced biting and vocalization at all dose-levels, while only the highest dose of acute alprazolam produced a similar effect, a result that is generally consistent with previous studies of the effects of BZPR full agonists on defensive behavior in animals. Thus, for several species and a variety of stimulus contexts, these compounds have been shown to

inhibit defensive threat and attack components (Rodgers and Waters, 1985; Blanchard et al., 1993b; Griebel et al., 1995b). In summary, in the light of recent findings from the MDTB with panic-modulating drugs (Blanchard et al., 1993a; Griebel et al., 1995a), the present data that only chronic alprazolam reduced the prey-predator distance that elicited flight and post-test potentiation of escape attempts, provides additional evidence that flight/escape responses may be an important element in panic. Furthermore, the present results indicate a general lack of effect on predator assessment activities after both acute and chronic treatments. However, they demonstrate that chronic alprazolam strongly reduced defensive threat/attack responses. With the exception of these latter measures, the present data demonstrate that alprazolam produces a behavioral profile which is somewhat different from that observed in the MDTB with the classic BZPR agonist chlordiazepoxide. However, the findings of similar drug effect of alprazolam and other panicolytic agents confirm that only a subset of defense behaviors (flight, escape attempts and defensive threat/ attack) are affected by such compounds. In conclusion, the present study with alprazolam suggest further that the MDTB may have utility for the investigation of both anxiolytic and panicolytic drugs. Acknowledgements-The research reported in this article was supported by USPHS Awards NIH MH42803, and RR03061. Guy Griebel was supported by funds from Institut de Recherches Intemationales Servier (Courbevoie, France).

REFERENCES Adami M., Balestrieri M., Fiorio R. and Bellantuono C. (1987) I1 trattamento farmacologico de1 disturb0 da attacchi di panico. (Pharmacological treatment of panic attacks.). Riv. Psichiatr. 22: 267-279. Blanchard D. C., Blanchard R. J., Tom P. and Rodgers, R. J. (1990) Diazepam changes risk assessment in an anxiety/ defense test battery. PsychopharmacologL 101:511-518. Blanchard R. J., Blanchard D. C., Rodgers J. and Weiss S. M. (1990a) The characterization and modelling of antipredator defensive behavior. Neurosci. Biobehav. Rev. 14: 463472. Blanchard R. J., Blanchard D. C., Weiss S. M. and Meyer S. (1990b) The effects of ethanol and diazepam on reactions to predatory odors. Pharmac. Biochem. Behav. 35: 775-780. Blanchard R. J., Taukulis H. K., Rodgers R. J., Magee L. K. and Blanchard D. C. (1993a) Yohimbine potentiates active defensive responses to threatening stimuli in Swiss-Webster mice. Pharmac. Biochem. Behav. 44: 673-681. Blanchard R. J., Yudko E. B., Rodgers R. J. and Blanchard D. C. (1993b) Defense system psychopharmacology: an ethological approach to the pharmacology of fear and anxiety. Behav. Brain Res. 58: 155-165. Bourin M., Hascoet M., Mansouri B., Colombel M. C. and Bradwejn J. (1992) Comparison of behavioral effects after single and repeated administrations of four benzodiazepines in three mice behavioral models. J. Psychiatry Neurosci. 17: 72-77.

G. Griebel et al. Charney D. S., Breier A., Jatlow P. I. and Heninger G. R. (1986) Behavioral, biochemical, and blood pressure responses to alprazolam in healthy subjects: interactions with yohimbine. Psychopharmacology 88: 133-140. Chouinard G., Annable L., Fontaine R. and Solyom L. (1982) AIprazolam in the treatment of generalized anxiety and panic disorders: a double-blind placebo-controlled study. Psychopharmacology 77: 229-233. Davis M. (1991) Animal models of anxiety based on classical conditioning: The conditioned emotional response and the fear-potentiated startle reflex. In: Psychopharmacology of Anxiolytics and Antidepressants (File S. E., Ed), pp. 187212. Pergamon Press, New York. Deakin J. F. W. and Graeff F. G. (1991) 5-HT and mechanisms of defense. J. Psychopharmac. 5: 305-315. Deakin J. F. W., Guimaraes F. S., Wang F., Hellewell J. and Hensman R. (1991) 5-HT receptor mechanisms in human anxiety. In: New Concepts in Anxiety (Briley M. and File S. E., Eds), pp. 74-93. CRC Press, New York. Dunner D. L., Ishiki D., Avery D. H., Wilson L. G. and Hyde T. S. (1986) Effect of alprazolam and diazepam on anxiety and panic attacks in panic disorder: a controlled study. J. Clin. Psychiatry 47: 458-460. Ellis D. M., Fontana D. J., McCloskey T. C. and Commissaris R. L. (1990) Chronic anxiolytic treatment effects on conflict behavior in the rat. Pharmac. Biochem. Behav. 37: 177-186. File S. E. (1985) Animal models for predicting clinical efficacy of anxiolytic drugs: social behaviour. Neuropsychobiology 13: 55-62. File S. E. and Pellow S. (1985) The effects of triazolobenzodiazepines in two animal tests of anxiety and in the holeboard. Br. J. Pharmac. 86: 729-735. Graeff F. G. (1990) Brain defense system and anxiety. In: Handbook of Arzxiety. The Neurobiology of Anxiety, (Burrows G. D., Roth M. and Noyes R., Eds), Vol. 3., pp. 307-343. Elsevier, Amsterdam. Graeff F. G. (1991) Neurotransmitters in the dorsal periaqueductal grey and animal models of panic anxiety. In: New Concepts in Anxiety (Briley M. and File S. E., Eds), pp. 288 307. CRC Press, New York. Griebel G., Blanchard D. C., Agnes R. S. and Blanchard R. J. (1995a) Differential modulation of antipredator defensive behavior in Swiss-Webster mice following acute or chronic treatment with imipramine and fluoxetine. Psychopharmacology. 120: 57-66. Griebel G., Blanchard D. C., Jung A. and Blanchard R. J. (1995b) A model of ‘antipredator’ defense in Swiss-Webster mice: effects of benzodiazepine receptor ligands with different intrinsic activities. Behav. Pharmac. In press. Griebel G., Blanchard D. C., Jung A., Masuda C. K. and Blanchard R. J. (1995~) 5-HTtA agonists modulate mouse antipredator defensive behavior differently from the 5-I-IT= antagonist pirenperone. Pharmac. B&hem. Behav. 51: 235244. Howard J. L. and Pollard G. T. (1991) Effects of drugs on punished behavior: preclinical test for anxiolytics. In: Pychopharmacology of Anxiolytics and Antidepressants (File S. E., Ed), pp. 131-154. Pergamon Press, New York. Jenck F., Moreau J. L., Bonetti E. P., Martin J. R. and Haefely W. E. (1992) Ro 19-8022, a nonbenzodiazepine partial agonist at benzodiazepine receptors: neuropharmacological profile of a potential anxiolytic. J. Pharmac. Exp. Ther. 262: 1121-1127.

Johnston A. L. and File S. E. (1988) Profiles of the antipanic compounds, triazolobenzodiazepines and phenelzine, in two animal tests of anxiety. Psychiatry Res. 25: 81-90. Jones G. H., Schneider C., Schneider H. H., Seidler J., Cole B. J. and Stephens D. N. (1994) Comparison of several benzodiazepine receptor ligands in two models of anxiolytic activity in the mouse: an analysis based on fractional receptor occupancies. Psychopharmacology 114: 191-199. Kaplan G. B., Tai N. T., Greenblatt D. J. and Shader R. I. (1990) Separate and combined effects of caffeine and alprazolam on motor activity and benzodiazepine receptor binding in vivo. Psychopharmacology 101: 539-544. Komecki E., Ehrlich Y. H. and Lenox R. H. (1986) Plateletactivating factor-induced aggregation of human platelets specifically inhibited by triazolobenzodiazepines. Science 226: 1454-1456. Kravitz H. M., Fawcett J. and Newman A. J. (1993) Alprazolam and depression: a review of risks and benefits. J. Clin. Psychiatry 54 (Suppl.): 78-84. Lister R. G. (1991) Ethologically based animal models of anxiety disorders. In: Psychopharmacology of Anxiolytics and Antidepressants, (File S. E., Ed.), pp. 155-186. Pergamon Press, New York. Lopez F., Miller L. G., Greenblatt D. J., Paul S. M. and Shader R. I. (1988) Low-dose alprazolam augments motor activity in mice. Pharmac. B&hem. Behav. 30: 511-513. Martin P. (1993) Effects of anxiolytic and antidepressant drugs in an animal model of panic. In: Anxiety: Neurobiology, Clinic and Therapeutic Perspectives, (Hamon M., Ollat H. and Thitbot M. H., Eds), pp. 203-204. Les editions INSERM/John Libbey Eurotext Ltd, Paris. Molewijk H. E., Vanderpoel A. M., Mos J., Vanderheyden J. A. M. and Olivier B. (1995) Conditioned ultrasonic distress vocalizations in adult male rats as a behavioural paradigm for screening anti-panic drugs. Psychopharmacology 117: 3240. Newman J. D. (1991) Vocal manifestations of anxiety and their pharmacological control. In: Psychopharmacology of Anxiolytics and Antidepressants, (File S. E., Ed.), pp. 251-260. Pergamon Press, New York. Pandey S. C., Dwivedi Y. and Pandey G. N. (1993) Effects of treatment with mCPP and alprazolam on serotonin receptor subtypes and phosphoinositide signalling. A.C.N.P., p. 167 (abstract). Pollack M. H. and Rosenbaum J. F. (1988) Benzodiazepines in panic-related disorders. Special Issue: Perspectives on panicrelated disorders. J. Anx. Disord. 2: 95-107. Prune11 M., Escorihuela R. M., Femandez Temel A., Nunez J. F. and Tobena A. (1994) Anxiolytic profiles of alprazolam and ethanol in the elevated plus-maze test and the early acquisition of shuttlebox avoidance. Pharmac. Rex 29: 3746. Rodgers R. J. and Waters A. J. (1985) Benzodiazepines and their antagonists: a pharmacoethological analysis with particular reference to effects on “aggression”. Neurosci. Biobehav. Rev. 9: 21-35. Seidel W. F., Cohen S. A. and Dement W. C. (1985) Effects of alprazolam and diazepam on the daytime sleepiness of nonanxious subjects. Psychopharmacology 87: 194-197. Sethy V. H. and Hodges D. H. (1982) Role of beta adrenergic receptors in the antidepressant activity of alprazolam. Res. Commun. Chem. Path. Pharmac. 36: 329-332. Sheehan D. V., Coleman J. H., Greenblatt D. J., Jones K. J.,

Alprazolam and antipredator defense Levine P. H., Orsulak P. J., Peterson M., Schildkraut J. J., Uzogara E. and Watkins D. (1984) Some biochemical correlates of panic attacks with agoraphobia and their response to a new treatment. J. Clin. Psychopharmac. 4: 66-75. Shehi M. and Patterson W. M. (1984) Treatment of panic attacks with alprazolam and propranolol. Am. J. Psychiatry 141: 900-901. Shimada T., Matsumoto K., Osanai M., Matsuda H., Terasawa K. and Watanabe H. (1995) The modified light/dark transition test in mice: Evaluation of classic and putative anxiolytic and anxiogenic drugs. Gen. Pharmac. 26: 205 210. Smith R. B. and Kroboth P. D. (1987) Influence of dosing

1633

regimen on alprazolam and metabolite serum concentrations and tolerance to sedative and psychomotor effects. Psychopharmacology 93: 105-112. Siiderpalm B., Eriksson E. and Engel J. A. (1989) Anticonflict and rotarod impairing effects of alprazolam and diazepam in rat after acute and subchronic administration. Prog. Neuropsychopharmac. Biol. Psychiatry 13: 269-283. Tesar G. E., Rosenbaum J. F., Pollack M. H., Otto M. W., Sachs G. S., Herman J. B., Cohen L. S. and Spier S. A. (1991) Double-blind, placebo-controlled comparison of clonazepam and alprazolam for panic disorder. J. Clin. Psychiatry 52: 69-76. Wilcox J. A. (1986) An open comparison of diazepam and alprazolam. J. Psychoactive Drugs 18: 159-160.