You are on page 1of 10

BR A I N R ES E A RC H 1 2 3 1 ( 2 00 8 ) 9 3 1 02

a v a i l a b l e a t w w w. s c i e n c e d i r e c t . c o m

w w w. e l s e v i e r. c o m / l o c a t e / b r a i n r e s

Research Report

Intra-periaqueductal gray matter injections of midazolam fail


to alter anxiety in plus-maze experienced mice

Luciana Maria dos Reis a,b , Azair Canto-de-Souza a,


a
Grupo de Psicobiologia/Depto de Psicologia/CECH UFSCar, Rod. Washington Lus, Km 235, 13565-905 So Carlos, SP, Brazil
b
Programa de Ps-Graduao em Cincias Fisiolgicas-PPG-CF/CCBS UFSCar, Rod. Washington Lus, Km 235,
13565-905 So Carlos, SP, Brazil

A R T I C LE I N FO AB S T R A C T

Article history: It is well known that prior experience to the elevated plus-maze increases the avoidance of
Accepted 15 June 2008 rodents to the open arms and impairs the anxiolytic-like effect of benzodiazepines evaluated
Available online 21 June 2008 during a subsequent exposure to the maze, a phenomenon known as one-trial tolerance.
Centrally injected benzodiazepine drugs attenuate anxiety in some limbic structures, such as
Keywords: hypothalamus, amygdala and the midbrain periaqueductal gray (PAG). This study
Anxiety investigated the effects of intra-PAG infusions of midazolam (MDZ) in maze-nave and
EPM maze-experienced mice. The antiaversive effects of MDZ (3.0 nmol and 30 nmol in 0.1 l) were
One-trial tolerance evaluated by prior injection of flumazenil (16 nmol/0.1 l), a benzodiazepine receptor
Midazolam antagonist, into the same midbrain site. Test videotapes were scored for conventional
Flumazenil measures of anxiety and locomotor activity, as well as a range of ethological measures
Periaqueductal gray related to risk assessment. In maze-nave mice, intra-PAG infusions of MDZ increased %
open arm entries (3.0 nmol) and % open arm time (3.0 and 30 nmol). These effects were
observed in the absence of significant changes in locomotor activity, indicating a selective
anxiolytic-like effect of MDZ. The antiaversive effects of MDZ were completely blocked by
prior injection of flumazenil which in turn did not alter any other behavioral measure. In
maze-experienced mice, intra-PAG infusion of MDZ did not modify any behavioral measure.
Taken together, present results corroborate previous studies demonstrating that GABA/
benzodiazepine receptor complex located within the PAG plays a role on anxiety modulation
in maze-nave mice as well as indicate its involvement in the OTT phenomenon.
2008 Elsevier B.V. All rights reserved.

1. Introduction Kennett et al., 1995, 1998; Pietraszek et al., 2005; Smith and Barret,
1997), Vogel conflict (Agmo et al., 1995; Millan, 2003) and elevated
Administration of benzodiazepines usually produces anxiolytic plus-maze (Cole and Rodgers, 1993; Griebel et al., 1996; Lister, 1987;
effects in animal models of anxiety such as social interaction test Nunes-de-Souza and Rodgers, 2000; Rosa et al., 2000; Treit, 1991).
(for review see File and Seth, 2003), lightdark exploration (for Since its introduction (Handley and Mithani, 1984), the
review see Hascot et al., 2001), open-field (for review see Prut and elevated plus-maze (EPM) became one of the most used animal
Belzung, 2003), shock-probe burying (De Boer and Koolhaas, 2003), models to detect the anxiolytic activity of drugs (Hogg, 1996). The
Geller conflict (Babbini et al., 1982; Howard and Pollard, 2004; test is based on the rodent natural aversion to open spaces

Corresponding author.
E-mail address: souzaalm@ufscar.br (A. Canto-de-Souza).

0006-8993/$ see front matter 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.brainres.2008.06.042
94 BR A IN RE S E A RCH 1 2 31 ( 20 0 8 ) 9 3 1 02

(Fernandes and File, 1996, Treit et al., 1993) and was validated in usually decreases open arm exploration during a second trial.
rats (Pellow et al., 1985) and mice (Lister, 1987; Stephens et al., Interestingly, anxiety-related behaviors recorded during trial 2
1986). The main index of anxiety in the elevated plus-maze is are also insensitive to anxiolytic drugs (File and Zangrossi, 1993;
related to space and time measures of open arms avoidance, Holmes and Rodgers, 1998; Rodgers and Shepherd, 1993).
whereas the locomotor activity is evaluated by the frequency of Although maze experience usually increases open arm avoid-
closed arms entries (File, 1992; Lister, 1987). An increase in the ance during a second trial (Bertoglio and Carobrez, 2000; Dawson
open arm activity (e.g., increase in percentage of open arm entries et al., 1994; Fernandes and File, 1996; Gonzalez and File, 1997;
or percentage of open arm time) indicates reduction of anxiety, Holmes and Rodgers, 1998; Holmes and Rodgers, 1999; Rodgers
whereas changes in the number of entries in closed arms indicate and Shepherd, 1993; Rodgers et al., 1996; Treit et al., 1993),
non-specific effects in the locomotor activity (Menard and Treit, previous findings have also failed to detect changes in anxiety
1999). In the EPM, animals also exhibit a set of behavioral indices in subsequent trials (File et al., 1990; Lister, 1987; Pellow
responses (e.g. stretched attend postures, head dipping) to gather et al., 1985).
information regarding the potential threat of the open arms. This The failure of well-known anxiolytic compounds (e.g.
behavioral strategy has been characterized as risk assessment benzodiazepines) to attenuate anxiety indices in maze-
behavior (Blanchard and Blanchard, 1989; Cruz et al., 1994; experienced animals has been interpreted as one-trial
Rodgers et al., 1997). Importantly, the efficacy of the EPM in tolerance (OTT) (Bertoglio and Carobrez, 2002a; File 1990;
discriminating anxioselective compounds has been increased File et al., 1990; Holmes and Rodgers, 1999; Lister, 1987;
with the adoption of a more ethological analysis (Cruz et al., 1994; Rodgers and Shepherd, 1993). Evidence that OTT also happens
Rodgers and Johnson, 1995; Carobrez and Bertoglio, 2005). with other anxiolytic compounds has been demonstrated (for
An important factor of the EPM is related to the effect on the a review see Carobrez and Bertoglio, 2005). In addition, OTT
previous experience to the test. A single experience in the maze seems to be independent of pharmacological treatment

Fig. 1 Schematic representation of microinfusion sites within the midbrain periaqueductal gray (PAG) of the mice (A). The
number of the points in the figure is less than the total number of mice because of the overlaps. (B) Photomicrograph of midbrain
coronal section from a representative subject showing an injection site into the PAG. Section corresponds to 4.24 mm from
bregma in the atlas of Franklin and Paxinos (1997).
BR A I N R ES E A RC H 1 2 3 1 ( 2 00 8 ) 9 3 1 02 95

animals have received on first trial, the material utilized to 2.1. Experiment 1: Intra-PAG infusion of MDZ (3.0 or
make the maze (File 1993), degree of room illumination as 30 nmol/0.1 l) in maze-nave and maze-experienced mice
well as the diurnal/nocturnal phases of day (Bertoglio and
Carobrez, 2002b). 2.1.1. Intra-PAG infusions of midazolam in maze-nave mice
Benzodiazepine drugs attenuate anxiety when admini- The effects of intra-PAG microinfusions of MDZ (0, 3.0 or
strated systemically (Albrechet-Souza et al., 2005, Audi and 30 nmol) on behavior in maze-nave mice are shown in Fig. 2
Graeff, 1984, Brando et al., 1982; Cole and Rodgers, 1993; Cruz- and Table 1. KruskalWallis one-way ANOVA followed by
Morales et al., 2002; Graeff et al., 1986; Lister, 1987; Motta and Dunn's multiple comparison test revealed that MDZ increased
Brando, 1993; Nunes-de-Souza et al., 2000; Rodgers and percentage open arm entries (H (2,32) = 7.11; P < 0.05) and
Shepherd, 1993; Rosa et al., 2000; Treit, 1991) and centrally percentage open arm time (H (2,32) = 10.56; P < 0.05) without
(Mendes-Gomes and Nunes-de-Souza, 2005; Menard and Treit, altering the locomotor activity [frequency of closed arms
2001; Motta and Brando, 1993; Nunes-de-Souza et al., 2000; entries (H (2,32) = 5,60; P N 0.05] (Fig. 2). Intra-PAG infusion of
Russo et al., 1993; Shah and Treit, 2004; Treit, 1991; Zangrossi midazolam also increased total entries (H (2,32) = 8.16; P < 0.05),
and Graeff, 1994) in rats and mice. The main brain structures and open arm entries (H (2,32) = 9.69; P < 0.05), while decreased
where benzodiazepine compounds have been shown to percentage of protected head-dips (H (2,32) = 6.73; P < 0.05).
attenuate anxiety are: hypothalamus, amygdala and the Furthermore, no significant effects were observed on any
midbrain periaqueductal gray (PAG) (File, 1992; Gonzalez and other behavioral measures recorded (see Table 1 for details).
File, 1997; Graeff et al., 1986; Green and Vale, 1992; Griebel et
al., 1996; Mendes-Gomes and Nunes-de-Souza, 2005; Motta 2.1.2. Intra-PAG infusions of midazolam in maze-experienced
and Brando, 1993; Nunes-de-Souza et al., 2000; Pesold and mice
Treit, 1994, 1995, 1996; Russo et al., 1993; Treit, 1991; Zangrossi Single-factor ANOVAs were initially used to assess between-
and Graeff, 1994). group differences during trial 1 (data not shown). Theses
Regarding the PAG, it has been widely demonstrated that analyses confirmed the reliability of the procedure employed
this midbrain structure belongs to the brain defensive system to randomly allocate subjects to experimental conditions. No
(for a review see Graeff et al., 1993 and Graeff, 1994). However, between-group behavioral differences were found on trial 1
not many studies with intra-PAG infusion of benzodiazepines (F(2,30) = 2.18; P = 0.13).
have investigated its anxiolytic-like effects in EPM-nave The effects of intra-PAG microinfusions of MDZ (0, 3.0 or
animals. Evidence implicating the involvement of the PAG 30 nmol) on behavior in maze-experienced (trial 2) mice are
on OTT has been recently reported by Bertoglio et al. (2005). shown in Fig. 3 and Table 2. KruskalWallis one-way ANOVA
These investigators have demonstrated that the functional did not reveal any significant effect of MDZ on conventional
integrity of the dorsolateral portion of the PAG (dlPAG) is measures of anxiety [% open arm entries (H (2,33) = 0.85; P N 0.05),
crucial to expression of the OTT phenomenon on trial 2, since % open arm time (H (2,32) = 0.15; P N 0.05] or locomotor activity
inactivation of this column with the local anesthetic lidocaine [closed arm entries (H (2,32) = 0.15; P N 0.05] (Fig. 3). Furthermore,
prior to trial 2 restores the anxiolytic effect of systemic
injections of midazolam. However, besides to alter action
potential at passage fibers such kind of lesion does not permit
to distinguish what systems (e.g., serotonergic, GABA/benzo-
diazepine complex, etc.) are modulating a particular emo-
tional state. To date, it is unknown the effects of intra-PAG
infusions of benzodiazepines on anxiety in maze-experienced
animals.
The present study investigated the effects of intra-PAG
infusions of the benzodiazepine midazolam (MDZ) on anxiety
in maze-nave and maze-experienced mice. The potential
anxiolytic effects of MDZ were investigated with local infusion
of the benzodiazepine receptor antagonist flumazenil (FMZ)
(Hunkeler et al., 1981).

2. Results

The histological analysis confirmed that a total of 108 mice


had accurate cannula placements in the PAG (Figs. 1A, B).
Thirty-two subjects were used to investigate the effects of
intra-PAG injections of midazolam on anxiety-like behavior
in maze-nave mice (n = 912), 33 animals were used to
investigate the effects of intra-PAG midazolam on anxiety in Fig. 2 Effects of midazolam (0, 3.0 and 30 nmol)
maze-experienced mice (n = 913) and 43 animals were used microinfusions into the PAG on the behavior of maze-nave
to evaluate the effect of the antagonist combined micro- mice. n = 912. Data are represented as mean S.E.M. *P < 0.05
injections FMZ (16 nmol) with MDZ (30 nmol) (n = 912). versus saline. See Table 1 for complementary data.
96 BR A IN RE S E A RCH 1 2 31 ( 20 0 8 ) 9 3 1 02

Table 1 Effects of MDZ (0, 3.0 or 30 nmol/0.1 l) Table 2 Effects of MDZ (0, 3.0 or 30 nmol/0.1 l)
microinfusions into the PAG on the behavior of maze-nave microinfusions into the PAG on the behavior of maze-
mice experienced mice
Behavior Midazolam (nmol/0.1 l) Behavior Midazolam (nmol/0.1 l)

0 3.0 30 H 0 3.0 30 H
(n = 11) (n = 12) (n = 9) (2,32) (n = 13) (n = 11) (n = 9) (2,33)

Total entries 9.2 1.4 8.5 1.5 16.7 2.2 8.16, P<0.05 Total entries 10.0 1.2 14.3 2.6 13.3 1.4 3.67, NS
Open entries 2.5 0.5 3.6 0.5 6.9 1.3 9,68, P<0.05 Open entries 2.4 0.6 3.5 1.2 2.3 0,8 0.01, NS
% Closed time 55.3 4.9 40.6 5.9 49,0 5.1 2.09, NS % Closed time 84.9 3.1 76.6 5.1 82.0 3.8 1.17, NS
% Centre time 35.6 5.8 30.6 4.4 24.5 6.7 3.18, NS % Centre time 7.4 1.3 16.3 3.8 13.2 3.7 2.52, NS
Total head-dips 9.8 1.1 9.7 1.1 10.0 2.6 0.01, NS Total head-dips 2.2 0.5 6.0 1.9 6.6 2.1 0.79, NS
% Protected head-dips 71.6 11 53.5 8.3 31.9 10 6.74, P<0.05 % Protected head-dips 74.7 12 69.0 14 79.2 11 1.76, NS
Total stretched 4.2 0.7 3.4 0.9 3.7 1.0 0.61, NS Total stretched 2.9 0.6 4.1 1.2 6.0 1.5 2.39, NS
attend postures attend postures
% Protected stretched 100 0 65.0 14 69.0 14 3.95, NS % Protected stretched 76.9 12 72.7 14 87.5 11 4.12, NS
attend postures attend postures
Total rear (frequency) 4.5 1.1 4.8 2.0 7.0 2.1 2.05, NS Total rear (frequency) 6.6 1.7 4.5 1.2 15.1 6.4 3.99, NS
% Protected rear 81.8 12 75.0 13 100 0 2.44, NS % Protected rear 92.3 7.6 72.7 14 88.9 11 1.99, NS
Total immobility 3.4 2.1 10.7 6.1 9.0 4.3 1.19, NS Total immobility 22.9 11 26.5 26 4.2 2.8 2.08, NS
% Protected immobility 45.5 16 41.7 15 55.5 18 0.39, NS % Protected immobility 61.5 14 45.5 16 44.4 18 0.83, NS

Data are represented as mean S.E.M. Data are represented as mean S.E.M.
See also Fig. 2. See also Fig. 3.
P < 0.05 versus saline.

entries (F(1,39) = 4.26, P < 0.05), percentage open arm time


no significant effects were observed on any other behavioral (F(1,39) = 4.82, P < 0.05) and percentage of protected head-dips
measures recorded (see Table 2 for details). (F(1,39) = 6.50, P < 0.05). No other behaviors were significantly

2.2. Experiment 2: Combined intra-PAG microinjections


of FMZ (16 nmol/0.1 l) and MDZ (30 nmol/0.1 l) in
maze-nave mice

The effects of intra-PAG combined microinfusions of FMZ (0


or 16 nmol) and MDZ (0 or 30 nmol), on behavior in maze-
nave mice are shown in Fig. 4 and Table 3. ANOVA revealed
significant main effects of FMZ pretreatment for open arm

Fig. 4 Effects of combined microinjection of vehicle + vehicle


(C + C), vehicle + midazolam 30 nmol (C + MDZ), flumazenil
16 nmol + vehicle (FMZ + C) and flumazenil 16 nmol +
Fig. 3 Effects of midazolam (0, 3.0 and 30 nmol) midazolam 30 nmol (FMZ + MDZ), in PAG of nave mice to
microinfusions into the PAG on the behavior of EPM. n = 912. Data are represented as mean S.E.M. *P < 0.05
maze-experienced mice. n = 913. Data are represented as versus C + C. #P < 0.05 versus C + MDZ. See Table 3 for
mean S.E.M. See Table 2 for complementary data. complementary data.
BR A I N R ES E A RC H 1 2 3 1 ( 2 00 8 ) 9 3 1 02 97

Table 3 Effects of combined microinfusion of vehicle + was evident for open arm entries, percentage open arm entries,
vehicle (C + C), vehicle + midazolam 30 nmol (C + MDZ), percentage open arm time, percent protected head-dips, total
flumazenil 16 nmol + vehicle (FMZ + C) and flumazenil stretched attend postures and percentage protected stretched
16 nmol + midazolam 30 nmol (FMZ + MDZ) into the PAG
attend postures.
on the behavior of maze-nave mice
Behavior C+C C + MDZ FMZ + C FMZ + MDZ
(n = 12) (n = 9) (n = 12) (n = 10)
3. Discussion
Total entries 7.1 1.1 9.8 1.7 5.3 0.8 7.9 1.7
Open entries 1.4 0.4 5.3 1.2 1.3 0.4 # 2.5 0.9 # The present study demonstrated that intra-PAG infusions of
% Closed time 49.9 9.1 36.8 8.5 59.4 10.8 57.6 9.9 midazolam attenuate selectively anxiety-like behaviors in
% Centre time 40.4 8.6 22.6 7.8 35.3 9.8 27.5 6.3
maze-nave mice, corroborating previous findings with mice
Total head-dips 9.0 1.8 7.2 1.9 7.2 1.5 5.7 1.2
(Mendes-Gomes and Nunes-de-Souza, 2005) and rats (Albre-
% Protected 72.2 10.3 30.0 9.4 83.3 11.2 # 74.7 11.7 #
head-dips chet-Souza et al., 2005; Bertoglio and Carobrez, 2002c; Bertoglio
Total stretched 4.3 0.9 0.6 0.3 3.8 0.9 3.6 1.5 # et al., 2005; Cruz-Morales et al., 2002; Motta and Brando, 1993;
attend postures Russo et al., 1993). Importantly, the anxiolysis produced by
% Protected 95.0 4.2 33.3 16.7 82.6 11.2 # 70.0 15.3 # midazolam was selective and completely blocked by prior
stretched injection into the same midbrain site of the benzodiazepine
attend postures
receptor antagonist flumazenil (Hunkeler et al., 1981). How-
Total rear 3.7 1.1 7.2 2.9 4.9 1.6 5.6 1.9
(frequency)
ever, in maze-experienced mice, midazolam failed to produce
% Protected rear 75.0 13.1 75.1 14.3 91.3 8.3 96.7 3.3 any significant effect on anxiety in the EPM, suggesting that
Total immobility 15.3 6.1 29.4 16.9 12.2 4.1 20.5 9.0 GABA/benzodiazepine receptor complex located within the
% Protected 50.0 15.1 44.4 17.6 83.3 11.2 50.0 16.7 PAG is involved in the OTT.
immobility The behavioral effect observed with intra-PAG injections of
Data are represented as mean S.E.M. See also Fig. 4. midazolam in maze-nave mice was selective on anxiety
P < 0.05 versus C + C. indices, since no changes on frequency of closed arm entries, a
# P < 0.05 versus C + MDZ. measure used to evaluate locomotor activity in the EPM (File,
1992; Lister, 1987), were obtained with this benzodiazepine
compound at both doses. At highest dose midazolam also
altered by FMZ per se (F(1,39) 3.25, P N 0.05). Post hoc increased total arm entries, an effect that could be related to
comparisons revealed that intra-PAG FMZ decreased open an increase of locomotor activity. However, 30 nmol mid-
arm entries and percentage open arm time and increased azolam also increased frequency open arm entries, indicating
protected head-dips. It is interesting to note that these the effect on total entries was likely due to a higher open arm
behavioral changes were evident only when FMZ pretreat- exploration. These results corroborate previous studies
ment (i.e., FMZ + C; FMZ + MDZ) was compared to MDZ demonstrating that midazolam attenuates anxiety indices
treated group (i.e., C + MDZ). when injected systemically (Albrechet-Souza et al., 2005; Audi
Significant main effects of MDZ treatment were found and Graeff, 1984; Bertoglio and Carobrez, 2002c; Brando et al.,
for open arm entries (F(1,39) = 12.66, P < 0.05), percentage open 1982; Cole and Rodgers 1993; Cruz-Morales et al., 2002; Graeff
arm entries (F(1,39) = 5.48, P < 0.05), percentage open arm time et al., 1986; Lister, 1987; Motta and Brando, 1993; Nunes-de-
(F(1,39) = 8.62, P < 0.05), total stretched attend postures (F(1,39) = 3.76, Souza et al., 2000; Rodgers and Shepherd, 1993; Rosa et al.,
P < 0.05), percentage protected stretched attend postures 2000; Treit, 1991) or intra-PAG in maze-nave rats (Motta and
(F(1,39) = 9.52, P < 0.05) and percentage head-dips (F(1,39) = 5.39, Brando, 1993; Russo et al., 1993) and mice (Mendes-Gomes
P < 0.05), while significant pretreatment treatment interactions and Nunes-de-Souza, 2005). Besides the conventional mea-
were obtained for percentage open arm entries (F(1,39) = 7.12, sures, ethological measures such as stretched attend postures
P < 0.05) and percentage protected stretched attend postures and head dipping have been employed to evaluate risk
(F(1,39) = 4.16, P < 0.05). Post hoc comparisons showed that, relative assessment in the EPM (Rodgers and Johnson, 1995; Griebel
to the C + C group, intra-PAG C + MDZ (30 nmol) significantly et al., 1997). It has been demonstrated that risk assessment
increased open arm entries (P < 0.01), percentage open arm behaviors are sensitive to benzodiazepine anxiolytics (Cole
entries (P < 0.01) and percentage open arm time (P < 0.01), while and Rodgers, 1993; Handley, 1991). In the present study, intra-
decreasing percent protected head-dips (P < 0.01), total stretched PAG infusions of midazolam reduced percentage of protected
attend postures (P < 0.05) and percentage protected stretched head-dips, corroborating previous studies in that midazolam
attend postures (P < 0.01). Importantly, these anxiolytic-like decreased the risk assessment parameter in the EPM when
effects of MDZ were observed in the absence of significant injected systemically (Bertoglio and Carobrez, 2002c), but in
changes to total arm entries, closed arm entries, percentage another study administration of midazolam produced
closed time, percentage centre time, total head-dips, total rear, increased the risk assessment (Cruz-Morales et al., 2002).
percentage protected rear, total immobility and percentage In the present work, the reduction of both conventional
protected immobility (F(1,39) = 3.91, P 0.05). Intra-PAG pretreat- and ethological measures of anxiety following intra-PAG
ment with FMZ completely blocked the effects of intra-PAG injection of midazolam was completely blocked by prior
MDZ treatment on anxiety-like behaviors. Complete anta- infusion with flumazenil in the same midbrain site in maze-
gonism (i.e., no significant difference FMZ + MDZ vs. C + C nave animals. Importantly, flumazenil did not affect any
coupled with a significant difference FMZ+ MDZ vs. C + MDZ) behavioral measure when injected alone (FMZ + C), suggesting
98 BR A IN RE S E A RCH 1 2 31 ( 20 0 8 ) 9 3 1 02

this benzodiazepine receptor antagonist is devoid of intrinsic and File, 1997) and PAG (Bertoglio et al., 2005; Mendes-Gomes
effects on anxiety in EPM. Although this failure of effect on and Nunes-de-Souza, 2005; Russo et al., 1993), have been
anxiety was not an expected result, similar evidence has been related to the phenomenon of OTT to anxiolytic drugs. The
previously reported. For instance, previous studies have results of the present study suggest that GABA/benzodiaze-
demonstrated that systemic administration of flumazenil pine receptor complex located within the PAG is involved in
produces anxiogenic-like effect (Lee and Rodgers, 1991; Savic the OTT.
et al., 2004) or does not alter anxiety (Pellow and File, 1986) In summary, the results of the present study confirm that
evaluated in the EPM. However, contrary to the present the midbrain PAG plays a key role in the modulation of anxiety
results, it has been emphasized that GABA exerts a tonic in maze-nave mice. A five-minute prior experience to the
action on defensive reactions within the PAG (Milani and maze renders the defensive responses insensitive to intra-
Graeff, 1987). Thus it would be expected an anxiogenic effect of PAG infusion of midazolam in mice, a phenomenon known as
flumazenil into the PAG. Contradictory results have also been OTT. The lack of effects of the midazolam on anxiety in maze-
reported following benzodiazepine antagonist plus agonist- experienced mice suggests that the GABA/benzodiazepine
combined injections in behavioral responses in anxiety tests. receptors located within the PAG somehow lose its action in
For instance, while systemic administration of flumazenil attenuate defensive behaviors in the EPM. Whether OTT might
antagonized the anxiolytic effect of diazepam (p.o.) (Wada and be due an alteration of the state of the ligand site and/or of the
Fukuda, 1991) and chlordiazepoxide (i.p.) (Ferris et al., 2001), it involved receptor complex (Bertoglio and Carobrez, 2002a,b,
was not capable to antagonize anxiolysis produced by mid- 2003; Gonzalez and File, 1997) within the midbrain PAG is a
azolam when both drugs were infused into the dorsal raphe hypothesis to be investigated.
nucleus in rats (Gonzalez and File, 1997). On the other hand
and in agreement with present results, intra-PAG injection of
flumazenil blocked the anxiolytic effects of local infusion of
4. Experimental procedures
midazolam in rats (Russo et al., 1993). A punctual inconsistent
result of intra-PAG infusion of midazolam in the measure 4.1. Subjects
stretched attend postures was observed between Exp.1 and
Exp. 2. While combined infusion of C + MDZ decreased both Seven-to-nine-week-old male Swiss mice (Federal University
total stretched attend postures and percentage protected of So Carlos-UFSCar), weighing 2327 g, were used. The
stretched attend postures (Exp. 2), single injection of mid- animals were housed in groups of 10 per cage (41 34 16 cm),
azolam did not produce a significant effect on this risk and maintained under controlled temperature (24 1 C), and
assessment behavior (Exp. 1). Although the reason for these humidity (55 5%), light (lightdark cycle of 12/12 h, lights on at
discrepant results is unclear, the complete blockade provoked 7:00 am and lights off at 7:00 pm) and free access to food and
by intra-PAG flumazenil on midazolam effects in stretched water, except during brief sessions of the test. The experi-
attend postures suggests this ethological measure of anxiety is mental sessions were carried out during the light phase of the
sensitive to benzodiazepine receptors activation within the cycle (09:0016:00 h).
mouse PAG.
In contrast to maze-nave mice, intra-PAG injections of 4.2. Drugs
midazolam (3.0 nmol and 30 nmol) did not attenuate anxiety
in maze-experienced mice. It has been demonstrated that Midazolam (3.0 nmol/0.1 l and 30 nmol/0.1 l; Roche, Brazil)
previous experience to EPM increases the avoidance of the and flumazenil (16 nmol/0.1 l; Sigma) were dissolved in 0.9%
open arms (Bertoglio and Carobrez, 2000; Fernandes and File, saline and in Tween-80 (1%), respectively. These doses were
1996; Holmes and Rodgers, 1998; Holmes and Rodgers, 1999), based in previous studies (Gonzalez and File, 1997; Nunes-de-
alters the natural behavior (File and Zangrossi, 1993; Holmes Souza et al., 2000; Russo et al., 1993).
and Rodgers, 1998) and impairs the anxiolytic effect of BDZs
in the subsequent reexposure to the maze (Bertoglio and 4.3. Surgery and microinjection
Carobrez, 2002c; File 1990; Holmes and Rodgers, 1999). Some
hypotheses have been suggested to explain these changes, Mice were implanted with 7-mm stainless steel guide cannula
including sensitization of the fear to the open arms (Rodgers (26-gauge; Insight Scientifics Equipment Ltda, Brazil) under
and Shepherd, 1993), qualitative change in the emotional sodium thiopental (90 mg/kg, i.p.) anesthesia. Guide cannula
state in the subsequent exposure to the maze (Holmes and was fixed to the skull using dental acrylic and jeweler's screws.
Rodgers, 1998), alteration of the state of the ligand site and/or Stereotaxic coordinates (Franklin and Paxinos, 1997) for the PAG
of the involved receptor complex (Bertoglio and Carobrez, were 4.1 mm posterior to bregma, 1.3 mm lateral to the midline,
2002a,b, 2003; Gonzalez and File, 1997) and locomotor and 1.2 mm ventral to the skull surface, with the guide cannula
habituation (Dawson et al., 1994). In the present study, angled 26 to the vertical. A dummy cannula (33-gauge stainless
midazolam failed to change locomotor activity (represented steel wire; Fishtex Industry and Commerce of plastics Ltda),
by the frequency of closed arms entries) during the reexpo- inserted into each guide cannula immediately after surgery,
sure (trial 2) to EPM, suggesting that the OTT phenomenon is served to reduce the incidence of occlusion. Before tests mice
not related to a locomotor habituation (Dawson et al., 1994). were allowed four to five days to recover from surgery.
Cerebral structures related to the defense system, including Solutions were injected into the PAG by microinjection unit
hippocampus (Nunes-de-Souza et al., 2002), median and (33-gauge stainless steel cannula Insight Scientifics Equipment
dorsal raphe nuclei (Canto-de-Souza et al., 2002; Gonzalez Ltda, Brazil), which extended 1.0 mm beyond the tips of the
BR A I N R ES E A RC H 1 2 3 1 ( 2 00 8 ) 9 3 1 02 99

guide cannula. The microinjection unit was connected to a conventional measures (Lister, 1987; Rodgers and Johnson,
10 l Hamilton microsyringe via polyethylene tubing (PE-10), 1995) as well the ethological measures (Cruz et al., 1994;
and administration was controlled by an infusion pump Rodgers and Johnson, 1995).
(Insight BI 2000 Scientific Equipment Ltd, Brazil) pro-
grammed to deliver a volume 0.1 l of each solution over a 4.5.1. Conventional measures
period of 60 s. The microinjection procedure consisted of The frequencies of total, open and closed entries (arm entry is
gently restraining the mice, inserting the injection unit, defined by all four paws into an arm), percentage open entries
infusing the solution for 60 s and keeping the injection unit [(open / total) 100] and percentage open, closed and central
for 90 s. The movement of a small air bubble in the PE-10 parts of the maze [(time / 300) 100] were also recorded.
tubing, during and after the microinjection confirmed the
4.5.2. Ethological measures
delivery of the solution.
Frequency scores for rearing (vertical movement against the
4.4. Apparatus and general procedure walls); head dipping (HDips: exploratory movement of head/
shoulders over the side of the maze) and stretched attend
The elevated plus-maze (EPM) comprised two open arms postures (SAP: exploratory posture in which the body stretched
(30 5 0.25 cm) and two enclosed arms (30 5 15 cm) that forward then retracted to the original position without any
extended from a common central platform (5 5 cm). The forward locomotion); as well as total duration scores in
apparatus was constructed from wood (floor) and transparent seconds for rearing and immobility. In view of importance of
glass (clear walls) and raised to a height of 38.5 cm above floor thigmotactic cues to patterns of EPM exploration, HDips and
SAP were further differentiated as a function of whereabouts
level. All testing was conducted under moderate illumination
on the maze they were displayed. Consistent with earlier
(77 lx; measured on the central platform of the EPM) during the
reports, the closed arms and the central platform were
light phase of the LD cycle. The EPM design was similar to that
originally described by Lister (Lister, 1987). together designated protected areas, while the open arms
were designated unprotected areas. Data for the above
4.4.1. Experiment 1: Intra-PAG infusion of MDZ (3.0 or measures are presented both as behavior totals (e.g. HDips)
30 nmol/0.1 l) in maze-nave and maze-experienced animals and as percent protective scores [e.g. %PHDips; (protected
On test day, animals were transported to the experimental HDips / total HDips) 100].
room with controlled temperature (24 1 C) and left undis-
turbed for at least 30 min prior to testing. After the intra-PAG 4.6. Histology
injection of saline or midazolam (3.0 nmol or 30 nmol), each
mouse was placed in an individual holding cage (33 15 3 cm) At the end of experiments all the animals received an infusion
for 1 min and then transported to the maze. Testing of 0.1 l solution of 1% of methylene blue in the PAG, according
commenced by placing each mouse on the central platform to the procedure described above. The animals were then
of the maze facing the left open arm. Test sessions were 5 min killed by the anesthetics overdose, and their brains were
in duration and, between subjects, the maze was thoroughly removed and accommodated in recipients containing forma-
cleaned with ethanol 20% and dry with towel paper. All lin solution (10%) for afterwards coronally sectioned by
sessions were video recorded by a camera (Panasonic-X12) microtome (criostato ANCAP 300). Injection sites were micro-
(positioned above and at ~45 to the maze) linked to a monitor scopically (Olympus B202) verified with reference to the Atlas
and VCR in the adjacent laboratory. A similar procedure was of Franklin and Paxinos (Franklin and Paxinos, 1997). Data
carried out for animals in the retest condition (i.e. prior maze from animals with injection sites outside the PAG were
experience) with the exception that, 24 h earlier, each animal excluded from analysis.
had been pre-exposed undrugged to the maze for 5 min (trial
4.7. Statistics
1). On the test day, they were infused with saline or
midazolam (3.0 nmol or 30 nmol) followed by a second 5-min
All results were initially submitted to Levene's test for
exposure to the maze (trial 2).
homogeneity of variance. When appropriate, the data were
4.4.2. Experiment 2: Combined intra-PAG microinjection of FMZ transformed to log, square root or cube root before being
(16 nmol/0.1 l) and MDZ (30 nmol/0.1 l) in maze-nave animals submitted to one-way analyses of variance (ANOVA) or two-
Four groups of animals received intracerebral infusion of way independent ANOVA [factor 1: pretreatment (flumazenil:
drugs in the following combination: vehicle + vehicle (control: 0 or 16 nmol); factor 2: treatment (midazolam: 0 or 30 nmol)].
C + C), MDZ 0 or 30 nmol (C + MDZ), FMZ 0 or 16 nmol (FMZ + C) Where indicated by significant F values, Duncan's multiple
and FMZ 16 nmol + MDZ 30 nmol (FMZ + MDZ). After the second comparisons tests were used. Where Levene's test remained
microinjection procedure each mouse was placed in the EPM significant even after data transformation non-parametric
as described in Experiment 1. Interval between intra-PAG drug tests (KruskalWallis test followed by Dunn's multiple com-
infusions was 10 min. parison test) were used. In all cases, a P value 0.05 was
required for significance.
4.5. Behavioral analysis
4.8. Ethics
Videotapes were scored by a highly trained observer using the
ethological analysis program X-Plo-Rat, version 3.0 (Becerra- The experiments reported in this study were performed in
Garcia et al., 2005). The analyzed behaviors involved the compliance with the recommendations of the Brazilian
100 BR A IN RE S E A RCH 1 2 31 ( 20 0 8 ) 9 3 1 02

Society of Neuroscience and Behavior (SBNeC) based on the US Canto-de-Souza, A., Nunes-de-Souza, R.L., Rodgers, R.J., 2002.
National Institutes of Health Guide for Care and Use of Anxiolytic-like effect of WAY-100635 microinfusions into the
median (but not dorsal) raphe nucleus in mice exposed to the
Laboratory Animals.
plus-maze: influence of prior test experience. Brain Res. 928, 5059.
Carobrez, A.P., Bertoglio, L.J., 2005. Ethological and temporal
analyses of anxiety-like behavior: the elevated plus-maze
Acknowledgments model 20 years on. Neurosci. biobehav. rev. 29, 11931205.
Cole, J.C., Rodgers, R.J., 1993. An ethological analysis of the effects
of chlordiazepoxide and bretazenil (Ro 16-6028) in the murine
The authors thank Professor R.L. Nunes-de-Souza (FCFar-
elevated plus maze. Behav. Pharmacol. 4, 573580.
UNESP, Brazil) for comments and suggestions. The authors
Cruz, A.P.M., Frei, F., Graeff, F.G., 1994. Ethopharmacological
thank Professor M.L. Brando (FFCLRP-USP, Brazil) for supply- analysis of rat behavior on the elevated plus-maze.
ing the midazolam. This study was supported by FAPESP (Proc. Pharmacol. Biochem. Behav. 49, 171176.
03/00086-0). L.M. Reis was a recipient of CAPES scholarship. Cruz-Morales, S.E., Santos, N.R., Brando, M.L., 2002. One-trial
tolerance to midazolam is due to enhancement of fear and
reduction of anxiolytic-sensitive behaviors in the elevated
REFERENCES plus-maze retest in the rat. Pharmacol. Biochem. Behav.r 72,
973978.
Dawson, G.R., Crawford, S.P., Stanhope, K.J., Iversen, S.D.,
Agmo, A., Galvan, A., Heredia, A., Morales, M., 1995. Naloxone Tricklebank, M.D., 1994. One-trial tolerance to the effects of
blocks the antianxiety but not the effects of benzodiazepines chlordiazepoxide on the elevated plus-maze may be due to
and pentobarbital experimental studies and literature. locomotor habituation, not repeated drug exposure.
Psychopharmacology 120 (2), 186194. Psycopharmacology 113, 570572.
Albrechet-Souza, L., Oliveira, A.R., De-Luca, M.C.Z., Tomazini, F.M., De Boer, S.F., Koolhaas, J.M., 2003. Defensive burying in
Santos, N.R., Brando, M.L., 2005. A comparative study with rodents: ethology, neurobiology and psychopharmacology
two types of elevated plus-maze (transparent vs, opaque walls) 463, 145161.
on the anxiolytic effects of midazolam, one-trial tolerance and Fernandes, C., File, S.E., 1996. The influence of open arm ledges
fear-induced analgesia. Prog. in Neuro-Psychopharmacol. Biol. and maze experience in the elevated plus-maze.
Psychiatry 29, 571579. Pharmacol. Biochem. Behav. 54, 3140.
Audi, E.A., Graeff, F.G., 1984. Benzodiazepine receptors in the Ferris, P., Seward, E., Dawson, G.R., 2001. Interactions between
periaqueductal grey mediate anti-aversive drug action. Eur. J. LY354740, a group II metabotropic agonist and the
Pharmacol. 103, 279285. GABA(A)benzodiazepine receptor complex in the rat elevated
Babbini, M., Gaiardi, M., Bartoletti, M., 1982. Benzodiazepine effects plus-maze. J. Psychopharmacol. 15, 7682.
upon GellerSeifter conflict test in rats: analysis of individual File, S.E., 1990. One-trial tolerance to the anxiolytic effects of
variability. Pharmacol. Biochem. Behav. 17 (1), 4348. benzodiazepines in the plus-maze. Psychopharmacology 100,
Becerra-Garcia, A.M., Cardenas, F.R., Morato, S., 2005. Effect of 281282.
different illumination levels on rat behavior in the elevated File, S.E., 1992. Behavioural detection of anxiolytic action. In:
plus-maze. Physiol. Behav. 85, 265270. Elliott, J.M., Heal, D.J., Marsden, C.A. (Eds.), Experimental
Bertoglio, L.J., Carobrez, A.P., 2000. Previous maze experience Approaches to Anxiety and Depression. J. Wiley, Chichester,
required to increase open arms avoidance in rats submitted to pp. 2544.
the elevated plus-maze model of anxiety. Behav. Brain Res. File, S.E., 1993. The interplay of learning and anxiety in the
108, 197203. elevated plus-maze. Behav. Brain Res. 58, 199202.
Bertoglio, L.J., Carobrez, A.P., 2002a. Anxiolytic effects of ethanol File, S.E., Seth, P., 2003. A review of 25 years of the interaction test.
and phenobarbital are abolished in test-experienced rats Eur. J. Pharmacol. 463, 3553.
submitted to the elevated plus maze. Pharmacol., Biochem. File, S.E., Zangrossi, H.J., 1993. One-trial tolerance to the anxiolytic
Behav. 73, 963969. actions of the benzodiazepines in the elevated plus-maze, or
Bertoglio, L.J., Carobrez, A.P., 2002b. Behavioral profile of rats the development of a phobic state? Psychopharmacology 110,
submitted to session 1-session 2 in the elevated plus-maze 240244.
during diurnal/nocturnal phases and under different File, S.E., Mabutt, P.S., Hitchott, P.K., 1990. Characterization of
illumination conditions. Behav. Brain Res. 132, 135143. phenomenon of one-trial tolerance to the anxiolytic effect
Bertoglio, L.J., Carobrez, A.P., 2002c. Prior maze experience of chlordiazepoxide in the elevated plus-maze
required to alter midazolam effects in rats submitted Psychopharmacology 102, 98101.
to the elevated plus-maze. Pharmacol. Biochem. Behav. 72, Franklin, K.B., Paxinos, G., 1997. The Mouse Brain in Stereotaxic
449455. Coordinates. Ed. USA, Academic Press. Califrnia.
Bertoglio, L.J., Carobrez, A.P., 2003. Anxiolytic-like effects of Gonzalez, L.E., File, S.E., 1997. A five-minute experience
NMDA/glicine-B receptor ligands are abolished during the in the elevated plus-maze alters the state of the benzodiazepine
elevated plus-maze Trial 2 in rats. Psychopharmacology 170, receptor in the dorsal raphe nucleus.
335342. J. Neurosci. 17, 505511.
Bertoglio, L.J., Anzini, C., Lino-de-Oliveira, C., Carobrez, A.P., 2005. Graeff, F.G., 1994. Neuroanatomy and neurotransmitter regulation
Enhanced dorsolateral periaqueductal gray activity of defensive behaviors and related emotions in mammals.
counteracts the anxiolytic response to midazolam on the elevated Braz. J. Med. and Biol. Res. 27 (4), 811829.
plus-maze Trial 2 in rats. Behav. Brain Res. 162, 99107. Graeff, F.G., Brando, M.L., Audi, E.A., Schutz, M.T.B., 1986.
Blanchard, R.J., Blanchard, D.C., 1989. Attack and defense in Modulation of the brain aversive system by GABAergic
rodents as ethoexperimental models for the study of emotion. and serotonergic mechanisms. Behav. Brain Res. 21,
Prog. in Neuro-psychopharmacol. Biol. Psychiatry 13 (suppl.), 6572.
S3S14. Graeff, F.G., Silveira, M.C.L., Nogueira, R.L., Audi, E.A., Oliveira,
Brando, M.L., Aguiar, J.C., Graeff, F.G., 1982. GABA mediation of R.M.W., 1993. Role of the amygdala and periaqueductal
antiaversive action of minor tranquilizers. Pharmacol., gray in anxiety and panic. Behav. Brain Res. 58 (12),
Biochem. Behav. 16, 397402. 123131.
BR A I N R ES E A RC H 1 2 3 1 ( 2 00 8 ) 9 3 1 02 101

Green, S., Vale, A., 1992. Role of amygdaloid nuclei in the antinociception in mice: effects of systemic and
anxiolytic effects of benzodiazepines in rats. Behav Pharmacol intra-amygdala administration of 8-OH-DPAT and midazolam.
3, 261264. Psychopharmacology 150 (3), 300310.
Griebel, G., Sanger, D.J., Perrault, G., 1996. The use of the rat plus- Nunes-de-Souza, R.L., Canto-de-Souza, A., Rodgers, R.J., 2002.
maze to discriminate non-selective and BZ-1 (w1) selective, Effects of intra-hippocampal infusion of WAY-1000635 on
benzodiazepine receptor ligands. Psychopharmacology 124, plus-maze behavior in mice. Influence of site of injection and
245254. prior test experience. Brain Res. 927, 8796.
Griebel, G., Rodgers, R.J., Perrault, G., Sanger, D.J., 1997. Risk Pellow, S., File, S.E., 1986. Anxiolytic and anxiogenic drug effects
assessment behaviour: evaluation of utility in the study of on exploratory activity in an elevated plus-maze: a novel
5-HT-related drugs in the elevated plus-maze. Pharmacol. test of anxiety in the rat. Pharmacol. Biochem. Behav. 24,
Biochem. Behav. 57, 817827. 525529.
Handley, S.L., 1991. Serotonin in animal models of anxiety: the Pellow, S., Chopin, P., File, S.E., Briley, M., 1985. Validation of open:
importance of stimulus and response. In: Idzikowski, C., closed arm entries in an elevated plus-maze as a measure of
Cowen, P.J. (Eds.), Serotonin, Sleep and Mental Disorder. anxiety in the rat. J. Neurosci. Methods 14, 149167.
Wrightson, London, pp. 89115. Pesold, C., Treit, D., 1994. The septum and amygdala differentially
Handley, S.L., Mithani, S., 1984. Effects of alpha-adrenoceptor mediate the anxiolytic effects of benzodiazepines. Brain Res.
agonists and antagonists in a maze exploration model of 638, 295301.
fear-motivated behaviour. Naunyn-Schmiedeberg's Arch. Pesold, C., Treit, D., 1995. The central and basolateral amygdala
Pharmacol. 327, 15. differentially mediate the anxiolytic effects of
Hascot, M., Bourin, M., Dhonnchadha, B.A.N., 2001. The mouse benzodiazepines. Brain Res. 671, 213221.
lightdark paradigm: a review. Prog. Neuro-Psychophamacol. Pesold, C., Treit, D., 1996. The neuroanatomical specificity of the
Biol. Psychiatry 25, 141166. anxiolytic effects of intra-septal infusions of midazolam. Brain
Hogg, S., 1996. A review of the validity and reliability of the Res. 710, 161168.
elevated plus-maze as a model of anxiety. Pharmacol. Pietraszek, M., Sukhanov, I., Maciejak, P., Szyndler, J., Gravius, A.,
Biochem. Behav. 54, 2130. Wislowska, A., Plaznik, A., Bespalov, A., Danysz, W., 2005.
Holmes, A., Rodgers, R.J., 1998. Responses of SwissWebster mice Anxiolytic-like effects of mGlu1 and mGlu5 receptor
to repeated plus-maze experience: further evidence for antagonists in rats. Eur. J. Pharmacol. 514, 2534.
qualitative shift in emotional state? Pharmacol. Biochem. Prut, L., Belzung, C., 2003. The open field as a paradigm to measure
Behav. 60, 473488. the effects of drugs on anxiety-like behaviors: a review. Eur. J.
Holmes, A., Rodgers, R.J., 1999. Influence of spatial and temporal Pharmacol. 463, 333.
manipulations on the anxiolytic efficacy of chlordiazepoxide in Rodgers, R.J., Johnson, N.J.T., 1995. Factor-analysis of spatiotemporal
mice previously exposed to the elevated plus maze. Neurosci. and ethological measures in the murine elevated plus-maze test of
Biobehav. Rev. 23, 971980. anxiety. Pharmacol. Biochem. Behav. 2, 297303.
Howard, J.L., Pollard, G.T., 2004. Effects of buspirone in the Rodgers, R.J., Shepherd, J.K., 1993. Influence of prior maze
GellerSeifter conflict test with incremental shock. Drug Dev. experience on behavior and response to diazepam in the
Res. 19 (1), 3749. elevated plus-maze and light/dark tests of anxiety in mice.
Hunkeler, W., Mohler, H., Pieri, L., Polc, P., Bonetti, E.P., Cumin, R., Psychopharmacology 113, 237242.
Schaffner, R., Haefely, W., 1981. Selective antagonists of Rodgers, R.J., Johnson, N.J.T., Cole, J.C., Dewar, C.V., Kidd, G.R.,
benzodiazepines. Nature 290, 514516. Kimpson, P.H., 1996. Plus-maze retest profile: importance of initial
Kennett, G.A., Bailey, F., Piper, D.C., Blackburn, T.P., 1995. Effects of SB stages of trial 1 and response to post-trial cholinergic receptor
2006464, a 5-HT2C 5-HT2B receptor antagonist, in 2 conflict blockade in mice. Pharmacol. Biochem. Behav. 54, 4150.
models of anxiety. Psychopharmacology 118 (2), 178182. Rodgers, R.J., Cutler, M.G., Jackson, J.E., 1997. Behavioural effects in
Kennett, G.A., Trail, B., Bright, F., 1998. Anxiolytic-like actions of mice of subchronic chlordiazepoxide, maprotiline and
BW 723C86 in the rat Vogel conflict test are 5-HT2B receptor fluvoxamine. II. The elevated plus-maze. Pharmacol. Biochem.
mediated. Neuropharmacology 37 (12), 16031610. Behav. 57, 127136.
Lee, C., Rodgers, R.J., 1991. Effects of benzodiazepine receptor Rosa, V.P., Vandressen, N., Calixto, A.V., Kovaleski, D.F., Faria, M.S.,
antagonist, flumazenil, on antinociceptive and behavioural 2000. Temporal analysis of the rat's behavior in the
responses to the elevated plus-maze in mice. plus-maze: effect of midazolam. Pharmacol. Biochem.
Neuropharmacology 30, 12631267. Behav. 67, 177182.
Lister, R.G., 1987. The use of a plus-maze to measure anxiety in the Russo, A.S., Guimares, F.S., De Aguiar, J.C., Graeff, F.G., 1993. Role
mouse. Psychopharmacology 92, 180185. of benzodiazepine receptors located in the dorsal
Menard, J., Treit, D., 1999. Effects of centrally administered periaqueductal gray of rats in anxiety. Psychopharmacology
anxiolytic compounds in animal models of anxiety. Neurosci. 110, 1982002.
Biobehav. Rev. 23, 591613. Savic, M.M., Obradovic, D.I., Ugresic, N.D., Cook, J.M., Yin, W.,
Mendes-Gomes, J., Nunes-de-Souza, R.L., 2005. Concurrent Bokonjic, D.R., 2004. Bidirectional effects of benzodiazepine
nociceptive stimulation impairs the anxiolytic effect of binding site ligands in the elevated plus-maze: differential
midazolam injected into the periaqueductal gray in mice. Brain antagonism by flumazenil and -CCt. Pharmacol. Biochem.
Res. 1047, 97104. Behav. 79, 279290.
Milani, H., Graeff, F.G., 1987. GABAbenzodiazepine modulation of Shah, A.A., Treit, D., 2004. Infusions of midazolam into the medial
aversion in the medial hypothalamus of the rat. Pharmacol. prefrontal cortex produce anxiolytic effects in the elevated
Biochem. and Behav. 28, 2127. plus-maze and shock-probe burying tests. Brain Res. 996,
Millan, M.J., 2003. The neurobiology and control of anxious states. 3140.
Prog.Neurobiol. 70 (2), 83244. Smith, R.L., Barret, R.J., 1997. Tolerance to the anticonflict effects of
Motta, V., Brando, M.L., 1993. Aversive and antiaversive effects of diazepam: importance of methodological considerations.
morphine in the dorsal periaqueductal gray of rats submitted Pharmacol. Biochem. Behav. 58 (1), 6166.
to the elevated plus-maze test. Pharmacol. Biochem. Behav. 44, Stephens, D.N., Meldrum, B.S., Weidmann, R., Schneider, C.,
119125. Grutzner, M., 1986. Does the excitatory amino acid
Nunes-de-Souza, R.L., Canto-de-Souza, A., Da-Costa, M., Fornari, receptor antagonist 2-APH exhibit anxiolytic activity?
R.V., Graeff, F.G., Pel, I.R., 2000. Anxiety-induced Psychopharmacology 90, 166169.
102 BR A IN RE S E A RCH 1 2 31 ( 20 0 8 ) 9 3 1 02

Treit, D., 1991. Anxiolytic effects of benzodiazepines and Wada, T., Fukuda, N., 1991. Effects of DN-2327, a new anxiolytic,
5-HT1A agonists: a review. In: Rodgers, R.J., Cooper, S.J. diazepam and buspirone on exploratory activity of the rat in an
(Eds.), 5-HT1A Agonists, 5-HT3 Agonists and elevated plus-maze. Psychopharmacology 104, 444450.
Benzodiazepines: Their Comparative Behavioural Zangrossi Jr., H., Graeff, F.G., 1994. Behavioural effects of
Pharmacology. J. Wiley, Chichester, pp. 107131. intra-amygdala injections of GABA and 5-HT acting drugs
Treit, D., Menard, J., Royan, C., 1993. Anxiogenic stimuli in the in the elevated plus-maze. Braz. J. Med. Biol Res. 27,
elevated plus-maze. Pharmacol. Biochem. Behav. 44, 463469. 24532456.