Introduction

Even though it is now well established that selective antidepressants (ADs) possess anxiolytic properties, their underlying neuropharmacological mechanism of action is still not understood (Zohar and Westenberg 2000; Bourin and Lambert 2002; Bourin et al. 2002; Nemeroff 2003; Vaswani et al. 2003). Owing to their 5-HT re-uptake inhibiting effects, SSRIs and SNRIs increase synaptic levels of 5-HT (Fuller 1994; Kreiss and Lucki 1995; Beyer et al. 2002; Lambert and Bourin 2002; Ables and Baughman 2003; Felton et al. 2003), leading to an increased activation of a multitude of specific postsynaptic serotonin (5-HT) receptors (e.g. 5-HT1A, 5-HT1D, 5-HT2A, 5-HT2C and 5-HT3 receptors). However, it is not yet known which 5-HT receptor subtypes mediate the therapeutic effects of antidepressants and whether different 5-HT receptor subtypes are involved in their anxiolytic actions.

The following study was carried out to explore the anxiolytic-like profile of two different classes of antidepressants; paroxetine, a selective serotonin reuptake inhibitor (SSRI), and venlafaxine, a serotonin and noradrenaline reuptake inhibitor (SNRI), in the mouse model of anxiety; the four plates test (FPT). This model is based on the suppression by punishment of a simple innate ongoing behaviour, i.e. the exploration of novel surroundings, of the mouse. Both compounds are licensed in most areas for the treatment of GAD and have become established as first-line therapies as shown by randomised, double-blind studies (Davidson et al. 1999; Gelenberg et al. 2000; Rickels et al. 2000) and a large placebo-controlled, flexible-dosage trial (Pollack et al. 2001).

Paroxetine, a phenylpiperidine derivative, is a chiral SSRI but is marketed as its active (S)-enantiomer. It is the most potent inhibitor of the reuptake of 5-HT, a very weak inhibitor of noradrenaline uptake but is still more potent at this site than the other SSRIs (Rasmussen and Brosen 2000; Bourin et al. 2001). Venlafaxine is a dual 5-HT and noradrenaline (NA) reuptake inhibitor which has been claimed to have a faster onset of antidepressant action and superior clinical efficacy than for other comparable drugs as shown by randomised, double-blind studies (Nierenberg 2001; Stahl et al. 2002) and a randomised clinical trial with placebo controls (Ninan 2000). The pharmacology of venlafaxine is dose dependent: namely, at low doses it is essentially an SSRI; at medium to high doses, additional NA reuptake inhibition occurs; and at very high doses, some DA reuptake inhibition also occurs (Stahl 1998). Venlafaxine is a bicyclic phenylthylamine and has no significant affinity for 5-HT1, 5-HT2, histamine H1, muscarinic receptors, α1, α2 and β adrenoceptors, dopamine or opiate receptors in the rat brain (Muth et al. 1986; Cusack et al. 1994; Millan et al. 2001a; Briley 2003).

Conflicting results have been reported for the anxiolytic-like effects of acute paroxetine and venlafaxine administration in animal models of anxiety (Hascoët et al. 2000a; Borsini et al. 2002; Prut and Belzung 2003; Sanchez 2003). However, few detailed comparative evaluations of the role of 5-HT receptors in the anxiolytic effects of these compounds using a broad range of selective 5-HT receptor ligands have been described.

Thus after exploring the potential anxiolytic-like profile of these two drugs in the anxiety model, we evaluated the contribution of particular 5-HT2 receptor subtypes in these anxiolytic actions. 5-HT2 receptor antagonists employed included the 5-HT2A receptor antagonist, SR 46349B, the 5-HT2B/2C receptor antagonist, SB 206553 and the 5-HT2C receptor antagonist RS 10-2221 (Nic Dhonnchadha et al. 2003).

Enhancement of the synaptic availability of 5-HT appears to be a critical component of the mechanism underlying the actions of ADs. The antidepressant effect of SSRIs can be enhanced by co-administration of non-specific 5-HT2 receptor antagonists including mianserin, olanzapine and trazodone using clinical double-blind studies (Rosen et al. 1999; Ferreri et al. 2001). Recently the addition of drugs with prominent but non-specific 5-HT2 receptor antagonist properties (risperidone, olanzapine, mirtazapine and mianserin) to SSRIs has been shown to enhance therapeutic responses in patients with treatment-refractory obsessive compulsive disorder (McDougle et al. 2000). It has been suggested that the simultaneous blockade of 5-HT2A receptors and an activation of an unknown constellation of other 5-HT receptors indirectly as a result of 5-HT uptake inhibition might have greater therapeutic efficacy than either action alone. Few animal studies have been carried out for the potential synergistic action of 5-HT2 ligands and ADs in anxiety models. Co-administration of the 5-HT2C receptor antagonist (irindalone) and paroxetine in the light/dark paradigm (Mork and Hogg 2002) was shown to augment the anxiolytic-like effects of acute paroxetine administration. It was equally suggested that the co-administration of mirtazapine (an α2, 5-HT2, 5-HT3 and H1 receptor antagonist) and paroxetine induces an earlier increase in 5-HT transmission (Besson et al. 2000).

In this study we equally investigated the effects of the co-administration of inactive doses of 5-HT2 receptor agonists and antagonists with sub-active doses of paroxetine and venlafaxine in the FPT. Locomotor activity studies were conducted in parallel to investigate the effects of drug administration on the spontaneous motor activity of mice and to discount stimulant or sedative doses.

Materials and methods

Animals

All studies employed male Swiss mice (from the Janvier breeding centre, France) housed in groups of 18 per cage for 4–6 days before experimentation. A standard 12:12 light cycle was employed with lights on at 0700 and lights off at 1900 hours. Naive animals weighing approximately 20±2 g (4 weeks old) on day of testing were randomly allotted to experimental cages. All experiments are carried out between 0700 and 1200 hours, in darkened quiet rooms. Animals are used once for each experiment. All experiments were conducted in accordance with the ethical rules of the French Ministry of Agriculture for experiments with laboratory animals (no. 87.848).

Compounds utilised

5-HT2 receptor agonists

DOI-hydrochloride [(±)-1-(2,5-dimethoxy-4-iodophenyl)-2-aminopropane] (Sigma, France); BW 723C86 hydrochloride [α-methyl-5-(2-thienylmethoxy)-1H-indole-3-ethanamine] (Tocris, France); RO 60-0175 hydrochloride [(s)-2-(6-chloro-5fluoroindol-1-yl)-1-methylethylamine hydrochloride] (Roche, Switzerland).

5-HT2 receptor antagonists

SR 46349B [2-propen-1-one, 1-(2-fluorophenyl)-3-(4-hydroxyphenyl)-O-[2-(dimethylamino)ethyl]oxime] (Sanofi Recherche, France); RS 10-2221 hydrochloride [8-[5-2,4-dimethoxy-5-(4-trifluoromethylphenylsulphonamido)phenyl-5-oxopentyl]-1,3,8-triazaspiro[4,5]decane-2,4-dione] (Tocris, France); SB 206553 hydrochloride [(N-3-pyridinyl-3,5-dihydro-5-methyl-benzo[1,2-b:4,5-b′]dipyrrole-1[2H]carboxamide) hydrochloride] (Sigma, France).

Antidepressants

Paroxetine HCl [(3S-trans)-3[(1,3-benzodioxol-5-yloxy)methyl-4-(4-floorophenyl)-piperidine] (SmithKline Beecham, France), venlafaxine [1-[2-(Dimethylamino)-1-(4-methoxyphenyl)ethyl]cyclohexanol] (Wyeth-Avert, France).

BDZ ligands

Alprazolam [8-Chloro-1-methyl-6-phenyl-4H-(1,2,4)-triazolo(4,3-a)(1,4)benzodiazepine] (Sigma, France), and diazepam [(7-chloro-1-methyl-5-phenyl-3H-1,4-benzodiazepine-2[1H]-one)] (Sigma, France).

Drug preparation

DOI, RO 60-0175, SB 206553, paroxetine and venlafaxine were dissolved in distilled water. BW 723C86, RS 10-2221, diazepam, and alprazolam were suspended in a solution of distilled water with 5% of Tween 80 (Merck, Germany).

For dose effect studies all compounds were administered 30 min before testing via the intraperitoneal (IP) route, in a dosing volume of 0.5 ml/20 g of bodyweight. For interaction and association studies pre-treatment compounds were administered IP 45 min before testing and treatment compounds IP 30 min before testing. Control animals received distilled water.

Behavioural models

General procedure

All tests were performed in a quiet, darkened room. The mice (n=10 for each group) were placed in holding cages in this room at least 1 h before the test in order to reduce any neophobic response to the test-room environment. After injection (vehicle or treatment), mice were placed in their holding cages. Mice were only used once and were not handled during the housing period. Mice were always tested in a soiled apparatus and no cleaning occurred between trials, but rather at the end of each daily test. For all procedures, observers were unaware of the treatments and all experiments were performed in a randomised manner.

Locomotor activity (actimeter test)

The spontaneous activity of naïve animals was recorded using a photoelectric actimeter (Boissier and Simon 1965). This actimeter consists of a stainless steel apparatus containing transparent cages in which the animals’ horizontal activity is measured by light beams connected to a photoelectric cell. The activity is recorded during a 10-min test period. The actimeter test is performed independently of the anxiety tests in order to examine the effect of the drugs on the spontaneous locomotor activity of mice.

Four plate test

This apparatus (Bioseb, Chaville, France) consists of a cage (18×25×16 cm) floored by four identical rectangular metal plates (8×11 cm) separated from one another by a gap of 4 mm. The plates are connected to a device that can generate electric footshocks (0.6 mA; 0.5 s). Following a 15-s habituation period, the animal is subjected to an electric shock when crossing (transition) from one plate to another, i.e. two legs on one plate and two legs on another (Boissier et al. 1968). The number of punished crossings is calculated for a period of 60 s. An anxiolytic substance is capable of augmenting the number of punished passages.

Statistical analysis

Drug alone studies

Statistical comparisons were performed initially via an one way analysis of variance (ANOVA) for independent groups, after verifying the normality of distribution by a Kolmogorof–Smirnov non-parametric test. If any statistical change was observed, data were further analysed using post hoc comparisons, with a Dunnett’s test, to detect eventual differences between control and treated groups. Data were deemed significant when P<0.05. The effects of diazepam and alprazolam, included as internal standards in the anxiety model, were compared to the control group via a Student’s t-test (P<0.05).

Interaction or association studies

For interaction and association studies, a two-way ANOVA (pre-treatment×treatment) was employed for global analysis purposes. If the ANOVA showed a significant difference between groups (P<0.05), a Sidak post hoc comparison test was performed to compare the effects of pre-treatment on treatment administered.

Results

Behavioural effects of paroxetine

Acute administration of paroxetine in the actimeter test

Paroxetine (1–32 mg/kg) did not increase locomotor activity [F(6,64)=2.112; P=0.064] in comparison to the control group (Table 1).

Table 1 Effects of acute administration of paroxetine and venlafaxine, IP 30 min in the actimeter test. All data are cited as means±SEM, (n=10). Statistical analysis was performed by a one-way ANOVA followed by a Dunnett’s test for comparisons between treated groups and control group **P<0.01
Full size table

Acute administration of paroxetine in the FPT

Paroxetine (0.25–8 mg/kg) increased the number of punished passages of mice in this test [F(6,69)=17.120; P<0.001] for the doses of 0.5, 2, 4 and 8 mg/kg. Alprazolam (0.25 mg/kg) also increased the number of punished passages of mice (P<0.001) (Fig. 1).

Fig. 1
figure 1

Effects of acute administration of paroxetine and venlafaxine, IP 30 min in the FPT. The results are cited as means±SEM (n=10). Statistical analysis was performed by a one-way ANOVA followed by a Dunnett’s test for comparisons between treated groups and control group ***P<0.001, **P<0.01, *P<0.05. A Student’s t-test was used for statistical analysis between the alprazolam group and control group: ^^^P<0.001

Full size image

Behavioural effects of venlafaxine

Acute administration of venlafaxine in the actimeter test

Venlafaxine (1–64 mg/kg) increased locomotor activity at the dose of 64 mg/kg [F(7,72)=5.506; P<0.001] in comparison to the control group (Table 1).

Acute administration of venlafaxine in the FPT

Venlafaxine (0.25–16 mg/kg) increased the number of punished passages of mice in this test [F(7,79)=7.445; P<0.001] for the doses of 2, 4, 8 and 16 mg/kg. Alprazolam (0.25 mg/kg) also increased the number of punished passages of mice (P<0.001) [(Fig. 1)].

Behavioural effects of 5-HT2 receptor antagonists and paroxetine

Acute administration of SR 46349B a 5-HT2A receptor antagonist and paroxetine in the FPT

The administration of paroxetine (8 mg/kg) significantly increased the number of punishments accepted by mice [F(1,54)=66.339; P<0.001] in this test. Although ANOVA showed an overall effect on the number of punishments accepted in this test [F(2,54)=17.241; P<0.001], SR 46349B (0.1 mg/kg and 1 mg/kg) had no effect on the number of punishments accepted in this test in comparison to the vehicle treated group after post hoc analysis (P>0.05). An interaction between the two treatments was observed [F(2,54)=9.181; P<0.001] and both doses of SR 46349B antagonised the anti-punishment effects of paroxetine. Diazepam (1 mg/kg) significantly increased the number of punished passages of mice (P<0.001) (Fig. 2).

Fig. 2
figure 2

Effects of acute administration of SR 46349B (IP 45 min pre-test) and paroxetine (IP 30 min pre-test), on the number of punishments accepted by mice in the FPT. The data are cited as means±SEM, (n=10). Statistical analysis was performed by a two-way ANOVA followed by a Sidak test (***P<0.001, versus control group and +++p<0.001 versus vehicle+Parox group). A Student’s t-test was used for statistical analysis between the diazepam group and control group: ^^^P<0.001

Full size image

Acute administration of SB 206553 a 5-HT2B/2C receptor antagonist and paroxetine in the FPT

The administration of paroxetine (8 mg/kg) significantly increased the number of punishments accepted by mice in this test, [F(1,54)=120.045; P<0.001]. SB 206553 (0.1 mg/kg and 1 mg/kg) had no effect on the number of punishments accepted in this test [F(2,54)=2.161; P>0.05]. Co-administration of SB 206553 with paroxetine failed to alter the anti-punishment effects of paroxetine [F(2,54)=2.654; P>0.05]. Diazepam (1 mg/kg) significantly increased the number of punished passages of mice (P<0.001) (Fig. 3).

Fig. 3
figure 3

Effects of acute administration of SB 206553 (IP 45 min pre-test) and paroxetine (IP 30 min pre-test), on the number of punishments accepted by mice in the FPT. The data are cited as means±SEM, (n=10). Statistical analysis was performed by a two-way ANOVA followed by a Sidak test (***P<0.001, versus control group). A Student’s t-test was used for statistical analysis between the diazepam group and control group: ^^^P<0.001

Full size image

Acute administration of RS 10-2221 a 5-HT2C receptor antagonist and paroxetine in the FPT

The administration of paroxetine (8 mg/kg) significantly increased the number of punishments accepted by mice in this test, [F(1,54)=118.905; P<0.001]. RS 10-2221 (0.1 mg/kg and 1 mg/kg) had no effect on the number of punishments accepted by mice in this test [F(2,54)=4.046; P<0.05] after post hoc analysis in comparison with the control group. Co-administration of RS 10-2221 with paroxetine failed to alter the anti-punishment effect of paroxetine [F(2,54)=3.037; P>0.05]. Alprazolam (0.25 mg/kg) significantly increased the number of punished passages of mice (P<0.001) (Fig. 4).

Fig. 4
figure 4

Effects of acute administration of RS 10-2221 (IP 45 min pre-test) and paroxetine (IP 30 min pre-test), on the number of punishments accepted by mice in the FPT. The data are cited as means±SEM, (n=10). Statistical analysis was performed by a two-way ANOVA followed by a Sidak test (***P<0.001, versus control group). A Student’s t-test was used for statistical analysis between the alprazolam group and control group: ^^^P<0.001

Full size image

Behavioural effects of 5-HT2 receptor antagonists and venlafaxine

Acute administration of SR 46349B a 5-HT2A receptor antagonist and venlafaxine in the FPT

The administration of venlafaxine (8 mg/kg) significantly increased the number of punishments accepted [F(1,54)=19.120; P<0.001]. SR 46349B (0.1 mg/kg and 1 mg/kg) had no effect on the number of punishments accepted in this test in comparison to the vehicle treated group after post hoc analysis (P>0.05), despite a significant effect of the ANOVA [F(2,54)=11.395; P<0.001]. An interaction between the two treatments was observed [F(2,54)=8.684; P<0.001] and both doses of SR 46349B antagonised the anti-punishment effects of venlafaxine. Alprazolam (0.25 mg/kg) significantly increased the number of punished passages of mice (P<0.001) [(Fig. 5)].

Fig. 5
figure 5

Effects of acute administration of SR 46349B (IP 45 min pre-test) and venlafaxine (IP 30 min pre-test), on the number of punishments accepted by mice in the FPT. The data are cited as means±SEM, (n=10). Statistical analysis was performed by a two-way ANOVA followed by a Sidak test (***P<0.001, versus control group and ++p<0.01, +++P<0.001 versus vehicle+Venla. group). A Student’s t-test was used for statistical analysis between the alprazolam group and control group: ^^^P<0.001

Full size image

Acute administration of SB 206553 a 5-HT2B/2C receptor antagonist and venlafaxine in the FPT

The administration of venlafaxine (8 mg/kg) significantly increased the number of punishments accepted [F(1,54)=24.680; P<0.001] (Fig. 6). SB 206553 (0.1 mg/kg and 1 mg/kg) had no effect on the number of punishments accepted by mice in this test [F(2,54)=1.415; P>0.05]. An interaction between the two treatments was observed [F(2,54)=11.868; P<0.001] and both doses of SB 206553 antagonised the anti-punishment effects of venlafaxine. Alprazolam (0.25 mg/kg) significantly increased the number of punished passages of mice (P<0.001).

Fig. 6
figure 6

Effects of acute administration of SB 206553 (IP 45 min pre-test) and venlafaxine (IP 30 min pre-test), on the number of punishments accepted by mice in the FPT. The data are cited as means±SEM, (n=10). Statistical analysis was performed by a two-way ANOVA followed by a Sidak test (***P<0.001, versus control group and ++P<0.01, versus vehicle+Venla. group). A Student’s t-test was used for statistical analysis between the alprazolam group and control group: ^^^P<0.001

Full size image

Acute administration of RS 10-2221 a 5-HT2C receptor antagonist and venlafaxine in the FPT

The administration of venlafaxine (8 mg/kg) significantly increased the number of punishments accepted [F(1,54)=38.976; P<0.001] (Fig. 7). RS 10-2221 (0.1 mg/kg and 1 mg/kg IP 45 min pre-test) had no effect on the number of punishments accepted in this test [F(2,54)=1.432; P>0.05] after sole administration. Co-administration of RS 10-2221 with venlafaxine failed to alter the anti-punishment effect of venlafaxine [F(2,54)=1.016; P>0.05]. Alprazolam (0.25 mg/kg) significantly increased the number of punished passages of mice (P<0.001).

Fig. 7
figure 7

Effects of acute administration of RS 10-2221 (IP 45 min pre-test) and venlafaxine (IP 30 min pre-test), on the number of punishments accepted by mice in the FPT. The data are cited as means±SEM, (n=10). Statistical analysis was performed by a two-way ANOVA followed by a Sidak test (***P<0.001, versus control group). A Student’s t-test was used for statistical analysis between the alprazolam group and control group: ^^^P<0.001

Full size image

Behavioural effects of 5-HT2 receptor agonists and paroxetine

Acute administration of DOI a 5-HT2A receptor agonist and paroxetine in the FPT

Although the ANOVA was statistically significant [F(2,81)=15.969; P<0.001 and F(2,81)=13.019; P<0.001, respectively], DOI (0.06 mg/kg and 0.25 mg/kg) and paroxetine (0.25 mg/kg and 1 mg/kg) had no effect on the number of punishments accepted in this test, after post hoc analysis (P>0.05), in comparison to the control group. No interaction between the two treatments was observed [F(4,81)=1.225; P>0.05]. Diazepam (1 mg/kg) significantly increased the number of punished passages of mice (P<0.001) (Table 2).

Table 2 Effects of acute administration of 5-HT2 receptor agonists (IP 45 min pre-test) and antidepressants (IP 30 min pre-test), on the number of punishments accepted by mice in the FPT. The data are cited as means±SEM, (n=10). No significant difference between groups was observed (two-way ANOVA). A Student’s t-test was used for statistical analysis between the diazepam/alprazolam group and control group: ^^^P<0.001
Full size table

Acute administration of BW 723C86 a 5-HT2B receptor agonist and paroxetine in the FPT

BW 723C86 (0.5 mg/kg and 2 mg/kg) and paroxetine (0.25 mg/kg and 1 mg/kg) had no effect on the number of punishments accepted by mice in this test [F(2,81)=16.864; P<0.05] and [F(2,81)=36.483; P<0.05], respectively, after post hoc analysis in comparison to the control group. There was an interaction however between the two treatments [F(4,81)=4.537; P<0.05]. Both doses of BW 723C86 (0.5 mg/kg and 2 mg/kg) and paroxetine 1 mg/kg administration increased the number of plate crossings in comparison to the appropriate paroxetine group. Diazepam (1 mg/kg) significantly increased the number of punished passages of mice (P<0.001) (Fig. 8).

Fig. 8
figure 8

Effects of acute administration of BW 723C86 (IP 45 min pre-test) and paroxetine (IP 30 min pre-test), on the number of punishments accepted by mice in the FPT. The data are cited as means±SEM, (n=10). Statistical analysis was performed by a two-way ANOVA followed by a Sidak test (***P<0.001, *P<0.05 versus control group and +++P<0.001 versus vehicle+Parox. group). A Student’s t-test was used for statistical analysis between the diazepam group and control group: ^^^P<0.001

Full size image

Acute administration of RO 60-0175 a 5-HT2C receptor agonist and paroxetine in the FPT

The administration of RO 60-0175 (0.25 mg/kg and 1 mg/kg) and paroxetine (0.25 mg/kg and 1 mg/kg) had no effect on the number of punishments accepted by mice in this test [F(2,81)=5.286; P<0.05] and [F(2,81)=14.792; P<0.05], respectively, in comparison to the control group after post hoc analysis. There was an interaction between the two treatments [F(4,81)=4.717; P<0.05]. The co-administration of RO 60-0175 (1 mg/kg) and paroxetine (0.25 mg/kg) significantly increased the number of punished passages accepted by mice, in comparison to the appropriate paroxetine control group. Alprazolam (0.25 mg/kg) significantly increased the number of punished passages of mice (P<0.001) (Fig. 9).

Fig. 9
figure 9

Effects of acute administration of RO 60-0175 (IP 45 min pre-test) and paroxetine (IP 30 min pre-test), on the number of punishments accepted by mice in the FPT. The data are cited as means±SEM, (n=10). Statistical analysis was performed by a two-way ANOVA followed by a Sidak test (***P<0.001, **P<0.01, *p<0.05 versus control group and +++P<0.001 versus vehicle+Parox. group). A Student’s t-test was used for statistical analysis between the alprazolam group and control group: ^^^P<0.001

Full size image

Behavioural effects of 5-HT2 receptor agonists and venlafaxine

Acute administration of DOI a 5-HT2A receptor agonist and venlafaxine in the FPT

The administration of DOI (0.06 mg/kg and 0.25 mg/kg) and venlafaxine (0.25 mg/kg and 1 mg/kg) had no effect on the number of punishments accepted by mice in this test, after post hoc analysis (P>0.05), in comparison to the control group despite significant ANOVA [F(2,81)=21.797; P<0.001] and [F(2,81)=9.933; P<0.001], respectively. An interaction between the two treatments was observed [F(4,81)=2.936; P<0.05]. The co-administration of both doses of DOI (0.06 mg/kg and 0.25 mg/kg) and venlafaxine 0.25 mg/kg increased the number of punished passages accepted by mice in comparison to the appropriate control group, as did DOI (0.25 mg/kg) in association with venlafaxine (1 mg/kg). Alprazolam (0.25 mg/kg) significantly increased the number of punished passages of mice (P<0.001) (Fig. 10).

Fig. 10
figure 10

Effects of acute administration of DOI (IP 45 min pre-test) and venlafaxine (IP 30 min pre-test), on the number of punishments accepted by mice in the FPT. The data are cited as means±SEM, (n=10). Statistical analysis was performed by a two-way ANOVA followed by a Sidak test (***P<0.001, *P<0.05 versus control group and +++P<0.001, ++P<0.01 versus vehicle+Venla. group). A Student’s t-test was used for statistical analysis between the alprazolam group and control group: ^^^P<0.001

Full size image

Acute administration of BW 723C86 a 5-HT2B receptor agonist and venlafaxine in the FPT

The administration of BW 723C86 (0.5 mg/kg and 2 mg/kg) and venlafaxine (0.25 mg/kg and 1 mg/kg) had no effect on the number of punishments accepted by mice in this test, [F(2,81)=1.186; P>0.05] and [F(2,81)=1.855; P>0.05], respectively. There was no interaction between the two treatments [F(4,81)=0.905; P>0.05]. Alprazolam (0.25 mg/kg) significantly increased the number of punished passages of mice (P<0.001) (Table 2).

Acute administration of RO 60-0175 a 5-HT2C receptor agonist and venlafaxine in the FPT

The administration of RO 60-0175 (0.25 mg/kg and 1 mg/kg) and venlafaxine (0.25 mg/kg and 1 mg/kg) had no effect on the number of punishments accepted by mice in this test [F(2,81)=1.141; P>0.05] and [F(2,81)=0.952; P>0.05], respectively, in comparison to the control group. There was no interaction between the two treatments [F(4,81)=1.298; P>0.05]. Alprazolam (0.25 mg/kg) significantly increased the number of punished passages of mice (P<0.001) (Table 2).

Discussion

This study compared the behavioural profiles of two clinically used ADs, the SSRI paroxetine and the SNRI venlafaxine in a mouse model of anxiety, the FPT. This research demonstrated that paroxetine (0.5, 2, 4, and 8 mg/kg) and venlafaxine (2, 4, 8 and 16 mg/kg) dose dependently increased plate crossing, the order of potency being similar to that found in a previous study (Hascoët et al. 2000a). The magnitude of response was slightly greater for paroxetine. This may be due to differences in the inhibitory potencies for 5-HT re-uptake of both compounds, i.e. in vitro IC50=1 nmol/l for paroxetine and IC50=39 nmol/l for venlafaxine. Differences also exist for the re-uptake inhibitory potencies for NA and DA, respectively: IC50=350 and 5100 nmol/l for paroxetine and IC50=213 and 2,800 nmol/l for venlafaxine (Kent 2000). The correlation between these in vitro values and the in vivo effects of each compound remains ambiguous. Recent microdialysis studies demonstrated a significantly greater increase in extracellular 5-HT in the prefrontal cortex of mice by venlafaxine (8 mg/kg, IP) in comparison to paroxetine (1, 4 and 8 mg/kg, IP) (David et al. 2003). The extrapolation of these results to behavioural responses remains difficult (Deakin et al. 1993), as the question persists as to whether or not the same reaction occurs when mice (subsequent to venlafaxine or paroxetine administration), are subjected to the FPT.

The acute administration of paroxetine has produced anxiolytic-like effects in a few animal models of anxiety; the mouse FPT, the mouse L/D paradigm; isolation-induced vocalisation in rats; shock-induced vocalisations in rats, schedule-induced polydipsia and the mouse defense test battery (Njung’e and Handley 1991; Winslow and Insel 1991; Woods et al. 1993; Hashimoto et al. 1996; Schreiber et al. 1998; Hascoët et al. 2000a,b; Beijamini and Andreatini 2003). No effects have been reported in conflict procedures in rats and the social interaction paradigm (Petersen and Lassen 1981; Lightowler et al. 1994; Duxon et al. 2000) while anxiogenic-like effects have been documented in the rat EPM and L/D test (Sanchez and Meier 1997; Kõks et al. 2001) and the mouse EPM (Nic Dhonnchadha et al., unpublished data).

The acute administration of venlafaxine has not been as extensively studied in anxiety models but has, however, produced anxiolytic-like effects in the mouse FPT; isolation-induced vocalisations in guinea pigs and marble burying behaviour in mice (Hascoët et al. 2000a; Rupniak et al. 2000; Millan et al. 2001b). One major problem in interpreting these results concerns the analgesic effect of AD drugs (they have been widely used in the treatment of chronic pain). However, in animal experiments, systemic administration of ADs have yielded confusing results in tests of nociception. Theoretically, a possible analgesic action could account for the effects observed in this procedure. However, at doses active in alleviating pain in various tests, morphine did not increase the number of shocks received in the FPT (Boissier et al. 1968). Concerning AD drugs, serotonin reuptake inhibitors produce analgesic effects in the hot plate test (Fasmer et al. 1989; Ardid et al. 1992), but the effect was not stronger than those of norepinephrine reuptake inhibitors, like desipramine (Fasmer et al. 1989). Furthermore, antinociceptive activity was not observed in the hot plate reaction test with citalopram (Hyttel 1994) except at high doses (Fasmer et al. 1989), in addition the antinociceptive effect of paroxetine was found for higher doses than the one producing anxiolytic-like effect and was not affected by 5HT2 mechanism (Duman et al. 2004).

Although most SSRIs and SNRIs have both been found to have analgesic properties, they were not all found to be active in a previous study (Hascoët et al. 2000a). For example fluoxetine did not induce any anti-punishment effects in the FPT. Thus, one may conclude that the effects found for the active ADs in the FPT were indeed anxiolytic-like and not analgesic effects.

Clinical studies have demonstrated SSRI-induced anxiety and even occasional panic attacks at initiation of SSRI-treatment, a mechanism suggested to be mediated by stimulation of 5-HT2 receptors in the serotonin pathway that projects to the hippocampus and limbic cortex. Equally, the 5-HT2A receptor antagonist properties are believed to enhance antidepressant and anti-anxiety activities of many ADs (Szabo and Blier 2002). In contrast, in a previous study (Nic Dhonnchadha et al. 2003), we have observed an anxiolytic-like action of 5-HT2A/2B receptor agonists in the FPT. Different brain areas may be implicated in the anxiolytic-like responses detected, with 5-HT2 receptors in the peri-aqueductal gray area (PAG) involved in the anxiolytic response observed in the mouse EPM (unconditioned fear), whereas the same receptors in the amygdala may be involved in the response which was provoked in the FPT (conditioned fear), thus explaining different or opposing effects being observed with the same molecule, depending on the paradigm used. Thus the second part of this study involved the identification of the potential involvement of the 5-HT2 receptor subtypes in the anxiety effects of both paroxetine and venlafaxine.

Conflicting results as to the in vitro affinities of paroxetine and venlafaxine for the 5-HT2 receptor subtypes in the rat brain have been documented. The receptor binding affinity (IC50, nM) of paroxetine and venlafaxine is 18,000 and >10,000 for the 5-HT2A receptor and 20,000 and 40,000 for the 5-HT2C receptors, respectively (Hyttel 1994). Another study reported a Ki (nM) value of 6.320 for the 5-HT2A receptor for paroxetine and >100,000 for venlafaxine (Owens et al. 1997). The binding affinity for the 5-HT2B receptor was not determined. However, the in vivo binding profile of many compounds may differ substantially from the in vitro profile, complicating the interpretation of results.

The 5-HT2A receptor antagonist SR 46349B blocked the anti-punishment activity of paroxetine, while the 5-HT2B/2C receptor antagonist, SB 206553 and the 5-HT2C receptor antagonist, RS 10-2221 did not alter the effects of paroxetine. These results suggest that the 5-HT2A and not the 5-HT2B or 5-HT2C receptors are involved in the anxiolytic-like action of paroxetine in the mouse FPT. A similar study reported the participation of the 5-HT2A receptor in the ultrasonic vocalisation-reducing activity of paroxetine in rats, as only the selective 5-HT2A receptor antagonist (MDL 100,907) reversed the effects of paroxetine, whereas compounds with less selectivity, i.e. ritanserin and ketanserin, failed to alter the effects of paroxetine (Schreiber et al. 1998), possibly due to their affinity for the 5-HT2C receptor subtype.

This is one of the first studies to examine the potential role of the 5-HT2 receptors in the anxiolytic effects of venlafaxine. The anti-punishment action of venlafaxine was eliminated by both doses of SR 46349B and SB 206553 (0.1 mg/kg and 1 mg/kg), which when administered alone were without effect. The 5-HT2C receptor antagonist RS 10-2221 failed to alter the effects of venlafaxine, implicating both the 5-HT2A and 5-HT2B receptor subtypes but not the 5-HT2C receptors in the anxiolytic-like action of venlafaxine in the FPT.

Our results strongly suggest that activation of 5-HT2A receptors is critically involved in the anxiolytic activity of paroxetine, whereas the 5-HT2A and 5-HT2B receptors are involved in the anti-punishment action of venlafaxine in the FPT. A more complete profile of the exact mechanisms involved in the anxiolytic-like effects of both ADs requires more extensive interaction studies employing various receptor antagonists of other 5-HT receptor subtypes, e.g. 5-HT1B/1D, 5-HT3, 5-HT4 and monoamine transporter blockers, before definite conclusions can be drawn about the role of 5-HT receptor subtypes in the therapeutic effects of these two ADs.

The next part of the present study was undertaken to determine whether the association of 5-HT2 receptor agonists or antagonists with the ADs, paroxetine and venlafaxine could act in synergy, to induce anti-anxiety effects. The co-administration of 5-HT2 ligands with paroxetine or venlafaxine failed to alter the spontaneous motor activity of mice in the actimeter test (data not shown).

The co-administration of 5-HT2 agonists with paroxetine in the FPT demonstrated potentiation effects. DOI (a 5-HT2A receptor agonist), increased the effects of paroxetine in this model, however, in a non-significant manner, whereas RO 60-0175 (a 5-HT2C receptor agonist), only augmented the paroxetine response at one dose (1 mg/kg) and only at the weakest dose of paroxetine (0.25 mg/kg). However, BW 723C86 (a 5-HT2B receptor agonist) significantly potentiated the effects of paroxetine at both doses employed.

RO 60-0175 at 1 mg/kg may occupy 5-HT2B receptors, as it has been recently revealed to possess potent affinity for this receptor (pKi=9.3) and has now been suggested to act as a non-selective agonist for the 5-HT2C receptor (Damjanoska et al. 2003). This activation in conjunction with the lower dose of paroxetine may lead to an increase of synaptic 5-HT, resulting in the observed anti-punishment effects. At the higher dose of paroxetine the increase in 5-HT may lead to the activation of 5-HT auto-receptors, inducing the attenuation of 5-HT release and consequently decreasing 5-HT levels. Dopaminergic neurotransmission may equally be implicated as several studies have reported the opposing influence of 5-HT2C receptor ligands on dopamine firing (Di Matteo et al. 2002) depending on the brain region examined.

The mechanisms involved in the synergistic effects of BW 723C86 and RO 60-0175 with paroxetine are not the same as those of venlafaxine, as both RO 60-0175 and BW 723C86 failed to alter the response of venlafaxine after combination treatment, whereas DOI significantly increased the effects of venlafaxine. It may be that the doses used for venlafaxine were too weak to induce adequate 5-HT synaptic release in combination with BW 723C86 and RO 60-0175 to produce an anxiolytic-like effect or that the doses of the 5-HT2 receptor agonists employed resulted, in inhibitory effects via auto-receptor stimulation. Only co-treatment with DOI induced anti-punishment action; however the effects were greater at the lower doses of venlafaxine. Different neurotransmitter systems may be implicated at these doses, e.g. NA and 5-HT. Comprehensive receptor binding in conjunction with microdialysis studies would possibly enlighten us on this subject.

In conclusion, these results demonstrate that 5-HT2A and 5-HT2B receptor agonism can augment the anxiolytic-like properties of venlafaxine and paroxetine, respectively, in the FPT (Table 3). It can thus be suggested that 5-HT2 receptor agonists have the capacity to potentiate anxiolytic-like effects of these drugs and such association studies may lead to the development of novel combination therapies, e.g. SSRI or SNRI and 5-HT2A/2B receptor agonist.

Table 3 Association effects of 5-HT2 receptor agonists and antagonist and antidepressants in the FPT: summary table
Full size table