Anti-aversive effects of the atypical antipsychotic,aripiprazole,in animal models of anxiety

Introduction

Schizophrenia is a devastating psychiatric disorder divided by groups of symptoms, such as positive (delusions and hallucinations), negative (emotional flattening and poverty of speech) and cognitive. Although the neurobiological basis of these alterations remains unclear, the prevailing hypothesis is that an over-functioning of the dopaminergic mesolimbic system may explain at least the occurrence of positive symptoms (Carlsson, 2001; van Os and Kapur, 2009).

Antipsychotic drugs are the cornerstone of treatment of this disorder. The most relevant mechanism that seems to account for their therapeutic effects is the antagonism of dopamine D2-receptors (Kapur, 2004). However, most antipsychotics bind dopamine as well as serotonin (5-hydroxytriptamine [5HT]) receptors, blocking 5-HT2 and activating 5HT1A (Meltzer et al., 2003). Based on their efficacy in treating each group of symptoms, as well as on their propensity to induce motor side-effects, these drugs have been classified as typical or atypical antipsychotics. The first group tends to have their efficacy limited to the positive symptoms and to induce extra-pyramidal side-effects, whereas at least some atypical antipsychotics are likely to be also effective against negative symptoms and induce less motor impairment (van Os and Kapur, 2009).

These characteristics, however, are not homogenous in each group. Nor are they the only concerned with the undesirable effects of these drugs. In addition to motor impairment, D2-receptor blockade may induce aversive states, aggravating the negative symptoms and interfering with adherence to treatment. This ‘neuroleptic-induced dysphoria’ is often described as feelings of depression, aversion, anxiety, despair and ‘worsening well-being’ (Bressan et al., 2002; Mizrahi et al., 2007; Voruganti and Awad, 2004). The occurrence of these side-effects highly correlates with the degree of striatal D2-receptor antagonism (Bressan et al., 2002; Mizrahi et al., 2007). Altogether, these data are in line with the large body of evidence implying dopamine in the modulation of mood, motivation, reward and hedonic states (Berridge, 2007; Iversen and Iversen, 2007; Wise, 2004). Furthermore, enhancing D2-mediated signalling generally results in anxiolytic-like effects, whereas antagonizing it tends to induce the opposite (de Oliveira et al., 2009; Karl et al., 2006). Thus, other strategies, rather than dopamine receptor blockade, have been sought in the search for new antipsychotic drugs that do not induce aversive states and extra-pyramidal side-effects.

Aripiprazole, for instance, is an antipsychotic drug that apparently acts through a distinct mechanism. Pre-clinical and clinical investigations have established that it may be effective in both positive and negative symptoms with low propensity to induce extra-pyramidal side-effects (Hirose et al., 2004; Leite et al., 2008; Potkin et al., 2003). Studies on molecular pharmacology indicate that, rather than an antagonist, aripiprazole seems to be a partial agonist at dopamine D2-receptors (Burris et al., 2002; Grunder et al., 2003; Shapiro et al., 2003). This probably explains its favourable profile, since it might normalize dopamine activity without significantly interfering with the physiological roles of this neurotransmitter (Tamminga, 2002). In this way, partial agonists would be less likely to induce motor- or emotion-related side-effects.

Furthermore, this antipsychotic also binds to distinct subtypes of serotonin receptors, such as 5HT1A, 5HT2 and 5HT7 (Shapiro et al., 2003), which have been strongly implicated on the modulation of anxiety states, antidepressant activity and reward (Joca et al., 2007; Millan, 2004). The role of serotonin on anxiety, however, seems to be quite complex. Depending on the brain regions and on the model employed, drugs that interfere with serotonin receptors may induce either pro- or anti-aversive effects. In an attempt to conciliate these apparently contrasting observations, it has been proposed that this neurotransmitter could reduce innate fear and facilitate learned aversive behaviour through distinct pathways and receptors (Graeff et al., 1996). By activating 5HT1A receptors in the dorsal periaqueductal grey (DPAG), serotonin would restrain innate fear and panic, whereas through 5HT2 receptor in the amygdale, it could facilitate learned aversive behaviours (Graeff et al. 1996). Considering the wide spectrum of molecular targets for aripiprazole, it could influence various anxiety-related responses.

The unique pharmacology of aripiprazole led us to the hypothesis that it could be an antipsychotic with anti-aversive properties. Thus, we sought to test this drug in models predictive of anxiolytic- and panicolytic-like effects in rats. First, we have screened for doses that would not impair locomotion, an otherwise major confounding factor in animal models of anxiety-related behaviours. Next, we tested whether these doses of aripiprazole would inhibit aversion in the elevated plus maze (EPM), a widely used animal model for investigating anxiolytic-like effects. We also tested this drug in the elevated T maze (ETM), in which it is possible to measure two defensive responses in the same rat, avoidance and escape. The former behaviour has been linked to generalized anxiety, whereas escape has been associated with panic (Graeff et al., 1998; Pinheiro et al., 2007). The advantage of the ETM is its higher predictive value for non-benzodiazepine anxiolytics, as compared with the EPM (Pinheiro et al., 2007). To further explore the effect of aripiprazole on learned aversive responses, we tested whether this drug would prevent the expression of contextual conditioned fear (Resstel et al., 2006). Collectively, these experiments provide a broad characterization of this antipsychotic in animal models relevant for studying drug anti-aversive effects.

Material and methods

Results

EPM

In this experiment, one-way ANOVA indicated a significant effect of drug treatment on the percentage of entries [F(3,26) = 5.90; p < 0.05; η²= 0.48] and a trend toward an effect on the percentage of time spent [F(3,26) = 2.67; p = 0.06] in the open arms. No significant effect on the number of entries in the enclosed arms was observed [F(3,26) = 0.9675; p = n.s.]. The post-hoc test indicated that the diazepam- and aripiprazole 1 mg/kg-treated groups were significantly different from the others (p < 0.05) in the percentage of entries onto the open arms, and the diazepam-treated group was also significantly different in the percentage of time spent in these arms, as seen in Figure 1. Thus, these data point to an anxiolytic-like effect of aripiprazole.

OPEN IN VIEWER

Figure 1.

Effect (mean ± SEM) of aripiprazole (ARI; 0.1 and 1 mg/kg) and diazepam (DZP, 2.5 mg/kg) injections in the elevated plus maze (EPM) (n = 7–9 per group).^* p < 0.05 compared with the vehicle-treated group.

ETM

The two-way ANOVA of avoidance indicated a significant effect of trial [F(2,27) = 12.91; p < 0.05; η²= 0.14], treatment [F(2,27) = 3.38; p < 0.05; η²= 0.09] and interaction [F(27,54) = 2.79; p < 0.05; η²= 0.06], as seen in Figure 2(a). In the escape analysis two-way ANOVA indicated no significant effects of trial [F(2,27) = 0.65; p > 0.05], but a significant effect of treatment [F(2,27) = 12.52; p < 0.05; η²= 0.32] and interaction [F(27,54) = 3.123; p < 0.05; η²= 0.05] showed in Figure 2(b). These data reinforce the observations from the EPM, suggesting an anti-aversive effect of aripiprazole.

OPEN IN VIEWER

Figure 2.

Effect (mean ± SEM) of aripiprazole (ARI; 1 and 1 mg/kg) injections in the inhibitory avoidance (a) and escape (b) latencies measured in the elevated T maze (ETM) (n = 10 per group); ^* p < 0.05 compared with the vehicle-treated group in the same trial.

Contextual fear conditioning

The administration of aripiprazole was able to reduce the percentage of time in freezing [F(2,21) = 10.03, p < 0.05; η²= 0.48]. The decrease in freezing was evoked by the 0.1 mg/kg dose (43.1 ± 7.1%, p < 0.01), but not by the dose of 1 mg/kg (69.3 ± 6.6%, p > 0.05), when compared with the vehicle-control group (78.7 ± 2.8%) (Figure 3).

OPEN IN VIEWER

Figure 3.

Effect (mean ± SEM) of aripiprazole (ARI; 0.1 and 1 mg/kg) injections in the freezing behaviour evoked by contextual conditioning to foot shock (n = 8 per group). ^* p < 0.05 compared with the vehicle-treated group.

Discussion

The present study demonstrates that aripiprazole, an atypical antipsychotic, exhibits anti-aversive properties, as revealed by diverse animal models. At the dose of 1 mg/kg, this drug induced an anxiolytic-like effect in the EPM and inhibited escape responses in the ETM. Furthermore, at the dose of 0.1 mg/kg, aripiprazole attenuated inhibitory avoidance in this latter model, in addition to inhibiting freezing response conditioned to a context previously paired with foot shock. None of these doses interfered with either the distance moved in the open field or the total number of entries in the enclosed arms of the EPM. Overall, these data suggest that this compound may induce anxiolytic- and panicolytic-like effects at doses that do not impair spontaneous locomotion.

Several antipsychotics may impair motor activity. In the present experiment with the open field, only the higher dose of aripiprazole (10 mg/kg) interfered with motor activity. Generally, this is in agreement with previous data from this and other groups showing that this drug does not tend to cause motor impairment at effective antipsychotic-like doses (Hirose et al., 2004; Kleven et al., 2005; Leite et al., 2008; Natesan et al., 2006). This profile contrasts with those observed with haloperidol and clozapine, which may impair locomotion in the open field even at low doses (Moreira and Guimarães, 2005). These animal studies are in line with clinical data showing that aripiprazole does not tend to induce significant sedation (Potkin et al., 2003), a major side-effect of both haloperidol and clozapine (Pretorius et al., 2001).

In the EPM, aripiprazole similarly to the clinically effective anxiolytic diazepam, increased open arm exploration, unveiling the anxiolytic-like property of this drug. This occurred without any modification in the number of entries in the enclosed arms, an index of motor activity (Carobrez and Bertoglio, 2005), further reinforcing that the effect is specific to aversive states, rather than secondary to any change in locomotor activity.

Notwithstanding the major importance of the EPM, this animal model of anxiety has some drawbacks. First, it may yield false negatives for several drugs, such as those that may interfere with serotonin-mediated neurotransmission (Pinheiro et al., 2007). Second, it does not distinguish between anti-aversive states relevant for generalized anxiety disorder or panic disorder. For this reason, we tested aripiprazole in the ETM, a more sophisticated model that detaches two components, the avoidance reactions and the escape response from the open arm, possibly reminiscent of anxiety and panic states, respectively (Graeff et al., 1998; Pinheiro et al., 2007). Neurobiological and pharmacological evidence has suggested that compounds able to reduce the avoidance time might be of relevance for the treatment of generalized anxiety or learned aversion, whereas inhibition of escape response are possibly relevant for panic disorder (Graeff et al., 1998). The present study shows that aripiprazole, at the lower dose, significantly impaired inhibitory avoidance acquisition while, at the higher dose, inhibited escape expression. This result supports the notion that this drug consistently induces anti-aversive effects, contrasting to other antipsychotics. This is also reinforced by the reduction of freezing observed in our contextual fear conditioning model.

The rationale why aripiprazole differs from other drugs of this class regarding the effects on mood and anxiety remains to be elucidated. Actually, even the mechanisms of its antipsychotic activity are not fully understood. This drug has a complex pharmacology, binding to a wide spectrum of receptors. It is believed to act as a partial agonist at dopamine receptors (Burris et al., 2002; Shapiro et al., 2003; Grunder et al., 2003), contrasting with all other antipsychotics, which are dopamine antagonists (Kapur, 2004). Furthermore, it may also bind to serotonin 5HT1A and 5HT2 receptors, acting as partial agonist and antagonist respectively (Newman-Tancredi et al., 2005; Shapiro et al., 2003).

One potential explanation for these anti-aversive properties is the partial agonistic property at dopamine D2 receptors. Accordingly, whereas D2-receptor antagonists induce aversive states, activation of this receptor with quinpirole (D2 agonist) in the ventral tegmental area (VTA) is able to decrease fear potentiated startle expression (de Oliveira et al., 2009). Although not a full agonist, aripiprazole could maintain a ‘dopaminergic tonus’ in this structure when levels of the neurotransmitter itself are low. Furthermore, given that the VTA projects also to the amygdaloid complex, where dopamine release is involved in the expression of conditioned emotional reactions (Pezze and Feldon, 2004), the activation of D2-receptors in the VTA by aripiprazole could reduce amygdala dopamine levels. In line with this view, lesions of cathecolaminergic neurons to the amygdaloid complex lead to impairment in the expression of conditioned aversive response (Selden et al., 1991).

In addition to dopamine, serotonin-related mechanisms might also explain the anti-aversive effects of aripiprazole. Considering the affinity of this drug for serotonin receptors (Shapiro et al., 2003), they could involve either direct blockade of 5HT2 receptors in the amygdala or partial agonist activity upon autossomic 5HT1A receptors located in the dorsal raphe nucleus (Graeff et al., 1996). Indeed, local blockade of 5HT1A receptors in this structure induces anxiolytic-like effects, possibly by increasing serotonin output onto this same receptor subtype located in the DPAG (Graeff et al., 1996). A similar effect is observed after 5HT2 blockade in the amygdale, where, in contrast, serotonin seems to facilitate aversion (Graeff et al., 1998; Pinheiro et al., 2007). Follow-up studies are required to test for these dopaminergic and serotonergic hypotheses, which are complementary, rather than mutually exclusive, to explain the effects of aripiprazole. In any case, despite this broad spectrum of actions of this drug, dopamine partial agonism seems to be the main mechanism accounting for its clinical effects in schizophrenia, given that the antipsychotic doses of aripiprazole in humans appear to bind preferentially to D2 rather than 5HT1A or 5HT2 receptors (Mamo et al., 2007).

Because the literature on the behavioural effects of aripiprazole is still scant, future experiments employing local intra-brain injections should be performed to test these possibilities and to try to identify the possible neuroanatomical sites for the anti-aversive effects of this drug. Nevertheless, the present data are in agreement with previous studies showing that this drug increased social interaction (Bruins Slot et al., 2005), inhibited marble-burying behaviour (Bruins Slot et al., 2008), reduced ultrasonic vocalizations (Dahan et al., 2009) and potentiated the effect of antidepressant drugs (Bourin et al., 2009; Chernoloz et al., 2009). This is relevant considering that schizophrenic patients suffer from anxious feeling as well as impaired humour and motivation (van Os and Kapur, 2009). Therefore, the anxiolytic-like and the antidepressant-potentiating effects of aripiprazole point to a favourable profile in terms of emotional responses, in contrast to most drugs of its class.

While these works may suggest the presence of anti-aversive properties, the present study is the first to characterize the effects of aripiprazole in elevated mazes and fear conditioning models. As discussed above, in contrast to this drug, most antipsychotics, such as haloperidol and risperidone, induce pro- rather than anti-aversive effects in these models (Karl et al., 2006), probably reflecting the fact that several drugs in this class may induce aversive states, dysphoria and depression, thereby aggravating, rather than alleviating, the negative symptoms of schizophrenia (Betensky et al., 2008). Aripiprazole, on the other hand, may even be useful against post-traumatic stress disorder as revealed by an open-label trial (Mello et al., 2008). It has also been investigated as a potential add-on therapy for major depression (Jarema 2007; Preskorn, 2009; Shelton and Papakostas, 2008). Quite interestingly, antipsychotic-induced dysphoria highly correlates with the levels of dopamine receptor occupancy for conventional antipsychotics, whereas aripiprazole does not interfere with ‘well-being’ even at high levels of occupancy, possibly due to its partial agonist activity (Mizrahi et al., 2009). In any case, further long-term clinical studies are required to support the advantage of partial agonists as compared with conventional antipsychotics in terms of efficacy and side-effects. So far, the main side-effects detected during aripiprazole treatment in clinical trials are weight gain and associated metabolic disorders, problems commonly seen with antipsychotic treatments (Potkin et al., 2003).

In conclusion, aripiprazole induced anti-aversive effects in the EPM and the ETM models of anxiety. In addition, it inhibits the expression of contextual fear conditioning. The study unveils new effects of this drug and may encourage the development of compounds with a similar pharmacological profile. They may be useful as possible adjunctive anxiolytic/antidepressant drugs or as antipsychotics that alleviate, rather than aggravate, the aversive states already present in schizophrenia.