Long-Term Administration of Citalopram Reduces Basal and Stress-Induced Extracellular Noradrenaline Levels in Rat Brain

Psychopharmacology (2007) 194:73–81 DOI 10.1007/s00213-007-0826-8

ORIGINAL INVESTIGATION

Long-term administration of citalopram reduces basal and stress-induced extracellular noradrenaline levels in rat brain

Yukie Kawahara & Hiroshi Kawahara & Fumi Kaneko & Masatoshi Tanaka

Received: 24 December 2006 /Accepted: 7 May 2007 /Published online: 30 May 2007 # Springer-Verlag 2007

Abstract suppressed only in the LC. The effect of local application of Rationale Panic disorders are commonly treated with clonidine was enhanced only in the BLA. selective serotonin reuptake inhibitors (SSRIs). However, Conclusion The present results indicate that chronic ad- the effect of SSRIs on noradrenaline systems in the brain ministration of citalopram strongly decreases the extracel- has not been fully elucidated at the present time. lular levels of noradrenaline in the brain. The anti-panic Objectives The effects of long-term administration of effect of citalopram might be due to sensitization of the α2- citalopram, an SSRI, on basal as well as stress-induced adrenoceptors leading to suppression of the stress response extracellular noradrenaline levels in the basolateral nucleus through noradrenergic activity. This mechanism is specific of the amygdala (BLA) and the locus coeruleus (LC) were for the BLA. determined. In addition, the responsiveness of noradrenaline transporters and α2-adrenoceptors were determined Keywords Citalopram . Microdialysis . Amygdala . after long-term administration of citalopram. Basolateral . Rat . Noradrenaline . Chronic treatment . Materials and methods Brain microdialysis was used to Locus coeruleus assess the extracellular levels of noradrenaline in conscious rats. Desipramine and clonidine were used to functionally evaluate the noradrenaline transporter and α2-adrenoreceptor, Introduction respectively. Results In rats treated daily for 14 days with citalopram Dysfunction of the central noradrenergic system has been −1 −1 (10 mg kg day s.c.), dialysate noradrenaline levels demonstrated in anxiety disorders, such as panic disorders, showed remarkable decreases in both the BLA and the LC in various studies (Brunello et al. 2003; Ressler and to about 25 and 45% of controls, respectively. The stress- Nemeroff 2001; Sullivan et al. 1999). The involvement of induced increase of noradrenaline was almost completely noradrenergic systems in stress responses, in particular in abolished in the BLA, but was relatively stable in the LC. the amygdala, locus coeruleus and hypothalamus, has been The effect of local application of desipramine tended to be reported by ourselves and others (Debiec and LeDoux 2006; Neophytou et al. 2001; Tanaka et al. 2000). As general activation of these brain regions seem to be involved in the pathology of panic disorders, it is of interest * : : Y. Kawahara ( ) F. Kaneko M. Tanaka to investigate whether selective serotonin reuptake inhibitor Department of Pharmacology, Kurume University School of Medicine, (SSRI) might affect the activity of noradrenergic neurons in Kurume 830 0011, Japan these regions (Gorman et al. 2000). e-mail: [email protected] Despite the fact that the pathology of anxiety disorders has been related to the noradrenergic system, SSRIs such as H. Kawahara Department of Dental Anesthesiology, Kyushu Dental College, citalopram, fluoxetine, and proxetine, are currently the first Kitakyushu 803 8580, Japan line treatment of anxiety disorders, i.e., panic and obses- 74 Psychopharmacology (2007) 194:73–81 sive-compulsive disorders (Tukel et al. 2006; Seedat et al. serotonin transporters, with little affinity for noradrenaline 2003; Pollock 2001; Wade 1999). transporters (Owens et al. 1997). The noradrenaline

It has been well established that SSRIs selectively block transporter and α2-adrenoceptors have been reported to be serotonin reuptake sites (Owens et al. 1997), and it is desensitized after chronic treatment with noradrenaline assumed that their clinical effects are attributed to increased reuptake inhibitors (Benmansour et al. 2004; Invernizzi levels of serotonin in the brain. However, accumulating and Garattini 2004). Therefore, we investigated whether evidence indicates that the noradrenergic system is also chronic citalopram affected the noradrenaline transporters implicated in the action of SSRIs. For instance, extracellu- and α2-adrenoceptors. For this purpose, the effects of local lar noradrenaline levels in the cortex are increased by infusion of desipramine and clonidine into the BLA or the SSRIs when administered both acutely and repeatedly LC by retrograde microdialysis in the chronically citalopram- (David et al. 2003; Thomas et al. 1998). Electrophysiolo- treated group were compared with saline-injected controls. gical studies have demonstrated that repeated administration of citalopram, as well as paroxetine, reduced the spontaneous firing rate of noradrenergic neurons (Szabo et al. 1999, 2000). In humans, significant normalization of Materials and methods noradrenergic function has been demonstrated during effective treatment with SSRIs in patients with panic Animals disorder in which a greater reduction of the major metabolite of noradrenaline, 3-methoxy-4-hydroxyphenyle- Male albino Wistar rats (280–320 g) were used for these thyleneglycol (MHPG) was observed compared to healthy experiments. The rats were maintained at 23±2°C under a volunteers (Coplan et al. 1997). Although the interaction of 12-h light–dark cycle (lights on at 08:00 A.M.) with free noradrenaline and serotonin has been demonstrated previ- access to food and water. After probe implantation, the rats ously (Svensson 2000; Shiekhattar and Aston-Jones 1993), were housed individually in plastic cages (30×30×40 cm). the mechanism by which SSRIs interact with the noradrenergic system is largely unknown. Based on these reports, Materials we presumed that the noradrenergic system could be functionally altered when serotonergic neuronal activity is 1-(3-Dimethylaminopropyl)-1-(4-fluorophenyl)-5-phtalan- enhanced by long-term administration of SSRIs. carbonitril (citalopram hydrobromide, generously provided The basolateral amygdaloid nucleus (BLA) serves as an by H. Lundbeck A/S, Copenhagen, Denmark) was dissolved integrator and relay center for the sensory and memory in saline (0.2 ml) or Ringer’s solution for systemic injection information necessary for anxiety and panic responses and local application, respectively. (Shekhar et al. 2003). Neuroimaging studies indicate that the amygdala is involved in the provocation of fear (Morris Surgery and brain dialysis et al. 1999). The BLA receives input from both noradrenergic and serotonergic neurons mainly from the brain stem, Microdialysis was performed with an I-shaped cannula. The e.g., A6 and B9 cell groups, respectively, and these neurons probe was implanted in the right BLA (exposed length are highly responsive to various types of aversive stimuli 1.0 mm) or the right LC (exposed length 1.5 mm) under (Uehara et al. 2005; Ferry and McGaugh 2000). Therefore, pentobarbital anesthesia (50 mg/kg i.p.) and local applica- it is likely that SSRIs modulate these neuronal inputs to the tion of 10% lidocaine. The coordinates of the implantation BLA, and this effect could be involved in the clinical were: A/P −2.8 mm, L/M 5.0 mm, V/D 8.8 mm from the efficacy of SSRIs in anxiety disorders. In addition, we bregma and dura for the BLA. For the LC, A/P −3.3 mm, selected the locus coeruleus (LC) to determine noradren- L/M 1.3 mm, V/D 8.3 mm at the angle of 15° in the sagittal aline levels, which may reflect a general effect of SSRIs on plane from lambda and the surface of the scull, respectively. the noradrenergic system, as most of the noradrenaline in Microdialysis experiments were conducted 24 h after the central nervous system is derived from the LC. implantation of the probes as described in previous reports The present microdialysis study was designed to studying noradrenaline (McKittrick and Abercrombie 2007; characterize the effect of long-term administration of Valentini et al. 2005) An online approach was used in citalopram on extracellular levels of noradrenaline in the which the probes were perfused with Ringer’s solution at a BLA and the LC. The effect of handling stimuli, which is flow rate of 2.0 ml/min by an infusion pump (EICOM, considered to elicit a mild emotional stressful state, on the Kyoto, Japan). Dialysate fractions were collected every release of noradrenaline was determined after long-term 20 min. administration of citalopram. Citalopram was selected, as it Noradrenaline was quantified by high-performance has the greatest selectivity among the known SSRIs for liquid chromatography using a reverse-phase column Psychopharmacology (2007) 194:73–81 75

Table 1 Basal noradrenaline levels in dialysates: control vs long-term Results citalopram

Control Long-term citalopram Effect of long-term administration of citalopram on basal levels of noradrenaline BLA 5.36±1.33 (n=16) 1.33±0.19 (n=21)** n n LC 4.62±0.81 ( =19) 2.04±0.43 ( =14)* The mean basal values for noradrenaline in the BLA and Data are expressed as mean±SEM (fmol/20 min). the LC corresponding to each experimental group are *Indicates p<0.05 provided in Table 1. **Indicates p<0.01 vs control The noradrenaline levels were significantly reduced, compared to baseline, after long-term administration of (150×4.6 mm; Supelco LC18, Belleofonte, PA, USA) with citalopram in both the BLA (to about 25% of controls) and electrochemical detection. An EICOM EP-300 pump (Kyoto, the LC (to about 45% of controls) as summarized in Fig. 1. Japan) was used in conjunction with an electrochemical detector (ESA; potential first cell, +180 mV; potential second Effect of long-term administration of citalopram cell, −180 mV). The mobile phase consisted of a mixture of on stress-induced noradrenaline in dialysates

4.1gsodiumacetateadjustedtopH5.5,50mgNa2EDTA and Handling for 20 min caused a rapid and significant increase of 140 mg octanesulfonic acid in 900 ml H2O, and 100 ml/ l methanol. The flow rate was 1.0 ml/min. The detection limit noradrenaline to about 200% of basal levels in both the BLA of the assay was about 0.3 fmol per sample (on-column). The and the LC in saline-injected controls. In the long-term composition of the Ringer’s solution was (in mM): NaCl citalopram-treated rats, the handling-induced increase of noradrenaline was almost completely abolished in the BLA 140.0, KCl 4.0, CaCl2 1.2, MgCl2 1.0. After the collection of three baseline samples, citalopram was administered system- (Fig. 2a). In contrast, the increased relative noradrenaline ically (s.c.). At the end of the experiments, the rats were levels in the LC caused by handling were similar in both the given an overdose of chloral hydrate, and the brains were controls and long-term citalopram-administered rats (Fig. 2b). fixed with 4% paraformaldehyde via intracardiac infusion. Effect of long-term administration of citalopram μ Coronal sections (16 m thick) were cut, and dialysis probe on the desipramine-induced increase of noradrenaline placement was localized according to the atlas of Paxinos and in dialysates Watson (1986). Local infusion of desipramine into the BLA (1 μM) or the Long-term administration of citalopram LC (0.1 μM) in saline-injected controls caused a gradual increase of noradrenaline to about 200 and 400% of basal Rats were treated with citalopram (10 mg/kg s.c.) once levels, respectively. Long-term administration of citalopram daily for 14 days. Microdialysis experiments were con- attenuated the desipramine-induced relative increase of ducted 48 h after the last injection. The control group noradrenaline only in the LC (Fig. 3b), with no effect in received saline (0.2 ml) once daily for 14 days. The the BLA (Fig. 3a). The AUC% of increases detected in the experiments were approved by the Committee for Animal LC between the controls and the long-term citalopram Experimentation, Kurume University School of Medicine. group did not reach statistical significance (Fig. 3c).

Expression of results and statistics

All values are expressed as a percentage of the average of three baseline samples. The average concentration of three stable baseline samples was set at 100%. Statistical analysis (Statview; SAS Institute, Cary, NC, USA) was performed using repeated measures one-way analysis of variance and Dunnett’s multiple comparison test for post hoc determina- tion of significant differences between baseline and post- drug administration. Two-way analysis of variance and Scheffe’s multiple comparison test for post hoc determina- tion were used for comparison between experimental groups expressed as the area under curve (AUC). AUC Fig. 1 Effects of long-term administration of citalopram on basal was presented as the total relative amount of noradrenaline extracellular levels of noradrenaline in the BLA and the LC. *p<0.05, measured. The level of significance was set at p<0.05. **p<0.01 vs control groups 76 Psychopharmacology (2007) 194:73–81

Fig. 2 Effects of long-term administration of citalopram on the n=8; long-term citalopram expressed with closed circles: n=7) (b). increased level of handling-induced noradrenaline in the BLA (control Open squares indicate the period of handling (20 min). Data are expressed as open circles: n=5; long-term citalopram expressed with expressed as the mean±SEM percent variation of basal levels. *p< closed circles: n=8) (a) and the LC (control expressed as open circles: 0.05, **p<0.01 vs basal levels

Effect of long-term administration of citalopram citalopram did not cause any effect until at least 120 min on the clonidine-induced decrease of noradrenaline after the injection in both saline-injected controls and the long- in dialysates term citalopram-treated group (data not shown). Despite the short half-life of citalopram (~3 h in rats when administered Local infusion of clonidine into the BLA (10 μM) or into systemically; Fredricson Overo 1982),thestableincreasesof the LC (10 μM) caused rapid and significant decreases to extracellular serotonin levels (to about 250% of basal level) about 60 and 20% of basal levels in saline-injected controls, immediately after systemic injection of citalopram (10 mg/kg) respectively. Long-term administration of citalopram en- have been demonstrated to last longer than 3 h (Millan et al. hanced the effect of clonidine in the BLA (Fig. 4a), but not 2001; Huang et al. 2006) even at lower doses (Boothman in the LC (Fig. 4b). The AUC% of the decreases between et al. 2006; Bosker et al. 2001). These reports indicate that the the control and citalopram-treated groups reached statistical increases in serotonin levels do not directly reflect the brain significance in the BLA (Fig. 4c). concentrations of citalopram. Long-term citalopram, at the same dose administered through the same route used in the present study (once daily injection of citalopram 10 mg/kg s.c.

Discussion for 14 days), has been demonstrated to upregulate serotonin2C receptors (Laakso et al. 1996).Basedontheseprevious Effect of long-term administration of citalopram on basal results, once daily injection of citalopram was selected in the levels of noradrenaline present study, as it was considered to be adequate for altering the serotonin systems. Therefore, the lack of effect with single Long-term administration of citalopram led to a remarkable administration of citalopram in both experimental groups was reduction of basal noradrenaline levels in dialysates in both unlikely to be due to the dose of citalopram administered. the BLA and the LC, whereas single administration of These results suggest that the decreased levels of noradren-

Fig. 3 a Effect of local infusion of desipramine (1 μM) into the BLA the infusion period of desipramine (140 min). Data are expressed as on extracellular noradrenaline in dialysates from the BLA (control the mean±SEM percent variation of basal levels within the same expressed with open circles: n=8; long-term citalopram expressed group. *p<0.05, **p<0.01 vs basal levels. c The comparison of the with closed circles: n=6). b Effect of local infusion of desipramine AUC for the relative increases of noradrenaline detected in the LC (1 μM) into the LC on extracellular noradrenaline in dialysates from between control and long-term citalopram group while desipramine the LC (control expressed with open circles: n=8; long-term was infused into the LC. The decreases between the control and long- citalopram expressed with closed circles: n=6). Open squares indicate term citalopram groups did not reach statistical significance Psychopharmacology (2007) 194:73–81 77

Fig. 4 a Effect of local infusion of clonidine (10 μM) into the BLA the infusion period of clonidine (140 min). Data are expressed as the on extracellular noradrenaline in dialysates from the BLA (control mean±SEM percent variation of basal levels within the same group. expressed with open circles: n=6; long-term citalopram expressed *p<0.05, **p<0.01 vs basal levels. c The comparison of the AUC for with closed circles: n=6). b Effect of local infusion of clonidine the relative decreases of noradrenaline detected in the BLA between (10 μM) into the LC on extracellular noradrenaline in dialysates from the control and long-term citalopram groups while clonidine was the LC (control expressed with open circles: n=6; long-term infused into the BLA. *p<0.05 vs control group citalopram expressed with closed circles: n=6). Open squares indicate aline after long-term administration of citalopram could be a reduced extracellular levels of noradrenaline after chronic consequence of long-term alteration of the neural system, but administration of citalopram in the present study. not due to acute effects. In the BLA, the reduction of basal noradrenaline might The long-term citalopram-induced decrease of noradren- be explained by the reduced LC activity. However, aline was greater in the BLA (to about 25%) compared with noradrenergic innervation of the amygdala originates not that in the LC (to about 45%). The results of decreased only from the LC but also from other noradrenergic groups noradrenaline level in the dialysates were supported by from the brain stem, e.g., A2 (Fallon 1981, Fallon et al. previous electrophysiological data demonstrating that 1978; McGaugh 2004). Therefore, the reduction of nor- chronic administration of citalopram caused a gradual adrenaline in the BLA is not likely a direct reflection of decrease in spontaneous firing activity of LC adrenergic reduced noradrenergic activity of the LC. The possibility of neurons in anesthetized rats (Szabo et al. 2000). Apart from regional differences of the mechanism responsible cannot citalopram, other antidepressants (e.g., paroxetine, fluox- be excluded, as regional differences of the interaction etine, desipramine, phenelzine, mianserin and more) have between serotonergic and noradrenergic systems have been been shown to reduce the activity of LC neurons after suggested previously (Baba et al. 1995). Indeed, the degree chronic treatment (Szabo et al. 1999; Sodero et al. 2004; of noradrenaline decrease observed in the BLA was larger Grant and Weiss 2001). Decreased activity of LC neurons than that in the LC, as shown in Fig. 1. causes a reduction of extracellular noradrenaline levels, as Although the enhancement of extracellular serotonin in extracellular noradrenaline in the LC is a result of the amygdala by citalopram administration has been neurogenic, impulse flow-dependent vesicular exocytosis demonstrated in a previous microdialysis study (Bosker et (Fernandez-Pastor et al. 2005). Furthermore, chronic al. 2001), there is no direct evidence of any pharmacolog- treatment of citalopram decreased plasma levels of nor- ical interaction between serotonin and noradrenaline in this adrenaline and adrenaline in rats (Jongsma et al. 2005). In region to date. Boyer et al. (1998) suggested that chronic humans, decreased levels of the noradrenergic metabolite, citalopram altered N-methyl-D-aspartate (NMDA) receptor MHPG, were measured in cerebrospinal fluid after therapy composition strictly in the amygdala, which strongly with antidepressants (Grant and Weiss 2001). Considering modulates the interneurons in the BLA (Rainnie 1999) that approximately 70% of the noradrenaline in the brain without affecting the NMDA receptors in the LC. Suppres- originates in the LC, the present findings are consistent sion of basal noradrenaline levels in the BLA likely occurs with these peripheral studies. The question arises how via a mechanism independent of the LC activity. chronic citalopram reduced the noradrenaline levels in the In this regard, it is of interest to note that chronically LC. Electrophysiology studies have demonstrated enhance- administered fluoxetine did not alter the basal release of ment of the inhibitory effects of dorsal raphe serotonergic noradrenaline in the hippocampus (Page and Abercrombie neurons on LC neurons (Haddjeri et al. 1997; Segal 1979) 1997), whereas LC neural activity was suppressed (Sodero after 2 weeks of administration of an SSRI (Millan et al. et al. 2004). Thus, the terminal release in the hippocampus 2000). A similar mechanism might be involved in the does not always correlate to the activity of LC neurons. 78 Psychopharmacology (2007) 194:73–81

Effect of long-term administration of citalopram delayed modulation of NMDA receptors on dendrites in on stress-enhanced noradrenaline the BLA (Gracy and Pickel 1995;Boyeretal.1998), likely via these cortical projections, could be related to the It is well known that the central noradrenergic system is suppression of noradrenergic activity without affecting highly responsive to various types of stressful stimuli GABAergic transmission as demonstrated in an in vitro (Vermetten and Bremner 2002), including the BLA (Tanaka study (Ferry et al. 1997). As most of the NMDA receptors et al. 2000; Ferry and McGaugh 2000)andtheLC in the BLA are located on dendrites (Gracy and Pickel (Kawahara et al. 1999). 1995), glutamate and noradrenaline could interact at the Herein, we presented the results of the stress experiments terminal levels, e.g., axon–axon interaction. as relative values, not absolute values between the control and In contrast, the noradrenaline in the LC was still citalopram-treated group. As general suppression of the responsive to the handling stimuli, while the basal level noradrenergic activity was considered in the citalopram- was significantly decreased in both the LC and the BLA. treated group, as shown in Fig. 1, to evaluate the responsive- These results suggest that noradrenergic function in these ness of neural functions, relative values were considered as two regions are different under stressful conditions, as appropriate indices during stressful conditions as well as supported by the recent study indicating that stress during pharmacological application. increases energy metabolism strictly in the BLA (Uehara As anticipated, handling stimuli considered to cause an et al. 2005). acute and emotional stressful state enhanced noradrenaline in dialysates in both the BLA and the LC of control Effect of long-term citalopram on the desipramine-induced animals. Long-term administration of citalopram abolished increase of noradrenaline the stress-induced increase of noradrenaline in the BLA alone. In contrast, the relative response of noradrenaline to To investigate whether the reduction of noradrenaline levels stress was still maintained in the LC. The result obtained in in these two regions are due to reduced release from intact the LC is in some agreement with the data using chronic neurons, the noradrenaline transporters and α2-adrenocep- sertraline, another highly selective SSRI, that demonstrated tors were functionally evaluated by local infusion of no change in the LC activity in response to stress in desipramine and clonidine. anesthetized rats (Valentino et al. 1990). The stress In the control group, the increases of noradrenaline response of noradrenaline has also been investigated after elicited by the noradrenaline transport blocker, desipramine, chronic fluoxetine treatment by microdialysis (Page and were greater in the LC than in the BLA, which was expected, Abercrombie 1997) in which the noradrenaline response to as the density of noradrenaline transporters is much higher in tail pinch stimuli was enhanced in the hippocampus. the LC than the amygdala (Kung et al. 2004). The present finding obtained in the BLA is supported by The effect of desipramine infused locally into the LC in previous behavioral and immunohistochemical investiga- the present study is similar to the previous data demonstrated tions showing that the target site of the anxiolytic effect of by Mateo et al. (2001) indicating that 0.1 μM desipramine, citalopram in the brain is the BLA (Inoue et al. 2004). at the same dose as applied in the present study, had little Furthermore, human volunteers treated with citalopram for effect on basal noradrenaline levels for 70 min after the 7 days exhibited decreased amygdala response to the initiation of drug infusion. Although the effect in the LC presentation of threat using functional magnetic resonance was greater than that in the BLA in the present study, this imaging (Harmer et al. 2006). Taken together, the BLA is increase did not reach statistical significance 80 min after apparently involved as the target for the anti-panic and anti- the initiation of drug infusion. A similarly delayed increase fear effects of citalopram, and delayed suppression of was also obtained in the BLA (Fig. 3). This study confirmed noradrenergic activity in this region is likely critical for that the increased levels of noradrenaline in the LC, as a the delayed onset of action of citalopram. The mechanism result of blockade of the noradrenaline transporters, could of suppression of the neurons in the BLA remains activate the α2-adrenoceptors in the LC. Such a mechanism unknown. However, the region-specific effects of citalo- might be involved in the delayed increase of extracellular pram within this region have been demonstrated by altered noradrenaline induced by local infusion of desipramine in NMDA receptor composition after 16 days of treatment the LC as well as the BLA. with citalopram (Boyer et al. 1998). The target effector In the long-term citalopram group, the relative response sites for acute citalopram have been shown by suppression of noradrenaline to the local application of desipramine of c-fos expression in the BLA and cortical areas, tended to be lower in the LC, but not in the BLA. including the somatosensory cortex (Izumi et al. 2006). This result implies that the tendency towards attenuation Such cortical excitatory projections drive BLA interneural of the effect of desipramine in the LC is attributed the firing (Rosenkranz and Grace 2001). Therefore, the sensitization of α2-adrenoceptors in the LC. However, the Psychopharmacology (2007) 194:73–81 79

functional evaluation of α2-adrenoceptors by retrograde transmission was due to activation of α2-adrenoceptors application of clonidine into the LC in the present study did suggests that the reduced extracellular basal levels of not confirm sensitization of the receptors. In the BLA, the noradrenaline in the BLA observed in the present study delayed increase of noradrenaline caused by local applica- might be involved in the latter phenomenon. tion of desipramine was not modified in the long-term In clinical studies of panic disorders, the modulation of citalopram group. However, retrograde application of α2-adrenoceptors is dependent on the brain region (Nutt clonidine into the BLA revealed sensitization of the α2- 1989). Electrical stimulation of the amygdala in humans has adrenoceptors in this region. Thus, the effects of local been reported to elicit symptoms of anxiety disorders, infusion of desipramine in the long-term citalopram group including fear and anxiety (Halgren 1982). In animal could not be explained by alteration of function of the α2- models, a high incidence of panic-like attacks was elicited adrenoceptors. by administration of yohimbine, an α2-adrenoceptor antag- The results obtained in the LC are inconsistent with onist that is known to increase synaptic availability of previous data demonstrating that chronic treatment with noradrenaline. Thus, a decrease of noradrenergic activity paroxetine did not exhibit any effect on noradrenaline likely due to supersensitive autoreceptors in the BLA is transporters in the LC (Benmansour et al. 2004). The probably essential to prevent the pathology of anxiety present results suggest that the noradrenaline transporters in disorders. Accordingly, the mechanism underlying the the LC are slightly downregulated or decreased in density therapeutic properties, such as anti-panic and anti-phobic after long-term citalopram, while the stress response to the effects of citalopram (Murphy et al. 2000) in anxiety handling stimuli was relatively stable after long-term disorders, might occur through the inactivation of the citalopram. The noradrenaline transporters in the LC might noradrenaline neurotransmission attributed to the supersen- not participate significantly in the mechanism of the stress sitive α2-adrenoceptors that occur specifically in the BLA. response, but may be related to the attenuated basal Such regional specificity of the actions of citalopram will noradrenaline levels specifically in the LC. be further revealed by future investigations in other brain areas, such as cortical areas where neurotransmission has Effect of long-term citalopram on the clonidine-induced been demonstrated to be affected by this drug (Valentini decrease of noradrenaline et al. 2005). Regarding the pharmacological application of desipra-

The function of α2-adrenoceptors was investigated by the mine and clonidine, it should be noted that treatment with local application of clonidine into the two studied brain greater doses of both drugs, ten times the doses used in the regions. In the control group, the decrease of noradrenaline present study, no longer differed in their relative effects in in the LC by local application of clonidine was greater than the BLA and the LC between the citalopram-treated and that observed in the BLA. These effects of clonidine are control groups (data not shown). As higher doses of these consistent with studies demonstrating a higher density of drugs are probably less specific (Robin and Burnstein 1998;

α2-adrenoceptors in the LC compared to the amygdala Gould et al. 2006), only the lower doses were taken into (Scheinin et al. 1994). The range of decreases caused by local account in the present study. application of clonidine into the LC on noradrenaline in the dialysates was similar to previous reports (Mateo et al. 2001; Pudovkina et al. 2001). The particularly interesting finding of the present study Conclusions is that chronic treatment with citalopram enhanced the effect of clonidine in the BLA alone. This result indicates The present findings demonstrate that long-term adminis- that long-term citalopram caused super-sensitization of the tration of citalopram leads to a decrease in the basal activity

α2-adrenoceptors, specifically in the BLA. Enhanced of noradrenergic neurons in the BLA and the LC, with a sensitivity of α2-adrenoceptors could be involved in the delay that is consistent with the therapeutic onset of this attenuation of basal noradrenaline, as well as the abolish- drug in patients, such as in panic disorder. The anti-panic ment of the stress response of noradrenaline in this region. effect of citalopram might be due to the loss of the stress This mechanism is apparently not present in the LC. The response of noradrenaline in the BLA caused by super- present results are consistent with an electrophysiological sensitivity of α2-adrenoceptors in this region. The present study using brain slice preparations of the BLA from naive study also indicates that the modulation of noradrenergic animals, which indicated that activation of α2-adrenoceptors system elicited by citalopram is region-dependent. impaired the induction of long-term potentiation and depres- sion and reduced synaptic transmission (DeBock et al. 2003). Acknowledgments This work was supported by the Kurume Based on this report demonstrating that the reduced synaptic University School of Medicine. 80 Psychopharmacology (2007) 194:73–81

References other antidepressants on serotonin and norepinephrine transporters. Biol Psychiatry 59:408–414 Gracy KN, Pickel VM (1995) Comparative ultrastructural localization Baba T, Takeda H, Lee JW, Takada K, Shibuya T (1995) Effect of of the NMDAR1 glutamate receptor in the rat basolateral fluvoxamine, a new antidepressant drug, on monoamine dynam- amygdala and bed nucleus of the stria terminalis. J Comp Neurol ics in the rat cerebral limbic system and its pharmacological 362:71–85 characteristic. Nippon Yakurigaku Zasshi 105:177–185 Grant MM, Weiss JM (2001) Effects of chronic antidepressant drug Benmansour S, Altamirano AV, Jones DJ, Sanchez TA, Gould GG, administration and electroconvulsive shock on locus coeruleus Pardon MC, Morilak DA, Frazer A (2004) Regulation of the electrophysiologic activity. Biol Psychiatry 49:117–129 norepinephrine transporter by chronic administration of antide- Haddjeri N, de Montigny C, Blier P (1997) Modulation of the firing pressants. Biol Psychiatry 55:313–316 activity of noradrenergic neurones in the rat locus coeruleus by Boothman LJ, Mitchell SN, Sharp T (2006) Investigation of the SSRI the 5-hydroxtryptamine system. Br J Pharmacol 120:865–875 augmentation properties of 5-HT(2) receptor antagonists using in Halgren E (1982) Mental phenomena induced by stimulation in the vivo microdialysis. Neuropharmacology 50:726–732 limbic system. Hum Neurobiol 1:251–260 Bosker FJ, Cremers TI, Jongsma ME, Westerink BH, Wikstrom HV, den Harmer CJ, Mackay CE, Reid CB, Cowen PJ, Goodwin GM (2006) Boer JA (2001) Acute and chronic effects of citalopram on Antidepressant drug treatment modifies the neural processing of postsynaptic 5-hydroxytryptamine 1A receptor-mediated feedback: nonconscious threat cues. Biol Psychiatry 59:816–820 a microdialysis study in the amygdala. J Neurochem 76:1645–1653 Huang M, Ichiwaka J, Li Z, Dai J, Meltzer HY (2006) Augmentation Boyer PA, Skolnick P, Fossom LH (1998) Chronic administration of by citalopram of risperidone-induced monoamine release in rat imipramine and citalopram alters the expression of NMDA prefrontal cortex. Psychopharmacology (Berl) 185:274–281 receptor subunit mRNAs in mouse brain. A quantitative in situ Inoue T, Li XB, Abekawa T, Kitaichi Y, Izumi T, Nakagawa S, hybridization study. J Mol Neurosci 10:219–233 Koyama T (2004) Selective serotonin reuptake inhibitor reduces Brunello N, Blier P, Judd LL, Mendlewicz J, Nelson CJ, Souery D, Zohar conditioned fear through its effect in the amygdala. Eur J J, Racagni G (2003) Noradrenaline in mood and anxiety disorders: Pharmacol 497:311–316 basic and clinical studies. Int Clin Psychopharmacol 18:191–202 Invernizzi RW, Garattini S (2004) Role of presynaptic α2-adrenoceptors Coplan JD, Papp LA, Pine D, Martinez J, Cooper T, Rosenblum LA, in antidepressant action: recent findings from microdialysis studies. Klein DF, Gorman JM (1997) Clinical improvement with Prog Neuropsychopharmacol Biol Psychiatry 28:819–827 fluoxetine therapy and noradrenergic function in patients with Izumi T, Inoue T, Kitaichi Y, Nakagawa S, Koyama T (2006) Target panic disorder. Arch Gen Psychiatry 54:643–648 brain sites of the anxiolytic effect of citalopram, a selective David DJ, Bourin M, Jego G, Przybylski C, Jolliet P, Gardier AM serotonin reuptake inhibitor. Eur J Pharmacol 534:129–132 (2003) Effects of acute treatment with paroxetine, citalopram and Jongsma ME, Bosker FJ, Cremers TI, Westerink BH, den Boer JA venlafaxine in vivo on noradrenaline and serotonin outflow: a (2005) The effect of chronic selective serotonin reuptake microdialysis study in Swiss mice. Br J Pharmacol 140:1128–1136 inhibitor treatment on serotonin 1B receptor sensitivity and Debiec J, LeDoux JE (2006) Noradrenergic signaling in the amygdala HPA axis activity. Prog Neuropsychopharmacol Biol Psychiatry contributes to the reconsolidation of fear memory: treatment 29:738–744 implications for PTSD. Ann NY Acad Sci 1071:521–524 Kawahara Y, Kawahara H, Westerink BH (1999) Comparison of DeBock F, Kurz J, Azad SC, Parsons CG, Hapfelmeier G, effects of hypotension and handling stress on the release of Zieglgansberger W, Rammes G (2003) α2-adrenoreceptor activa- noradrenaline and dopamine in the locus coeruleus and medial tion inhibits LTP and LTD in the basolateral amygdala: prefrontal cortex of the rat. Naunyn Schmiedebergs Arch involvement of Gi/o-protein-mediated modulation of Ca2+- Pharmacol 360:42–49 channels and inwardly rectifying K+-channels in LTD. Eur J Kung MP, Choi SR, Hou C, Zhuang ZP, Foulon C, Kung HF (2004) Neurosci 17:1411–1424 Selective binding of 2-[125I]iodo-nisoxetine to norepinephrine Fallon JH (1981) Histochemical characterization of dopaminergic, transporters in the brain. Nucl Med Biol 31:533–541 noradrenergic and serotonergic projections to the amygdala. In: Laakso A, Palvimaki EP, Kuoppamaki M, Syvalahti E, Hietala J Ben-Ari Y (ed) The amygdaloid complex. Elsevier, Amsterdam, (1996) Chronic citalopram and fluoxetine treatments upregulate pp 175–183 5-HT2c receptors in the rat choroid plexus. Neuropsychophar- Fallon JH, Koziell DA, Moore RY (1978) Catecholamine innervation macology 15:143–151 of the basal forebrain. II. Amygdala, suprarhinal cortex and Mateo Y, Fernandez-Pastor B, Meana JJ (2001) Acute and chronic entorhinal cortex. J Comp Neurol 180:509–532 effects of desipramine and clorgyline on α2-adrenoceptors Ferry B, Magistretti PJ, Pralong E (1997) Noradrenaline modulates regulating noradrenergic transmission in the rat brain: a dual- glutamate-mediated neurotransmission in the rat basolateral probe microdialysis study. Br J Pharmacol 133:1362–1370 amygdala in vitro. Eur J Neurosci 9:1356–1364 McGaugh JL (2004) The amygdala modulates the consolidation of Ferry B, McGaugh JL (2000) Role of amygdala norepinephrine in memories of emotionally arousing experiences. Annu Rev mediating stress hormone regulation of memory storage. Acta Neurosci 27:1–28 Pharmacol Sin 21:481–493 McKittrick CR, Abercrombie ED (2007) Catecholamine mapping Fernandez-Pastor B, Mateo Y, Gomez-Urquijo S, Javier Meana J within nucleus accumbens: differences in basal and amphet- (2005) Characterization of noradrenaline release in the locus amine-stimulated efflux of norepinephrine and dopamine in shell coeruleus of freely moving awake rats by in vivo microdialysis. and core. J Neurochem 100:1247–1256 Psychopharmacology (Berl) 180:570–579 Millan MJ, Gobert A, Lejeune F, Newman-Tancredi A, Rivet JM, Fredricson Overo K (1982) Kinetics of citalopram in test animals; Auclair A, Peglion JL (2001) S33005, a novel ligand at both drug exposure in safety studies. Prog Neuropsychopharmacol serotonin and norepinephrine transporters: I. Receptor binding, Biol Psychiatry 6:297–309 electrophysiological, and neurochemical profile in comparison Gorman JM, Kent JM, Sullivan GM, Coplan JD (2000) Neuroanatomical with venlafaxine, reboxetine, citalopram, and clomipramine. J hypothesis of panic disorder, revised. Am J Psychiatry 157:493–505 Pharmacol Exp Ther 298:565–580 Gould GG, Altamirano AV, Javors MA, Frazer A (2006) A Millan MJ, Gobert A, Rivet JM, Adhumeau-Auclair A, Cussac D, comparison of the chronic treatment effects of venlafaxine and Newman-Tancredi A, Dekeyne A, Nicolas JP, Lejeune F (2000) Psychopharmacology (2007) 194:73–81 81

Mirtazapine enhances frontocortical dopaminergic and cortico- α2-adrenergic receptor subtype gene expression in rat brain. limbic adrenergic, but not serotonergic, transmission by blockade Brain Res Mol Brain Res 21:133–149 of α2-adrenergic and serotonin2C receptors: a comparison with Shekhar A, Sajdyk TJ, Gehlert DR, Rainnie DG (2003) The citalopram. Eur J Neurosci 12:1079–1095 amygdala, panic disorder, and cardiovascular responses. Ann Morris JS, Ohman A, Dolan RJ (1999) A subcortical pathway to the NY Acad Sci 985:308–325 right amygdala mediating “unseen” fear. Proc Natl Acad Sci Shiekhattar R, Aston-Jones G (1993) Sensory responsiveness of brain USA 96:1680–1685 noradrenergic neurons is modulated by endogenous brain Murphy TK, Bengtson MA, Tan JY, Carbonell E, Levin GM (2000) serotonin. Brain Res 623:72–76 Selective serotonin reuptake inhibitors in the treatment of Sodero AO, Valdomero A, Cuadra GR, Ramirez OA, Orsingher OA paediatric anxiety disorders: a review. Int Clin Psychopharmacol (2004) Locus coeruleus activity in perinatally protein-deprived 2:S47–S63 rats: effects of fluoxetine administration. Eur J Pharmacol Neophytou SI, Aspley S, Butler S, Beckett S, Marsden CA (2001) 503:35–42 Effects of lesioning noradrenergic neurones in the locus Sullivan GM, Coplan JD, Kent JM, Gorman JM (1999) The coeruleus on conditioned and unconditioned aversive behaviour noradrenergic system in pathological anxiety: a focus on panic in the rat. Prog Neuropsychopharmacol Biol Psychiatry 25: with relevance to generalized anxiety and phobias. Biol Psychi- 1307–1321 atry 46:1205–1218 Nutt DJ (1989) Altered central α2-adrenoceptor sensitivity in panic Svensson TH (2000) Brain noradrenaline and the mechanisms of disorder. Arch Gen Psychiatry 46:165–169 action of antidepressant drugs. Acta Psychiatr Scand Suppl Owens MJ, Morgan WN, Plott SJ, Nemeroff CB (1997) Neurotrans- 402:18–27 mitter receptor and transporter binding profile of antidepressants Szabo ST, de Montigny C, Blier P (1999) Modulation of noradren- and their metabolites. J Pharmacol Exp Ther 283:1305–1322 ergic neuronal firing by selective serotonin reuptake blockers. Br Page ME, Abercrombie ED (1997) An analysis of the effects of acute J Pharmacol 126:568–571 and chronic fluoxetine on extracellular norepinephrine in the rat Szabo ST, de Montigny C, Blier P (2000) Progressive attenuation of hippocampus during stress. Neuropsychopharmacology 16: the firing activity of locus coeruleus noradrenergic neurons by 419–425 sustained administration of selective serotonin reuptake inhib- Paxinos G, Watson C (1986) The rat brain in stereotaxic coordinates. itors. Int J Neuropsychopharmacol 3:1–11 Academic, Sydney Tanaka M, Yoshida M, Emoto H, Ishii H (2000) Noradrenaline Pollock BG (2001) Citalopram: a comprehensive review. Expert Opin systems in the hypothalamus, amygdala and locus coeruleus are Pharmacother 2:681–698 involved in the provocation of anxiety: basic studies. Eur J Pudovkina OL, Kawahara Y, de Vries J, Westerink BH (2001) The Pharmacol 405:397–406 release of noradrenaline in the locus coeruleus and prefrontal Thomas DN, Nutt DJ, Holman RB (1998) Sertraline, a selective cortex studied with dual-probe microdialysis. Brain Res 906: serotonin reuptake inhibitor modulates extracellular noradrena- 38–45 line in the rat frontal cortex. J Psychopharmacol 12:366–370 Rainnie DG (1999) Serotonergic modulation of neurotransmission in Tukel R, Bozkurt O, Polat A, Genc A, Atli H (2006) Clinical the rat basolateral amygdala. J Neurophysiol 82:69–85 predictors of response to pharmacotherapy with selective seroto- Ressler KJ, Nemeroff CB (2001) Role of norepinephrine in the nin reuptake inhibitors in obsessive-compulsive disorder. Psychi- pathophysiology of neuropsychiatric disorders. CNS Spectr atry Clin Neurosci 60:404–409 6:663–666 Uehara T, Sumiyoshi T, Matsuoka T, Tanaka K, Tsunoda M, Itoh H, Robin AL, Burnstein Y (1998) Selectivity of site of action and Kurachi M (2005) Enhancement of lactate metabolism in the systemic effects of topical alpha agonists. Curr Opin Ophthalmol basolateral amygdala by physical and psychological stress: role 9:30–33 of benzodiazepine receptors. Brain Res 1065:86–91 Rosenkranz JA, Grace AA (2001) Dopamine attenuates prefrontal Valentini V, Cacciapaglia F, Frau R, Di Chiara G (2005) Selective cortical suppression of sensory inputs to the basolateral amygdala serotonin reuptake blockade increases extracellular dopamine in of rats. J Neurosci 21:4090–4103 noradrenaline-rich isocortical but not prefrontal areas: depen- Seedat S, van Rheede van Oudtshoorn E, Muller JE, Mohr N, Stein DJ dence on serotonin-1A receptors and independence from norad- (2003) Reboxetine and citalopram in panic disorder: a single- renergic innervation. J Neurochem 93:371–382 blind, cross-over, flexible-dose pilot study. Int Clin Psychophar- Valentino RJ, Curtis AL, Parris DG, Wehby RG (1990) Antidepressant macol 18:279–284 actions on brain noradrenergic neurons. J Pharmacol Exp Ther Segal M (1979) Serotonergic innervation of the locus coeruleus from 253:833–840 the dorsal raphe and its action on responses to noxious stimuli. J Vermetten E, Bremner JD (2002) Circuits and systems in stress. I. Physiol 286:401–415 Preclinical studies. Depress Anxiety 15:126–147 Scheinin M, Lomasney JW, Hayden-Hixson DM, Schambra UB, Wade AG (1999) Antidepressants in panic disorder. Int Clin Caron MG, Lefkowitz RJ, Fremeau RT Jr (1994) Distribution of Psychopharmacol 14(Suppl 2):S13–S17