J Psychopharmacol OnlineFirst, published on November 25, 2009 as doi:10.1177/0269881109348178

Original Paper

Effects of acute and sustained administration of the catecholamine reuptake inhibitor nomifensine on the firing activity of monoaminergic neurons

Journal of Psychopharmacology 0(00) 1–13 ! The Author(s) 2009 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0269881109348178 jop.sagepub.com

Noam S Katz1,2, Bruno P Guiard1,3, Mostafa El Mansari1 and Pierre Blier1,2

Abstract Nomifensine potently inhibits the reuptake of norepinephrine and dopamine in vitro. It is one of few antidepressants with marked potency to block dopamine reuptake that has ever been used clinically. Acute and sustained administration of nomifensine was investigated on the firing of monoaminergic neurons to understand its mechanism of action. In vivo extracellular recordings of locus coeruleus, ventral tegmental area and dorsal raphe nucleus neurons were obtained from male Sprague-Dawley rats. The intravenous injection of nomifensine in the locus coeruleus and ventral tegmental area yielded ED50 values of 40  1 and 450  41 mg/kg, respectively, suggesting that nomifensine directly acted upon dopamine and norepinephrine neurons, since these values are proportional to its affinities for norepinephrine and dopamine transporters. There was no effect on 5-HT neurons. Nomifensine (5 mg/kg/day, subcutaneous, using minipumps) potently and significantly inhibited dopamine neuronal firing in the ventral tegmental area after 2 days, with recovery to normal after the 14-day treatment due to D2 autoreceptor desensitization. Norepinephrine neuronal firing in the locus coeruleus was significantly decreased after 2 and 14 days. A significant increase in dorsal raphe nucleus 5-HT neuronal firing was seen after a two-day regimen, and remained elevated after 14 days. Desensitization of the 5-HT1A receptor on 5-HT neurons of the dorsal raphe nucleus occurred after two days of nomifensine administration. Nomifensine likely treated depression by acting on dopamine, norepinephrine and 5-HT neurons, highlighting the importance of the functional connectivity between these three monoaminergic systems.

Keywords depression, dopamine, nomifensine, norepinephrine, serotonin

Introduction There is a great deal of interplay between monoaminergic systems, highlighted by reciprocal interactions linking the dopaminergic neurons of the ventral tegmental area (VTA), noradrenergic neurons of the locus coeruleus (LC) and serotonergic neurons of the dorsal raphe nucleus (DRN) (Guiard et al., 2008b). Activation of D2 receptors induces an increase in the firing rate of DRN 5-HT neurons (Martin-Ruiz et al., 2001), demonstrating the excitatory role of dopamine (DA) on these neurons. This is further evidenced by the fact that bath application of the D2/3 receptor agonist quinpirole induces a concentration-dependent membrane depolarization of 5-HT neurons in vitro, an effect that is blocked by the D2 receptor antagonist sulpiride (Aman et al., 2007). On the other hand, norepinephrine (NE) neurons of the LC have an excitatory effect on 5-HT neurons mediated by a direct pathway to a1-adrenoceptors on 5-HT neurons (Baraban and Aghajanian, 1980). Together, these data demonstrate that both dopaminergic and noradrenergic inputs to the DRN stimulate the neuronal activity of 5-HT neurons. Conversely, 5-HT neurons appear to exert an overall inhibitory influence on the firing rate of DA neurons in the VTA (De Deurwaerdere et al., 2004; Guiard et al., 2008b; Prisco et al., 1994), mainly by activating 5-HT2C receptors; (Di Giovanni et al., 2008). There is contradictory evidence regarding the effect of NE neuronal

projections from the LC to DA neurons of the VTA; however, lesioning the LC actually enhanced DA firing as well as burst activity, indicating the net effect of LC neurons on the VTA is inhibitory (Guiard et al., 2008b). Finally, 5-HT neurons are known to indirectly induce an inhibitory effect on NE neurons of the LC through 5-HT2A receptors activating GABAergic interneurons (Szabo and Blier, 2001a, 2002). Compelling evidence suggests that descending dopaminergic pathways also inhibit the firing rate of LC NE neurons through the activation of D2-like (Piercey et al., 1994) and a2-adrenergic receptors (Cedarbaum and Aghajanian, 1977; Elam et al., 1986; Guiard et al., 2008a). Until recently, the role of DA in the etiopathology and treatment of depression was largely ignored. This occurred 1 University of Ottawa Institute of Mental Health Research, Ottawa, ON, Canada. 2 University of Ottawa, Department of Cellular and Molecular Medicine, Ottawa, ON, Canada. 3 Universite´ Paris Sud, Faculte´ de Pharmacie, Paris, France.

Corresponding author: Pierre Blier, MD, PhD, University of Ottawa Institute of Mental Health Research, 1145 Carling Avenue, Ottawa, Ontario, Canada K1Z 7K4. Email: [email protected]

2 despite a dopaminergic theory of depression being proposed more than 30 years ago (Randrup et al., 1975), as well as many lines of evidence implicating that deficiencies in the dopaminergic system play a significant role in major depression (Dunlop and Nemeroff, 2007). Nomifensine is an isoquinoline antidepressant that was distinct from all other available antidepressants. Although it was removed from the market due to its hematopoietic toxicity, it was unique in its clinical properties. It is a potent reuptake inhibitor of both NE and DA, producing an increase in extracellular levels while only minimally affecting their release (McKillop and Bradford, 1981; Schacht and Heptner, 1974; Tatsumi et al., 1997). Although only minimal acute effects were found in the accumulation of 5-HT in response to nomifensine (Schacht and Heptner, 1974), the M1 metabolite of nomifensine did in fact induce a potent inhibition of 5-HT reuptake (Kruse et al., 1977). Nomifensine is the only antidepressant ever used routinely in the clinic with a marked potency to block DA reuptake, aside from amineptine (Bonnet et al., 1987). To our knowledge, no studies have been done on the sustained administration of a catecholamine reuptake inhibitor as potent as nomifensine, and its effect on monoaminergic systems. In the present study, the effects of acute administration of nomifensine on NE, DA and 5-HT neurons of the LC, VTA and DRN, respectively, were assessed in vivo. Furthermore, the effects of sustained administration of this dual reuptake inhibitor for two or 14 days on the firing properties of the aforementioned neurons were investigated. Due to the high degree of functional connectivity between these systems, it was deemed important to understand the mechanism of action of nomifensine, which was beneficial in the treatment depression in its brief period of use. This research endeavor is all the more relevant as several pharmaceutical firms have been developing triple reuptake inhibitors, drugs that simultaneously inhibit the reuptake of 5-HT, NE and DA (Chen and Skolnick, 2007).

Materials and methods Animals Adult male Sprague-Dawley rats (Charles River, SaintConstant, QC, Canada) weighing 260–320 g at the time of the experiments were used. Animals were housed two per cage at standard experimental conditions (12 h light/dark cycle, with lights on at 07 : 00; temperature 21  1 C, 40–50% relative humidity) with access to food and water ad libitum. All procedures were approved by the local Animal Care Committee and were in accordance with the guidelines set by the Canadian Council for Animal Care.

Sustained treatment Rats were anesthetized with an inhalant mixture of isofluorane and oxygen. Subcutaneously implanted osmotic minipumps (Alzet, Durect Corporation, Cupertino, CA) were preloaded with nomifensine (5 mg/kg/day), GBR 12909 (7.5 mg/kg/day), reboxetine (2.5 mg/kg/day) or 20% hydroxypropyl-beta-cyclodextrin (b-OH) which provided delivery of

Journal of Psychopharmacology 0(00) each drug for two or 14 days. Minipumps remained in situ throughout the recordings in order to ensure the steady-state levels of the drug were maintained, thereby mimicking clinical conditions.

In vivo electrophysiological recordings Single-unit, extracellular recordings of presumed monoaminergic neurons were performed in order to examine the firing rates of NE, DA and 5-HT neurons in the LC, VTA and DRN, respectively. Recordings were taken using singlebarreled glass micropipettes (Stoelting, Wood Dale, IL, USA) pulled on a Narashige pipette puller (Tokyo, Japan) and preloaded with fiberglass strands to promote filling with a 2 M NaCl solution. Pipette tips were broken to 1–3 mm, with impedances ranging from 2.5–5 MV. Rats were anesthetized with chloral hydrate (400 mg/kg i.p.) and mounted in a stereotaxic frame. Additional anesthesia (50–100 mg/kg i.p.) was given as necessary to maintain a full anesthetic state, characterized by the absence of response to a nociceptive paw or tail pinch. Stereotaxically defined coordinates were used to locate the appropriate neural regions and neurons were individually identified by their spike shape, duration and frequency. Electrodes were lowered using a hydraulic micropositioner (Kopf Instruments), and all neuronal activity was recorded in real-time using Spike2 software (Cambridge Electronic Design, Cambridge, UK), which was also used to analyze neurons offline. Body temperature was maintained at 37 C throughout the experiments using a thermistor-controlled heating pad (Seabrook Medical Instruments, SaintHyacinthe, Quebec, Canada). Prior to recordings, a catheter was inserted into the lateral tail vein for systemic injections of any drug required. For all dose-response curves, only one neuron was used from each animal.

Recording of LC NE neurons: The following coordinates were used in order to record from the LC (in mm from lambda): AP 1.0 to 1.2, L 1.0 to 1.3, V 5 to 7. NE neurons were identified by the following criteria: regular firing rate (1–5 Hz), a positive action-potential of long duration (0.8–1.2 ms), and a robust response to a nociceptive contralateral pinch of the hind paw, eliciting a volley of spikes followed by a brief quiescent period (Aghajanian and Vandermaelen, 1982a).

Recording of VTA DA neurons: The following coordinates were used for micropipette descent into the VTA (in mm from lambda): AP + 3.2 to 3.6, L 0.9 to 1.1, V 7 to 9. DA neurons were identified by the well-established in vivo properties of spontaneously firing neurons: biphasic spikes (often with an inflection or ‘notch’ in the rising phase) with a prominent negative phase, long spike duration (2–4 ms), irregular spontaneous firing rate (2–9 Hz) often demonstrating bursts of 3–10 spikes with decreasing amplitude within the burst (Grace and Bunney, 1984a). A criterion of duration of >1.1 ms from the beginning of the action potential to the negative trough was also used for identification of DA neurons (Ungless et al., 2004).

3

Katz et al.

Recording of DRN 5-HT neurons: In order to record 5-HT neurons in the DRN, a burr hole was drilled in order to descend the glass micropipette at the following coordinates (in mm from lambda): AP + 1 to 1.2, L 0, V 5 to 7. Presumed 5-HT neurons were identified by their characteristic slow, regular firing rate (0.5–2.5 Hz) and long duration action potential (2–5 ms) (Aghajanian and Vandermaelen, 1982b).

Firing rate and burst analysis DA and NE neurons: The bursting activity of both DA and NE neurons was determined using interspike interval (ISI) analysis as previously described (Grace and Bunney, 1984b). Briefly, the onset of a burst was defined as two spikes occurring 160 ms apart. 5-HT neurons: The characteristics for a distinct population of bursting 5-HT neurons have been known to exist for some time and were recently confirmed (Hajos et al., 2007). As such, this was the criterion used to measure bursting activity in 5-HT neurons of the DRN. Briefly, the onset of a spike was considered to be an ISI of 24 ms, with spikes occurring as a doublet or triplet (two or three spikes, respectively) of decreasing amplitudes. Whole-blood 5-HT measurement In order to determine if nomifensine (or a metabolite) resulted in significant 5-HT reuptake inhibition, blood samples from control and 14-day treated animals were collected following recording, and frozen in tubes containing 7.5 mg/ml ascorbic acid. Whole-blood 5-HT levels were measured by highperformance liquid chromatography using a method described previously (Blier et al., 2007). As blood platelets lack the ability to synthesize 5-HT, any 5-HT present in these cells has been taken up by the 5-HT transporters located on these cells. Following blockade of this reuptake process, there is a depletion of whole-blood 5-HT levels, due to the inability of 5-HT to enter platelets. As whole-blood 5-HT levels are also decreased by selective serotonin reuptake inhibitors (SSRIs) such as paroxetine (Blier et al., 2007; Petersen et al., 1978) which are known to affect the neuronal serotonin transporter, this method provides an indirect measurement of neuronal 5-HT reuptake activity.

Drugs Nomifensine, GBR 12909 and DSP-4 (Sigma-Aldrich Canada, Oakville, ON, Canada) and paliperidone (provided by Janssen Pharmaceutica; Titusville, NJ) were dissolved in a 20% (2 g/10 ml distilled H2O) solution of hydroxypropylbeta-cyclodextrin, 97% (b-OH), and sonicated until completely dissolved. The D2 receptor agonist apomorphine hydrochloride, the D2 receptor antagonist haloperidol, the a2-adrenoceptor antagonist idazoxan hydrochloride, the NE reuptake inhibitor reboxetine mesylate hydrate, the 5-HT autoreceptor agonist lysergic acid diethylamide (LSD) and

the 5-HT1A receptor antagonist WAY100635 (SigmaAldrich), were all dissolved in distilled H2O. Escitalopram provided by Lundbeck (Lundbeck, Copenhagen, Denmark) was dissolved in 0.9% NaCl.

Statistical analysis All data are reported as mean values  SEM obtained from a single neuron for the following parameters: firing rate expressed in Hz (number of spikes/second), percentage of total spikes in 60-s intervals occurring as bursts (% of spikes as bursts), percentage of neurons displaying any burst activity (% neurons in burst mode), the number of spikes occurring per burst, and the number of cells found in each electrode descent (number cells/track). For the construction of dose-response curves, the assessment of neuronal responsiveness was taken as the percent of baseline firing rate measured 60 s after the systemic administration of the drug. A two-way analysis of variance (ANOVA) was conducted for the firing rate of each monoaminergic neuron population, with ‘drug treatment’ and ‘duration of treatment’ used as factors. The Bonferroni corrected post-hoc analysis was conducted when significant ANOVA results were obtained. A one-way ANOVA was conducted in order to determine significant differences in the burst activity and the number of cells per track. The drug treatment was used as the factor in this case, and separate ANOVAs were run for the 2-day and 14-day treatment groups. Again, the Bonferroni corrected post-hoc analysis was conducted when significant ANOVA results were obtained. All statistical analyses were completed using SigmaStat software (Systat, Chicago, IL, USA), aside from dose-response curves which were completed using GraphPad Prism 5.0 (GraphPad Software Inc, La Jolla, CA).

Results Acute dose-response curves in response to nomifensine In order to assess the potency of nomifensine to inhibit NE and DA neurons in vivo, acute dose-response curves (i.v.) were constructed in the LC and VTA, respectively. As previously reported, the in vitro receptor affinities for nomifensine on the NE, DA and 5-HT transporters are 15.6  0.4 nM, 56  3 nM and 1010  30 nM, respectively (Tatsumi et al., 1997), indicating the high affinity for nomifensine at both the NE and DA transporter sites. In agreement with these data, nomifensine inhibited the NE neurons in the LC with an ED50 of 40  1 mg/kg, and a dose of 100 mg/kg was required to produce a complete, lasting inhibition in all neurons (ED100; Figure 1A, 1B). In the VTA, the ED50 was much greater at 450  41 mg/kg, with an extremely high dose of 10 mg/kg required to induce a nearly complete inhibition (Figure 1C, 1D). The nomifensine-induced inhibition was reversed with 1 mg/kg of the a2-adrenoceptor antagonist idazoxan for NE neurons, and 200 mg/kg of the D2 receptor antagonist haloperidol for DA neurons (Figure 1A, 1C). As expected, due to the low affinity of nomifensine for the 5-HT transporter, acute administration of nomifensine produced no

4

Journal of Psychopharmacology 0(00)

Idazoxan (1 mg/kg)

Nomifensine (1 mg/kg)

Idazoxan Idazoxan (1 mg/kg) (1 mg/kg)

(A)

(B) 100 % Inhibition of LC NE neurons

Spikes/10sec

16

8

50 25 0

0

0 120 s

300 Nomifensine (10 mg/kg)

(C)

600

-1

900

0

1

Log [nomifensine] (D)

Haloperidol (0.2 mg/kg)

40

100 % Inhibition of VTA DA neurons

Spikes/10sec

75

20

0 120 s

200

400

600

75 50 25 0 –1.5 –1.0 –0.5

0.0

0.5

1.0

1.5

Log [nomifensine]

Figure 1. Acute dose-response curves of intravenous (i.v.) administration of nomifensine in norepinephrine (NE) and dopamine (DA) neurons. (A) Representative histogram of the firing rate of a NE neuron in the locos coeruleus (LC). Each bin represents the number of spikes/10 s. Arrows indicate the compounds administered and the time at which the injection of the specified doses was completed. (B) Dose-response curve of NE neurons in the LC in response to i.v. administration of nomifensine. (C) Representative histogram of the firing rate of a DA neuron in the ventral tegmental area (VTA). (D) Dose-response curve of DA neurons in the VTA in response to i.v. administration of nomifensine.

significant change in the firing rate of DRN 5-HT neurons (see Figure 9B, 5 mg/kg, n ¼ 5).

Sustained administration of nomifensine, GBR 12909 and reboxetine Sustained administration was examined in all three monoaminergic neuronal types using subcutaneously implanted osmotic minipumps for two or 14 days, delivering 5 mg/kg of nomifensine per day. Due to nomifensine being both a NE and DA reuptake inhibitor, the effects of sustained administration of GBR 12909 – a selective DA reuptake inhibitor – and reboxetine – a selective NE reuptake inhibitor – were also examined. As the effect of sustained administration of reboxetine had previously been studied in both the DRN and LC (Szabo and Blier, 2001b), only the response to the drug in the VTA was examined. The effect of sustained

administration of GBR 12909 was examined in all three neuronal populations.

Ventral tegmental area A two-way ANOVA showed significant main effects for both drug treatment (F(3,266) ¼ 8.78, p < 0.001), as well as treatment duration (F(1,266) ¼ 9.81, p < 0.01), while also eliciting a significant drug  duration interaction (F(3,266) ¼ 4.36, p < 0.01). For the drug treatment, post-hoc analysis revealed significant differences between the control group and all three treatment groups (reboxetine and GBR 12909 versus controls: p < 0.001, nomifensine versus controls: p < 0.01), while none of the treatment groups significantly differed from one another. A significant decrease of 39% was seen following the two-day administration of nomifensine in the firing rate of DA neurons in the VTA relative to controls (p < 0.001).

5

Katz et al.

Control Nomifensine (5 mg/kg) GBR 12909 (7.5 mg/kg) Reboxetine (2.5 mg/kg)

5

Firing rate of VTA DA neurons (Hz ± SEM)

4.5 4 **

3.5 3

***

***

2.5 2 1.5 1 0.5

40 (4)

33 (5)

30 (3)

33 (3)

32 (4)

33 (4)

33 (4)

0 Control 2-day regimens

14-day regimens

Figure 2. Average firing rate of dopamine (DA) neurons in the ventral tegmental area (VTA). A significant decrease in firing rate of dopamine neurons was found in the group treated with nomifensine for two days, with a complete recovery to control levels after 14 days. Both reboxetine and GBR 12909 produced a significant decrease in firing after two days which no longer remained significant at 14 days. The number at the bottom of each column represents the number of neurons. Number of rats is indicated in parentheses. **p < 0.01, ***p < 0.001.

However, a complete recovery in the firing rate of these neurons to control levels was observed after 14-day administration (p > 0.05; Figure 2). Burst analysis revealed a 52% decrease in the percentage of total spikes occurring as bursts in the 2-day treatment group relative to controls (p < 0.05), again recovering to control levels after the 14-day treatment (p > 0.05; Figure 3). The previously established effective regimen of reboxetine (2.5 mg/kg/day) on; NE; neuronal firing (Szabo and Blier, 2001b) was used in order to examine the effect of the drug on DA neurons in the VTA. Following the two-day treatment, a significant decrease of 31% in the firing rate of these neurons was found relative to controls (p < 0.01; Figure 2). Following the 14-day treatment with reboxetine, the average firing rate of these neurons was decreased by 22% relative to controls, thus no longer remaining significantly different from control levels (p > 0.05). A decrease of 39% and 56% was found in the percentage of spikes occurring as bursts in both two- and 14-day treated animals, respectively; however, only the 14-day treated group was significantly different from control levels (p < 0.05; Figure 3). No difference was found in the number of spikes per burst (data not shown), nor the number of cells found per track (data not shown) in any of the groups.

As the ED50 of GBR 12909 to inhibit DA neurons of the VTA had previously been found to be 7.5 mg/kg (Einhorn et al., 1988), this dose was first tested. The administration of GBR 12909 for two days produced a 26% decrease in the firing rate of DA neurons in the VTA (p < 0.01) recovering to a decrease of only 22% after 14 days, no longer being significantly different to control levels (p > 0.05; Figure 2). When coupled with the 31% decrease seen in firing following the two-day treatment of reboxetine, it was determined that 7.5 mg/kg/day was an effective dose of GBR 12909 for sustained treatment, therefore this dose was used for all subsequent experiments. A significant decrease of 66% was also found in the percentage of total spikes occurring as bursts after the two-day treatment relative to controls (p < 0.01), which did not remain significant after treatment for 14 days (p > 0.05; Figure 3). None of the other burst parameters, nor the number of cells found per track were significantly different from control levels (data not shown).

Assessment of D2 autoreceptor sensitivity: As a complete recovery in firing rate was observed following the 14-day treatment with nomifensine, the sensitivity of the D2 autoreceptor was tested in the VTA using the D2 agonist

6

Journal of Psychopharmacology 0(00)

Control Nomifensine (5 mg/kg) GBR 12909 (7.5 mg/kg) Reboxetine (2.5 mg/kg)

% Spikes occurring as bursts in VTA DA neurons

35

30

25

20 *

*

15 ** 1.0

5 40

33

30

33

32

33

33

0 Control

2-day regimens

14-day regimens

apomorphine. In control animals, a complete attenuation of DA neuronal firing was achieved with a dose of 30 mg/kg, with an ED50 of 12  0.5 mg/kg (Figure 4). Following the 14-day treatment with nomifensine, the ED50 increased to 21  3 mg/kg, with administration of a more than threefold increase in the ED100 of control animals to 100 mg/kg. In addition, the slopes of the dose-response curves were significantly different at 2.4  0.4 versus 0.7  0.1 for controls and the 14-day nomifensine group, respectively; F(1,15) ¼ 15.02, p ¼ 0.001). This rightward shift of the dose-response curve indicated that a significant desensitization of the D2 autoreceptor occurred with sustained treatment of nomifensine.

% Inhibition of VTA DA neurons

Figure 3. Effect of nomifensine, GBR 12909 and reboxetine administration (two day and 14 day) on burst activity of ventral tegmental area (VTA) dopamine (DA) neurons. Data are expressed as means  S.E.M. of the percentage of spikes occurring in burst. The number at the bottom of each column represents the number of neurons * indicates a significant effect in comparison to the control group, **p < 0.01; ***p < 0.001.

100

50 Control 14 days

0 0

10

20

30

40

50

60

70

80

90 100

Apomorphine (µg/kg; iv)

Locus coeruleus A two-way ANOVA yielded a significant main effect of drug treatment (F(2, 269) ¼ 85.32, p < 0.001) as well as a significant drug  duration interaction (F(2, 269) ¼ 4.358, p < 0.01). Posthoc analyses revealed significant differences between nomifensine and both the control group and GBR 12909 (p < 0.001), with no difference between GBR 12909 and controls. A significant decrease of 71% in the firing rate of NE neurons in the LC was found following the two-day treatment with nomifensine and remained decreased by 62% after 14 days, relative to controls (p < 0.001 for both groups; Figure 5). Accordingly, no recovery in the firing rate of NE LC neurons occurred following sustained treatment with nomifensine, contrary to what was observed in DA cells of the VTA using the same regimen. No significant differences

Figure 4. Relationship between the degree of suppression of ventral tegmental area (VTA) firing activity and doses of apomorphine administered intravenously in controls and rats treated with nomifensine for 14 days. Inhibition of the neurons by apomorphine was reversed with the D2 receptor antagonist haloperidol. The rightward shift of the doseresponse curve clearly indicates a desensitization of the D2 autoreceptor has occurred. Due to steepness of the curves, the confidence intervals were removed for the sake of clarity of the graph Controls: n ¼ 11, 14-day treated: n ¼ 8.

were observed shown), while found in the (p < 0.05; data

for any of the bursting parameters (data not only half the number of cells per track was 14-day treated group relative to controls not shown).

7

Katz et al.

Control Nomifensine (5 mg/kg) GBR 12909 (7.5 mg/kg)

###

Firing rate of LC NE neurons (Hz ± SEM)

1.8 1.6

### 1.4 *** 1.2 1 0.8 *** 0.6

***

0.4 0.2 0

95 (4) Control

43 (4)

52 (4)

2-day regimens

48 (4)

46 (4)

14-day regimens

Figure 5. Average firing rate of norepinephrine (NE) neurons in the locus coeruleus (LC). Treatment with nomifensine potently inhibited firing of neurons relative to controls after both two-day and 14-day treatment. GBR 12909 did not affect the firing rate of NE neurons after 2-day treatment, but produced a potent decrease in firing rate relative to controls after 14 days. The number at the bottom of each column represents the number of neurons. Number of rats is indicated in parentheses. **p < 0.01, ***p < 0.001.

Following treatment with GBR 12909 for two days, no significant deviation from control levels for any of the measured parameters in NE neurons in the LC were observed. After 14-day treatment with GBR 12909, a 26% decrease was seen in the firing rate of LC NE neurons relative to controls (p < 0.001; Figure 5). Despite this decrease, the firing rate of the LC NE neurons remained significantly different from the firing rate of nomifensine-treated animals (p < 0.001) within both the two- and 14-day groups (i.e. nomifensine  two days versus GBR 12909  2 days), as well as a significant difference within the GBR 12909 treated groups (p < 0.001). None of the burst parameters significantly differed from control levels (data not shown), while a 45% decrease in the number of cells found per track was observed in the 14-day GBR 12909-treated group relative to controls (p < 0.05; data not shown).

Dorsal raphe nucleus A two-way ANOVA showed a significant main effect of drug treatment (F(2, 441) ¼ 19.47, p < 0.001), as well as a significant drug  duration interaction (F(2, 449) ¼ 6.715, p < 0.01). Posthoc analyses revealed significant differences between nomifensine and both the control group and GBR 12909 (p < 0.001), with no difference found between GBR and controls. Contrary to the other regions examined, a significant increase in the firing rate of 5-HT neurons in the DRN by

50% was observed following the two-day nomifensine treatment (p < 0.001) and remained elevated by 32% after 14 days (p < 0.01) relative to controls (Figure 6). None of the burst analyses elicited any significant differences within the two-day or 14-day-treated groups (data not shown). No significant difference was observed in the number of cells per track in any group (data not shown). Following GBR 12909 administration for 2 days, no significant deviation from control levels for any of the measured parameters in 5-HT neurons in the DRN occurred. A 23% increase in the firing rate of DRN 5-HT neurons in the 14-day-treated GBR 12909 group was observed, but this increase did not reach significance relative to controls (p ¼ 0.10). No significant differences were detected for any of the burst parameters (data not shown) or the number of cells per track (data not shown) relative to controls.

Assessment of 5-HT1A autoreceptor sensitivity: As a significant increase in firing rate was observed following the two-day nomifensine regimen, the sensitivity of the 5-HT1A autoreceptor was tested in the DRN using the 5-HT autoreceptor agonist lysergic acid diethylamide (LSD). LSD is a reliable indicator of 5-HT1A autoreceptor sensitivity as it produces the same effect when applied systemically, or locally onto 5-HT neurons via iontophoresis, unlike other 5-HT1A receptor agonists (Blier and De Montigny, 1987). In control

8

Journal of Psychopharmacology 0(00)

Control Nomifensine (5 mg/kg) GBR 12909 (7.5 mg/kg)

2

***

Firing rate of DRN 5-HT neurons (Hz ± SEM)

1.8 ** 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0

70 (5)

80 (5)

70 (5)

82 (3)

65 (4)

Control 2-day regimens

14-day regimens

Figure 6. Average firing rate of 5-HT neurons in the dorsal raphe nucleus. A significant increase in the firing rate of 5-HT neurons was found in the group treated with nomifensine for two and 14 days. GBR 12909 did not significantly alter firing rates. The number at the bottom of each column represents the number of neurons. Number of rats is indicated in parentheses. **p < 0.01, ***p < 0.001.

% Inhibition of DRN 5-HT neurons

100 80 60 40

Control 2 days

20 0 0

50

100

150

LSD (µg/kg; iv)

Figure 7. Relationship between the degree of suppression of dorsal raphe nucleus firing activity and doses of lysergic acid diethylamide (LSD) administered intravenously in controls and rats treated with nomifensine for two days. The LSD-induced inhibition was reversed with the 5-HT1A receptor antagonist WAY100635. The rightward shift of the dose-response curve and the large differences in the slopes of the curves clearly indicates a desensitization of the 5-HT1A autoreceptor has occurred. Due to steepness of the curves, the confidence intervals were removed for the sake of clarity of the graph. Controls: n ¼ 10, 2-day treated: n ¼ 8.

animals, a complete attenuation of 5-HT neuronal firing was achieved with a dose of 10 mg/kg, with an ED50 of 6  0.2 mg/kg (Figure 7). Following the two-day treatment with nomifensine, the ED50 increased to 36  2 mg/kg, while administration of an increased dose of 150 mg/kg unreliably

induced a complete suppression in firing. In addition, the slopes of the dose-response curves were significantly different at 14.2  2.7 versus 0.5  0.1 for controls and the 14-day nomifensine group, respectively (F(1,14) ¼ 22.36, p ¼ 0.001). This rightward shift of the dose-response curve coupled to

9

Katz et al.

**

1.8 **

Firing rate of 5-HT Neurons (spikes/s)

1.5

1.2

0.9

0.6

0.3

0

57 (5)

34 (3)

47 (3)

32 (3)

50 (4)

Control

Nomifensine

Nomifensine + paliperidone

DSP-4

DSP-4 + nomifensine

Figure 8. Histograms of the firing rates of 5-HT neurons in control rats and in rats given nomifensine, nomifensine and paliperidone, the norepinephrine (NE) neurotoxin DSP-4, and nomifensine in NE lesioned rats. The number at the bottom of each column represents the number of neurons. Number of rats is indicated in parentheses. **p < 0.001.

the significant difference between the two slopes indicated that a desensitization of the 5-HT1A autoreceptor occurred after only two days of the nomifensine regimen.

Whole-blood 5-HT levels Analysis of whole-blood 5-HT levels indicated a significant increase in the 14-day-treated group (1743  113 ng/ml; n ¼ 11) relative to controls (1217  67 ng/ml; n ¼ 9; p < 0.001). This indicates that nomifensine did not inhibit reuptake, as would be expected if levels of the M1 metabolite reached effective levels.

Catecholaminergic regulation of 5-HT neuronal firing rate in the DRN In order to determine a possible involvement of LC NE neurons in the effect of nomifensine on the firing activity of 5-HT neurons, NE neurons were lesioned using the neurotoxin DSP-4. While DSP-4 has no effect by itself, it; abolished the increase in the spontaneous firing activity of DRN 5-HT neurons in nomifensine-treated rats (Figure 8), indicating an involvement of the NE system in the enhancing effect of nomifensine treatment on 5-HT neuronal firing activity. Previous studies have shown that activation of D2 receptors has a stimulating effect on 5-HT neurons (Aman et al., 2007; Chernoloz et al., 2009a). Consequently, this effect might also underlie the increase of 5-HT neurons firing following two days of nomifensine administration. To test this hypothesis, paliperidone, a drug that by itself has no effect on 5-HT

neurons firing, was used to block D2 receptors (Dremencov et al., 2007). The administration of paliperidone to rats treated with nomifensine for two days did not block the increase in firing activity of 5-HT neurons (Figure 8), suggesting that D2 receptor activation cannot explain per se the increase of firing produced by nomifensine. In order to further investigate the interactions between catecholamines and 5-HT in modulating 5-HT neuronal firing, dose-response curves were constructed in rats treated with the SSRI escitalopram, the NE reuptake inhibitor reboxetine, the DA reuptake inhibitor GBR 12909, or the NE/DA reuptake inhibitor nomifensine. A two-way ANOVA for the percentage of baseline firing rate of DRN 5-HT neurons indicated an overall significant inhibitory effect of pre-treatment (F(3,28) ¼ 16.4, p < 0.001) and dose of escitalopram (F(2,28) ¼ 10.5, p < 0.001) factors, while no significant interaction between the independent variables was detected (F(6,28) ¼ 0.4, p > 0.05). At each dose tested, the inhibitory effects of escitalopram on DRN 5-HT basal firing rates were not significantly different from those obtained in rats pre-treated with the selective DA reuptake inhibitor GBR 12909, or the selective NE reuptake inhibitor reboxetine relative to vehicle (Figure 9). However, in rats administered with the dual NE/DA reuptake inhibitor nomifensine, a significant attenuation of the effect of escitalopram was detected (Figure 9B and 9C). Catecholamine reuptake inhibitors did not modify the spontaneous firing activity of DRN 5-HT neurons when given alone (before and after the administration of GBR 12909: 1.3  0.4 Hz versus 1.2  0.4, p > 0.05; before and after the administration of reboxetine: 1.5  0.3 Hz versus 1.3  0.2, p > 0.05; before and after the administration of nomifensine: 1.5  0.2 Hz versus

10

Journal of Psychopharmacology 0(00)

(A)

Escitalopram

Spikes/10 sec.

100

WAY100635 (100 µg/kg; iv)

200 µg/kg; iv

20

0 Escitalopram

Spikes/10 sec.

(B)

Nomifensine

100

200

300

WAY100635 400 (100 µg/kg; iv)

20

0 60 sec.

(C)

100

Vehicle + escitalopram

***

Reboxetine + escitalopram Nomifensine + escitalopram

Percent basal firing rate of DR 5-HT neurons

***

GBR 12909 + escitalopram

75 *** 50

25

0 100

200

300

Cumulative dose of escitalopram (µg/kg; iv)

Figure 9. Effect of single- or dual-acting catecholaminergic reuptake inhibitors on escitalopram-induced decrease in dorsal raphe nucleus (DRN) 5-HT neuronal activity. A, B: Examples of integrated firing histograms showing the effects of cumulative intravenous (i.v.) doses of escitalopram on the spontaneous activity of DRN 5-HT neurons in presence of vehicle (A), or of the dopamine/norepinephrine reuptake inhibitor nomifensine (5 mg/kg; i.v.) (B). The arrows indicate the compounds administered and the time at which the injection of the specified doses was completed. The escitalopraminduced inhibition of firing rate was reversed with the 5-HT1A antagonist WAY100635. (C) Symbols represent the mean (SEM) of percent of baseline firing rate of DRN 5-HT neurons observed at each dose of escitalopram after the administration of vehicle (n ¼ 9); GBR 12909 (n ¼ 5), reboxetine (n ¼ 5) and nomifensine (n ¼ 5). These means were calculated on the 60 s period preceding each drug administration. ***p < 0.001.

1.7  0.3, p > 0.05). These results put into evidence that an elevation of both catecholamines in the DRN was a prerequisite to counterbalance the inhibitory action of the 5-HT1A autoreceptor, exerted indirectly via 5-HT reuptake inhibition.

Discussion The results of the present study showed that sustained administration of the NE and DA reuptake inhibitor nomifensine

induced robust alterations in the firing properties of all three types of monoaminergic neurons. Acute, i.v. administration of nomifensine produced a robust inhibition in the firing rate of both NE and DA neurons of the LC and VTA, respectively. This suggests that nomifensine acted directly on both the NE transporter (NET) and DA transporter (DAT) in vivo. Additionally, the lack of acute effect of nomifensine on 5-HT neurons of the DRN again correlated to the low affinity of nomifensine in vitro for the 5-HT

Katz et al. transporter, indicating that this drug does not directly affect these neurons. Nomifensine markedly decreased the firing rate of NE neurons in the LC after two days with no recovery in the firing rate of these neurons after 14 days, similar to previous results found with the norepinephrine reuptake inhibitor reboxetine. The absence of a recovery in the firing rate of these neurons is likely due to a lack of desensitization of somatodendritic a2-adrenoceptors, which are known to inhibit NE neuronal firing (Szabo and Blier, 2001b). The absence of recovery of firing of NE neurons indicates that levels of nomifensine towards the end of sustained administration could not explain apparent adaptive changes. Furthermore, nomifensine significantly decreased the number of spontaneously active NE cells found in the LC, although it had little effect on the burst-firing of these neurons. The lack of recovery in the firing rate of these NE neurons is dissimilar to another antidepressant, bupropion, which also has a dual mechanism affecting both NE and DA. Previous studies have found that bupropion attenuates NE neuronal firing in the LC after two days, similar to other norepinephrine reuptake inhibitors (NRIs), such as desipramine and reboxetine (Szabo and Blier, 2001b; Szabo et al., 2000), although unlike what is seen with these drugs, the firing rate of these neurons recovers to normal after a 14-day treatment (El Mansari et al., 2008). This may in fact be due to the different mechanism of action of bupropion relative to nomifensine, as bupropion has been postulated to be a NE releaser rather than a reuptake inhibitor (Dong and Blier, 2001). Similar to NE neurons, the firing rate of DA neurons in the VTA was markedly decreased with the short-term nomifensine regimen; however, unlike for NE neurons, there was also a significant decrease in the percentage of the total burst activity. Contrary to NE neurons, both the firing rate and burst activity of the DA neurons completely recovered to control levels following the 14-day treatment. Through assessing the sensitivity of the somatodendritic D2 receptor, the dopamine autoreceptor controlling firing activity, it was found that the recovery of the firing rate of these neurons was due to the desensitization of this autoreceptor. This is similar to what was previously described by Pitts et al. (1995) after repeated intraperitoneal administration of the D2-like receptor agonist quinpirole for 14 days. Nomifensine induced a significant increase in the firing rate of 5-HT neurons after two days, which remained significantly elevated after the 14-day regimen. It is well documented that increases in synaptically available NE and DA cause an excitation of DRN 5-HT neurons via a1-adrenoceptors (Baraban and Aghajanian, 1980) and D2-receptors (Aman et al., 2007; Martin-Ruiz et al., 2001), respectively. Although it has been previously reported that acute administration of the NRI reboxetine increases 5-HT firing (Linner et al., 2004), sustained treatment with reboxetine does not alter 5-HT neuronal firing rates (Szabo and Blier, 2001b). Regardless, the increase in 5-HT firing following nomifensine administration indicates that dopaminergic alterations likely contribute to increasing 5-HT neuronal firing. Additionally, there was a desensitization of 5-HT1A autoreceptors on DRN 5-HT neurons after only the two-day nomifensine regimen as observed by the blunted response of 5-HT neurons to

11 systemic; injection of LSD, similar to what was found with a two-day bupropion administration (El Mansari et al., 2008). This desensitization generally occurs after sustained SSRI administration, but only after a much longer period of 14 days. Again, this may occur in such a short timeframe with nomifensine due to the excitation input that 5-HT neurons receive via the activation of the postsynaptic a1-adrenergic and D2 heteroreceptors. Furthermore, activation of the a1-adrenergic heteroreceptors likely plays a role in this desensitization, as activation of this receptor has been shown to alter 5-HT1A receptor responsiveness mediated by the intracellular phosphoinositide pathway (Hensler et al., 1996). In contrast to a robust decrease in whole-blood 5-HT levels seen with SSRI administration, a significant increase was observed following treatment with nomifensine for 14 days, thus suggesting that there is no inhibition of 5-HT reuptake. Therefore, despite the fact that the M1 metabolite of nomifensine is a potent 5-HT reuptake inhibitor, this metabolite was thus unlikely to have reached significant levels in plasma and consequently the rat brain. There was a significant increase in 5-HT neuronal firing after two days, whereas a decrease would have been expected in the presence of 5-HT reuptake inhibition. It remains to be determined whether this change is due to increased reuptake activity or to increased b-adrenoceptor stimulation, which increases circulating levels of tryptophan (Claustre et al., 2008). On the other hand, there is also the possibility that this metabolite may be produced in humans but not in rats. The increase in 5-HT neuronal firing following two days of nomifensine administration implies a synergistic effect of the dual action of this drug since GBR 12909 and selective NE reuptake inhibitors (Szabo and Blier, 2001b) produced no such effect. However, the elevation of 5-HT neurons firing following two days of nomifensine administration was abolished by lesion of NE neurons, but not by the administration of the D2 receptor antagonist paliperidone. This indicates on the one hand that NE plays a significant role in this increase, and on the other hand enhanced DA levels may also play a role but may be not sufficient alone to induce such an increase. For instance, sustained administration of the D2/3 receptor agonist pramipexole for two days does not alter the firing of 5-HT neurons, but does so after 14 days (Chernoloz et al., 2009a). In contrast, the stimulatory action of the D2/; 5-HT1A partial agonist aripiprazole on the firing activity of 5-HT neurons after two days of administration is reversed by paliperidone (Chernoloz et al., 2009b). To further explore the mechanism of action of DA and NE in the DRN, the effects of single- or dual-acting NE/DA reuptake inhibitors such as reboxetine, GBR 12909 or nomifensine were examined on SSRI (escitalopram)-induced inhibition of DRN 5-HT neuronal activity. It was found that neither pre-treatment with the NRI reboxetine nor pre-treatment with the dopamine reuptake inhibitor GBR 12909 at doses previously shown to elevate extracellular NE (Page and Lucki, 2002) or DA levels (Baumann et al., 1994) attenuated the escitalopram-induced decrease in DRN 5-HT firing rates. However, when NE and DA levels were simultaneously increased by systemic administration of the dualacting reuptake inhibitor nomifensine, an upward shift of the dose–response curve of escitalopram was observed,

12 demonstrating that both catecholamines were required to counteract the inhibitory electrophysiological effect of escitalopram. These results suggested that dual reuptake inhibitors such as nomifensine enhanced brain NE and DA transmission, which in turn can activate 5-HT neurons. Since the DRN receives noradrenergic and dopaminergic innervations and expresses both the NET (Javitch et al., 1985), and DAT (Fujita et al., 1994), it can be anticipated that the acute systemic administration of nomifensine does effectively increase catecholamine levels around cell bodies of 5-HT neurons, which would potentiate their neurochemical effects at nerve terminals. Following the two-day GBR 12909 treatment, a significant decrease was observed in both the firing rate and total burst activity of VTA DA neurons, as was expected due to the high affinity of GBR 12909 to block the DAT. Neither of these parameters remained significantly different from control levels after 14 days, similarly to nomifensine. No significant deviations from control levels for any of the measured parameters in LC NE neurons or DRN 5-HT neurons were observed in animals given GBR 12909 for two days. This indicates that in the short term, the primary site of action of GBR 12909 is solely in the DA neurons, and produces no effect in either of the other two monoaminergic systems. After the 14-day treatment with GBR 12909, a significant decrease was observed in the firing rate of LC NE neurons. A rather unexpected result of this study was the decrease in DA neuronal firing following the two-day treatment with the NRI reboxetine, because in vitro application of NE has been found to be excitatory on midbrain DA neurons inducing a membrane depolarization via a1-adrenoceptors. In a recent study by Guiard et al. (2008b), it was found that selective lesioning of LC NE neurons increased the firing rate of VTA DA neurons by 70%, indicating that the noradrenergic projections from the LC to the VTA exert a net inhibitory effect. This is strengthened by the fact that it has been demonstrated that local application of NE on VTA DA neurons generally induces an inhibition of the firing rate of these neurons (Aghajanian and Bunney, 1977). Moreover, the data from Guiard et al (2008a) showed that this inhibition is significantly attenuated by both the a2-adrenoceptor antagonist idazoxan and the D2 antagonist raclopride. Based on such in vivo results, it was thus not surprising to have observed a decrease in the firing rate of VTA DA neurons following a sustained reboxetine regimen. In conclusion, this study demonstrated that the catecholamine reuptake inhibitor nomifensine produced marked changes in all three monoaminergic systems. The progressive recovery of DA neuron firing, as well as putative delayed a2adrenoceptor desensitization on NE terminals, known to occur with NRIs (Invernizzi and Garattini, 2004) may have accounted for the delayed antidepressant action of nomifensine. Future endeavors will investigate the effects of this drug in postsynaptic structures.

Acknowledgements The authors thank Jonathan James for his assistance in analyzing blood samples, Frank Chenu, PhD, for his assistance during experiments and statistical analysis. We also thank Lundbeck

Journal of Psychopharmacology 0(00) Pharmaceuticals for supplying escitalopram and Janssen Pharmaceutica for supplying paliperidone. This study was supported by the Canadian Institutes for Health Research grant (77838) to P. Blier, a Canada Chair and an endowed chair from the University of Ottawa Institute of Mental Health Research. This study was presented in part at the 63rd annual scientific convention of the Society for Biological Psychiatry, Washington, DC, USA, May 1–3, 2008.

References Aghajanian GK, Bunney BS (1977) Dopamine autoreceptors: pharmacological characterization by microiontophoretic single cell recording studies. N–S Arch Pharmacol 297: 1–7. Aghajanian GK, Vandermaelen CP (1982a) Alpha 2-adrenoceptormediated hyperpolarization of locus coeruleus neurons: intracellular studies in vivo. Science 215: 1394–1396. Aghajanian GK, Vandermaelen CP (1982b) Intracellular recordings from serotonergic dorsal raphe neurons: pacemaker potentials and the effect of LSD. Brain Res 238: 463–469. Aman TK, Shen RY, Haj-Dahmane S (2007) D2-like dopamine receptors depolarize dorsal raphe serotonin neurons through the activation of nonselective cationic conductance. J Pharmacol Exp Ther 320: 376–385. Baraban JM, Aghajanian GK (1980) Suppression of firing activity of 5-HT neurons in the dorsal raphe by alpha-adrenoceptor antagonists. Neuropharmacology 19: 355–363. Baumann MH, Char GU, De Costa BR, Rice KC, Rothman RB (1994) GBR12909 attenuates cocaine-induced activation of mesolimbic dopamine neurons in the rat. J Pharmacol Exp Ther 271: 1216–1222. Blier P, De Montigny C (1987) Modification of 5-HT neuron properties by sustained administration of the 5-HT1A agonist gepirone: electrophysiological studies in the rat brain. Synapse 1: 470–480. Blier P, Saint-Andre´ E, He´bert C, De Montigny C, Lavoie N, Debonnel G (2007) Effects of different doses of venlafaxine on serotonin and norepinephrine reuptake in healthy volunteers. Int J Neuropsychop 10: 41–50. Bonnet JJ, Chagraoui A, Protais P, Costentin J (1987) Interactions of amineptine with the neuronal dopamine uptake system: neurochemical in vitro and in vivo studies. J Neural Transm 69: 211–220. Cedarbaum JM, Aghajanian GK (1977) Catecholamine receptors on locus coeruleus neurons: pharmacological characterization. Eur J Pharmacol 44: 375–385. Chen Z, Skolnick P (2007) Triple uptake inhibitors: therapeutic potential in depression and beyond. Expert Opin Inv Drug 16: 1365–1377. Chernoloz O, El Mansari M, Blier P (2009a) Sustained administration of pramipexole modifies the spontaneous firing of dopamine, norepinephrine, and serotonin neurons in the rat brain. Neuropsychopharmacology 34: 651–661. Chernoloz O, El Mansari M, Blier P (2009b) Electrophysiological studies in the rat brain for aripiprazole augmentation of antidepressants in major depressive disorder. Psychopharmacology 206: 335–344. Claustre Y, Leonetti M, Santucci V, et al. (2008) Effects of the beta3adrenoceptor (Adrb3) agonist SR58611A (amibegron) on serotonergic and noradrenergic transmission in the rodent: relevance to its antidepressant/anxiolytic-like profile. Neuroscience 156: 353–364. De Deurwaerdere P, Navailles S, Berg KA, Clarke WP, Spampinato U (2004) Constitutive activity of the serotonin2C receptor inhibits in vivo dopamine release in the rat striatum and nucleus accumbens. J Neurosci 24: 3235–3241.

Katz et al. Di Giovanni G, Di Matteo V, Pierucci M, Esposito E (2008) Serotonin-dopamine interaction: electrophysiological evidence. Prog Brain Res 172: 45–71. Dong J, Blier P (2001) Modification of norepinephrine and serotonin, but not dopamine, neuron firing by sustained bupropion treatment. Psychopharmacology 155: 52–57. Dremencov E, El Mansari M, Blier P (2007) Distinct electrophysiological effects of paliperidone and risperidone on the firing activity of rat serotonin and norepinephrine neurons. Psychopharmacology 194: 63–72. Dunlop BW, Nemeroff CB (2007) The role of dopamine in the pathophysiology of depression. Arch Gen Psychiat 64: 327–337. Einhorn LC, Johansen PA, White FJ (1988) Electrophysiological effects of cocaine in the mesoaccumbens dopamine system: studies in the ventral tegmental area. J Neurosci 8: 100–112. El Mansari M, Ghanbari R, Janssen S, Blier P (2008) Sustained administration of bupropion alters the neuronal activity of serotonin, norepinephrine but not dopamine neurons in the rat brain. Neuropharmacology 55: 1191–1198. Elam M, Clark D, Svensson TH (1986) Electrophysiological effects of the enantiomers of 3-PPP on neurons in the locus coeruleus of the rat. Neuropharmacology 25: 1003–1008. Fujita M, Shimada S, Fukuchi K, Tohyama M, Nishimura T (1994) Distribution of cocaine recognition sites in rat brain: in vitro and ex vivo autoradiography with [125I] RTI-55. J Chem Neuroanat 7: 13–23. Grace AA, Bunney BS (1984a) The control of firing pattern in nigral dopamine neurons: single spike firing. J Neurosci 4: 2866–2876. Grace AA, Bunney BS (1984b) The control of firing pattern in nigral dopamine neurons: burst firing. J Neurosci 4: 2877–2890. Guiard BP, El Mansari M, Blier P (2008a) Crosstalk between dopaminergic and noradrenergic systems in the rat ventral tegmental area, locus coeruleus, and dorsal hippocampus. Mol Pharmacol 74: 1463–1475. Guiard BP, El Mansari M, Merali Z, Blier P (2008b) Functional interactions between dopamine, serotonin and norepinephrine neurons: an in-vivo electrophysiological study in rats with monoaminergic lesions. Int J Neuropsychop 11: 625–639. Hajos M, Allers KA, Jennings K, et al. (2007) Neurochemical identification of stereotypic burst-firing neurons in the rat dorsal raphe nucleus using juxtacellular labelling methods. Eur J Neurosci 25: 119–126. Hensler JG, Cervera LS, Miller HA, Corbitt J (1996) Expression and modulation of 5-hydroxytryptamine1A receptors in P11 cells. J Pharmacol Exp Ther 278: 1138–1145. Invernizzi RW, Garattini S (2004) Role of presynaptic alpha2-adrenoceptors in antidepressant action: recent findings from microdialysis studies. Prog Neuro–Psychoph 28: 819–827. Javitch JA, Strittmatter SM, Snyder SH (1985) Differential visualization of dopamine and norepinephrine uptake sites in rat brain using [3H] mazindol autoradiography. J Neurosci 5: 1513–1521. Kruse H, Hoffmann I, Gerhards HJ, Leven M, Schacht U (1977) Pharmacological and biochemical studies with three metabolites of nomifensine. Psychopharmacology (Berlin) 51: 117–123. Linner L, Wiker C, Arborelius L, Schalling M, Svensson TH (2004) Selective noradrenaline reuptake inhibition enhances serotonergic

13 neuronal activity and transmitter release in the rat forebrain. J Neural Transm 111: 127–139. Martin-Ruiz R, Ugedo L, Honrubia MA, Mengod G, Artigas F (2001) Control of serotonergic neurons in rat brain by dopaminergic receptors outside the dorsal raphe nucleus. J Neurochem 77: 762–775. McKillop D, Bradford HF (1981) Comparative effects of benztropine and nomifensine on dopamine uptake and release from striatal synaptosomes. Biochem Pharmacol 30: 2753–2758. Page M, Lucki I (2002) Effects of acute and chronic reboxetine treatment on stress-induced monoamine efflux in the rat frontal cortex. Neuropsychopharmacology 27: 237–247. Petersen EN, Bechgaard E, Sortwell RJ, Wetterberg L (1978) Potent depletion of 5HT from monkey whole blood by a new 5HT uptake inhibitor, paroxetine (FG 7051). Eur J Pharmacol 52: 115–119. Piercey MF, Smith MW, Lum-Ragan JT (1994) Excitation of noradrenergic cell firing by 5-hydroxytryptamine1A agonists correlates with dopamine antagonist properties. J Pharmacol Exp Ther 268: 1297–1303. Pitts DK, Wang L, Kelland MD, Freeman AS, Chiodo LA (1995) Repeated stimulation of dopamine D2-like receptors: reduced responsiveness of nigrostriatal and mesoaccumbens dopamine neurons to quinpirole. J Pharmacol Exp Ther 275: 412–421. Prisco S, Pagannone S, Esposito E (1994) Serotonin-dopamine interaction in the rat ventral tegmental area: an electrophysiological study in vivo. J Pharmacol Exp Ther 271: 83–90. Randrup A, Munkvad I, Fog R, et al. (1975) Mania, depression and brain dopamine. Curr Dev Psychopharmacol 2: 206–248. Schacht U, Heptner W (1974) Effect of nomifensine (HOE 984), a new antidepressant, on uptake of noradrenaline and serotonin and on release of noradrenaline in rat brain synaptosomes. Biochem Pharmacol 23: 3413–3422. Szabo ST, De Montigny C, Blier P (2000) Progressive attenuation of the firing activity of locus coeruleus noradrenergic neurons by sustained administration of selective serotonin reuptake inhibitors. Int J Neuropsychop 3: 1–11. Szabo ST, Blier P (2001a) Functional and pharmacological characterization of the modulatory role of serotonin on the firing activity of locus coeruleus norepinephrine neurons. Brain Res 922: 9–20. Szabo ST, Blier P (2001b) Effect of the selective noradrenergic reuptake inhibitor reboxetine on the firing activity of noradrenaline and serotonin neurons. Eur J Neurosci 13: 2077–2087. Szabo ST, Blier P (2002) Effects of serotonin (5-hydroxytryptamine, 5-HT) reuptake inhibition plus 5-HT(2A) receptor antagonism on the firing activity of norepinephrine neurons. J Pharmacol Exp Ther 302: 983–991. Tatsumi M, Groshan K, Blakely RD, Richelson E (1997) Pharmacological profile of antidepressants and related compounds at human monoamine transporters. Eur J Pharmacol 340: 249–258. Ungless MA, Magill PJ, Bolam JP (2004) Uniform inhibition of dopamine neurons in the ventral tegmental area by aversive stimuli. Science 303: 2040–2042.