Sustained pharmacological blockade of NK 1 substance P receptors causes functional desensitization of dorsal raphe 5-HT 1A autoreceptors in mice

Transcription

1 Journal of Neurochemistry, 25, 95, doi:1.1111/j x Sustained pharmacological blockade of NK 1 substance P receptors causes functional desensitization of dorsal raphe 5-HT 1A autoreceptors in mice Bruno P. Guiard,*,1 Nicolas Froger,,1 Michel Hamon, Alain M. Gardier* and Laurence Lanfumey UMR 677 INSERM/UPMC, Neuropsychopharmacologie, CHU Pitié-Salpêtrière, Paris, France *Laboratoire de Neuropharmacologie, EA 3544 MJENR, Faculté de Pharmacie, Université Paris-Sud, Châtenay-Malabry, France Abstract Antagonists at NK 1 substance P receptors have demonstrated similar antidepressant properties in both animal paradigms and in human as selective serotonin reuptake inhibitors (SSRIs) that induce desensitization of 5-HT 1A autoreceptors within the dorsal raphe nucleus (DRN). We investigated whether this receptor adaptation also occurs upon NK 1 receptor blockade. C57B/L6J mice were treated for 21 days with the selective NK 1 receptor antagonist GR (1 mg/kg daily) through subcutaneously implanted osmotic mini pumps, and DRN 5-HT 1A autoreceptor functioning was assessed using various approaches. Recording of DRN serotonergic neurons in brainstem slices showed that GR treatment reduced (by 1.5 fold) the potency of the 5-HT 1A receptor agonist, ipsapirone, to inhibit cell firing. In parallel, the 5-HT 1A autoreceptor-mediated [ 35 S]GTP-c-S binding induced by 5-carboxamidotryptamine onto the DRN in brainstem sections was significantly decreased in GR treated mice. In vivo microdialysis showed that the cortical 5-HT overflow caused by acute injection of the SSRI paroxetine (1 mg/kg) was twice as high in GR treated as in vehicle-treated controls. In the DRN, basal 5-HT outflow was significantly enhanced by GR treatment. These data supported the hypothesis that chronic NK 1 receptor blockade induces a functional desensitization of 5-HT 1A autoreceptors similar to that observed with SSRIs. Keywords: 5-HT 1A autoreceptor, desensitization, dorsal raphe nucleus, NK 1 receptors, substance P receptor antagonists. J. Neurochem. (25) 95, A large body of evidence supports the idea that a deficit in central serotonin (5-hydroxytryptamine, 5-HT) neurotransmission is associated with depression (Coppen et al. 1973; Asberg et al. 1984; Delgado et al. 199). In line with this concept, effective treatment of the disease is achieved with selective serotonin reuptake inhibitors (SSRIs) that enhance central 5-HT neurotransmission (Bel and Artigas 1993; Kreiss and Lucki 1995), mainly through a functional desensitization of somatodendritic 5-HT 1A autoreceptors located on dendrites and somata of 5-HT neurons within the dorsal raphe nucleus (DRN) (Blier and De Montigny 1983; Le Poul et al. 2). Because 5-HT 1A autoreceptor activation inhibits 5-HT neuron firing (Aghajanian et al. 1987), this desensitization may promote 5-HT neurotransmission, thereby representing an adaptive change relevant to the therapeutic effect of SSRIs. Although SSRIs are effective in most depressive episodes, they do not alleviate symptoms in 3% of depressed patients and some adverse effects have been reported after long-term treatment with these drugs (Stahl 1998). In addition, the therapeutic efficacy of SSRIs is blunted by the long delay in Received March 1, 25; revised manuscript received July 21, 25; accepted August 16, 25. Address correspondence and reprint requests to Laurence Lanfumey, UMR 677 INSERM/UPMC, Neuropsychopharmacologie, Faculté de Médecine Pitié-Salpêtrière, 91, Boulevard de l Hôpital, Paris Cedex 13, France. lanfumey@ext.jussieu.fr 1 These authors contributed equally to this work. Abbreviations used: acsf, artificial cerebrospinal fluid; AUC, area under curve; DRN, dorsal raphe nucleus; 5-HT, serotonin; NS, not significant; n.s., non-specific; OD, optical density; PLSD, protected least significance difference; SSRIs, selective serotonin reuptake inhibitors. Ó 25 International Society for Neurochemistry, J. Neurochem. (25) 95,

2 1714 B. P. Guiard et al. clinical benefit. Accordingly, different strategies are under way in order to develop novel antidepressant therapies. Recent clinical and preclinical studies have pointed out the potential therapeutic action of substance P receptor antagonists (NK 1 receptor antagonists) in major depressive disorders (Rupniak 22a,b). The implication of substance P neurotransmission in mood disorders had been previously evoked by Rimon et al. (1984), who reported an increase in substance P levels in the cerebrospinal fluid of depressed patients. Substance P is a neuropeptide which belongs to the tachykinin family. Tachykinins bind to three distinct receptors designated NK 1, NK 2 and NK 3. In rodent brain, substance P actions are preferentially mediated via the NK 1 receptors, although a mediation via NK 2 and NK 3 receptors is also possible as substance P has some affinity for those receptors (Saria 1999; Strand 1999). In various animal species, pharmacological blockade of NK 1 receptors has been shown to attenuate behavioural and neurochemical responses to stress (Ebner et al. 24; Hutson et al. 24), to reduce ultrasonic vocalization in pups isolated from their mother (Kramer et al. 1998), to decrease aggressiveness (Shaikh et al. 1993), to restore responsiveness to rewarding stimuli in rats subjected to chronic mild stress (Papp et al. 2) and to reduce immobility time in the forced swimming test (Zocchi et al. 23). In humans, a clear-cut antidepressant action, similar to that of the SSRI paroxetine, was reported for the NK 1 receptor antagonists MK-869 and L by Kramer et al. (1998, 24). However, the precise neurobiological mechanisms underlying the antidepressant action of these NK 1 receptor antagonists have yet to be elucidated. As emphasized above, DRN 5-HT 1A autoreceptors appear to be critically involved in the antidepressant action of SSRIs. Accordingly, whether NK 1 receptor antagonists also act, albeit indirectly, through 5-HT 1A autoreceptors is an important question to be addressed. Recently, this issue was assessed in NK 1 receptor knock-out mice in which a marked functional desensitization of 5-HT 1A autoreceptors was observed in the DRN (Froger et al. 21). Whether or not sustained pharmacological blockade of NK 1 receptors could mimic the effect of NK 1 receptor gene knock-out has now been investigated in mice. For this purpose, wild-type C57BL/6J mice were chronically treated with GR 25171, a selective and brain penetrating NK 1 receptor antagonist (Gardner et al. 1996), and electrophysiological, autoradiographic and biochemical approaches were used to assess possible changes in 5-HT 1A receptor functioning caused by the treatment. Materials and methods Animals Experiments were performed on male C57BL/6J mice (CER Janvier, Le Genest-St Isle, France) weighing 2 25 g (8 1 weeks old) and housed six per cage under standard conditions (12-h light/dark cycle, 22 ± 1 C ambient temperature, 6% relative humidity, food and water ad libitum). Procedures involving animals and their care were conducted in conformity with the institutional guidelines that are in compliance with national and international laws and policies (Council directive no , 19 October 1987, Ministère de l Agriculture et de la Forêt, Service Vétérinaire de la Santé et de la Protection Animale, permissions no to MH, no to LL and no to AMG). Pharmacological treatments Acute treatments Mice were injected subcutaneously with either GR (1 mg/ kg) or its vehicle (.9% NaCl), and were decapitated 6 min later. Brains were rapidly removed in order to proceed with electrophysiological experiments. For microdialysis experiments, mice received a challenge dose of paroxetine (1 mg/kg, i.p.) or a combined treatment with paroxetine plus WAY1635 (.5 mg/kg, s.c.). Control animals were injected with the vehicle (saline) through the same routes of administration. Chronic treatments Mice were treated for 21 days with the NK 1 receptor antagonist GR The compound was dissolved in.9% NaCl/dimethylsulfoxide (8/2) at a concentration of 2 mg/ml to produce an average dose of 1 mg/kg/day. Then, 1 ll of this solution were filled into the osmotic mini pump (Alzet Ò, 17D, Charles River Laboratories Inc., Wilmington, MA, USA) ensuring constant delivery (.5 ll/h) for 7 days. The mini pump was implanted subcutaneously on the back under light anaesthesia with ether; this procedure was repeated two times for the 3-week treatment of each animal. Control mice were implanted with mini pumps filled with vehicle only. Control and treated mice were decapitated 24 h after removal of the third mini pump, and brains were rapidly taken out to perform electrophysiological and [ 35 S]GTP-c-S binding experiments (see below). For microdialysis experiments, the third mini pump was removed on day 21, just before implantation of the microdialysis probe either in the frontal cortex or in the DRN. Animals were then returned to their home cage for recovery until the start of dialysate sample collection. Electrophysiological experiments Immediately after removal from the skull, the brain was immersed in an ice-cold artificial cerebrospinal fluid (acsf1) of the following composition (mm): NaCl 126, KCl 3.5, NaH 2 PO 4 1.2, MgCl 2 1.3, CaCl 2 2., NaHCO 3 25, D-glucose 11, continuously bubbled with carbogen (95% O 2 /5% CO 2 ) to maintain ph value at 7.3. A block of tissue containing the DRN was cut into sections (4-lm thick) in the same ice-cold acsf1 using a vibratome. Brain stem slices were immediately immersed in oxygenated acsf1 at room temperature (22 C). A single slice was then placed on a nylon mesh, completely submerged in the recording chamber and continuously superfused with oxygenated acsf1 (34 C) at a constant flow rate of 2 3 ml/min (Haj-Dahmane et al. 1991). Extracellular recordings of the firing of DRN serotonergic neurons were made using glass microelectrodes filled with 2 M NaCl (12 15 MW). Cells were identified as 5-HT neurons according Ó 25 International Society for Neurochemistry, J. Neurochem. (25) 95,

3 NK 1 receptor antagonism on 5-HT 1A receptor function 1715 to previously described criteria (Lanfumey et al. 1999). Firing was evoked in the otherwise silent neurons by adding the a 1 -adrenoceptor agonist phenylephrine (3 lm) into the superfusing acsf1 (VanderMaelen and Aghajanian 1983). Electrical signals were fed into a high-input impedance amplifier (VF 18, BioLogic, Claix, France), an oscilloscope and an electronic ratemeter triggered by individual action potentials, connected to an A/D converter and a personal computer (Haj-Dahmane et al. 1991). Using dedicated software, the integrated firing rate was recorded, computed, and displayed on a chart recorder as consecutive 1-s samples. Baseline activity was recorded for at least 1 min prior to perfusion of drugs into the chamber, via a three-way tap system. Because complete exchange of fluids occurred within 2 min following the arrival of a new solution into the chamber, the duration of each drug application was 3 min. The effects of a given drug were evaluated by comparing the mean discharge frequency during the 2 min prior to its application with that recorded at the peak action of the drug, i.e. 2 3 min after its removal from the perfusing acsf1. Data are expressed as percentage of the baseline firing rate ± SEM. Non-linear regression fitting was carried out using Prism 4. (GraphPad Software Inc., San Diego, CA, USA) software for the calculation of EC 5 values of drugs. Quantitative autoradiography of 5-HT 1A receptor-mediated [ 35 S]GTP-c-S binding Immediately after removal from the skull, brains were frozen by immersion in isopentane chilled at )3 C with dry ice, then stored at )8 C for 1 3 weeks. Coronal sections (2-lm thick) were cut at )2 C, thaw-mounted onto gelatin-coated slides, and stored at )8 C until use. The protocol for autoradiographic measurement of 5-HT 1A receptor-stimulated [ 35 S]GTP-c-S binding was adapted from Fabre et al. (2). Briefly, brain sections were pre-incubated at room temperature for an initial 15-min period in 5 mm HEPES, ph 7.5, supplemented with 1 mm NaCl, 3 mm MgCl 2,.2 mm EGTA and 2mM dithiothreitol, and then for another 15 min in the same buffer with 2 mm GDP and 1 lm 8-cyclopentyl-1,3-dipropylxanthine (DPCPX, an A 1 adenosine receptor antagonist) to decrease background labelling (Laitinen and Jokinen 1998). Thereafter, sections were incubated for 1 h at 3 C in the same buffer with.5 nm [ 35 S]GTP-c-S (1 Ci/mmol) either in the absence (basal conditions) or the presence (stimulated conditions) of 5-carboxamidotryptamine (5-CT, a non-selective potent 5-HT1 receptor agonist; Hamon 1997) at four different concentrations (1 )8,1 )7,1 )6, 1 )5 M). Non-specific binding was determined in the presence of 1 )5 M WAY 1635 to block 5-HT 1A receptors (Fletcher et al. 1996). Incubation was stopped by two 2-min washes in ice-cold 5 mm HEPES, ph 7.5, and a brief immersion in ice-cold distilled water. Sections were dried and exposed to a Biomax MR autoradiographic film (Kodak). Optical density (OD) was measured on autoradiographic films using computerized image software (Samba Technologies, Meylan, France). 5-CT-stimulated [ 35 S]GTP-c-S binding is expressed as percentage over the baseline {[(stimulated-basal)/ basal]*1) ± SEM}. Intracerebral in vivo microdialysis in freely moving mice Concentric dialysis probes were made of cuprophan fibres and set up as described previously (Malagié et al. 1996). Animals were anaesthetized with chloral hydrate (4 mg/kg, i.p.) and placed in a stereotaxic frame. The probes were implanted into either the frontal cortex or the DRN (active length: mm outer diameter and 1..3 mm outer diameter, respectively) according to the Mouse brains atlas of Hof et al. (2). Stereotaxic coordinates from bregma were (in mm) for the frontal cortex: anterior ¼ +1.6, lateral ¼ +1.3, ventral ¼ )1.6, and for the DRN: anterior ¼ )4.5, lateral ¼, ventral ¼ )3.. Animals were allowed to recover from surgery overnight and the next day, i.e. 2 h after the surgery, continuous perfusion of the probe with acsf2 (composition in mm: NaCl 147, KCl 3.5, CaCl , MgCl 2 1.2, NaH 2 PO 4 1., ph 7.4) started at a flow rate of 1.5 and.5 ll/min in the frontal cortex and the DRN, respectively, using a CMA/1 pump (Carnegie Medicine, Stockholm, Sweden). Dialysates were collected in Eppendorf tubes for the measurement of extracellular levels of 5-HT ([5-HT] ext ) using a high-performance liquid chromatography (HPLC) system (XL-ODS, mm, particle size 3 lm; Beckman, Roissy-Charles-de-Gaulle, France) coupled to an amperometric detector (HP149A, Hewlett-Packard, Les Ulis, France). The dialysate fractions were collected every 3 min in the DRN, and every 15 min in the frontal cortex, so as to adapt to the detection limit of our HPLC system and to allow reliable quantification of 5-HT accumulated per sample. The basal value of [5-HT] ext was calculated as the mean ± SEM of the first four samples collected before systemic administration of drugs. Extracellular 5-HT levels were not corrected for probe recovery and/or differences in probe length. The limit of sensitivity for the quantification of [5-HT] ext was.5 fmol per sample (signal-tonoise ratio ¼ 2). At the end of microdialysis experiments, mice were killed by cervical dislocation, brains were removed and stored at )4 C. To determine the exact probe implantation into the frontal cortex or the DRN, brains were placed in a Kryomat apparatus and kept at )25 C. Coronal frozen sections were cut serially at 4 lm intervals (Guiard et al. 24). Each section was photographed using a digital camera, and brain regions were identified according to Hof et al. (2). Only mice with probes confined to either the frontal cortex or the DRN were used for subsequent data analyses. Statistical analyses All data are given as means ± SEM. Extracellular recordings and [ 35 S]GTP-c-S binding were analyzed by two-way ANOVA followed by Bonferroni post-hoc tests. Unpaired two-tailed Student s t-test was used also to compare the treated groups with their control. Microdialysis results were analyzed by the computer software StatView 5. (Abacus Concepts Inc., Berkeley, CA, USA), used to compare area under the curve (AUC; mean ± SEM) values expressed as percentage of basal [5-HT] ext levels. These AUC values were calculated as the amount of 5-HT outflow collected during the 18-min period after the administration of a challenge dose of saline, paroxetine alone or paroxetine plus WAY For each brain region, a two-way ANOVA on AUC values was performed with the chronic drug treatment (vehicle or GR at 1 mg/kg/day for 21 days) and the single challenge injection performed on day 21 (saline, paroxetine alone or paroxetine plus WAY 1635) as main factors, followed by Fisher protected least significance difference (PLSD) post-hoc test. Statistical significance was set at p <.5. Ó 25 International Society for Neurochemistry, J. Neurochem. (25) 95,

4 1716 B. P. Guiard et al. Drugs GR [2-methoxy-5-(5-trifluoromethyl-tetrazol-1-yl)-(2-phenyl-piperidin-3-yl)-amine] and paroxetine were generously given by GlaxoSmithKline (Harlow, UK). [ 35 S]GTP-c-S was purchased from Amersham Pharmacia Biotech (Buckinghamshire, UK). Other compounds were: N-[2-[4-(2-methoxyphenyl)1-piperazinyl]- ethyl]-n-(2-pyridinyl)-cyclohexanecarboxamide (WAY 1635, Wyeth-Ayerst, Princeton, NJ, USA), GDP dilithium salt (Boehringer-Mannheim, Meylan, France), DPCPX and 5-CT (Research Biochemicals Inc., Natick, MA, USA), ipsapirone (Troponwerke Bayer, Cologne, Germany). Results Effects of the NK 1 antagonist, GR 25171, on the electrical activity of DRN 5-HT neurons In vitro application of GR DRN serotonergic neurons recorded in vitro in brainstem slices from C57BL/6J mice displayed a characteristic slow (1.74 ±.32 spikes/s, mean ± SEM, n ¼ 6) and regular pattern of discharge (Fig. 1), as previously described for other mouse strains (Lanfumey et al. 1999). The typical example illustrated in Fig. 1 shows that superfusion of brainstem slices with increasing concentrations of GR (1-9 )1-5 M) did not significantly modify the firing rate of DRN serotonergic neurons. In contrast, the 5-HT 1A receptor agonist ipsapirone produced a Spikes per 1 sec 2 2 GR (nm) GR (nm) ipsapirone (nm) min Fig. 1 Effects of the NK 1 receptor antagonist GR (1 nm)1 lm) and the 5-HT 1A receptor agonist ipsapirone (3 1 nm) on DRN 5-HT neuron firing in vitro. Typical integrated firing rate histograms (in spikes per 1 s) showing the absence of effect of GR on the electrical activity of DRN 5-HT neurons. The inhibitory effect of ipsapirone on neuron firing demonstrated the serotonergic nature of the recorded cell. Similar data have been obtained on at least three cells for each pharmacological condition tested. concentration-dependent inhibition of the electrophysiological activity of the same neurons (Fig. 1). Acute in vivo administration of GR Neither the frequency nor the pattern of the discharge of DRN serotonergic neurons were significantly altered in mice that had been given GR (1 mg/kg s.c.) or saline 6 min before death compared with naive, untreated, mice (not shown). Furthermore, the dose response curve of ipsapirone-induced inhibition was similar in GR treated mice (EC 5 ¼ 46.8 ± 5. nm, mean ± SEM, n ¼ 6) and in paired saline-treated controls (EC 5 ¼ 36.1 ± 3.3 nm, mean ± SEM, n ¼ 6; F 1,42 ¼ 3.4, NS; Fig. 2). In both groups, complete inhibition of DRN 5-HT neuron firing was obtained by tissue superfusion with 1 nm ipsapirone (Figs. 2a and b). Chronic administration of GR Chronic treatment with GR (1 mg/kg/day s.c. for 21 days), such as that noted after acute administration, did not alter the electrophysiological activity of DRN 5-HT neurons recorded in brainstem slices. The firing rate of these neurons was similar in mice that had been treated for 3 weeks with either GR (1.8 ±.2 spikes/s, n ¼ 15) or saline (1.8 ±.1 spikes/s, n ¼ 21) and in the same range as that of naive, untreated, animals (see above). Although ipsapirone caused a concentration-dependent inhibition of the firing of DRN serotonergic neurons in both saline- and GR treated mice (Figs. 3a and b), this effect appeared significantly less pronounced in the latter group. Indeed, the concentration response curve of ipsapirone in GR treated mice was shifted to the right compared with that for paired control mice (F 1,29 ¼ 11.11, p <.1), with an EC 5 value of the 5-HT 1A receptor agonist 1.5 times higher in GR treated animals (57.5 ± 8.3 nm, n ¼ 15) than in the saline-treated group (37. ± 4.4 nm, n ¼ 21, p <.5) (Fig. 3b). Furthermore, complete inhibition of DRN 5-HT neuron firing required only 1 nm ipsapirone in the latter group, but up to 3 nm of the drug in GR treated mice (Fig. 3a). Effects of chronic administration of GR on 5-HT 1A receptor-mediated [ 35 S]GTP-c-S binding Because electrophysiological experiments showed the occurrence of DRN 5-HT 1A autoreceptor desensitization in mice that had been treated for 3 weeks with GR (1 mg/kg/day s.c.), further investigations were made using autoradiographic measurement of [ 35 S]GTP-c-S binding in order to assess G-protein coupling of this receptor. Under basal conditions, i.e. in the absence of 5-CT, [ 35 S]GTP-c-S labelling within the DRN did not differ between GR treated mice (OD 9.6 ±.5, n ¼ 5) and paired vehicle-treated controls (9.2 ±.8, n ¼ 4; Fig. 4a). In both groups, 5-CT induced a concentration-dependent increase in Ó 25 International Society for Neurochemistry, J. Neurochem. (25) 95,

5 NK 1 receptor antagonism on 5-HT 1A receptor function 1717 (a) 2 ipsapirone (nm) (a) 2 ipsapirone (nm) Spikes per 1 sec 2 vehicle ipsapirone (nm) Spikes per 1 sec 2 vehicle ipsapirone (nm) GR (1 mg/kg) 3 min GR (1 mg/kg/d) 3min (b) 1 (b) 1 Inhibition of firing (%) ( ) vehicle ( ) GR (1 mg/kg) Inhibition of firing (%) ( ) vehicle * * ( ) GR (1 mg/kg) log [ipsapirone] M Fig. 2 Effects of ipsapirone on the firing of DRN 5-HT neurons in mice injected acutely with GR (1 mg/kg s.c ) or its vehicle. Electrophysiological recordings were performed in brainstem slices from mice decapitated 6 min after the treatment. (a) Integrated firing rate histograms (in spikes per 1 s) showing ipsapirone-induced inhibition of 5-HT neuron firing in both a GR and a vehicle-injected mouse (bottom and top recordings, respectively). (b) Concentration response curves of ipsapirone-induced inhibition of the firing of DRN 5-HT neurons in brainstem slices from GR and vehicleinjected mice. Ipsapirone-induced inhibition is expressed as a percentage of the baseline firing rate. Each point is the mean ± SEM of data obtained from five to six individual cells. The dotted lines illustrate the EC 5 values of ipsapirone (abscissa) in GR and vehicleinjected mice. [ 35 S]GTP-c-S labelling, which could be completely prevented by the selective 5-HT 1A receptor antagonist WAY 1635 (at 1 lm; Fig. 4a). However, the 5-CTinduced increase in [ 35 S]GTP-c-S binding was significantly less in the GR treated group (F 1,34 ¼ , p <.1). Thus, the concentration response curve of 5- CT-evoked [ 35 S]GTP-c-S binding in GR treated mice (EC 5 ¼ 47.4 ± 16.1 nm, n ¼ 5) was shifted to the right compared with that for vehicle-treated mice (EC 5 ¼ 13.6 ± 1.2 nm, n ¼ 4; p <.5; Table 1; Fig. 4b). Furthermore, the maximal [ 35 S]GTP-c-S binding (E max ) obtained with 1 1 lm 5-CT was significantly lower in GR treated mice than in paired controls (Fig. 4b; Table 1) log [ipsapirone] M Fig. 3 Effects of ipsapirone on the firing of DRN 5-HT neurons in mice after a 3-week treatment with GR (1 mg/kg/d s.c.) or its vehicle. Electrophysiological recordings were performed in brainstem slices from mice decapitated 24 h after removal of osmotic mini pump delivering GR or the vehicle. (a) Integrated firing rate histograms (in spikes per 1 s) showing ipsapirone-induced inhibition of 5-HT neuron firing in both a GR and a vehicle-treated mouse (bottom and top recordings, respectively). (b) Concentration response curves of ipsapirone-induced inhibition of the firing of DRN 5-HT neurons in brainstem slices from GR and vehicle-treated mice. Ipsapirone-induced inhibition is expressed as a percentage of the baseline firing rate. Each point is the mean ± SEM of data obtained from 15 to 21 individual cells. The dotted lines illustrate the EC 5 values of ipsapirone (abscissa) in GR and vehicletreated mice. *p <.5, as compared with the respective inhibition in control mice (Student s t-test). Effects of chronic administration of GR on [5-HT]ext in freely moving mice Basal outflow in the frontal cortex and the DRN Under our microdialysis conditions, [5-HT] ext was relatively stable for the whole duration of experiments (18 min) allowing calculations of mean values in the frontal cortex and the DRN. The basal [5-HT] ext levels in these two brain regions were measured 2 h after the removal of the osmotic mini pumps. The mean basal [5-HT] ext levels in the frontal cortex across all treatment groups were not Ó 25 International Society for Neurochemistry, J. Neurochem. (25) 95,

6 1718 B. P. Guiard et al. (a) vehicle GR (1 mg/kg/d) Table 1 Effects of chronic treatment with GR (1 mg/kg/d, 21 days) or its vehicle on the characteristics of 5-CT-evoked [ 35 S]GTP-c-S binding onto the DRN in brain sections 5-CT-evoked [ 35 S]GTP-c-S binding (b) Percentage over baseline (%) n.s. BASAL 1-8 M 5-CT 1-7 M 5-CT 1-6 M 5-CT 1-5 M 5-CT 5-CT WAY 1635 ( ) vehicle ** ** log [5-CT] M * * ( ) GR (1 mg/kg/d) Fig. 4 Effects of long-term treatment with GR (1 mg/kg/d, 21 days) on 5-HT 1A receptor-mediated increase in [ 35 S]GTP-c-S binding onto the DRN in brain sections. Autoradiographic labelling by [ 35 S]GTP-c-S was performed using brain sections from mice decapitated 24 h after removal of osmotic mini pump delivering GR or the vehicle. (a) Representative autoradiograms of brain sections (2-lm thick) at the level of the DRN, labelled by [ 35 S]GTP-c-S (5 pm) in the absence (Basal) or the presence of four different concentrations (1 )8,1 )7,1 )6,1 )5 M) of 5-CT. Non-specific (n.s.) labelling was obtained from adjacent sections exposed to 1 lm 5-CT plus 1 lm WAY Similar autoradiograms were obtained from four to five mice per group. (b) Concentration response curves of 5-CT-induced [ 35 S]GTP-c-S binding onto the DRN in brain sections from GR and vehicle-treated mice. 5-CT-induced increase in [ 35 S]GTP-c-S binding is expressed as percentage over baseline (no 5- CT). Each point is the mean ± SEM of data obtained from four to five mice. *p <.5, **p <.1 as compared with the respective increase (% over baseline) in vehicle-treated controls (Student s t-test). Vehicle GR EC 5 (nm) 13.6 ± 1.2 (4) 47.4 ± 16.1 (5)** E max (%) ± 9.2 (4) 11.3 ± 5.1 (5)* In vitro binding experiments and quantitative autoradiographic determinations were as described in the legend to Fig. 4. EC 5 is the concentration of 5-CT inducing half-maximal [ 35 S]GTP-c-S binding, and E max is the maximal 5-CT-evoked increase expressed as percentage over basal [ 35 S]GTP-c-S binding. Results are the means ± SEM of independent determinations in the number of mice indicated in parentheses. *p <.5, **p <.1, as compared with vehicle-treated mice (unpaired two-tailed Student s t-test). statistically different between GR treated mice (4.4 ±.2 fmol/15 min fraction, n ¼ 23) and paired control mice treated with the vehicle (4.1 ±.1 fmol/15 min fraction, n ¼ 25; F 1,46 ¼.53, p ¼.46) (Fig. 5a). By contrast, the mean basal [5-HT] ext levels in the DRN were significantly higher in GR treated animals (21.7 ± 2.2 fmol/3 min fraction, n ¼ 15) than in those treated with the vehicle (13.2 ± 1. fmol/3 min fraction, n ¼ 1; F 1,23 ¼ 6.9, p ¼.1; Fig. 5d). The acute injection of saline did not modify basal [5-HT]ext during the following 18 min in the frontal cortex and the DRN in both GR25171-treated mice and paired vehicle-treated controls. Effects of acute administration of paroxetine A two-way ANOVA analysis of treatments (chronic challenge) on AUC values revealed significant main effects of chronic treatment factor (F 1,25 ¼ 11.3, p <.1; F 1,23 ¼ 6.9, p <.5) and challenge treatment factor (F 1,25 ¼ 16.1, p <.1; F 1,23 ¼ 15.6, p <.1) both in the frontal cortex and in the DRN (Figs 5c and f). In the frontal cortex, the challenge of paroxetine (1 mg/kg, i.p.) caused a significant increase in [5-HT]ext in GR treated mice (231 ± 23 vs. 113 ± 19%; p <.1) but not in paired controls that had been treated with the vehicle, in which the enhancing effect of paroxetine was too small to reach statistical significance (125 ± 1 vs. 11 ± 3%; p ¼.15; Fig. 5c). However, in the DRN, the challenge of paroxetine (1 mg/kg, i.p.) produced a significant increase in [5-HT]ext in both GR treated mice (165 ± 17 vs. 111 ± 5%; p <.1) and vehicle-treated control animals (124 ± 5 vs. 13 ± 5%; p <.5) (Fig. 5f). Thus, chronic treatment with the NK 1 receptor antagonist GR potentiated the enhancing effect of a challenge dose of paroxetine on [5-HT]ext, but this effect was higher in the frontal cortex than in the DRN (Figs 5c and f). Ó 25 International Society for Neurochemistry, J. Neurochem. (25) 95,

7 NK 1 receptor antagonism on 5-HT 1A receptor function 1719 Frontal Cortex Dorsal Raphe Nucleus (a) 3 (d) 3 saline extracellular 5-HT levels (fmol/15 min fraction) 2 saline (b) 3 2 Prx 1 extracellular 5-HT levels (fmol/3 min fraction) 2 1 (e) Prx Time (min) Time (min) AUC (% of basal values) (c) saline ** paroxetine chronic vehicle chronic GR AUC (% of basal values) (f) saline ** * paroxetine Fig. 5 Effects of long-term treatment with GR (1 mg/kg/ day for 21 days) on basal and paroxetine-affected outflow of 5-HT in the frontal cortex and the dorsal raphe nucleus of freely moving mice. Intracerebral in vivo microdialysis was started 2 h after removal of the osmotic mini pump delivering GR (m) or the vehicle (h). Data are expressed as [5-HT] ext in fmol/15 min fraction in the frontal cortex (a, n ¼ 23 25; b, n ¼ 5 11) and in fmol/3 min fraction in the dorsal raphe nucleus (d, n ¼ 1 15; e, n ¼ 5 1). Each point represents the mean ± SEM after acute saline (a and d) or paroxetine (1 mg/kg; i.p.) administration (b, e). (c, f) AUC values expressed as percentage of basal values. Each bar represents the mean ± SEM of [5-HT] ext measured from t to t18 min after saline or paroxetine administration. A two way-anova analysis of treatments (chronic challenge) followed by Fisher PLSD multiple comparison test was used to compare the effect of saline to paroxetine challenge in mice chronically treated with vehicle or GR 25171, in each brain region studied. *p <.5 and **p <.1, significantly different from the corresponding chronic vehicle- or GR treated group of mice receiving a challenge dose of saline; p <.5 and p <.1, significantly different from the corresponding chronic vehicle-treated group of mice receiving a challenge dose of paroxetine. Injection of the selective 5-HT 1A receptor antagonist, WAY 1635 (.5 mg/kg, s.c.), 15 min prior to the paroxetine challenge, potentiated the increasing effect of the latter drug on cortical [5-HT] ext in control mice that received the vehicle for 21 days, but not in GR treated mice (Fig. 6a). Statistical analysis showed that cortical AUC values in control mice treated acutely with WAY 1635 and paroxetine reached those obtained in GR treated mice (184 ± 17 vs. 23 ± 14%) (F 1,17 ¼.7; p ¼.4; Fig. 6b). The acute administration of WAY 1635 alone did not alter AUC values for the frontal cortex in these two groups of mice (data not shown). Discussion Because (i) 5-HT neurotransmission, and especially 5-HT 1A autoreceptors, are critically involved in the central mechanisms of action of several classes of antidepressant drugs (Blier and De Montigny 1983; Le Poul et al. 2), and (ii) NK 1 receptor antagonists are endowed with antidepressant properties (Kramer et al. 1998, 24), our objective was to assess whether 5-HT 1A autoreceptors might also be involved in the central effects of this novel class of potential antidepressants. Our first series of in vitro electrophysiological experiments clearly showed that the selective NK 1 receptor antagonist GR (Zocchi et al. 23) exerted no direct influence on DRN 5-HT neuron activity in mice, in line with previous data obtained with another NK 1 receptor antagonist, L-76735, in guinea pigs (Conley et al. 22). Together with immunocytochemical observations which demonstrated that NK 1 receptors are not expressed by DRN 5-HT cells in mice (Froger et al. 21), these electrophysiological results strongly suggest that, if it Ó 25 International Society for Neurochemistry, J. Neurochem. (25) 95,

8 172 B. P. Guiard et al. (a) extracellular 5-HT levels (fmol/15 min fraction) (b) AUC (% of basal values) WAY 1635 Prx Frontal Cortex Time (min) NS WAY paroxetine chronic vehicle chronic GR Fig. 6 Effect of challenge dose of paroxetine in combination with WAY 1635 on extracellular 5-HT levels in the frontal cortex of mice which received either GR (1 mg/kg/day) or its vehicle for 21 days. Intracortical in vivo microdialysis was started 2 h after removal of osmotic mini pump delivering GR (m) or the vehicle (h). (a) Data are expressed as [5-HT] ext in fmol/15 min fraction. Each point represents the mean ± SEM of nine to 1 determinations per group. (b) AUC values expressed as percentage of basal values. Each bar represents the mean ± SEM of [5-HT] ext measured from t to t18 min after WAY 1635 (.5 mg/kg, s.c)/paroxetine (1 mg/kg, i.p.) administration. WAY 1635 (first arrow) was administered 15 min prior to paroxetine (second arrow).anova followed by Fisher PLSD multiple comparison test was used to compare the effect of WAY 1635/paroxetine challenge in mice treated chronically with GR versus its vehicle. NS, not significant. exists, substance P regulation of 5-HT neuronal activity should be indirect. Our second series of in vitro electrophysiological experiments assessed the effects of acute in vivo administration of GR on 5-HT neuron firing and 5-HT 1A autoreceptor function in the DRN. The choice of GR was especially relevant for such experiments because this drug is one of the most potent, selective (Gardner et al. 1996), and centrally active (Goadsby et al. 1998; Rupniak et al. 23) NK 1 receptor antagonists in rodents. However, acute injection of the effective dose of 1 mg/kg s.c. of GR (Rupniak et al. 23; Hutson et al. 24) affected neither the basal firing rate nor the potency of the 5-HT 1A autoreceptor agonist, ipsapirone, to inhibit 5-HT neuron discharge in the DRN. In contrast, previous in vivo studies showed that acute administration of NK 1 receptor antagonists caused a delayed increase in DRN neuronal firing rate (Santarelli et al. 21; Conley et al. 22). However, investigations in rats led to the conclusion that such a treatment did not modify the 5-HT 1A autoreceptor-mediated inhibitory effect of 8-OH- DPAT on DRN 5-HT neuron firing in vivo (Haddjeri and Blier 2). All these data support the idea that acute NK 1 receptor blockade does not directly affect 5-HT 1A autoreceptor functioning. However, under in vivo conditions, indirect network-dependent mechanisms (which are obviously lost in brainstem slices) might account for the delayed activation of DRN neurons (Santarelli et al. 21; Conley et al. 22). In contrast, in mice that had been chronically treated with GR 25171, we found that the potency of ipsapirone to inhibit 5-HT neuron discharge in brainstem slices was clearly reduced, as expected from the occurrence of a functional desensitization of DRN 5-HT 1A autoreceptors. This observation agrees with previous data obtained in NK 1 receptor knock-out (NK 1 / ) mice (Froger et al. 21), although the amplitude of the adaptive change noted here after chronic blockade of NK 1 receptors was less than that reported after genetic deletion of NK 1 receptors. Previous studies in anaesthetized rats also showed such a desensitization of DRN 5-HT 1A autoreceptors, associated with an increase in basal firing rate of 5-HT neurons, after short (2 days) and long (14 days) term treatments with another NK 1 receptor antagonist, CP (Haddjeri and Blier 21). However, in guinea pigs, a 28-day treatment with L-76735, another potent NK 1 receptor antagonist, was found to increase in vivo basal firing rate of 5-HT neurons without modifying the functional characteristics of DRN 5-HT 1A autoreceptors (Conley et al. 22). These discrepancies might be explained, at least in part, by well-established species differences (rat and mouse vs. guinea pig) in the pharmacological properties of NK 1 receptors (see Conley et al. 22). In order to probe further 5-HT 1A autoreceptors, we measured 5-HT 1A autoreceptor-mediated binding of [ 35 S]GTP-c-S onto the DRN in brainstem sections. In both vehicle- and GR treated mice, the 5-HT 1 receptor agonist 5-CT produced a concentration-dependent increase in [ 35 S]GTP-c-S autoradiographic labelling, which could be completely prevented by WAY 1635, as expected from its mediation through 5-HT 1A autoreceptor activation (see also Fabre et al. 2; Pejchal et al. 22). Clearly, chronic administration of GR induced a shift to the right of the 5-CT concentration response curve, with a significant decrease in the maximal stimulation of [ 35 S]GTP-c-S binding, which could indicate that the number of G proteins to be recruited by 5-HT 1A receptor activation was reduced in GR treated compared with vehicle-treated mice. This alteration is similar to that previously found in NK 1 / mutants, where a reduction in the maximal 5-CT-evoked stimulation of [ 35 S]GTP-c-S binding onto the DRN was observed in comparison with NK 1 +/+wild-type mice (Froger Ó 25 International Society for Neurochemistry, J. Neurochem. (25) 95,

9 NK 1 receptor antagonism on 5-HT 1A receptor function 1721 et al. 21). Interestingly, it also corresponds to the adaptive change in 5-HT 1A autoreceptor-mediated [ 35 S]GTP-c-S binding already noted after 5-HT transporter (5-HTT) inactivation, either by pharmacological blockade with SSRI antidepressants (Pejchal et al. 22) or by gene deletion (Fabre et al. 2). Because 5-HT 1A autoreceptors mediate a negative feedback control on both synthesis and release of 5-HT (Hamon 1997), their desensitization occurring after chronic NK 1 receptor blockade may lead to an increased serotonergic neurotransmission, such as that found after chronic SSRI treatment (Bel and Artigas 1993) and in 5-HTT / knock-out mice (Fabre et al. 2). We directly assessed this inference using in vivo intracerebral microdialysis in freely moving animals. Indeed, we found no evidence for changes in basal cortical [5-HT] ext in mice after a 3-week treatment with GR Similarly, no differences in cortical 5-HT outflow were previously noted between NK 1 / and wildtype mice (Froger et al. 21). However, chronic GR treatment was found to almost double the basal [5-HT]ext levels in the DRN and to potentiate paroxetineinduced increase in [5-HT]ext levels in both the frontal cortex and the DRN. To date, microdialysis studies have yielded rather inconsistent results regarding the desensitization of somatodendritic 5-HT 1A autoreceptors following chronic treatment with SSRI. Several studies demonstrated increases in cortical [5-HT] ext levels after sustained treatments with these drugs (Bel and Artigas 1993; Invernizzi et al. 1994; Hervas et al. 21; Dawson et al. 22), but we found no changes in basal cortical 5-HT outflow following chronic paroxetine treatment in rats (Malagié et al. 2) and in mice (Gardier et al. 23), such as that found here after a long-term treatment with GR Although no data are available in mice, SSRI treatments have been shown to either increase (Malagie et al. 2) or leave unchanged (Bel and Artigas 1993) [5-HT] ext levels in the rat DRN. However, in the latter study, microdialysis experiments were performed when osmotic mini pumps were still in place (Bel and Artigas 1993). Here, in contrast, as no pharmacokinetic data concerning GR have yet been reported, pumps were removed 2 h before performance of our microdialysis experiments in order to manage a significant washout period. Interestingly, the selective 5-HT 1A receptor antagonist WAY 1635 (Fletcher et al. 1996), failed to alter paroxetine-induced increase in cortical [5-HT]ext levels in GR treated mice, further suggesting that 5-HT 1A autoreceptors were desensitized in the latter animals as in SSRI-treated rodents (see Le Poul et al. 2). Because our electrophysiological data suggest that NK 1 receptor-mediated regulation of 5-HT neuron activity is indirect, it can be inferred that the changes in 5-HT neurotransmission observed after long-term blockade of NK 1 receptors probably involved complex neuronal networks. Interestingly, Chen et al. (2) provided evidence that NK 1 receptors are expressed by noradrenergic neurons in the locus coeruleus, thus raising the possibility that substance P regulates the activity of these neurons and, indirectly, that of 5-HT neurons. In contrast, the lateral habenula and the amygdala, two regions which have close anatomo-functional relationships with DRN 5-HT neurons and are endowed with NK 1 receptors (Quartara and Maggi 1997; Peyron et al. 1998; Levita et al. 23), might also be involved in NK 1 receptor antagonist-induced modulations of 5-HT neurotransmission (see Conley et al. 22). Finally, local acting NK 1 receptor antagonists within the DRN should also be considered. Indeed, Liu et al. (22) found that substance P acts primarily on local glutamatergic neuronal afferents to produce excitatory postsynaptic currents in 5-HT neurons. In addition, NK 1 receptors have been clearly identified on GABAergic neurons surrounding 5-HT cell bodies in the DRN (Ma and Bleasdale 22), and their blockade might also contribute to enhanced 5-HT neurotransmission in mice chronically treated with GR All these hypotheses should be addressed in order to elucidate the mechanisms responsible for 5-HT 1A autoreceptor desensitization which probably contributes to the possible antidepressant action of chronic NK 1 receptor blockade, like that evidenced for SSRI antidepressants. Acknowledgements This research was supported by the Institut National de la Santé et de la Recherche Médicale (France) and European Community (6th FWP, contract no. LSHM-CT ). NF and BPG were recipients of a fellowship from the Ministère de l Education Nationale et de la Recherche (France). NF was also recipient of a fellowship from La Fondation pour la Recherche Médicale (FRM) during the performance of this work. We are grateful to pharmaceutical companies (GlaxoSmithKline, Sanofi-Synthélabo, Troponwerke-Bayer and Wyeth-Ayerst) for generous gifts of drugs. References Aghajanian G. K., Sprouse J. M. and Rasmussen K. (1987) Physiology of the midbrain serotonin system, in The Third Generation of Progress (Meltzer, H. Y., ed.), pp Raven Press, New York. Asberg M., Bertilsson L., Martensson B., Scalla-Tomba G. P., Thoren P. and Träksman-Bendz L. (1984) CSF monoamine metabolites in melancholia. Acta Psychiatr. Scand. 69, Bel N. and Artigas F. (1993) Chronic treatment with fluvoxamine increases extracellular serotonin in frontal cortex but not in raphe nuclei. Synapse 15, Blier P. and De Montigny C. (1983) Electrophysiological investigations on the effect of repeated zimelidine administration on serotoninergic neurotransmission in the rat. J. Neurosci. 3, Chen L. W., Wei L. C., Liu H. L. and Rao Z. R. (2) Noradrenergic neurons expressing substance P receptor (NK 1 ) in the locus coeruleus complex: a double immunofluorescence study in the rat. Brain Res. 873, Ó 25 International Society for Neurochemistry, J. Neurochem. (25) 95,

10 1722 B. P. Guiard et al. Conley R. K., Cumberbatch M. J., Mason G. S. et al. (22) Substance P (neurokinin 1) receptor antagonists enhance dorsal raphe neuronal activity. J. Neurosci. 22, Coppen A., Eccleston E. and Peet M. (1973) Total and free tryptophan concentration in the cerebrospinal fluid of depressed patients. Lancet 1, Dawson L. A., Nguyen H. Q., Smith D. L. and Schechter L. E. (22) Effect of chronic fluoxetine and WAY-1635 treatment on serotonergic neurotransmission in the frontal cortex. J. Psychopharmacol. 16, Delgado P. L., Charney D. S., Price L. H., Aghajanian G. K., Landis H. and Heninger G. R. (199) Serotonin function and the mechanism of antidepressant action: reversal of antidepressant-induced remission by rapid depletion of plasma tryptophan. Arch. Gen. Psychiatry 47, Ebner K., Rupniak N. M., Saria A. and Singewald N. (24) Substance P in the medial amygdala: emotional stress-sensitive release and modulation of anxiety-related behavior in rats. Proc. Natl Acad. Sci. USA 11, Fabre V., Beaufour C., Evrard A., Rioux A., Hanoun N., Lesch K. P., Murphy D. L., Lanfumey L., Hamon M. and Martres M. P. (2) Altered expression and functions of serotonin 5-HT 1A and 5-HT 1B receptors in knock-out mice lacking the 5-HT transporter. Eur. J. Neurosci. 12, Fletcher A., Forster E., Bill D. et al. (1996) Electrophysiological, biochemical, neurohormonal and behavioural studies with WAY- 1635, a potent, selective, and silent 5-HT 1A receptor antagonist. Behav. Brain Res. 73, Froger N., Gardier A. M., Moratalla R. et al. (21) 5-Hydroxytryptamine (5-HT) 1A autoreceptor adaptive changes in substance P (neurokinin 1) receptor knock-out mice mimic antidepressantinduced desensitization. J. Neurosci. 21, Gardier A. M., David D. J., Jego G. et al. (23) Effects of chronic paroxetine treatment on dialysate serotonin in 5-HT 1B receptor knockout mice. J. Neurochem. 86, Gardner C., Armour D., Beattie D., Gale J., Hawcock A., Kilpatrick G., Twissell D. and Ward P. (1996) GR 25171: a novel antagonist with high affinity for the tachykinin NK 1 receptor, and potent broad-spectrum anti-emetic activity. Regul. Pept. 65, Goadsby P. J., Hoskin K. L. and Knight Y. E. (1998) Substance P blockade with the potent and centrally acting antagonist GR does not effect central trigeminal activity with superior sagittal sinus stimulation. Neuroscience 86, Guiard B. P., Przybylski C., Guilloux J. P., Seif I., Froger N., De Felipe C., Hunt S. P., Lanfumey L. and Gardier A. M. (24) Blockade of substance P (neurokinin 1) receptors enhances extracellular serotonin when combined with a selective serotonin reuptake inhibitor: an in vivo microdialysis study in mice. J. Neurochem. 89, Haddjeri N. and Blier P. (2) Effect of neurokinin-1 receptor antagonist on the function of 5-HT and noradrenaline neurons. Neuroreport 11, Haddjeri N. and Blier P. (21) Sustained blockade of neurokinin-1 receptors enhances serotonin neurotransmission. Biol. Psychiatry 5, Haj-Dahmane S., Hamon M. and Lanfumey L. (1991) K + channel and 5 hydroxytryptamine 1A autoreceptor interactions in the rat dorsal raphe nucleus: an in vitro electrophysiological study. Neuroscience 41, Hamon M. (1997) The main features of central 5-HT 1A receptors, in Serotoninergic Neurons and 5-HT Receptors in the CNS, Handbook of Experimental Pharmacology (Baumgarten, H. G. and Göthert, M., eds), vol. 29, pp Springer-Verlag, Berlin. Hervas I., Vilaro M. T., Romero L., Scorza M. C., Mengod G. and Artigas F. (21) Desensitization of 5-HT(1A) autoreceptors by a low chronic fluoxetine dose: effect of the concurrent administration of WAY Neuropsychopharmacology 24, Hof P., Young W., Bloom F., Belichenko P. and Celio M. (2) Mouse Brains. Comparative Cytoarchitectonic Atlas of the C57BL/6 and 129/SV. Elsevier Science, Amsterdam. Hutson P. H., Patel S., Jay M. T. and Barton C. L. (24) Stressinduced increase of cortical dopamine metabolism: attenuation by a tachykinin NK 1 receptor antagonist. Eur. J. Pharmacol. 484, Invernizzi R., Bramante M. and Samanin R. (1994) Chronic treatment with citalopram facilitates the effect of a challenge dose on cortical serotonin output: role of presynaptic 5-HT 1A receptors. Eur. J. Pharmacol. 26, Kramer M. S., Cutler N., Feighner J. et al. (1998) Distinct mechanism for antidepressant activity by blockade of central substance P receptors. Science 281, Kramer M. S., Winokur A., Kelsey J. et al. (24) Demonstration of the efficacy and safety of a novel substance P (NK 1 ) receptor antagonist in major depression. Neuropsychopharmacology 29, Kreiss D. S. and Lucki I. (1995) Effects of acute and repeated administration of antidepressant drugs on extracellular levels of 5-hydroxytryptamine measured in vivo. J. Pharmacol. Exp. Ther. 274, Laitinen J. T. and Jokinen M. (1998) Guanosine 5 -(c-[ 35 S]thio) triphosphate autoradiography allows selective detection of histamine H3 receptor-dependent G protein activation in rat brain tissue sections. J. Neurochem. 71, Lanfumey L., Pardon M. C., Laaris N., Hanoun N., Joubert C. and Cohen-Salmon C. (1999) 5-HT 1A autoreceptor desensitization by chronic-ultra-mild-stress in mice. Neuroreport 1, Le Poul E., Boni C., Hanoun N., Laporte A. M., Laaris N., Chauveau J., Hamon M. and Lanfumey L. (2) Differential adaptation of brain 5-HT 1A and 5-HT 1B receptors and 5-HT transporter in rats treated chronically with fluoxetine. Neuropharmacology 39, Levita L., Mania I. and Rainnie D. G. (23) Subtypes of substance P receptor immunoreactive interneurons in the rat basolateral amygdala. Brain Res. 981, Liu R., Ding Y. and Aghajanian G. K. (22) Neurokinins activate local glutamatergic inputs to serotonergic neurons of the dorsal raphe nucleus. Neuropsychopharmacology 27, Ma Q. P. and Bleasdale C. (22) Modulation of brain stem monoamines and c-aminobutyric acid by NK 1 receptors in rats. Neuroreport 13, Malagié I., Trillat A. C., Douvier E., Anmella M. C., Dessalles M. C., Jacquot C. and Gardier A. M. (1996) Regional differences in the effect of the combined treatment of WAY 1635 and fluoxetine: an in vivo microdialysis study. Naunyn Schmiedebergs Arch. Pharmacol. 354, Malagié I., Deslandes A. and Gardier A. M. (2) Effects of acute and chronic tianeptine administration on serotonin outflow in rats: comparison with paroxetine by using in vivo microdialysis. Eur. J. Pharmacol. 43, Papp M., Vassout A. and Gentsch C. (2) The NK 1 -receptor antagonist NKP 68 has an antidepressant-like effect in the chronic mild stress model of depression in rats. Behav. Brain Res. 115, Pejchal J., Foley M. A., Kosofsky B. E. and Waeber C. (22) Chronic fluoxetine treatment selectively uncouples raphe 5-HT(1A) receptors as measured by [(35)S]-GTP-c-S autoradiography. Br. J. Pharmacol. 135, Ó 25 International Society for Neurochemistry, J. Neurochem. (25) 95,