Neuropathic Pain And The Id System In The Dorsal Raphe

European Journal of Neuroscience, Vol. 24, pp. 2011–2020, 2006

doi:10.1111/j.1460-9568.2006.05086.x

Neuropathic pain and the endocannabinoid system in the dorsal raphe: pharmacological treatment and interactions with the serotonergic system Enza Palazzo,1,* Vito de Novellis,1,* Stefania Petrosino,2 Ida Marabese,1 Daniela Vita,1 Catia Giordano,1 Vincenzo Di Marzo,2 Giuseppe Salvatore Mangoni,3 Francesco Rossi1 and Sabatino Maione1 1

Department of Experimental Medicine, Section of Pharmacology ‘L. Donatelli’, Faculty of Medicine and Surgery, The Second University of Naples, via Costantinopoli 16, 80138 Naples, Italy 2 Endocannabinoid Research Group, Institute of Biomolecular Chemistry, Consiglio Nazionale delle Ricerche, Pozzuoli, Italy 3 Department of Anaesthesiological and Surgical Sciences and Intensive Care, Faculty of Medicine and Surgery, The Second University of Naples, Naples, Italy Keywords: chronic constriction injury of the sciatic nerve, N-(4-hydroxyphenyl)-5Z, 8Z, 11Z, 14Z-eicosatetraenamide, (R)-(+)[2, 3-dihydro-5-methyl-3-(4-morpholinylmethyl)pyrrolo[1, 2, 3-de]-1, 4-benzoxazin-6-yl]-1-naphthalenylmethanone mesylate, rat, serotonin

Abstract We used a model of neuropathic pain consisting of rats with chronic constriction injury (CCI) of the sciatic nerve, in order to investigate whether endocannabinoid levels are altered in the dorsal raphe (DR) and to assess the effect of repeated treatment with (R)-(+)-[2,3-dihydro-5-methyl-3-(4-morpholinylmethyl)pyrrolo[1,2,3-de]-1,4-benzoxazin-6-yl]-1-naphthalenylmethanone mesylate, a synthetic cannabinoid agonist, or N-(4-hydroxyphenyl)-5Z,8Z,11Z,14Z-eicosatetraenamide (AM404), an inhibitor of endocannabinoid reuptake, on DR serotonergic neuronal activity and on behavioural hyperalgesia. CCI resulted in significantly elevated anandamide but not 2-arachidonoylglycerol levels in the DR. Furthermore, as well as thermal and mechanical hyperalgesia, CCI caused serotonergic hyperactivity (as shown by the increase of basal activity of serotonergic neurones, extracellular serotonin levels and expression of 5-HT1A receptor gene). Repeated treatment with either (R)-(+)-[2,3-dihydro-5-methyl-3-(4-morpholinylmethyl)pyrrolo[1,2,3-de]-1,4-benzoxazin-6-yl]-1-naphthalenylmethanone mesylate or AM404 reverted the hyperalgesia and enhanced serotonergic activity induced by CCI in a way attenuated by N-piperidino-5-(4-chlorophenyl)-1-(2,4dichlorophenyl)-4-methyl-3pyrazolecarboxamide, a selective cannabinoid subtype 1 (CB1) receptor antagonist. Despite the elevated levels of anandamide following CCI, N-piperidino-5-(4-chlorophenyl)-1-(2,4dichlorophenyl)-4-methyl-3-pyrazolecarboxamide did not produce hyperalgesia or any other effect on serotonergic neuronal activity when administered alone. Furthermore, the effects of AM404 were not accompanied by an increase in endocannabinoid levels in the DR. In conclusion, following CCI of the sciatic nerve, the endocannabinoid and serotonergic systems are activated in the DR, where repeated stimulation of CB1 receptors with exogenous compounds restores DR serotonergic activity, as well as thermal and mechanical nociceptive thresholds, to pre-surgery levels. However, an elevated level of endogenous anandamide in the DR does not necessarily contribute to the CB1-mediated tonic control of analgesia and serotonergic neuronal activity.

Introduction Recent studies have shown the effectiveness of cannabinoids in neuropathic pain (Herzberg et al., 1997; Costa et al., 2004, 2005; Scott et al., 2004; La Rana et al., 2006). The mechanisms by which cannabinoids reduce neuropathic pain remain unclear. Interactions between the endocannabinoid and other neurotransmitters have been reported (Di Marzo et al., 1998b; Schlicker & Kathmann, 2001), thus the effectiveness of cannabinoids in alleviating hyperalgesia could be due to their ability to modify major neurotransmitter systems playing a major role in the pathophysiology of chronic pain, such as the

Correspondence: Dr Sabatino Maione, as above. E-mail: [email protected] *E.P. and V.d.N. contributed equally to this work. Received 30 March 2006, revised 28 July 2006, accepted 31 July 2006

serotonergic system (Barkin & Fawcett, 2000; Mattia & Coluzzi, 2003). Tricyclic antidepressants blocking reuptake of monoamines are also effective in alleviating neuropathic pain (MacFarlane et al., 1997; Wang et al., 1997, 1999; Watson, 2000). It is known that serotonin (5HT) is involved in the inhibitory control of pain and that 5-HT antagonists block the analgesic effect of some antidepressants (Sindrup et al., 1990). The majority of neurones that contain 5-HT lie in the dorsal raphe (DR) (Dahlstrom & Fuxe, 1964). The role of the DR in modulating pain (Reynolds, 1969; Wang & Nakai, 1994; Cucchiaro et al., 2005) is probably mediated through its connection with the nucleus raphe magnus (Fields & Basbaum, 1978; Basbaum & Fields, 1984; Wang & Nakai, 1994) or due to direct projections from the DR to the spinal cord (Basbaum & Fields, 1979) and thalamus (Andersen & Dafny, 1983; Dong et al., 1991). So far, the DR has mainly been associated with the control of emotional states (Graeff

ª The Authors (2006). Journal Compilation ª Federation of European Neuroscience Societies and Blackwell Publishing Ltd

2012 E. Palazzo et al. et al., 1996). As a reciprocal relationship exists between persistent pain and negative affective states, as with the amygdala (Neugebauer et al., 2004), the DR might also be an important site for such interaction. To date there have been few reports on the effect of cannabinoids on 5-HT release (Nazaki et al., 2000; Tzavara et al., 2003) and none of them have monitored serotonergic cell activity in neuropathic pain following cannabinoid treatment. Furthermore, there has been no report on the activity of the endocannabinoid in the DR during neuropathic pain. The finding of elevated endocannabinoid levels with analgesic action in the DR would open the way to the use of agents inhibiting endocannabinoid inactivation (i.e. cellular reuptake and enzymatic hydrolysis) as possible antihyperalgesic agents (Di Marzo et al., 2004). With this background in mind, we considered that it was worth investigating changes in the DR serotonergic system in rats with chronic constriction injury (CCI) of sciatic nerve, before and after 7-day repeated treatment with (R)-(+)-[2,3-dihydro-5-methyl-3-(4-morpholinylmethyl)pyrrolo[1,2,3-de]-1,4-benzoxazin-6-yl]-1-naphthalenylmethanone mesylate (WIN55,212-2), a direct cannabinoid agonist, or N-(4hydroxyphenyl)-5Z,8Z,11Z,14Z-eicosatetraenamide (AM404), an inhibitor of anandamide cellular reuptake and inactivation. In particular, we performed in vivo and ex vivo experiments using single-unit electrophysiological recordings, microdialysis, semiquantitative analysis of the expression of 5-HT1A receptors and quantification of the levels of the endocannabinoids anandamide and 2-arachidonoylglycerol (2AG) in the DR. In addition, we evaluated the ability of WIN55,212-2 or AM404 treatments to relieve thermal and mechanical hyperalgesia associated with neuropathic pain.

Materials and methods Animals and treatment Male Wistar rats (Harlan, Italy) weighing 220–250 g were housed three per cage under controlled illumination (12-h light ⁄ 12-h dark cycle; light on 06.00 h) and standard environmental conditions (ambient temperature 20–22 C, humidity 55–60%) for at least 1 week before the commencement of experiments. Rat chow and tap water were available ad libitum. All surgery and experimental procedures were performed during the light cycle and were approved by the Animal Ethics Committee of The Second University of Naples. Animal care was in compliance with Italian (D.L. 116 ⁄ 92) and EC (O.J. of E.C. L358 ⁄ 1 18 ⁄ 12 ⁄ 86) regulations on the protection of laboratory animals. All efforts were made to reduce animal numbers. Neuropathic pain was induced by CCI of the sciatic nerve. Briefly, animals were anaesthetized with sodium pentobarbital (60 mg ⁄ kg i.p.), the right sciatic nerve was exposed and four ligatures were loosely tied around the nerve just proximal to the trifurcation. Control rats underwent a sham surgery with exposure of the sciatic nerve without ligature. WIN55,212-2 (0.1 mg ⁄ kg s.c.) and AM404 (10 mg ⁄ kg s.c.) alone or in combination with N-piperidino-5-(4chlorophenyl)-1-(2,4dichlorophenyl)-4-methyl-3-pyrazolecarboxamide (SR14176A) (1 mg ⁄ kg s.c.) and respective vehicle (10% dimethyl sulphoxide in 0.9% NaCl) were administered to sham and CCI groups of rats (n ¼ 10) for 7 days, starting the day after surgery (day 1). For acute dosing studies, a single AM404 (10 mg ⁄ kg s.c.) was administered 7 days after surgery.

Nociceptive behaviour Changes in thermoceptive responses were evaluated using plantar test apparatus (Ugo Basile, Varese, Italy). On the day of microdialytic

perfusate collection, each animal was simultaneously placed in a plastic cage (22 · 17 · 14 cm; length · width · height) with a glass floor. After a 1-h adaptation period, the plantar surface of the hindpaw was exposed to a beam of radiant heat through the glass floor. The radiant heat source consisted of an infrared bulb (Osram halogenbellaphot bulb; 8 V, 50 W). A photoelectric cell detected light reflected from the paw and turned off the lamp when paw movement interrupted the reflected light. The paw withdrawal latency was automatically displayed to the nearest 0.1 s; the cut-off time was 30 s in order to prevent tissue damage. The mechanical paw withdrawal threshold was measured by a dynamic plantar aesthesiometer (Ugo Basile). On the day of microdialytic perfusate collection, rats were allowed to move freely in one of the two compartments of the enclosure positioned on the metal mesh surface. Rats were adapted to the testing environment before any measurement was taken. The mechanical stimulus was delivered to the plantar surface of the hindpaw of the rat from below the floor of the test chamber by an automated testing device. A steel rod (2 mm) was pushed against the hindpaw with ascending force (1–30 g in 10 s). When the rat withdrew its hindpaw, the mechanical stimulus was automatically withdrawn and the force recorded to the nearest 0.1 g. Nociceptive responses (thermal paw withdrawal latency and mechanical paw withdrawal threshold) were measured in s and in g every 30 min (within the time for changing the perfusate samples) for 3 h and averaged in order to establish the baseline for each differently treated group of rats. For single vehicle or AM404 administrations, the thermal paw withdrawal latency and mechanical paw withdrawal threshold were measured every 30 min for 2 h before treatment and averaged in order to establish the pre-treatment baselines. The paw withdrawal latency and paw withdrawal threshold were subsequently monitored for a further 3 h. Groups of 10 rats per treatment were used, with each animal used for one treatment only.

Dorsal raphe single-unit extracellular recordings On the day of the experiment (day 7) rats were initially anaesthetized with chloral hydrate (400 mg ⁄ kg i.p.) and placed in steretaxic apparatus (David Kopf Instruments, Tujunga, CA, USA) with the animal warmed on a homeothermic temperature control blanket (Harvard Apparatus, Edembridge, Kent, UK). Subsequent i.v. anaesthetic maintenance injections were given through lateral tail vein cannulation. The electrocardiogram and electroencephalogram were recorded and monitored throughout the experiment via two screw electrodes fixed onto the skull to monitor the level of anaesthesia. The skull was exposed and a small craniotomy allowed the glass insulated tungsten filament electrodes (3–5 MW) (Frederick Haer and Co., Brunswick, ME, USA) to be lowered into the DR. Extracellular recordings were made from single neurones. Serotonergic neurones were identified by broad action potentials with positive–negative or positive–negative–positive deflections, broad action potential (2 ms), and regular and slow firing pattern (0.4–2.6 Hz firing rate), as described by Aghajanian & Vandermaelen (1982). The recorded signals were amplified and displayed on analogue and digital storage oscilloscopes. Signals were also fed into a window discriminator, whose output was processed by an interface (CED 1401, Cambridge Electronic Design Ltd, UK) connected to a Pentium III PC. spike2 software (version 4, Cambridge Electronic Design Ltd) was used to create peristimulus rate histograms on-line and to store and analyse digital records of single-unit activity off-line. The configuration, shape and height of the recorded action potentials were monitored and recorded continuously, using a window discriminator and spike2

ª The Authors (2006). Journal Compilation ª Federation of European Neuroscience Societies and Blackwell Publishing Ltd European Journal of Neuroscience, 24, 2011–2020

Cannabinoid and dorsal raphe changes in the neuropathic rat 2013 software for on-line and off-line analysis. This study only included neurones whose spike configuration remained constant and could clearly be discriminated from activity in the background throughout the experiment, indicating that the activity from one neurone only and from the same neurone was measured. Five rats per group were used with up to three neurones per rat recorded. At the end of the experiment, each animal was killed with a lethal dose of chloral hydrate and the recording site was marked with a 20-lA DC current for 20 s. The recording sites were identified after fixation by immersion in 10% formalin.

Microdialysis Microdialysis experiments were performed on awake and freely moving rats on day 7. In brief, on day 6 rats were anaesthetized with chloral hydrate (400 mg ⁄ kg i.p.) and concentric dialysis probes were implanted in the DR (7.8 mm posterior to bregma, 7.1 mm ventral to it and 0.2 mm lateral to the midline) (Paxinos & Watson, 1986) with stereotaxic apparatus. Dialysis probes were constructed with 25G (0.3 mm ID, 0.5 mm OD) stainless steel tubing (A-M Systems, Everett, WA, USA). Inlet and outlet cannulae (0.04 mm ID, 0.14 mm OD) consisted of fused silica tubing (Scientific Glass Engineering, Melbourne, Australia). The microdialysis probes had a 1.5-mm-long tubular dialysis membrane (Enka AG, Wuppertal, Germany). Animals were allowed to recover from surgery for 20–24 h and then probes were perfused with artificial cerebrospinal fluid (composition in mm: NaCl, 125; KCl, 2.5; MgCl2, 1.18 and CaCl2, 1.26) at a rate of 0.8 lL ⁄ min using an infusion pump (model 22, Harvard Apparatus). After an initial 60-min equilibration period, dialysate samples were collected every 30 min for 3 h to establish the baseline release of 5-HT. At the end of the experiments, all the rats were anaesthetized with pentobarbital and transcardially perfused with 0.9% NaCl solution followed by 10% formaldehyde solution. The brain was dissected out and fixed in a 10% formaldehyde solution for 48 h. The brain was cut in 40-lm-thick slices and observed in a light microscope to identify the probe tip. The concentration of 5-HT was determined using high performance liquid chromatography equipment fitted with an electrochemical detector. The composition of the mobile phase was 0.15 mm NaH2PO4, 0.01 mm octyl sodium sulphate, 0.5 mm EDTA (pH 3.8 adjusted with phosphoric acid) and 12.5% methanol. The mobile phase was delivered (flow rate 1 mL ⁄ min) by a model 590 pump (Waters Associates, Milford, MA, USA) into an Ultrasphere 3-lm octadecylsilane column (4.6 mm · 7.5 cm; Beckman Ltd, San Ramon, CA, USA). The electrochemical detector was an ESA Coulochem model 5100A with a dual electrode analytical cell (model 5011). The conditioning cell was set at )0.05 V, electrode 1 at +0.10 V and electrode 2 at +0.25 V with respect to palladium reference electrodes. The limit of detection for 5-HT was 2–3 fmol ⁄ sample injected with a signal-to-noise ratio of 2. The mean dialysate concentration of 5-HT in the six samples of each differently treated group represented the basal release of 5-HT (fmol ⁄ 20 lL). In vitro recovery of the microdialysis probe for 5-HT was 23–25%.

RNA extraction and reverse transcription-polymerase chain reaction Rats were decapitated and their brains rapidly removed and immersed in ice-cold artificial cerebrospinal fluid. A block of tissue containing the DR was cut using a vibrotome (Vibratome 1500, Warner

Instruments, Hamden, CT, USA). A brainstem slice of 1.2 mm was cut throughout the rostral part of the DR (interaural from +1.9 mm to +0.7 mm, Paxinos & Watson, 1986) and the DR was then isolated under optical microscope (M650, Wild Heerbrugg, Switzerland), homogenized and total RNA extracted using an RNA Tri-Reagent (Molecular Research Center Inc., Cincinnati, OH, USA) according to the manufacturer’s protocol. The extracted RNA was subjected to DNase I treatment at 37 C for 30 min. The total RNA concentration was determined by UV spectrophotometer. The mRNA levels of the genes for 5-HT1A were measured by reverse transcription-polymerase chain reaction amplification, as previously reported (Galderisi et al., 1999). Sequences for rat 5-HT1A mRNAs from GeneBank (DNASTAR Inc., Madison, WI, USA) were used to design primer pairs for reverse transcription-polymerase chain reactions (oligo 4.05 software, National Biosciences Inc., Plymouth, MN, USA). Each reverse transcription-polymerase chain reaction was repeated at least four times. A semiquantitative analysis of mRNA levels was carried out by the Gel Doc 2000 UV System (Bio-Rad, Hercules, CA, USA). The measured mRNA levels were normalized with respect to hypoxanthine-guanine phosphoribosyltransferase, chosen as housekeeping gene, and the gene expression values were expressed as arbitrary units ± SEM.

Endocannabinoid extraction and quantification Procedure of extraction The DR from treated sham or CCI rats was extracted immediately after killing as described above. The tissue was homogenized in 5 vol of chloroform ⁄ methanol ⁄ Tris HCl 50 mm (2 : 1 : 1) containing 100 pmol of d8-anandamide and d8-2-AG. Deuterated standards were synthesized from d8 arachidonic acid and ethanolamine or glycerol as described, respectively, in Devane et al. (1992) and Bisogno et al. (1997). Homogenates were centrifuged at 13 000 g for 16 min (4 C) and the aqueous phase plus debris was collected and extracted again twice with 1 vol of chloroform. The organic phases from the three extractions were pooled and the organic solvents evaporated in a rotating evaporator. Lyophilized samples were then stored frozen at )80 C under nitrogen atmosphere until analysed. Analysis of endocannabinoid contents Lyophilized extracts were resuspended in chloroform ⁄ methanol (99 : 1 by volume). The solutions were then purified by open bed chromatography on silica as described in Bisogno et al. (1997). Fractions eluted with chloroform ⁄ methanol (9 : 1 by volume) (containing anandamide and 2-AG) were collected, the excess solvent evaporated with a rotating evaporator and aliquots analysed by isotope dilution-liquid chromatography ⁄ atmospheric pressure chemical ionization ⁄ mass spectrometry carried out under the conditions previously described (Marsicano et al., 2003) and allowing the separation of 2AG and anandamide. Mass spectrometry detection was carried out in the selected ion monitoring mode using m ⁄ z-values of 356 and 348 (molecular ions +1 for deuterated and undeuterated anandamide), 384.35 and 379.35 (molecular ions +1 for deuterated and undeuterated 2-AG). The area ratios between signals of deuterated and undeuterated anandamide varied linearly with varying amounts of undeuterated anandamide (30 fmol)100 pmol). The same applied to the area ratios between signals of deuterated and undeuterated 2-AG in the 100 pmol)20 nmol interval. Anandamide and 2-AG levels in unknown samples were therefore calculated on the basis of their area ratios with the internal deuterated standard signal areas. The amounts of endocannabinoids were expressed as pmol or nmol ⁄ g of wet tissue

ª The Authors (2006). Journal Compilation ª Federation of European Neuroscience Societies and Blackwell Publishing Ltd European Journal of Neuroscience, 24, 2011–2020

2014 E. Palazzo et al. extracted and were compared by anova followed by the Bonferroni test.

Drugs The WIN55,212-2 and AM404 were purchased from Tocris Cookson Ltd (Bristol, UK). SR141716A was kindly provided by SanofiSynthelabo Recherche (Montpellier, France). These drugs were dissolved in 10% dimethyl sulphoxide in 0.9% NaCl.

Statistics Statistical analysis of the data was performed by anova at each point using one-way anova followed by Student’s t-test. For molecular data anova followed by Student-Newman-Keuls post-hoc test was used to determine the statistical significance between groups or pre- and posttreatment values for acute administrations. For the levels of endocannabinoids anova followed by the Bonferroni tests was used. P < 0.05 was considered statistically significant.

Results Endocannabinoid levels We examined the levels of the cannabinoid subtype 1 (CB1)-selective endocannabinoid, 2-AG, and of the endogenous transient receptor potential vanilloid type 1 (TRPV1) ⁄ CB1 ‘hybrid’ agonist, anandamide, in sham and CCI rats following repeated treatments with vehicle (10% dimethyl sulphoxide in 0.9% NaCl) or AM404 (10 mg ⁄ kg s.c.) for 7 days starting the day after surgery. For acute studies, a single administration of vehicle (10% dimethyl sulphoxide in 0.9% NaCl) or AM404 (10 mg ⁄ kg s.c.) was performed in CCI rats 7 days after surgery and 30 min before the dissection of the DR. The DR content of anandamide was 57.1 ± 7.3 pmol ⁄ g whereas the levels of 2-AG were 5.9 ± 0.8 nmol ⁄ g in sham rats receiving 7 days of treatment with vehicle. The DR content of endocannabinoids was not modified by 7day treatment with AM404 (10 mg ⁄ kg s.c.) in sham rats (anandamide, 42.9 ± 4.5 pmol ⁄ g; 2-AG, 5.0 ± 0.2 nmol ⁄ g). CCI of the sciatic nerve induced a significant enhancement of anandamide levels (145.1 ± 10 pmol ⁄ g, P < 0.001) but not of 2-AG levels (4.3 ± 0.4 nmol ⁄ g). In CCI rats, repeated 7-day treatment with AM404 (10 mg ⁄ kg s.c.) did not significantly modify endocannabinoid contents (anandamide, 153.4 ± 3.0 pmol ⁄ g; 2-AG, 3.0 ± 0.3 nmol ⁄ g). Interestingly, however, acute administration of AM404 did produce a significant increase in 2-AG, but not anandamide, levels in CCI rats (from 2.4 ± 0.2 to 3.7 ± 0.4 nmol ⁄ g, P < 0.05 for 2-AG and from 158.2 ± 21.2 to 182.8 ± 21.8 pmol ⁄ g, P > 0.005 for anandamide).

Nociceptive behaviour Thermal hyperalgesia The basal (pre-surgery) thermal withdrawal latency was 11.2 ± 0.6 s (mean ± SEM). Seven days of treatment with vehicle (10% dimethyl sulphoxide in 0.9% NaCl) did not change thermal withdrawal latency (10.8 ± 0.8 s) in sham-operated rats compared with the naives. CCI of the sciatic nerve significantly reduced thermal withdrawal latency (4.3 ± 1.1 s) in rats receiving 7 days of treatment with vehicle. Seven days of treatment with WIN55,212-2 (0.1 mg ⁄ kg s.c.) or AM404 (10 mg ⁄ kg s.c.) did not change thermal withdrawal latency in the shams (10.6 ± 0.7 and 10.3 ± 0.5 s, respectively). Seven days of treatment with WIN55,212-2 (0.1 mg ⁄ kg s.c.) or AM404 (10 mg ⁄ kg

s.c.) reduced thermal hyperalgesia (9.9 ± 0.8 and 10.5 ± 0.8 s, respectively) in the neuropathic rats. These effects were antagonized when SR141716A (1 mg ⁄ kg s.c.) was administered daily in combination with WIN55,212-2 or AM404 (4.7 ± 0.5 and 6.6 ± 0.5 s, respectively) (Fig. 1A). Seven days of treatment with SR141716A (1 mg ⁄ kg s.c.) did not modify thermal withdrawal latency in either the shams or CCI rats (9.7 ± 0.5 and 8.9 ± 0.9 s). No changes in thermal withdrawal latency were observed in the contralateral paw in CCI rats, with or without any treatment (data not shown). We also measured the effect of a single administration of vehicle or AM404 (10 mg ⁄ kg s.c.) on the thermal withdrawal latency 7 days after the sham or CCI surgery. A single injection of vehicle did not change thermal withdrawal latency (pre-treatment, 10.5 ± 0.6 s; 30 min post-treatment, 11.2 ± 0.8 s) in the shams. A single injection of AM404 did not change thermal withdrawal latency in the shams (pre-treatment, 11.0 ± 0.5 s; 30 min post-treatment, 12.2 ± 1.5 s). A single injection of vehicle did not modify thermal hyperalgesia in CCI rats (pretreatment, 4.5 ± 0.7 s; 30 min post-treatment, 4.2. ± 0.9 s). A single administration of AM404 (10 mg ⁄ kg s.c.) relieved thermal hyperalgesia (pre-treatment, 4.5 ± 0.7 s; 30 min post-treatment, 6.8 ± 0.3 s) when administered on day 7 after ligation of the sciatic nerve. The maximum antihyperalgesic effect induced by AM404 in CCI rats was already apparent 30 min after administration and was observable up to the end of the experimentation (3 h). Mechanical allodynia The basal (pre-surgery) mechanical withdrawal threshold was 27.7 ± 2.3 g (mean ± SEM). Seven days of treatment with vehicle did not change the mechanical withdrawal threshold (26.3 ± 3.1 g) in the sham-operated rats compared with the naives. CCI of the sciatic nerve significantly reduced the mechanical withdrawal threshold (13.2 ± 3 g) in rats receiving 7 days of treatment with vehicle. Seven days of treatment with WIN55,212-2 (0.1 mg ⁄ kg s.c.) or AM404 (10 mg ⁄ kg s.c.) did not change the mechanical threshold in the shams (28.0 ± 2 and 26.2 ± 3 g, respectively). Seven days of treatment with WIN55,212-2 (0.1 mg ⁄ kg s.c.) or AM404 (10 mg ⁄ kg s.c.) reverted the mechanical allodynia (25.3 ± 2 and 26.9 ± 3 g, respectively) in the neuropathic rats. These effects were antagonized when SR141716A (1 mg ⁄ kg s.c.) was administered daily in combination with WIN55,212-2 (0.1 mg ⁄ kg s.c.) or AM404 (10 mg ⁄ kg s.c.) (14.7 ± 0.5 and 19.2 ± 1 g, respectively) (Fig. 1B). Seven days of treatment with SR141716A (1 mg ⁄ kg s.c.) did not affect the mechanical threshold in either the sham or CCI rats (26.5 ± 1.5 and 13.2 ± 1.1 g, respectively). No changes in the mechanical withdrawal threshold were observed in the contralateral paw in CCI rats, with or without any treatment (data not shown). When administered on day 7 after surgery, a single injection of vehicle did not change the mechanical withdrawal threshold (pre-treatment, 26.5 ± 1.9 g; 30 min post-treatment, 28.2 ± 1.2 g) in the shams. A single injection of AM404 did not change the mechanical withdrawal threshold in the shams (pre-treatment, 27.0 ± 1.5 s; 30 min post-treatment, 27.9 ± 1.5 g). A single injection of vehicle did not modify the mechanical allodynia in CCI rats (pre-treatment, 14.3 ± 2.7 g; 30 min post-treatment, 15.9 ± 2.9 g). A single administration of AM404 (10 mg ⁄ kg s.c.) relieved the mechanical allodynia (pre-treatment, 13.9 ± 0.7 g; 30 min post-treatment, 21.6 ± 1.3 g) when administered on day 7 after ligation of the sciatic nerve. Single-unit electrophysiological recordings We studied the basal ⁄ spontaneous activity of DR neurones. A total of 149 neurones in the DR were recorded from 55 rats. The majority of

ª The Authors (2006). Journal Compilation ª Federation of European Neuroscience Societies and Blackwell Publishing Ltd European Journal of Neuroscience, 24, 2011–2020

Cannabinoid and dorsal raphe changes in the neuropathic rat 2015

Fig. 1. Effects of 7-day treatment with vehicle (veh) (10% dimethyl sulphoxide in 0.9% NaCl s.c.), (R)-(+)-[2,3-dihydro-5-methyl-3-(4-morpholinylmethyl)pyrrolo[1,2,3-de]-1,4-benzoxazin-6-yl]-1-naphthalenylmethanone mesylate (WIN) (0.1 mg ⁄ kg s.c.) or N-(4-hydroxyphenyl)-5Z,8Z,11Z,14Z-eicosatetraenamide (AM) (10 mg ⁄ kg s.c.) alone or in combination with N-piperidino-5-(4-chlorophenyl)-1-(2,4dichlorophenyl)-4-methyl-3-pyrazolecarboxamide (SR) (1 mg ⁄ kg s.c.) on thermal withdrawal latency (A) and mechanical withdrawal threshold (B) in sham or chronic constriction injury (CCI) rats. Each point represents the mean ± SEM of 10 animals per group. *P < 0.05 vs. sham ⁄ veh; +P < 0.05 vs. CCI ⁄ veh; and #P < 0.05 vs. CCI ⁄ WIN or AM.

these neurones, encountered at a depth of 5.65 and 6.5 mm from the surface of the brain, had a spontaneous firing activity in the range of 0.4–2.6 Hz and a biphasic action potential of 1.9–2.1 ms (Fig. 2, upper left panel). These cells corresponded to 67% of the total cells found. These cells fit in with the classic properties of serotonergic DR neurones (slow rhythmic activity in spontaneously active cells, broad action potential) as previously described by Aghajanian & Vandermaelen (1982). Seven days of treatment with vehicle did not alter the firing activity of DR neurones in the sham-operated rats compared with the naives (2.5 ± 0.8 vs. 2.6 ± 0.9 spikes ⁄ s) and nor did it change the percentage (65%) of slow and rhythmic cell activity. CCI of the sciatic nerve determined a massive, significant increase in DR firing activity in rats receiving daily vehicle treatment; 73% of recorded neurones showed a higher but nevertheless rhythmic firing activity (7.8 ± 1.1 spikes ⁄ s). Although these cells showed higher frequency, they kept a broad action potential (2 ms) and a regular firing pattern. Likewise, 7 days of treatment with WIN55,212-2 (0.1 mg ⁄ kg s.c.) or AM404 (10 mg ⁄ kg s.c.) significantly increased DR firing activity (6.3 ± 0.9 and 5.9 ± 0.5 spikes ⁄ s, respectively) in the shams. Neurones from rats receiving 7 days of treatment with WIN55,212-2 or AM404 showed increased spontaneous firing activity of up to 71 and 69%, respectively. These effects were antagonized when SR141716A (1 mg ⁄ kg s.c.) was administered daily in combination with WIN55,212-2 (0.1 mg ⁄ kg s.c.) or AM404 (10 mg ⁄ kg

s.c.) (3.1 ± 0.7 and 4.4 ± 0.1 spikes ⁄ s, respectively). WIN55,212-2 (0.1 mg ⁄ kg s.c.) or AM404 (10 mg ⁄ kg s.c.) reduced firing activity in CCI rats to 3.1 ± 0.5 and 3.3 ± 0.3 spikes ⁄ s, respectively. The effects of repeated treatment with WIN55,212-2 or AM404 in CCI rats were antagonized when SR141716A (1 mg ⁄ kg s.c.) was administered daily in combination with cannabinoids (7.5 ± 0.5 and 5.5 ± 0.6 spikes ⁄ s, respectively) (Figs 2 and 3A). Seven days treatment with SR141716A (1 mg ⁄ kg s.c.) did not modify DR cell spontaneous activity in either the shams or CCI animals (data not shown). Finally, we measured the effect of a single administration of vehicle or AM404 (10 mg ⁄ kg s.c.) on DR cell spontaneous activity in either the shams or CCI animals. When administered on day 7 after surgery, a single injection of vehicle did not change DR serotonergic firing (pre-treatment, 2.3 ± 0.4 spikes ⁄ s; 30 min post-treatment, 2.5 ± 0.5 spikes ⁄ s) in the shams. A single injection of AM404 (10 mg ⁄ kg s.c.) was unable to significantly increase DR serotonergic firing activity (pre-treatment, 2.4 ± 0.7 spikes ⁄ s; 30 min post-treatment, 3.2 ± 0.8 spikes ⁄ s) in the shams. A single injection of vehicle did not modify the increased serotonergic firing activity observed in CCI rats (pre-treatment, 7.7 ± 0.7 spikes ⁄ s; 30 min post-treatment, 7.2. ± 0.3 spikes ⁄ s). A single administration of AM404 (10 mg ⁄ kg s.c.) did reduce DR serotonergic firing (pre-treatment, 7.9 ± 0.5 spikes ⁄ s; 30 min post-treatment, 5 ± 0.4 spikes ⁄ s) when administered on day 7 after ligation of the sciatic nerve.

ª The Authors (2006). Journal Compilation ª Federation of European Neuroscience Societies and Blackwell Publishing Ltd European Journal of Neuroscience, 24, 2011–2020

2016 E. Palazzo et al.

Fig. 2. A regular slow-firing putative serotonin (5-HT) neurone is shown at the top. The bar indicates a period of 2 ms. Integrated firing rate histograms (in spikes ⁄ 10 s) show the effects of 7-day treatment with vehicle (10% dimethyl sulphoxide in 0.9% NaCl s.c.) and (R)-(+)-[2,3-dihydro-5-methyl-3-(4morpholinylmethyl)pyrrolo[1,2,3-de]-1,4-benzoxazin-6-yl]-1-naphthalenylmethanone mesylate (WIN) (0.1 mg ⁄ kg s.c.) alone or in combination with N-piperidino5-(4-chlorophenyl)-1-(2,4dichlorophenyl)-4-methyl-3-pyrazolecarboxamide (SR) (1 mg ⁄ kg s.c.) on the spontaneous firing of dorsal raphe 5-HT neurones in sham and chronic constriction injury (CCI) rats. Briefly, CCI ⁄ vehicle rats show increased firing activity compared with sham ⁄ vehicle rats. Repeated chronic treatment with WIN (0.1 mg ⁄ kg s.c.) reduced the firing rate and this last effect was antagonized when SR was administered in combination with WIN.

Microdialysis In order to further ascertain the involvement of serotonergic cells in CCI-induced electrophysiological changes, we performed in vivo microdialysis experiments in the DR. The mean basal value (not corrected for probe recovery) of extracellular 5-HT levels in the DR was 27 ± 6 fmol ⁄ 20 lL. Seven days of treatment with vehicle did not change DR extracellular 5-HT (30.5 ± 5 fmol ⁄ 20 lL) in the shamoperated rats compared with the naives. CCI of the sciatic nerve determined a massive, significant increase in DR 5-HT (91.5 ± 7 fmol ⁄ 20 lL) in rats receiving 7 days of treatment. Seven days of treatments with WIN55,212-2 (0.1 mg ⁄ kg s.c.) or AM404 (10 mg ⁄ kg s.c.) significantly increased DR 5-HT release (79.1 ± 6 and 72 ± 6 fmol ⁄ 20 lL, respectively) in the shams. These effects were antagonized when SR141716A (1 mg ⁄ kg s.c.) was administered daily in combination with WIN55,212-2 or AM404 (26.6 ± 6 and 34.1 ± 4 fmol ⁄ 20 lL). Seven days of treatment with WIN55,212-2 (0.1 mg ⁄ kg s.c.) or AM404 (10 mg ⁄ kg s.c.) reduced the CCI-induced increase of DR 5-HT to 31.6 ± 11 and 36.2 ± 4 fmol ⁄ 20 lL. The latter effects were antagonized when SR141716A (1 mg ⁄ kg s.c.) was administered daily in combination with WIN55,212-2 or AM404 (89 ± 9 and 68.2 ± 6 fmol ⁄ 20 lL, respectively) (Fig. 3B). Seven days treatment with SR141716A (1 mg ⁄ kg s.c.) did not modify DR 5HT in either the shams or CCI animals (data not shown).

5-HT1A receptor gene expression The semiquantitative analysis of mRNA levels within the DR measured by reverse transcription-polymerase chain reaction amplification showed a dramatic over-expression (170%) of the 5-HT1A receptor gene (mean ± SE of arbitrary units, 1.77 ± 0.21) in the neuropathic rats compared with the sham rats (mean ± SE of arbitrary units, 0.65 ± 0.02). Seven days of treatment with WIN55,212-2

(0.1 mg ⁄ kg s.c.) or AM404 (10 mg ⁄ kg s.c.) significantly decreased (52%) the DR 5-HT1A receptor gene expression in CCI rats to 0.92 ± 0.09 and 0.89 ± 0.10 (mean ± SE of arbitrary units). These effects were antagonized when SR141716A was administered in combination with WIN55,212-2 or AM404 (mean ± SE of arbitrary units, 1.75 ± 0.03 and 1.52 ± 0.01, respectively).

Discussion In the present study we investigated the possible role of the endocannabinoid system in the control of serotonergic signalling in the DR of neuropathic rats. The first finding of the study was that CCI of the sciatic nerve is accompanied by considerable enhancement of anandamide (although not 2-AG) levels in this brainstem region. Evidence of an increase in the release of anandamide following pain has been reported for the periaqueductal grey (Walker et al., 1999) but had never been shown before within the DR, and never in relation to neuropathic pain. In view of the analgesic effects induced by endocannabinoids at supra-spinal, spinal and peripheral levels (Iversen & Chapman, 2002; Walker et al., 2005, for reviews), we hypothesized that this rise in anandamide levels in the DR could be an adaptive response aimed at producing analgesic effects, possibly by influencing serotonergic signalling in the DR. Therefore, we subsequently analysed the effect of CCI of the sciatic nerve on the activity of DR neurones, and on 5-HT release and 5-HT1A receptor expression in this region. We found that neuropathic rats exhibit an increase of the firing rate of a large number of cells (67%) recorded in the DR. This brain region contains approximately one- to two-thirds of 5-HT neurones (Descarries et al., 1982; Jacobs & Azmitia, 1992). The finding of an increase in spontaneous neuronal firing of still regular and rhythmic neuronal firing activity, as well as of a broad neuronal action potential (in agreement with the classic

ª The Authors (2006). Journal Compilation ª Federation of European Neuroscience Societies and Blackwell Publishing Ltd European Journal of Neuroscience, 24, 2011–2020

Cannabinoid and dorsal raphe changes in the neuropathic rat 2017

Fig. 3. Effects of 7-day treatment with vehicle (veh) (10% dimethyl sulphoxide in 0.9% NaCl s.c.), (R)-(+)-[2,3-dihydro-5-methyl-3-(4-morpholinylmethyl)pyrrolo[1,2,3-de]-1,4-benzoxazin-6-yl]-1-naphthalenylmethanone mesylate (WIN) (0.1 mg ⁄ kg s.c.) or N-(4-hydroxyphenyl)-5Z,8Z,11Z,14Z-eicosatetraenamide (AM) (10 mg ⁄ kg s.c.) alone or in combination with N-piperidino-5-(4-chlorophenyl)-1-(2,4dichlorophenyl)-4-methyl-3-pyrazolecarboxamide (SR) (1 mg ⁄ kg s.c.) on the neuronal spontaneous firing (A) and release of serotonin (5-HT) (B) of the dorsal raphe in sham and chronic constriction injury (CCI) rats. Each point represents the mean ± SEM of five or 10 animals per group for electrophysiological and microdialysis experiments, respectively. *P < 0.05 vs. sham ⁄ veh; •P < 0.05 vs. sham ⁄ WIN or AM; +P < 0.05 vs. CCI ⁄ veh; and #P < 0.05 vs. CCI ⁄ WIN or AM.

serotonergic cell properties, see Aghajanian & Vandermaelen, 1982), leads us to reason that these cells might be hyperactive serotonergic cells. Nevertheless, the possibility that different populations of neurones become spontaneously active following CCI of the sciatic nerve cannot be excluded. Indeed, there is evidence that phenotype changes in rostral ventromedial medulla neutral cells occur after chronic inflammation as a possible mechanism to counterbalance pain (Miki et al., 2002). In agreement with the possible involvement of hyperactive serotonergic neurones, our microdialysis experiments indicate that 7-day CCI of the sciatic nerve also induced an increase in DR 5-HT release. Changes in extracellular 5-HT release following chronic pain have scarcely been investigated in the DR (Palazzo et al., 2004). Previous evidence showed that chronic noxious stimuli increased 5-HT release in the periaqueductal grey (Zhang et al., 2000), an area with morphological properties similar to the DR. Moreover, it has been shown that chronic pain enhances the activity of DR neurones (Sanders et al., 1980; Porro et al., 1991). The DR contains intrinsic neuronal circuitry in which serotonergic neurones are inhibited by GABAergic interneurones, which are in turn inhibited by opioidergic neurones (Wang & Nakai, 1994; Zhang et al., 2000). Apart from opioids, endocannabinoids inhibit GABAergic synaptic transmission, producing analgesia through distinct, partially overlapping, disinhibitory mechanisms in the rostral ventromedial medulla (Meng et al., 1998) and periaqueductal grey (Vaughan et al., 2000). Thus, the increase of DR anandamide content during neuropathic pain

could also be responsible for a disinhibitory action on neurones at the DR level, and this might be a naturally occurring mechanism to counteract the establishment of pain, e.g. via serotonergic-mediated antinociception. We also observed that the expression of 5HT1A receptor mRNA in the DR significantly increased 7 days after CCI of the sciatic nerve. A similar observation was previously reported in the DR under conditions of carrageenan-induced chronic inflammatory pain by Zhang et al. (2000). DR 5-HT1A receptors are mainly autoreceptors, which negatively regulate 5-HT release. It is therefore reasonable to assume that the increased expression of 5HT1A receptor within the DR may be required to enhance the feedback regulation of cell activity as a mechanism to counterbalance the CCI-induced neural over-activity described above. In order to substantiate the hypothesis of a role of the endocannabinoid system in the control of serotonergic signalling, we analysed the effect of pharmacological stimulation or blockade of CB1 receptors on the activity of DR neurones, 5-HT release and 5-HT receptor expression in the DR. Doses of cannabinoids used in the study were chosen according to previous studies in which repeated treatments with cannabinoids were carried out in neuropathic rats (Costa et al., 2004; Rodella et al., 2005; La Rana et al., 2006). It has been reported (Costa et al., 2004; La Rana et al., 2006) and further confirmed in this study that these doses do not cause any overt change in animal behaviour. In sham-operated rats, the cannabinoid receptor agonist WIN55,212-2 enhanced both DR neuronal firing and 5-HT release

ª The Authors (2006). Journal Compilation ª Federation of European Neuroscience Societies and Blackwell Publishing Ltd European Journal of Neuroscience, 24, 2011–2020

2018 E. Palazzo et al. without changing behavioural nociceptive responses (Herzberg et al., 1997; Costa et al., 2004). In agreement with these studies, cannabinoid treatment did not change the pain threshold in the shams. However, the same treatment was able to enhance DR cell activity, suggesting that there does not seem to be a direct correlation between changes in spontaneous DR activity and the thermal and mechanical pain thresholds in the sham animal. The increased DR cell activity that we observed supports the possible disinhibitory and GABA-mediated effect of CB1 receptor stimulation on serotonergic signalling in this region, as hypothesized above. The alternative possibility of inhibition of 5-HT uptake by WIN55,212-2, via a receptor-independent mechanism, with a subsequent massive increase in 5-HT levels, has been proposed in the neocortex by experiments carried out in vitro by Steffens & Feuerstein (2004) and this effect could also be important within the DR, where the 5-HT transporter is located on cell bodies and dendrites of serotonergic neurones (Dahlstrom & Fuxe, 1964). However, we found here that the effects of WIN55,212-2 on both neuronal firing and 5-HT release in the DR were entirely dependent on CB1 receptors, as they were fully antagonized by SR141716A, which was inactive per se. Importantly, unlike sham-operated rats, neuropathic rats responded to 7-day treatment with WIN55,212-2 with reduced DR neurone firing activity and 5-HT extracellular release, as well as reduced 5HT1A receptor mRNA expression. All these parameters were brought back to the normal physiological levels, i.e. to levels identical to those found in pre-surgery conditions, and these effects were accompanied by a complete reversal of behavioural hyperalgesia (both thermal hyperalgesia and mechanical allodynia), in agreement with previous studies (Costa et al., 2004; La Rana et al., 2006). Thus, this study shows a different effect of repeated treatment with cannabinoid in sham and CCI rats, i.e. an increase in DR cell activity with no change in pain threshold and a normalization of DR cell activity with a complete reversal of hyperalgesia, respectively. This finding suggests that neural DR circuitry is not directly involved in pain modulation, whereas it may play a major role in the integration of emotional states. Indeed, it seems reasonable to assume that changes in DR activity may reflect a highly compromised emotional state during neuropathic pain conditions. All the actions of WIN55,212-2 were antagonized by coadministration with SR141716A, which exerted no effect when given alone throughout the development of neuropathic pain. The lack of any effect of SR141716A would appear to indicate against a possible tonic inhibition of serotonergic signalling by elevated anandamide in the DR during neuropathic pain, as in this case we should have observed an enhancement or non-effect of WIN55,212-2 on neuronal firing and 5-HT release, and a reduction of these parameters with SR141716A alone. By contrast, as the CB1-mediated restoration of DR serotonergic signalling produced by WIN55,212-2 was accompanied by a strong CB1-mediated antihyperalgesic action, we hypothesize that

enhanced 5-HT might be an adaptive response to neuropathic pain in the DR and that CB1 stimulation, by producing a strong antihyperalgesic effect, normalizes serotonergic signalling in this region of the brain. In this sense, the enhanced anandamide levels observed following CCI of the sciatic nerve might be a consequence (rather than the cause, as hypothesized at the beginning of this study) of enhanced serotonergic signalling, or represent a totally unrelated event. In a previous study, repeated treatment with SR141716A was found to induce behavioural analgesia in neuropathic rats (Costa et al., 2005). This discrepancy with our present data might be due to: (i) the difference in dosage and administration routes; (ii) the onset of treatment related to the establishment of neuropathic pain or (iii) the number of days following surgery (day 7 vs. day 14). Using the same dose used in this study (1 mg ⁄ kg), La Rana et al. (2006) recently reported the absence of any effect of SR141716A at 3 or 7 days postligation. In order to further investigate the role of endocannabinoids in the DR in relation to serotonergic signalling, we studied here for the first time the effect of repeated administration of AM404, an inhibitor of endocannabinoid cellular re-uptake and hence of endocannabinoid inactivation, on the activity of DR neurones, 5-HT release and 5-HT receptor expression in the DR. The compound, at a dose previously reported to elevate brain anandamide levels and to enhance anandamide pharmacological action (Beltramo et al., 1997), behaved exactly like WIN55,212-2, inasmuch as it elevated DR neuronal firing, 5-HT release and 5-HT1A receptor expression in sham-operated rats, whereas it normalized such parameters in neuropathic rats. Furthermore, in agreement with the recent data of La Rana et al. (2006), both repeated treatment with, and single administration of, AM404 in fully established neuropathic pain also counteracts hyperalgesia and allodynia. Moreover, a single administration of AM404 was able to transiently decrease the higher DR neuronal firing in the CCI animals, without changing it in the shams. However, under the same conditions of chronic administration, AM404 was shown here to exert no effect on endocannabinoid levels in the DR, thus indicating that the actions observed with this compound when administered repeatedly during the development of neuropathic pain were not due to the elevation of endocannabinoid levels but instead to direct interaction with CB1 receptors. Conversely, the antihyperalgesic effect of acute administration of AM404 was accompanied by enhancement of 2-AG levels, but not of the possibly already maximally elevated anandamide levels, in agreement with the concept that 2-AG is also a substrate for the putative anandamide membrane transporter (Di Marzo et al., 2004). The effects of chronic AM404 were sensitive to SR141716A, although the CB1 antagonist did not completely revert them at a dose shown here to be fully efficient against WIN55,212-2. As AM404 is a weaker agonist to CB1

Table 1. Observed in vitro activity of AM404 on anandamide uptake, fatty acid amide hydrolase inhibition and transient TRPV1 and CB1 receptor stimulation

References

Inhibition of anandamide cellular uptake (AM404 IC50 ¼ 1–5 lm)

Stimulation of TRPV1 receptors (EC50 ¼ 25–60 nm)

Inhibitor of fatty acid amide hydrolase (IC50 ¼ 0.5–5 lm)

Agonist at cannabinoid CB1 receptors (Ki ¼ 1.8 lm)

Beltramo et al. (1997) Di Marzo et al. (1998a) De Petrocellis et al. (2000) Zygmunt et al. (2000) Ross et al. (2001) Jarrahian et al. (2000) Fowler et al. (2004)

Yes Yes – – – – –

– – Yes Yes Yes – –

– – – – – Yes Yes

Yes – – – – – –

ª The Authors (2006). Journal Compilation ª Federation of European Neuroscience Societies and Blackwell Publishing Ltd European Journal of Neuroscience, 24, 2011–2020

Cannabinoid and dorsal raphe changes in the neuropathic rat 2019 receptors than WIN55,212-2 (Beltramo et al., 1997), it is therefore possible that the AM404-induced effects may also involve other receptors, such as TRPV1 receptors, which are highly expressed in the DR (Mezey et al., 2000) and towards which AM404 behaves as a full agonist (De Petrocellis et al., 2000; Zygmunt et al., 2000) (Table 1). In support of this possibility, it has been shown that the effect of AM404, in exactly the same animal model of neuropathic pain used here, is antagonized by both CB1 and TRPV1 receptor antagonists (Costa, 2005). Interestingly, stimulation of TRPV1 receptors was previously shown to be coupled to enhanced glutamatergic signalling in other brain areas (Marinelli et al., 2002, 2003) and this property might lead to enhanced firing activity of DR neurones, as well as to enhanced 5HT release, although further studies are necessary to support this hypothesis. Although the experiments with chronic administration of AM404 did not allow us to reach any definitive conclusion as to the possible role of elevated anandamide levels in the DR during neuropathic pain, the lack of any effect observed with this type of treatment on anandamide levels might also be interpreted as the sum of two opposing actions on this end-point, i.e. (i) an enhancement of anandamide levels due to inhibition of anandamide cellular re-uptake and (ii) an inhibition of anandamide levels due to the normalization of serotonergic signalling if, as hypothesized above, the elevation of anandamide levels observed in the DR of neuropathic rats is the consequence, rather than the cause, of enhanced serotonergic signalling. Nevertheless, this possibility is strongly opposed by our findings in sham-operated rats, where repeated treatment with AM404 (but not a single administration) enhances serotonergic signalling but still produces no enhancement of anandamide levels. This observation, together with the other findings described here, strongly suggests that: (i) the strong elevation of anandamide levels observed in the DR of neuropathic rats has no relevance to CB1-mediated tonic control of pain and no relation to CB1-mediated serotonergic signalling and (ii) repeatedly administered AM404 does not influence endocannabinoid levels in the DR, and its effects observed in this study are due to direct interaction with CB1 and, perhaps, TRPV1 receptors. It is possible that enhanced anandamide during neuropathic pain plays a role in other consequences of this disorder, as the DR, rather than pain, controls mood, sleep, memory, cognition and feeding. It is also possible that, in the DR of neuropathic rats, endogenous anandamide acts on one of its many non-CB1 receptors (including TRPV1, see Di Marzo et al., 2002 for review), and that consequently, we could not unmask its action as in this study we used a selective CB1 antagonist and an agent that was not capable of further increasing its levels in this area of the brain. However, the antihyperalgesic effects observed in CCI rats with a single administration of AM404, which was also accompanied by a decrease in serotonergic signalling in the DR, did correlate with an enhancement of 2-AG levels in this brain area and therefore might be due (at least in part) to enhanced ‘indirect’ activation of CB1 receptors via reduced 2-AG inactivation. Thus, another possible interpretation of our data is that 2-AG, rather than anandamide, is the endocannabinoid that participates in pain signalling in the DR when its levels are pharmacologically elevated. In conclusion, we have shown that the endocannabinoid system is activated in the DR following the development of neuropathic pain, although the physiopathological meaning of this finding has yet to be assessed and does not appear to be relevant to the CB1-mediated control of pain and serotonergic signalling. Nevertheless, we have also reported that repeated stimulation of CB1 receptors with an exogenous synthetic agonist proves very effective in changing nociceptive thresholds during neuropathic pain, while normalizing the serotonergic activity that we have suggested here for the first time to be dramatically

activated in this pathological condition. Normalization of pain-induced over-activation of DR extracellular 5-HT opens up the perspective that cannabinoid treatment could resolve neuropathic pain and some of its negative emotional consequences, thus representing a possible and very promising advance in the clinical management of this highly impairing condition. Finally, acute pharmacological elevation of 2-AG levels in the DR with inhibitors of 2-AG inactivation appears to correlate with antihyperalgesic effects and might be pursued as an additional therapeutic strategy to treat chronic pain.

Acknowledgements Financial support was provided by MIUR, Italy (PRIN 2003). We thank SanofiSynthelabo (Montpellier, France) for the generous gift of SR141716A.

Abbreviations 2-AG, 2-arachidonoylglycerol; AM404, N-(4-hydroxyphenyl)-5Z,8Z,11Z,14Zeicosatetraenamide; CB1, cannabinoid subtype 1; CCI, chronic constriction injury; DR, dorsal raphe; 5-HT, serotonin; SR141716A, N-piperidino-5-(4chlorophenyl)-1-(2,4dichlorophenyl)-4-methyl-3-pyrazolecarboxamide; TRPV1, transient receptor potential vanilloid type 1; WIN55,212-2, (R)-(+)[2,3-dihydro-5-methyl-3-(4-morpholinylmethyl)pyrrolo[1,2,3-de]-1,4-benzoxazin-6-yl]-1-naphthalenylmethanone mesylate.

References Aghajanian, G.K. & Vandermaelen, C.P. (1982) Intracellular identification of central noradrenergic and serotonergic neurons by a new double labelling procedure. J. Neurosci., 2, 1786–1792. Andersen, N. & Dafny, N. (1983) Dorsal raphe stimulation reduces responses of parafascicular neurons to noxious stimulation. Pain, 15, 323–331. Barkin, R.L. & Fawcett, J. (2000) The management challenges of chronic pain: the role of antidepressants. Am. J. Ther., 7, 31–47. Basbaum, A.I. & Fields, H.L. (1979) The origin of descending pathways in the dorsolateral funiculus of the spinal cord of the cat and rat; further studies on the anatomy of pain modulation. J. Comp. Neurol., 187, 513–532. Basbaum, A.I. & Fields, H.L. (1984) Endogenous pain control system: brainstem spinal pathways and endorphin circuitry. Annu. Rev. Neurosci., 7, 309–338. Beltramo, M., Stella, N., Calignano, A., Lin, S.Y., Makriyannis, A. & Piomelli, D. (1997) Functional role of high-affinity anandamide transport, as revealed by selective inhibition. Science, 277, 1094–1097. Bisogno, T., Sepe, N., Melck, D., Maurelli, S., De Petrocellis, L. & Di Marzo, V. (1997) Biosynthesis, release and degradation of the novel endogenous cannabimimetic metabolite 2-arachidonoylglycerol in mouse neuroblastoma cells. Biochem. J., 322, 671–677. Costa, B. (2005) Cannabinoids in neuropathic pain. In Second European Workshop on Cannabinoid Research ICRS, April 29–30 2005, Busto Arsizio, Varese, Italy, p. 29. Costa, B., Colleoni, M., Conti, S., Trovato, A.E., Bianchi, M., Sotgiu, M.L. & Giagnoni, G. (2004) Repeated treatment with the synthetic cannabinoid WIN 55,212-2 reduces both hyperalgesia and production of pronociceptive mediators in a rat model of neuropathic pain. Br. J. Pharmacol., 141, 4–8. Costa, B., Trovato, A.E., Colleoni, M., Giagnoni, G., Zarini, E. & Croci, T. (2005) Effect of the cannabinoid CB1 receptor antagonist, SR141716, on nociceptive response and nerve demyelination in rodents with chronic constriction injury of the sciatic nerve. Pain, 116, 52–61. Cucchiaro, G., Chaijale, N. & Commons, K.G. (2005) The dorsal raphe nucleus as a site of action of the antinociceptive and behavioural effects of the a4 nicotinic receptor agonist epibatine. J. Pharmacol. Exp. Ther., 313, 389–394. Dahlstrom, A. & Fuxe, K. (1964) Localization of monoamine in the lower brainstem. Experientia, 20, 398–399. De Petrocellis, L., Bisogno, T., Davis, J.B., Pertwee, R.G. & Di Marzo, V. (2000) Overlap between the ligand recognition properties of the anandamide transporter and the VR1 vanilloid receptor: inhibitors of anandamide uptake with negligible capsaicin-like activity. FEBS Lett., 483, 52–56. Descarries, L., Watkins, K.C., Garcia, S. & Beaudet, A. (1982) The serotonin neurons in nucleus raphe dorsalis of adult rat: a light and electron microscope radioautographic study. J. Comp. Neurol., 207, 239–354.

ª The Authors (2006). Journal Compilation ª Federation of European Neuroscience Societies and Blackwell Publishing Ltd European Journal of Neuroscience, 24, 2011–2020

2020 E. Palazzo et al. Devane, W.A., Hanus, L., Breuer, A., Pertwee, R.G., Stevenson, L.A., Griffin, G., Gibson, D., Mandelbaum, A., Etinger, A. & Mechoulam, R. (1992) Isolation and structure of a brain constituent that binds to the cannabinoid receptor. Science, 258, 1946–1949. Di Marzo, V., Bisogno, T., Melck, D., Ross, R., Brockie, H., Stevenson, L., Pertwee, R. & De Petrocellis, L. (1998a) Interactions between synthetic vanilloids and the endogenous cannabinoid system. FEBS Lett., 436, 449– 454. Di Marzo, V., Melk, D., Bisogno, T. & De Petrocellis, L. (1998b) Endocannabinoids: endogenous cannabinoid receptor ligands with neuromodulatory action. Trends Neurosci., 12, 521–528. Di Marzo, V., De Petrocellis, L., Fezza, F., Ligresti, A. & Bisogno, T. (2002) Anandamide receptors. Prostaglandins Leukot. Essent. Fatty Acids, 66, 377– 391. Di Marzo, V., Bifulco, M. & De Petrocellis, L. (2004) The endocannabinoid system and its therapeutic exploitation. Nat. Rev. Drug Discov., 3, 771– 784. Dong, W.Q., Qiao, J.T., Skolnick, M. & Dafny, N. (1991) Focal dorsal raphe stimulation and pinnal electrical stimulation modulation modulate spontaneous and noxious evoked responses in thalamic neurons. Int. J. Neurosci., 57, 123–140. Fields, H.L. & Basbaum, A.I. (1978) Brainstem control of spinal pain transmission neurons. Ann. Rev. Phys., 40, 217–248. Fowler, C.J., Tiger, G., Ligresti, A., Lopez-Rodriguez, M.L. & di Marzo, V. (2004) Selective inhibition of anandamide cellular uptake versus enzymatic hydrolysis ) a difficult issue to handle. Eur. J. Pharmacol., 492, 1–11. Galderisi, U., Di Bernardo, G., Cipollaro, M., Peluso, G., Cascino, A., Cotrufo, R. & Melone, M.A. (1999) Differentiation and apoptosis of neuroblastoma cells: role of N-myc gene product. J. Cell. Biochem., 73, 97–105. Graeff, F.G., Guimara˘es, F.G., De Andrade, T.G.S.C. & Deakin, J.F.W. (1996) Role of 5-HT in stress, anxiety and depression. Pharmacol. Biochem. Behav., 54, 129–141. Herzberg, U., Eliav, E., Bennett, G.J. & Kopin, I.J. (1997) The analgesic effect of R(+)-WIN 55,212-2 mesylate, a high affinity cannabinoid agonist, in a rat model of neuropathic pain. Neurosci. Lett., 221, 157–160. Iversen, L. & Chapman, V. (2002) Cannabinoids: a real prospect for pain relief? Curr. Opin. Pharmacol., 2, 50–55. Jacobs, B.L. & Azmitia, E.C. (1992) Structure and function of the brain serotonin system. Physiol. Rev., 72, 165–229. Jarrahian, A., Manna, S., Edgemond, W.S., Campbell, W.B. & Hillard, C.J. (2000) Structure-activity relationships among N-arachidonylethanolamine (anandamide) head group analogues for the anandamide transporter. J. Neurochem., 74, 2597–2606. La Rana, G., Russo, R., Campolongo, P., Bortolato, M., Manieri, R.A., Cuomo, V., Iacono, A., Mattace Raso, G., Meli, R., Pomelli, D. & Malignano, A. (2006) Modulation of neuropathic and inflammatory pain by the endocannabinoid transport inhibitor AM404. J. Pharmacol. Exp. Ther., 317, 1365– 1371. MacFarlane, B.V., Wright, A., O’Callaghan, J. & Benson, H.A.E. (1997) Chronic neuropathic pain and its control by drugs. Pharmacol. Ther., 75, 1– 19. Marinelli, S., Vaughan, C.W., Christie, M.J. & Connor, M. (2002) Capsaicin activation of glutamatergic synaptic transmission in the rat locus coeruleus in vitro. J. Physiol., 543, 531–540. Marinelli, S., Di Marzo, V., Berretta, N., Matias, I., Maccarrone, M., Bernardi, G. & Mercuri, N.B. (2003) Presynaptic facilitation of glutamatergic synapses to dopaminergic neurons of the rat substantia nigra by endogenous stimulation of vanilloid receptors. J. Neurosci., 23, 3136–3144. Marsicano, G., Goodenough, S., Monory, K., Hermann, H., Eder, M., Cannich, A., Azad, S.C., Cascio, M.G., Gutierrez, S.O., van der Stelt, M., LopezRodriguez, M.L., Casanova, E., Schutz, G., Zieglgansberger, W., Di Marzo, V., Behl, C. & Lutz, B. (2003) CB1 cannabinoid receptors and on-demand defense against excitotoxicity. Science, 302, 84–88. Mattia, C. & Coluzzi, F. (2003) Antidepressants in chronic neuropathic pain. Mini Rev. Med. Chem., 7, 773–784. Meng, I.D., Manning, B.H., Martin, W.J. & Fields, H.L. (1998) An analgesia circuit activated by cannabinoids. Nature, 395, 381–383. Mezey, E., Toth, Z.E., Cortright, D.N., Arzubi, M.K., Krause, J.E., Elde, R., Guo, A., Blumberg, P.M. & Szallasi, A. (2000) Distribution of mRNA for vanilloid receptor subtype 1 (VR1), and VR1-like immunoreactivity, in the central nervous system of the rat and human. Proc. Natl Acad. Sci. U.S.A., 97, 3655–3660.

Miki, K., Zhou, Q.-Q., Guo, W., Guan, Y., Terayama, R., Dubner, R. & Ren, K. (2002) Changes in gene expression and neuronal phenotype in brain stem pain modulatory circuitry after inflammation. J. Neurophysiol., 87, 750–760. Nazaki, M., Bauer, U., Nickel, T., Kathmann, M. & Sclicker, E. (2000) Inhibition of serotonin release in the mouse brain via presynaptic cannabinoid CB1 receptors. Naynyn-Schmiedeberg’s Arch. Pharmacol., 361, 19–24. Neugebauer, V., Li, W., Bird, G.C. & Han, J.S. (2004) The amygdala and persistent pain. Neuroscientist, 10, 221–234. Palazzo, E., Genovese, R., Mariani, L., Siniscalco, D., Marabese, I., de Novellis, V., Rossi, F. & Maione, S. (2004) Metabotropic glutamate receptor 5 and dorsal raphe serotonin release in inflammatory pain in rat. Eur. J. Pharmacol., 492, 169–176. Paxinos, G. & Watson, C. (1986) The Rat Brain in Stereotaxic Coordinates. Academic Press, London. Porro, C., Cavazzuti, M., Galetti, A. & Passatelli, L. (1991) Functional activity mapping of the rat brainstem during formalin-induced noxious stimulation. Neuroscience, 41, 667–680. Reynolds, D.V. (1969) Surgery in the electrical analgesia induced by focal brain stimulation. Science, 164, 444–445. Rodella, L.F., Borsani, E., Rezzani, R., Ricci, F., Buffoli, B. & Bianchi, R. (2005) AM404, an inhibitor of anandamide reuptake decreases Fosimmunoreactivity in the spinal cord of neuropathic rats after non-noxious stimulation. Eur. J. Pharmacol., 508, 139–146. Ross, R.A., Gibson, T.M., Brockie, H.C., Leslie, M., Pashmi, G., Craib, S.J., Di Marzo, V. & Pertwee, R.G. (2001) Structure-activity relationship for the endogenous cannabinoid, anandamide, and certain of its analogues at vanilloid receptors in transfected cells and vas deferens. Br. J. Pharmacol., 132, 631–640. Sanders, K.H., Klein, C.E., Mayer, T.E., Heym, C. & Handweker, H.O. (1980) Differential effects of noxious and non noxious input on neurons according to location in ventral periaqueductal grey or dorsal raphe nucleus. Brain Res., 186, 83–97. Schlicker, E. & Kathmann, M. (2001) Modulation of transmitter release via presynaptic cannabinoids receptors. Trends Pharmacol. Sci., 22, 565–572. Scott, D.A., Wright, C.E. & Angus, J.A. (2004) Evidence that CB1 and CB2 cannabinoid receptors mediate antinociception in neuropathic pain in the rat. Pain, 109, 124–131. Sindrup, S.H., Gram, L.F., Brosen, K., Eshoj, O. & Mogensen, E.F. (1990) The selective serotonin reuptake inhibitor paroxetine is effective in the treatment of diabetic neuropathy symptoms. Pain, 42, 135–144. Steffens, M. & Feuerstein, T. (2004) Receptors-independent depression of DA and 5-HT uptake by cannabinoids in rat neocortex-involvement of Na+ ⁄ K+ATPase. Neurochem. Int., 44, 529–538. Tzavara, E.T., Davis, R.J., Perry, K.W., Li, X., Salhoff, C., Bymaster, F.P., Witkin, J.M. & Nomikos, G.G. (2003) The CB1 receptor antagonist SR141716A selectively increases monoaminergic neurotransmission in the medial prefrontal cortex: implications for therapeutic actions. Br. J. Pharmacol., 138, 544–553. Vaughan, C.W., Connor, M., Bagley, E.E. & Macdonald, J.C. (2000) Actions of cannabinoids on membrane properties and synaptic transmission in rat periaqueductal gray neurons in vitro. Mol. Pharmacol., 57, 288–295. Walker, J.M., Huang, S.M., Strangman, N.M., Tsou, K. & Sanudo-Pena, M.C. (1999) Pain modulation by release of the endogenous cannabinoid anandamide. Proc. Natl Acad. Sci. U.S.A., 96, 12 198–12 203. Walker, J.M., Krey, J.F., Chen, J.S., Vefring, E., Jahnsen, J.A., Bradshaw, H. & Huang, S.M. (2005) Targeted lipodomics: fatty acid amides and pain modulation. Prostaglandins Other Lipid Med., 77, 35–45. Wang, Q.P. & Nakai, Y. (1994) The dorsal raphe: an important nucleus in pain modulation. Brain Res. Bull., 34, 575–585. Wang, Y.X.B.V., Wright, A., O’Callaghan, J. & Benson, H.A.E. (1997) Chronic neuropathic pain and its control by drugs. Pharmacol. Ther., 75, 1–19. Wang, Y.X., Bowersox, S.S., Pettus, M. & Gao, D. (1999) Antinociceptive properties of fenfluramine, a serotonin reuptake inhibitor, in a rat model of neuropathy. J. Pharmacol. Exp. Ther., 291, 1008–1016. Watson, C.P.N. (2000) The treatment of neuropathic pain: antidepressant and opioids. Clin. J. Pain, 16, S49–S55. Zhang, Y.Q., Gao, X., Zhang, L.M. & Wu, G.C. (2000) The release of serotonin in rat spinal dorsal horn and periaqueductal gray following carrageenan inflammation. Neuroreport, 11, 3539–3543. Zygmunt, P.M., Chuang, H., Movahed, P., Julius, D. & Ho¨gesta¨tt, E.D. (2000) The anandamide transport inhibitor AM404 activates vanilloid receptors. Eur. J. Pharmacol., 396, 39–42.

ª The Authors (2006). Journal Compilation ª Federation of European Neuroscience Societies and Blackwell Publishing Ltd European Journal of Neuroscience, 24, 2011–2020