In the ventrolateral periaqueductal gray (PAG), activation of excitatory output neurons projecting monosynaptically to OFF cells in the rostral ventromedial medulla (RVM) causes antinociceptive responses and is under the control of cannabinoid receptor type-1 (CB1) and vanilloid transient receptor potential vanilloid type 1 (TRPV1) receptors. We studied in healthy rats the effect of elevation of PAG endocannabinoid [anandamide and 2-arachidonoylglycerol (2-AG)] levels produced by intra-PAG injections of the inhibitor of fatty acid amide hydrolase URB597 [cyclohexylcarbamic acid-3′-carbamoyl-biphenyl-3-yl ester] on 1) nociception in the “plantar test” and 2) spontaneous and tail-flick-related activities of RVM neurons. Depending on the dose or time elapsed since administration, URB597 (0.5–2.5 nmol/rat) either suppressed or increased thermal nociception via TRPV1 or CB1 receptors, respectively. TRPV1 or cannabinoid receptor agonists capsaicin (6 nmol) and (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 (4 nmol)] also suppressed or enhanced nociception, respectively. URB597 dose dependently enhanced PAG anandamide and 2-AG levels, with probable subsequent activation of TRPV1/CB1 receptors and only CB1 receptors, respectively. The TRPV1-mediated antinociception and CB1-mediated nociception caused by URB597 correlated with enhanced or reduced activity of RVM OFF cells, suggesting that these effects occur via stimulation or inhibition of excitatory PAG output neurons, respectively. Accordingly, several ventrolateral PAG neurons were found by immunohistochemistry to coexpress TRPV1 and CB1 receptors. Finally, at the highest doses tested, URB597 (4 nmol/rat) and, as previously reported, WIN55,212-2 (25–100 nmol) also caused CB1-mediated analgesia, correlating with stimulation (possibly disinhibition) of RVM OFF cells. Thus, endocannabinoids affect the descending pathways of pain control by acting at either CB1 or TRPV1 receptors in healthy rats.
Two G-protein-coupled receptors for Cannabis psychotropic component Δ9-tetrahydrocannabinol have been cloned to date (Matsuda et al., 1990; Munro et al., 1993) and are implicated in the control of nociception under both physiological and pathological conditions (for review, see Iversen and Chapman, 2002). The cannabinoid receptor type-1 (CB1) modulates, by inhibiting glutamate or GABA release (for review, see Schlicker and Kathmann, 2001), both spinal nociceptive circuits and the descending supraspinal pathways at the level of the brainstem and medulla. In the spinal cord, CB1 receptor stimulation causes analgesia by inhibiting glutamate and proalgesic neuropeptide release (Richardson et al., 1998). Inhibition of GABA release from opioidergic/GABAergic interneurons of the periaqueductal gray (PAG) causes disinhibition of output neurons in the ventrolateral PAG (Vaughan et al., 2000). These neurons project monosynaptically to OFF cells of the rostral ventromedial medulla (RVM) (Sandkuhler and Gebhart, 1984; Moreau and Fields, 1986). This latter mechanism, together with inhibition of GABA release from RVM interneurons, underlies those antinociceptive effects of CB1 receptor activation that are common to opiate-induced analgesia in these midbrain and medulla areas (Meng et al., 1998; Meng and Johansen, 2004). On the other hand, direct stimulation of excitatory PAG output neurons under the control of opioidergic/GABAergic interneurons produces antinociceptive actions that are even more efficacious than those induced by morphine and correlates again with increased activity of the RVM OFF cells that these neurons directly innervate (Aimone and Gebhart, 1986; Wiklund et al., 1988; Tortorici and Morgan, 2002).
In addition, the stimulation of transient receptor potential vanilloid type 1 (TRPV1) receptors in the PAG produces analgesic effects by acting on the same RVM-projecting neurons that mediate glutamate-induced analgesia in the PAG or by desensitizing the activity of other neurons involved in nociception in this region (Palazzo et al., 2002; McGaraughty et al., 2003). TRPV1 receptors are nonselective cation channels known to be expressed mostly by peripheral sensory afferents and spinal cord neurons and to act as polymodal transducers of the nocifensive effects of some plant toxins, low pH, and noxious heath (Szallasi and Blumberg, 1999). Recent evidence suggests that these receptors are also expressed in the brain (Mezey et al., 2000), including the PAG (McGaraughty et al., 2003), and that they can be activated by endogenous compounds known as “endovanilloids” (van der Stelt and Di Marzo, 2004).
The lipid mediator anandamide (Devane et al., 1992) activates both CB1 and TRPV1 receptors (van der Stelt and Di Marzo, 2004), and its levels are elevated in the rat PAG after noxious stimulation (Walker et al., 1999). The other major endocannabinoid, 2-arachidonoyl glycerol (2-AG) (Mechoulam et al., 1995), is a full agonist at CB1 receptors but lacks activity at TRPV1 channels (van der Stelt and Di Marzo, 2004). Both compounds are inactivated by fatty acid amide hydrolase (FAAH) (Cravatt et al., 1996; Goparaju et al., 1998), whereas another enzyme, the monoacylglycerol lipase, also catalyzes 2-AG hydrolysis (Hohmann et al., 2005). Recent reports have shown that administration of FAAH inhibitors to mice and rats in vivo can enhance both anandamide and 2-AG levels (de Lago et al., 2005). Therefore, it is possible that the injection of FAAH inhibitors into the PAG may offer the opportunity to directly investigate the role of both endocannabinoids and of their respective molecular targets in the control of descending pain pathways. For these reasons, in this study, we have looked at the effect on rats of intra-PAG injections of selective FAAH inhibitor URB597 (Hohmann et al., 2005) on 1) nocifensive responses to heat in the plantar test, 2) PAG levels of the two major endocannabinoids, anandamide and 2-AG, and 3) spontaneous and tail-flick-related activities of OFF and ON cells of the RVM. Furthermore, we have investigated the presence and possible colocalization of TRPV1 or CB1 receptors in the PAG by immunohistochemistry.
Materials and Methods
Animals and Surgical Preparation
Male Wistar rats (250–300 g) were housed three per cage under controlled illumination (12-h light/12-h dark cycle; lights on at 6:00 AM) 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 European regulations on the protection of laboratory animals (Official Journal of E.C. L358/1 18/12/86). In agreement with the Ethical Guidelines of the International Association for the Study of Pain, all efforts were made to reduce both animal numbers and suffering during the experiments. On the day of the experiment, rats were anesthetized with pentobarbital (60 mg/kg, i.p.) and a 23-gauge, 12 mm-long stainless steel guide cannula was stereotaxically lowered until its tip was 1.4 mm above the ventrolateral PAG by applying coordinates from Paxinos and Watson (1986) (A, –7.8 mm from bregma; L, 0.5 mm; V, 4.3 mm below the dura). Ventrolateral PAG was considered in this study, because previous studies have shown in that area the presence of glutamatergic output neurons projecting monosynaptically to OFF neurons in the RVM (Sandkuhler and Gebhart, 1984; Aimone and Gebhart, 1986; Wiklund et al., 1988; Tortorici and Morgan, 2002). The cannula was anchored with dental cement to a stainless steel screw in the skull. We used a David Kopf stereotaxic apparatus (David Kopf Instruments, Tujunga, CA) with the animal positioned on a homeothermic temperature control blanket (Harvard Apparatus Limited, Edenbridge, Kent, UK). The jugular vein was also cannulated to allow intravenous anesthetic administration. At the end of the experiment, a volume of 200 nl of Neutral Red (0.1%) was also injected in the ventrolateral PAG 40 to 50 min before killing the rat. Rats were then perfused intracardially with 20 ml of phosphate-buffered solution followed by 20 ml of 10% formalin solution in phosphate-buffered solution. The brains were removed and immersed in a saturated formalin solution for 2 days. The injection sites were ascertained by using two consecutive sections (40 μm), one stained with Cresyl Violet to identify nuclei and the other one unstained to determine dye spreading. Only those rats whose microinjection site and diffusion were located within ventrolateral PAG were included in the results.
Changes in thermoceptive responses were evaluated according to Hargreaves et al. (1988) using a Plantar Test Apparatus (Ugo Basile, Varese, Italy). On the day of experiment, each animal was placed in a plastic cage (22 × 17 × 14 cm; length × width × height) with a glass floor. After a 1-h habituation period, the plantar surface of the hind paw was exposed to a beam of radiant heat through the glass floor. The radiant heat source consisted of an infrared bulb (Osram halogen-bellaphot bulb; 8 V, 50 watts). 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 latency of nociceptive reaction was measured in seconds under basal condition and after intra-PAG treatment with drugs. Each rat served as its own control, the latency to nociceptive response being measured both before and after treatments. At time 0, five baseline responses were already obtained for each animal at 30-min intervals and averaged. Groups of 10 animals per treatment were used, with each animal used for one treatment only. The results are expressed as a percentage of the maximal possible effect (% MPE), using the following formula:
Statistical analysis of the data were performed by analysis of variance (ANOVA) followed by the Student-Newman-Keuls multiple comparison test. Differences were considered significant at the P < 0.05.
RVM Extracellular Recording and Intra-PAG Microinjections
After implantation of the guide cannula into the ventrolateral PAG matter, a tungsten microelectrode was stereotaxically (Paxinos and Watson, 1986) lowered through a small craniotomy into the RVM to record the activity of ON and OFF cells. These neurons were identified by the characteristic OFF cell pause and ON cell burst of activity just before tail-flick responses (Fields et al., 1983). Anesthesia was maintained with a constant continuous infusion of sodium methohexital (40–60 mg/kg/h i.v.). Anesthesia was adjusted so that tail-flicks were elicited with a constant latency of 4 to 5 s. From 35°C, the temperature increased linearly to 53°C and it was adjusted at the beginning of each experiment to elicit a constant tail-flick latency of 4 to 5 s. The thermal stimulus was elicited by a radiant heat source of a tail-flick unit (Ugo Basile) focused on the rat tail approximately 3 to 5 cm from the tip. Tail-flicks were elicited every 3 min for at least 15 to 20 min before microinjecting drugs or respective vehicles into the PAG.
Extracellular single-unit recordings were made in the RVM with glass-insulated tungsten filament electrodes (3–5 MΩ) (FHC Inc., Bowdoinham, ME) using the following stereotaxic coordinates (Paxinos and Watson, 1986): 2.8 to 3.3 mm caudal to lambda, 0.4 to 0.9 mm lateral, and 8.9 to 10.7 mm depth from the surface of the brain. The recorded signals were amplified and displayed on analog and digital storage oscilloscope to ensure that the unit under study was unambiguously discriminated throughout the experiment. Signals were also fed into a window discriminator, whose output was processed by an interface (CED 1401; Cambridge Electronic Design Ltd., Cambridge, UK) connected to a Pentium III PC. Spike2 software (CED, version 4) was used to create peristimulus rate histograms on-line and to store and analyze digital records of single-unit activity off-line. Configuration, shape, and height of the recorded action potentials were monitored and recorded continuously, using a window discriminator and Spike2 software for on-line and off-line analysis. Once an ON or OFF cell was identified from its background activity, we optimized spike size before all treatments. This study only included neurons whose spike configuration remained constant and could clearly be discriminated from activity in the background throughout the experiment, indicating that the activity from one neuron only and from the same one neuron was measured. In each rat, only one neuron was recorded.
Direct intra-PAG administration of drugs, or respective vehicle 10% dimethyl sulfoxide in artificial cerebrospinal fluid (ACSF; composition of 2.5 mM KCl, 125 mM NaCl, 1.18 mM MgCl2, and 1.26 mM CaCl2), were conducted with a stainless steel cannula connected by a polyethylene tube to a Hamilton 1-μl syringe (Hamilton Company, Reno, NV), inserted through the guide cannula, and extended 1.4 mm beyond the tip of the guide cannula to reach the ventrolateral PAG. Volumes of 200 nl of drug solutions or vehicles were injected into the ventrolateral PAG over a period of 60 s, and the injection cannula was gently removed 5 min later. At the end of the experiment, each animal was killed with a lethal dose of urethane, the microinjection site was marked with 0.2 μl of a Cresyl Violet solution and the recording site was marked with a 20-μA direct current for 20 s. After fixation by immersion in 10% formalin, the microinjection and recording sites were identified.
For behavioral studies, groups of rats (n = 10) received, after five baseline responses at intervals of 30 min each, intra-PAG administrations of 200 nl of drugs or respective vehicle (10% dimethyl sulfoxide in ACSF) as follows: 1) intra-PAG microinjections of vehicle, capsaicin (6 nmol/rat), WIN55,212-2 (2, 4, 25, and 100 nmol/rat) alone, or WIN55,212-2 (4, 25, and 100 nmol/rat) in combination with capsaicin (6 nmol); 2) intra-PAG microinjections of vehicle, URB597 (0.5, 2.5, and 4 nmol/rat) alone, or in combination with AM251 (2.5 nmol/rat), capsazepine (6 nmol/rat), or URB597 (0.5 nmol/rat) in combination either with AM251 (2.5 nmol/rat) or capsazepine (6 nmol/rat); and 3) intra-PAG microinjections of vehicle, URB597 (0.5 nmol/rat) alone, or in combination with THL (0.4 nmol/rat) or O-3841 (0.45 nmol/rat).
For endocannabinoid extraction and quantification studies, groups of rats (n = 4) received, 15 min before PAG dissection, intra-PAG injections of URB597 (0.5, 2.5, and 4 nmol/rat). For electrophysiological studies, groups of rats (n = 10) received intra-PAG administrations of vehicle, URB597 (0.5, 2.5, and 4 nmol/rat) alone, or in combination with capsazepine (6 nmol/rat) or AM251 (2 mg/kg i.p.)
WIN55,212-2, AM251, and capsazepine were purchased from Tocris Cookson Inc. (Bristol, UK). URB597 was purchased from Alexis Biochemicals (San Diego, CA). Tetrahydrolipstatin was obtained in our laboratory, and O-3841 was a kind gift from Dr. Raj Razdan (Organix, MA). Drugs were dissolved in 10% dimethyl sulfoxide in ACSF.
Behavioral data are represented as means ± S.E. ANOVA, followed by Student-Newman-Keuls post hoc test, was used to determine the statistical significance among differently treated groups of rats. Extracellular recording single-unit activity (action potentials) was analyzed off-line from peristimulus rate histograms using Spike2 software (CED, version 4). The responses of the neurons before and after intra-PAG drug microinjections were measured and expressed as spikes/seconds (Hertz). Baseline activities of neurons were measured between tail-flicks. In particular, basal values were obtained by averaging the activities recorded 30 to 50 s before the application of three to four thermal stimulations (each stimulation trial was performed every 3 min). Data are presented as mean ± S.E. either of changes in time latencies (tail-flick test) or changes in neuron responses (extracellular recordings). Statistical comparisons of values from differently treated groups of rats were made using the Wilcoxon Signed Rank test.
To analyze tail-flick-related ON cell activities (before and after drug treatment), the ongoing activity (spikes/seconds) was determined 40 to 50 s before tail-flick application and then the peak of the ON cell activity related to the tail-flick (peak firing) was quantified. Tail-flick-related ON cell firing was calculated as the number of spikes in the 2-s interval beginning 0.5 s before the tail-flick. Comparisons between pretreatment and post-treatment ongoing activity and tail-flick-related cell burst were performed by applying the nonparametric Wilcoxon matched-pairs test. Moreover, we calculated the ON cell burst latency; that is the interval between the onset of the applied noxious radiant heat and the beginning of the tail-flick-related cell burst. Burst latency was analyzed with a two-way analysis of variance for repeated measures with the Tukey-Kramer test for post hoc comparisons.
We also performed analysis of tail-flick-related OFF cell activities before and after drug treatments. The ongoing activity (spikes/seconds, 40 to 50 s before radial heat application), the latency to onset of the OFF cell pause (the interval between the onset of thermal stimulus and the last spike), and the duration of the cell pause (the interval between the pause onset and the first spike after the tail-flick) were determined. Comparisons between cell pause related to tail-flick and ongoing activity from differently treated groups of rats were performed by applying the nonparametric Wilcoxon matched-pairs test. The interval between the onset of the applied noxious radiant heat and the beginning of the tail-flick-related cell pause (pause latency) was also calculated. The latency to the onset of the cell pause was analyzed with a two-way analysis of variance for repeated measures with the Tukey-Kramer test for post hoc comparisons.
Endocannabinoid Extraction and Quantification
Procedure of Extraction. The PAG from treated animals was extracted immediately after sacrifice. 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 in Devane et al. (1992) and Bisogno et al. (1997), respectively. Homogenates were centrifuged at 13,000g for 16 min (4°C), and the aqueous phase plus debris were 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 analyzed.
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, and the excess solvent evaporated with a rotating evaporator. In addition, aliquots were analyzed by isotope dilution-liquid chromatography/atmospheric pressure chemical ionization/mass spectrometry, carried out under conditions described previously (Marsicano et al., 2003) and allowing the separations of 2-AG and anandamide. MS 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) and 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 to 100 pmol). The same applied to the area ratios between signals of deuterated and undeuterated 2-AG in the 100-pmol to 20-nmol interval. Therefore, anandamide and 2-AG levels in unknown samples were calculated based on their area ratios with the internal deuterated standard signal areas. The amounts of endocannabinoids were expressed as picomoles or nanomoles per gram of wet tissue extracted and were compared by ANOVA followed by the Bonferroni's test.
Two anesthetized male Wistar rats were transcardially perfused with 4% paraformaldehyde in 0.1 M phosphate buffer, pH 7.4. The brains were removed, cryoprotected in sucrose, and cut on a cryostat into 20-μm-thick frozen coronal sections. Double immunofluorescence was carried out on free floating sections incubated for 2 days in a mixture of the following antibodies: goat anti-TRPV1 N terminus (1:100; Santa Cruz Biotechnologies, Santa Cruz, CA) and rabbit anti-CB1 N terminus (1:400; Calbiochem, San Diego, CA). Thereafter, the sections were incubated for 4 h at room temperature in a mixture of secondary IgG antibodies, including goat anti-rabbit Alexa 488 and rabbit anti-goat Alexa 546 (1:100; Invitrogen, Carlsbad, CA). For single ABC immunohistochemistry, the same primary antibodies as those for immunofluorescence were used (anti-TRPV1, 1:200; anti-CB1, 1:800) followed by biotin-conjugated IgG secondary antibodies and avidin-biotin-peroxidase solution (ABC; Vector Laboratories, Burlingame, CA).
Effects on Nociception of CB1 and TRPV1 Receptor Agonists. These experiments were carried out with the aim of mimicking a situation in which both TRPV1 and CB1 receptors are simultaneously stimulated and thus of determining what might be the contribution of each receptor type to nociception in subsequent experiments with a FAAH inhibitor (see below). For these reasons, the agonists were also coinjected. The latency of thermoceptive responses measured by the Hargreaves method (Hargreaves et al., 1988) was 10.2 ± 0.6 s in rat receiving 200 nl of vehicle in the ventrolateral PAG. Injection of the cannabinoid receptor agonist WIN55,212-2 into the ventrolateral PAG caused a delayed (45–75 min) hyperalgesic effect (6.6 ± 0.4 s) in the plantar test at a low dose (4 nmol/rat) and an immediate, long-lasting (5–60 min) analgesic effect (17.1 ± 0.3 and 19.6 ± 0.4 s) in the same test at higher doses (25–100 nmol/rat) (Fig. 1A). The TRPV1 agonist capsaicin only exhibited an immediate and long-lasting (5–60 min) analgesic action at the one dose tested of 6 nmol (Fig. 1B). Interestingly, this analgesic effect of capsaicin erased the hyperalgesic effect of 4 nmol/rat WIN55,212-2 when the two compounds were coinjected. However, when the dose of WIN55,212-2 was augmented to 25 nmol/rat, the initial analgesic phase of capsaicin disappeared and only the delayed hyperalgesia typical of the low dose of WIN55,212-2 was observed (Fig. 1B). Finally, coinjection of 100 nmol/rat WIN55,212-2 with capsaicin (6 nmol) produced a strong initial analgesia (5–30/45 min) followed by a delayed hyperalgesia (60–105/120 min), which was significantly less marked than that observed with 25 nmol/rat WIN55,212-2 (Fig. 1B).
Effects on Nociception of URB597. The latency of thermoceptive responses was 10.2 ± 0.6 s in rats receiving 200 nl of vehicle in the ventrolateral PAG. Injection of URB597 into the ventrolateral PAG produced an immediate and prolonged (5–75 min) hyperalgesic effect in the plantar test at the lowest dose (0.5 nmol/rat), an immediate (5–30/45 min) analgesic effect at the highest dose (4 nmol/rat), and a biphasic effect on nociception at an intermediate dose (2.5 nmol/rat) (Fig. 2, A and B). The hyperalgesic effect of low-dose URB597 was transformed into an analgesic action if the compound was coadministered with the selective CB1 antagonist AM251 (2.5 nmol/rat). The TRPV1 antagonist capsazepine (6 nmol/rat) did not antagonize the hyperalgesic effect of the lowest dose of URB597 (Fig. 2A). Moreover, coinjection of AM251 and capsazepine abolished any effect of this dose of URB597 (Fig. 2A). AM251, but not capsazepine, antagonized the highest dose of URB597, which therefore no longer caused analgesia when CB1 was blocked (Fig. 2A). The analgesia caused by the intermediate dose (2.5 nmol/rat) of URB597 became hyperalgesia in the presence of capsazepine, whereas the hyperalgesic action of this dose of the FAAH inhibitor was blocked by AM251 (Fig. 2B). Neither AM251 alone (2.5 nmol/rat) nor capsazepine (6 nmol/rat) modified thermal threshold per se (data not shown). Coinjection of 0.5 nmol/rat URB597 with 0.45 nmol/rat O-3841 or 0.4 nmol/rat tetrahydrolipstatin, two inhibitors of 2-AG biosynthesis with no effect on anandamide biosynthesis (Bisogno et al., 2004), totally reversed the hyperalgesic effect of the FAAH inhibitor (Fig. 3).
Effect of URB597 on Endocannabinoid Levels in the PAG. We examined the levels of the CB1-selective endocannabinoid, 2-AG, and the endogenous TRPV1/CB1“hybrid” agonist anandamide after intra-PAG injection of different doses (0.5, 2.5, and 4 nmol/rat) of URB597 20 min after injection. The compound caused a dose-dependent enhancement of anandamide levels, the effect being maximal (2-fold enhancement) at the highest dose. Conversely, the effect on 2-AG levels was already statistically significant and maximal at the lowest dose (Fig. 4).
Effect of URB597 on the Ongoing Activities of RVM OFF and ON Cells. The results are based on RVM neurons (group size = 10; one cell recorded from each animal per treatment) at a depth of 8875 to 10,730 μm from the surface of the brain, the estimated location of the neurons being in nucleus raphe magnus, nucleus reticularis gigantocellularis pars α, and nucleus reticularis paragigantocellularis. All recorded neurons were spontaneously active and discharged with a mean frequency of 6.9 ± 0.6 (ON cells) and 8.5 ± 0.7 (OFF cells) spikes/seconds. These neurons were identified by the characteristic OFF cell pause and ON cell burst of activity just before tail-flick responses. Injection of low-dose URB597 (0.5 nmol/rat) into the ventrolateral PAG caused a stimulation of the firing activity of the pronociceptive ON cells in the RVM, which was significant between 6 and 21 min and maximal at 15 min from administration (Fig. 5B), and a very rapid reduction of the firing activity of the antinociceptive OFF cells, significant already after 3 min and maximal after 6 min from administration (Fig. 5A). By contrast, the intermediate and highest doses of URB597 (2.5 and 4 nmol/rat) first caused a stimulation of the firing activity of OFF cells, which was significant after 3 min and maximal after 9 min from administration, and then a reduction of the activity of ON cells, which was significant after 5 to 6 min and maximal after 12 to 15 min from administration (Fig. 5, C–F). The effects of the 2.5 nmol/rat dose were blocked by either pretreatment of the animals with AM251 (2 mg/kg i.p.) or by coinjection with capsazepine (6 nmol/rat), which were inactive per se at these doses (data not shown). Conversely, the effects of the lowest and highest dose of URB597 were blocked by AM251 only (Fig. 5, A–F).
Effect of URB597 on Tail-Flick-Related Activities of RVM OFF and ON Cells. Tail-flicks were elicited every 3 to 4 min for at least 20 min before microinjecting drugs or respective vehicle into the PAG. AM251 was administered systemically via intraperitoneal injection. Data related to pretreatment interval were considered as basal tail-flick latencies (4.5 ± 0.6 s). Intra-PAG microinjections of vehicle did not change the tail-flick latency compared with basal values (4.7 ± 0.5 s). To clarify the opposite effects of URB597 at the doses used, we performed further studies on tail-flick-related RVM cell activities. Tail-flick latency was shortened to 2.7 ± 0.4 s and delayed to 7.8 ± 0.6 s by intra-PAG microinjections of URB597 at doses of 0.5 and 4 nmol/rat (p < 0.05), respectively. The intermediate dose (2.5 nmol/rat) of URB597 generated a biphasic time-dependent change in tail-flick latency. Indeed, the latency was first (5 min post-treatment) increased to 7.2 ± 0.5 s and then (45 min post-treatment) reduced to 2.9 ± 0.6s(p < 0.05). Systemic pretreatment with AM251 (2 mg/kg i.p.), a selective CB1 receptor antagonist, reversed the decrease in tail-flick latency induced by URB597 (0.5 nmol) (from 2.7 ± 0.4 to 7.7 ± 0.5 s) (p < 0.05). Injections of both systemic AM251 (2 mg/kg i.p.) and intra-PAG capsazepine (6 nmol/rat) prevented any effect on tail-flick latency generated by 0.5 nmol URB597. Intra-PAG injection of capsazepine did not change the decrease in tail-flick latency induced by URB597 (0.5 nmol/rat) (from 2.7 ± 0.4 to 3.3 ± 0.6 s). Moreover, systemic administration of AM251 (2 mg/kg i.p.) prevented both the delayed decrease (from 2.9 ± 0.6 to 5.1 ± 0.6 s, p < 0.05) and the increase (from 7.8 ± 0.6 to 4.6 ± 0.5 s, p < 0.05) in tail-flick latencies induced by 2.5 and 4 nmol of URB597, respectively. Finally, when coinjected in the PAG with URB597, capsazepine transformed into a decrease the increase in tail-flick latency induced by the intermediate dose of the inhibitor (from 7.2 ± 0.5 to 3.1 ± 0.3 s, p < 0.05) but did not change its delayed decrease in tail-flick latencies, nor did it alter the increase in latencies caused by the 4 nmol/rat dose. Intra-PAG capsazepine and systemic AM251, at the doses used, did not modify per se tail-flick latencies (data not shown), similar to the findings obtained with the plantar test.
The lower dose of URB597 modified the tail-flick-related cell activities in both ON and OFF cells. URB597 (0.5 nmol/rat) increased the OFF cell pause (from 27.5 ± 4.2 to 41.7 ± 6.4 s; p < 0.05) (Figs. 6 and 7), whereas it did not significantly change the ON cell onset of burst (4.8 ± 0.6 versus 3.7 ± 0.7 s). In addition, low-dose URB597 did not significantly reduce the tail-flick-induced ON-cell peak firing (from 29.5 ± 3.4 to 24.7 ± 5.2 spikes/s) (Figs. 6 and 7) and shortened the onset of OFF cell pause (from 4.4 ± 0.6 to 2.7 ± 0.3 s, p < 0.05). The intermediate dose of URB597 (2.5 nmol/rat) modified the tail-flick-related cell activities in both ON and OFF cells in a biphasic and time-dependent manner. 1) Five minutes postinjection, this dose delayed the ON cell onset of burst from 4.2 ± 0.5 to 7.1 ± 0.3 s (p < 0.05) and reduced the tail-flick-induced ON cell peak firing from 29.5 ± 3.4 to 13.2 ± 3.9 spikes/s (p < 0.05) (Fig. 7). This same dose of URB597 also reduced the OFF cell pause from 27.5 ± 4.2 to 14.2 ± 5.3 s (p < 0.05) (Fig. 7) and delayed the onset of OFF cell pause from 4.4 ± 0.6 to 7.8 ± 0.5 s (p < 0.05). 2) Forty-five minutes postinjection, 2.5 nmol/rat URB597 decreased the ON cell onset of burst (from 4.8 ± 0.6 to 2.1 ± 0.3 s, p < 0.05) and increased the tail-flick-induced ON cell peak firing from 29.5 ± 3.4 to 38.9 ± 3.7 spikes/s (p < 0.05). Moreover, this dose of URB597 increased the OFF cell pause from 27.5 ± 4.2 to 37.7 ± 3.5 s (p < 0.05) and shortened the onset of OFF cell pause (from 4.4 ± 0.6 to 2.4 ± 0.5 s, p < 0.05).
The higher dose of URB597 (4 nmol/rat) also modified the tail-flick-related neuronal activities in both ON and OFF cells. In particular, this dose reduced the OFF cell pause from 27.5 ± 4.2 to 14.7 ± 6.7 s (p < 0.05) (Figs. 6 and 7) and delayed the ON cell onset of burst from 4.8 ± 0.6 to 7.8 ± 0.5 s (p < 0.05). URB597 (4 nmol/rat) also reduced the tail-flick-induced ON cell peak firing from 29.5 ± 3.4 to 18.3 ± 4.5 spikes/s and delayed the onset of the OFF cell pause from 4.4 ± 0.6 to 7.3 ± 0.6 s (p < 0.05) (Figs. 6 and 7).
The biphasic effect of the intermediate dose of URB597 (2.5 nmol/rat) was modified by intraperitoneal AM251 and intra-PAG capsazepine as follows. 1) The decrease induced by URB597 on tail-flick ON cell burst of firing at 5 and 15 min was not modified by systemic administration of AM251 (2 mg/kg i.p.) (from 13.2 ± 3.9 to 15.1 ± 3.2 spikes/s at 5 min) (Fig. 8A). Likewise, AM251 did not affect the decrease in the OFF cell pause (from 14.2 ± 5.3 to 13.9 ± 2.6 s at 5 min) (Fig. 8B). Conversely, at 30 and 45 min post-URB597, AM251 (2 mg/kg i.p.) reversed the increase in the tail-flick-related ON cell burst (from 38.9 ± 3.7 to 24.5 ± 3.2 spikes/s at 45 min) (Fig. 8A) and also shortened the OFF cell pause (from 37.7 ± 3.5 to 26.7 ± 4.2 s at 45 min) (Fig. 8B). 2) Intra-PAG capsazepine (6 nmol/rat) reversed the decrease of the tail-flick-related ON cell burst (from 13.2 ± 3.9 to 37.1 ± 3.1 spikes/s at 5 min) (Fig. 8A) and of the decrease the OFF cell pause (from 14.2 ± 5.3 to 28.2 ± 2.9 s at 5 min) induced by URB597 at 5 and 15 min postinjection (Fig. 8B). However, capsazepine did not affect URB597-induced tail-flick-related changes in the activities of ON and OFF cells from 30 min post-URB597 onwards (Fig. 8, A and B).
Localization of CB1 and TRPV1 Receptors in Neurons of the Rat PAG. The immunohistochemical expression of CB1 and TRPV1 receptors in rat ventrolateral PAG as determined by immunofluorescence (A–D, double immunostaining in the same section) and ABC immunohistochemistry techniques (E–F, single immunostaining in consecutive sections) is shown in Fig. 9. CB1 immunoreactivity was mostly found at the level of terminal processes on cell membranes (probably including presynaptic processes; arrowhead in E), initial exon segments (arrows in A), and axons (arrowheads in A). Expression of TRPV1 was mostly found in cytoplasm, cell membranes (possibly including presynaptic processes), and initial exon segments, but also in some axons (arrowheads in D and F). Coexpression of CB1/TRPV1 receptors was found in several cell bodies and initial exon segments (A–D) and in some axons (arrowheads in C and D). Some axons were found to be CB1-positive (arrowheads in A) and TRPV1-negative (B), and many cell bodies were TRPV1-positive and CB1-negative (B and D). Figure 9, G and H, shows a comparison between CB1 and TRPV1 receptor immunohistochemical expression in a section of the whole PAG region. TRPV1 receptors seem to be significantly more abundant than CB1 receptors in this midbrain area. Similar findings were observed in the mouse PAG (data not shown), where the specificity of the antibodies could be checked by the use of PAG from CB1 and TRPV1 null mice (a generous gift from Dr. David Baker, University College of London, London, UK) in which no immunoreactivity for either CB1 or TRPV1 receptors was found.
In the present study, we wanted to determine whether the enhancement of endocannabinoid levels in the ventrolateral PAG caused by FAAH inhibition modifies the threshold of thermal pain sensitivity in healthy rats. In particular, since among FAAH substrates, only anandamide is able to activate both CB1 and TRPV1 receptors (van der Stelt and Di Marzo, 2004), we assessed whether elevation of the levels of this compound results in activation of these receptors and pain modulation. A recent study has shown that FAAH inhibition in the dorsal PAG does indeed modulate nociception, although in a pathological context (stress-induced analgesia) where the involvement of TRPV1 receptors was not investigated (Hohmann et al., 2005).
We propose the existence of three mechanisms for pain modulation in the PAG-RVM pathway (see Scheme 1; Table 1): 1) CB1-mediated pronociception with low concentrations of WIN 55,212-2, an effect that has never been reported before and might be due to inhibition of those excitatory output neurons that directly stimulate RVM OFF cells (Aimone and Gebhart, 1986; Wiklund et al., 1988; Maione et al., 1998; Tortorici and Morgan, 2002; de Novellis et al., 2003); 2) TRPV1-mediated rapid analgesia, which our data suggest are due to either presynaptic or postsynaptic stimulation of the same output neurons mentioned above (although possible alternative mechanisms are discussed below); and 3) CB1-mediated rapid analgesia at higher concentrations of WIN 55,212-2, a well known effect ascribed to the inhibition of GABA release from PAG interneurons and subsequent disinhibition of the antinociceptive PAG output neurons mentioned above (Moreau and Fields, 1986; Meng et al., 1998; Vaughan et al., 2000; Tortorici and Morgan, 2002). We found that CB1 and TRPV1 receptors are indeed colocalized in several neurons of the ventrolateral PAG in the rat and that stimulation of TRPV1 with capsaicin overrides the pronociceptive effect of a low dose of WIN55,212-2, thus suggesting that capsaicin stimulates and low-dose WIN55,212-2 inhibits the same neurons, with opposite effects on their activity in the RVM (Scheme 1). However, costimulation with capsaicin and high doses of WIN55,212-2 modified the effects observed with the two substances alone by giving rise to either a delayed hyperalgesia or a mild analgesia followed by a delayed hyperalgesia (with 25 or 100 nmol/rat WIN55,212-2, respectively). This phenomenon might be explained by the previous evidence that, if CB1 and TRPV1 receptors are coexpressed on the same cell, their simultaneous stimulation, depending on the dose of agonists used, apart from merely exerting opposing actions, may also dramatically modify the effect that each receptor would elicit alone (Hermann et al., 2003). In particular, it is known that prior stimulation of CB1 sensitizes TRPV1 receptors (Hermann et al., 2003), thus causing a more efficacious Ca2+ influx. This might lead to a more rapid desensitization of antinociceptive stimulatory output neurons and hence to hyperalgesia (Table 1). Indeed, previous data suggested that capsaicin per se, if injected into the dorsal PAG, does activate and desensitize TRPV1-expressing pronociceptive, rather than antinociceptive output neurons, thereby inducing hyperalgesia first and analgesia later (McGaraughty et al., 2003). However, we never observed any hyperalgesia with capsaicin alone, probably because we injected a different PAG region. Thus, our observation of immediate capsaicin-induced analgesia might suggest that desensitization is not involved in this effect, whereas it might instead explain why capsaicin causes delayed hyperalgesia in the presence of high doses of WIN55,212-2. At any rate, we cannot entirely exclude that, also in the absence of WIN55,212-2, desensitization of pronociceptive output neurons occurs earlier in our experiments, thus contributing to the analgesic effect of capsaicin (Table 1). We also cannot rule out that some of the effects observed here with capsaicin or the lowest dose of WIN55,212-2 are due to drug diffusion to close areas (i.e., dorsal PAG, dorsal raphe, or RVM), although we only included in our study those rats whose injection area was located within the ventrolateral PAG, as assessed with the help of the Neutral Red dye. At least the early effects of the drugs, however, should not be due to diffusion because previous experiments (Lichtman et al., 1996; Walker et al., 1999; McGaraughty et al., 2003; Meng and Johansen, 2004; Hohmann et al., 2005) showed that TRPV1 or CB1 receptor stimulation in these areas does not cause immediate analgesia or hyperalgesia, respectively.
As with the “direct” CB1 agonist WIN55,212-2, the lowest dose of the FAAH inhibitor URB597 caused rapid and prolonged hyperalgesia, whereas an almost 10-fold dose caused rapid analgesia. Both effects were due to activation of CB1 receptors, as shown by the fact that they were abolished by the CB1 antagonist AM251. Upon coadministration of AM251, the hyperalgesic action of low-dose URB597 became an analgesic effect. This was mediated by TRPV1 receptors, because it was blocked by the TRPV1 antagonist capsazepine. Accordingly, 0.5 nmol/rat URB597 was barely sufficient to enhance the PAG concentrations of anandamide, which acts at both TRPV1 and CB1 receptors, and it mostly enhanced 2-AG, which is selective for CB1, thus explaining why the TRPV1-mediated analgesic effect of anandamide could be unmasked only after blockade of CB1 receptors. The highest dose of URB597 greatly enhanced both 2-AG and anandamide levels, thus possibly resulting in the strong activation only of those CB1 receptors coupled to antinociception. This effect was erased by AM251 but not by capsazepine, suggesting that the TRPV1-mediated analgesic effect undergoes desensitization with high levels of anandamide (Toth et al., 2005). Our data do not explain why 0.5 nmol/rat URB597 results in an immediate pronociceptive action, whereas the effect of 4 nmol/rat WIN55,212-2 was delayed. It is possible that locally elevated endocannabinoid levels act more efficaciously than exogenous CB1 receptor agonists (for example, see Marsicano et al., 2003). In fact, the pronociceptive effect of the low dose of URB597 is not due uniquely to 2-AG action at CB1 receptors, as suggested by the fact that selective inhibitors of the sn-1-selective diacylglycerol lipase responsible for 2-AG biosynthesis in the brain (Bisogno et al., 2004) abolished this effect but, unlike AM251, did not cause analgesia. This suggests again that, with 0.5 nmol/rat URB597, anandamide may activate both CB1 and TRPV1 and produce, in the absence of 2-AG, a net null effect on pain sensitivity.
In agreement with the elevation of both pronociceptive and antinociceptive endocannabinoids after FAAH inhibition, an intermediate dose (2.5 nmol/rat) of URB597 caused a biphasic response on nociception. Accordingly, with this dose of URB597, the ratio between the concentrations of anandamide and 2-AG was higher than with the low dose of the inhibitor, thus explaining why 2.5 nmol/rat URB597 caused a TRPV1-mediated analgesia even in the absence of CB1 blockade. The transient antinociceptive response of this dose of URB597 became an immediate and prolonged pronociception in the presence of capsazepine, or it remained unaltered in the presence of AM251. Based on these observations, we hypothesize the following points. 1) URB597 increases brain anandamide levels earlier than 2-AG, as suggested by previous data (de Lago et al., 2005). As a result, TRPV1 would be stimulated earlier than CB1, thus explaining not only the biphasic response but also why in this case there is analgesia and not hyperalgesia as with cotreatment of rats with capsaicin and high doses of WIN55,212-2 (Table 1). 2) When TRPV1 is blocked, anandamide concurs to induce CB1-mediated nociception. 3) When CB1 receptors are blocked, one observes only the TRPV1-mediated antinociception, which never lasts more than 30 to 45 min also when it is induced by exogenous capsaicin (Fig. 1).
To investigate the involvement of excitatory output neurons in URB597-induced TRPV1-mediated analgesia, we measured the effect of intra-PAG injections of 2.5 nmol/rat URB597 on spontaneous activity of RVM OFF and ON cells, which receive stimulatory inputs from CB1- and/or TRPV1-regulated PAG neurons (Palazzo et al., 2002; McGaraughty et al., 2003). URB597 produced an immediate stimulatory effect on the OFF cells and a more delayed and prolonged inhibitory effect on ON cells. Although both effects were erased by capsazepine, only the stimulatory action on OFF cells preceded the inhibitory effect of URB597 on nociception, which under the same conditions was maximal 5 min after administration. This suggests that the TRPV1-mediated analgesia that follows FAAH inhibition is due to the activation of PAG neurons directly stimulating OFF cells. However, the effect of 2.5 nmol/rat URB597 on both OFF and ON cells was also reversed by systemic AM251, whereas the analgesic effect of 2.5 nmol/rat URB597 was reversed only by capsazepine. This finding and the observation that, unlike the behavioral activities, the RVM cell ongoing activities occur and terminate within 30 min after injection and are not biphasic indicate that some dissociation exists between the complex response elicited by this dose of URB597 in the plantar test and the ongoing activities of RVM cells. Indeed, these data represent one of several examples (McGaraughty and Heinricher, 2002; McGaraughty et al., 2003; Meng and Johansen, 2004) of how spontaneous RVM activity is not always a predictor of behavioral effects. On the other hand, 2.5 nmol/rat URB597 did induce a tail-flick-related biphasic RVM cell response that correlated with the plantar test data in terms of both time response and, particularly, sensitivity to antagonists (Figs. 2 and 8). These findings suggest that, to fully correlate drug-induced analgesia/hyperalgesia with the activity of RVM cells, the responses of the latter to noxious stimuli need to be monitored in addition to their ongoing activities.
The CB1-mediated pronociceptive or antinociceptive effects of low or high doses of URB597 were immediately preceded by AM251-sensitive (but not capsazepine-sensitive) decreases and increases, respectively, of both spontaneous and tail-flick-related OFF cell activities, which were in turn immediately followed by opposing changes in the activity of ON cells. These data support the hypothesis that, although the pronociceptive effect of low doses of direct and “indirect” CB1 agonists are due to direct inhibition of excitatory antinociceptive PAG output neurons, the analgesic effect of the high doses are caused by depression of GABAergic interneuron activity and subsequent disinhibition of these neurons (Meng et al., 1998). These data also support the concept that, under physiological conditions, descending pain pathways are controlled by the activity of OFF rather than ON RVM cells and hence by the activity of PAG output neurons making direct synapses with these RVM neurons (Aimone and Gebhart, 1986; Wiklund et al., 1988; Tortorici and Morgan, 2002). It is possible that OFF cells inhibit the activity of ON cells, thus explaining why increased firing activity of the former is followed by decreased activity of the latter in this and in a previous study (Tortorici and Morgan, 2002). However, hyperalgesia induced by selective activation of ON cells has also been reported in naive rats (Neubert et al., 2004). Moreover, whereas in several experiments of the current study, OFF cells might control ON cell burst, this does not hold true for the 2.5 nmol/rat dose of URB597 where ON cell burst occurred simultaneously with or even preceded OFF cell pause, thus indicating that cessation of OFF cell firing was not required for the increase of the ON cell firing. Thus, the involvement of different PAG outputs modulating distinct RVM cells should not be excluded (Heinricher et al., 1987) (Table 1).
In conclusion, we have shown that, in healthy animals under conditions such as after FAAH inhibition leading to enhanced levels of anandamide in the PAG, this mediator not only participates together with 2-AG in CB1-mediated analgesia, it also produces TRPV1-mediated analgesia. Furthermore, we have reported for the first time that low doses of both direct and indirect CB1 agonists after injection into the ventrolateral PAG exert a pronociceptive effect, in agreement with the concept that brain CB1 receptors are presynaptically located not only in GABAergic interneurons, as originally believed, but also in excitatory (glutamatergic) neurons (Marsicano et al., 2003). Our data suggest multiple pathways and molecular targets for endocannabinoid control of descending pain pathways in rodents.
We thank Dr. Luciano De Petrocellis, Institute of Cybernetics “E. Caianiello” C.N.R. [Pozzuoli (Napoli), Italy] for valuable help and advice.
- Received July 27, 2005.
- Accepted November 10, 2005.
This work was partly supported by the VolkswagenStiftung (to V.D.M.) and Progretti di Ricerca di Rilevante Interesse Nazionale 2003 (Ministero dell'Istruzione dell'Università e della Ricerca, Rome, Italy) (to S.M.).
ABBREVIATIONS: CB1, cannabinoid receptor type 1; CB2, cannabinoid receptor type 2; ACSF, artificial cerebrospinal fluid; 2-AG, 2-arachidonoylglycerol; FAAH, fatty acid amide hydrolase; % MPE, percentage of the maximal possible effect; PAG, periaqueductal gray; RVM, rostral ventromedial medulla; TRPV1, transient receptor potential vanilloid type-1; capsaicin, (E)-N-[(4-hydroxy-3-methoxyphenyl)methyl]-8-methyl-6-nonenamide; capsazepine, N-[2-(4-chlorophenyl)ethyl]-1,3,4,5-tetrahydro-7,8-dihydroxy-2H-2-benzazepine-2-carbothioamide; URB597, cyclohexylcarbamic acid-3′-carbamoyl-biphenyl-3-yl ester; O-3841, octadec-9-enoic acid 1-methoxymethyl-2-(fluoro-methyl-phosphinoyloxy)-ethyl ester; 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; ANOVA, analysis of variance.
- The American Society for Pharmacology and Experimental Therapeutics