Nociceptin (1–13)NH 2 Inhibits Stimulated Calcitonin-Gene-Related-Peptide Release From Primary Cultures of Rat Trigeminal Ganglia Neurones

Abstract

In this work we have developed and characterized primary cultures of neonatal rat trigeminal ganglia neurones; calcitonin-gene-related-peptide (CGRP) released from cells was taken as a marker of neuronal function. A significant and consistent increase in CGRP secretion was elicited by non-specific (56 mM KCl or veratridine) or specific (capsaicin) depolarizing stimuli. This paradigm was subsequently used to investigate the effects of nociceptin, an opioid-like peptide involved in central and peripheral control of nociception. We found that the nociceptin analogue nociceptin (1–13)NH₂ (NOC) did not affect baseline CGRP release, but it reduced in a concentration-dependent manner CGRP release induced by all tested stimuli. NOC-induced reduction was statistically significant from 0.01 nM onward and achieved maximal effects at 10 nM. Such effects of NOC were seemingly mediated by the activation of specific ORL1 receptors, as a well-known nociceptin antagonist, N(Phe¹)nociceptin (1–13)NH₂, was able to completely revert NOC inhibition of capsaicin-stimulated CGRP release.

Introduction

The heptadecapeptide nociceptin, also referred to as orphanin FQ (N/OFQ), has been identified as the endogenous ligand for the opioid receptor-like receptor ORL1 (OP₄) (1, 2). Although structurally related to the opioid peptide dynorphin A, N/OFQ does not bind to µ-, δ- or κ- opioidergic receptors, nor are its effects antagonized by naloxone (3, 4). However, similarly to opioids, N/OFQ activates inwardly rectifying K⁺ channels and inhibits both voltage-gated (N/P type) Ca⁺⁺ channels and adenylate cyclase activity (3), and, like dynorphin A, it exerts opioid modulating actions (5). Supraspinal injection of N/OFQ decreases morphine-induced analgesia (anti-opioid activity), and spinal administration induces analgesia or potentiates the analgesic effects of morphine (pro-opioid activity) (6–8), suggesting an involvement of this peptide-receptor system in the regulation of nociception. Indeed, in the central nervous system (CNS), N/OFQ has been found to exert both anti-analgesic, and analgesic and/or hyperalgesic effects, depending on the experimental models and route of administration (9–11). In the peripheral nervous system (PNS), N/OFQ mediates mainly antinociceptive effects similarly to classic opioids but via activation of its specific ORL1 receptor (12, 13).

The ORL1 receptor is widely expressed throughout the CNS in rodents and humans (14–16), supporting a role of this system in several CNS functions (17), such as the stress response, anxiety, learning and memory processes, and mechanisms of reinforcement and reward during drug addiction, besides pain control (18). An involvement of N/OFQ has been also hypothesized in the pathogenesis of migraine, in which the activation of the trigeminal afferents causes vasodilatation of dural vessels leading to neurogenic inflammation and pain. N/OFQ was shown to inhibit neurogenic dural vasodilatation in the rat (19); however, the exact mechanism of action underlying such an effect was not determined in this study. ORL1 mRNA expression and N/OFQ immunoreactivity have been detected in rat and human trigeminal ganglia, which indicates a possible role of N/OFQ in the control of trigeminal nociception (20, 21). Interestingly, low levels of N/OFQ were recently demonstrated in migraine patients, suggesting that deregulation of the opioid-like system may be associated with increased susceptibility to migraine (22). During migraine attacks, trigeminal ganglia innervation of meningeal vessels mediates the release of vasoactive peptides such as calcitonin gene-related peptide (CGRP) and substance P, followed by dural vasodilatation and subsequent activation of nociceptive trigeminal sensory fibres (23, 24). In fact, CGRP was found to be elevated in jugular blood samples of patients during migraine attacks (25), and effective antimigraine drugs, such as 5HT1_B/D agonists (triptans), can reduce CGRP secretion from trigeminal nerve endings (26).

N/OFQ attenuates peripheral neuropeptide release in a variety of tissues, including the bronchi (27), vas deferent and skin (28, 29), as well as from peripheral sensory nerve endings in in vivo models of neuroinflammation (30, 31). Therefore, we hypothesize that N/OFQ can regulate trigeminal nociception via the modulation of neuropeptide release, particularly reducing CGRP secretion. To test this hypothesis, we used primary cultures of rat trigeminal ganglia neurones. This is a well-established tool to investigate the patho-physiological mechanisms of migraine, as well as to test putative antimigraine drugs. We have investigated the effects of a full ORL1 agonist, nociceptin (1–13)NH₂ (NOC) (32), on CGRP-release from trigeminal neurones, under basal conditions and after stimulation by non-specific or specific depolarizating agents. In these experimental conditions, we have found that NOC reduced CGRP-stimulated release without affecting basal secretion. The inhibitory effects of NOC were mediated by the ORL1 receptor activation, and were almost completely reversed by the selective ORL1 antagonist, N(Phe¹)nociceptin (1–13)NH₂ (NPheNOC) (33). Altogether, evidence presented in this manuscript indicates that trigeminal nociception may be reduced by the activation of the N/OFQ-ORL1 system, and the latter may be thought of as a novel candidate target for antimigraine drugs.

Methods

Drugs

Veratridine and capsaicin were purchased from Sigma (Sigma Chemicals Co., St. Louis, MO, USA), and dissolved in 100% ethanol at 10 mm concentration; further dilutions were made in the release-medium. NOC and the selective ORL1 antagonist, NPheNOC, were purchased from Tocris (Cookson, UK), dissolved in distilled water at 100 mm and aliquots stored at −35°C, avoiding repeated cycles of freezing and thawing. Working dilutions were made using the release-medium. For immunocytochemistry, mouse monoclonal anti-rat phosphorilated neurofilament 200 (pNF200) was from Sigma. Secondary antibodies were from Jackson Immuno Research Laboratories, Inc. (West Grove, PA, USA). Vectashield mounting fluid was from Vector Laboratories Inc (Burlingame, CA, USA).

Trigeminal neurone cultures

Trigeminal neuronal cultures were prepared from 6- to 7-day-old Wistar rats as previously described (34), with some modifications introduced in our laboratory. The use of animals for this experimental work has been approved by the Italian Ministry of Health (licensed authorization to P. Navarra). In brief, animals were decapitated and, after removal of the brain, basal skulls were exposed and trigeminal ganglia from both sides were aseptically removed. Tissues were collected in a Petri dish containing 3–5 ml of ice-cold phosphate buffer saline without Ca⁺⁺ and Mg⁺⁺ (PBS w/o; Sigma), supplemented with antibiotics (100 IU/ml penicillin and 100 μg/ml streptomycin; Sigma) and d-glucose (6 g/l), and cut into small fragments. At this point, tissues were transferred into a 15-ml tube and further digested in 5 mg/ml collagenase (Biospa, Milan, Italy) for 20 min at 37°C, followed by 5 min incubation in 0.125% trypsin (EuroClone, Pero, Milan, Italy). Ten µl of DNAse I 1 mg/ml (2320 Kunitz/ml) was added in the last 5 min of incubation. After 2 min centrifugation at 1200 g , digested tissues were resuspended in 5 ml Ham's F12 medium (Biospa, Milan, Italy), containing 10% heat-inactivated endotoxin-free fetal calf serum (FCS; Gibco, Milan, Italy) and antibiotics (100 IU/ml penicillin and 100 µg/ml streptomycin), and cells were mechanically dissociated using a Pasteur pipette. Cell suspension was plated on a 25-cm² flask (Corning, Turin, Italy) and incubated at 37°C in a humidified atmosphere containing 5% CO₂ for 2–3 h (preplating incubation).

Because sensory ganglia consist of neuronal cells and satellite cells, the latter being part of a sheath that surrounds neurones (35), satellite cells may represent an important contaminant of neuronal cultures. These cells share several properties of glial cells (35), and glia tend to adhere to the bottom of uncoated culture flasks (36). Thus, we introduced into the standard procedure of isolation of trigeminal neurones (34) a preplating incubation step, and let non-neuronal cells adhere to the uncoated tissue culture flasks for a few hours. At the end of the preplating incubation, neurones were harvested from the flask and plated in 24-well tissue culture plates, previously coated with poli-D-lysine (40 µg/ml; MW 70000–130000, Sigma) at a density of 100 000 cells/well. The incubation volume was 1 ml/well of complete culture medium (see above), enriched with 50 ng/ml of 2.5 S murine Nerve Growth Factor (Alexis-Vinci Biochem, Vinci, Florence, Italy). The culture medium was changed within 24 h from seeding and 10 µm cytosine arabinoside (ARA-C) was added to further reduce non-neuronal cell growth (37). All experiments were carried out 6–7 days after dissection, when cells reached complete maturation (Figure 1a).

Figure 1

Rat primary cultures of trigeminal neurones after 7 days in vitro. (a) Morphology of trigeminal neurones after maturation in vitro. Image taken in phase contrast microscopy (20×). (b) Cultures were stained with the antibody against the phosporilated neurofilament 200 (pNF200), a specific marker of sensitivity neurones. Scale bars are 50 µm in both panels.

CGRP release studies

For these studies, at the beginning of experiments, the incubation medium was replaced and cells were incubated with 300 μL of complete Ham's F12 medium (containing 10% FCS, 1 μM bacitracin, 10 μg/ml aprotinin, without NGF) at 37°C. In these conditions we assessed a basal release of CGRP, which could be increased by incubating the cultures with medium containing different stimulating agents. In the experiments in which KCl was used as depolarizing agent, 56 mm KCl solutions were prepared. These solutions consisted of 56 mm KCl, 67 mm NaCl and other ions at the same concentrations as in the incubation medium, plus aprotinin, bacitracin and 0.1% bovine serum albumin (BSA). When used, NOC was dissolved directly into 56 mm KCl solutions.

In the experiments with the selective ORL1 antagonist, a 5-min pre-incubation in the presence of this compound was performed. After pre-incubation, the medium was replaced with fresh release-medium containing testing drugs. At the end of the experiments, media were collected and stored at −35°C until the assays were performed.

CGRP radioimmunoassay (RIA)

CGRP release was measured by a radioimmunoassay technique validated in our laboratory. We generated the antiserum used in this assay by immunization of New Zealand white rabbits with synthetic human α-CGRP (hα-CGRP) coupled to BSA with carbodiimide as previously described (38). The cross-reactivity of the anti-hα-CGRP serum with several other peptides and CGRP analogues was tested by adding up to 1 µg/tube of hα-CGRP, hβ-CGRP, rat α-CGRP (rα-CGRP), rβ-CGRP, rα-CGRP (8–37), rα-CGRP (29–37), h-amylin, r-amylin, cholecystokinin-8 (CCK-8) sulphated, neurokinin A (NKA), neurokinin B (NKB), physalaemin, somatostatin, substance P (SP), thyrocalcitonin, and vasoactive intestinal polypeptide (VIP). Iodinated hα-CGRP was purchased from Amersham (Amersham, UK). The RIA was performed in a buffer containing 10 mm sodium phosphate, 154 mm NaCl, 25 mm ethylene diamine tetraacetic acid (EDTA), 0.01% thimerosal, 0.5% BSA and 0.03% Tween 20, pH 7.2. The anti-hα-CGRP serum was used at the final dilution of 1:120 000, which binds approximately 40% of the iodinated peptide (≈6000 cpm per tube). The RIA was performed as follows: an initial incubation was carried out at 4°C in disposable plastic tubes containing 100 μl of anti-hα-CGRP serum, 100 μl of sample or standard solutions and 100 μl of RIA buffer. After 24 h, 100 μl of [¹²⁵I]-hα-CGRP (6000 cpm/tube) was added and the incubation continued at 4°C for 48 h. Separation of free from bound α-CGRP was achieved by adding 100 μl of anti-rabbit goat serum, diluted 1:20 in RIA buffer, and 500 μl of 6.6% polyethylene glycol solution in 50 mm phosphate buffer. After 2 h incubation at 4°C, the tubes were centrifuged at 3000 g for 30 min at 4°C, the supernatant was discarded and the pellet counted in a γ-counter. The standard curve ranged from 1.95 to 1000 pg/tube of rα-CGRP. Each sample was assayed in duplicate. The release findings were expressed as pg/well. The anti-hα-CGRP serum produced in our laboratory proved to be highly specific. Cross-reactivity with CCK-8 sulphated, NKA, NKB, physalaemin, somatostatin, SP, thyrocalcitonin and VIP (each added at concentrations of ≤1 µg/tube) was less than 0.01%. Cross-reactivity with h- and r-amylin was, respectively, 0.1% and 0.22%. Taking as 100% the cross-reactivity with hα-CGRP, cross-reactivity with hβ-CGRP, rα-CGRP, rβ-CGRP and rα-CGRP (8–37) was, respectively, 68.2%, 45.5%, 67.4% and 157.5%. Rα-CGRP (29–37) partially displaced antigen-antibody binding at the maximal concentration (3 ng/tube) (by 31.1%).

On the whole, these findings indicate that: (i) the polyclonal antibody is mainly directed towards the middle region of the peptide; and (ii) the presence of the amino-terminal sequence reduces the affinity of the antibody towards the peptide. The mean IC50 of the standard curve was 50.8 ± 2.0 pg/tube (n = 10), and non-specific binding of the labelled ligand was not different from the background of the γ-counter. The intra-assay (n = 6) and inter-assay (n = 10) coefficients of variation were, respectively, 1.27% and ±0.95% at the lowest (1.95 pg/tube) and 13.7% and ±12.6% at the highest (1000 pg/tube) levels of standard. The lowest concentration that could be measured with 95% confidence (i.e. 2 s.d. at zero) was 1.95 pg/tube. Thus, the detection limit was 19.5 pg/ml of the original sample (as 100 µl of each sample were used in the RIA system).

Immunocytochemistry

For immunocytochemical (ICC) characterization of neuronal cultures, cells were plated on glass coverslips coated with 1 mg/ml poli-D-lysine, at a density of 5000 cells/well, and kept in culture for 7 days. At this time, cells were washed three times with PBS containing Ca⁺⁺ and Mg⁺⁺ (PBS-w) and fixed in 4% paraformaldehyde for 20 min. After fixation, cells were washed twice with PBS-w and blocked with 0.5% BSA in PBS-w for 30 min. Primary antibodies were diluted in 0.1% BSA and 0.2% Triton X100 in PBS-w (anti-pNF200 1:400) and cells incubated overnight at 4°C. Secondary antibodies were donkey anti-mouse conjugated with Texas-red. Secondary antibodies were incubated for 1 h at 37°C and diluted 1:200 in PBS-w in 0.1% BSA. Coverslips were mounted using Vectashield mounting media. Images were obtained on a Nikon Eclipse TE300 inverted fluorescence microscope equipped with a Cool SNAP professional digital camera and LUCIA-G/F imaging software.

Statistical analysis

All data are presented as mean ± s.e.m., unless otherwise noted. Significant differences between groups were assessed by one-way analysis of variance (anova) with Newman-Keuls' multiple comparison post-test, unless otherwise stated. All data were analysed using a Prism^TM computer program (GraphPad, San Diego, CA, USA).

Results

Primary rat trigeminal neurones release CGRP

In preliminary experiments, we characterized the release of CGRP-LI from primary cultures of rat trigeminal ganglia neurones. Cells were cultured for 7 days in vitro in the presence of 50 ng/ml NFG (as described in Methods), to reach complete maturation (Figure 1a). At this time, neurones highly expressed the specific marker for sensory neurones (39), pNF200 (Figure 1b). In these conditions, it was possible to detect CGRP-LI immunoreactivity in the incubation media. The basal secretion of CGRP ranged from 53.11 ± 3.29 pg/well after 10 min incubation to 76.26 ± 9.45 pg/well after 20 min incubation (data expressed as means ± s.e.m.). Basal CGRP-LI release could be increased by depolarizing solutions containing 56 mm KCl. As shown in Figure 2, the stimulation induced by KCl was statistically significant after 10-min incubation (P < 0.001) and persisted for up to 40 min, the later time point studied. Regardless of whether or not cells were stimulated with KCl, the amount of CGRP-LI in the incubation media did not significantly vary with time during the time frame studied. Thus, all subsequent experiments were carried out at 10 min, with the exception of KCl experiments, carried out at 20 min as this stimulation proved to be the most effective (about +90% increase vs controls, compared with +65% after 10 min of incubation). CGRP-LI release was also increased in a dose-dependent manner by a different depolarizing agent, veratridine (Figure 3). The stimulatory effect of veratridine was statistically significant from 0.01 µm (P < 0.001) onward, and reached maximal effect (+100% increase vs controls) at 1 µm, used in subsequent experiments. Figure 4 shows that CGRP-LI release was also significantly stimulated by capsaicin given in the range 0.01–10 µm. The dose–response curve of capsaicin effect shows a typical bell-shaped profile, with maximal stimulation obtained at 0.1–0.3 μM concentrations (about +150% vs controls), whereas higher concentrations were less effective.

Figure 2

Basal and 56 mm KCl-evoked CGRP-LI release after 10, 20 and 40 min incubations. 56 mm KCl induced a statistically significant increase in CGRP-LI release with respect to basal controls. Data are expressed as pg of CGRP-LI/well, means ± s.e.m. of n = 4 replicates for each experimental group. ∗∗∗P < 0.001 vs controls at each time point.

Figure 3

Effect of veratridine on CGRP-LI release. Data are expressed as pg of CGRP-LI/well/10 min, means ± s.e.m. of n = 4 replicates for each experimental group. ∗∗∗P < 0.001 vs controls.

Figure 4

Effect of capsaicin on CGRP-LI release. Data are expressed as pg of CGRP-LI/well/10 min, means ± s.e.m. of n = 8 replicates for each experimental group. ∗∗∗P < 0.001 vs controls.

Nociceptin reduces CGRP-LI release

After the initial characterization, we tested the effects of N/OFQ on CGRP release, using the selective ORL1 agonist NOC. In resting neurones, NOC, given in the range 0.01 nm to 10 μM, did not exert any significant effect (data not shown). In contrast, when NOC was used with 56 mm KCl, a concentration-dependent reduction of KCl-evoked CGRP-LI release was observed. Such inhibitory effect of NOC was statistically significant (P < 0.01) from 0.01 nm onwards (Figure 5) and reached the maximal inhibitory effect at 10 nm; in fact, we did not observe any additional inhibition at higher concentrations (up to 1 μM, not shown). As shown in Figure 6, NOC was also able to reduce the stimulatory effect of 1 µm veratridine. In a similar manner to the inhibitory effects observed using KCl, NOC inhibition on CGRP-LI stimulated by veratridine was concentration-dependent in the range 0.01–10 nm, with maximal effect observed at 10 nm (−52.5% reduction vs veratridine alone). Finally, we observed a similar inhibitory effect of NOC using the more specific stimulus, 0.1 μM capsaicin, to increase CGRP-LI release. And also in this paradigm, NOC reduced CGRP-LI stimulated release in a concentration-dependent manner (Figure 7), with maximal reduction observed at 10 nm (−31% vs capsaicin alone, P < 0.01). In order to demonstrate that the inhibitory effects of NOC were mediated by the activation of its specific receptor, we performed a series of experiments using a competitive and selective antagonist to the ORL1 receptor, NPheNOC (33). This antagonist had no significant intrinsic effects on capsaicin-stimulated CGRP-LI release when tested by itself (1–10 µm) (Figure 8a), while it was able to reverse the inhibitory effect of NOC over capsaicin stimulation in a concentration-dependent manner. As shown in Figure 8(b), NOC inhibition was completely reverted by 10 µm NPheNOC (P < 0.05 vs NOC plus CAP alone).

Figure 5

Effect of NOC on KCl-evoked CGRP-LI release. Data are expressed as pg of CGRP-LI/well/20 min, means ± s.e.m. of n = 5 replicates for each experimental group. ∗∗∗P < 0.001 vs controls; °°P < 0.01 vs 56 mm KCl alone.

Figure 6

Effect of NOC on veratridine-evoked CGRP-LI release. Data are expressed as pg of CGRP-LI/well/10 min, means ± s.e.m. of n = 5 replicates for each experimental group. ∗∗∗P < 0.001 vs controls; °°P < 0.01 vs veratridine alone.

Figure 7

Effect of NOC on capsaicin-evoked CGRP-LI release. Results are expressed as pg of CGRP-LI/well/10 min. Data are the means ± s.e.m. of two independent experiments, representative of six experiments, each carried out with n = 3–6 replicates for each experimental group. ∗∗∗P < 0.001 vs controls; °°P < 0.01 vs capsaicin alone.

Figure 8

Effects of the NOC selective antagonist NPheNOC. (a) NPheNOC does not modify capsaicin-stimulated CGRP-LI release. (b) NPheNOC counteracts NOC inhibition of capsaicin-stimulated CRRP-LI release; representative of two independent experiments with similar findings. Results are expressed as pg of CGRP-LI/well/10 min. Data are the means ± s.e.m. of n = 5 replicates per group. ∗∗P < 0.001 vs capsaicin alone; °P < 0.05 vs capsaicin plus NOC.

Discussion

In the present study, primary cultures of rat trigeminal ganglia neurones were used to investigate the involvement of the N/OFQ-ORL1 system in the regulation trigeminal nociception. In preliminary experiments, we have first characterized the in vitro model. As a specific marker of neuronal functions we assessed in the incubation medium the release of CGRP. Interestingly, this is a critical vasoactive peptide, which mediates in vivo neurogenic inflammation; therefore the assessment of CGRP in vitro also represents a useful tool to investigate the pain mechanisms involving trigeminal ganglia. In our cultures, we were able to detect a baseline CGRP secretion, which was significantly and consistently increased by different types of stimuli, namely 56 mm KCl solutions, veratridine and capsaicin. The first two should be considered as non-specific depolarizating agents, with veratridine acting primarily through the opening of Na⁺ channels, whereas 56 mm KCl solutions elicit directly Ca⁺⁺ ion influx within the cells (40, 41). In contrast, the C-fibre stimulating agent capsaicin is considered a specific stimulus, acting via binding to and activation of vanilloid receptors (42). In this paradigm, the ORL1 full agonist NOC had no significant effects on baseline CGRP release, but it consistently inhibited CGRP-stimulated release, independent from the underlying mechanism of stimulation. These data thus confirm that N/OFQ can modulate the permeability of several ion channels in neurones, including K⁺–, Ca⁺⁺ and also Na⁺ channel-dependent currents (43).

Experiments with the N/OFQ antagonist indicate that NOC inhibitory effects are indeed mediated by the activation of its specific ORL1 receptors. NPheNOC is a selective and competitive antagonist to ORL1 (33). NPheNOC binds to ORL1 and prevents the effects of N/OFQ in the CNS and PNS. In our model, NPheNOC did not have any effect per se, but it was able to reverse NOC inhibition of capsaicin-evoked CGRP release when given at 10 µm. Consistent with our results, NPheNOC had no significant effect in a model of neurogenic meningeal vasodilatation (19), which led these authors to postulate that there is no activation of ORL1 in trigeminal nerve afferents under basal conditions. However, NPheNOC was found to exert an intrinsic analgesic effect when administered centrally, which is reminiscent of the opposite effects displayed by N/OFQ in the CNS vs the periphery. In this regard, while different effects were described when N/OFQ was centrally administered (see Introduction), the activation of this peptide-receptor system has unambiguous antinociceptive effects in peripheral ganglia pain transmission. N/OFQ and ORL1 are constitutively expressed in dorsal root ganglia neurones (DRG), and they undergo up-regulation during experimentally induced neuropathic pain (44). Functional evidence shows that N/OFQ inhibits neuropeptide release from peripheral tissues such as airways (27), gastrointestinal tract (29), and vas deferens (28). Moreover, N/OFQ is able to block antidromic neurogenic vasodilatation via the suppression of CGRP release from peripheral nerve endings (30, 31).

More recent evidence also suggests a possible role of the N/OFQ-ORL1 system in the modulation of trigeminal nociception. N/OFQ and ORL1 were found expressed in human trigeminal ganglia, colocalized with substance P and CGRP (21); in small trigeminal ganglia neurones from the mouse, N/OFQ was able to inhibit Ca⁺⁺ channel currents (45). Finally, CGRP is the major constituent of sensory neurones and is considered a marker of activation of the trigeminal-vascular system (25). Bartsch and colleagues have demonstrated that N/OFQ is able to reduce neurogenic dural vasodilatation in the rat (19), suggesting that this effect could be mediated by the inhibition of CGRP release from trigeminal fibres innervating meningeal vessels. Our findings represent a strong in vitro correlation to the above hypothesis and shed more light on the mechanisms of N/OFQ antinociceptive activity in peripheral pain. In fact, most of the previous findings involving N/OFQ (30, 31) were obtained in experimental models where the observation was limited to nerve endings, thereby missing nociceptin effects on the neuronal cell body. In contrast, here for the first time we were able to investigate the effect of N/OFQ on trigeminal neurones taken as a whole; our findings indicate that the N/OFQ-ORL1 system is definitely involved in trigeminal pain transmission and may represent a new target for antimigraine therapeutic agents.

Acknowledgements

This work was supported in part by Grünenthal–Formenti.