Abstract
It was demonstrated that both nociceptin, a novel opioid neuropeptide, and its receptor are present in trigeminovascular neurons. In an animal model nociceptin dose-dependently inhibited neurogenic dural vasodilatation. These results suggest that nociceptin may be involved in neurovascular headaches such as migraine. To test this hypothesis, we studied circulating nociceptin levels in 18 patients suffering from migraine without aura and in 24 controls. Headache-free migraineurs had significantly lower nociceptin levels than controls (5.79 ± 1.82 vs. 9.74 ± 2.43 pg/ml, P < 0.0001, Student's t-tests). Nociceptin levels were further reduced in six patients studied in the first 3 h of typical migraine attacks (1.04 ± 0.17 pg/ml). Nociceptin levels correlated with the frequency of attacks in this group of migraineurs. Lower interictal nociceptin levels may contribute to a defective regulation of trigeminovascular neurons in migraineurs which might be important in the pain process of migraine.
Introduction
Migraine is currently considered a primary neurovascular headache in which the activation of intracranial nociceptors and an alteration in the function of the endogenous pain control system lead to paroxysmal head pain with characteristic accompanying symptoms. There is extensive evidence that the small-diameter unmyelinated fibres of the first trigeminal division innervating the supratentorial dura and its vessels (the trigeminovascular system) have a central role in the generation of migraine pain (1). According to animal studies the activation of the trigeminovascular system results in the release of vasoactive neuropeptides (CGRP and substance P) (2) with consequent vasodilatation and protein extravasation (1), a phenomenon termed neurogenic inflammation. Electrical stimulation of the human trigeminal ganglion also leads to increased CGRP and substance P levels in the cranial venous outflow (2). More specific stimulation of the intracranial pain producing structures such as the superior sagittal sinus causes the cranial venous release of CGRP but not substance P (3). That the trigeminovascular system is activated in migraine is suggested by elevated levels of CGRP in the ipsilateral external jugular vein during the attack, which normalize after successful treatment (4). Trigeminal input elicited by painful stimuli is relayed to the trigeminocervical complex (5); higher order processing occurs in the ventroposteromedial thalamus, in the medial nucleus of the posterior complex, and in the intralaminar nuclei (6).
Rostral dorsomedial brainstem structures are activated during the migraine attack and immediately after its successful treatment (7). The area of activation most likely corresponds to the dorsal raphe nucleus and/or locus coeruleus. Interestingly, electrical stimulation of the same region can cause migraine-like headaches in patients who underwent electrode implantation for the control of intractable pain (8). The stimulation of the periaqueductal grey inhibits superior sagittal sinus-evoked trigeminal neuronal activity in the cat (9) and also inhibits the neuronal activity in the spinal trigeminal nucleus (10). These data raise the possibility that the dysfunction of the rostral brainstem might lead to the disinhibition of trigeminal afferents and, ultimately, to migraine pain.
Nociceptin is an endogenous ligand for the opiate-4 (OP-4) receptor (11), previously termed orphan opioid receptor-like (ORL-1) receptor (12, 13). The activation of OP-4 receptors reduces neuronal excitability and inhibits presynaptic neurotransmitter release by inhibiting high-voltage-activated Ca2+ channel currents and activating an inwardly rectifying K+ conductance (14). The OP-4 receptor is abundantly expressed in various CNS structures in rodents (15), nonhuman primates (16) and in humans (17), supporting a role for nociceptin in a multitude of CNS functions, including motor and balance control, reinforcement and reward, nociception, the stress response, sexual behaviour, aggression, and autonomic control of physiologic processes. In the sensory system OP-4 receptors are expressed by the dorsal horn of the spinal cord, trigeminal sensory nuclei, spinothalamic and trigeminothalamic tracts, locus coeruleus, the periaqueductal grey, raphe complex, thalamic structures (intralaminar and medial nuclei) and the sensory cortex (16, 18). Extensive experimental evidence suggests that the nociceptin/OP-4 system is involved in sensory processing. In the peripheral nervous system nociceptin has antinociceptive effects, similarly to classical opioids (19), although its action is independent of antinociception produced at the classical opioid receptors (20). In the CNS nociceptin can have antianalgesic, but also analgesic and hyperalgesic effects, depending on the experimental settings, testing paradigm, animal species, doses, and route of administration (21–23).
Approximately 70% of neurons in the human trigeminal ganglion exhibit nociceptin immunoreactivity and express OP-4 receptor mRNA (24). The majority of nociceptin-positive neurons are medium-sized (30–60 ìm). In these cells nociceptin is colocalized with CGRP and substance P, marker peptides of the trigeminovascular system. This distribution suggests that nociceptin may be involved in the regulation of neuropeptide release from trigeminal nerve terminals and, perhaps, in neurovascular headaches such as migraine. To test this hypothesis, we measured plasma nociceptin levels in migraine without aura patients.
Patients and methods
Patients
Eighteen migraineurs whose attacks were not preceded by aura (13 females and 5 males, mean age: 35.4 ± 14.7 years) with a mean disease duration of 15.4 ± 15.3 years, and a group of 24 age-matched healthy controls were enrolled. Subjects also suffering from other significant headaches (migraine with aura, chronic tension type headache, frequent episodic tension type headache and cluster headache) were excluded. We did not exclude minor concurrent diseases, but subjects receiving regular medical treatment (other than acute treatment of migraine) were excluded. We also excluded patients receiving prophylactic treatment for migraine. The demographic data of patients and controls are summarized in Table 1. Informed consent was obtained. The study was approved by the local Ethics Committee.
Demographic data of migraineurs and controls
Data are presented as mean ± SD (range) if not otherwise indicated. na, not applicable.
Blood samples and nociceptin measurements
Blood was sampled from the antecubital vein of fasting headache-free subjects, between 0830h and 1000 h. This was at least 7.5 h before or after the migraine attack. Six patients were also studied during the first hours of a typical migraine attack; their samples were drawn in the same time window. Blood samples were collected in 6 ml vacutainers containing K-EDTA and aprotinin (0.6 TIU/ml of blood). The plasma, collected in mini-sorb tubes, was kept frozen at −80°C until analysis. The extraction and radioimmunoassay of nociceptin were carried out as previously described (25), using a commercially available nociceptin kit (Phoenix Pharmaceuticals, Inc., Belmont, CA 94002, USA).
Statistics
The distribution of data was tested using the Kolmogorov-Smirnov test with Lilliefors’ method. As the distribution was Gaussian, we compared the plasma nociceptin levels of headache-free migraineurs and controls using Student's t-tests. Due to the small number of samples, nociceptin levels obtained during migraine attacks were not included in the statistical analysis. Correlation between nociceptin levels and patient characteristics was evaluated by Pearson's tests. Linear and nonlinear regression models were used to further clarify any association between nociceptin levels and the time span to/from the nearest attacks, using Graph Pad's Prism software.
Results
Interictal nociceptin levels of migraineurs were 5.79 ± 1.82 pg/ml (range 3.75–9.25), significantly lower (P < 0.0001) than those of healthy controls (9.74 ± 2.43 pg/ml, range: 7.11–15.86). During the attacks, nociceptin levels further decreased to 1.04 ± 0.17 pg/ml (range 0.91–1.37) (Fig. 1). Nociceptin levels showed a significant correlation with the number of attacks during the last month (r = −0.625, P = 0.013) but did not correlate with age, length of disease or length of attacks in migraineurs and with age in controls. Male and female subjects’ nociceptin values were not statistically different within each group. There was no association (either linear or nonlinear) between the time span to/from the nearest migraine attack and nociceptin levels (Fig. 2).
Plasma nociceptin levels in migraineurs in the interictal phase and during attacks compared to controls. Nociceptin levels are expressed in pg/ml. Mean values are represented with a rectangle inside the box. Boxes represent the standard error of mean (SEM) and whiskers show the standard deviation (SD). Headache-free migraineurs have significantly lower nociceptin levels than controls (P < 0.0001, Student's t-test). Nociceptin levels are further decreased during attacks. Due to small sample numbers nociceptin levels obtained during migraine attacks were not included in the statistical analysis.
Correlation of plasma nociceptin levels in attack-free migraineurs with the time span to/from the nearest migraine attack. Nociceptin levels are expressed in pg/ml. ▪ nociceptin levels in relation to time from last attack; ▴ nociceptin levels in relation to time to the next attack. There was no correlation between the time span to/from the nearest migraine attack and nociceptin levels. Regression calculations (linear and nonlinear models) failed to detect any relationship between these two variables. Linear regression results (—— and ------ lines for the respective groups) are presented as an example; the slopes do not differ significantly from zero.
Discussion
In this study migraine without aura patients had significantly lower circulating nociceptin levels than controls and nociceptin levels further decreased during the early phase of the attacks. The latter finding needs to be confirmed on larger samples.
Although nociceptin and its receptor are subjects of intense research, surprisingly few human data have been published to date. Thus, we do not have any direct evidence about the effects of endogenous nociceptin release or exogenous nociceptin challenge and no information about its kinetics. We also lack information about the effect of gender variation or hormonal changes (e.g. menstrual cycle) on nociceptin levels or about any circadian changes these levels may display. In this study, the male/female ratio was different in the migrainous and control groups, but there was no significant difference between males’ and females’ nociceptin levels within each group. So it seems that gender per se does not influence circulating nociceptin levels.
Animal data suggest that the biological effect of intracerebroventricular nociceptin lasts 10–120 min in various assays (22). In the present study headache-free blood samples were taken at least 10.5 h after and (with one exception) at least 7.5 h before an attack, so it seems improbable that the lower nociceptin levels in migraineurs should be the consequence of the pain itself. The nociceptin level of the patient who developed migraine an hour after blood sampling was 4.87 pg/ml, much higher than levels obtained during acute attacks.
The hitherto published studies of nociceptin in humans give conflicting results. Some acute pain conditions may not influence the quantity of circulating nociceptin as both CSF and plasma nociceptin levels were unchanged during labour (26). Decreased nociceptin levels were observed in fibromyalgia (27) and recently, in cluster headache patients when headache-free during the cluster period (28); elevated ones in aetiologically heterogeneous groups of acute, subacute and chronic pain patients (29). Thus, in contrast to previous hypotheses that nociceptin concentration would either represent a response to chronic painful stress or reflect the existence and duration of pain syndromes, it seems that this neuropeptide may have a more complex role in pain processing.
The human temporal artery and dermal tissue do not show nociceptin immunoreactivity (30). Nociceptin injected into intact muscles of healthy persons does not cause pain, although it does increase local tenderness when injected to already tender muscles of otherwise healthy persons (30). Moreover, neither nociceptin nor the OP-4 receptor could be demonstrated in synovial tissue and fluid of arthritis patients (31). These findings suggest that nociceptin may not have a major algesic effect on the periphery.
The demonstration of OP-4 receptor mRNA and nociceptin immunoreactivity in CGRP- and SP-containing trigeminovascular neurons in humans (24) suggests that nociceptin might be involved in the regulation of neuropeptide release from trigeminal sensory nerve terminals. A similar effect of nociceptin, with the inhibition of CGRP and SP release from capsaicin-sensitive sensory fibres, was described in the rat (32); nociceptin pretreatment also attenuated neurogenic inflammation, evoked by electrical stimulation of the saphenous nerve, or by topical application of mustard oil (32, 33). Nevertheless, we do not yet have any direct evidence to suggest that locally released nociceptin could also modulate the release of CGRP and SP from trigeminovascular neurons.
Stimulation of the trigeminal nerve results in neurogenic dural vasodilatation (NDV). In an animal model intravenous nociceptin dose-dependently suppressed NDV, while it had no effect on baseline vessel diameter (34). In the same experiment, pretreatment with [Nphe1]NC-(1–13)-NH2, a selective OP-4 receptor antagonist, dose-dependently reversed the effect of nociceptin. Nociceptin immunoreactivity has not been observed in basilar and middle cerebral arteries and nociceptin does not modify the contractile properties of human cortical arteries (24). Therefore the suppression of NDV by nociceptin is probably mediated by OP-4 receptors on trigeminal neurons and nociceptin may not have a direct effect on dural vessels.
Opioid peptides inhibit neurotransmitter release from the capsaicin sensitive primary afferent nerve terminals (35). This effect is thought to mediate part of their analgesic action.
Opioids also inhibit neurogenic dural vasodilation via an action on mu-opioid receptors located on trigeminal sensory fibres innervating dural blood vessels (36). As nociceptin does not activate classical opioid receptors, and opioid peptides do not elicit functional responses at OP-4 receptors (37), the role of nociceptin in migraine might be quite different from that of classical opioids. This notion is also underscored by the conflicting results of endogenous opioid peptide plasma levels found in the headache-free state of migraine (38), the only consistent result being a significant reduction in peripheral blood mononuclear cell beta-endorphin concentrations (39).
Nociceptin may also be involved in trigeminal sensory processing, as intracerebroventricular or microiontophoric nociceptin administration inhibits NMDA-evoked responses in the trigeminal nucleus caudalis (40). Nociceptin might also change the activity of rostral brainstem structures involved in pain modulation: the microinjection of nociceptin into the ipsilateral PAG facilitates C-fibre evoked response and postdischarge of wide dynamic range (WDR) neurons in the spinal dorsal horn (41) and nociceptin microinjection to the PAG blocks analgesia caused by morphine or kainate administration (42). It is not clear, however, whether the modulation, by nociceptin, of these rostral brainstem areas could be relevant for the pathogenesis of migraine.
Opioids are a therapeutic option in the abortive treatment of migraine (43), although their use is hampered by their side-effects. In this respect, drugs targeting the OP-4 receptor might be a promising alternative, as nociceptin may not have an addictive potential and could reduce the tolerance to morphine (44).
Based on its inhibitory effect on NDV, nociceptin has been implicated in the pathophysiology of migraine (24, 35). The present finding of lower circulating nociceptin levels in migraine without aura seems to underscore this hypothesis. Lower nociceptin levels could, by lack of inhibition of the trigeminovascular system, predispose for the generation of migraine pain. This possibility is underscored by the finding that higher nociceptin levels were associated with a lower attack frequency in this group of migraineurs. To date, the amount of evidence in this field is too limited to arrive to firm conclusions. Animal studies about the effect of nociceptin on trigeminovascular activation and pain transmission in the trigeminal pathways as well as clinical studies of nociceptin levels during migraine attacks and in experimental pain could help us better understand the importance of the present finding.