Different processing of meningeal
and cutaneous pain information in
the spinal trigeminal nucleus caudalis
Ce´line Melin
1,2, Florian Jacquot
1, Nicolas Vitello
1,
Radhouane Dallel
1,2and Alain Artola
1Abstract
Introduction: Within superficial trigeminal nucleus caudalis (Sp5C) (laminae I/II), meningeal primary afferents project exclusively to lamina I, whereas nociceptive cutaneous ones distribute in both lamina I and outer lamina II. Whether such a relative absence of meningeal inputs to lamina II represents a fundamental difference from cutaneous pathways in the central processing of sensory information is still unknown.
Methods: We recorded extracellular field potentials in the superficial Sp5C of anesthetised rats evoked by electrically stimulating the dura mater, to selectively assess the synaptic transmission between meningeal primary afferents and second-order Sp5C neurons, the first synapse in trigeminovascular pathways. We tested the effect of systemic morphine and local glycinergic and GABAAergic disinhibition.
Results: Meningeal stimulation evokes two negative field potentials in superficial Sp5C. The conduction velocities of the activated primary afferents are within the Ad- and C-fibre ranges. Systemic morphine specifically suppresses meningeal C-fibre-evoked field potentials, and this effect is reversed by systemic naloxone. Segmental glycinergic or GABAAergic
disinhibition strongly potentiates meningeal C-fibre-evoked field potentials but not Ad-fibre ones. Interestingly, the same segmental disinhibition conversely potentiates cutaneous Ad-fibre-evoked field potentials and suppresses C-fibre ones. Conclusion: These findings reveal that the different anatomical organization of meningeal and cutaneous inputs into superficial Sp5C is associated with a different central processing of meningeal and cutaneous pain information within Sp5C. Moreover, they suggest that the potentiation upon local disinhibition of the first synapse in trigeminovascular pathways may contribute to the generation of headache pain.
Keywords
Migraine, synaptic inhibition, glycine, GABA, morphine, rat
Date received: 29 April 2016; revised: 23 June 2016; 13 July 2016; accepted: 19 July 2016
Introduction
During the past few decades, great progress has been made in understanding the pathophysiological mechan-isms of migraine. Nevertheless, important questions remain including the mechanisms of headache pain. It is generally recognised that the development of migraine headache depends on the activation of nocicep-tive afferent fibres of the ophthalmic division of the tri-geminal nerve, which convey pain information from intracranial structures, such as the dura mater and large vessels, to the trigeminal nucleus pars caudalis (Sp5C) and the two uppermost divisions of the cervical spinal cord (C1 and C2), referred to as the trigeminocer-vical complex (TCC). However, the mechanisms of
pain generation after activation of the trigeminovascular system remain incompletely understood (see review, (1–4)).
1
Clermont University, University of Auvergne, Clermont-Ferrand, France 2
Clermont-Ferrand University Hospital, Department of Odontology, Clermont-Ferrand, France
Corresponding author:
Alain Artola, Inserm/UdA, U1107, Neuro-Dol, Trigeminal Pain and Migraine Faculty of Dentistry, 2 rue de Braga, F-63100 Clermont-Ferrand, France.
Email: alain.artola@udamail.fr
Cephalalgia
2017, Vol. 37(12) 1189–1201
!International Headache Society 2016
Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0333102416673204 journals.sagepub.com/home/cep
It is possible that the specificity of the central pro-cessing of meningeal pain information contributes to the generation of migraine headache. It is now well recognised that the laminar distribution of meningeal primary afferents within superficial TCC is different from that of cutaneous ones. Thus, Strassman et al. (5) showed that dural stimulation-induced Fos label-ling, an anatomical marker of neuronal activation, within superficial TCC is primarily restricted to lamina I while facial stimulation typically induces label-ling in both lamina I and outer lamina II (IIo). Of note,
such restricted central projections to lamina I, with almost no projection into lamina IIo, is actually similar
to that of the primary afferents from most visceral organs, muscles and joints into superficial spinal dorsal horn (SDH) (5). Given such difference in the pattern of meningeal and cutaneous inputs into super-ficial TCC, the question arises as to whether this repre-sents a fundamental difference in the central processing of meningeal and cutaneous pain information.
Unit recording studies in cats (6–10) and rats (11– 21), have already provided some information on the properties of the responses of TCC neurons to dura mater stimulation. However, these dural-responsive neurons were found almost entirely in deep laminae (III-V) of TCC and not in lamina I, where meningeal primary afferents terminate, suggesting that the responses of these TCC neurons to dural stimulation were polysynaptic. Therefore, the present study exam-ined extracellular field potentials in superficial (laminae I-II) Sp5C, evoked by meningeal stimulation in vivo, in anaesthetised rats. Field potentials were used with respect to their capability to specifically assess the first synapse in meningeal pain pathways, between pri-mary afferents and Sp5C second-order neurons. Indeed, field potentials within SDH appear to reflect, under control conditions, the population monosynaptic EPSPs of nearby neurons (22–24). We characterised the field potentials evoked within superficial Sp5C by electrical stimulation of the dura mater and examined the effect of (i) systemic morphine and (ii) locally-applied glycinergic or GABAAergic receptor
antagon-ists, on these field potential responses.
Material and methods
Animals
Altogether, 25 adult male Sprague–Dawley rats (250– 300 g) were used in the present experiments. They were obtained from Charles River (L’Arbresle, France). Rats were housed in plastic cages (three to four rats per cage) with soft bedding and free access to food and water. They were maintained in climate (23 1C) and
light-controlled (12:12 h dark:light cycle) protected units
(Iffa-Credo) for at least one week before experiments. All efforts were made to minimise the number of ani-mals used. Experiments followed the ethical guidelines of the International Association for the Study of Pain (25) of the Directive 2010/63/UE of the European Parliament and of the Council on the protection of animals used for scientific purpose. Protocols applied in this study were approved by the local animal experi-mentation committee: CEMEAA ‘‘Comite´ d’Ethique en Matie`re d’Expe´rimentation Animale Auvergne’ (no. CE 28–12).
Animal preparation
Animals were prepared as previously described (26,27). Briefly, animals were anesthetised with 2% halothane (66% NO2/33% O2). After intramuscular injection of
100 mg atropine sulfate, the trachea was cannulated and the carotid artery and external jugular vein cathe-terised. Animals were then paralysed by an intravenous perfusion of vecuronium bromide (2.4 mg.h1) and artificially ventilated with a volume-controlled pump (54–55 strokes.min1). Levels of halothane, O2, N2O,
and end-tidal CO2 (3.5–4.5%) were measured by an
anaesthetic gas analyser (Vamos, Dra¨ger, Lu¨beck, Germany) during the entire experimental period. These parameters were digitally displayed and under the con-trol of alarms. The arterial catheter was attached to a calibrated pressure transducer (UFI, Morro Bay, CA, USA) connected to an amplifier (Stoelting, Wood Dale, IL, USA) for continuous monitoring of the mean arterial blood pressure (90–110 mm Hg). The ana-logue output from the blood pressure amplifier was con-nected to a computer data sampling system (Cambridge Electronics Design 1401 computer interface; Cambridge, UK). Heart rate was continuously monitored and cuta-neous vascularisation periodically checked by observing the color of the paw extremities and the rapidity with which they regained normal color after pressure applica-tion. Colorectal temperature was kept constant at 38 0.5C by means of a feedback-controlled heating
blanket. Surgical level of anesthesia was verified by a stable mean arterial blood pressure and a constant heart rate during noxious stimulation.
The animals were placed in a stereotaxic frame with the head fixed in a ventroflexed position (incisor bar dropped 5 mm under the standard position) by means of an adapted metallic bar. To expose the Sp5C, the overlying musculature, atlanto-occipital membrane and dura mater were removed. Large portions of the par-ietal bones were removed on one side to expose the dura overlying the medial ipsilateral transverse sinus and allow the introduction of the bipolar stimulating electrode. The dura mater overlying the surface of the brain was covered with warm saline. After surgery,
halothane was reduced to 0.6 – 0.7% and maintained at this level during recordings. All exposed nervous tissues were maintained wet with warm artificial cerebrospinal fluid (ACSF; pH 7.4) containing (in mM): 125 NaCl, 26 NaHCO3, 2.5 KCl, 1.25 NaH2PO4, 2 CaCl2, 2 MgSO4,
and 25 glucose).
Stimulation and electrophysiological recordings
Two pairs of electrodes, one of silver electrodes with a convex contact surface (1 mm diameter) and another one of steel needle electrodes, were placed on the dura mater and within the supra orbital skin, respect-ively. Dura mater and facial skin were alternatively stimulated by 1 ms rectangular pulses at a frequency of 0.03 Hz. Dural and facial stimulation-evoked extra-cellular field potentials were recorded using a tungsten microelectrode (0.5 M, World Precision Instruments, Sarasota, Florida). The electrode was lowered into the brain stem within a region from 0.5 mm rostral to 2.3 mm caudal to the obex. Recording sites within Sp5C were adjusted to yield maximal meningeal C-fibre-evoked field potentials while the recording elec-trode was slowly moved through the trigeminal nucleus. Potentials were amplified and filtered at two band-widths to concomitantly record field potentials (10– 1000 Hz) and action potentials (0.3–3 kHz) of nearby neurons at the same site. Data were analysed using a CED 1401 interface coupled to a Pentium computer with Spike 2 software (Cambridge Electronic Design, Cambridge, UK).
Stimulus intensity was never set above 8 mA to avoid directly stimulating the underlying neocortex. Nevertheless, in most rats, the response curve to elec-trical stimuli of increasing intensity had already plat-eaued such that greater stimuli would have not produced larger responses. When stimulus intensity was high enough to recruit both meningeal A- and C-fibres, test stimuli evoked two negative field poten-tials. Extracellular field potentials evoked within SDH by afferent stimulation have been shown to reflect, under physiological conditions, population monosynaptic EPSPs (22–24). Therefore, conduction velocities of the stimulated meningeal afferent fibres could be easily estimated on the basis of response laten-cies, assuming a distance between the stimulation and recording sites of 25–27 mm (11,13) and a synaptic delay of 1 ms. The relative contribution of the Sp5C population postsynaptic potentials and afferent volley to meningeal A fibre-evoked potentials was tested in four rats by examining the effects of a microinjection with 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) (1 mM), an antagonist of AMPA/kainate glutamate receptors, on meningeal stimulation-evoked field poten-tials. CNQX microinjection resulted in a dramatic
decrease in the amplitudes of the two meningeal stimu-lation-evoked field potentials, including the meningeal A fibre-evoked one, indicating that such early potential largely reflects a population monosynaptic EPSPs of nearby neurons rather than afferent volley.
Drug preparation and delivery
The following test substances were used: strychnine, bicuculline, CNQX, morphine chlorhydrate and nalox-one chlorhydrate. All test substances, except for mor-phine (CMD Lavoisier, Paris, France) and naloxone (NarcanÕ, Bristol-Myers-Squibb, Italia), were purchased
from Sigma-Aldrich (Saint-Louis, USA). Stock solutions were obtained by dissolving drug powder in distilled water (strychnine, bicuculline) or DMSO (CNQX). They were diluted in ACSF immediately before delivery at final concentrations (strychnine: 4 mM; bicuculline: 1 mM; CNQX: 1 mM).
Strychnine, bicuculline and CNQX were micro-injected into the Sp5C using 3-barrel glass micropip-ettes fixed on a micromanipulator and connected to Hamilton syringes (0.5 ml) by means of polyethylene tubing (28). Microinjections (0.25 ml) were performed manually over a period of 2 min and monitored by observing the movement of an air bubble in the tubing. The injection rate was slow to reduce tissue damage as much as possible. Micropipettes were pos-itioned within the Sp5C, 1 mm below the pial surface, as close as possible to the recording electrodes, 1 h before injection. The medulla was then covered with 2% Ringer-agar gel. A single rat was microinjected with only one pharmacological agent.
Morphine (3 mg.kg1) and naloxone (0.4 mg.kg1) were both intravenously (i.v.) applied (29).
Histology
At the end of the experiments, electrolytic lesions were performed at the recording sites by passing a direct anodal current (20 mA for 20 s) through the tip of the recording electrode. Rats were killed by an overdose of halothane, brains removed and put into fixative (for-maldehyde: 25%). Ninety mm-thick transverse sections were cut from brainstems using a freezing microtome and observed under a microscope.
Data analysis and statistics
Samples sizes were based on previous experience (26–30), such numbers reflecting a balance between com-monly used sample sizes in the field and a desire to reduce the use of animals in pain experiments. The amp-litudes of meningeal A- and C-fibre-evoked field potentials and cutaneous Ab- and Ad-fibre-evoked
ones were measured at peak. Cutaneous C-fibre field potentials were assessed as the area between the recorded wave-form signal and a straight line tangen-tial to it, before and after the field potentangen-tial (30). The effect of strychnine or bicuculline microinjection was assessed at a delay of 10–12.5 min after micro-injection, corresponding to the maximum effect of local disinhibition (Figure 3c). Data were analysed using a two-way ANOVA for repeated measures fol-lowed by a Bonferroni post hoc test. Means are given s.e.m.
Results
Meningeal stimulation-evoked field potentials
in superficial Sp5C
Field potentials to electrical stimulation of the dura mater were recorded in the ventro-lateral part of super-ficial Sp5C, corresponding to the region where primary afferents of the ophthalmic division of the trigeminal nerve terminate (Figure 1a). At low intensity, electrical stimulation of the meninges evoked a single negative field potential (latencies to onset and to peak:
Amplitude (mV) 1.0 0.5 0.0 Aδ fibers C fibers Intensity (mA) 0 2 4 6 8 100 50 0 Nor m aliz ed amplitude (%) Aδ fibers C fibers Intensity (mA) 0 2 4 6 8 (d) (e) 0.2 mV 25 ms C Aδ 0.5 mm 0.03 mV 0.2 mV 50 ms 1 mA 5 mA (a) (b) (c)
Figure 1. Field potentials evoked in superficial Sp5C by electrical stimulation of the dura mater: (a) Light micrograph of the Sp5C (frontal plane). Lamina II appears as a clear band. The electrolytic lesion corresponding to the recording site (arrow) is located in superficial Sp5C, right in the projection area of V1 primary afferents. Scale bar ¼ 0.5 mm; (b) superimposed field potentials evoked in superficial Sp5C by electrical stimulation of the dura mater at increasing intensity (1, 2, 5 and 8 mA x 1 ms). Each trace is the average of 10 successive sweeps (0.05 Hz). On increasing stimulation strength, the amplitude of the early, Ad-fibre-evoked field potential pro-gressively increases while a second, late C-fibre-evoked field potential appears; (c) simultaneously recorded action potentials (upper recordings) and field potentials (lower recordings) to the same 1 and 5 mA stimulations of the dura mater in another animal than (b). Note that, at the delay of the meningeal C-fibre-evoked field potential, the frequency of action potentials evoked in neighboring second-order neurons increases concomitantly with the amplitude of the field potential; (d, e) Average peak amplitudes (d: absolute values; e: normalised to the amplitude of the response to 8 mA stimulation, as described under Material and Methods) of Ad-fibre- and C-fibre-evoked field potentials are plotted against stimulation strength. Note that the amplitudes of both Ad- and C-fibre-evoked field potentials increase steeply with stimulation intensity and then plateau. The amplitudes of Ad-fibre-evoked field potentials to stimu-lation intensities 3mA and of C-fibre-evoked field potentials to stimustimu-lation intensities 4mA are not different from the corres-ponding maximal ones (to 8 mA stimulation; two-way ANOVA for repeated measures followed by Bonferroni post-hoc test). In addition, the threshold for C-fibre-evoked field potentials is higher than that for Ad-fibre-evoked ones.
4.4 0.2 ms and 7.0 0.1 ms, respectively; n ¼ 7). On increasing stimulus intensity, a second negative, longer latency field potential was observed (latencies to onset and to peak: 18.8 0.8 ms and 33.3 1.3 ms, respectively; n ¼ 7) (Figure 1b).
Meningeal stimulation-evoked field potentials and action potentials of neighbouring neurons within superfi-cial Sp5C were recorded with the same electrode but using different filter settings. In three recordings, action poten-tials were large enough to be dissociated from noise (Figure 1c; see also Figure 3b). These recordings exhibited a clear increase in discharge frequency at the delay of the second negative, longer latency field potential.
The stimulation-response curves (Figures 1d, e) for the two meningeal stimulation-evoked field potentials were very steep once the respective thresholds were exceeded, but then plateaued such that progressively greater stimuli did not produce greater responses (max-imum amplitudes of the first and second field potentials at 8 mA: 0.8 0.1 mV and 0.4 0.1 mV, respectively; Figure 1d). The thresholds for eliciting the first and second field potentials were 0.4 0.1 mA and 1.1 0.2 mA, respectively (Figure 1d).
The conduction velocities of stimulated meningeal afferent fibres were estimated to be 4.0 0.1 and 0.8 0.0 m.s1for the first and second field potentials, respectively (n ¼ 7; see Material and methods); corres-ponding to those expected for Ad- and C-fibre primary afferents. Therefore, our results suggest that these first and second field potentials reflect the monosynaptic excitatory response of nearby neurons in superficial
Sp5C to activation of meningeal Ad- and C-fibre primary afferents, respectively. These field potentials will thus be referred to as meningeal Ad- and C-fibre-evoked field potentials in the remainder of the paper.
Systemic morphine selectively abolishes meningeal
C-fibre-evoked field potentials
Systemic administration of the m opioid receptor agon-ist, morphine, has been shown to reduce the neuronal as well as field potential responses to cutaneous, specif-ically C-fibre, activation within both the SDH (31–33) and Sp5C (29,30). As the second meningeal stimula-tion-evoked field potential has a high threshold and long delay, it is likely evoked by meningeal C-fibre acti-vation. This raises the question as to whether this late, meningeal likely C-fibre-evoked field potential within superficial Sp5C is also sensitive to systemic morphine. Morphine (i.v.; 3 mg.kg1) quickly reduced, if not completely suppressed (see Figure 2a), meningeal C-fibre-evoked field potentials in all animals (10 min after morphine: 14.6 11.1% of baseline; two-way ANOVA for repeated measures followed by Bonferroni post-hoc test: p < 0.001; n ¼ 4; Figure 2b). This effect was specific for meningeal C-fibre-evoked field potentials, as Ad-fibre ones were not affected (10 min after injection: 105.4 6.9% of baseline; two-way ANOVA for repeated measures followed by Bonferroni post-hoc test: p > 0.05; n ¼ 4; Figure 2b). Depression of the meningeal C-fibre-evoked field potential was quickly reversed by systemic injection of
Morphine *** *** *** *** Nor m aliz ed amplitude (%) Naloxone Aδ fibers (4) C fibers (4) 0 10 20 30 Time (min) 150 100 50 0 Aδ C Baseline 25 ms 0.1 mV Morphine Naloxone (a) (b)
Figure 2. Systemic morphine selectively depresses meningeal C-fibre-evoked field potentials in superficial Sp5C: (a) Superimposed field potentials evoked in superficial Sp5C by electrical stimulation of the dura mater before (baseline), 10 min after i.v. injection of
morphine (3 mg.kg-1morphine) and 10 min after i.v. injection of naloxone (0.4 mg.kg-1naloxone). Each trace is the average of five
successive sweeps (0.03 Hz); (b) average time course of the effect of systemic (i.v.) morphine and naloxone (n ¼ 4) on the amplitude of meningeal Ad-fibre- and C-fibre-evoked field potentials. Each point is the average amplitude of five successive responses (0.03 Hz). Systemic morphine leads to a fast and complete suppression of meningeal C-fibre-evoked field potentials that is reversed upon systemic naloxone, the Ad-fibre-evoked ones remaining unaffected (two-way ANOVA for repeated measures followed by Bonferroni post-hoc test: ***p < 0.001; p > 0.05; n ¼ 4).
the opioid receptor antagonist, naloxone (10 min after injection: 106.5 24.9% of baseline; two-way ANOVA for repeated measures followed by Bonferroni post-hoc test: p > 0.05; n ¼ 4; Figure 2b). These results thus indi-cate that only meningeal C-fibre-evoked field potentials are sensitive to systemic morphine.
Segmental glycinergic and GABA
Aergic
disinhibition selectively potentiates meningeal
C fibre-evoked field potentials
The effect of selectively blocking segmental glycinergic inhibition was examined in nine animals (Figure 3). Figures 3a, b, show an example of the effect of a microinjection of strychnine (4 mM) into Sp5C on men-ingeal stimulation-evoked field potentials. Strychnine microinjection strongly potentiated meningeal C-fibre-evoked field potentials in all animals. This potenti-ation peaked at 10–12.5 min after microinjection (426.7 81.3% of baseline; two-way ANOVA for repeated measures followed by Bonferroni post-hoc test: p < 0.01; n ¼ 9) and then progressively reversed (Figure 3c). These changes were selective for meningeal C-fibre-evoked field potentials as Ad-fibre ones were not changed (10–12.5 min after microinjection: 97.2 5.6% of baseline; two-way ANOVA for repeated meas-ures followed by Bonferroni post-hoc test: p > 0.05; n ¼ 9) (Figures 3a–c).
The effect of selectively blocking segmental GABAAergic inhibition on the field potential response
to meningeal stimulation was also examined in eight
other animals (Figure 4). Of these, six exhibited a (measurable) meningeal C-fibre-evoked field potential. Bicuculline microinjection (1 mM) led to a strong, reversible increase in meningeal C-fibre-evoked field potentials (see Figure 4a; on average, 10–12.5 min after injection: 230.5 20.6% of baseline; two-way ANOVA followed by Bonferroni post-hoc test: p <0.001; n ¼ 6; Figure 4c). These changes were again selective for meningeal C-fibre field potentials as Ad-fibre ones were not affected (10–12.5 min after injec-tion: 96.3 2.9% of baseline; two-way ANOVA for repeated measures followed by Bonferroni post-hoc test: p > 0.05; n ¼ 6; Figures 4b, c). Interestingly, in the two other animals, in which only an early meningeal Ad-fibre-evoked field potential could be recorded in control conditions, bicuculline microinjection led to the appearance of a second, late meningeal stimulation-evoked field potential, at the delay of C-fibre-stimulation-evoked ones in other animals (see example in Figures 4d, e).
Segmental disinhibition concomitantly
potentiates meningeal C-fibre-evoked field
potentials and depresses cutaneous C-fibre-evoked
field potentials
Together, these results indicate that local disinhibition, whether it is glycinergic or GABAAergic, consistently
potentiates meningeal C-fibre-evoked field potentials, whereas Ad-fibre ones are not changed. Such changes are thus different from disinhibition-induced ones in
Baseline 0.1 mV 25 ms Aδ C Strychnine (4mM) (a) 0.04 mV 0.4 mV 25 ms
Baseline Strychnine (4mM) Aδ fibers (9)
C fibers (9) 0 10 20 30 40 Time (min) Normalized amplitude (%) *** *** *** * *** Strychnine (4mM) 600 500 400 300 200 100 (b) (c)
Figure 3. Microinjection of strychnine into the Sp5C selectively potentiates meningeal C-fibre-evoked field potentials: (a) Super-imposed field potentials evoked in superficial Sp5C by electrical stimulation of the dura mater before (baseline) and 10–12 min after intracerebral microinjection of strychnine (4 mM; strychnine). Each trace is the average of five successive sweeps (0.03 Hz). Note that glycinergic disinhibition within the Sp5C strongly potentiates the late C-fibre, but not the early Ad-fibre, evoked field potential; (b) simultaneously recorded action potentials (upper recordings) and field potentials (lower recordings) before (baseline) and 10–12 min after the intracerebral microinjection of strychnine (4 mM) in another animal than (a). Note that the frequency of action potentials evoked in neighboring second-order neurons increases concomitantly with the potentiated meningeal C-fibre-evoked field potential; (c) average time course of the effect of an intracerebral microinjection of strychnine (4 mM; n ¼ 9) on the amplitudes of Ad-fibre and C-fibre-evoked field potentials. Each point is the average amplitude of five successive responses (0.03 Hz). (two-way ANOVA for repeated measures followed by Bonferroni post-hoc test: ***p < 0.001; p > 0.05; n ¼ 9).
cutaneous stimulation-evoked field potentials within Sp5C: potentiation of Ad-fibre-evoked ones and depression, if not suppression, of C-fibre-evoked ones (30). However, in this previous work, glycine and GABAA receptor antagonists were intracisternally
applied, whereas here they were microinjected. Might such variations in experimental conditions account for the opposite effects of segmental disinhibition on
meningeal and cutaneous stimulation-induced field potentials? To test this possibility, we compared the effect of strychnine and bicuculline microinjections on concomitantly recorded cutaneous and meningeal sti-mulation-evoked field potentials with the same record-ing electrode. Because, in this series of experiments, recording sites were adjusted to yield maximal menin-geal C-fibre-evoked field potentials while the recording
(a) 0.1 mV 30 ms Aδ 0.1 mV 25 ms Aδ C Baseline 0.1 mV 60 ms Bicuculline (1mM) Aδ fibers (6) (6) (6) (6) C fibers 300 200 100 0 Baseline Normalized amplitude (%) *** Bicuculline (1mM) Aδ C 0.1 mV 30 ms Baseline Bicuculline (1mM) (b) (c) (d) (e)
Figure 4. Microinjection of bicuculline into the Sp5C selectively potentiates meningeal C-fibre-evoked field potentials: (a, b) Superimposed field responses ((b) at extended timescale to clearly show meningeal Ad-fibre-evoked field potentials) recorded within the superficial Sp5C to electrical stimulation of the dura mater before (baseline) and 10–12 min after intracerebral injection of
bicuculline (1 mM). Each trace is the average of five successive sweeps (0.03 Hz). Note that GABAAergic disinhibition within the Sp5C
potentiates the meningeal C-fibre, but not the Ad-fibre, evoked field potential; (c) bar histograms of the effect of intracerebral microinjection of bicuculline (1 mM) on the amplitudes of Ad- and C-fibre-evoked field potentials, 10–12 min after bicuculline microinjection, in the six rats in which baseline meningeal C-fibre-evoked field potentials could be recorded (two-way ANOVA followed by Bonferroni post-hoc test: ***p < 0.001; p > 0.05; n ¼ 6); (d, e) superimposed field potentials evoked in superficial Sp5C by electrical stimulation of the meninges before (baseline) and 10–12 min after intracerebral microinjection of bicuculline (1 mM). Each
trace is the average of five successive sweeps (0.03 Hz). Note that, in this rat, a C-fibre-evoked potential appears under GABAAergic
disinhibition. In (e), the field responses to dura mater stimulation are superimposed at extended time and amplitude scales at the delay of the C-fibre-evoked field potential.
electrode was slowly moved through the Sp5C, a con-sistent cutaneous C-fibre-evoked field potential could be concomitantly recorded in only three out of the nine strychnine-treated animals and one of the eight bicucul-line-treated rats (see above). Strychnine microinjection, which produced a selective potentiation of meningeal C-fibre-evoked field potentials, led to enhanced cutane-ous Ad-fibre-evoked field potentials and decreased, and even suppressed, C-fibre-evoked ones in all three animals (Figure 5). A similar result was obtained in the bicucul-line-treated rat (not shown). This indicates that the opposite effects of segmental disinhibition on meningeal and cutaneous stimulation-induced field potentials are
not due to variations in experimental conditions, but represent a clear difference in the central processing of cutaneous and meningeal sensory information within Sp5C.
Discussion
We show that electrical stimulation of the dura mater evokes two negative field potentials in superficial Sp5C. The conduction velocities of the activated primary afferents are within the Ad- and C-fibre ranges. Systemic morphine suppresses specifically meningeal C-fibre-evoked field potentials and this is reversed by
Aδ fibers C fibers Normalized amplitude (%) 600 400 200 0 600 400 200 0
Before After Before After
Cutaneous Meningeous Animal 1 Animal 2 Animal 3 (d) (a) Aδ C 0.1 mV 40 ms 0.2 mV 40 ms Aδ Aβ C 0.2 mV 40 ms Baseline Strychnine (4mM) (b) (c)
Figure 5. Microinjection of strychnine into the Sp5C produces opposite changes in meningeal and cutaneous Ad- and C-fibre-evoked field potentials: (a, b, c) Superimposed field responses recorded in the superficial Sp5C to electrical stimulation of the dura mater (a) and face skin (above the eye; b, c), before (baseline) and 10–12 min after intracerebral microinjection of strychnine (4 mM) in a single rat. In (b), arrows point to Ab-, Ad- and C-fibre-evoked field potential responses: Note that at this timescale, the Ab-fibre-evoked field potential response is very close to the stimulus artefact. In (c), the field responses to stimulation of the facial skin are superimposed at extended time and amplitude scales at the delay of cutaneous C-fibre-evoked field potentials. Note that, under glycinergic disinhibition, the cutaneous C-fibre-evoked field potential is completely suppressed while the meningeal one is potentiated. Each trace is the average of five successive sweeps (0.03 Hz); (d) amplitudes of concomitantly recorded Ad- (left) and C-fibre-evoked field potential responses (right) to electrical stimulation of the dura mater (stippled lines) and facial skin (full lines) in the same three rats (as indicated in the Figure). Upon glycinergic disinhibition, cutaneous Ad-fibre-evoked field potentials are potentiated whereas meningeal ones do not change, and cutaneous C-fibre-evoked field potentials are inhibited, if not abolished, whereas meningeal ones are potentiated.
systemic naloxone. Segmental glycinergic or GABAAergic disinhibition strongly potentiates
menin-geal C-fibre-evoked field potentials, Ad-fibre ones remaining unaffected. Interestingly, the same segmental disinhibition potentiates cutaneous Ad-fibre-evoked field potentials and suppresses C-fibre ones.
Meningeal A- and C-fibre-evoked field potentials
There are very few reports of meningeal stimulation-evoked trigeminal field potentials, and they were all recorded in cats (10,34). We show for the first time that meningeal electrical stimulation evokes two nega-tive field potentials within the rat Sp5C: early (delay to peak &7 ms) and late (delay to peak &33 ms). Based on the computed conduction velocities, we concluded that the early and late trigeminal field potentials result from the monosynaptic activation of second-order Sp5C neurons by meningeal Ad- and C-fibres. Consistently, meningeal stimulation also evokes a biphasic responses in Sp5C neurons: an early, Ad-fibre-evoked phase (11– 14 ms latency) and a late C-fibre-evoked one (34–45 ms latency) (13,14,15). Of note, there are only few reports of meningeal C-fibre-evoked neuronal responses (11,14–16) as most studies examined only Ad-fibre-evoked ones (12,13,17–21,35).
In agreement with our previous findings (30), cuta-neous electrical stimulation evoked three negative field potentials within the rat Sp5C elicited by stimulation of cutaneous Ab-, Ad- and C-fibres. Comparison between meningeal (present report) and cutaneous (30) stimula-tion-evoked field potentials within the rat Sp5C reveals that the meningeal and cutaneous Ad- and C-fibres dis-play rather similar conduction velocities (Ad-fibres: 4.1 and 2.8 m.s1, respectively; C-fibres: 0.9 and 0.6 m.s1, respectively) and activation thresholds (Ad-fibres: 0.4 and 0.2 mA, respectively; C-fibres: 1.1 and 1.9 mA, respectively). Meningeal fast-A afferent neurons with conduction velocities >5 m.s1 have been reported in the trigeminal ganglion (36,37) and myelinated axons classified as Ab in nerves supplying the dura (38). Given the short latency to onset of our early field potential, it likely includes responses to such fast A-fibres. Nevertheless, a meningeal fast A-evoked field potential could not be isolated.
Systemic morphine selectively abolishes meningeal
C-fibre-evoked field potential
In rats, systemic morphine at 3 mg.kg1 completely suppressed meningeal C-fibre-evoked field potential responses, but had no effect on Ad-fibre ones. This is in contrast to previous results in cats, showing that systemic morphine (39) or iontophoretic applied DAMGO, a m opioid receptor agonist (40), attenuates
meningeal Ad-fibre-evoked neuronal firing. However, such inhibition appears to only occur at high con-centrations of locally or systemically applied m opioid receptor agonists: For instance, whereas morphine at 10 mg.kg1 i.v. inhibits 65% of neuronal activity, at 3 mg.kg1, it only produces a hardly significant &33% attenuation of Ad-fibre-evoked responses in cats (see Figure 4, (39)). Moreover, these previous reports (39,40) only examined meningeal Ad-fibre-evoked responses. The present results provide evidence for systemic morphine preferentially inhibiting synaptic transmission between meningeal C-fibres and second-order Sp5C neurons. That morphine had no effect on meningeal Ad-fibre-evoked field potentials is thus likely due to the relatively low dose we used here. Nevertheless, since previous results were obtained in cats and the present ones in rats, we cannot eliminate the possibility that the differences described above are due, at least in part, to species differences.
Systemic morphine has previously been shown to preferentially depress C-fibre-evoked neuronal responses within (i) Sp5C to facial stimulation (29,30) and (ii) in lumbar SDH in response to sciatic nerve stimulation (32). Thus, that C-fibre-evoked responses are more sen-sitive than A-fibre-evoked ones to m opioid receptor agonists appears to be a general characteristic of the somatosensory system. This could be due to segmental mechanisms. Locally-applied morphine also selectively inhibits C-fibre-evoked neuronal responses, within both Sp5C (28) and lumbar SDH (33). Moreover, in slices of adult rat SDH, C-fibre inputs to second-order neurons are more sensitive to m opioid receptor agonists than Ad-fibre ones (41,42). However, given the wide-ranging CNS distribution of m opioid receptors, their activation in supraspinal areas, including the periaque-ductal gray (43), may also account for the selective suppression of meningeal C-fibre-evoked field potentials. Periaqueductal gray projects to the rostral ventral medulla, which in turn sends outputs to Sp5C/SDH, where they control spinal processing of noxious, but not innocuous, inputs (see review (44,45)).
That systemic morphine preferentially inhibits meningeal C-fibre-evoked responses compared with Ad-fibre-evoked ones is also consistent with the conclu-sion that meningeal Ad- and C-fibre inputs to Sp5C can be independently modulated. Similarly, systemic orexin A has been shown to inhibit meningeal Ad-fibre-evoked responses, but not C-fibre-evoked ones.16
Central processing of meningeal information
Meningeal C-fibre-evoked field potentials, but not Ad-fibre-evoked ones, recorded in the Sp5C are potentiated upon glycinergic as well as GABAAergic
as the major fast inhibitory neurotransmitter in adult rat lamina I (46,47). But GABAA receptors have
been shown to also modulate (i) the response of lamina I neurons to mechanical stimulation within the rat SDH (48) and (ii) neurons activated by supramax-imal electrical stimulation of the superior sagittal sinus within the TCC (35), suggesting that both GABAAand
glycine receptors contribute to synaptic inhibition onto lamina I Sp5C neurons. It was surprising that menin-geal C-fibre-evoked field potentials alone were poten-tiated by such disinhibition. Indeed, it is now well recognised that all classes of sensory primary afferents (Ab-, Ad- and C-fibres) can contact inhibitory inter-neurons within SDH (see review (49,50)). Therefore, disruption of feedforward (as feedback) inhibition is expected to facilitate meningeal Ad- as well as C-fibre inputs onto second-order trigeminovascular neurons. Our results rather suggest that the inhibitory controls of the synapses between meningeal C-fibre or Ad-fibre and second order neurons are different. In cutaneous inputs too, synaptic inhibition appears to be fibre-spe-cific: GABAAreceptors were shown to control Ad-, but
neither Ab- nor C-fibre-mediated mechanical responses in lamina I neurons (48) and GABAA and glycine
receptors, to modulate A-d and C-, but not Ab-fibre-evoked field potentials in the Sp5C (30).
Disinhibition-induced changes in meningeal stimula-tion-evoked field potentials are different from those in cutaneous stimulation-evoked ones in the Sp5C: facili-tation of the polysynaptic components of Ab- and Ad-fibre-evoked field potentials but inhibition of C-fibre ones (present results and (30)). Such cross-modal inhibition between cutaneous A- and C-fibre inputs is likely due to the unmasking of polysynaptic A-fibre inputs onto last-order GABAergic neurons, leading to activation of GABAB receptors on C-fibre
terminals and, in turn, inhibition of C-fibre inputs onto superficial Sp5C/SDH neurons (see Figure 8, (30)). Of note, cross-modal interactions can only be assessed by synchronising afferent inputs and thus simultaneously activating functionally-different afferent fibres using electrical stimuli. Nevertheless, the fact that simultan-eous activation of cutansimultan-eous, but not meningeal, A- and C-fibres results in an inhibitory influence of one input on the other indicates a clear difference in the intrinsic Sp5C circuitry of cutaneous and trigemi-novascular inputs.
It is now well recognised that central sensitisation – an enhancement in the function of neurons and circuits in nociceptive pathways in response to nociceptor activity (see review, (51)) – develops over the course of a migraine attack. Most patients experience cephalic cutaneous allodynia (pain in response to normally
innocuous stimuli of the skin), a marker of central sensitisation, during their migraine episode (52–54). In animals, local application of an inflammatory soup to the dura also produces cutaneous allodynia (26, 55– 57) and concomitantly sensitises second-order trigemi-novascular neurons (11,26). There is now clear evidence for disruption of local inhibitory controls underlying mechanical allodynia (see review (49,51)). Cephalic mechanical allodynia can be consistently replicated simply by impairing glycine- or GABAA-mediated
inhibition within the Sp5C (58–60). Moreover, the use of anatomical markers of activation has identified, associated with behavioral mechanical allodynia, poly-synaptic pathways driving up cephalic cutaneous A-fibre low-threshold mechanosensitive afferents to Sp5C lamina I projection neurons (58–60). Once inhibition is lifted, such A-fibre inputs can gain access to the pain transmission circuitry of superficial Sp5C, producing pain. Altogether, this suggests that a barrage of incoming signals from meningeal nociceptors results in Sp5C sensitisation, disrupting local inhibitory controls. According to the present results, Sp5C disin-hibition can in turn further potentiate synaptic trans-mission between meningeal primary afferents and Sp5C second-order neurons. Such a mechanism might likely enhance headache pain after activation of the trigemi-novascular system. At the same time, existing polysy-naptic excitatory circuits become unmasked: cutaneous A-fibre inputs can then drive (i) lamina I projection neurons, producing cephalic mechanical allodynia (58–60), and (ii) last order inhibitory interneurons (30) underlying A-fibre-mediated inhibition of C-fibre inputs to Sp5C. Of note, other mechanisms are likely involved in pain symptoms during migraine attacks. Thus, the development of extracephalic allodynia has been attributed to sensitisation of third-order trigemino-vascular thalamic nociceptive neurons that receive conver-gent input from the cranial meninges and extracephalic skin (57), activation of descending pain-facilitating pro-cesses from the rostral ventromedial medulla (56), or impaired descending inhibition of pain (26).
While the molecular mechanisms of disinhibition and its consequences for cellular excitability have been the object of intense investigations, much less attention has been given to the effects on network function. Nevertheless, such effects have important functional consequences. Thus impaired central inhibition can unmask existing interconnections between separate cuta-neous sensory pathways, underlying the development of pain symptoms during the course of a migraine attack. The present findings indicate that such cross-modal interactions do not exist between meningeal Ad- and C-fibre afferents, and that synaptic transmission between
meningeal C-fibre afferents and second-order trigemino-vascular neurons is potentiated by reduced inhibition. Such selective potentiation might specifically contribute
to the generation of headache pain after activation of the trigeminovascular system, offering a therapeutical target for the management of headache pain.
Article highlights
. Electrical stimulation of the dura mater evokes two negative Ad- and C-fibre-evoked field potentials in the superficial spinal trigeminal nucleus caudalis.
. Systemic morphine specifically suppresses meningeal C-fibre-evoked field potentials.
. Segmental glycinergic or GABAAergic disinhibition strongly potentiates meningeal C-fibre-evoked field
potentials but not Ad-fibre ones.
. Such disinhibition-induced changes in meningeal stimulation-induced field potentials indicate a difference from cutaneous pathways in the intrinsic trigeminal circuitry.
Acknowledgments
We thank Jean-Louis Molat for technical help during electro-physiological experiments and Anne-Marie Gaydier for sec-retarial assistance.
Declaration of conflicting interests
The authors declare no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial sup-port for the research, authorship, and/or publication of this article: Funding from Institut National de la Sante´ et de la Recherche Me´dicale (Inserm), Universite´ Clermont1 (France), and Re´gion Auvergne (France).
References
1. Olesen J, Burstein R, Ashina M, et al. Origin of pain in migraine: Evidence for peripheral sensitisation. Lancet
Neurol2009; 8: 679–690.
2. Akerman S1, Holland PR and Goadsby PJ. Diencephalic and brainstem mechanisms in migraine. Nat Rev Neurosci 2011; 12: 570–584.
3. Pietrobon D and Moskowitz MA. Pathophysiology of migraine. Annu Rev Physiol 2013; 75: 365–391.
4. Burstein R, Noseda R and Borsook D. Migraine: Multiple processes, complex pathophysiology. J Neurosci 2015; 35: 6619–6629.
5. Strassman AM, Mineta Y and Vos BP. Distribution of fos-like immunoreactivity in the medullary and upper cer-vical dorsal horn produced by stimulation of dural blood vessels in the rat. J Neurosci 1994; 14: 3725–3735. 6. Davis KD and Dostrovsky JO. Activation of trigeminal
brainstem nociceptive neurons by dural artery stimulation.
Pain1986; 25: 395–401.
7. Davis KD and Dostrovsky JO. Responses of feline trigeminal spinal tract nucleus neurons to stimulation of the middle meningeal artery and sagittal sinus.
J Neurophysiol1988; 59: 648–666.
8. Davis KD and Dostrovsky JO. Cerebrovascular applica-tion of bradykinin excites central sensory neurons. Brain
Res1988; 446: 401–406.
9. Strassman A, Mason P, Moskowitz M, et al. Response of brainstem trigeminal neurons to electrical stimulation of the dura. Brain Res 1986; 379: 242–250.
10. Lambert GA, Lowy AJ, Boers PM, et al. The spinal cord processing of input from the superior sagittal sinus: Pathway and modulation by ergot alkaloids.
Brain Res1992; 597: 321–330.
11. Burstein R, Yamamura H, Malick A, et al. Chemical stimulation of the intracranial dura induces enhanced responses to facial stimulation in brain stem trigeminal neurons. J Neurophysiol 1998; 79: 964–982.
12. Ellrich J, Schepelmann K, Pawlak M, et al. Acetylsa-licylic acid inhibits meningeal nociception in rat. Pain 1999; 81: 7–14.
13. Schepelmann K, Ebersberger A, Pawlak M, et al. Response properties of trigeminal brain stem neurons with input from dura mater encephali in the rat.
Neuroscience1999; 90: 543–554.
14. Panteleev SS, Sokolov AY, Kartus DE, et al. Responses of neurons in the spinal nucleus of the trigeminal nerve to electrical stimulation of the dura mater of the rat brain.
Neurosci Behav Physiol2005; 35: 555–559.
15. Bolton S, O’Shaughnessy CT and Goadsby PJ. Properties of neurons in the trigeminal nucleus caudalis responding to noxious dural and facial stimulation. Brain Res 2005; 1046: 122–129.
16. Holland PR, Akerman S and Goadsby PJ. Modulation of nociceptive dural input to the trigeminal nucleus caudalis via activation of the orexin 1 receptor in the rat. Eur J
Neurosci2006; 24: 2825–2833.
17. Martin VT, Lee J and Behbehani MM. Sensitization of the trigeminal sensory system during different stages of the rat estrous cycle: implications for menstrual migraine.
Headache2007; 47: 552–563.
18. Charbit AR, Akerman S, Holland PR, et al.
Neurons of the dopaminergic/calcitonin gene-related peptide A11 cell group modulate neuronal firing in the trigeminocervical complex: An electrophysiological
and immunohistochemical study. J Neurosci 2009; 29: 12532–12541.
19. Charbit AR, Akerman S and Goadsby PJ. Trigeminocer-vical complex responses after lesioning dopaminergic A11 nucleus are modified by dopamine and serotonin mech-anisms. Pain 2011; 152: 2365–2376.
20. Sokolov AY, Lyubashina OA and Panteleev SS. Spinal trigeminal neurons demonstrate an increase in responses to dural electrical stimulation in the orofacial formalin test. J Headache Pain 2012; 13: 75–82.
21. Andreou AP, Holland PR, Lasalandra MP, et al. Mod-ulation of nociceptive dural input to the trigeminocervical complex through GluK1 kainate receptors. Pain 2015; 156: 439–450.
22. Baldissera F, Hultborn H and Illert M. Integration in spinal neuronal systems. In: Brooks, VB (Ed), Handbook of Physiology, Section I: The Nervous System. Williams & Wilkins, Baltimore, 1981, pp. 509–595.
23. Schouenborg J. Functional and topographical properties of field potentials evoked in rat dorsal horn by cutaneous C-fibre stimulation. J Physiol 1984; 356: 169–192. 24. Shreckengost J, Calvo J and Quevedo J. Bicuculline
sen-sitive primary afferent depolarization remains after greatly restricting synaptic transmission in the mamma-lian spinal cord. J Neurosci 2010; 30: 5283–5288. 25. Zimmermann M. Ethical guidelines for investigations of
experimental pain in conscious animals. Pain 1983; 16: 109–110.
26. Boyer N, Dallel R, Artola A, et al. General trigeminosp-inal central sensitization and impaired descending pain inhibitory controls contribute to migraine progression.
Pain2014; 155: 1196–1205.
27. Lapirot O, Melin C, Modolo A, et al. Tonic and phasic descending dopaminergic controls of nociceptive trans-mission in the medullary dorsal horn. Pain 2011; 152: 1821–1831.
28. Dallel R, Duale´ C and Molat JL. Morphine administered in the substantia gelatinosa of the spinal trigeminal nucleus caudalis inhibits nociceptive activities in the spinal trigeminal nucleus oralis. J Neurosci 1998; 18: 3529–3536.
29. Dallel R, Luccarini P, Molat JL, et al. Effects of systemic morphine on the activity of convergent neurons of spinal trigeminal nucleus oralis in the rat. Eur J Pharmacol 1996; 314: 19–25.
30. Melin C, Jacquot F, Dallel R, et al. Segmental disinhib-ition suppresses C-fiber inputs to the rat superficial
medullary dorsal horn via the activation of GABAB
receptors. Eur J Neurosci 2013; 37: 417–428.
31. Le Bars D, Rivot JP, Guilbaud G, et al. The depressive effect of morphine on the C fibre response of dorsal horn neurones in the spinal rat pretreated or not by pCPA.
Brain Res1979; 176: 337–353.
32. Tanabe M, Murakami H, Honda M, et al. Gabapentin depresses C-fiber-evoked field potentials in rat spinal dorsal horn only after induction of long-term potenti-ation. Exp. Neurol 2006; 202: 280–286.
33. Buesa I, Urrutia A, Bilbao J, et al. Morphine-induced depression of spinal excitation is not altered following
acute disruption of GABA(A) or GABA(B) receptor activity. Eur J Pain 2008; 12: 677–685.
34. Goadsby PJ and Hoskin KL. Inhibition of trigeminal neurons by intravenous administration of the serotonin (5HT)1B/D receptor agonist zolmitriptan (311C90): Are brain stem sites therapeutic target in migraine? Pain 1996; 67: 355–359.
35. Storer RJ1, Akerman S and Goadsby PJ. GABA recep-tors modulate trigeminovascular nociceptive neurotrans-mission in the trigeminocervical complex. Br J Pharmacol 2001; 134: 896–904.
36. Strassman AM, Raymond SA and Burstein R. Sensitiza-tion of meningeal sensory neurons and the origin of head-aches. Nature 1996; 384: 560–564.
37. Levy D and Strassman AM. Mechanical response properties of A and C primary afferent neurons innervat-ing the rat intracranial dura. J Neurophysiol 2002; 88: 3021–3031.
38. Strassman AM, Weissner W, Williams M, et al. Axon diameters and intradural trajectories of the dural innerv-ation in the rat. J Comp Neurol 2004; 473: 364–376. 39. Williamson DJ, Shepheard SL, Cook DA, et al. Role of
opioid receptors in neurogenic dural vasodilation and sensitization of trigeminal neurones in anaesthetized rats. Br J Pharmacol 2001; 133: 807–814.
40. Storer RJ, Akerman S and Goadsby PJ. Characterization of opioid receptors that modulate nociceptive neurotrans-mission in the trigeminocervical complex. Br J Pharmacol 2003; 138: 317–324.
41. Ikoma M, Kohno T and Baba H. Differential presynaptic effects of opioid agonists on Ad- and C-afferent
glutama-tergic transmission to the spinal dorsal horn.
Anesthesiology2007; 107: 807–812.
42. Heinke B, Gingl E and Sandku¨hler J. Multiple targets of m-opioid receptor-mediated presynaptic inhibition at pri-mary afferent Ad- and C-fibers. J Neurosci 2011; 31: 1313–1322.
43. Yaksh TL, Yeung JC and Rudy TA. Systematic examin-ation in the rat of brain sites sensitive to the direct appli-cation of morphine: Observation of differential effects within the periaqueductal gray. Brain Res 1976; 114: 83–103.
44. Millan MJ. Descending control of pain. Prog Neurobiol 2002; 66: 355–474.
45. Heinricher MM, Tavares I, Leith JL, et al. Descending control of nociception: Specificity, recruitment and plas-ticity. Brain Res Rev 2009; 60: 214–225.
46. Che´ry N and de Koninck Y. Junctional versus extrajunc-tional glycine and GABA(A) receptor-mediated IPSCs in identified lamina I neurons of the adult rat spinal cord.
J Neurosci1999; 19: 7342–7355.
47. Keller AF, Coull JA, Chery N, et al. Region-specific developmental specialization of GABA-glycine cosy-napses in laminas I-II of the rat spinal dorsal horn.
J Neurosci2001; 21: 7871–7880.
48. Seagrove LC, Suzuki R and Dickenson AH. Electrophy-siological characterisations of rat lamina I dorsal horn neurones and the involvement of excitatory amino acid receptors. Pain 2004; 108: 76–87.
49. Zeilhofer HU, Wildner H and Ye´venes GE. Fast synaptic inhibition in spinal sensory processing and pain control.
Physiol Rev2012; 92: 193–235.
50. Braz J, Solorzano C, Wang X, et al. Transmitting pain and itch messages: A contemporary view of the spinal cord circuits that generate gate control. Neuron 2014; 82: 522–536.
51. Latremoliere A and Woolf CJ. Central sensitization: A generator of pain hypersensitivity by central neural plasticity. J Pain 2009; 10: 895–926.
52. Burstein R, Cutrer MF and Yarnitsky D. The develop-ment of cutaneous allodynia during a migraine attack clinical evidence for the sequential recruitment of spinal and supraspinal nociceptive neurons in migraine. Brain 2000; 123: 1703–1709.
53. Lipton RB, Bigal ME, Ashina S, et al. Cutaneous allo-dynia in the migraine population. Ann Neurol 2008; 63: 148–158.
54. Guy N, Marques AR, Orliaguet T, et al. Are there dif-ferences between cephalic and extracephalic cutaneous allodynia in migraine patients? Cephalalgia 2010; 30: 881–886.
55. Oshinsky ML and Gomonchareonsiri S. Episodic dural stimulation in awake rats: A model for recurrent head-ache. Headache 2007; 47: 1026–1036.
56. Edelmayer RM, Vanderah TW, Majuta L, et al. Medullary pain facilitating neurons mediate allodynia in headache-related pain. Ann Neurol 2009; 65: 184–193. 57. Burstein R, Jakubowski M, Garcia-Nicas E, et al.
Thalamic sensitization transforms localized pain into widespread allodynia. Ann Neurol 2010; 68: 81–91. 58. Miraucourt LS, Dallel R and Voisin DL. Glycine inhibitory
dysfunction turns touch into pain through PKCgamma interneurons. PLoS ONE 2007; 2: e1116.
59. Miraucourt LS, Moisset X, Dallel R, et al. Glycine inhibitory dysfunction induces a selectively dynamic, morphine-resistant, and neurokinin 1 receptor-
indepen-dent mechanical allodynia. J Neurosci 2009; 29:
2519–2527.
60. Peirs C, Bourgois N, Artola A, et al. Protein kinase C g interneurons mediate C-fiber-induced orofacial second-ary static mechanical allodynia, but not C-fiber-induced
nociceptive behavior. Anesthesiology 2016; 124:
