OUP user menu

Calcitonin gene-related peptide receptor antagonist olcegepant acts in the spinal trigeminal nucleus

Marie-Luise Sixt, Karl Messlinger, Michael J. M. Fischer
DOI: http://dx.doi.org/10.1093/brain/awp168 3134-3141 First published online: 8 September 2009


Several lines of evidence suggest a major role of calcitonin gene-related peptide (CGRP) in the pathogenesis of migraine and other primary headaches. Inhibition of CGRP receptors by olcegepant and telcagepant has been successfully used to treat acute migraine and to reduce the activity of spinal trigeminal neurons involved in meningeal nociception in rodents. The site of CGRP receptor inhibition is unclear, however. In adult Wistar rats anaesthetized with isofluorane systemic intravenous infusion (0.9 mg/kg) or unilateral facial injection (1 mM in 100 µl) of capsaicin was used to induce activity in the trigeminal nociceptive system. Animals were pre-treated either by saline or olcegepant. In comparison with vehicle infusion or the non-injected side of the face, capsaicin significantly increased the expression of the activation markers Fos in the spinal trigeminal nucleus and phosphorylated extracellular signal-regulated kinase in the trigeminal ganglion. Pre-treatment with olcegepant (900 µg/kg) inhibited the capsaicin-induced expression of Fos throughout the spinal trigeminal nucleus by 57%. In contrast, the expression of phosphorylated extracellular signal-regulated kinase in the trigeminal ganglion was not changed by olcegepant pre-treatment. CGRP receptor inhibition, which has been shown to decrease spinal trigeminal activity, is likely to occur in the central nervous system rather than in the periphery including the trigeminal ganglion. This may be important for future therapeutic interventions with CGRP receptor antagonists in migraine.

  • migraine
  • headache
  • BIBN4096BS
  • MK-0974


Meningeal (dural and pial) blood vessels in rat and man are innervated by a dense network of trigeminal sensory fibres, ∼40% of which contain calcitonin gene-related peptide (CGRP) (Edvinsson et al., 1987a, b; Keller and Marfurt, 1991; Messlinger et al., 1993). Most of these are polymodal nociceptive afferents, activated by a variety of noxious stimuli including capsaicin, the response to which is mediated by the transient receptor potential cation channel, subfamily V, member 1 (TRPV1) (Caterina et al., 1999; Davis et al., 2000). CGRP is released upon stimulation from both the peripheral and central processes of trigeminal afferents (Messlinger et al., 1995; Jenkins et al., 2004).

Electrical stimulation of trigeminal afferents in humans and animals caused release of CGRP, suggesting a major role of CGRP in the trigeminal system (Zagami et al., 1990; Goadsby and Edvinsson, 1993). Infusion of CGRP in migraineurs triggered headaches (Lassen et al., 2002). The CGRP receptor antagonists olcegepant (BIBN4096BS) and telcagepant (MK-0974) were successfully tested in the treatment of acute migraine (Olesen et al., 2004; Ho et al., 2008b). CGRP seems to be involved in the control of neuronal activity in the spinal trigeminal nucleus (STN), which integrates nociceptive afferent inputs from trigeminal tissues including intracranial afferents. Microinjection of olcegepant into the spinal trigeminal nucleus reduced the activity of central trigeminal neurons evoked by glutamate microiontophoresis or by electrical stimulation of the superior sagittal sinus (Storer et al., 2004). Systemic infusion of olcegepant decreased spontaneous and heat-evoked activity in the spinal trigeminal nucleus (Fischer et al., 2005).

However, the essential site of action of the CGRP receptor inhibition to reduce spinal trigeminal activity is unknown. CGRP receptor inhibitors are known to barely penetrate the blood brain barrier (Edvinsson et al., 2007). Compared with the potency in cellular models, high concentrations of CGRP inhibitors were necessary for a therapeutic effect in humans, indicating that the site of action may be central. Using activity markers for primary afferent and secondary neurons, we asked at which site a systemically applied CGRP receptor antagonist can reduce activity evoked by a noxious stimulus in the trigeminal system.

Materials and methods

The experiments were performed in accordance with the ethical guidelines of the International Association for the Study of Pain (Zimmermann, 1983). The protocol for in vivo experiments was reviewed by an ethics committee and approved by the local district government. For all experiments adult male Wistar rats were used.

Experimental procedures

Rats were anaesthetized with isofluorane (2.5%) applied through a mask close to but not touching the forehead. The animals were breathing spontaneously and their body temperature was monitored and held at 37°C with a feedback controlled homoeothermic system (Foehr Medical Instrument, Seeheim, Germany). A catheter was inserted into the femoral vein and then the depth of anaesthesia was decreased by reducing the isofluorane concentration to 1.5%. This dose was maintained throughout the experiment and sufficient to suppress nociceptive reflexes evoked by noxious pinch stimuli of the hind paw. Ointment was used to prevent drying of the eyes and care was taken to avoid any input to trigeminal tissues.

Experimental protocol

The animals were intravenously infused with 0.33 ml saline or 900 µg/kg olcegepant dissolved in 0.33 ml saline within 10 min. Five minutes after this pre-treatment, capsaicin was applied.

In experiments intended for the analysis of spinal trigeminal Fos expression, 1 ml of either 915 µg/kg (3 µmol/kg) capsaicin or vehicle was infused within 30 min. Two hours after start of this infusion the anaesthesia was deepened (5% isofluorane) and the experiment was terminated by transcardial perfusion of the animal with saline followed by 4% paraformaldehyde. In several pilot experiments, blood pressure and heart rate were recorded via a catheter in the femoral artery.

In experiments designed for the analysis of pERK in primary trigeminal neurons, capsaicin was injected subcutaneously unilateral at three sites in the facial area corresponding to the three trigeminal branches. The injection sites were 3 mm above the lateral corner of the eye, midway between the lateral corners of mouth and eye and 3 mm below the lateral corner of the mouth. At each site 33 µl of 1 mM capsaicin diluted in isotonic saline were deposited. The animals were transcardially perfused in deep anaesthesia with saline 7.5 min after the capsaicin injection followed by 4% paraformaldehyde 2.5 min later. In additional experiments, 50 µl lidocaine (2%) was injected unilaterally at the three facial spots mentioned above. Fifteen minutes later capsaicin was injected into the same sites but on both sides of the animal.


Capsaicin was stored at 3 mM in pure ethanol and diluted in isotonic saline. The final ethanol dose in the capsaicin solution and the vehicle was 24 µg/kg. Olcegepant, kindly provided by Boehringer-Ingelheim (Biberach, Germany), was diluted in isotonic saline.


The dissected brainstem samples were processed as published previously (Offenhauser et al., 2005). Briefly, after overnight immersion in 4% paraformaldehyde and 1 day in 15% sucrose the brainstem was frozen in methylbutane at −40°C and stored at −20°C. With a cryostat serial sections of 50 µm were cut in a 6 mm range caudal to the obex. One section per 500 µm was processed for immunohistochemistry.

Trigeminal ganglia were removed, immersed in 4% paraformaldehyde for 1 h, placed in 30% sucrose, frozen in methylbutane and cut into serial sections of 25 µm, so that comparable layers from both ganglia could be obtained. Eight slices evenly spaced throughout the ganglion were processed for pERK as recommended by the supplier.

Dilution of chemicals and antibodies as well as all washes were made with 0.1 molar phosphate buffered saline (pH 7.4). The free floating sections were kept for 30 min in phosphate buffered saline with 0.3% hydrogen peroxide for Fos or in phosphate buffered saline with 10% methanol and 3% hydrogen peroxide for pERK, placed for at least 1 h in blocking solution containing 5% normal goat serum and 0.3% Triton X-100 and after washing incubated overnight with the primary antibody for Fos [polyclonal rabbit anti-c-Fos (Ab-5) (4–17), 1:8000, http://www.merckbiosciences.co.uk/product/PC38, Calbiochem, USA] or pERK [monoclonal rabbit anti-p44/42 MAPK (Thr202/Thr204) (20G11), 1:1000, http://www.cellsignal.com/products/9101.html, Cell Signaling Technology, MA, USA] dissolved in the blocking solution. After washing, incubation for 2 h with the secondary biotinylated antibody (1:600, goat anti-rabbit IgG, Vector Laboratories, USA) followed. Then the rinsed sections were immersed in the avidin-biotin complex (Vectastain Elite ABC Kit, Vector Laboratories) for 1 h (Fos) or 30 min (pERK) and developed in diaminobenzidine tetrahydrochloride peroxidase substrate solution enhanced by cobalt for 7.5 min (Sigma Fast 3,3′-Diaminobenzidine tetrahydrochloride with Metal Enhancer Tablet Set, Sigma, USA). After washing, mounting on slides and air-drying, the slices were dehydrated and coverslipped. The Fos immunoreactive neuronal nuclei within the trigeminal nucleus caudalis were identified by their typical dark staining and counted naïvely to the experimental protocol using bright field microscopy at 200- to 400-fold magnification. Twelve brainstem sections per animal were analysed for Fos. In the trigeminal ganglion, cells with an intense staining, especially visible in the nucleus, were counted as positive for pERK. All cell bodies had a slight background staining in contrast to the surrounding tissue. The bifurcation of the trigeminal nerve was used as a landmark to distinguish V1/2 from V3 neurons. At least five sections per trigeminal ganglion were analysed. Sections processed exactly like those for Fos or pERK staining but without the first antibody did not show specific staining of cells. Omitting both first and secondary antibody did neither show any unspecific staining.

Data analysis

Experiments were compared using analysis of variance (ANOVA) followed by a LSD post hoc test. All tests were performed by the STATISTICA software package (StatSoft, Tulsa, OK, USA). All values are given as mean ± SE of the mean. Differences were considered significant at P < 0.05.


Fos immunoreactivity in the spinal trigeminal nucleus

Within the investigated 6-mm segment caudal to the obex significant differences in the numbers of Fos immunoreactive cells, indicated by dark nuclei, were observed between the experimental groups [ANOVA, F(2,14) = 9.3, P = 0.003, see representative photomicrographs in Fig. 1]. In control rats treated with saline and vehicle, Fos immunoreactivity was detected in 7.0 ± 1.0 cells per slice of the spinal trigeminal nucleus (seven rats with 12 sections each). Infusion of 915 µg/kg capsaicin caused a significant increase in the number of Fos immunoreactive nuclei to 33.7 ± 8.5 per section compared with vehicle (P < 0.001, LSD post hoc test). The nuclear Fos staining in these neurons was generally more intense compared with the vehicle treated specimens (compare Fig. 1C and D). Rats infused with 900 µg/kg olcegepant prior to capsaicin infusion had 17.5 ± 2.0 immunpositive cells per section, which was significantly less compared to rats infused with saline prior to capsaicin (P = 0.030, LSD post hoc test, Fig. 2A). No difference was found between left and right trigeminal nucleus (P = 0.35, LSD post hoc test) in either the groups. In a sub-analysis of superficial laminae (I/IIa) similar results were observed [ANOVA, F(2,14) = 27.5, P < 0.001, vehicle versus capsaicin: P < 0.001, capsaicin versus olcegepant/capsaicin: P = 0.011, LSD post hoc tests]. In deeper laminae (IIb-IV) a similar pattern but no significant differences were found (Fig. 2B). Further on, the cranio-caudal distribution of the reduction of Fos-positive nuclei after olcegepant infusion was examined in superficial laminae. A homogenous effect was found within 6 mm of the brainstem caudal to the obex (Fig. 2C). For the three experimental groups, the number of positive cells per slice was cumulated and fitted (Fig. 2D). In vehicle controls the brainstem slices showed less than 20 Fos-positive cells. Slices from capsaicin treated animals were mostly above this limit, and there was a leftward shift for the olcegepant pre-treated group. As previously reported, the capsaicin infusion transiently increased blood pressure and heart rate (Offenhauser et al., 2005). The body temperature slightly decreased to the level held by the feedback controlled homoeothermic system. This thermoregulatory effect of capsaicin is long known (Issekutz et al., 1950).

Figure 1

Specimen of Fos immunoreactivity in the spinal trigeminal nucleus. Cross-sections through the medulla with the trigeminal nucleus caudalis (darker background) stained for the activation marker Fos in neuronal nuclei. (A) Representative specimen fixed 2 h after a 30 min i.v. capsaicin infusion. (B) Representative specimen fixed 2 h after a 30 min i.v. vehicle infusion. Magnification bars 100 µm. (C and D) Details of A and B, respectively, at higher magnification (bars 50 µm).

Figure 2

Summary of Fos immunoreactive neurons in the spinal trigeminal nucleus. Intravenous capsaicin infusion for 30 min increased the number of Fos-positive neurons per slice 2 h after start of the infusion. Pre-treatment of animals with olcegepant (900 µg/kg i.v.) reduced this increase significantly. (A) All laminae were considered. No differences between both sides of the trigeminal nucleus were observed. (B) A similar pattern was found in superficial and deeper laminae. In superficial laminae, more Fos-positive neurons were observed. (C) Fos-positive cells within the superficial laminae were analysed due to their distance to the obex, two slices per millimetre were counted in each animal. (D) Histogram for the number of observations of Fos-positive cells per slice in the superficial laminae.

pERK immunoreactivity in the trigeminal ganglion

Capsaicin was injected unilaterally into the area innervated by the three trigeminal branches. Compared with the contralateral side, the percentage of pERK positive cells in the trigeminal ganglion was significantly increased [ANOVA, F(1,12) = 11.0, P = 0.006, Fig. 4A; representative specimens in Fig. 3]. This increase in pERK positive cells was found both in animals pre-treated with saline (P = 0.036, n = 7, LSD post hoc test) and olcegepant (P = 0.038, n = 7, LSD post hoc test). The number of pERK positive neurons in the ganglia of the capsaicin-treated side after pre-treatment with olcegepant was not different from the number of positive neurons after pre-treatment with saline (P = 0.072, LSD post hoc test). The ratio of pERK positive cells of the capsacicin treated versus the control side was analysed separately for the trigeminal branches V1/2 and branch V3 [ANOVA, F(1,12) = 0.2, Fig. 4C]. Ratios of corresponding slices of the left and the right trigeminal ganglion were calculated. The distribution of these ratios differed from the null hypothesis (vehicle pre-treatment P = 0.002 and olcegepant pre-treatment P < 0.001, one-sample t-tests against one) and did not depend on the pre-treatment (P = 0.82, t-test, independent samples, Fig. 4D). In an additional five animals, lidocaine was pre-injected on one side of the head followed by bilateral injection of capsaicin 15 min later into the same facial sites as in the main experimental group. The local anaesthesia reduced the average number of pERK-positive cells (P = 0.034, n = 5, t-test dependent samples, Supplementary Fig. 1). However, the ratio of phosphorylated ERK comparing capsaicin versus capsaicin with lidocaine pre-treatment and capsaicin versus control was similar.

Figure 3

Specimen of pERK immunoreactivity in the trigeminal ganglion. Longitudinal sections through trigeminal ganglia stained for the activation marker pERK visible most clearly in neuronal nuclei. All animals were unilaterally injected with capsaicin in all three trigeminal branches and 10 min later fixed by perfusion. Representative specimens from the injected (A and C) and the control side (B and D) showing the mandibular (V1) region (A and B) and the ophthalmic (V3) region (C and D). Magnification bars 100 µm in A, B and 50 µm in C and D.

Figure 4

Summary of pERK immunoreactive neurons in the trigeminal ganglion. Number of pERK immunoreactive trigeminal ganglion neurons after unilateral injections of capsaicin (33 µl, 1 mM) in facial areas innervated by the three trigeminal branches. (A) Trigeminal ganglia from the capsaicin-injected side showed a significantly higher proportion of pERK-positive cells compared with the control side. No difference was found for the pre-treatment with saline or olcegepant (900 µg/kg). (B) Distribution of percentages of pERK-positive cells from all counted slices. Note that the histogram underestimates the intra-animal difference evoked by capsaicin treatment due to the inter-animal variation. (C) Percentage of pERK-positive cells as ratio of the capsaicin-injected versus the control side. No difference was found between the branches of the trigeminal nerve or between animals pre-treated with saline or olcegepant. (D) Distribution of ratios between corresponding slices of the capsaicin-injected versus the control side across all animals. The histogram shows a similar rightward shift with olcegepant and saline pre-treatment.


Using immunohistochemical methods, we visualized the activation of first and second order trigeminal neurons after noxious stimulation with capsaicin. Systemic pre-treatment with the CGRP receptor antagonist olcegepant reduced the activation of second order but not of first order trigeminal neurons.

CGRP receptor inhibition in the treatment of migraine

The sensory neuropeptide CGRP is present in a considerable proportion of trigeminal afferent fibres and their cell bodies (Edvinsson et al., 1987b; Messlinger et al., 1993). Direct measurements supported by functional evidence have shown that trigeminal afferents release CGRP on stimulation from both the peripheral (Geppetti et al., 1990) and central processes (Samsam et al., 1999; Jenkins et al., 2004). In peripheral tissues including intracranial blood vessels CGRP causes potent vasodilatation (Messlinger et al., 1995; Williamson et al., 1997). At central terminals CGRP seems to potentiate synaptic transmission (Storer et al., 2004). Both effects, which could be inhibited by CGRP receptor antagonists, are attributed a role in the generation of migraine pain (Link et al., 2008), rendering them a potential target for anti-migraine drugs (Doods et al., 2007; Durham, 2008). Three drugs that potently and selectively inhibit CGRP receptors have been presented so far (Doods et al., 2000; Degnan et al., 2008; Salvatore et al., 2008).

Site of action

Potential sites of origin of migraine pain include intracranial blood vessels, the trigeminal neurons, the spinal trigeminal nucleus caudalis, pathways of descending inhibition or facilitation, and the cerebral cortex. For all these sites hypotheses of migraine generation have been formulated (Pietrobon, 2005) with some supporting experimental but not unequivocal evidence. No matter whether a peripheral or a central site is the prevailing source for activation, all painful trigeminal sensations are conveyed by the spinal trigeminal nucleus, serving as rationale to examine neuronal activation at this site. Imaging studies have shown that an area corresponding to this brainstem region is activated in trigeminal pain (DaSilva et al., 2002).

For olcegepant and telcagepant a therapeutic efficacy in migraine has been shown in clinical trials (Olesen et al., 2004; Ho et al., 2008a). In both clinical trials the effective drug concentrations were unexpectedly high compared with the potency of the CGRP receptor antagonists in established cellular and in in vitro models. The effective and recommended 2.5 mg dose of olcegepant led to a maximum plasma level of ∼240 nM (Iovino et al., 2004) compared with an IC50 at human cerebral arteries of 0.08 nM (Edvinsson et al., 2002). This raises questions about the site of action. The central nervous system is assumedly reached by only small amounts of the potent drugs due to their low penetration of the blood brain barrier (Doods et al., 2007). This may serve as an explanation for the discrepancy between the potency at the receptor level and the doses needed to treat migraine.

The low permeability of the blood brain barrier for olcegepant is also supported by the difference in the IC50 of this drug for pial compared with dural blood vessels in rat (Petersen et al., 2004). Further support comes from in vitro experiments on pressurized rat cerebral arteries, in which luminal administration of CGRP receptor antagonists was less effective than abluminal administration (Edvinsson et al., 2007). Details for the blood brain barrier permeability of olcegepant have not been published by the developers. Two previous reports have found no support for a peripheral site of action; the peripheral processes of primary trigeminal neurons seem to lack the CGRP receptor (Lennerz et al., 2008) and the application of CGRP was reported to neither sensitize nor activate rat trigeminal neurons in vivo (Levy et al., 2005).

The possible sites of action are discussed in the ‘perspective’ to the first paper reporting clinical efficacy of olcegepant (Durham, 2004) and in most reviews thereafter including a recent one dedicated to this subject (Edvinsson, 2008); however, no conclusive evidence has been reported so far. The present study is the first to address the site of action of systemically administered olcegepant. Since a reduction in pERK in the trigeminal ganglion indicative for a peripheral action of olcegepant was not observed, the reduced activation of Fos in the brainstem after noxious stimulation with capsaicin indicates a central site of action. In contrast to CGRP receptor inhibition, local anaesthesia by lidocaine reduced capsaicin-induced pERK formation.

Different models of activation

We have chosen different routes of capsaicin administration to investigate activation of primary and secondary trigeminal neurons. For the evaluation of central trigeminal activity with Fos immunochemistry, a previously established model using intravenous capsaicin infusion was applied (Offenhauser et al., 2005). This allowed us to select a capsaicin dose, which was known to create a robust signal and low variation. The observed activation by capsaicin in our study is similar to the previous results. Capsaicin can activate nociceptive trigeminal afferents through TRPV1 receptors at both peripheral and central nerve endings (Hershey et al., 2005; Price et al., 2005). Since capsaicin is a lipophilic compound, also central TRPV1 receptors might have been activated and contributed to the Fos signal indicative for an activation of second order neurons. In this case the reduction of the Fos signal in the experimental group treated with olcegepant would provide further support for a central action of the CGRP receptor antagonist. According to a recent immunohistochemical investigation from our group, CGRP receptors in the rat spinal trigeminal nucleus are located presynaptically, i.e. most likely on central primary afferent endings (Lennerz et al., 2008). This would imply that centrally acting CGRP receptor antagonists inhibit presynaptic functions, especially reducing the release of glutamate and neuropeptides, whereas iontophoretic application of CGRP may facilitate neurotransmitter release to strengthen synaptic transmission (Storer et al., 2004).

To examine modulation of primary trigeminal activity by olcegepant, local activation at a peripheral site by facial injections of capsaicin was used, as previously established in a similar preparation with injections of capsaicin (Noma et al., 2008). In their experiment, pERK activation was not detected after 60 min compared with controls; we neither found pERK activation in the trigeminal ganglia of animals used for the Fos staining after systemic capsaicin infusion. Therefore, in an independent set of experiments a stimulation by capsaicin was chosen to activate the same types of trigeminal neurons as during systemic application. This approach allowed us to use the contralateral trigeminal ganglion of the same animal as a control. All trigeminal branches were stimulated by capsaicin to yield a maximum effect in activating trigeminal afferents compared to the contralateral ganglion. The local capsaicin injection provides a better control of neuronal activation restricted to the trigeminal system in comparison with the systemic infusion, which on the other hand is necessary for the fast time course of pERK. Since the rationale was to compare the side of maximal activation with the contralateral not activated side, a vehicle injection was not made.

In additional experiments we demonstrated that, in contrast to olcegepant, local anaesthesia by lidocain blocks the activation by local injection of capsaicin.

Cellular activation pathways including extracellular signal-regulated kinases have been reviewed extensively (Roux and Blenis, 2004; Murphy and Blenis, 2006). The extracellular signal-regulated protein kinases are activated in a stimulus-intensity dependent manner and restricted to the activated subpopulation of neurons, e.g. after capsaicin stimulation (Dai et al., 2002). Phosphorylated extracellular signal-regulated kinases serve not only as reporters of evoked activity, but have also been found to be critically involved in sensitization (Zhuang et al., 2004). The results show clearly that pre-administration of olcegepant did not change the robust activation. Therefore inhibition of CGRP receptors, irrespective of the site of their expression (Lennerz et al., 2008), did not affect the activation of primary trigeminal neurons indicated by phosphorylation of extracellular signal-regulated kinase.

Phosphorylation of the MAP kinase extracellular signal-regulated kinase fosters the expression of CGRP via an enhancer of the CGRP/calcitonin gene promoter specific for CGRP producing cell lines (Durham and Russo, 1998). This MAP kinase pathway is suppressed by activation of 5-HT1 receptors resulting in a decrease in the CGRP promoter activity in primary trigeminal neurons, which is regarded as a possible mechanism of the anti-migraine effects of 5-HT1 receptor agonists (Durham and Russo, 2003). Therefore, extracellular signal-regulated kinase phosphorylation might serve as an index for trigeminal nociception, which is relevant for the development of migraine headaches.

This study confirmed earlier results with a different approach showing that inhibition of CGRP receptors reduces trigeminal activity. We further presented the first evidence that the decrease in spinal trigeminal activity after systemic infusion of a CGRP receptor antagonist is likely to occur at the central branch of trigeminal neurons.


Supported by the Federal Ministry of Education and (German Headache Consortium).

Supplementary material

Supplementary material is available at Brain online.


We acknowledge Birgit Vogler and Jana Schramm for their technical assistance.


  • Abbreviations:
    calcitonin gene related peptide
    protein encoded by the c-Fos immediate early gene
    phosphorylated extracellular signal-regulated kinase
    transient receptor potential vanilloid 1


View Abstract