Nitric oxide synthase,calcitonin gene-related peptide and NK-1 receptor mechanisms are involved in GTN-induced neuronal activation

Abstract

Background and aim

Infusion of glyceryltrinitrate (GTN), a nitric oxide (NO) donor, in awake, freely moving rats closely mimics a universally accepted human model of migraine and responds to sumatriptan treatment. Here we analyse the effect of nitric oxide synthase (NOS) and calcitonin gene-related peptide (CGRP) systems on the GTN-induced neuronal activation in this model.

Materials and methods

The femoral vein was catheterised in rats and GTN was infused (4 µg/kg/min, for 20 minutes, intravenously). Immunohistochemistry was performed to analyse Fos, nNOS and CGRP and Western blot for measuring nNOS protein expression. The effect of olcegepant, L-nitro-arginine methyl ester (L-NAME) and neurokinin (NK)-1 receptor antagonist L-733060 were analysed on Fos activation.

Results

GTN-treated rats showed a significant increase of nNOS and CGRP in dura mater and CGRP in the trigeminal nucleus caudalis (TNC). Upregulation of Fos was observed in TNC four hours after the infusion. This activation was inhibited by pre-treatment with olcegepant. Pre-treatment with L-NAME and L-733060 also significantly inhibited GTN induced Fos expression.

Conclusion

The present study indicates that blockers of CGRP, NOS and NK-1 receptors all inhibit GTN induced Fos activation. These findings also predict that pre-treatment with olcegepant may be a better option than post-treatment to study its inhibitory effect in GTN migraine models.

Keywords

Nitric oxide (NO)nitric oxide synthase (NOS)glyceryltrinitrate (GTN)awake rat model migraine neuronal activation L-nitro-arginine methyl ester (L-NAME)olcegepant trigeminal vascular system sumatriptan

Introduction

Glyceryltrinitrate (GTN) provokes migraine attacks in humans, and infusion of GTN is the most widely accepted human model of migraine. GTN diffuses freely and liberates nitric oxide (NO) in peripheral and cerebral structures, but the exact neurobiological mechanism of GTN-induced headache is yet to be understood. In migraineurs, GTN causes an immediate headache that is followed by a migraine-like delayed headache five to six hours later (1).

The involvement of endogenous NO production during migraine attacks is substantiated by the anti-migraine effect of the nitric oxide synthase (NOS) inhibitor L-N^G-monomethyl arginine citrate (L-NMMA) (2). This clinical observation has been supported by electrophysiological studies showing that long-lasting infusion of NO donors in rats stimulates ongoing neuronal activity in the spinal trigeminal nucleus caudalis (TNC) (3). This activity was decreased by inhibiting endogenous NO (4), suggesting that the NO tone partially modulates the activity of the spinal TNC. NO also modulates the release of vasoactive neuropeptides including calcitonin gene-related peptide (CGRP) in vitro (5) and in vivo (6). Moreover, CGRP receptor antagonists are clinically effective against acute migraine (7,8). Accordingly in rats, intravenous (i.v.) injection of olcegepant was effective in preventing the vasodilatory actions of endogenously released CGRP following transcranial electrical stimulation (9) and lowering the neuronal activity of the spinal TNC (10). Also, microiontophoretic injection of olcegepant inhibited trigeminal neuronal activity evoked by stimulation of the superior sagittal sinus (SSS) in cats (11), indicating the importance of CGRP in controlling the activity of spinal trigeminal neurons. An inflammatory response may be one of the mechanism by which GTN induces headache, since induction of inducible nitric oxide synthase (iNOS) by NO was observed along with interleukin 1β granular changes of the mast cells in the dura mater (12).

Recently we have developed an experimental model of migraine using GTN infusion in unanaesthetised, freely moving rats. It is the closest possible simulation of the human GTN model. This model circumvents confounders like high doses, anaesthetics, acute surgical stress and hypotensive effects associated with earlier GTN-infusion studies in animals. Sumatriptan treatment significantly decreased GTN-induced Fos expression in this model (13). The aim of the present study was to investigate the role of NOS and CGRP systems in our rat migraine model and therefore to investigate the effect of L-NAME and CGRP receptor antagonist olcegepant, a drug with known efficacy in the treatment of migraine attacks in humans. To evaluate the role of neurokinin (NK)-1 receptor mechanisms, we also investigated the effect of L-733,060 a highly selective NK-1 antagonist.

Materials and methods

Animals

Seventy-eight male Sprague-Dawley rats weighing 320–340 g (Taconic M&B, Denmark) were used for the study. The rats were maintained in cages with a 12-hour light/dark cycle and free access to food and water. All the experimental protocols were approved by the Danish committee for experiments with animals (2009/561-1664).

Surgical procedures

Surgery was performed according to the procedures reported in our recent study (13). In brief, rats were anaesthetized by injecting ketamine (100 mg/kg) and xylazine (7.5 mg/kg) and the femoral vein was cannulated. The sealed cannula was then placed subcutaneously (s.c.) and pulled out at the neck nape. Baytril (10 mg/kg s.c.) and Rimadyl (5 mg/kg s.c.) treatment was given for 72 hours and temgesic treatment for 48 hours (two days) post-surgery with an interval of 24 hours. Seven days later the rats were connected to a tether and allowed to move freely in the cage. After two days, GTN (4 µg/kg/min) was administered i.v. for 20 minutes and the rats were perfused at different time points. Saline- or vehicle- (0.18% ethanol) treated rats were perfused after two hours and four hours. Sumatriptan (0.6 mg/kg for three minutes) or L-NAME (40 mg/kg) was infused over 10 minutes and five minutes later was followed by GTN infusion. Olcegepant (1 mg/kg) was infused over three minutes and 10 minutes later was followed by GTN infusion. L-733060 (1 mg/kg) was infused over three minutes and 10 minutes later was also followed by GTN infusion. All the drugs were infused over three minutes except for L-NAME, which was infused over 10 minutes because L-NAME increases systemic blood pressure. For post-treatment studies, olcegepant was infused over 30 minutes after the start of the GTN infusion.

Blood pressure measurements

The right femoral vein and carotid artery were cannulated in three rats. Post-operative care was performed as described above. The day after surgery, rats were connected to a tether and allowed to move freely in the cage. Baseline mean arterial blood pressure (MABP) was recorded for a minimum of 30 minutes through the carotid artery. Saline was infused via the femoral vein for 10 minutes, followed by 40 mg/kg L-NAME over 10 minutes. Blood pressure was monitored continuously during the L-NAME infusion and up to one hour after infusion. MABP was analysed using Perisoft (Version 2.5.5; Perimed AB, Järfälla, Sweden).

Drugs

The drugs used in our research included: GTN stock solution (5 mg/ml in 95% ethanol) (Nycomed, Roskilde, Denmark), L-NAME (Sigma-Aldrich, Schnelldorf, Germany), sumatriptan (kindly provided by John Andrews from Neur Axon (Mississauga, Canada)), olcegepant (Boehringer Ingelheim Pharma K.G., Biberach, Germany), L-733060 hydrochloride (Tocris bioscience, Bristol, UK), Baytril (50 mg/ml) and Xylazine (20 mg/ml) (Rompun®, Bayer Inc, Germany), Rimadyl (50 mg/ml) (Pfizer Inc, NY, USA), Temgesic (0.3 mg/ml) (Schering Plough, Europe, Belgium) and ketamine (100 mg/ml) (Intervet, Skovlunde, Denmark).

Protein expression studies

Immunohistochemistry (IHC)

Dura mater, trigeminal ganglion (TG) and TNC were isolated after perfusion and post-fixed overnight in 4% paraformaldehyde (PFA) and transferred to 30% sucrose solution. TG sections were cut longitudinally with a thickness of 12 µm and were collected on glass slides and processed further for immuno-fluorescence staining. Cross-sections of TNC were cut, with each section having a thickness of 40 µm. Every tenth section was subjected to Fos or CGRP immunostaining in the TNC. The whole dura mater or TNC was stained for neuronal nitric oxide synthase (nNOS), Fos or CGRP in a free-floating manner.

c-fos and nNOS staining

The immunolabeling procedures were performed according to the avidin-biotin-peroxidase complex (ABC) protocol as described (14). Sections or free-floating dura were incubated overnight with polyclonal rabbit anti-Fos antibody or polyclonal rabbit anti-nNOS antibody. Biotinylated goat anti-rabbit immunoglobulin G (IgG) was used as the secondary antibody. Buffer controls omitted either primary or secondary antibodies. The sections or dura were subsequently mounted onto the glass slides and coverslipped, and were observed through a Leica DMR HCS microscope using a 20 × objective lens. An observer blinded to the treatment counted the Fos-positive cells in the region of interest. Fos-positive cells were counted in lamina I and II in every 0.72 mm starting from 0.80 mm from the obex and ending at 5.12 mm. This 5.12-mm continuum covered the subnucleus caudalis zone (Vc), C1 and the rostral part of C2. Fos-positive cells were also evaluated in the tractus nucleus solitarius (NTS).

nNOS fibre density

For nNOS nerve fibre-density analysis, using a fixed magnification of 20 × in a light microscope, nNOS-positive nerve fibres were counted in three different visual fields around MMA. The mean value was calculated for each rat and data were presented as number of nerve fibres/visual field.

CGRP immuno-fluorescence staining

Dura mater and TNC were processed for CGRP immunostaining using monoclonal mouse anti-CGRP. Alexa Flour 594 donkey anti-mouse IgG was used as a secondary antibody. Buffer controls were included as mentioned above. Sections stained with fluorescent-tagged secondary antibody were observed under the fluorescence microscope. Four pictures from each rat were captured under fixed magnification of 20 × in the regions around the MMA in dura mater. Similarly, three pictures of the left and right TNC were captured from each rat at 20 × magnification. An observer blinded to the treatment analysed the mean intensity of CGRP-positive fibres using Image J software. The pictures were converted to binary images and, using a rectangular tool of the fixed area, the mean intensity of nerve fibres and nerve endings were determined using the ‘analyse particle’ parameter. Mean values were calculated for each rat.

Western blotting

Dura was homogenised with a mechanical stirrer in 50–100 µl of ice-cold lysis buffer (10 mM Tris (pH = 7.4), 0.5% Triton X, phosphate and proteinase inhibitor tablets). The samples were centrifuged at 10,000 × g for 10 minutes at 4℃, and the supernatants were collected. The protein concentrations were determined using a Bio-Rad DC protein assay (Bio-Rad, CA, USA). Fifteen µg of protein was loaded into the wells. The proteins were separated by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) and then transferred to a polyvinylidene difluoride (PVDF) membrane (Bio-Rad). The PVDF blots were blocked in 5% non-fat dry milk in Tris-buffered saline (TBS) with 1% Tween-20 (TBS-T) for one hour at room temperature. Blots were cut in halves according to the expected molecular weights of the target antigens. The antibodies (Table 1) were diluted in TBS-T containing 3% non-fat dry milk.

Table 1.

Details of the antibodies used in the immunohistochemistry (IHC) and Western blot (WB) experiments. Tissues or blots were incubated with the primary and secondary antibody combinations as indicated in each row.

Molecular marker	Primary antibody	Dilution	Secondary antibody	Dilution
Fos (IHC)	Rabbit polyclonal anti-c-fos (Calbiochem, Darmstadt, Germany)	1:4000	Biotin-SP-conjugated F(ab’)2 fragment donkey anti-rabbit IgG (Jackson ImmunoResearch, Suffolk, UK)	1:2000
nNOS (IHC)	Rabbit, polyclonal anti-nNOS (NOSI) (Alexis, Enzo Life Sciences, Lörrach, Germany)	1:4000	Biotin-SP-conjugated F(ab’)2 fragment donkey anti-rabbit IgG (Jackson ImmunoResearch, Suffolk, UK)	1:2000
nNOS (WB)	Mouse IgG polyclonal (BD Biosciences, San Jose, CA, USA)	1:1000	Stabilized goat anti-mouse/HRP (Pierce, Rockford, IL, USA)	1:10,000
CGRP (IHC)	Mouse monoclonal anti-calcitonin gene-related peptide (C7113) (Sigma-Aldrich, USA)	1:1000	Alexa Fluor® 594 donkey anti-mouse IgG (Invitrogen, Oregon, USA)	1:200
β-actin (WB)	(A544) Mouse monoclonal clone AC-15, ascites fluid (Sigma-Aldrich, USA)	1:10,000	Stabilised goat anti-mouse/HRP (Pierce, Rockford, IL)	1:10,000

HRP: horseradish peroxidase; IgG: immunoglobulin G; nNOS: neuronal nitric oxide synthase.

The upper halves of the blots were incubated with the nNOS primary antibody, and the lower halves were incubated with the β-actin primary antibodies. The blots were left overnight at 4℃. The secondary antibodies, all conjugated to horseradish peroxidase (HRP), were incubated with the PVDF blot for one hour at room temperature. The PVDF blots were then processed for analysis using an enhanced chemiluminescence (ECL) detection kit (Pierce). The blots were scanned by an electronically cooled CCD Camera System (Fujifilm LAS-1000) and the chemifluorescent images were captured and stored digitally. To compare the expression profile between the tissues, the densitometry of each antibody signal was measured in Image Gauge 4.0 and related to the densitometry counts from the β-actin signals (loading controls). Densitometry data were graphically visualized.

Statistical analysis

All values are presented as mean ± SEM. Experiments were compared using analysis of variance (ANOVA) (Kruskal-Wallis test), which was followed by Dunn’s post-hoc test to determine the significant difference from the saline-treated group. Two-way ANOVA was used followed by Bonferroni post-hoc test for the comparisons of Fos-positive cells between the treatment groups in different levels of brain stem sections. Differences were considered significant at p < 0.05. GraphPad Prism (GraphPad Prism Software Inc, San Diego, CA, USA) was used for statistical analysis.

Results

Effect of GTN infusion on Fos protein expression in TNC and upper cervical dorsal horn with or without L-NAME pre-treatment

In the group treated with GTN and sacrificed after four hours, a significant increase of Fos-positive cells was seen in the superficial laminae, from 0.80 mm to 4.40 mm as compared to the vehicle (two-way ANOVA, F(2,14) = 29.24, p < 0.001) (Figure 1). Fos-positive cells were more concentrated in the regions between 2.24 mm to 4.40 mm from the obex. No differences were detected between the left and the right TNC in any of the groups. Furthermore, the Fos-positive cells were also observed from the dorsal to ventral axis with the cells slightly higher towards the ventrolateral axis. Pre-treatment with L-NAME (40 mg/kg over 10 minutes) showed a significant decrease in Fos-positive cells in the superficial laminae when compared with the GTN-treated rats (p < 0.001).

Figure 1.

Analysis of total number of Fos-positive cells along different levels from the obex following different treatments. GTN infusion showed an increase in Fos expression at four hours when compared with saline. Pre-treatment with L-NAME significantly reduced the Fos expression induced by GTN. Two-way ANOVA was followed by Bonferroni’s post-hoc test, *p < 0.001 when compared to saline, ^#p < 0.001 when compared to GTN treatment after four hours. (N = 6–7).

We also evaluated the NTS, which contains the neurons involved in arterial baroreflex afferent processing for Fos expression. We did not observe any Fos expression after treatment with GTN or saline in the NTS (data not shown). Moreover, we measured the MABP in three rats after L-NAME infusion, which showed an increase by 48% compared to the vehicle treatment. However, we also analysed Fos expression in these rats, and did not observe any increase in the NTS or TNC regions, suggesting that L-NAME per se did not modulate Fos expression (data not shown).

Effect of pre- and post-treatment with olcegepant on GTN-induced Fos expression

We hypothesised that CGRP might play a major role in GTN-induced Fos activation. In order to mimic previous human studies in which olcegepant was given after GTN infusion, a group of rats were infused with olcegepant (1 mg/kg over three minutes), 30 minutes after the start of the GTN infusion and analysed for the Fos expression in TNC and upper cervical dorsal horn. The dose of CGRP antagonist used (1 mg/kg) in the present study has previously been shown to act directly on the TNC (15). Analysis of each level did not show any significant differences (Figure 2). Therefore, we included a group in which olcegepant was given as a pre-treatment. Interestingly, this group showed a significant decrease (two-way ANOVA, F(2,13) = 7.01, p = 0.008) in Fos expression at four hours in the superficial laminae (Figure 2).

Figure 2.

Total Fos-positive cells along the rostrocaudal axis from obex (per 0.72 mm). Post-treatment with olcegepant did not inhibit the GTN-induced Fos expression, whereas pre-treatment with olcegepant significantly reduced Fos expression. Two-way ANOVA was followed by Bonferroni’s post-hoc test, *p < 0.05 when compared to GTN treatment after four hours (N = 5–7).

Effect of the NK-1 receptor antagonist L-733,060 on GTN-induced Fos expression

Rats were infused with 1 mg/kg of L-733,060 10 minutes prior to GTN infusion. The analysis of Fos in the superficial laminae of this group showed a significant decrease when compared with the GTN-treated rats (two-way ANOVA, F(1,9) = 26.35, p < 0.001) (Figure 3).

Figure 3.

Rostrocaudal analysis of Fos expression after pre-treatment with NK-1 antagonist. L-733,060 showed a significant reduction in Fos expression induced by GTN. Two-way ANOVA was followed by Bonferroni’s post-hoc test, *p < 0.001 when compared to GTN treatment after four hours (N = 5–6).

nNOS expression in trigeminovascular system

IHC of the whole mount dura mater showed an increase in nNOS fibre density (Figure 4(a)) at four hours following GTN infusion (7.18 ± 1.42) when compared with the control group (1.83 ± 0.3). A beaded structure of nNOS immunopositive fibres was observed in dura mater after two hours. Pre-treatment with L-NAME did not show any appreciable difference in nNOS immunopositive fibres (6.5 ± 0.76) when compared with the four-hour GTN-treated group. Our previous study has shown an inhibitory effect with sumatriptan (0.6 mg/kg over three minutes), a very specific anti-migraine drug, on GTN-induced Fos activation. Therefore, in this study we also evaluated the effects of sumatriptan on nNOS expression. However, sumatriptan did not reduce dural nNOS expression induced by GTN (6.16 ± 1.14).

Figure 4.

Effect of GTN infusion on protein expression of nNOS in dura mater. Representative images of nNOS nerve fibres in the vehicle- and GTN-treated group and analysis of nNOS-immunoreactive fibre-density/visual field in dura mater four hours after GTN infusion and after pre-treatment with L-NAME and sumatriptan (a). Western blot and quantified protein expression of nNOS in dura mater (b). The signal was observed at approximately 155 kDa. Pre-treatment with L-NAME or sumatriptan or L733060 did not significantly inhibit the nNOS expression at four hours. *p < 0.05 (ANOVA (Kruskal Wallis) followed by Dunn’s post hoc test) compared to vehicle. (N = 4–6).

Western blot with dura mater lysate for nNOS-immunoreactivity (nNOS-IR) confirmed the results obtained by IHC. Two characteristic bands of the nNOS protein were identified. The upper band at 155 kDa corresponds to nNOS and the second band refers to lower molecular mass variant βNos 1 (16). The physiological role of βNOS 1 is unknown and hence we chose to examine only the intensity of the top band, which so far is known to be functional. Densitometric analyses of the upper band confirmed that nNOS was upregulated four hours after GTN infusion (Figure 4(b)). Pre-treatment with L-NAME, sumatriptan or L-733060 did not show any significant decrease of the nNOS in dura mater at four hours as observed in the IHC. No differences in nNOS expression were observed in the TNC or TG after GTN treatment.

CGRP protein expression in the trigeminovascular system

Total CGRP expressed in the dural nerve fibres were quantified. IHC showed an increase in CGRP immunopositive fibres in dura mater at four hours (Figure 5). Furthermore, the nerve terminals visually exhibited a bulb-like structure in the GTN-treated group. The increase in CGRP immunopositive dural nerve fibres was significantly inhibited by L-733060 but not by L-NAME or sumatriptan. Except for sumatriptan other drugs inhibited the bulb-like structure observed after GTN infusion.

Figure 5.

Representative images of CGRP immunoreactivity in dura at two hours (b) and four hours (c) after GTN or vehicle infusion (a) by IHC. CGRP expression was analysed using Image J software. An increase in CGRP expression was seen in the nerve fibres four hours after GTN infusion. Pre-treatment with sumatriptan (d) or L-NAME (e) did not show any reduction in CGRP expression, whereas L733060 (f) reduced CGRP expression in the nerve fibres of dura. However, a bulb-like structure was seen at the nerve terminals four hours after GTN infusion (arrows). Mean fluorescence intensity of the total CGRP expression in dura (g). ***p < 0.001 (ANOVA (Kruskal-Wallis) followed by Dunn’s post hoc test) compared with vehicle. ^#p < 0.05 compared to four-hour GTN treatment (N = 4).

CGRP immunopositive fibres in TNC at two and four hours were increased in the superficial layers of the TNC (Figures 6 and 7). Pre-treatment with sumatriptan or L-NAME or L-733060 decreased CGRP expression in the TNC at four hours. TG did not show any signs of increase in CGRP expression.

Figure 6.

Rat TNC section nuclei stained with the fluorescent dye DAPI (a), showing the region of interest that was magnified in the subsequent images. The superficial laminae I and II over mid dorsal to ventrolateral region (shown in the box in (a)) was magnified at 20 × to compare the CGRP expression between different groups. Representative images of CGRP immunoreactivity in TNC at two hours (c) and four hours (d) after GTN infusion and compared with the vehicle-treated group (b). Pre-treatment with sumatriptan (e), L-NAME (f) or L733060 (g) reduced the CGRP expression in TNC. Mean fluorescence intensity of the CGRP expression in TNC (h). *p < 0.05 ANOVA (Kruskal Wallis) followed by Dunn’s post hoc test) compared to vehicle. ^#p < 0.05 compared to four-hour treatment (N = 4).

Figure 7.

Possible mechanisms of GTN in the TVS: GTN can interact at different levels of the TVS by releasing NO which activate multiple signalling cascades. At the peripheral level (dura mater) CGRP-ir and nNOS-ir are increased and their release may augment inflammatory mediators. It may also sensitise sensory nerve endings and via TG increase the input to TNC. Activation of TG by increased peripheral signals and/or direct NO causes enhanced activation of second-order neurons in the superficial laminae I/II of the TNC. In TNC the amplified signal together with direct NO-induced activation increases c-fos and upregulates CGRP in nerve fibres that may lead to pain signalling in higher centres of the CNS.

Discussion

CGRP antagonism and NOS inhibition are validated targets for new anti-migraine drugs. In the present study, we explore these targets in our naturalistic animal model of migraine showing that GTN-induced Fos activation is reversed using the CGRP receptor antagonist olcegepant, the non-selective NOS inhibitor L-NAME and the NK-1 receptor antagonist L-733060, suggesting the involvement of CGRP, endogenous NO production and NK-1 receptor activation after GTN infusion in the rat. This is the first study to demonstrate the effect of different drugs on neuronal activation in the trigeminovascular pathways, NOS and CGRP systems with a clinically relevant dose of GTN in awake, freely moving rats.

Animal model

In humans, migraine-provoking experiments are conducted in the conscious state. In our model, rats were unanaesthetised and freely moving with no influence of acute surgery or anaesthetic interventions, which are known to upregulate Fos (17) and inhibit NO synthesis (18). The GTN dose (4 µg/kg/min, for 20 min, i.v.) chosen for this study was only eight times higher than the dose given in humans (19) and was given through the same route as in human migraine-provocation studies. In anaesthetized rats this dose caused a significant drop in MABP but not in our model (13). Hypotension can result in baroreceptor activation leading to Fos expression in the barosensitive neurons of the tractus nucleus solitarius (NTS) (20), which may contribute to the activation of brain stem nuclei and higher centres. We did not observe Fos expression in NTS, suggesting that GTN infusion did not have any systemic side effects as observed in the GTN infusion model with extremely higher doses (21). The rats used in these studies were normal and most probably migraine free. However, previous studies suggest that long-term GTN infusion can change naïve rats into a state similar to migraine in humans (22). Further, intraperitonial infusion of GTN in awake rats has shown to reduce the threshold of pain and facial allodynia in naïve rats (23). In normal subjects, infusion of GTN induces a bilateral headache, which supports well our findings of fos upregulation bilaterally. All these observations together suggest that GTN infusion in naïve rats might be a useful animal model.

Peripheral activation

The peripheral effects activated after GTN infusion in our rat model could be seen as a time-dependent increase in CGRP and nNOS expression in dura mater. The NK-1 antagonist L-733060 inhibited the increase in CGRP expression, which suggests that peripheral neurogenic inflammation might be upstream from this phenomenon. The bulb-like structures of the CGRP immunopositive nerve terminals observed in the GTN-treated group were distinct from the beaded structure seen along the nerve fibres. The presence of such bulb-like varicosities at CGRP nerve terminals has been reported after electrical stimulation; however, they were not reduced by sumatriptan as observed in our study (24). GTN infusion in our model also showed an increase in nNOS protein in dura that was probably the result of an increased endogenous production of NO (25) and thus, NO may contribute to peripheral and central sensitisation. We suggest that there is a basal ongoing activity in the periphery due to the 20-minute infusion of GTN. Factors like CGRP release and neurogenic inflammation in the periphery may potentiate this ongoing activity, for example, sodium nitroprusside when applied topically potentiated the responses to mechanical stimuli (26). Also, GTN infusion potentiated the responses to dural and facial stimulation in another model (27), suggesting that NO can amplify the afferent signals. It was recently suggested that NO acts on dural vessels, which subsequently contributes to the amplification of these afferent signals by releasing pro-inflammatory factors leading to neurogenic inflammation (28). The role of specific NOSs is unclear so far. An increase in nicotinamide adenine dinucleotide phosphate-oxidase (NADPH) diaphorase was previously shown after GTN infusion (29). The results from this and the present study suggest that endogenous NO may act as a chemical stimulant to sensitise peripheral afferent nerves, which in turn activate second-order neurons. Prior to the upregulation of nNOS, IHC revealed a structural transition from normal to beaded nerve fibres. This may point to the reorganisation/upregulation of nNOS in the fibres and has been described earlier (30). Whilst our study provides compelling evidence of the involvement of NO, pre-treatment with L-NAME did not suppress the increase in nNOS protein in dura mater after GTN infusion. This is in line with the fact that L-NAME inhibits the production of NO without affecting the expression of NOS enzymes and is supported by studies showing that NOS activity can be increased without an increase in NOS expression (31). Therefore, the nNOS expression observed at four hours may be a feed-forward mechanism of GTN.

Central activation

GTN infusion caused Fos upregulation in the nociresponsive neurons of the superficial laminae I and II in the TNC and upper spinal cord, confirming our previous findings (13). Activation of peripheral afferents transmits the nociceptive signal to second-order neurons and might be involved in the upregulation of Fos observed in our study. Maintenance of persistent hyperalgesia through central sensitisation is another possibility (32) since NO-producing cells are abundant in the superficial laminae (33). Significant increase in cortical NO over a long time after the termination of GTN infusion suggests that NO by itself can activate endogenous NOS to maintain the NO level (25). Supporting this, we found that the non-selective NOS inhibitor L-NAME reduced Fos expression in the TNC after GTN infusion. These results are in accordance with our previous clinical findings that spontaneous migraine was relieved by administration of the non-selective NOS inhibitor L-NMMA (2). In this regard, activation of Fos in our model could be a direct effect of NO on second-order neurons as well as caused by amplified afferent nociceptive signals. This is supported by previous studies that showed an ongoing discharge was found in dorsal horn neurons of the TNC after infusion of GTN and other NO donors (27,3). However, the dose of GTN used in these studies was higher as compared to our dose. Interestingly, such an ongoing discharge was not found in TG neurons receiving input from meningeal nociceptors after GTN infusion (28).

The CGRPergic pathways show intricate relation with migraine pathogenesis in preclinical and clinical settings. In migraineurs, infusion of CGRP caused migraine-like headache (34) and the CGRP receptor antagonists olcegepant (i.v.) (7) and telcagepant (orally) (8) were effective against acute migraine, thus establishing the role of CGRP in migraine. CGRP receptor antagonism in our model did not inhibit Fos activation induced by GTN when given as post-treatment. Although this finding was in complete agreement with our previous clinical findings (35), we further investigated the effect of olcegepant in this model when given as a pre-treatment and inhibition was seen. It is likely that GTN may act deeper in the cascades immediately after its availability in the system. This could explain why post-treatment with olcegepant given 30 minutes after the GTN infusion did not show any inhibitory effect in the human GTN model (35). Moreover, GTN in our model also increased CGRP expression in the TNC which was attenuated by all the three antagonists. GTN-induced neuronal activation may be accompanied by changes in CGRP expression, which may be differentially affected by anti-migraine interventions, depending on whether the site of interest lies in or out of the central nervous system.

Inflammatory mediators have recently been suggested to be involved during GTN-induced delayed headache (36). Our current findings with NK-1 receptor antagonism inhibiting GTN-induced Fos responses support the above speculations. In support of a peripheral mode of action, the dose of NK-1 antagonist used in this study (1 mg/kg) was shown to significantly inhibit neurogenic plasma extravasation in rat dura produced by electrical stimulation of trigeminal nerves (37). The present study, however, lacks direct evidence of neurogenic inflammation in the dura, and therefore the central effects of NK-1 antagonist cannot be ignored as it can easily cross the blood-brain barrier and probably inhibit c-fos by NK-1 receptor blockade in central neurons.

For a decade, animal models of neurogenic inflammation seemed promising as a model for migraine. However, the neurogenic inflammation theory has failed in migraine since eight different drugs antagonising plasma protein extravasation (PPE) proved to be ineffective in migraine (38). Several studies led to the speculation that the PPE component of neurogenic inflammation may not even exist in humans except in rare and extreme situations (39,40). Rodent receptors for tachykinins appear to be more sensitive than human receptors and this could be one of the limitations of this model. The physiological events inducing PPE in rodents are a very well-defined process, but we lack data supporting such events in humans.

Conclusion

CGRP and NOS systems along with NK-1 receptor mechanisms are all involved in GTN-induced neuronal activation in rat. These factors individually and together may explain migraine after GTN infusion. Further animal work may dissect these mechanisms, and they need validation in additional human experimental studies.

Clinical implications

Calcitonin gene-related peptide (CGRP) and nitric oxide synthase (NOS) systems along with neurokinin (NK)-1 receptor mechanisms are involved in glyceryltrinitrate GTN-induced activation.

Pre-treatment with olcegepant might be a better option than post-treatment to see an inhibitory effect of GTN-induced headache.

Footnotes

Funding

This work was supported by the Candy’s Foundation, Lundbeck Foundation, Danish Agency of Science, Technology and Innovation (271-07-0773), Fonden for Neurologisk Forskning and the Augustinus Foundation.

Conflict of interest

None declared.

References

Thomsen

Olesen

. Nitric oxide in primary headaches. Curr Opin Neurol 2001; 14: 315–321.

Lassen

Ashina

Christiansen

. Nitric oxide synthase inhibition in migraine. Lancet 1997; 349: 401–402.

Koulchitsky

Fischer

Messlinger

. Calcitonin gene-related peptide receptor inhibition reduces neuronal activity induced by prolonged increase in nitric oxide in the rat spinal trigeminal nucleus. Cephalalgia 2009; 29: 408–417.

De Col

Koulchitsky

Messlinger

. Nitric oxide synthase inhibition lowers activity of neurons with meningeal input in the rat spinal trigeminal nucleus. Neuroreport 2003; 14: 229–232.

Strecker

Messlinger

. Neuropeptide release in the dura mater encephali in response to nitric oxide – relevance for the development of vascular headaches? [in German]. Schmerz 2003; 17: 179–184.

Dun

Hao

. Protective effects of nitroglycerin-induced preconditioning mediated by calcitonin gene-related peptide in rat small intestine. Eur J Pharmacol 2001; 430: 317–324.

Olesen

Diener

Husstedt

. Calcitonin gene-related peptide receptor antagonist BIBN 4096 BS for the acute treatment of migraine. N Engl J Med 2004; 350: 1104–1110.

Mannix

Fan

. Randomized controlled trial of an oral CGRP receptor antagonist, MK-0974, in acute treatment of migraine. Neurology 2008; 70: 1304–1312.

Petersen

Birk

Doods

. Inhibitory effect of BIBN4096BS on cephalic vasodilatation induced by CGRP or transcranial electrical stimulation in the rat. Br J Pharmacol 2004; 143: 697–704.

10.

Fischer

Koulchitsky

Messlinger

. The nonpeptide calcitonin gene-related peptide receptor antagonist BIBN4096BS lowers the activity of neurons with meningeal input in the rat spinal trigeminal nucleus. J Neurosci 2005; 25: 5877–5883.

11.

Storer

Akerman

Goadsby

. Calcitonin gene-related peptide (CGRP) modulates nociceptive trigeminovascular transmission in the cat. Br J Pharmacol 2004; 142: 1171–1181.

12.

Reuter

Bolay

Jansen-Olesen

. Delayed inflammation in rat meninges: Implications for migraine pathophysiology. Brain 2001; 124: 2490–2502.

13.

Ramachandran

Bhatt

Ploug

. A naturalistic glyceryl trinitrate infusion migraine model in the rat. Cephalalgia 2012; 32: 73–84.

14.

Hundahl

Hannibal

Fahrenkrug

. Neuroglobin expression in the rat suprachiasmatic nucleus: Colocalization, innervation, and response to light. J Comp Neurol 2010; 518: 1556–1569.

15.

Sixt

Messlinger

Fischer

. Calcitonin gene-related peptide receptor antagonist olcegepant acts in the spinal trigeminal nucleus. Brain 2009; 132: 3134–3141.

16.

Sullivan

Giulumian

Pollock

. Functional NOS 1 in the rat mesenteric arterial bed. Am J Physiol Heart Circ Physiol 2002; 283: H658–H663.

17.

Stenberg

Ovlisen

Svendsen

. Effect of local anaesthesia on neuronal c-fos expression in the spinal dorsal horn and hypothalamic paraventricular nucleus after surgery in rats. Basic Clin Pharmacol Toxicol 2005; 96: 381–386.

18.

Chen

Tai

. Propofol reduces nitric oxide biosynthesis in lipopolysaccharide-activated macrophages by downregulating the expression of inducible nitric oxide synthase. Arch Toxicol 2003; 77: 418–423.

19.

Ashina

Bendtsen

Jensen

. Nitric oxide-induced headache in patients with chronic tension-type headache. Brain 2000; 123: 1830–1837.

20.

Graham

Hoffman

Sved

. c-Fos expression in brain in response to hypotension and hypertension in conscious rats. J Auton Nerv Syst 1995; 55: 92–104.

21.

Tassorelli

Joseph

. Systemic nitroglycerin induces Fos immunoreactivity in brainstem and forebrain structures of the rat. Brain Res 1995; 682: 167–181.

22.

Covasala

Stirn

Albrecht

. Calcitonin gene-related peptide receptors in rat trigeminal ganglion do not control spinal trigeminal activity. J Neurophysiol 2012; 108: 431–440.

23.

Oshinsky

Gomonchareonsiri

. Episodic dural stimulation in awake rats: A model for recurrent headache. Headache 2007; 47: 1026–1036.

24.

Knyihar-Csillik

Tajti

Samsam

. Effect of a serotonin agonist (sumatriptan) on the peptidergic innervation of the rat cerebral dura mater and on the expression of c-fos in the caudal trigeminal nucleus in an experimental migraine model. J Neurosci Res 1997; 48: 449–464.

25.

Read

Manning

McNeil

. Effects of sumatriptan on nitric oxide and superoxide balance during glyceryl trinitrate infusion in the rat. Implications for antimigraine mechanisms. Brain Res 1999; 847: 1–8.

26.

Levy

Strassman

. Modulation of dural nociceptor mechanosensitivity by the nitric oxide-cyclic GMP signaling cascade. J Neurophysiol 2004; 92: 766–772.

27.

Jones

Lever

Bingham

. Nitric oxide potentiates response of trigeminal neurones to dural or facial stimulation in the rat. Cephalalgia 2001; 21: 643–655.

28.

Zhang

Kainz

Zhao

. Vascular ERK mediates migraine-related sensitization of meningeal nociceptors. Ann Neurol 2013; 73: 741–750.

29.

Schlechtweg

Röder

Fischer

. Increase in NADPH-diaphorase-positive and neuronal NO synthase immunoreactive neurons in the rat spinal trigeminal nucleus following infusion of a NO donor – evidence for a feed-forward process in NO production involved in trigeminal nociception. Cephalalgia 2009; 29: 566–579.

30.

Knyihar-Csillik

Vecsei

. Effect of a nitric oxide donor on nitroxergic nerve fibers in the rat dura mater. Neurosci Lett 1999; 260: 97–100.

31.

Zochodne

Verge

Cheng

. Nitric oxide synthase activity and expression in experimental diabetic neuropathy. J Neuropathol Exp Neurol 2000; 59: 798–807.

32.

Meller

Gebhart

. Nitric oxide (NO) and nociceptive processing in the spinal cord. Pain 1993; 52: 127–136.

33.

Valtschanoff

Weinberg

Rustioni

. NADPH diaphorase in the spinal cord of rats. J Comp Neurol 1992; 321: 209–222.

34.

Lassen

Haderslev

Jacobsen

. CGRP may play a causative role in migraine. Cephalalgia 2002; 22: 54–61.

35.

Tvedskov

Tfelt-Hansen

Petersen

. CGRP receptor antagonist olcegepant (BIBN4096BS) does not prevent glyceryl trinitrate-induced migraine. Cephalalgia 2010; 30: 1346–1353.

36.

Tfelt-Hansen

Daugaard

Lassen

. Prednisolone reduces nitric oxide-induced migraine. Eur J Neurol 2009; 16: 1106–1111.

37.

Seabrook

Shepheard

Williamson

. L-733,060, a novel tachykinin NK1 receptor antagonist; effects in [Ca2+]i mobilisation, cardiovascular and dural extravasation assays. Eur J Pharmacol 1996; 317: 129–135.

38.

Peroutka

. Neurogenic inflammation and migraine: Implications for the therapeutics. Mol Interv 2005; 5: 304–311.

39.

Weidner

Klede

Rukwied

. Acute effects of substance P and calcitonin gene-related peptide in human skin – a microdialysis study. J Invest Dermatol 2000; 115: 1015–1020.

40.

Petersen

Winge

Brodin

. No release of histamine and substance P in capsaicin-induced neurogenic inflammation in intact human skin in vivo: A microdialysis study. Clin Exp Allergy 1997; 27: 957–965.