NOP receptor agonist attenuates nitroglycerin-induced migraine-like symptoms in mice

Abstract

Migraine is an extraordinarily prevalent and disabling headache disorder that affects one billion people worldwide. Throbbing pain is one of several migraine symptoms including sensitivity to light (photophobia), sometimes to sounds, smell and touch. The basic mechanisms underlying migraine remain inadequately understood, and current treatments (with triptans being the primary standard of care) are not well tolerated by some patients. NOP (Nociceptin OPioid) receptors, the fourth member of the opioid receptor family, are expressed in the brain and periphery with particularly high expression known to be in trigeminal ganglia (TG). The aim of our study was to further explore the involvement of the NOP receptor system in migraine. To this end, we used immunohistochemistry to examine NOP receptor distribution in TG and trigeminal nucleus caudalus (TNC) in mice, including colocalization with specific cellular markers, and used nitroglycerin (NTG) models of migraine to assess the influence of the selective NOP receptor agonist, Ro 64-6198, on NTG-induced pain (sensitivity of paw and head using von Frey filaments) and photophobia in mice. Our immunohistochemical studies with NOP-eGFP knock-in mice indicate that NOP receptors are on the majority of neurons in the TG and are also very highly expressed in the TNC. In addition, Ro 64-6198 can dose dependently block NTG-induced paw and head allodynia, an effect that is blocked by the NOP antagonist, SB-612111. Moreover, Ro 64-6198, can decrease NTG-induced light sensitivity in mice. These results suggest that NOP receptor agonists should be futher explored as treatment for migraine symptoms.

1. Introduction

Migraine is an extraordinarily prevalent and disabling headache disorder that affects one billion people worldwide (Molana et al., 2014). Throbbing pain (lasting 4–72h) is one of several migraine symptoms that also might include nausea, vomiting, as well as sensitivity to light (photophobia), sounds, smell and touch. Migraine can be seasonal or hormone-related and is highly sex dependent (Buse et al., 2013), with roughly 70% of migraineurs being women.

Several rodent assays have been developed to model migraines in humans. In particular, the systemic injection of the nitric oxide donor, nitroglycerin (NTG) induces vasodilation and significant allodynia throughout the body that can be measured with von Frey filaments either in the paw or in the head (Bates et al., 2010; Moye et al., 2019; Pradhan et al., 2014a; Pradhan et al., 2014b). Furthermore, NTG induces migraines in people prone to migraine (Ashina et al., 2013; Iversen et al., 1989). Importantly, NTG-induced pain in rodents can be blocked by specific compounds, such as the most effective human migraine treatment, triptans (Bates et al., 2010; Burstein and Jakubowski, 2004; Pradhan et al., 2014b) and CGRP blockers (Christensen et al., 2019). Also important is that this migraine model produces not only cutaneous allodynia, but also induces light sensitivity (photophobia), another characteristic of migraines in people (Chou and Chen, 2018; Pradhan et al., 2014a; Tang et al., 2018).

The basic mechanisms underlying migraine remain inadequately understood, and current treatments (with triptans being the primary standard of care) are not well tolerated by some patients. Calcitonin Gene Related Peptide (CGRP) is known to be a major contributor to migraine, and although CGRP receptor antagonists are not clinically approved treatments, CGRP antibodies have shown considerable promise and are now available clinically. Migraine pain can also be attenuated by activation of opioid receptors. Mu-opiates are often prescribed for severe migraine pain, but due to opioid side effects, this is not recommended (Tepper, 2012). The delta-opioid receptor has also been extensively examined for treatment of migraine, and several delta-opioid receptor agonists are currently in development for clinical use (Moye et al., 2019; Pradhan et al., 2014b). Both long-acting and short-acting kappa opioid receptor antagonists have also been demonstrated to reduce cephalic pain and plasma CGRP levels in a medication overuse headache model, presumably through inhibition of endogenous dynorphin (Xie et al., 2017).

NOP (Nociceptin OPioid) receptors (originally called ORL1) are the fourth member of the opioid receptor family and are highly expressed in dorsal root ganglia (DRG), spinal cord, and in brain regions involved in pain, drug reward, anxiety, memory, and feeding, as well as others (Ozawa et al., 2015). NOP receptors are Gi coupled and receptor activation stimulates G protein Inwardly Rectifying Potassium (GIRK) channels, thereby universally hyperpolarizing cells and reducing cellular activation (Chiou et al., 2004). Using NOP-eGFP knock-in mice, it has been determined that NOP receptors are particularly highly expressed in pain-processing pathways, as well as in circuitries regulating reward and stress (Ozawa et al, 2015). In situ hybridization studies on nociceptin/orphanin FQ (N/OFQ), the endogenous ligand, has shown a similar distribution (Neal et al., 1999a; Neal et al., 1999b).

Initial results indicated that N/OFQ, was pro-nociceptive as its administration intracerebroventricularly (i.c.v.) into mice decreased hot plate and tail flick latencies (Meunier et al., 1995; Reinscheid et al., 1995). However, N/OFQ is anti-nociceptive when injected intrathecally (i.t.) in both acute and chronic models of pain (Courteix et al., 2004; Tian et al., 1997; Yamamoto et al., 1997). When administered systemically, selective small molecule NOP receptor agonists are not particularly effective for treatment of acute pain in standard tail flick or hot plate testes in rodents (Jenck et al., 2000; Khroyan et al., 2011). Nevertheless, systemic administration of small molecule NOP receptor agonists, which are ineffective in the tail flick, block mechanical allodynia in mice that have chronic neuropathic pain due to spinal nerve ligation surgery (Khroyan et al., 2011). Therefore, the effectiveness of NOP receptor agonists, for treatment of pain, appears to depend upon both the site of drug administration, as well as the state of the animal.

NOP receptor expression is also very high in the trigeminovascular system (TGVS), including the trigeminal ganglia (TG) and trigeminal nucleus caudalis (TNC) (Hou et al., 2003; Xie et al., 1999) (see Figures 1 and 2). The TGVS plays a pivotal role in regulating cerebral blood flow and head pain-processing, and as such is considered to be a primary effector in migraine (Tardiolo et al., 2019). We hypothesized that inhibition of cellular activity, through NOP receptor activation in the TG and TNC will attenuate migraine pain. In this regard, the aim of our study was to validate the NOP receptor system as a target for cephalic pain by determining the NOP receptor distribution in the TG and TNC, and examining the effect of a NOP receptor agonist on NTG-induced pain and light sensitivity.

Figure 1. NOP-eGFP receptor expression in the trigeminal nucleus caudalis.

Representative images show NOP-eGFP expression in the TNC (Sp5C, spinal trigeminal nucleus caudalis). CGRP was used as a maker to visualize spinal trigeminal tract (Sp5). Scale bar, 250 μm.

Figure 2. Neuroanatomical analysis of TG neurons using NOP-eGFP mice.

To characterize the trigeminal neurons expressing NOP-eGFP, sections were incubated with anti-GFP antibody together with cellular markers. A, The NOP-eGFP containing trigeminal neurons were quantified by determining the percentage of eGFP-positive cells (green) compared with the total number of trigeminal sensory neurons (n = 5264 neurons). Nuclei were stained with DAPI (blue). The total number of trigeminal sensory neurons was determined by counting the total number of DAPI-stained cells and excluding those from satellite cells. Tissue sections were also co-stained with anti-CGRP and -NF200 antibodies (B), anti-nNOS and -CGRP (C), biotinylated IB4 together with streptavidin conjugated with Alexa fluor (D), anti-Ret and -NF200 (E) and anti-TrkC and -NF200 (F). Yellow arrowheads indicate the trigeminal neurons where co-staining occurs. Scale bars, 100 μm.

2. Material and Methods

2.1. Animals

All behavioral studies were conducted in C57BL6/J wild-type mice (male and female) from the Jackson Laboratory (Jackson Laboratories, Bar Harbor, ME, USA), while histology used Oprl1^egfp/egfp (NOP-eGFP) mice on a C57BL6/J background (as previously described in (Ozawa et al., 2015), weighing 20–30 g. Animals were housed in a 12h light–dark cycle, group-housed (three to five mice per cages) and provided with food and water ad libitum. All experiments were carried out in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. All methods used were pre-approved by the Institutional Animal Care and Use Committees at Florida Atlantic University and the University of Chicago at Illinois.

2.2. Immunohistochemistry

Immunohistochemistry was performed by following the procedure described previously (Ozawa et al., 2018; Ozawa et al., 2015). Briefly, 6-week old mice were anesthetized with overdose of isoflurane for 1 min and transcardially perfused in 4% paraformaldehyde in PBS. The brain and TG were then dissected and cryoprotected in 30% sucrose in PBS and frozen in OCT (Sakura Finetek, INC., Torrence, CA, USA). Tissue sections (40 μm for brain; and 10 μm for TG) were prepared by using a cryostat (Leica Biosystems, Buffalo Grove, IL, USA) and blocked with PBS containing 5% normal donkey serum and 0.3% Triton X-100 for 1 h at room temperature and then incubated with primary antibodies at 4°C overnight. For the incubation with the chicken anti-GFP antibody, the incubation with primary antibodies was performed at 37°C for 2 h. After washing primary antibodies 3 times for 10 min with PBS containing 1% normal donkey serum and 0.3% Triton X-100, tissue sections were incubated with secondary antibodies for 2 h at room temperature. Sections were then washed with PBS 3 times for 10 min, and mounted with Fluoromount (Southern Biotech). Images were collected under an A1R confocal microscope and NIS-Elements Imaging software (Nikon, Melville, NY, USA) and analyzed by using Fiji (National Institute of Health). Images from 4–6 sections per animals were collected for analysis with each cellular marker. Cell counts and percentage obtained for each section were then averaged. To quantify the trigeminal sensory neurons expressing NOP-eGFP, the number of eGFP-positive cells was compared with the total number of sensory neurons. The total number of trigeminal sensory neurons was determined by counting the total number of DAPI-stained cells and excluding those from satellite cells. We used the following primary antibodies: rabbit anti-GFP (1:1000, Life Technologies); chicken anti-GFP (1:2500, Abcam); rabbit anti-nNOS (1:1000, Thermo Fisher Scientific); sheep anti-CGRP (1:1500, Abcam); mouse anti-NF200 (1:20,000, Sigma-Aldrich); goat anti-TrkC (1:500, R&D Systems); goat anti-Ret (1:60, R&D Systems). For identifying IB4-binding cells, biotinylated IB4 (1:500, Sigma-Aldrich) and fluorophore conjugated streptavidin (1:2000, Thermo Fisher Scientific) were used.

2.3. Drugs

Nitroglycerin was prepared from a stock solution of 5.0 mg/ml NTG in 30% alcohol, 30% propylene glycol and water (American Regent, Shirely, NY, USA). NTG was freshly (before each experiment) diluted in 0.9% saline and used at a dose of 10 mg/kg. The final vehicle solution used for NTG in these experiments was 6% alcohol, 6% propylene glycol in 0.9% saline. NTG or vehicle solution were administered intraperitoneally (i.p.) as a 10 ml/kg volume. Ro 64-6198 (NOP receptor agonist) and SB-612111 (NOP receptor antagonist) were provided by the National Institute on Drug Abuse Drug Supply Program. SB-612111 and Ro 64-6198 were suspended in a vehicle containing 5% DMSO and 95% hydroxypropyl cellulose (0.5% HPC in distilled water) and were administered in a 5 ml/kg volume injection (i.p.).

2.4. Periorbital mechanical hypersensitivity test

Different groups of animals were used for periorbital and hind paw experiment. Cephalic/periorbital responses were determined as described in (Ben Aissa et al., 2018). Mice were counter-balanced into groups following baseline assessment. Mice were tested in a behavior room, separate from the vivarium, with low light (~35–50 lux) and low-noise conditions, between 09:00 and 16:00. Animals were habituated to the testing racks for 30 minutes for 2 days before the initial test day, and on the test day for 20 min before the baseline measurement. For cephalic measures, mice were tested in 4 oz paper cups that were placed in the plexiglass box. The periorbital region caudal to the eyes and near the midline was tested. Testing of mechanical thresholds to punctate mechanical stimuli was tested using the up-and-down method. The selected region of interest was stimulated using a series of manual von Frey hair filaments (bending force ranging from 0.008 g to 2 g). A response of the head was defined as shaking, repeated pawing, or cowering away from the filament. The first filament used was 0.4 g. If there was no response a heavier filament (up) was used, and if there was a response a lighter filament (down) was tested. The up-down pattern persisted for 4 filaments after the first response. On the test day mice were habituated to the rack/cup, baseline mechanical responses were determined, and mice were weighed and injected with NTG or vehicle and placed back on the testing rack. One hour and fifteen minutes later they were injected with NOP agonist or vehicle (VEH), returned to the rack and tested 45 min later (2h post-NTG/VEH).

2.5. Paw withdrawal threshold test

Sensitivity of paws was measured with the von Frey filaments at 1 h after NTG injection as described previously (Pradhan et al., 2014a; Pradhan et al., 2014b). For the pain sensitivity test, mice were placed on an elevated surface made of wired mesh and constrained by cylindrical Plexiglas containers. Prior to injections, mice were habituated for 30 minutes on the mesh surface. Subsequently, NTG or VEH injections were administered and after a 30 min period NOP agonist, Ro 64-6198, was injected. 30 min after drug injection, von Frey filaments were used to stimulate the plantar surface of the mouse’s hind paw. To determine whether NOP antagonists block the effect of NOP agonist, SB-612111 was injected 15 min before Ro 64-6198 administration. The seven calibrated plastic monofilaments delivered a target force of 0.04, 0.07, 0.16, 0.40, 0.6, 1.0, or 2.0 grams. A positive response was described as rapid pawl withdrawal, licking or shaking of the paw, during stimulation or immediately after the removal of the filament. The first filament applied was the middle force instrument (0.40 g). The hind paw was stimulated five times, for no longer than 2 seconds and in resting interims of 10 s. A positive response of three or more of the five stimulations was considered a positive response for that filament. In the presence of a response, a smaller force filament (down) was used in the subsequent test; while the absence of a response led to a heavier force filament (up). This process was repeated, for a total of four responses. The paw withdrawal threshold was calculated on the basis of the four responses. This up-and-down method and threshold calculation was first described by Chaplan et al. (1994).

2.6. Light-aversive behavior

Photophobia (light-aversive behavior) is an assay that can be adapted for rodents using a light/dark box (Vuralli et al., 2019). In light-aversion behavior, a Med Associates Activity Monitor box with Dark Box Insert was used. The testing chambers were placed in a ventilated and Sound Attenuating Med Associates Cubicle (56 cm wide × 38 cm deep × 36 cm height). For the photophobia assay, NOP agonist or VEH were administered 20 min before NTG or VEH injection. Then, 10 min later, each mouse was placed in the box and allowed to freely explore both chambers for 20 minutes. The mice were tracked using the Activity Monitor v6.02 software (Med Associates). The eight Plexiglas chambers consisted of three axes infrared arrays fixed to the four sides, used for open-field activity. The infrared beams allowed for tracking the animal’s movements between chambers and time spent inside. The chambers were equivalent in dimensions, one side was covered with black Plexiglas to prevent light entrance, while the other was lit by a 1000 lux lighting unit (surgical lamp, Fisher Scientific, USA). The chambers were separated by a black Plexiglas insert, and a cutout made both sides available to the mouse.

2.7. Elevated Plus Maze

Elevated plus maze (EPM) is a method used to assess anxiogenic or anxiolytic responses. EPM test was conducted as previously reported (Brunori et al., 2018). The EPM is a Plexiglas platform consisting of a surface in the form of a ‘plus’ sign with two open arms (35.6 × 7.6 cm) and two closed arms (35.6 × 7.6 × 20.3 cm), connected by a square central platform. The closed arms were only vertically concealed by black Plexiglas (20.3 cm), to ensure similar light intensities, while the open arms are not concealed by Plexiglas. The platform was at a raised height of 50 cm above ground and tested with room lighting of 500 lux. For EPM test the same treatment schedule and drug doses were used as described above for photophobia assays. The mouse was placed on the center platform in the direction of a closed arm, and was free to roam for 5 minutes, as first described by Lister (1987). An entry of an arm was considered when the animal placed all four paws into an arm, as determined by the lines enclosing the center square. The number of entries into the open arms and the time spent in the open arms were measured and their respective percentages was calculated. An increase of the time spent in the open arms was considered an anxiolytic response, with consideration of the number of entries in the open arms. Anxiogenic responses were determined by the opposite behavior, with consideration of entries into the closed arms. The number of closed arm entries was also recorded as an indicator of the motor activity of tested animals.

2.8. Statistical analysis

GraphPad Prism software (GraphPad Software, San Diego, California, USA) was used to perform statistical analysis and to generate graphs. Data (mean ± standard error of the mean (SEM)) were analyzed by paired t test or by means of a two-way analysis of variance (ANOVA). Where appropriate ANOVAs were followed by the Newman-Keuls’s post-hoc test, used to compare differences between groups.

3. Results

3.1. Various types of trigeminal neurons express NOP receptors

Taking advantage of NOP-eGFP mice, we determined the distribution of the NOP receptor in TNC (Figure 1) and characterized neuronal subsets expressing NOP receptors in the TG (Table 2 and Figure 2) to obtain a better understanding of the functional properties of the NOP receptor system in the transmission of cephalic pain. Strong immunostaining of NOP-eGFP was observed throughout the TNC, but was particularly prevalent, colocalized with CGRP, in the spinal trigeminal tract (sp5), where peptidergic C-fibers in the TG send their projections (Figure 1).

Table 2.

Neurochemistry of NOP-expressing trigeminal neurons.

Neuronal Marker	% of neurons co-expressing NOP receptors	Total number of neurons counted
CGRP+ NF200−	33.1±1.9	279
CGRP+ NF200+	5.0±0.7	523
nNOS+ CGRP+	28.0±7.5	216
IB4+	10.8±4.1	860
NF200+	85.4±2.5	2347

Percentage of NOP receptor-expressing neurons is calculated as (%=[Number of NOP-eGFP positive neurons that coexpress the marker / Total number of neurons expressing the cellular marker] × 100). Data are represented as mean ± SEM (n=4 mice).

Approximately 72% of all TG neurons express NOP-eGFP (Figure 2A). This is a slightly higher percentage compared to the dorsal root ganglion, in NOP-eGFP mice (Ozawa et al., 2015). Next, using cellular biomarkers for the TG, we investigated the potential functional properties of NOP-eGFP-positive neurons. There are the small unmyelinated NOP-eGFP positive neurons expressing CGRP (33% of CGRP+ NF200- neurons co-express NOP receptors, Figure 2B) that belong to C-peptidergic nociceptors. Interestingly, about 28% of nitric oxide synthase (nNOS)-positive neurons, which always co-stained with CGRP, also express NOP-eGFP receptors (Figure 2C). Because we identified negligible NOP-eGFP expression in the myelinated trigeminal neurons that express CGRP, we assume that these nNOS+CGRP+ neurons expressing NOP-eGFP are unmyelinated C-fibers (Figure 2B and Table 2).

Interestingly, the majority of the NOP-eGFP+ cells (88%, Figure 2B) are co-stained with neurofilament 200 (NF200), a marker for neurons with myelinated axons (A-fibers), indicating that NOP receptors are highly expressed in trigeminal mechanoreceptors (Ret+ NF200+) and proprioceptors (TrkC+ NF200+). While most of the TrkC- and myelinated-Ret-positive neurons co-express NOP-eGFP (Figures 2E and F), the immunoreactivity of NOP-eGFP receptors are rarely found in unmyelinated Ret-positive neurons. These small unmyelinated Ret-positive neurons are presumably non-peptidergic that are linked to the perception of thermal-noxious stimuli, and are mostly IB4-positive (10% of IB4+ neurons are co-stained with NOP-eGFP, Figure 2C). In addition, similar to DRG neurons, there is very little NOP-eGFP receptor expression in the myelinated trigeminal neurons that express CGRP (Table 2). Altogether, our immunohistochemical studies indicate the NOP receptor system might regulate various types of supraspinal somatosensations, including cephalic pain.

3.2. Mechanical allodynia induced by acute NTG is reversed by NOP agonist, Ro 64-6198

It was previously shown that an acute systemic administration of NTG produce mechanical allodynia in mice (Bates et al., 2010; Pradhan et al., 2014b). Using the NTG model of migraine we found that selective and high affinity NOP receptor agonist, Ro 64-6198, fully blocks allodynia, as measured by von Frey filament stimulation of the periorbital region and hind paw in mice. We initially tested the effects or Ro 64-6198 in NTG-induced cephalic allodynia. Prior to drug administration, mice had comparable baseline periorbital responses as determined by two-way RM ANOVA (Figure 3A, Baseline). Two-way ANOVA conducted for von Frey hairs stimulation of the head following combined treatments of NTG and Ro 64-6198 revealed significant “pretreatment (NTG or VEH)” x “treatment” (Ro 64-6198 or VEH) interaction F [5,42] = 6.594, p = 0.0001]. NTG (10 mg/kg) produced a significant decrease in cephalic mechanical responses (***p<0.001, as compared to VEH-VEH, Figure 3A, Post-treatment). This severe allodynia was reversed by the administration of Ro 64-6198, and both 1 and 3 mg/kg showed similar, maximal responses (###p<0.001 as compared to NTG-vehicle).

Figure 3. The NOP receptor agonist Ro 64-6198 attenuates NTG-induced mechanical allodynia, an effect blocked by the NOP antagonist SB-612111.

Effect of Ro 64-6198 (0.3 to 3.0 mg/kg) on NTG-induced mechanical hypersensitivity in (A) the periorbital region and (B) the paw. (C) SB-612111 (10 mg/kg) effect on Ro 64-6198 in mice with NTG-induced pain. NTG-induced hypersensitivity in the head and the paw was measured by von Frey filaments. (A) NTG (10 mg/kg) produced marked mechanical response (***p<0.001). Ro 64-6198 was effective in blocking NTG-induced allodynia at both 1.0 and 3.0 mg/kg doses (###p<0.001 for both, n=6–12 mice per dose). (B) NTG (10 mg/kg) also produced marked mechanical response in the paw (**p<0.01), with Ro 64-6198 being effective in reversing allodynia at 1.0 mg/kg dose (### p<0.001). (C) In a separate experiment, Ro 64-6198 blocked mechanical allodynia induced by NTG (^^ p<0.01). This effect was reversed by the antagonist SB-612111 (## p<0.01, two-way ANOVA followed by Newman-Keuls’s post hoc test). Experiments shown in B and C were carried out using n=10–12 mice (4–6 males and 5–6 females) per group and n=8 per group, respectively.

Ro 64-6198 was also effective for treatment of NTG-induced allodynia in the paw. Two-way ANOVA conducted for von Frey hairs stimulation of the hind paw following combined treatments of NTG and Ro 64-6198 revealed significant “pretreatment (NTG or VEH)” x “treatment” (Ro 64-6198 or VEH) interaction [F(2,59) = 3.76, p = 0.028], suggesting altered allodynic responses to the treatments. Specifically, treatments with both 0.3 and 1 mg/kg Ro 64-6198 significantly reversed mechanical allodynia induced by NTG (* p<0.05 and ** p<0.001, respectively) in a combined group of male and female mice (Figure 3B).

We then investigated NTG treated mice to determine whether the anti-allodynic effect of Ro 64-6198 was blocked by pretreatment with the NOP receptor antagonist SB-612111 (10 mg/kg). We found significant “pretreatment (SB-612111 or VEH)” x “treatment” (Ro 64-6198 or VEH) interaction [F(1,28) = 9.93, p = 0.004], suggesting that the NOP antagonist could reverse the anti-allodynic response induced by 1 mg/kg Ro 64-6198. On post hoc analysis, Ro 64-6198 produced a significant block of mechanical hypersensitivity induced by administration of NTG (^^ p < 0.01), an effect that was reversed by pretreatment with SB-612111 (## p < 0.01, Figure 3C).

3.3. NTG-induced light-aversive behavior is reversed by Ro-64-6198 but not sumatriptan in male and female mice

NTG (10 mg/kg) leds animals to spend significantly less time in the light chamber of a light/dark box (paired t test; t (6) = 2.948, p = 0.0257), as shown in Figure 4A. Two-way ANOVA conducted for light-aversive behavior following combined treatments of Ro 64-6198 and NTG revealed significant “pretreatment (Ro 64-6198 or VEH)” x “treatment” (NTG or VEH) interaction both in male [F(1,32) = 12.44, p = 0.001] (Figure 4B) and female [F(1,29) = 6.33, p = 0.018] (Figure 4C) mice. Newman-Keuls’s post hoc test indicated significant NTG-induced aversion to light in male and female mice (* p < 0.05 for both sexes), whereas treatment with 0.3 mg/kg Ro 64-6198 significantly attenuated the aversion to light, as demonstrated by increased percentage of time spent in the light compartment (## p < 0.01 and # p <0.05 for male and female mice, respectively).

Figure 4. NTG-induced light aversion is blocked by Ro 64-6198.

Effects of Ro 64-6198 and sumatriptan on NTG-induced light sensitivity in the light/dark box in male and female mice. (A) NTG-treated animals showed decrease in the percent (%) of time spent in light chamber (p < 0.05, n = 7 per group). In a separate experiment, Ro 64-6198 (0.3 mg/kg, i.p.) blocked NTG-induced light aversion (*p <0.05) in (B) male and (C) female mice (n = 8–10 per group, ## p < 0.01 and # p < 0.05, respectively, when compared to NTG-treated mice; two-way ANOVA followed by Newman-Keuls’s post hoc test). Sumatriptan (0.6 mg/kg, i.p.) did not reverse NTG-induced light aversion in male (D) and female (E) mice (main NTG effect: +p<0.05 for both, n = 7 per group).

In our experiments, sumatriptan (0.6 mg/kg) failed to reverse NTG-induced light-aversive behavior in male (Figure 4D) and female (Figure 4E) mice. Two-way ANOVAs conducted separately for each sex revealed significant main effect of NTG treatment (males: [F(1,24) = 4.96, p = 0.03], females: [F(1,24) = 6.20, p = 0.02]) accompanied by not significant sumatriptan pretreatment effect (males: [F(1,24) = 0.79, p = 0.38], females: [F(1,24) = 0.26, p = 0.61]) and interaction of the two factors( [F(1,24) = 0.02, p = 0.90] for males, [F(1,24) = 1.60, p = 0. 22] for females).

3.4. Ro-64-6198 induces anxiolytic activity in male but not female mice but is not anxiolytic in the presence of NTG

There was concern that the protective effect of Ro 64-6198 to light sensitivity observed in the light dark box, was rather due to anxiolytic activity of the NOP agonist, as previously demonstrated in the elevated plus maze and other anxiety models (Jenck et al., 2000; Wichmann et al., 2000). In this experiment, we found a considerable difference between male and female mice. In male mice, NTG had a clear anxiogenic effect. Two-way ANOVA conducted for the variable “percent of time spent on the open arms” indicated significant pretreatment effect of Ro 64-6198 [F(1,24) = 6.56, p = 0.02], significant main effect of NTG [F(1,24) = 9.08, p = 0.006] and lack of interaction between the two factors [F(1,24) = 2.44, p = 0.13; Figure 5A]. These results indicated that, in male mice, NTG was anxiogenic and Ro 64-6198 was effective as an anxiolytic when tested alone, but not in the presence of NTG.

Figure 5. Treatment of NTG does not increase anxiety-like behavior.

Anxiety-like behavior was measured by the elevated plus maze test in male (A and B) and female mice (C and D). The percent (%) of time spent in the open arms and % of open-arm entries was calculated (mean ± standard error of the mean (SEM)) for each group. (A) % of time spent in the open arms was increased by Ro 64-6198 0.3 mg/kg (main Ro 64-6198 effect: ⁺ p < 0.05) and decreased by NTG (10 mg/kg) in male mice (main NTG effect: ⁺⁺ p < 0.01). (B) % open arm entries was decreased by NTG (10 mg/kg) in male mice (main NTG effect: ⁺ p < 0.05, n = 7 per group). (C, D) Anxiety-like response was not altered by Ro 64-6198 (0.3 mg/kg) in female mice (n = 8–9 per group).

Open arm entries were also significantly reduced [F(1,24) = 4.47, p = 0.045] by NTG treatment, but in this case Ro 64-6198 had no significant effect [F(1,24) = 0.96, p = 0.337] or interaction [F(1,24) = 0.44, p = 0.51, Figure 5B]. Closed arm entries did not change following NTG treatment [F(1,24) = 0.31, p = 0.58], however were increased following Ro 64-6198 administration [F(1,24) = 6.71, p = 0.016], suggesting that increased locomotor activity in the presence of the NOP agonist (Table 2).

We observed female mice spending considerably more time in the open space than male mice, consistent with other studies (Kokras and Dalla, 2014). Furthermore, female mice were not sensitive to Ro 64-6198 or NTG administrations with no effect on open arm time, open arm entries or closed are entries. (% open arm time: main pretreatment effect [F(1,30) = 0.49, n.s.], main treatment effect [F(1,30 ) = 1.89, n.s], interaction [F(1,30 ) = 0.88, n.s], Figure 5C; % open arm entries: pretreatment [F(1,30) = 0.37,n.s.], treatment [F(1,30) = 0.26,n.s.], interaction [F(1,30) = 0.11,n.s.], Figure 5D;closed arm entries: pretreatment [F(1,30) = 0.39,n.s.], treatment [F(1,30) = 0.047, n.s.], interaction [F(1,30) = 0.30, n.s.], Table 2). Overall, these results indicate that Ro 64-6198 has anxiolytic activity only in male mice in the absence of NTG and is not effective as an anxiolytic in the presence of NTG at the dose (0.3 mg/kg) that it reversed photophobia.

4. Discussion

The head pain associated with a migraine attack, is thought to be due to activation of the trigeminovascular system. A large number of nociceptive nerve fibers that originate in the trigeminal ganglion innervate the pial, arachnoid, and dural blood vessels and release several vasoactive neuropeptides, including CGRP, substance P, neurokinin A, and pituitary adenylate cyclase-activating peptide (PACAP) upon stimulation, causing vasodilation of dural and pial vessels. Nociceptive information from these craniovascular structures is relayed through the TNC, and via ascending connections to other areas of the brain stem and further cortical regions involved in the processing of pain and other sensory information (Goadsby et al., 2017).

The first line treatment for migraine are triptans, which work by activating 5-HT1B and 5-HT1D receptors leading to vasoconstriction. Triptans also activate 5-HT1D receptors in TNC, inhibiting the activation of nociceptive receptors and the release of vasoactive peptides including substance P and CGRP (Ahn and Basbaum, 2005). CGRP is known to be an important contributor to migraine pain, and anti-CGRP monoclonal antibodies are now available for treatment of chronic migraine. Despite these advances, a large number of migraineurs are unresponsive or only partially responsive to current treatments.

There is reason to believe that the NOP receptor system could be a viable target for treatment of migraine or other cephalic pain. The level of the endogenous ligand, N/OFQ, is greatly reduced in the plasma of migraineurs compared to controls and it was further reduced during the first three hours of an attack (Ertsey et al., 2005). Furthermore, N/OFQ blocks smooth muscle contractions in a variety of tissues (Berzetei-Gurske et al., 1996; Calo et al., 1997) including inhibition of electrically-induced dilation of the middle meningeal artery in the rat (Bartsch et al., 2002). In addition, as we have demonstrated here, NOP receptors are on the vast majority of TG neurons (~72%) and are highly expressed in the TNC. Since NOP receptor activation uniformly hyperpolarizes neurons (Chiou et al., 2004; Connor and Christie, 1999), receptor activation would most likely reduce the pain signal.

Our neuroanatomical analysis demonstrated that NOP-eGFP receptors are highly expressed in the TG with immunoreactivity found on both myelinated (NF200+) and unmyelinated (NF200−) TG neurons. NOP-eGFP receptors are on one third of the peptidergic trigeminal neurons (CGRP+NF200-). Previous electrophysiological studies have demonstrated that N/OFQ preferentially inhibits calcium channel currents in a subpopulation of small nociceptive trigeminal neurons but not larger nociceptive trigeminal nociceptors (Borgland et al., 2001). These small trigeminal neurons are also sensitive to mu receptor agonists. Nociceptive primary afferents in the TG are generally classified into two different neuronal groups that either express TRPV1 (vanilloid-sensitive transient receptor potential channel) or bind to IB4 (Price and Flores, 2007; Price et al., 2005; Vedder et al., 1993). The TRPV1 positive neurons are the peptidergic neurons in which CGRP is the most prominent and have an important role in the pathogenesis of cephalic pain. As IB4-binding trigeminal neurons expressing NOP-eGFP are rarely observed in our analysis, the electrophysiologically identified N/OFQ sensitive small trigeminal neurons are probably CGRP-expressing C-nociceptors. In addition, we also observed immunoreactivity of NOP-eGFP in nNOS-positive neurons that also co-express CGRP. These peptidergic nociceptors tend to innervate intracranial blood vessels rather than facial skin (O’Connor and van der Kooy, 1988). In the trigeminovascular system, N/OFQ inhibits neurogenic dural vasodilatation in rats (Bartsch et al., 2002), and also the activation of precontracted vessels due to an induced release of neurotransmitters (Meunier, 1997). These data and our findings support the idea that the NOP receptor system plays a role in modulating neurotransmitter release such as CGRP and nitric oxide in the intraganglionic signaling system to regulate the extent of vessel dilation and contraction under a cephalic pain condition. However, in human cerebral vessels, which are another vasodilation contributor in the pathophysiology of migraine, N/OFQ was reported not have any effect on inducing either dilation or contraction (Hou et al., 2003). In this case, the NOP receptor system might play a role in modulating neurotransmitter release that regulates nociception rather than vessel dilation to block cephalic pain.

NOP-eGFP immunoreactivity was found on 85% of myelinated neurons in the TG, which is considerably higher than what we observed in the DRG (~58%, (Ozawa et al., 2015)). In the TG, a large number of myelinated NOP-eGFP+ neurons are co-stained with TrkC and Ret. TrkC is required mostly for various types of cutaneous mechanosensory neurons and for some proprioceptive neurons (Funfschilling et al., 2004). As myelinated RET+ neurons are known to partially coexpress TrkB but not TrkA or TrkC (Luo et al., 2009), these neurons are mechanoreceptors and not nociceptors nor proprioceptors. These findings provide a cellular basis for the functional properties of NOP receptors in trigeminal primary afferent neurons. Further molecular characterization with a combination of cellular markers will be performed to fully resolve the trigeminal somatosensories influenced by the NOP receptor system.

Based upon extensive anatomical and behavioral experimentation, NOP receptors activation can be considered as potential target for pain, per se. Although, in rodents, there is a dichotomy in the effect on nociception when N/OFQ is administered i.c.v. (Meunier et al., 1995; Reinscheid et al., 1995) or i.t. (Courteix et al., 2004; Yamamoto et al., 1997), and systemically administered selective NOP receptor agonists or antagonists are not effective in most acute pain models (Toll et al., 2009), this is different in chronic pain, in which systemic administration of NOP receptor agonists block mechanical allodynia in a variety of neuropathic pain models (Khroyan et al., 2011). In fact, in humans, NOP receptor agonists may be effective for treating both acute and chronic pain, as compounds such as Ro 64-6198 have been demonstrated to have potent NOP receptor mediated antinociceptive activity in non-human primates (Ko et al., 2009). Currently, NOP/mu agonists are farthest along. The NOP/mu partial agonist AT-121 has potent antinociceptive activity in rhesus monkeys, with no apparent adverse side effects (Ding et al., 2018). Moreover, in animal models, the NOP/mu full agonist cebranopadol is particularly effective in treating chronic pain (Tzschentke et al., 2017), and this compound is now in Phase III clinical trials for acute pain. It is very important to note that NOP receptor agonists, including the endogenous ligand N/OFQ and Ro 64-6198, appear to be neither rewarding nor aversive in conditioned place preference (CPP) assays (Devine et al., 1996; Le Pen et al., 2002).

Regardless of whether NOP receptor agonists are effective for treatment of peripheral or neuropathic pain, our results described herein suggest that they might be most relevant for treatment of headache disorders. Consistent with the immunohistochemistry discussed above, we demonstrated that the selective and high affinity NOP agonist, Ro 64-6198, can block NTG-induced mechanical allodynia, as measured by von Frey filament stimulation of both periorbital region and paw in mice, at doses that do not have antinociceptive activity in the tail flick assay in mice (Jenck et al., 2000). In addition, the anti-allodynic action is blocked by the selective NOP receptor antagonist, SB-612111.

It is also well known that migraine induces sensitivity to light (photophobia) in humans. Yet, the mechanisms by which neural circuits regulate migraine symptoms, including photophobia, through the NOP receptor system have not been explored. Our findings confirmed that NTG induces animals to spend significantly less time in the light chamber of a light/dark box (as previously published (Mason et al., 2017; Tang et al., 2018). In this experiment, Ro 64-6198 decreased the NTG-induced light sensitivity at a low dose (0.3 mg/kg). This experiment is somewhat complicated since NOP receptor agonists are known to be anxiolytic (Jenck et al., 1997; Jenck et al., 2000), and an anxiolytic might be expected to reduce the time in the dark side of the light/dark box. Although we found Ro 64-6198 to be anxiolytic in male but not female mice in the elevated plus maze, it was not effective in either male or female mice as an anxiolytic in the presence of NTG, at the dose that it reversed the photophobia in both sexes (0.3 mg/kg). Therefore, we believe that the photophobia experiment can give another indication of anti-migraine activity of Ro 64-6198. Furthermore, treatment with NTG produced weak anxiogenic-like activity in the elevated plus maze assay in male but not female mice. This weak anxiogenic effect, compared to the strong reliable signal of NTG-induced aversion to light observed in both sexes in the light dark box suggests that the photophobic and anxious components of the NTG-induced migraine model can be experimentally separated and detected.

Most opioid analgesics in common clinical use for pain primarily target the mu-opioid receptor. However, mu-opioid receptor agonists have relatively poor efficacy as analgesics for migraine headache and they are not recommended for the treatment of migraine (Tepper, 2012). It was reported that in migraine patients, chronic opioids upregulate peripheral expression of CGRP in primary afferent neurons, to precipitate bad clinical outcomes (Tepper, 2012), and transform episodic migraine to daily headache and chronic migraine (Bigal and Lipton, 2009). Interestingly, delta-opioid receptor agonists appear to be promising for migraine treatment. In particular, both SNC80 and ARM390 effectively reversed hyperalgesia induced by NTG in both acute and chronic migraine models in mice (Pradhan et al., 2014b), although some selective delta receptor agonists are known to cause seizures (Haffmans and Dzoljic, 1983). Delta-opioid receptor agonists are currently in development for clinical use. As discussed above, since migraine may be partially induced by a reduction in levels of circulating N/OFQ and the addition of an exogenous agonist appears to attenuate migraine symptoms, including pain and photophobia, both by blocking vasodilation and by blocking the pain signal through the trigeminal region, there is good reason to believe that the NOP receptor could represent an equal or even better target for cephalic pain. Finally, since Ro 64-6198 blocks both NTG-induced pain and photophobia at doses that are ineffective in increasing tail flick latency, NOP agonists might be most relevant for treatment of headache versus other types of pain states.

Table 1.

List of Reagents

Reagent	Source	Identifier
Antibodies
Chicken anti-GFP	Abcam	Cat# ab13970; RRID: AB_300798
Rabbit anti-GFP	Life Technologies	Cat# A11122; RRID: AB_221569
Rabbit anti-nNOS	Thermo Fisher Scientific	Cat# 61–7000; RRID: AB_2313734
Sheep anti-CGRP	Abcam	Cat# ab22560; RRID: AB_725809
Mouse anti-NF200	Sigma-Aldrich	Cat# N0142; RRID: AB_477257
Goat anti-TrkC	R&D Systems	Cat#AF1404; RRID: AB_2155412
Goat anti-Ret	R&D Systems	Cat# AF482; RRID: AB_2301030
Donkey anti-rabbit IgG	Thermo Fisher Scientific	Cat# A31572; RRID: AB_162543
Alexa Fluor 555
Donkey anti-sheep IgG	Thermo Fisher Scientific	Cat# A21436; RRID:
Alexa Fluor 555		AB_10376163
Donkey anti-goat IgG	Thermo Fisher Scientific	Cat# A21432; RRID: AB_2535853
Alexa Fluor 555
Donkey anti-rabbit IgG	Jackson ImmunoResearch	Cat# 711–545–152; RRID:
Alexa Fluor 488	Laboratories	AB_2313584
Donkey anti-chicken	Jackson ImmunoResearch	Cat# 703–545–155; RRID:
IgY Alexa Fluor 488	Laboratories	AB_2340375
Donkey anti-mouse IgG	Jackson ImmunoResearch	Cat# 715–295–15; RRID:
Rhodamine RedTM-X	Laboratories	AB_2340832
Donkey anti-mouse IgG	Jackson ImmunoResearch	Cat# 715–605–15;
Alexa Fluor 647	Laboratories	RRID: AB_2340863
Donkey anti-sheep IgG	Jackson ImmunoResearch	Cat# 713–605–147; RRID:
Alexa Fluor 647	Laboratories	AB_2340751
Chemicals and Recombinant Proteins
IB4 biotinylated	Sigma-Aldrich	Cat# L2140
Streptavidin conjugated	Thermo Fisher Scientific	Cat# S21381
with Alexa Fluor 555
Streptavidin conjugated	Thermo Fisher Scientific	Cat# S21374
with Alexa Fluor 647

Table 3.

Effect of acute treatment of studied compounds on the closed arm entries (mean ± SEM) in EPM test in male and female mice.

Drug	Male	Female
vehicle + vehicle	14.42 ± 1.62	16.40 ± 1.33
vehicle + 10 mg/kg NTG	16.29 ± 1.27	15.25 ± 1.28
0.3 mg/kg Ro-64-6198 + vehicle	19.43 ± 1.36	14.62 ± 1.27
0.3 mg/kg Ro-64-6198 + 10 mg/kg NTG	19.29 ± 1.86	15.12 ± 2.03

Mice (7–8 mice/condition) were administered with Ro-64-6198 (0.3 mg/kg), or vehicle 20 min prior to NTG (10 mg/kg) or vehicle and 10 min later the EPM test was conducted.

Highlights.

• NOP receptors are highly expressed in the trigeminal ganglia and trigeminal nucleus caudalis

• The selective NOP receptor agonist Ro 64-6198 can block nitroglycerin-induced paw and head sensitivity

• Ro 64-6198 can also block nitroglycerin-induced light sensitivity (photophobia)

• NOP agonists might be useful treatments for migraine pain and other symptoms