Wistar rats of both sexes from the Animal House of the University of Eastern Finland were used in this study. P9–P13 rats were used for trigeminal cell culture preparation and mast cell degranulation assay, whereas adult (P35–P36) rats were used for electrophysiology and immunohistochemistry.

Preparation of male rat (P35–P36) hemiskulls was performed as published earlier (23, 24). Briefly, after decapitation, the skin and muscles were removed from the skull, which was dissected mid-sagitally, and both brain hemispheres were gently removed leaving dura mater untouched with preserved trigeminal innervations. A hemiskull was placed into a recording chamber and was continuously superfused with oxygenated (5% CO₂/95% O₂) artificial cerebrospinal fluid (ACSF) containing (in millimolars): NaCl 120, KCl 3, CaCl₂ 2, MgCl₂ 1, NaH₂PO₄ 1, NaHCO₃ 25, and glucose 10, pH 7.3 (6–7 ml/min). The nervus spinosus, which is a part of the mandibular branch of the trigeminal nerve (25) was exposed by a careful dissection of dura mater and thereafter was placed into a fire-polished recording glass microelectrode filled with ACSF. Investigated substances were applied to the recording chamber for 10–20 min via superfusion. Low-noise digital amplifier (ISO 80, WPI Inc., Sarasota, FL, USA) was used for recordings. Signals were visualized and analyzed with the WinEDR V3.4.6 software (Strathclyde University, Glasgow, UK). Spontaneous and drug-induced action potentials generated in the distal parts of the transected nervus spinosus were recorded at room temperature. Spike frequencies (number of spikes per every 2 min) were estimated and analyzed using OriginPro 9.1 software (Microcal, Northampton, MA, USA). For control values, the frequency of spikes during 10 min prior to drug application was taken.

Isolated male rat dura mater preparations were fixed during 60 min in 2% p-formaldehyde (Sigma-Aldrich, USA) and rinsed three times for 30 min in the phosphate buffer (PBS, Sigma-Aldrich). Then, the preparations were consecutively incubated for 30 min in 0.3% Triton X-100 (Sigma-Aldrich) for 15 min in a solution containing 5% goat serum (Sigma-Aldrich), 1% bovine serum albumin (BSA, Sigma-Aldrich), and 0.3% Triton X-100 and for 15 min in a solution of 1% BSA and 0.3% Triton X-100 (solution A). These solutions were based on PBS.

Next, preparations were incubated for 15 h at 4°C in the solution A containing goat polyclonal antibody to AChE (dilution 1:1,000, Santa Cruz Biotechnologies, USA), or mouse monoclonal antibody to BuChE (dilution 1:1,000) and rabbit polyclonal antibodies to the neurofilament NF-H (1:1,000, Santa Cruz Biotechnologies, USA). The preparations were rinsed three times in the solution A and then incubated for 1 h at room temperature with the appropriate secondary antibodies conjugated to Alexa Fluor 488 or 647 (Invitrogen, USA) diluted 1:1,000 in the solution A.

After washing in PBS, the preparations were exposed to PBS solution diluted by glycerol (1:1) and used for microscopy (63×, oil immersion, NA = 1.4) with a laser scanning confocal microscope Zeiss LSM 510 Meta (Carl Zeiss, Germany).

To study the action of cholinergic drugs on degranulation of dural mast cells, hemiskulls of Wistar rats of both sexes were used. One control hemiskull was filled for 20 min with the basic saline solution (BSS), containing (in millimolars) 152 NaCl, 5 KCl, 10 HEPES, 10 glucose, 2.6 CaCl₂, 2.1 MgCl₂ (pH adjusted to 7.4 with NaOH), whereas others were filled with BSS containing 50 µM carbachol or 100 µM nicotine. Then, the hemiskulls were left in 4% PFA solution for at least 4 h. After that meninges were dissected carefully and placed on a microscope slide, where they were stained with 0.1% toluidine blue (26). Images of meninges were acquired with Olympus AX70 microscope (20× objective). Homogeneously stained and well-shaped mast cells were classified as non-degranulated, whereas pale poorly stained mast cells as well as mast cells with distorted borders were classified as degranulated (22, 26). To study an effect of carbachol and nicotine on mast cells degranulation, 10 randomly chosen, same in different animals, perivascular areas of meninges enriched by mast cells were analyzed (27). Then, total number and number of degranulated mast cells were counted in a blind manner.

Trigeminal ganglion (TG) cell cultures were prepared as described previously (28). Briefly, trigeminal ganglia of male rats were isolated, and cells were enzymatically dissociated by trypsin (0.5 mg/ml, Sigma-Aldrich), collagenase (1 mg/ml, Sigma-Aldrich), and DNAse (0.2 mg/ml, Sigma-Aldrich) for 20 min. Isolated cells were plated on coverslips coated with poly-l-lysine (0.2 mg/ml, Sigma-Aldrich, MO, USA) and were cultured in F12 medium (Gibco, USA) at 37°C, 5% CO₂ for 48 h prior to an experimental treatment.

Primary TG cells kept in culture for 2 days were rinsed with basic solution (BS) containing (in millimolar) 152 NaCl, 2.5 KCl, 10 HEPES, 10 glucose, 2 CaCl₂, 1 MgCl₂ (pH adjusted to 7.4) followed by loading with 2 µM fluo-3 acetoxymethyl ester (Invitrogen, USA) at 37°C for 45 min in BS. After 20 min postincubation, dishes were transferred to TILL Photonics imaging system (TILL Photonics GmbH, Munich, Germany) and were constantly perfused (at 1.2 ml/min) with BS. The setup was equipped by fast perfusion system (Rapid Solution Changer RSC-200, BioLogic Science Instruments, Grenoble, France), which allowed rapid (exchange time ~30 ms) application of various compounds. Cells were viewed via Olympus IX-70 (Tokyo, Japan) microscope with a specific filter of 488 nm wavelength using 10× objective. Images were collected using CCD camera (SensiCam, PCO imaging, Kelheim, Germany) at sampling frequency set to 2 fps. Cells were further characterized by their responsiveness to a brief application of high potassium (50 mM KCl with compensated osmolarity) as a marker for neurons.

Nociceptive firing was statistically analyzed using OriginPro 9.1 software (Microcal, Northampton, MA, USA) and SigmaPlot 11.0 (Systat Software Inc., CA, USA). One-way repeated measures ANOVA with Tukey post hoc test was used. Mast cell degranulation data were analyzed using two-tailed Mann–Whitney U test with GraphPad Prism software (La Jolla, CA, USA). Two-tailed Student’s t-test [OriginPro 9.1 software (Microcal, Northampton, MA, USA)] was applied to evaluate the significance of increase in basal meningeal firing after 1 µM CGRP pre-application. Results are expressed as mean of data ± SE unless otherwise stated. In all tests, p < 0.05 was taken as significant.

To study if cholinergic agents can stimulate or inhibit meningeal nociceptive signaling, we applied various AChR agonists directly to the rat hemiskull preparation with preserved dura mater innervation. To this end, we recorded spontaneous nociceptive firing of meningeal trigeminal nerves in basal conditions, then the preparation was superfused with a ligand solution, and changes in spike frequency were monitored.

First, we applied the natural agonist ACh to rat hemiskull preparation. Application of 250 µM ACh significantly increased nociceptive firing (Figures 1A–C). Figures 1A,B show representative spikes in control and after ACh application, respectively. Notice that the amplitude and the shape of spikes were not significantly changed after ACh treatment (Figures 1A,B, inset). On average, the frequency of spikes was increased by 2.5-fold by 10 min of 250 µM ACh application (n = 6, p < 0.05, Figure 1C). Lower (50 µM) concentration of ACh slightly increased spike frequency, but this effect was not significant (n = 6, p > 0.05, Figure 1C).

Because ACh, an endogenous agonist of both mAChRs and nAChRs, quickly degrades in live tissues, we next used the stable agonist carbachol, to test whether it can induce similar or higher level of firing. Although ACh demonstrates comparable activity towards both classes of AChRs, carbachol is more potent to mAChR, to which it demonstrates nanomolar affinity in contrast to micromolar constants for its binding to nAChRs (29–32). Like low doses of ACh, 50 µM carbachol showed a clear tendency to increase firing, but this effect was not significant (n = 10, p > 0.05, Figure 1D). However, application of 250 µM carbachol to the rat hemiskull preparation significantly increased the number of nociceptive spikes peaking by 5–10 min of drug application (n = 6, p < 0.05, Figure 1D). Moreover, the significantly increased number of spikes was detected even 6 min after carbachol withdrawal. Notice that the same concentration of ACh induced the significant changes in firing only after 10 min of the compound application consistent with higher pro-nociceptive efficacy of carbachol.

The pro-nociceptive action of carbachol was largely reduced by pretreatment with broad spectrum muscarinic antagonist, atropine. Thus, in the presence of 1 µM atropine, 250 µM carbachol only slightly increased nociceptive firing, but this effect was non-significant (n = 7, p > 0.05, Figure 1D).

Essential evidence suggests that ligand-gated nicotinic receptors (nAChRs) are also expressed by nociceptive sensory neurons to modulate the nociceptive signaling (33–37). During carbachol application, the role of nAChRs could be masked by simultaneous activation of multiple mAChRs often producing functionally opposite effects on nociceptive neurotransmission (38–40). Therefore, to investigate directly a contribution of nAChRs to the nociceptive signaling in meninges, we used the specific nAChRs agonist nicotine. Nicotine (100 µM) induced a strong and prolonged nociceptive firing in trigeminal nerves (Figure 2A). This effect was peaking at 8–10 min (n = 8, p < 0.05, Figure 2B) of nicotine application. Consistent with the involvement of nAChRs, pre-application of 50 µM d-tubocurarine completely prevented the pro-nociceptive effect of nicotine (n = 9, p > 0.05, Figure 2C). TRPA1 receptors have also been suggested as molecular targets sensitive to this agonist (18, 19). However, in rat meninges, TRPA1 receptors are unlikely the target for nicotine, as the selective TRPA1 blocker 25 µM HC-030031 did not prevent the pro-nociceptive action of nicotine (n = 7, Figure 2D).

To study a role of endogenous meningeal cholinergic system in rat dura mater, we explored the expression and localization of AChE and BuChE, two key enzymes implicated in cholinergic signaling by termination of ACh action. The cholinesterase-positive nerve fibers were found mainly in perivascular regions (n = 7 hemiskulls), although spare labeling was observed across the dura mater. Figures 3A–F present a co-expression of AChE and BuChE with neurofilaments (NF-H) in nerve fibers located nearby meningeal blood vessels. Such localization of both enzymes in dura mater suggests their role in control of ACh-mediated signaling.

Further, to test the involvement of endogenous ACh in regulation of meningeal nociceptive firing, we applied the classical inhibitor of AChE neostigmine to the hemiskull preparation, to increase the concentration of endogenous ACh around nerve fibers. Neostigmine (12.5 µM) did not change the nociceptive firing per se (n = 5, p > 0.05, Figures 4A,B). However, a significant neostigmine-induced increase in nociceptive firing was observed after 2 h exposure of meninges to the migraine mediator calcitonin gene-related peptide (CGRP, 1 µM). Thus, in this case, the number of spikes was increased by 6 min action of neostigmine by 2.7-fold (n = 7, p < 0.05, Figures 4A,C). The basal level of trigeminal firing was not changed after 2 h preincubation with 1 µM CGRP (372 ± 124 spikes in 10 min recording, n = 12) versus 385 ± 75 spikes in control (n = 12, p > 0.05). Importantly, co-application of 50 µM d-tubocurarine (non-selective nAChR antagonist) prevented the facilitatory pro-nociceptive action of neostigmine (n = 5, p < 0.05, Figure 4D), suggesting the major role of nicotinic AChRs in mechanism of neostigmine-induced firing.

As degranulation of mast cells has been recently suggested as one of potential triggers of migraine (22, 41), we next tested the ability of carbachol and nicotine to induce degranulation of dural mast cells. In these experiments, we also explored a possible role of gender in the ability of these cholinergic agents to induce degranulation. Figure 5 shows representative samples of mast cells in control (Figure 5A) and after application of 50 µM carbachol (Figure 5B). Pooled data in Figure 5C show that the application of 50 µM carbachol to the whole mount meningeal preparation for 20 min induced a significant degranulation of mast cells in both sexes. Thus, in males, the level of degranulation in control (vehicle treatment) was 10 ± 8% (n = 4), whereas it was largely enhanced up to 62 ± 5% after 50 µM carbachol treatment (n = 4, p < 0.05). In females, there was 19 ± 5% of degranulation in control (n = 4), and the large enhancement of degranulation up to 61 ± 8% was observed after 50 µM carbachol exposure (n = 4, p < 0.05).

In contrast to carbachol, the concentrations of nicotine, which induced the nociceptive firing, did not promote degranulation of mast cells (Figure 5D) suggesting only a direct action of this agonist on nerve terminals. Thus, in males, the level of mast cells degranulation was 22 ± 4% in control (n = 4) and 20 ± 4% after application of 100 µM nicotine (n = 4, p > 0.05). Likewise, in females, these values were 24 ± 8% for vehicle (n = 4) and 24 ± 5% for nicotine treatment (n = 4, p > 0.05).

Thus, the pro-nociceptive action of nicotine was clearly mast cells independent, whereas the action of carbachol could potentially be, at least partly, mediated by products released from degranulated mast cells.

The somas of TG neurons are important components of peripheral trigeminal nociceptive system, and they could be implicated in development of migraine pain (1). To study the effect of cholinergic agents in this part of the trigeminal nociceptive system, we used the live imaging technique to monitor Ca²⁺ signals in neurons isolated from TG. Application of 50 mM KCl at the end of protocol was used to identify neurons (42). To study the neurochemical profile of the neurons responding to cholinergic agonists, we tested if they were activated by the classical algogen capsaicin operating via TRPV1 receptors (43). We found a variable profile of trigeminal neurons as some cells responded to nicotine, some only to capsaicin and the fraction of cells responded to both agonists.

Figure 6 shows example of two neurons exposed to 100 µM nicotine. In one trigeminal neuron (Figure 6A), nicotine induced a large fast Ca²⁺ transient, and later this neuron responded also to capsaicin. In the other neuron (Figure 6B), nicotine apparently did not evoke response; however, this cell also responded to capsaicin suggesting the nociceptive phenotype. On average, almost half of neurons (43%, 70/163 cells) responded to nicotine suggesting that nAChRs can have an important role in trigeminal nociception. Consistent with previous data, 34% of neurons (56/163) were sensitive to application of 200 nM capsaicin. Part of them (12/56 cells) were also sensitive to nicotine.

We also tested the action of carbachol to address the presence of mAChRs and their co-expression with nicotinic receptors in individual trigeminal neurons (Figure 7). For this purpose we first applied 100 µM nicotine with subsequent application of 50 µM carbachol. We found that 71% of neurons (82/116) responded to carbachol, whereas 70% of these cells were responsive also to nicotine application. Thus, there were neurons that responded only to carbachol application (Figure 7B) suggesting functional expression of mAChRs in trigeminal neurons.