Male C57BL/6N mice aged 9–11 weeks, weighing 24–30 g (Kyudo, Saga, Japan), and TRPM8 homozygous knockout (TRPM8KO) mice (Prof. Makoto Tominaga, Thermal Biology Group, Exploratory Research Center on Life and Living Systems, Okazaki, Japan) were housed in groups of five mice per cage and kept under standard laboratory conditions in a temperature- and humidity-controlled room (25 ± 2°C and 55 ± 5%, respectively) on a 12-h light/dark cycle (lights on at 8:00 a.m.) (Dhaka et al., 2007). The animals had free access to food and water. The animal care and experimental procedures were approved by the Experimental Animal Care and Use Committee of Yamaguchi University School of Medicine, Japan. All experiments were performed in accordance with the guidelines of the Japan Association for Laboratory Animal Facilities of National University Corporations.

Pentylenetetrazol (PTZ; Tokyo Chemical Industry, Japan) and penicillin G potassium (PG: Meiji, Tokyo, Japan) were dissolved in saline. WS-3 (Funakoshi, Japan) was dissolved in 1% dimethyl sulfoxide (DMSO; Merck KGaA) in saline.

To determine whether TRPM8 channel deficiency exacerbates generalized and focal seizures in mice, we compared PTZ- and PG-induced seizures in WT and TRPM8KO mice. PTZ-induced seizure mice were generated according to a previously reported method (Mizoguchi et al., 2011). Mice were intraperitoneally (i.p.) administered a single dose of PTZ (40 mg/kg). A model with focal epilepsy originating in the left cerebral cortex was generated by modifying a previously reported method (Moriyama et al., 2019). The animals were anesthetized with urethane (1.75 g/kg i.p.) or through inhalation of sevoflurane (Pfizer Japan, Tokyo, Japan) (3% for induction, 1.5% for maintenance) and immobilized using a stereotaxic apparatus. To maintain the body and brain temperature in each rat, rectal temperature was adjusted to 37.5 ± 0.5°C using a heat pad. We created three burr holes measuring 1.0 mm in diameter, two for the ECoG recording and the other for the reference electrode, above the bilateral sensorimotor cortices and cerebellum using the following coordinates: 1.0 mm posterior; ± 2.0 mm lateral from the bregma, and 2.0 mm posterior from the lambda. We also placed a thin thermocouple (IT-24, Physitemp, Tokyo, Japan), and two ECoG and reference electrodes in the bilateral somatosensory cortices and cerebellar area between the skull and dura. We placed a ground electrode in the tail. A small slit in the dura was created to insert an injection cannula (0.4 mm diameter × 40 mm length, and 3 µl volume; EIM-40, Eicom, Kyoto, Japan). The injection cannula was inserted to a depth of 0.75 mm from the brain surface. PG at a concentration of 200 IU/µl was injected intracortically for 10 min at a rate of 0.1 µl/min using a 10 µl Hamilton syringe and a microinjection pump (ESP-64, Eicom, Japan).

WS-3 was administered intracortically for 10 min, with WS-3 administration starting 90 min after the PG injection at a rate of 0.1 µl/min using a microinjection pump. In the prevention of epileptic seizures induced by TRPM8 activation, DMSO or WS-3 was administered to mice 30 min before PG injection.

The behavior of mice was observed for 1 h following PTZ injection (Figure 1). Because anesthesia was continuously maintained for at least 5 h by urethane, we did not perform seizure monitoring (Figure 2A). The behavior of mice was observed for 1 h after the end of ECoG recording for 2.5 h (Figure 3A). Referring to previous studies (Mizoguchi et al., 2011), PTZ-, or PG-induced seizure intensity was scored as follows: stage 0, no response; stage 1, ear and facial twitching; stage 2, convulsive twitching axially through the body; stage 3, myoclonic jerks and rearing; stage 4, turning over onto the side, wild running, and wild jumping; stage 5, generalized tonic-clonic seizures; and stage 6, death.

ECoG was recorded by modifying a previously reported method (Moriyama et al., 2019). EDs developed 10–20 min after PG injection and reached a steady state 1.5 h after PG injection (Moriyama et al., 2019), remaining stable for at least 5 h (data not shown). Because icilin, a TRPA1 and TRPM8 agonist, eliminated all EDs for only a few minutes (Moriyama et al., 2019), we regarded the drug efficacy period as 5 min from 5 to 10 min after the start of TRPM8 agonist injection. ECoGs were continuously recorded for 2.5 h (0.5 h for ECoG stabilization and 1.5 h for stabilization of PG-induced EDs, and 0.5 h for drug efficacy evaluation; Figure 2A; and 0.5 h for ECoG stabilization and 0.5 h for ECoG stabilization after 1% DMSO or WS-3 injection, and 1.5 h for stabilization of PG-induced EDs; Figure 3A).

ECoGs were amplified by a bio-amplifier (EX-1; Dagan Corporation, Minneapolis, MN) using an analogue-to-digital converter at a sampling rate of 2 kHz (PowerLab 8/30; AD Instruments, Castle Hill, Australia). The conditions for recording ECoGs were as follows: low-frequency filter, 0.1 Hz; high-frequency filter, 10 kHz; and notch filter, off.

To evaluate the effects of TRPM8 channel activation on PG-induced EDs, the left ECoG was fast Fourier-transformed. We calculated the power of the low beta band (14–24 Hz) during the subsequent 5 min periods using Lab Chart Pro v. 8.1.16 (AD Instruments), as follows: basal activity, just before PG injection; pre-injection, just before DMSO or WS-3 injection; and post-injection, just before the end of the latest injection. The changes in the low beta band power in ECoG were reported to be a good indicator of the degree of focal neocortical seizure (Ludvig et al., 2009) and suppression of EDs by FBC (Kida et al., 2012).

Statistical analyses were performed using StatMate version Ⅲ software (Advanced Technology for Medicine and Science, Chiba, Japan). All results are expressed as mean ± standard error of the mean (SEM). Statistically significant differences were evaluated by Tukey’s test, Dunnett’s test, Student’s t-test, or Welch’s t-test with p < 0.05 indicating statistical significance.

We first investigated whether a lack of TRPM8 channels results in deteriorated epileptic seizures. Wildtype (WT) mice intraperitoneally administered a single dose of PTZ at 40 mg/kg exhibited ear and facial twitching, turning over onto the side, myoclonic jerks and rearing, and wild running (stages 1, 3, or 4). Administration of PTZ (40 mg/kg) to TRPM8KO mice induced generalized turning over onto the side, wild running and jumping, tonic-clonic seizures, or lethal seizures (stages 4, 5, or 6). PTZ administration resulted in a greater increase in seizure score in TRPM8KO mice compared with that in WT mice (p < 0.001, Student’s t-test; Figure 1).

Icilin, a TRPM8 and TRPA1 agonist, suppressed PG-induced EDs (Moriyama et al., 2019), but it remains unknown whether selective TRPM8 agonists suppress EDs. Therefore, we investigated the effects of WS-3, a selective TRPM8 agonist, on PG-induced EDs. We continuously recorded ECoGs for 150 min in mice anesthetized with urethane (Figure 2A). Mice were injected with PG 30 min from the start of ECoG recording and were injected with vehicle or WS-3 90 min after PG injection (Figure 2A). EDs plateaued 90 min after PG injection (Figures 2B–G). The higher the dose of WS-3 (0.1–1 μM), the greater the amplitude of EDs was decreased in WT mice (Figures 2B–E). DMSO or low and middle doses of WS-3 did not affect the power of the low beta band in WT mice, whereas a high dose of WS-3 (1.0 mM) significantly suppressed the power of the low beta band (p = 0.042 vs. pre-injection group, Student’s t-test; Figure 2H).

To elucidate the effects of a lack of TRPM8 channels on EDs, we compared EDs in WT and TRPM8KO mice. A lack of TRPM8 channels increased the intensity of low beta band power during pre-injection by approximately double compared with WT mice at the same dose of DMSO or WS-3 (p < 0.034 and p < 0.031, respectively, Welch’s t-test; Figure 2H). Administration of vehicle to TRPM8KO mice did not affect the PG-induced EDs, and this was not suppressed by 1.0 mM WS-3 (Figures 2F–H). In WT mice, EDs were focally induced by epilepsy-inducer, and were rarely propagated in the contralateral side (Figures 2B–E, p = 0.035, p = 0.064, respectively, Welch’s t-test; Figure 2I), whereas the EDs were easily propagated in the contralateral side by the lack of TRPM8 channels (Figures 2F,G, p = 0.003, p = 0.015, respectively, Welch’s t-test; Figure 2I). The intensity of low beta band that was observed in the contralateral side in was more increased than that in WT mice (p = 0.018, p = 0.042, respectively, Welch’s t-test; Figure 2I). In addition, the lack of TRPM8 channels resulted in a shorter firing latency of EDs by the epilepsy inducer compared with that in WT mice (p < 0.001, p < 0.001, respectively, Welch’s t-test; Figure 2J).

To reveal the effects of TRPM8 agonists on the development of EDs and epileptic seizures, mice were injected with WS-3 30 min before receiving an epilepsy-inducer injection (Figure 3A). In WT mice, PG induced EDs when 1% DMSO was injected before the PG injection, whereas EDs were rarely detected when WS-3 was injected before the PG injection (Figures 3B,C). DMSO administration before PG injection did not affect the increased intensity of the beta band, whereas WS-3 administration before PG injection suppressed the increased intensity of the beta band by PG (p = 0.038, Welch’s t-test; Figure 3F). Injection of vehicle before PG injection in WT mice promoted myoclonic jerks and rearing or turning over onto the side, wild running and jumping (stage 3 or 4). In contrast, WS-3 pre-injection of WT mice resulted in no response or ear and facial twitching or convulsive twitching axially through the body (stage 0 or 2). The seizure score was 3.40 ± 0.24 by PG, which was reduced by WS-3 pre-injection (p = 0.002, Student’s t-test; Figure 3G).

To elucidate the effects of TRPM8 channel absence on the development of EDs and epileptic seizures, we compared EDs and seizure scores between WT and TRPM8KO mice. In TRPM8KO mice, PG injection after 1% DMSO or WS-3 injection induced EDs (Figures 3D,E), increasing the intensity of the beta band compared with basal activity (p < 0.001 and p < 0.001, respectively, Tukey’s test; Figure 3F). A lack of TRPM8 channels increased the intensity of low beta band power during post-injection, compared with WT mice at the same dose of DMSO (p < 0.045, Welch’s t-test; Figure 2H). In addition, PG injection after 1% DMSO or WS-3 injection of TRPM8KO mice resulted in turning over onto the side, wild running, and wild jumping or generalized tonic-clonic seizures (stages 4 or 5). The seizure score was higher in TRPM8KO mice compared with WT mice following PG injection after 1% DMSO injection, whereas WS-3 injection before PG injection did not reduce the seizure score in TRPM8KO mice (p = 0.020 and p = 1, respectively, Student’s t-test; Figure 3G).