Specific pathogen free (SPF) and conventional (CV) NC mice were purchased from Japan SLC, Inc. (Shizuoka, Japan). Male mice (8–30 weeks old) were housed at a temperature of 25°C ± 1°C with a 12 h light-dark cycle (light from 7:00–19:00) and water and food (CE-2, CLEA Japan, Inc., Tokyo, Japan) were provided ad libitum. All animal experiments were conducted according to relevant national and international guidelines contained in the “Act on Welfare and Management of Animals” (Ministry of Environment of Japan), and procedures used in the animal experiments were approved by the Committee for Animal Experiments at the University of Toyama (A2019PHA-12, A2019PHA-13, A2022PHA-2, A2023PHA-13). Efforts were made to minimize animal suffering and counts.

MGB besylate (code number: DS-5565) and PGB were provided by Daiichi Sankyo Co., Ltd. (Tokyo, Japan). GBP was obtained from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). The drugs were dissolved in distilled water, which was procured from Otsuka Pharmaceutical Factory, Inc (Tokushima, Japan). In this study, dose levels were presented to reflect those of the free form. When given perorally, the drugs were administered at a volume of 0.1 mL/10 g body weight. To act locally on the cervical spinal cord, where NC mice scratch most often, we conducted intracisternal injection. For intracisternal injection, the drugs were dissolved in saline (Otsuka Pharmaceutical Factory, Inc., Tokushima, Japan) and administered at a volume of 5 μL/site via a disposable 27-gauge needle attached to a Hamilton microsyringe (Hamilton Company, Nevada, United States) after the mice were lightly anesthetized using isoflurane (FUJIFILM Wako Pure Chemical Corp., Osaka, Japan). We chose the dose for intracisternal injection of MGB for reference previously described (Kitamura et al., 2014).

All behavioral tests were conducted during the light period (7:00–19:00). The scratching behavior was recorded using the SCLABA^®-Next (Noveltec Inc., Kobe, Japan) real-time scratch counting system. The animals were placed in an acrylic cage (approximately 150 mm wide, 200 mm deep, and 350 mm high) for at least 30 min before scratching behavior was measured. Locomotor activity was measured from image data recorded by SCLABA^®-Next.

The running wheel test was performed for 30 min using activity wheel (SW-20; Melquest Ltd., Toyama, Japan) (Akiyama et al., 2018). The number of revolutions of the wheel cage was determined. The animals were placed in a wheel cage and trained for at least 30 min/d for 3 days prior to testing.

The rotarod (Acceler Rota-Rod for mice 7,650; Ugo Basile, Italy) test was performed for reference previously described (Kayser and Christensen, 2000). The animals were placed on the rod, and a timer switch was simultaneously activated to rotate the rod from approximately 3.3 rpm to approximately 38 rpm for a maximum 5 min. The timer was stopped when the animals fell to the surface or performed three rotations while holding the rod. The animals were tested for two sessions per day with a resting period of approximately 30 min between sessions. The animals were trained 3 min/d for 3 days before the test day. On the test day, only the animals that remained balanced on the rotating rod for 3 min (cut-off time) were selected for testing.

Mice were anesthetized with intraperitoneal injection of a mixture of three anesthetic agents: 0.75 mg/kg medetomidine hydrochloride (Domitor, Nippon Zenyaku Kogyo, Koriyama, Japan), 4.0 mg/kg midazolam (Midazolam (SANDOZ), Sandoz K.K., Tokyo, Japan), and 5.0 mg/kg butorphanol tartrate (Betorphal, Meiji Seika Pharma, Tokyo, Japan) and perfused transcardially with 10 mL phosphate-buffered saline (PBS) at pH 7.4, followed by 10 mL of 4% paraformaldehyde (Hiramatsu et al., 2021). The spinal cord (cervical1-5) was removed, immersed in the same fixative overnight, and then immersed in 25% sucrose in PBS for 48 h at 4°C. Tissues were cut into 30 μm thick sections by use of a cryostat (CM3050 S; Leica, Germany). The sections were then rinsed three times with PBS for 5 min each.

These methods were refined from previous reports (Takanami et al., 2010). For antigen activation, spinal cord sections were treated with HistoVT One (Nacalai Tesque, Kyoto, Japan) for 20 min at 70°C in PBS. The sections were then rinsed three times with PBS containing 0.3% Triton X-100 (PBS-T, Wako, Osaka, Japan) for 5 min each. Nonspecific binding components were blocked with 1% normal goat serum and 1% bovine serum albumin in PBS-T for 1 h at room temperature. Subsequently, the sections were treated with an antibody against α₂δ-1 subunit (1:1,000; C5105; Sigma) in the blocking buffer for 2 days at 4°C. The sections were rinsed with PBS-T three times for 5 min each and treated for 120 min at room temperature with Alexa Fluor 488-labeled anti-rabbit IgG (1:1,000, Thermo Fisher Scientific, Waltham, MA, United States). Sections were rinsed with PBS-T, and mounted on glass slides (Matsunami Glass, Kishiwada, Japan), and coverslipped (Matsunami Glass) with mounting medium (FLUOROSHIELD; ImmunoBioScience Corp, CA, United States). Fluorescence images were captured using a microscope (BZ-X800; Keyence Corporation, Osaka, Japan).

Results are expressed as the mean ± standard error of the mean. Statistical differences between the two groups for various behavior tests were analyzed using the unpaired t-test and Mann Whitney test. If the normality test or equal variance test failed, the unpaired t-test was replaced with the Mann Whitney test. Multiple-group comparisons were performed using a one way analysis of variance (ANOVA) followed by the Tukey’s test and Dunnett’s test, and Kruskal–Wallis test followed by Dunnett’s test. All tests were considered statistically significant at p < 0.05. Prism 5 (GraphPad Software Inc., La Jolla, CA, United States) was used for the statistical analyses.

Comparisons between healthy SPF and CV mice with AD are shown in Figure 1. Dermatitis was observed all over the body, including the face, in CV mice compared to SPF mice (Figure 1A). There was a significant difference in scratching bouts within 1 h between the CV and SPF groups (p < 0.001, Figure 1B). The scratching duration per scratching was significantly longer in the CV group than in the SPF group (p = 0.003, Figure 1C). Moreover, there was no difference in the total distance moved between the SPF and CV group (Figure 1D).

We investigated whether MGB has an antipruritic effect and its action time. When the vehicle (distilled water) or MGB (10 mg/kg) was administered orally to CV mice and analyzed for 12 h, MGB produced an antipruritic effect (p = 0.031, Figure 2A; Supplementary Figure S1). In addition, when we analyzed chronologically, MGB had the effect for up to 6 h after administration (p = 0.012, Figure 2B). Further analysis indicated that the antipruritic effect of MGB appeared to be present at least between 2.5 and 3.5 h (p = 0.002, Figure 2C). Thereafter, we decided that behavioral experiments were conducted 2.5–3.5 h after the administration of MGB. When the vehicle or MGB (1, three or 10 mg/kg) was administered orally, MGB produced the antipruritic effect in a dose-dependent manner (vehicle vs. 10 mg/kg, p < 0.05, Figure 2D). Additionally, locomotor activity among these four groups did not show a significantly change (Figure 2E).

Gabapentinoids are associated with the sedation as a side effect (Gou et al., 2021). In order to investigate more details of the effect of MGB on motor function, we conducted three behavioral tests, namely, the rotarod test, the running wheel test, and the locomotor activity test (using SCLABA^®-Next) in SPF mice. Oral administration of the vehicle or MGB (10 mg/kg) caused no significant changes (Figure 3). These results suggest that MGB (10 mg/kg) does not have a sedative effect.

The conventional gabapentinoids, GBP and PGB, were originally developed as anticonvulsants, and then currently are indicated for chronic pain disorders, such as neuropathic pain as well (Kremer et al., 2016). First, we investigated the antipruritic effects of GBP, PGB, and MGB and the duration of their activity. We orally administered the vehicle, MGB (10 mg/kg), GBP (100 mg/kg), or PGB (30 mg/kg) to CV mice and monitored them for 6 h. Figure 4A shows the chronological rate of change in the scratch bouts as a heatmap. Between 1.5 and 2.5 h after administration, differences were observed in the MGB, GBP, and PGB groups compared to the vehicle group. Therefore, we decided to further analyze this period. Compared with the vehicle group, all gabapentinoids showed significantly suppressed scratch bouts (vs. vehicle group, p < 0.05, Figure 4B). However, we observed a significant decrease in locomotor activity in GBP and PGB groups (vs. vehicle group, p < 0.05, Figure 4C). This indicates that GBP and PGB may have sedative effects.

In this study, we investigated the sedative effects of GBP and PGB in SPF mice. We conducted two behavioral tests, the locomotor activity test and rotarod test between 1.5 and 2.5 h after administration. In the locomotor activity test, GBP did not affect the total distance moved by SPF mice, whereas in the rotarod test, it significantly reduced the latency to fall (Figure 5A; p = 0.006; Figure 5B). In contrast, PGB decreased the total distance moved, but did not affect the latency to fall. (p = 0.017, Figures 5C, D). These data suggest that GBP and PGB exert sedative effects.

To elucidate whether MGB exerts an antipruritic effect via the CNS, including the SDH, we intracisternally administered MGB (10 μg/site) to CV mice and measured the number of spontaneous scratching behavior for 6 h (Supplementary Figure S2). When we analyzed from 0.5 to 1.0 h after intracisternal administration, MGB (10 µg/site) significantly inhibited spontaneous scratching bouts compared to the saline group (p = 0.007, Figure 6A) and there was no significant change in locomotor activity (Figure 6B). Then, we performed the locomotor activity test using SPF mice during this period. When SPF mice were intracisternally injected with MGB (10 µg/site) and subjected to the locomotor activity test, no difference was observed between the saline and MGB groups (Figure 6C). The α₂δ-1 subunit of VGCC, the target of MGB, is expressed at primary afferent fiber terminals in the SDH. Here, we used immunohistochemical staining to analyze the expression of the α₂δ-1 subunit in the SDH. We observed the α₂δ-1 subunit expression in the superficial layer of the SDH in CV mice (Figure 6D; Supplementary Figure S3). These data indicate that MGB exerts an antipruritic effect via the α₂δ-1 subunit in the SDH.