We used l-carrageenan (carrageenan; WAKO Pure Chemical Industries Ltd., Osaka, Japan), ropinirole (Sigma-Aldrich Japan, Tokyo, Japan) as a D2-like receptor agonist, haloperidol (WAKO Pure Chemical Industries Ltd.) as D2-like receptor antagonist, and carrier-free recombinant mouse IL-17A (rmIL-17A; R&D Systems, Guthrie, MN, USA). In the preliminary in vitro experiments, we have found that 10 mg/ml is more effective than 1 mg/ml (data not shown). It is enough to conduct the in vivo experiment at one dose (20). According to a paper by Hashimoto et al (15), the endogenous IL-17A concentration is approximately 1.0 to 1.5 ng/ml, and we used this concentration as a reference for deciding on the dose of IL-17A to use in this experiment. We decided on the ropinirole concentrations by referring to previous reports (17,21).

Cells of the murine gingival epithelial cell line GE1 (RCB1709; RIKEN BioResource Center Cell Bank, Japan) were cultured in SFM101 medium (Nissui, Tokyo, Japan) supplemented with 1% fetal bovine serum (Sigma-Aldrich Japan), 100 U/ml penicillin G (Sigma-Aldrich Japan), 100 µg/ml streptomycin (Sigma-Aldrich Japan), and 10 ng/ml epithelial growth factor (Sigma-Aldrich Japan) (22). The cells were incubated in a humidified atmosphere of 5% CO₂ at 33̊C. For the in vitro assay, we seeded GE1 cells into 24-well plates at 30x10⁴ cells per well. We added carrageenan at the concentration of 50 or 100 µg/ml and/or rmIL-17A at the concentration of 0.5 or 2 ng/ml to the wells. Ropinirole was added to the wells at the concentration of 1, 10, or 50 µg/ml prior to the addition of carrageenan and/or rmIL-17A. After 24 h of incubation, the cells and supernatants were harvested.

The harvested cells were rinsed with ice-cold phosphate-buffered saline (PBS). QIAzol Lysis Reagent (QIAGEN, Hilden, Germany) was added to the samples, and the total RNA was extracted. The reverse-transcription reaction was performed with a High Capacity cDNA Reverse Transcription kit (Thermo Scientific). TaqMan Gene Expression Assays for CXCL1 (assay identification number: Mm04207460_m1), IL17-RA (assay identification number: Mm00434214_m1), and glyceraldehyde-3-phosphate dehydrogenase (assay identification number: Mm99999915_g1), which was used as an endogenous control, were obtained from Thermo Scientific. As Taqman probes do not provide the sequence information, we cannot show the sequence information. All experiments were performed in quadruplicate, i.e., each reaction was performed in quadruplicate on four individual samples. Values were normalized to those of glyceraldehyde-3-phosphate dehydrogenase using the 2^-ΔΔCt method. This experiment was repeated at least three times.

CXCL1 protein in the harvested supernatants was measured using a mouse CXCL1 ELISA kit (Proteintech, Rosemont, IL USA) according to the manufacturer's instructions. All experiments were performed in quadruplicate, i.e., each reaction was performed in quadruplicate on four individual samples. This experiment was repeated at least three times.

The bilateral maxillary bone including the second molar of two rats were used in each group. We used 6 rats in total. The experiments were approved by the Animal Research Committee of Saitama Medical University (approval number: 2877), and conducted according to the institutional guidelines for ethical animal experiments. The humane endpoint is 20% body-weight reduction and severe suffering. All rats were housed in appropriate animal care facilities at Saitama Medical University, Japan. Five-week-old male Wistar rats (200 to 300 g in weight) were obtained from Japan SLC (Shizuoka, Japan). The rats had access to food and water ad libitum, and were maintained on a 12-h light/dark cycle at 23±1˚C with 60±10% humidity. We used a mixture of medetomidine hydrochloride (0.15 mg/kg; Nippon Zenyaku Kogyo, Fukushima, Japan), midazolam (2 mg/kg; Astellas Pharma, Tokyo, Japan), and butorphanol tartrate (2.5 mg/kg; Meiji Seika Pharma, Tokyo, Japan) as an anesthetic mixture, which was administered intraperitoneally (23-25). These drugs do not affect dopamine signaling (26). For euthanasia, sodium pentobarbital (200 mg/kg; Kyoritsu Seiyaku, Tokyo, Japan) was administered intraperitoneally. We verified cardiac and respiratory arrest. We used a previously reported carrageenan-induced rat model of periodontitis (27). While the rats were under anesthesia, buccal and lingual gingiva were exfoliated using an explorer. We divided the rats into three groups: a PBS group, a carrageenan (CA) group, and a carrageenan and ropinirole (CA/RP) group. For the PBS group, silk ligature, cut to the length of the mesiodistal distance of the mandibular second molar, was immersed in PBS, which is the solvent for carrageenan. For the CA group, the silk ligature was immersed in 1% carrageenan. For the CA/RP group, the silk ligature was immersed in 1% carrageenan and 10 µg/ml ropinirole. These ligatures were inserted into the periodontal pocket once a week for 4 consecutive weeks. The ligatures were inserted once a week and removed until the animal was sacrificed. The ligatures did not affect the animals' quality of life because 20% body-weight reduction has not been observed. The experimental protocol is shown in Fig. S1.

The rats were euthanized 4 weeks after the first operation. The maxillary bone was dissected, and subsequently fixed with 70% ethanol for analysis on a micro-CT 35 (SCANCO Medical, Brüttisellen, Switzerland). After preparing three-dimensional images (Fig. S2), the perpendicular distance between the cemento-enamel junction and the alveolar margin on the palatal side of the maxillary second molar was measured as shown in Fig. S3(26). This experiment was repeated at least three times.

Differences between more than three groups were analyzed by one-way analysis of variance with Tukey's post-hoc tests. P-values <0.05 were considered to indicate statistical significance. All data are presented as the mean ± standard error of the mean. The analyses were performed using EZR software version 1.52 which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria) (28). These experiments were repeated at least three times.

Previously, we demonstrated that tannic acid, which is a D2-like receptor agonist, inhibited alveolar bone resorption in a carrageenan-induced rat model of periodontitis (20). In the present study, we used the same model to analyze the function of ropinirole, and conducted micro-CT analysis. The results revealed that the perpendicular distance was increased in the CA group when compared to the PBS group, and the perpendicular distance of the CA/RP group was significantly shorter than that of the CA group (Fig. 1).

As an inflammatory cytokine, IL-8 plays a key role for neutrophil migration in human inflammatory diseases (29). Since carrageenan induces IL-8 secretion in human colonic epithelial cells (30), we examined whether carrageenan modulates CXCL1 expression in murine gingival epithelial cells. As shown in Fig. 2A, CXCL1 expression in GE1 cells was enhanced by the addition of carrageenan in a dose-dependent manner. As the level of IL-17 is significantly higher in gingival crevicular fluid from patients with periodontitis than that from healthy participants (31), we evaluated the influence of carrageenan on IL-17RA mRNA expression in gingival epithelial cells. We found that carrageenan induced IL-17RA expression in GE1 cells (Fig. 2B).

We next examined whether IL-17A, which is the prototypical member of the IL-17 family, modulates CXCL1 and IL-17RA mRNA expression in the presence of carrageenan. As shown in Fig. 3A, CXCL1 expression was increased by carrageenan in a dose-dependent manner. In contrast, IL-17RA expression was not affected by the addition of carrageenan (Fig. 3B). These results suggest that carrageenan modulates IL-17A-mediated CXCL1 expression in gingival epithelial cells.

We next examined whether ropinirole, which is a D2-like receptor agonist, inhibits CXCL1 and IL-17A mRNA expression in GE1 cells in the presence of carrageenan in vitro. As expected, CXCL1 expression was significantly suppressed by the addition of ropinirole in a dose-dependent manner (Fig. 4A). In addition, IL-17A expression was significantly suppressed by the addition of ropinirole at a dose of 10 µg/ml (Fig. 4B).

Furthermore, we examined whether ropinirole at a high dose modulates CXCL1 and IL-17RA mRNA expression in the presence or absence of IL-17A with or without carrageenan. Ropinirole at 50 µg/ml in the presence of carrageenan significantly inhibited the upregulation of CXCL1 and IL-17RA (Fig. 5A and B). In addition, we confirmed that the CXCL1 protein level was also suppressed by the addition of ropinirole in the presence of carrageenan and rmIL-17A (Fig. 5C). These results indicated that ropinirole inhibits the IL-17A-mediated CXCL1 expression induced by carrageenan in gingival epithelial cells. We speculate that carrageenan binds to unknown receptor and subsequently activates IL-17RA expression, resulting in CXCL1 upregulation (Fig. S4). Thus, carrageenan modulates IL-17RA as a stimulator.

To examine whether the modulation of CXCL1 production by carrageenan is dependent on D2-like receptors, we lastly explored the effect of haloperidol, which is a D2-like receptor antagonist, on CXCL1 production. As anticipated, haloperidol significantly promoted CXCL1 expression (Fig. 6). This result confirmed that carrageenan enhances CXCL1 production via D2-like receptors.