To the best of our knowledge, this is the first study to comprehensively analyze the expression of genes involved in the healing process of periodontitis in diabetic animal model and in clinical samples from patients with T2D. During the healing process of experimental periodontitis, we observed a decreased mRNA expression of PPARα in the gingiva of the ZDF group. Immunofluorescence staining revealed a decreased PPARα expression in the inner pocket epithelium and connective tissue. In the clinical study, after undergoing non-surgical periodontal therapy, the reduction in PD was significantly decreased in the T2D group, and the mRNA expression of PPARα in the gingiva at the residual periodontal pockets was lower in the T2D group than in the control group. Similar findings were observed for PPARγ. This study suggested that the downregulation of PPAR expression is involved in impairing responsiveness following periodontal therapy in individuals with DM.
ZDF rats are valuable models for identifying molecules contributing to obesity-related insulin resistance and elucidating the molecular pathogenesis of T2D. ZDF rats exhibit an age-dependent diabetic phenotype, developing obesity at age of five weeks and starting to show the metabolic states of DM, such as hyperinsulinemia and insulin resistance29. In this study, we used eight-week-old ZDF rats and conducted experiments after confirming the induction of hyperglycemia and insulin resistance.
Ligation-induced experimental periodontitis has been widely used as a rodent model. Both oral microbiome alterations and physical stimuli by ligation involve the induction of periodontal tissue inflammation16,30. In this model, significant inflammation is well-established at seven days after ligature placement31, and the ligation-induced periodontal tissue destruction reaches an approximate plateau after 14 days22. This experimental model can be also used to observe periodontal tissue healing after ligature removal because tissue attachment recovers with the disappearance of inflammation after ligature removal18. Previous study reported that IL-1β expression is significantly reduced two days after ligature removal compared with that during ligature removal18. In our pilot study, IL-1β expression was significantly reduced one day after ligature removal compared with that during ligature removal, and we observed a significant decrease in inflammatory cytokines such as IL-6 and TNFα in the gingiva one day after ligature removal compared with that during ligation (data not shown). Therefore, we evaluated the tissue one day after removal of the ligature, which had been ligated for seven days in this study.
In this study, bone resorption was significantly higher in the ZDF group than in the control group after ligature placement. Bone recovery after ligature removal was worse in the ZDF group. This finding is consistent with the findings of a previous report that the new bone formation after ligature removal was significantly lower in the ZDF group24. They also reported that the number and activity of osteoclasts were significantly higher after ligature removal in the ZDF rats than in the controls24.
To date, no study has comprehensively evaluated gene expression during the healing process in diabetic animal models or patients with DM. As a result of KEGG pathway analysis, we found decreased PPAR signaling, which regulates glucose and lipid metabolism and transport in the DM group. PPAR is a subfamily of the nuclear receptor superfamily that regulates gene expression by binding to retinoid X receptors as PPAR response elements32. To date, three subtypes called PPARα, PPARβ/δ, and PPARγ have been identified33. Dysfunction of the Krüppel-like factor 16-PPARα pathway has been shown to promote hepatic insulin resistance and fatty liver.34 Agonists of PPARγ have been clinically applied as thiazolidinedione antidiabetic agents because they simultaneously enhance insulin signaling via insulin receptor substrate 1 (IRS-1) and inhibit the mitogen-activated protein kinase pathway35,36. PPAR signaling has been reported to have many functions related to inflammation37. PPARα has anti-inflammatory effects and alters cytokine secretion by T-cell lines38. PPARγ has a wide range of biological roles, including the inhibition of inflammatory signaling via NF-κB.
Recently, PPAR signaling has been reported to be protective against progression of periodontitis39. In ligature models, PPARα expression in the gingiva was decreased40 and bone loss was significantly greater in PPARα knockout mice than that in wild-type mice, suggesting that decreased PPARα expression associates with the increased bone resorption40. In another in vivo study with a rodent ligature model, treatment with the PPARα agonist WY14643 reduced the inflammatory markers37. Systemic administration of GW0742, a PPARβ/δ agonist, suppressed ligature-induced periodontitis and tissue damage41. Furthermore, in vitro studies reported that PPARβ/δ activation inhibited Porphyromonas gingivalis-derived LPS-stimulated activation of matrix metalloproteinase-2 in the human gingival fibroblasts42. Some reports have linked increased PPARγ expression to osteoclast activation in ligature models43, and the administration of PPARγ agonists has been suggested to promote anti-inflammatory effects44,45. However, few basic and clinical studies have addressed the PPAR-mediated effects on the healing process of periodontitis in diabetic condition. This study first revealed that PPAR signaling may be involved in the DM-induced delayed recovery from periodontal inflammation during the healing process.
Previous clinical reports have suggested that the outcomes of non-surgical periodontal therapy are impaired in patients with DM compared with those in non-diabetic patients14,15. In the present study, PD reduction was significantly lower in the T2D group than in the control group. Currently, the mechanisms for impaired healing include hyperglycemia-induced low-level inflammation5, alteration of the immune response7,46 and angiogenesis6, and oxidative stress47. The association between insulin resistance and periodontitis have been proposed48,49. Hyperglycemia worsens wound healing by exacerbating Akt phosphorylation in the insulin signaling pathway9 and oral administration of metformin restores gingival wound healing13. Recently, Shinjo et al. demonstrated that insulin signaling enhances C-X-C motif chemokine ligand 1 (CXCL1) production and modulates neutrophil mobilization7.
In the present study, the expression of PPAR signaling in gingiva was compared in participant with T2D who had shown significantly less PD reduction after non-surgical periodontal therapy compared to patients without DM. A decreased expression of PPARα and γ in the gingiva were observed in the periodontally involved site in participants with T2D. This finding suggests that PPAR signaling may be suppressed by DM in the periodontally involved tissue. In this study, gingival connective tissue, except for the outer keratinized epithelium, was collected from both periodontally involved and healthy sites. Previous studies investigating gingival gene expression have not strictly defined the parts of the gingiva for clinical sample. As inflammatory and non-inflammatory regions show different clusters of gene expression between the epithelium and connective tissue50, the site-specific sampling seems reasonable to assess specific gene expression in the connective tissue rather than including whole gingival epithelium. The expression of PPARα and γ in periodontally involved sites was significantly decreased in participants with T2D compared to those without DM, while no significant differences were observed in periodontally healthy sites. It is quite possible that decreased PPAR signaling may be involved with the adverse effects of periodontal disease in patients with DM.
This study has several limitations. First, the principal analysis in vivo experiments were performed only one day after ligature removal when upregulated inflammatory cytokines were suppressed. On the other hand, in the clinical study, inflammation was not fully controlled in the remaining periodontal pockets, even after non-surgical therapy. The inflammatory levels in the clinical samples can be different pathologically compared to the periodontal tissues evaluated in the in vivo model. Since different gene expression patterns might be observed in the respective stages of healing, further analysis of the various stages of periodontal healing will provide more detailed information. Second, we observed decrease PPAR expression in the tissues; however, the results require further validation by in vitro and in vivo experiments addressing the intracellular signals and protein expression. Finally, due to the small sample size in the clinical study, the association between PPAR expression levels and the progression of DM and periodontal disease could not be statistically analyzed. Thus, further studies with larger sample sizes and more comprehensive analyses are required to confirm these results.