Gene expression dynamics during diabetic periodontitis

Abstract

Diabetes impairs the resolution of periodontal inflammation. We explored pathways altered by inflammation in the diabetic periodontium by using ligatures to induce periodontitis in type-2 diabetic Goto-Kakizaki rats. Ligatures were removed after 7 days, and rats were then treated with TNF inhibitor (pegsunercept) or vehicle alone and euthanized 4 days later. RNA was extracted from periodontal tissue, examined by mRNA profiling, and further analyzed by functional criteria. We found that 1,754 genes were significantly up-regulated and 1,243 were down-regulated by pegsunercept (p < 0.05). Functional analysis revealed up-regulation of neuron-associated and retina-associated gene clusters as well as those related to cell activity and signaling. Others were down-regulated by TNF inhibition and included genes associated with host defense, apoptosis, cell signaling and activity, and coagulation/hemostasis/complement. For selected genes, findings with microarray and rt-PCR agreed. PPAR-α was investigated further by immunohistochemistry due to its anti-inflammatory function and was found to be up-regulated in the gingiva during the resolution of periodontal inflammation and suppressed by diabetes. The results indicate that diabetes-enhanced inflammation both up- and down-regulates genes involved in cellular activity and cell signaling, while it predominantly up-regulates genes involved in the host response, apoptosis, and coagulation/homeostasis/complement and down-regulates mRNA levels of neuron, retina, and energy/metabolism-associated genes.

Figure 1.

Diagram of the approach used for functional analysis of genes regulated by high levels of TNF in the diabetic periodontium. RNA was isolated from diabetic rats treated with vehicle alone or pegsunercept during the resolution of periodontal inflammation and subjected to mRNA profiling. The distribution of the functional gene clusters was determined by Database for Annotation, Visualization and Integrated Discovery (DAVID) and also by gene set enrichment analysis. The approach identified gene clusters that are up- or down-regulated when TNF is inhibited with pegsunercept in diabetic animals and provides insight as to which gene sets are specifically modulated by diabetes-enhanced inflammation.

Figure 2.

Functional distribution of gene clusters up- or down-regulated by TNF inhibition in diabetic specimens. (A) Functional distribution of gene clusters up- or down-regulated as determined by DAVID. RNA was isolated from diabetic rats treated with vehicle alone or pegsunercept during the resolution of periodontal inflammation and subjected to mRNA profiling. Bars represent the distribution of the functional gene clusters determined by Database for Annotation, Visualization and Integrated Discovery (DAVID). Numbers represent the percentage that each category contributed to the total number of up- or down-regulated gene clusters. Gene clusters that are up-regulated when TNF is inhibited with pegsunercept are shown as positive numbers and those that are down-regulated as negative. Only functional annotation clusters with enrichment scores ≥ 1.5 and biological functional ontologies with adjusted p < 0.05 were selected and are displayed in the Fig. (B) Functional distribution of significantly up- or down-regulated gene sets as determined by GSEA. mRNA profiling data described in Fig. 1 were further analyzed by gene set enrichment analysis (GSEA). Only gene sets with FDR < 25% and nominal p value < 0.05 were considered significant and are displayed in the Fig.

Figure 3.

Up-regulation of PPAR-α protein levels in periodontal tissues is reduced by diabetes during the resolution phase of inflammation. Periodontal tissue samples at days 0, 7, and 11 were obtained from normal wild-type rats (blue bars), diabetic rats treated with vehicle (red bars), and diabetic rats treated with TNF-inhibitor (green bars). Immunohistochemistry with antibody specific for PPAR-α was carried out and analyzed according to the following scale: 0, no positive staining in the field; 1, 1 to 10% immunopositive cells with light immunostaining; 2, 11 to 25% immunopositive with moderate immunostaining; 3, 26 to 40% immunopositive with moderate immunostaining; 4, 41 to 60% immunopositive with dark immunostaining; and 5, 61% or higher immunopositive with dark immunostaining. (A) Immunostaining in gingival connective tissue. (B) Immunostaining in periodontal ligament space. Data are mean ± SEM. *Significant difference (p < 0.05). Each group at each time-point included from 6 to 8 animals. WT, Wistar rats; GK + V, GK rats treated with vehicle only; GK+TI, GK rats treated with TNF-specific inhibitor.