Integrative Transcriptomics Reveals a Mitochondria-Related Gene Signature in Periodontitis

Mitochondrial dysfunction contributes to periodontal inflammation and tissue destruction, yet the mitochondria-related transcriptional changes associated with periodontitis remain incompletely characterized. Using an integrative bioinformatics study design with experimental validation in cellular and animal models, we sought to identify a mitochondria-related gene signature in affected periodontal tissues and evaluate selected candidates experimentally. Differential expression analysis and weighted gene co-expression network analysis of GSE16134 were integrated with MitoCarta3.0 annotation. Candidate genes were evaluated using 101 feature-selection and classification combinations across GSE16134, GSE10334, GSE33774, and GSE173078, followed by functional enrichment, pathway-level analysis, and immune-cell deconvolution. Selected candidates were assessed by reverse-transcription quantitative polymerase chain reaction in lipopolysaccharide-stimulated RAW 264.7 macrophages and by immunohistochemistry in a ligature-induced rat periodontitis model. The integrated analysis identified 25 candidate genes. A backward stepwise generalized linear model combined with random forest selected CBR3, CYP24A1, SLC25A45, SSBP1, and TMEM205 and achieved the highest mean area under the receiver operating characteristic curve (AUC = 0.819) across the four datasets. The corresponding AUCs were 0.995 in GSE16134, 0.894 in GSE10334, 0.661 in GSE33774, and 0.727 in GSE173078. The model achieved an AUC of 0.870 in the independent validation cohort GSE223924. The selected genes exhibited distinct expression patterns in affected gingival sites and were associated with pathways involving mitochondrial metabolism, cellular respiration, and inflammatory signaling. Immune-cell deconvolution further revealed associations between the selected genes and multiple B- and T-cell subsets. Experimental analyses consistently reproduced decreased CBR3 and increased CYP24A1 and SSBP1 expression in both the macrophage and rat models. These findings delineate a mitochondria-centered transcriptional landscape associated with periodontitis and prioritize CBR3, CYP24A1, and SSBP1 as candidate molecular markers for further investigation. By integrating cross-dataset transcriptomic analysis with experimental evidence, this study provides a focused basis for subsequent mechanistic investigation and clinical validation of mitochondria-related candidate markers in periodontitis.