Development of an miRNA-Array-Based Diagnostic Signature for Periodontitis

Su-Han Jin; Jian-Guo Zhou; Xiao-Yan Guan; Guo-Hui Bai; Jian-Guo Liu; Liang-Wen Chen

doi:10.3389/fgene.2020.577585

Development of an miRNA-Array-Based Diagnostic Signature for Periodontitis

Front Genet. 2020 Dec 16:11:577585. doi: 10.3389/fgene.2020.577585. eCollection 2020.

Authors

Su-Han Jin¹, Jian-Guo Zhou², Xiao-Yan Guan¹, Guo-Hui Bai^{3

4}, Jian-Guo Liu^{3

4}, Liang-Wen Chen⁵

Affiliations

¹ Department of Orthodontics, Affiliated Stomatological Hospital of Zunyi Medical University, Zunyi, China.
² Department of Oncology, Affiliated Hospital of Zunyi Medical University, Zunyi, China.
³ School of Stomatology, Zunyi Medical University, Zunyi, China.
⁴ Special Key Laboratory of Oral Diseases Research, Higher Education Institution, Zunyi, China.
⁵ Hubei-MOST KLOS & KLOBM, Department of Oral Implantology, School and Hospital of Stomatology, Wuhan University, Wuhan, China.

Abstract

Periodontitis progression is accompanied by irreversible alveolar bone absorption and leads to tooth loss. Early diagnosis is important for tooth stability and periodontal tissue preservation. However, there is no recognized miRNA diagnostic signature with convincing sensitivity and specificity for periodontitis. In this study, we obtained miRNA array expression profiles of periodontitis from the Gene Expression Omnibus (GEO) database. After screening for differentially expressed miRNAs, the least absolute shrinkage and selection operator (LASSO) method was performed to identify and construct a 17-miRNA-based diagnostic signature (hsa-miR-3917, hsa-mir-4271, hsa-miR-3156, hsa-miR-3141, hsa-miR-1246, hsa-miR-125a-5p, hsa-miR-671-5p, hcmv-mir-UL70, hsa-miR-650, hsa-miR-497-3p, hsa-miR-145-3p, hsa-miR-141-3p, hsa-miR-210-3p, hsa-miR-204-3p, hsa-miR-203a-5p, hsa-miR-99a-3p, and hsa-miR-30a-3p). Periodontal tissue samples with higher risk scores were more likely to show symptoms of periodontitis. Then, the receiver operating characteristic (ROC) curves were used to assess the diagnostic value of the miRNA signature, which indicated that the optimum cutoff value in periodontitis diagnosis was 0.5056 with an area under the ROC curve (AUC) of 0.996, a sensitivity of 97.3%, a specificity of 100.0% in the training cohort; in the testing cohort, the corresponding values were as follows: an AUC of 0.998, a sensitivity of 97.9%, and a specificity of 91.7%. We next evaluated the efficacy of the signature in differentiating disease subtype and affected range. Furthermore, we conducted functional enrichment analysis of the 17 miRNA-targeted mRNAs, including the regulation of mTOR activity and cell autophagy, Th1/Th2 cell balance and immunoregulation, cell apoptosis, and so on. In summary, our study identified and validated a 17-miRNA diagnostic signature with convincing AUC, sensitivity, and specificity for periodontitis.

Keywords: LASSO method; ROC curve; diagnostic signature; miRNA; periodontitis.