Identification of a potential diagnostic signature for postmenopausal osteoporosis via transcriptome analysis

Rui Zeng; Tian-Cheng Ke; Mao-Ta Ou; Li-Liang Duan; Yi Li; Zhi-Jing Chen; Zhi-Bin Xing; Xiao-Chen Fu; Cheng-Yu Huang; Jing Wang

doi:10.3389/fphar.2022.944735

Identification of a potential diagnostic signature for postmenopausal osteoporosis via transcriptome analysis

Front Pharmacol. 2022 Aug 29:13:944735. doi: 10.3389/fphar.2022.944735. eCollection 2022.

Authors

Affiliations

¹ Department of Physiology, School of Medicine, Jinan University, Guangzhou, China.
² Department of Orthopedics, The First Affiliated Hospital of Jinan University, Guangzhou, China.
³ Department of Radiology, The First Affiliated Hospital of Jinan University, Guangzhou, China.
⁴ Department of Plastic Surgery, The First Affiliated Hospital of Jinan University, Guangzhou, China.

Abstract

Purpose: We aimed to establish the transcriptome diagnostic signature of postmenopausal osteoporosis (PMOP) to identify diagnostic biomarkers and score patient risk to prevent and treat PMOP. Methods: Peripheral blood mononuclear cell (PBMC) expression data from PMOP patients were retrieved from the Gene Expression Omnibus (GEO) database. Differentially expressed genes (DEGs) were screened using the "limma" package. The "WGCNA" package was used for a weighted gene co-expression network analysis to identify the gene modules associated with bone mineral density (BMD). Least absolute shrinkage and selection operator (LASSO) regression was used to construct a diagnostic signature, and its predictive ability was verified in the discovery cohort. The diagnostic values of potential biomarkers were evaluated by receiver operating characteristic curve (ROC) and coefficient analysis. Network pharmacology was used to predict the candidate therapeutic molecules. PBMCs from 14 postmenopausal women with normal BMD and 14 with low BMD were collected, and RNA was extracted for RT-qPCR validation. Results: We screened 2420 differentially expressed genes (DEGs) from the pilot cohort, and WGCNA showed that the blue module was most closely related to BMD. Based on the genes in the blue module, we constructed a diagnostic signature with 15 genes, and its ability to predict the risk of osteoporosis was verified in the discovery cohort. RT-qPCR verified the expression of potential biomarkers and showed a strong correlation with BMD. The functional annotation results of the DEGs showed that the diagnostic signature might affect the occurrence and development of PMOP through multiple biological pathways. In addition, 5 candidate molecules related to diagnostic signatures were screened out. Conclusion: Our diagnostic signature can effectively predict the risk of PMOP, with potential application for clinical decisions and drug candidate selection.

Keywords: WGCNA; biomarkers; diagnostic signature; network pharmacology; postmenopausal osteoporosis (PMOP).