Uncovering the Molecular Networks of Extracellular Vesicles in the Pathogenesis of Periodontitis

Objectives

The aim of the study is to explore EV-related genes (ERGs) networks underlying periodontitis by integrating bulk RNA-seq, single-cell RNA-seq, and causal genetic analysis.

Methods

Using GEO and eQTLGen datasets, we identified periodontitis-associated ERGs through SMR and differential gene analysis. Potential therapeutics were predicted via drug repurposing and molecular docking, while machine learning built validated predictive models. Functional enrichment revealed pathogenic pathways, and scRNA-seq explored ERGs in tissue, peripheral blood, and diabetes-comorbid periodontitis cases

Results

SMR prioritised nine causal ERGs (CD34, TLR4, TIMD4, ENPP4, HLA-DRA, HLA-B, ENDOD1, PSMA4 and HSPB1) that were mainly expressed in NK cells and CD8⁺ T cells in the subsequent single cell sequence and deconvolution analysis analysis. Dexamethasone emerged as the most promising drug via binding to HSPB1. The predictive models developed using machine learning achieved robust accuracy (XGBoost: AUC = 0.836) in validation. Key biological pathways included blood coagulation, hemostasis cell, and adhesion molecules, and so on.

Conclusion

Through multiple analyses integration, we defined an ERG network central to periodontitis pathogenesis. The diagnostic model and repurposed drugs offer translational avenues, requiring further experimental validation.

Clinical Relevance

This study identified and prioritised periodontitis-associated ERGs, elucidated their molecular mechanisms, and established a foundation for future functional validation and clinical risk assessment.