Exploring the common ferroptosis-related genes and molecular mechanisms in periodontitis and systemic sclerosis via integrated bioinformatics and experimental analysis.

BACKGROUND

Accumulating evidence suggests reciprocal risk factors between periodontitis (PD) and systemic sclerosis (SSc). Ferroptosis, an iron-dependent and immune-related form of cell death, has been implicated in both diseases, yet its shared molecular mechanisms remain largely unclear.

METHODS

Bidirectional Mendelian randomization (MR) analysis was first conducted to evaluate the potential causal relationship between PD and SSc. Gene expression datasets for PD and SSc were retrieved from the GEO database, and ferroptosis-related genes were obtained from FerrDb. Differential expression and WGCNA identified common ferroptosis-related differentially expressed genes (Co-FRDEGs) and common ferroptosis-related module genes (Co-FRMGs). Functional enrichment analyses were subsequently performed. The intersection of Co-FRDEGs and Co-FRMGs yielded candidate genes, which were further screened by three machine learning algorithms (LASSO, SVM-RFE, and Random Forest) to identify shared hub genes. Immune infiltration, single-cell RNA sequencing, regulatory network analysis (TF-miRNA), and drug prediction with molecular docking were further performed. In addition, preliminary in vitro experiments were conducted to validate the expression and potential ferroptosis-associated roles of the identified hub genes.

RESULTS

MR revealed an asymmetric causal association between PD and SSc. A total of 28 Co-FRDEGs and 63 Co-FRMGs were identified, and their intersection yielded nine candidate genes. Machine learning analysis predicted FNDC3B and NNMT as shared hub genes, exhibiting good diagnostic performance (AUC >0.75) in both discovery and validation cohorts. Immune infiltration analysis revealed multifaceted immune dysregulation in both diseases, while single-cell analysis confirmed cell type-specific expression of the two hub genes. Regulatory network analysis predicted GTF2E2 and three miRNAs as potential co-regulators. Drug prediction and molecular docking suggested thapsigargin as a potential lead compound. Furthermore, in vitro experiments demonstrated that FNDC3B and NNMT were significantly upregulated in PD- and SSc-like cellular models, and their silencing alleviated ferroptosis-associated cellular injury.

CONCLUSION

This study highlights potential shared ferroptosis-related genes, regulatory networks, and candidate therapeutic compounds associated with PD and SSc, providing new insights into their molecular connections. However, as these findings are largely derived from bioinformatics analyses with preliminary experimental validation, further mechanistic and clinical studies are required to confirm them.