Identification of Ferroptosis-related Genes for Diabetic Nephropathy by Bioinformatics and Experimental Validation

Abstract

Objective: The present study delves into the exploration of diagnostic biomarkers linked with ferroptosis in the context of diabetic nephropathy, unraveling their underlying molecular mechanisms.

Methods: In this study, we retrieved datasets GSE96804 and GSE30529 as the training cohort, followed by screening for Differentially Expressed Genes (DEGs). By intersecting these DEGs with known ferroptosisrelated genes, we obtained the differentially expressed genes related to ferroptosis (DEFGs). Subsequently, Weighted Correlation Network Analysis (WGCNA) was carried out to identify key modules associated with Diabetic Nephropathy (DN), culminating in the identification of a significant gene. Enrichment analysis and Gene Set Enrichment Analysis (GSEA) were then carried out on the DEFGs and genes linked to the significant gene. To validate our findings, we employed cohorts GSE30528 and GSE43950, utilizing ROC curve analysis to assess diagnostic efficacy for DN, as measured by the area under the curve (AUC). Immune cell infiltration was analyzed and compared between groups using the CIBERSORT algorithm. Bayesian colocalization analysis was performed to examine the co-location of DEFGs and DN. Finally, to validate the hub genes identified, we conducted quantitative real-time polymerase chain reaction (qRT-PCR) experiments in vitro.

Results: FUZ, GLI1, GLI2, GLI3, and DVL2 were identified as the hub genes. Functional enrichment analysis demonstrated that ferroptosis and immune response play an important role in DN. ROC analysis showed that the identified genes had good diagnostic efficiency in DN. The results of the immune infiltration analysis showed that there may be crosstalk between ferroptosis and immune cells in DN. Bayesian co-localization analysis revealed the genetic correlation between the hub genes and DN. The outcomes of the qRT-PCR analyses corroborated the reliability of the identified hub genes as robust molecular markers for targeted therapy in DN.

Conclusion: The interplay between immune inflammatory reactions and ferroptosis emerges as a crucial pathogenic mechanism, offering novel insights into the molecular therapy of DN. Furthermore, the identification of FUZ, GLI1, GLI2, GLI3, and DVL2 as potential targets holds promise for future therapeutic interventions aimed at treating DN.

Keywords: Bayesian co-localization analysis; Bioinformatics; diabetic nephropathy; experimental validation.; ferroptosis; immune-inflammatory reaction.