Unravelling gut-kidney axis alterations in BPA-related CKD: Integrating network toxicology and machine learning with...

Chronic kidney disease (CKD) has a complex pathogenesis. Exposure to bisphenol A (BPA) is closely associated with its occurrence and development. Although it is known that BPA can affect the gut microbial community and that the gut-kidney axis serves a crucial role in CKD, the molecular mechanism by which BPA mediates CKD via the gut microbiota and its metabolites still remains uncertain. Our study aims to systematically explore potential molecular mechanisms underlying the association between BPA exposure and CKD involving the gut‑kidney axis, and to screen core targets.

In this study, we integrated multi-omics approaches to investigate the underlying potential mechanism. Firstly, the physicochemical properties of BPA were evaluated using SwissADME. A systematic review and 16S rRNA sequencing were conducted to investigate the influence of BPA exposure on the gut microbiota. Subsequently, the gutMGene, SEA, SwissTargetPrediction (STP), Comparative Toxicogenomics Database (CTD) databases were used to screen and predict the metabolites generated by the gut microbiota and their related targets. At the same time, genes related to CKD were extracted from GSE104954. The core genes were determined by taking the intersection of them, and functional enrichment and PPI network analysis were conducted. To further optimization, 130 machine learning algorithms were used to identify key genes, and the CIBERSORTx algorithm was utilized to analyze their correlation with immune cell infiltration. Molecular docking was employed to evaluate the binding affinities of BPA-induced gut microbiota metabolites with the core targets. Finally, the expression of target proteins was verified through both in vivo and in vitro experiments.

Based on the results of the systematic review and 16S rRNA sequencing, we found that BPA exposure significantly reduced the abundance of Lactobacillus and decreased the levels of its corresponding metabolites, butyrate and trimethylamine-N-oxide. A total of 228 genes related to gut microbiota metabolites were identified. 574 differentially expressed genes (DEGs) related to CKD were identified, and 22 intersecting genes were obtained. These genes were further refined by machine learning to seven key target genes: ALB, CCL2, COL1A1, MYC, FOS, PCK1, and APOC3. These seven core genes are mainly implicated in inflammatory response, lipid‑glucose metabolic homeostasis and extracellular matrix remodeling. They were highly correlated with the levels of immune cell infiltration. Molecular docking confirmed that most core proteins possessed moderate binding capacity toward the target metabolites. In vivo experiments verified the expression of core genes in BPA-induced renal injury. In vitro results further demonstrated that butyrate could modulate the expression of these genes, thereby affecting the progression of CKD.

Our study suggests that BPA may be linked to the onset and progression of CKD, and the gut-kidney axis may play a role in this process. These findings provide a theoretical foundation for the prevention and intervention of CKD.