Integrative multi-omics analysis reveals the PFAS-gut microbiome-FN1 axis in chronic kidney disease: From exposome to therapeutic target

Per- and polyfluoroalkyl substances (PFAS) are persistent environmental pollutants associated with chronic kidney disease (CKD), yet the molecular mechanisms linking PFAS exposure to renal fibrosis remain incompletely understood. We developed a Chemical-Microbiome-Metabolite-Target-Pathway-Disease (C-M-M-T-P-D) network framework integrating differential expression analysis of real GEO transcriptomic data (GSE104954, n = 195 kidney biopsies), gene set enrichment analysis across four pathway databases, two-sample Mendelian randomization with four complementary methods, molecular docking and molecular dynamics simulation of deoxycholic acid (DCA) to fibronectin 1 (FN1), and transcription factor and protein-protein interaction network analyses. FN1 was significantly upregulated in CKD (log2FC = 0.322, FDR = 6.67e-11), accompanied by elevated COL1A2, VIM, and POSTN, whereas TGFBR3 was downregulated (log2FC = -0.746, FDR = 1.67e-14). GSEA identified ECM-receptor interaction (NES = 2.23), interferon gamma response (NES = 2.70), and epithelial-mesenchymal transition (NES = 1.99) as top enriched pathways. MR analysis using real MiBioGen and CKDGen GWAS data (19 SNPs for Akkermansia, 12 SNPs for Desulfovibrio) revealed contrasting findings: Akkermansia to CKD showed no significant genetic association (0/4 methods, all p > 0.30), whereas Desulfovibrio to CKD demonstrated a robust positive genetic association across all four methods (IVW p = 0.008, WM p = 4.70e-04, MR-RAPS p = 1.46e-05, MR-cML p = 2.08e-06), though the significant Egger intercept (p = 0.005) for Desulfovibrio warrants cautious interpretation regarding pleiotropy. Molecular docking and 15-ns MD simulation confirmed stable DCA binding to FN1 (PDB: 2CG7, 1.20 A X-ray; MM/GBSA ΔGbind = -32.4 kcal/mol). The C-M-M-T-P-D network positioned FN1 as the third highest-degree node, bridging metabolite and pathway layers. This integrative multi-omics analysis provides multi-omics evidence supporting the PFAS-gut microbiome-FN1 axis as a hypothesis-generating framework for understanding CKD pathogenesis, supported by transcriptomic, structural, and network-level evidence, with genetic associations warranting further experimental validation.