CYP1B1 is a potential target of kaempferol in renal fibrosis: a multi‑omics and single‑cell transcriptomics study

Abstract Background Renal fibrosis (RF) is a progressive and irreversible pathological process for which effective treatments are limited. This study explores the molecular targets and mechanisms through which kaempferol (Kae) ameliorates RF, with particular emphasis on cytochrome P450 1B1 (CYP1B1). Methods Using a combination of transcriptomic profiling, weighted gene co‑expression network analysis (WGCNA), multi‑algorithm machine learning [least absolute shrinkage and selection operator (LASSO), support vector machine‑recursive feature elimination (SVM‑RFE), and Random Forest], and single‑cell transcriptomics, we identified the core targets of Kae intervention in RF. shapley additive explanations (SHAP) analysis evaluated gene contributions to the predictive model, while cell‑type identification by estimating relative subsets of RNA transcripts (CIBERSORT) estimated the immune microenvironment. The cell type-specific distribution of core targets was further characterized through single-cell RNA sequencing (scRNA-seq), and the scTenifoldKnk (a machine learning-based virtual gene knockout tool) algorithm simulated the effects of CYP1B1 loss-of-function on fibroblast activation and extracellular matrix (ECM) deposition. Finally, molecular docking was performed to predict the binding conformations of kaempferol with the proteins encoded by the six prioritized genes. Results Six core genes (VCAM1, NFKBIA, SFN, CCL5, CYP1B1, and AXL) were identified from 25 Kae-RF-related genes, with CYP1B1 as an upstream regulatory hub. Immune infiltration analysis revealed increased plasma cells and eosinophils, alongside decreased M0/M1 macrophages and activated NK cells. Single-cell analysis localized CYP1B1 to smooth muscle cells and fibroblasts, with significant upregulation during late-stage fibroblast-to-myofibroblast differentiation. Virtual knockdown of CYP1B1 upregulated 46 genes associated with fibrosis-related biological processes, including cell adhesion, G protein-coupled receptor (GPCR) signaling, ion transport, and cytoskeletal remodeling. The six-gene diagnostic model achieved area under the curves (AUCs) of 0.889 (training) and 0.759 (external validation). Molecular docking showed favorable Kae–CYP1B1 binding (− 6.5 kcal/mol). Conclusion CYP1B1 may serve as a potential functional target of Kae in RF, likely through modulation of fibroblast activation and ECM remodeling. These findings are hypothesis‑generating and provide a computational basis for future experimental validation of CYP1B1‑targeted strategies in RF.