Abstract
INTRODUCTION: Renal fibrosis is a common pathological feature of chronic kidney disease and a major driver of progression to end-stage renal disease, but its molecular mechanisms remain incompletely understood. METHODS: We integrated multi-omics datasets from GEO and published studies, including mRNA, protein, miRNA, and circRNA data from unilateral ureteral obstruction (UUO) models, TGF-β-induced in vitro fibrosis models, and human umbilical cord mesenchymal stem cell-derived exosomes (HucMSC-Exo). Differential expression analysis, functional enrichment, immune infiltration analysis, fuzzy c-means clustering, weighted gene co-expression network analysis, and ceRNA network construction were performed, with selected findings further validated experimentally. RESULTS: We identified stable fibrosis-associated genes and proteins, with metabolic dysregulation emerging as a prominent feature of renal fibrosis. Time-series analysis revealed dynamic transcriptional changes during UUO progression. Comparative analysis showed that in vitro fibrosis models reproduced only part of the in vivo molecular landscape. Immune analyses consistently highlighted macrophages, especially M2-like macrophages, and also suggested a potential role for B cells. In addition, we identified immune-related hub genes and constructed fibrosis-associated ceRNA networks linked to macrophage regulation. Several miRNAs enriched in HucMSC-Exo, particularly miR-30a-5p, were predicted to counteract fibrosis, and exosome treatment alleviated renal injury, macrophage infiltration, and fibrotic marker expression. CONCLUSION: These findings provide a comprehensive view of the molecular and immune landscape of renal fibrosis, clarify key differences between in vivo and in vitro fibrosis models, and suggest potential therapeutic targets for antifibrotic intervention.