Abstract
BACKGROUND: Kidney fibrosis (KF) represents a critical pathological alteration in the end stage of chronic kidney disease (CKD) and is the ultimate cause of mortality. Lipid metabolism plays a significant role in the pathogenesis of KF. Therefore, biomarkers associated with lipid metabolism will be identified to guide the treatment and management of CKD. METHODS: Three datasets obtained from the GEO database, along with 760 lipid metabolism-related genes sourced from two databases, were utilized to identify lipid metabolism-associated differentially expressed genes (LMDEGs) in KF. Subsequently, we performed GO, KEGG and ssGSEA enrichment analysis to elucidate the characteristics of LMDEGs. Then, machine learning was applied to identify core LMDEGs, Least Absolute Shrinkage and Selection Operator (LASSO) was utilized to construct a diagnostic model, and Receiver Operation Curve (ROC) was operated to evaluate the diagnostic performance. We used unsupervised hierarchical clustering to identify subtypes of KF associated with lipid metabolism and employed Gene Set Variation Analysis (GSVA) to examine differences among clusters. Finally, transcription factor and miRNA regulatory networks upstream of core LMDEGs were constructed using Cytoscape software. RESULTS: We identified 54 LMDEGs and constructed a six core LMDEGs (UGCG, SFRP1A6, OSBPL6, INPP5J, PNPLA3, and GK) predictive model by LASSO regression, achieving area under the curve (AUC) values ranging from 0.723 to 0.774. ssGSEA confirmed that these six core LMDEGs exhibited significant positive or negative correlations with immune cell infiltration. Based on the expression profiles of these core LMDEGs, KF samples were categorized into three distinct subtypes. One subtype is predominantly characterized by enhanced lipid and energy metabolism, another exhibits features of inflammation and immune response activation, while the third displays an intermediate pattern between the two extremes. Moreover, the regulatory network of these core LMDEGs shared several common transcription factors, suggesting a potential interplay between lipid metabolism and immune responses in the pathogenesis of KF. CONCLUSION: We have identified six core LMDEGs that are significantly associated with KF. Based on this, we have established three distinct clusters related to lipid metabolism in KF, which may provide valuable insights into the treatment and management of CKD.