Chronic kidney disease (CKD) is defined as the presence of structural or functional abnormalities in the kidneys persisting for more than 3 months, resulting in adverse effects on overall health. Clinically, CKD is diagnosed when the estimated glomerular filtration rate (eGFR) falls below 60 mL/min/1.73 m² or when the albumin-to-creatinine ratio (ACR) reaches or exceeds 30 mg/g. Current data indicate that the global prevalence of CKD is estimated to range from 10% to 14% (Chen et al., 2019). However, due to the asymptomatic nature of early-stage CKD, the true prevalence may be substantially underestimated. Irrespective of etiology, CKD progresses through a series of molecular mechanisms including apoptosis, chronic inflammation, oxidative stress, fibrosis, and metabolic disturbances, culminating in irreversible loss of nephrons, microvascular damage, end-stage renal disease (ESRD), and premature mortality (GBD Chronic Kidney Disease Collaboration, 2020; Ruiz-Ortega et al., 2020). Projections indicate that by 2040, CKD will rank as the fifth leading cause of death globally (Foreman et al., 2018), posing a significant challenge to public health systems and necessitating heightened awareness and substantial investment in research (Francis et al., 2024; Jadoul et al., 2024).

KF represents a ubiquitous pathological progression and ultimate manifestation of chronic kidney disease, characterized by glomerular sclerosis, tubular atrophy, chronic interstitial inflammation, fibrogenesis, and vascular rarefaction (Yuan et al., 2022). Researches has elucidated that the pathogenesis of renal fibrosis is multifaceted, involving interactions among various cell types including mesenchymal cells, immune cells, and specific tubular epithelial cells (Li et al., 2022; Kuppe et al., 2021), as well as multiple signaling pathways such as Notch, Wnt, and Hedgehog (Huang et al., 2023). Moreover, with advancements in epigenetic studies, the mechanisms underlying the interaction between aberrant gene expression and environmental changes have been progressively uncovered, emerging as a novel focus for the prevention, diagnosis, and therapeutic targets of renal fibrosis (Feinberg, 2018). Despite some anti-fibrotic agents (Huang et al., 2023), including RAS blockers, SGLT2 inhibitors, vasopressin receptor 2 antagonists, and non-steroidal anti-mineralocorticoids, being capable of delaying the progression of chronic kidney disease to a certain extent, they are unable to reverse the established pathological outcomes of renal fibrosis.

Lipid metabolism encompasses a series of biochemical processes, including synthesis, degradation, transport, and storage of lipids within the body. These processes are crucial for maintaining cell membrane integrity, energy storage, and signal transduction. Disruptions in lipid metabolism have been implicated in the pathogenesis and progression of various diseases, such as cancer (Terry and Hay, 2024), cardiovascular and cerebrovascular disorders (Sakers et al., 2022), and non-alcoholic fatty liver disease (Scorle et al., 2022). Research has demonstrated that excessive lipid accumulation and lipid-induced toxicity are frequently associated with CKD and fibrosis (Chen et al., 2022). This pathological process can be effectively mitigated by enhancing fatty acid oxidation in renal tubular epithelial cells or inhibiting fatty acid transporters (Li et al., 2025; Chen et al., 2020). Consequently, an in-depth investigation into the specific mechanisms of lipid metabolism in KF and the exploration of potential intervention targets could facilitate the development of novel therapeutic strategies, ultimately benefiting patients with CKD. However, current research predominantly focuses on analyzing differences in lipid metabolism markers in KF or examining it as a complication of other diseases, while studies targeting lipid metabolism pathways specifically for diagnosing or treating KF remain relatively limited.

Based on this, we employed bioinformatics tools to identify potential targets associated with lipid metabolism that influence the occurrence and progression of KF. Initially, three transcriptome datasets were retrieved from the GEO database for analysis. After identifying DEGs using sva, we intersected these DEGs with a previously reported set of 760 lipid metabolism-related genes to obtain LMDEGs. Subsequently, GO functional analysis and KEGG pathway enrichment analysis were performed on the LMDEGs. Importantly, through various machine learning algorithms, we identified six biomolecules with diagnostic significance: UGCG, SERPINA6, OSBPL6, INPP5J, PNPLA3, and GK, which were subsequently validated. To further elucidate the role of these six core lipid metabolism-related genes in KF development, we conducted immune infiltration analysis, gene correlation analysis, REACTOME pathway enrichment analysis, and constructed their upstream regulatory networks. Additionally, based on these six core genes, unsupervised clustering was performed using ConsensusClusterPlus on all samples, resulting in three distinct clusters, for which pathway scores were calculated. In conclusion, our study has revealed lipid metabolism genes implicated in the progression of renal fibrosis, which may serve as potential targets for guiding clinical diagnosis and treatment.

The incidence of CKD and the overall progression rate to ESRD have been rising annually (Romagnani et al., 2025). Tissue fibrosis, a common and critical pathological pathway leading to ESRD, represents an ideal focus for clinical and scientific research teams aiming to address the challenges of delaying or reversing its progression (Miguel et al., 2024). To date, current treatments primarily aim to slow the progression of KF, manage the underlying disease and associated complications (Ruiz-Ortega et al., 2022), while no specific drug has been developed to reverse this process (Chen et al., 2019). Research has demonstrated that fatty acid oxidation, which generates over three times the energy produced by aerobic glucose metabolism, serves as the predominant energy-obtaining pathway for renal tubules (Noels et al., 2021; Mitrofanova et al., 2023). However, in CKD, fatty acid oxidation is reduced, resulting in diminished energy acquisition, fat accumulation, and subsequent infiltration of inflammatory cells along with oxidative damage (Lee et al., 2024; Afshinnia et al., 2018). Therefore, fatty acid metabolic disorders are crucial in the progression of KF, and targeted modulation of lipid metabolism holds promise as a novel strategy for reversing this condition. A deeper investigation into its underlying mechanisms will not only provide a robust theoretical foundation for the development of specific drugs but also facilitate breakthroughs in clinical treatment.

However, there are currently no well-defined targets or established therapeutic mechanisms for KF in the context of lipid metabolism (Chen et al., 2022; Wei et al., 2023). Consequently, we performed a bioinformatics analysis to investigate the correlation between lipid metabolism and KF. The analysis revealed that in KF tissues, a total of 54 LMDEGs exhibited differential expression, including 14 upregulated genes such as UGCG and 40 downregulated genes such as INPP5J. Further investigation demonstrated that these genes are closely associated with fatty acid oxidation, lipid synthesis, and immune-inflammatory responses. Their aberrant expression not only disrupts energy metabolism but also facilitates immune cell infiltration, thereby exacerbating the progression of KF. Using the top five machine-learning algorithms, the six core LMDEGs were identified. Subsequent ROC analysis of these genes yielded AUC values exceeding 0.7, indicating high diagnostic and predictive accuracy.

Lipid metabolism disorders influence multiple critical pathways in KF, including extracellular matrix deposition and inflammatory responses (Qu and Jiao, 2023), sharing analogous mechanisms with fibrotic diseases such as diabetic nephropathy and hypertensive nephropathy (Kurano et al., 2023; Tanaka et al., 2022). A total of six core LMDEGs were identified: UGCG, SERPINA6, OSBPL6, INPP5J, PNPLA3, and GK. These genes exhibit close associations with the infiltration of immune cells, such as mast cells, activated dendritic cells, type 1 T helper cells, and natural killer T cells, and participate in the activation of immune signaling pathways, including interferon alpha/beta signaling, immunoregulatory interactions between lymphoid and non-lymphoid cells, and cytokine signaling in the immune system. This suggests that abnormal expression of these core LMDEGs promote the infiltration of inflammatory cells and extracellular matrix deposition, thereby accelerating the progression of KF. Literature supports that similar mechanisms are observed in other fibrotic diseases, such as diabetic nephropathy, hepatofibrosis (Horn and Tacke, 2024; Zhong et al., 2024). Notably, five core genes—SERPINA6, OSBPL6, INPP5J, PNPLA3, and GK—are negatively correlated with cells and pathways involved in the inflammatory response, whereas UGCG demonstrates a positive correlation trend, implying its potentially complex role in modulating the inflammatory response. Considering inflammatory responses, we hypothesize that UGCG exerts antagonistic effects or forms feedback regulatory mechanisms with one or more of the other five inflammatory suppressor molecules, which contributes to the worsening of inflammatory reactions during KF progression.

SERPINA6 is the gene encoding corticosteroid-binding globulin (CBG), which plays a crucial role in maintaining corticosteroid homeostasis in the plasma. Research indicates that CBG influences the negative feedback regulation of glucocorticoid receptors by modulating corticosterone availability in the adrenal glands of rodents. This function supports the normal growth, development, and functional integrity of the adrenal glands in females, highlighting its significance in sexual dimorphism (Toews et al., 2022; Yasmine Nei and rijnck, 2022). In the context of inflammation, CBG serves as a link between lipid metabolism dysregulation and inflammatory responses. Mice deficient in KLF15 exhibit a marked decrease in CBG levels, reduced plasma corticosteroid binding capacity, and increased mortality under inflammatory stress conditions, such as lipopolysaccharide (LPS) exposure. These abnormalities can be fully corrected by CBG supplementation, suggesting that KLF15 primarily contributes to the regulation of plasma corticosteroid homeostasis and inflammatory responses through its modulation of CBG expression (Zhen et al., 2022). INPP5J is a critical tumor suppressor that primarily inhibits tumor cell proliferation through the regulation of cellular signaling pathways. Research has demonstrated that both the mRNA and protein expression levels of INPP5J are markedly reduced in ovarian cancer (Zhu et al., 2015), breast cancer (Toker and Rameh, 2015), and pancreatic cancer (Zhou et al., 2024). Targeting and modulating INPP5J expression have been shown to effectively suppress tumor cell growth, indicating its potential therapeutic significance in cancer treatment. Herold, C et al. performed a family-based genome-wide association analysis and meta-analysis involving approximately 3,500 individuals. Utilizing a multivariate phenotype that integrates disease status and age at onset and identified a significant association between genetic variation in OSBPL6 (rs1347297) and the risk of Alzheimer’s disease (AD). These results underscore the potential involvement of lipid metabolism in the pathogenesis of AD (Herold et al., 2016). Metabolic-associated fatty liver disease (MAFLD) is closely related to PNPLA3 gene variations. The risk of developing ESRD in MAFLD patients is more than twice that of non-MAFLD patients (Chen et al., 2023). Studies (Targher et al., 2019)have shown that children and adolescents with the PNPLA3 rs738409 G/G genotype have significantly lower eGFR than those with G/C and C/C genotypes, and their 24-h urine protein excretion also significantly increases. These changes are independent of the severity of liver disease. Therefore, PNPLA3 gene variations not only increase the risk of ESRD in patients with MAFLD, but also may promote renal lipid accumulation and fibrosis, directly participating in the process of kidney injury (Chen et al., 2023; Mantovani and Targher, 2024). In this study, the patients with renal fibrosis included may carry a genetic variant of PNPLA3, which could influence the measured AUC value associated with PNPLA3. Therefore, future experimental studies should aim to disentangle the effect of the PNPLA3 variant and clarify its interaction with KF. UGCG is one of the key molecules involved in metabolic remodeling and has become a key therapeutic target in tumors and cardiovascular diseases, driving clinical drug development. After B-cell receptor (BCR) activation, UGCG expression is promoted. UGCG catalyzes the conversion of pro-apoptotic ceramide to anti-apoptotic glucosylceramide, which mediates chemotherapy resistance in chronic lymphocytic leukemia (CLL) (Schwamb et al., 2012). Moreover, in highly malignant melanoma, studies have shown that UGCG significantly inhibits tumor cell apoptosis induced by autophagy-lysosome inhibitors by increasing the level of sphingolipids (Jain et al., 2023). Notably, the absence of UGCG leads to the impairment of β1-adrenergic receptor (β1-AR) endocytosis in cardiomyocytes and disrupts retrograde transport from the endoplasmic reticulum to lysosomes in transgenic mice. These defects result in impaired β-adrenergic signal transduction, decreased myocardial contractility, and ultimately the development of dilated cardiomyopathy and premature death (Andersson et al., 2021). There is evidence suggesting that the expression of UGCG is upregulated in diabetic nephropathy and is associated with the downregulation of kidney-protecting genes. Especially in the condition where Fabry disease and diabetes coexist, the high expression of UGCG may lead to more severe kidney phenotypes. This indicates that UGCG may also play a certain role in kidney fibrosis, but the specific mechanism still requires further study (Sanchez-Niño et al., 2023).

In addition, within the interaction network, UGCG exhibits significant interactions with transcription factors such as PAX5, GATA2, and CUX1, as well as miRNAs including miR-374a and miR-374b. Notably, miR-374b regulates both UGCG and GK simultaneously. It may upregulate one gene (e.g., by promoting its expression) while downregulating the other (e.g., by inhibiting its expression), thereby demonstrating its bidirectional regulatory effects on UGCG and GK. This dual regulation disrupts lipid homeostasis in patients with CKD, either by reducing fatty acid oxidation or accelerating lipid synthesis, which subsequently activates inflammatory responses, promotes extracellular matrix deposition, and exacerbates KF. In glomerular podocyte injury (Luo et al., 2024), lipid accumulation leads to cytoskeleton rearrangement, insulin resistance, mitochondrial oxidative stress, and inflammatory activation. These processes, in turn, promote glomerular sclerosis and fibrosis. The Klotho gene exerts a protective effect on chronic kidney disease by reducing inflammatory responses and improving lipid metabolism (Liu et al., 2024). However, during the progression of CKD to KF, lipid metabolism disorders and inflammatory responses do not have a simple cause-and-effect relationship. Instead, they may present a cyclic relationship, where the dominant role—either lipid metabolism disorder or inflammatory response—alternates at different stages of disease progression and progresses in a synergistic manner. This understanding offers a new direction for the clinical management of chronic kidney disease. Early intervention targeting these core LMDEGs and their regulatory networks could potentially inhibit or delay the fibrotic process, offering novel therapeutic targets for the management of CKD.

From the results of GO and KEGG enrichment analysis of DEGs, we identified multiple biological signaling pathways that are closely associated with CKD and KF. These include cytokine-mediated signaling pathways, collagen-containing extracellular matrices, leukocyte-mediated immunity, and regulation of cell-cell adhesion. These findings are consistent with the current understanding of the pathogenesis of KF (Jia et al., 2025). Notably, pathways such as Phagosome and vesicle lumen suggest that intercellular signal interactions may play a pivotal role in the progression of renal fibrosis, warranting further in-depth investigation (Eirin et al., 2017; Liu et al., 2020). Moreover, the GSEA results demonstrated that DEGs were significantly enriched in pathways associated with immune infiltration, including the T cell activation pathway and B cell receptor signaling pathway. Furthermore, all six core LMDEGs exhibited significant correlations with immune cell infiltration. Specifically, UGCG was primarily enriched in natural killer T cells and natural killer cells, showing a positive correlation with immune signaling pathways such as cytokine signaling in immune system and interferon Alpha/Beta signaling. This suggests that UGCG plays a pro-inflammatory role in the progression of KF. The remaining five core LMDEGs (SERPINA6, OSBPL6, INPP5J, PNPLA3, and GK) displayed negative correlations with immune signaling pathways, such as interferon Alpha/Beta signaling, while being positively correlated with lipid metabolism pathways, including the citric acid (TCA) cycle and respiratory electron transport. These findings indicate that these core LMDEGs may contribute to promoting fatty acid metabolism, correcting lipid metabolic disorders, inhibiting inflammatory cells, and counteracting immune infiltration reactions during KF. Through the analysis of the aforementioned results, we have drawn the conclusion that immune infiltration is intricately involved in the entire progression of KF. Furthermore, genes associated with lipid metabolism exert regulatory effects on immune infiltration (He et al., 2024), indicating a potential correlation between immune infiltration and lipid metabolic disorders, both of which collectively influence the process of KF. This area will constitute one of our key research focuses in the future. Additionally, we have observed certain conclusions from other studies that are inconsistent with our findings. For instance, some pathways were not significantly enriched in our study, which may be attributable to variations in sample selection, experimental conditions, or analytical methodologies and thus warrant further investigation and consideration.

The unsupervised clustering analysis identified three distinct molecular subtypes of KF, each with potential therapeutic implications. Cluster B, marked by elevated UGCG expression and activation of immune-related pathways, may represent an “inflammatory” subtype that could potentially benefit from immunomodulatory therapies. In contrast, the lipid metabolism-dominant signature of Cluster A suggests a possible responsiveness to PPAR agonists or other lipid-modulating agents. The intermediate phenotype of Cluster C might necessitate combination therapies targeting both metabolic and inflammatory pathways. These findings are consistent with the emerging paradigm of precision medicine in KF, where molecular subtyping could inform therapy selection. However, the clinical significance of these subtypes remains unclear without longitudinal data linking them to disease progression or treatment response. Furthermore, the underlying biological mechanisms driving these subtype distinctions warrant further investigation, particularly to determine whether they reflect distinct disease etiologies or different stages along a common pathogenic continuum.

Despite the robust bioinformatics methodologies utilized in this study, several limitations must be acknowledged. First, the findings lack experimental validation via functional assays such as gene knockout or overexpression models, which are essential for establishing causal relationships between the identified LMDEGs (particularly UGCG and SERPINA6) and fibrotic progression. Second, although the machine learning models exhibited strong predictive performance (AUC 0.723–0.774), the relatively small sizes of the validation cohorts (GSE22459: n = 65; GSE65326: n = 22) and potential batch effects, even after sva/limma correction, may constrain their generalizability. Third, the absence of longitudinal clinical data prevents an evaluation of whether the three molecular subtypes demonstrate differential prognostic outcomes or therapeutic responses. Future studies should integrate single-cell RNA sequencing to elucidate immune-metabolic crosstalk at a cellular level and validate subtype-specific treatment strategies in preclinical models.

This study identifies the dysregulation of the lipid metabolism-immune network as a hallmark feature of KF, with UGCG functioning as a central pro-inflammatory hub. Furthermore, genes such as SERPINA6 and OSBPL6, along with other LMDEGs, play modulatory roles in lipid metabolic pathways. The six-gene signature derived from machine learning not only stratifies patients into clinically relevant subtypes—characterized by lipid-dominant versus immune-dominant phenotypes—but also reveals shared transcriptional regulators, including PPAR and the miR-27 family, which may coordinately drive fibrotic progression. These findings provide a foundational framework for the development of precision therapeutics targeting specific components of the LMDEG-immune axis. However, translational applications will require validation in larger, prospectively collected cohorts with matched histopathological and functional data.