Identification of PANoptosis hub genes driving immune activation and tubulointerstitial injury in diabetic kidney disease by integrative bioinformatics and machine learning

作者：Yu Chen, Feifei Yuan, S. Li, Lerong Liu, Xuankun Peng, Xiangrong Zeng, Siyong Chen, Nianping Liu, Tongfeng Zhao · 发表于：Frontiers in Immunology · 年份：2026 · DOI：10.3389/fimmu.2026.1759781 · 被引用次数：3 · 研究领域：Inflammasome and immune disorders、Ferroptosis and cancer prognosis、GDF15 and Related Biomarkers

Background Diabetic kidney disease (DKD) is characterized by chronic inflammation and immune dysregulation. Multiple programmed cell death pathways contribute to tubulointerstitial injury, but their perturbations, crosstalk, and integrative impact in DKD remain unclear. PANoptosis—a coordinated program integrating pyroptosis, apoptosis, and necroptosis—has emerged as a key mechanism in inflammatory disorders, yet its role in DKD is not defined. Methods We integrated multiple renal tubulointerstitial transcriptomic datasets from DKD and control cohorts to identify differentially expressed genes, followed by functional enrichment analysis. PANoptosis-related gene sets were curated from MSigDB, and pathway crosstalk was evaluated using independent single-cell RNA-seq datasets. Hub genes were prioritized by combining weighted gene co-expression network analysis (WGCNA) with five machine-learning algorithms, and a PANoptosis-related risk score (PRS) was constructed and correlated with clinical parameters and immune infiltration. miRNA–mRNA and transcription factor–hub gene regulatory networks were inferred using ENCORI and hTFtarget, respectively. Druggability of hub genes was assessed using DrugnomeAI, and candidate compounds were retrieved from DGIdb. Key findings were validated in diabetic mouse models. Results Apoptosis, pyroptosis, necroptosis and the integrated PANoptosis program were markedly activated in DKD. At the single-cell level, these pathways were frequently co-acti...