Zhu Enyi, Zhong Ming, Liang Tiantian, Liu Yu, Wu Keping, Zhang Zhijuan, Zhao Shuping, Guan Hui, Chen Jiasi, Zhang Li-Zhen, Zhang Yimin
The Division of Nephrology, The Sixth Affiliated Hospital, Sun Yat-sen University, Guangdong Provincial Key Laboratory of Colorectal and Pelvic Floor Diseases, Guangzhou, Guangdong, 510000, People's Republic of China.
Biomedical Innovation Center, The Sixth Affiliated Hospital, Sun Yat-sen University, Guangdong Provincial Key Laboratory of Colorectal and Pelvic Floor Diseases, Guangzhou, Guangdong, 510000, People's Republic of China.
J Inflamm Res. 2024 Feb 3;17:693-710. doi: 10.2147/JIR.S440374. eCollection 2024.
Diabetic nephropathy (DN) represents the principal cause of end-stage renal diseases worldwide, lacking effective therapies. Fatty acid (FA) serves as the primary energy source in the kidney and its dysregulation is frequently observed in DN. Nevertheless, the roles of FA metabolism in the occurrence and progression of DN have not been fully elucidated.
Three DN datasets (GSE96804/GSE30528/GSE104948) were obtained and combined. Differentially expressed FA metabolism-related genes were identified and subjected to DN classification using "ConsensusClusterPlus". DN subtypes-associated modules were discovered by "WGCNA", and module genes underwent functional enrichment analysis. The immune landscapes and potential drugs were analyzed using "CIBERSORT" and "CMAP", respectively. Candidate diagnostic biomarkers of DN were screened using machine learning algorithms. A prediction model was constructed, and the performance was assessed using receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA). The online tool "Nephroseq v5" was conducted to reveal the clinical significance of the candidate diagnostic biomarkers in patients with DN. A DN mouse model was established to verify the biomarkers' expression.
According to 39 dysregulated FA metabolism-related genes, DN samples were divided into two molecular subtypes. Patients in Cluster B exhibited worse outcomes with a different immune landscape compared with those in Cluster A. Ten potential small-molecular drugs were predicted to treat DN in Cluster B. The diagnostic model based on PRKAR2B/ANXA1 was created with ideal predictive values in early and advanced stages of DN. The correlation analysis revealed significant association between PRKAR2B/ANXA1 and clinical characteristics. The DN mouse model validated the expression patterns of PRKAR2B/ANXA1.
Our study provides new insights into the role of FA metabolism in the classification, immunological pathogenesis, early diagnosis, and precise therapy of DN.
糖尿病肾病（DN）是全球终末期肾病的主要病因，缺乏有效的治疗方法。脂肪酸（FA）是肾脏的主要能量来源，其代谢失调在DN中经常出现。然而，FA代谢在DN发生和发展中的作用尚未完全阐明。
获取并合并三个DN数据集（GSE96804/GSE30528/GSE104948）。鉴定差异表达的FA代谢相关基因，并使用“ConsensusClusterPlus”进行DN分类。通过“WGCNA”发现DN亚型相关模块，对模块基因进行功能富集分析。分别使用“CIBERSORT”和“CMAP”分析免疫图谱和潜在药物。使用机器学习算法筛选DN的候选诊断生物标志物。构建预测模型，并使用受试者工作特征（ROC）曲线、校准曲线和决策曲线分析（DCA）评估其性能。使用在线工具“Nephroseq v5”揭示候选诊断生物标志物在DN患者中的临床意义。建立DN小鼠模型以验证生物标志物的表达。
根据39个失调的FA代谢相关基因，将DN样本分为两种分子亚型。与A组相比，B组患者的预后较差，免疫图谱不同。预测有10种潜在的小分子药物可治疗B组的DN。基于PRKAR2B/ANXA1创建的诊断模型在DN的早期和晚期具有理想的预测价值。相关性分析显示PRKAR2B/ANXA1与临床特征之间存在显著关联。DN小鼠模型验证了PRKAR2B/ANXA1的表达模式。
我们的研究为FA代谢在DN的分类、免疫发病机制、早期诊断和精准治疗中的作用提供了新的见解。
