糖尿病肾病肾小管病变关键基因的预测

Prediction of hub genes in renal tubular lesions in diabetic kidney disease

目的:基于生物信息学探索糖尿病肾病(DKD)肾小管差异表达基因(DEGs)与相关信号通路,结合蛋白互作(PPI)网络分析与比较毒理基因组学数据库(CTD)筛选在DKD肾小管病变中发挥关键作用的基因。方法:选取基因表达公共数据库(GEO)中芯片数据集GSE30529与Karolinska肾脏研究中心RNA-seq数据集,采用R4.03软件的“limma”和“DESeq2”包分析两个数据集共有的DEGs,设定筛选阈值为差异倍数≥2,P<0.05。应用clusterProfiler包进行GO分析和KEGG信号通路富集分析,STRING数据库建立PPI网络,Cytoscape和CTD筛选DKD核心基因。结果:共得到277个DEGs,DEGs的GO分析主要表现于细胞外基质组织,涉及免疫反应、中性粒细胞激活、免疫效应调节等生物学过程。KEGG信号通路分析结果表明,吞噬小体、补体及凝血级联反应、趋化因子信号通路、糖尿病并发症AGE-RAGE信号通路及NF-κB信号通路参与DKD肾小管病变的发生、发展。通过PPI网络及CTD数据库联合分析筛选出CXCL1、CXCL8、CCL5、FN1及EGF共5个关键基因。结论:本研究从转录组水平对两个不同来源数据集进行联合分析,有利于了解DKD肾小管病变发生的潜在分子机制,为进一步研究提供了有意义的线索。

Objective:To identify hub differentially expressed genes(DEGs)and potential signaling pathways of tubular lesions in diabetic kidney disease(DKD)based on multiple bioinformatics analysis and databases,protein-protein interaction(PPI)network analysis and Comparative Toxicogenomics Database(CTD)were used to screen genes that play key roles in renal tubular lesions in DKD.Methods:Microarray dataset GSE30529 from Gene Expression Public Database(GEO)and RNA-seq dataset from Karolinska were selected.DEGs common to the two datasets were screened using"limma"and"DESeq2"package in R4.03,the screening threshold was set as fold change≥2,P<0.05.GO and KEGG signaling pathway enrichment analysis of DEGs were performed using"clusterProfiler"package.PPI network was constructed using STRING database,DKD hub genes were obtained using Cytoscape and CTD.Results:A total of 277 overlapping DEGs were identified,which mainly manifested in extracellular matrix tissues and enriched in humoral immune response,neutrophil activation and regulation of immune effector process.KEGG pathway analysis showed that phagosome,complement and coagulation cascades,chemokine signaling pathway,AGE-RAGE signaling pathway in diabetic complications and NF-κB signaling pathway were involved in occurrence and development of renal tubular lesions in DKD.Five hub genes(CXCL1,CXCL8,CCL5,FN1 and EGF)were obtained from PPI network and CTD.Conclusion:Conjoint analysis of two datasets from different sources at transcriptome level in this study is helpful to understand potential molecular mechanism of renal tubular lesions in DKD and provide meaningful clues for further research.