基于生物信息学构建绝经后子宫内膜癌患者生存预后模型及验证

Construction and validation of PKD1 involved survival prognosis model of postmenopausal endometrial cancer patients

目的通过构建临床预测模型探究在绝经后子宫内膜癌(EC)患者中发挥关键作用的基因并进行验证,评价其在预测EC患者生存预后中的应用价值。方法选取2022年1月—2023年10月在西北妇女儿童医院妇产科就诊的EC患者45例,通过手术获得EC患者癌组织及癌旁正常组织。从基因表达综合数据库(GEO)和癌症基因组图谱(TCGA)数据库中下载EC样本及正常组织样本的转录组数据,通过加权基因共表达网络分析(WGCNA)确定与EC患者绝经后状态相关的重要基因模块和中心基因,并利用GO和KEGG富集重叠基因所涉及的信号通路。通过在线工具(STRING)分析蛋白质相互作用关系,并在Cytoscape软件中进行可视化,利用Degree算法评价每个节点的重要性,选取排名前5位的节点。构建Logistic回归模型和受试者工作特征(ROC)曲线,分析关键基因在预测EC患者生存情况中的作用。利用免疫印迹法和实时荧光定量PCR(RT-qPCR)验证关键基因在绝经后EC患者组织中的表达。结果WGCNA分析发现,在GSE17025数据集中共识别出17个模块,“红色”模块与EC呈高度正相关(r=0.650,P<0.001),包括1019个基因。在TCGA_UCEC中共鉴定出5个模块,其中“蓝色”模块与EC呈高度正相关(r=0.380,P<0.001),包括336个基因。其中有195个基因重叠,重叠基因主要与细胞周期、能量代谢等有关。通过STRING数据库对KEGG富集的前10位通路中126个基因进行蛋白质相互作用分析,并基于Degree算法将PKD1、ACTB、SRC、CDH1和COL1A1作为潜在的核心基因。通过构建Logistic回归,筛选出PKD1(OR=2.930,P=0.047)、SRC(OR=0.656,P=0.041)和CDH1(OR=0.612,P=0.023)均可有效预测EC患者生存情况。绘制ROC曲线,发现在PKD1、SRC和CDH1中,PKD1对预测EC的生存情况具有较好的诊断价值(AUC=0.634,95%CI 0.540~0.727,P=0.006)。免疫印迹法和RT-qPCR结果显示,癌组织中PKD1蛋白及RNA水平均显著高于癌旁组织(t=10.090,11.257,P<0.001)。结论PKD1可作为影响绝经后EC患者预后生存情况的关键基因,其机制可能是通过调控细胞周期或PI3K-Akt信号通路发挥作用。

Objective To explore the key genes in postmenopausal endometrial cancer(EC)patients by constructing and validating a clinical prediction model,and to discuss its role in predicting the survival prognosis of EC patients.Methods A total of 45 patients with EC in the Department of Obstetrics and Gynecology of Hospital from January 2022 to October 2023 were selected.Tumor tissues and adjacent normal tissues of EC patients were obtained by surgery.The transcriptome data of EC samples and normal tissue samples were downloaded from GEO data and the Cancer Genome Atlas(TCGA)database.The important gene modules and hub genes related to postmenopausal status of EC patients were determined by weighted gene co-expression network analysis(WGCNA),and the signaling pathways involved in overlapping genes were enriched by GO and KEGG.The protein interaction was analyzed by the online tool(STRING)and visualized in Cytoscape software.The importance of each node was evaluated by the Degree algorithm in the CytoHubba plug in,and the top five nodes were selected.The Logistic regression model and ROC curve were constructed to analyze the role of key genes in predicting the survival of EC patients.Finally,Western blotting and real-time fluorescence quantitative PCR(RT qPCR)were used to verify the expression of key genes in postmenopausal EC tissues.Results WGCNA analysis of genes obtained by sequencing in GSE17025 dataset and TCGA_UCEC transcriptome samples identified 17 modules in GSE17025 dataset.The"red"module was highly positively correlated with EC(r=0.650,P<0.001).1019 genes were included.A total of 5 modules were identified in TCGA_UCEC,of which the"blue"module was highly positively correlated with EC(r=0.380,P<0.001),including 336 genes.GO and KEGG enrichment analysis were performed on overlapping genes.Protein interaction analysis was performed on 126 genes in the top ten KEGG enriched pathways through STRING database.PKD1,ACTB,SRC,CDH1 and COL1A1 were selected as potential core genes based on Degree algorithm.By constructing logistic regression,PKD1(OR=2.930,P=0.047),SRC(OR=0.656,P=0.041)and CDH1(OR=0.612,P=0.023)could effectively predict the survival of EC patients.The ROC curve showed that among PKD1,SRC and CDH1,PKD1 had a better diagnostic value for predicting the survival of EC(AUC=0.634,95%CI=0.540-0.727,P=0.006).The results of Western blot and RT qPCR showed that PKD1 protein and RNA levels in cancer tissues were significantly higher than those in adjacent tissues(protein:3.17±1.09 vs.0.98±0.36,t=10.090,P<0.001;RNA:2.15±0.84 vs.0.99±0.31,t=11.257,P<0.001).Conclusion PKD1 may be a key gene affecting the prognosis and survival of postmenopausal EC patients by regulating the cell cycle or PI3K Akt signaling pathway.It provides a basis for further exploring the molecular mechanism of PKD1 in EC,which has important scientific value and significance.