COMMD7在脑低级别胶质瘤发生发展中作用机制的生物信息学分析

Bioinformatics analysis on mechanism of COMMD7 in occurrence and development of brain low-grade glioma

目的:通过生物信息学方法分析COMM域包含蛋白质7 (COMMD7)的表达水平与脑低级别胶质瘤(LGG)患者预后之间的关系,以期寻找脑LGG新的潜在生物标志物,并建立预后预测模型,为患者预后预测及个性化治疗提供依据。方法:从UCSC基因组数据库下载523个脑LGG样本和1 152正常样本的数据。采用Mann-Whitney U检验分析COMMD7在脑LGG样本和正常样本中的表达差异,并采用人类蛋白图谱(HPA)数据库进行验证。使用R语言的DESeq软件包对COMMD7低表达组和COMMD7高表达组脑LGG样本进行差异表达基因(DEGs)鉴定,采用R语言的pROC软件包进行受试者工作特征(ROC)曲线分析,采用单因素和多因素Cox回归分析COMMD7表达与脑LGG患者临床病理特征的关系及对脑LGG患者1、3和5年生存率的影响,采用R语言的ggplot2软件包构建森林图,采用R语言的RMS软件包和Survival软件包构建列线图和Calibration预测模型,采用比例风险回归模型分析COMMD7用于区分脑LGG样本和正常样本的诊断价值。采用GEPIA数据库及Oncolnc数据库进一步验证COMMD7与脑LGG患者预后的关系。使用基因本体(GO)功能注释和京都基因与基因组百科全书(KEGG)对DEGs进行功能和通路富集分析,使用基因集富集分析(GSEA)获得DEGs显著富集的基因集,采用TISIDB数据库分析COMMD7表达与脑LGG患者免疫细胞浸润之间的相关性。结果:COMMD7在脑LGG样本中表达明显上调,HPA检测结果显示COMMD7在脑LGG样本中高表达。分析得到了764个DEGs (|log_(2)FC|>1,P<0.05),包括654个上调和110个下调DEGs。基于多因素分析的预测模型显示COMMD7是一个独立的预后因素。ROC分析,COMMD7用于区分癌组织和癌旁组织具有较高的诊断价值[ROC曲线下面积(AUC)=0.835,95%CI:0.816~0.853]。Cox回归分析,COMMD7高表达组脑LGG患者生存率明显低于COMMD7低表达组,COMMD7高表达与LGG患者的预后不良呈明显正相关关系(P<0.001)。GEPIA数据库514个脑LGG样本分析和Oncolnc数据库510个脑LGG样本分析,COMMD7高表达组脑LGG患者的生存率明显低于COMMD7低表达组(P=0.003,P=0.006);GO功能富集分析,上述DEGs主要富集于DNA结合转录激活子活性、RNA聚合酶Ⅱ特异性、增强子序列特异性DNA结合和增强子绑定等方面;KEGG通路分析,DEGs仅富集于癌症中的转录失调通路。GSEA分析,DEGs主要富集于与KEGG数据库中细胞周期(N=1.950,P=0.012)、World press (WP)数据库中细胞周期(N=1.944,P=0.012)和G_(2)/M检查点(N=2.118, P=0.0012)和G_(2)/M DNA损伤检查点(N=1.879,P=0.012)密切相关的基因集,COMMD7高表达与肿瘤密切相关的p53稳定有关(N=1.793,P=0.012),可以激活p53下游通路(N=1.782, P=0.012)和p53信号通路(N=1.762,P=0.012),激活Tp53调控细胞周期基因的转录(N=1.766, P=0.018)。免疫浸润分析,COMMD7表达与活化的CD8+T淋巴细胞、中枢记忆CD8+T淋巴细胞和CD56^(bright)自然杀伤细胞等15种细胞数量呈明显正相关关系(P<0.05),与吲哚胺2, 3-加双氧酶1 (IDO1)、半乳凝素9(LGALS9)和CD244等11个关键的免疫检查点呈明显正相关关系(P<0.05),与CD274 (PD-L1)、激酶插入区域受体(KDR)和带有Ig及ITIM结构域的T淋巴细胞免疫受体(TIGIT)呈明显负相关关系(P<0.05),与CD40、CD276和CD48等8个关键的免疫刺激因子及C-C基序趋化因子配体2(CCL2)、C-C基序趋化因子配体5 (CCL5)和C-X-C基序趋化因子配体9 (CXCL9)等7个免疫趋化因子呈明显正相关关系(P<0.05)。结论:COMMD7的高表达可能是脑LGG患者不良预后的因素、潜在生物标志物和治疗靶点,通过调节癌症中的转录失调通路,激活p53信号通路和p53下游通路相关基因的失调促进脑LGG发生发展。COMMD7可能在LGG患者免疫检查点抑制剂治疗中发挥关键作用。

Objective:To analyze the relationship between the expression level of COMM domaincontaining protein 7(COMMD7)and prognosis of the brain low-grade glioma(LGG)patients by bioinformatics method,and to find new potential biomarkers of brain LGG,and to establish the prognostic prediction model,and to provide the basis for the prognostic prediction and individualized treatment of the patients.Methods:A total of 523 brain LGG samples and 1152 normal samples were downloaded from the UCSC Genome Database.Mann-whitney U test was used to analyze the expression difference of COMMD7 between brain LGG samples and normal samples.The Human Protein Atlas(HPA)Database was used to verify the results.DESeq software package of R language was used to screen the differentially expressed genes(DEGs)in the brain LGG samples in low expression of COMMD7 group and high expression of COMMD7 group;pROC software package of R language was used to perform the receiver operating characteristic(ROC)curve,univariate and multivariate Cox regression analysis were used to analyze the relationship between the COMMD7 expression and clinicopathological features of the brain LGG patients and its effects on the 1,3,and 5-year survival rates of the brain LGG patients;ggplot2 software package of R language was used to construct the forestplot;RMS software package of R language and survival package were used to construct the Nomogram and Calibration prediction model;the proportional risk regression model was used to analyze the diagnostic value of COMMD7 in distinguishing the brain LGG samples from normal samples.GEPIA and Oncolnc Databases were used to further verify the relationship between COMMD7 and prognosis of the brain LGG patients.Gene Ontology(GO)and Kyoto Encyclopedia of Genes and Genomes(KEGG)enrichment analysis were used to analyze the function and pathway enrichment of the DEGs;Gene Set Enrichment Analysis(GSEA)was used to obtain the significantly enriched gene sets of DEG;TISIDB Database was used to analyze the correlation between the COMMD7 expression and immune cell infiltration of the brain LGG patients.Results:The expression of COMMD7 in the brain LGG samples was significantly increased,and the HPA detection results showed that COMMD7 was highly expressed in the brain LGG samples.A total of 764 DEGs were identified(|log_(2)FC|>1,P<0.05),including 654 up-regulated DEGs and 110 down-regulated DEGs.The predictive model based on the multivariate analysis results suggested that COMMD7 was an independent prognosticfactor. The ROC analysis results showed that COMMD7 had a high diagnostic value in distinguishing thecancer tissue from the adjacent tissue[ area under curve( AUC) = 0. 835, 95%CI = 0. 816-0. 853]. TheCox regression results showed that the survival rate of the brain LGG patients in high expression ofCOMMD7 group was significantly lower than that in low expression of COMMD7 group, and the highexpression of COMMD7 was negatively related with poor prognosis of the brain LGG patients(P<0. 01).The analysis results of 514 LGG samples from GEPIA Database and 510 brain LGG samples from OncolncDatabase showed that the survival rate of the brain LGG patients in high expression of COMMD7 groupwas significantly lower than that in low expression of COMMD7 group (P=0. 003, P=0. 006);the GOenrichment analysis results showed that the above DEGs were mainly enriched in the DNA-bindingtranscription activator activity, RNA polymerase Ⅱ -specificity, enhancer sequence specific DNA binding,and enhancer binding and so on;the KEGG enrichment analysis results showed that DEGs were mainlyenriched in the transcriptional misregulation pathway in cancer. The GSEA results showed that DEGs weremainly enriched in the cell cycle in KEGG Database(N=1. 950, P=0. 012),cell cycle in World Press(WP) Database(N=1. 944, P=0. 012), G_(2) /M checkpoints (N=2. 118,P=0. 0012),and G_(2)/M DNAdamage checkpoints (N=1. 879, P=0. 012). High expression of COMMD7 was not only associated withstabilization of P53 closely realted to tumor(N=1. 793,P=0. 012),but also activating P53 downstreampathway(N=1. 782,P=0. 012),and p53 signaling pathway (N=1. 762,P=0. 012), and activating Tp53regulated transcription of cell cycle genes (N=1. 766, P=0. 018). The immune infiltration analysis resultsshowed that the expression of COMMD7 was significantly positively correlated with the numbers of 15kinds of cells including activated CD8+T lymphocytes,central memory CD8+T lymphocyte and CD56^(bright) natural killer cells, and so on(P<0. 05), the expression of COMMD7 was significantly positivelycorrelated with 11 kinds of key immune checkpoints including indoleamine 2, 3-dioxygenase 1 (IDO1),galactin 9 (LGALS9), and CD244, and so on(P<0. 05),and the expression of COMMD7 wassignificantly negatively correlated with CD274 (PD-L1), kinase insert domain receptor (KDR), and Tlymphocytes immunoreceptor with Ig and ITIM domains (TIGIT) (P<0. 05),and the expression ofCOMMD7 was positively correlated with 4 kinds of key immune stimulators including CD40, CD276, andCD48,and 5 kinds of immune chemokines including C-C motif chemokine ligand 2 (CCL2),C-C motifchemokine ligand 5 (CCL5), C-X-C motif chemokine ligand (CXCL9) and so on. Conclusion:The highexpression of COMMD7 may be a factor of the poor prognosis and a potential biomarker and therapeutictarget of the brain LGG patients, and it can promote the occurrence and development of brain LGG byregulating the transcriptional misregulation pathway in cancer, activating p53 signaling pathway and thedysregulation of p53 downstream pathway related genes. COMMD7 may play a key role in the immunecheckpoint inhibitor therapy in the brain LGG patients.