目的 基于公共数据库探究亚洲人群肝细胞癌(hepatocellular carcinoma,HCC)与细胞自噬的特征关系.方法 从PubMed上文献检索收集 232 个自噬相关基因(autuphagy-related genes,ARGs).从癌症基因组图谱(The Cancer Genome Atlas,TCGA)和国际癌症基因组联盟(International Cancer Genome Consortium,ICGC)数据库分别纳入 148 例亚洲人种HCC样本(TCGA队列)和 228 例HCC样本(ICGC队列).采用Kaplan-Meier法在TCGA队列中分析得到 24 个对HCC患者具有预后价值的ARGs.对TCGA队列采用聚类分析鉴定ARGs分子亚型.对TCGA队列应用LASSO-Cox回归方法建立基于ARGs的风险评分模型,采用ICGC队列作验证.采用单因素和多因素Cox回归分析筛选出具有统计学意义(P<0.01)的指标,用以构建包含ARGs风险评分和临床因素的列线图预测HCC患者的生存预后.采用受试者工作特征(receiver operating characteristic,ROC)曲线评估分析列线图的预测性能.分别采用WebGestalt数据库、TIMER数据库和相应R语言包在ARGs风险评分亚组中分析免疫浸润、富集途径和基因组改变状态.结果 基于ARGs的HCC分子亚型[聚类A(Cluster-A)和聚类B(Cluster-B)]患者的总生存期比较,差异具有统计学意义(P<0.01).使用LASSO-Cox回归方法筛选出特征性的 5 个预后相关ARGs[5-aminoimidazole-4-carboxamide ribonucleotide formyltransferase/inosine monophosphate cyclohydrolase(ATIC)、eukaryotic translation initiation factor 2 subunit alpha(EIF2S1)、histone deacetylase 1(HDAC1)、Rac family small GTPase 1(RAC1)和autophagy related 3(ATG3)]构建的风险评分模型将TCGA队列患者分为高风险组和低风险组.两组HCC患者总生存期比较,差异具有统计学意义(P<0.01).单因素和多因素Cox分析显示,ARGs风险评分和TNM肿瘤分期与HCC患者总生存期相关(均P<0.01).纳入ARGs风险评分和TNM肿瘤分期的列线图用于预测HCC患者1、3和5年总生存状态具有优秀的预测能力.免疫浸润分析显示,与低风险组比较,高风险组在多数固有免疫细胞和适应免疫细胞浸润水平更高(均P<0.05).基因本体论(Gene Ontology,GO)和京都基因与基因组百科全书(Kyoto Encyclopedia of Genes and Genomes,KEGG)分析显示,高风险组上调基因主要富集于白细胞迁移、细胞运动和分化等多种通路和途径(均P<0.05).基因组改变差异分析显示,高风险组患者tumor protein p53(TP53)突变频率更高(P=0.003),低风险组患者lysine methyltransferase 2D(KMT2D)突变率更高(P=0.031).结论 本研究建立并验证了基于ARGs的风险评分模型,为预测亚洲人种HCC患者的预后状态提供了一项有效的参考指标.
Identification of prognostic signature of autophagy-related genes in Asian patients with hepatocellular carcinoma based on RNA-seq in public databases
Objective To explore the characteristic relationship between autophagy and hepatocellular carcinoma(HCC)in Asian popu-lation based on public databases.Methods A total of 232 autophagy-related genes(ARGs)were collected by literature search.From The Cancer Genome Atlas(TCGA)and the International Cancer Genome Consortium(ICGC)databases,148 Asian HCC samples(TCGA cohort)and 228 HCC samples(ICGC cohort)were included.Kaplan-Meier method was used to screen 24 ARGs with prognostic value for HCC in the TCGA cohort.The molecular subtypes of ARGs were identified by cluster analysis in the TCGA cohort.The LASSO-Cox regres-sion method was used to establish an ARG-based risk scoring model for the TCGA cohort,and the ICGC cohort was used for verification.Univariate and multivariate Cox analyses were used to select statistically significant indicators(P<0.01),which were used to construct a prognotic nomogram based on ARG risk score and clinical factors for HCC.The receiver operating characteristic(ROC)curve was used to evaluate the prediction performance of the nomogram.The WebGestalt and TIMER databases and corresponding R language packages were used to analyze immune infiltration,enrichment pathways,and genome alteration status in ARG risk score subgroups.Results The overall survival of HCC subtypes(Cluster-A and Cluster-B)based on ARGs was significantly different(P<0.01).Five prognostic ARGs were iden-tified by LASSO-Cox regression,including 5-aminoimidazole-4-carboxamide ribonucleotide formyltransferase/inosine monophosphate cyclohydrolase(ATIC),eukaryotic translation initiation factor 2 subunit alpha(EIF2S1),histone deacetylase 1(HDAC1),Rac family small GTPase 1(RAC1),and autophagy related 3(ATG3).A risk score model was constructed based on these five ARGs,and divided the TCGA cohort into high-and low-risk groups.The overall survival of HCC patients in these two groups was significantly different(P<0.01).Uni-variate and multivariate Cox analysis showed that ARG risk score and TNM stage were correlated with the overall survival of HCC patients(both P<0.01).The nomogram based on the ARG risk score and TNM stage had excellent performance for predicting the 1-,3-,and 5-year overall survival of HCC patients.Immunoinfiltration analysis showed that,compared with the low-risk group,the high-risk group had high-er levels of most innate and adaptive immune cell infiltration(all P<0.05).Gene Ontology(GO)analysis and Kyoto Encyclopedia of Genes and Genomes(KEGG)analysis showed that the up-regulated genes in the high-risk group were mainly enriched in the pathways of leuko-cyte migration,cell motility and differentiation(all P<0.05).The analysis of differences in genomic alterations showed that patients in the high-risk group had a higher frequency of tumor protein p53(TP53)mutations(P=0.003),and patients in the low-risk group had a higher rate of lysine methyltransferase 2D(KMT2D)mutations(P=0.031).Conclusions This study established and validated the ARG-based risk score model which provided a practical reference index for predicting the prognosis of Asian HCC patients.
国家科技期刊平台
NSTL
万方数据
