GRSF1 在结肠癌组织中的表达及其与预后的关系

目的:本研究旨在讨论富含鸟嘌呤的序列结合因子1（G-rich RNA sequence binding factor 1，GRSF1）在结肠癌组织中的表达水平及其与临床病理特征和预后的关系。方法:从基因表达综合数据库（GEO，包括GSE39582、GSE110225、GSE113513三个数据集)以及癌症基因组图谱(TCGA)中下载相关数据，收集565例结肠癌组织的基因表达谱以及相关临床数据，利用Graphpad prism 5分析GRSF1在结肠癌与癌旁组织中的表达情况，并将其分为GRSF1低表达组(n=396)及GRSF1高表达组(n=169)。利用IBM SPSS Statistics 25对样本中GRSF1基因表达数据及对应的临床信息进行统计分析。计量资料和计数资料的组间比较分别采用t检验和χ2检验; 生存分析采用 log-rank 检验; 生存资料单因素及多因素分析采用Cox比例风险模型。利用基因集富集分析（gene set enrichment analysis，GSEA）预测GRSF1可能的相关功能及基因通路。结果:TCGA、CPTAC及GEO数据库分析提示，与正常组织相比，结肠癌组织GRSF1的mRNA及蛋白表达水平显著高于癌旁组织（P＜0.000 1）。GRSF1 表达水平与T分期(χ2=5.797,P=0.016)、M分期(χ2=7.365,P=0.007)、N分期(χ2=6.664,P=0.010)、临床TNM分期（χ2=4.264,P=0.039）有关，GRSF1高表达组总生存率及无复发生存率显著高于GRSF1低表达组（P=0.031,P=0.039）。Cox单因素分析结果显示，结肠癌患者预后和GRSF1表达水平(P=0.001)、T分期(P=0.002)、M分期( P=0.000)、N分期(P=0.000)、临床TNM分期（P=0.000）相关; Cox多因素分析结果显示，结肠癌患者预后与GRSF1表达水平(P=0.033)、T分期(P=0.012)、M分期(P=0.000)、临床TNM分期（P=0.023）相关。基因相互作用分析热图提示，与GRSF1正相关的基因中，EIF4E（r=0.854 2）、LARP1B(r=0.847 6)、ABCE1(r=0.843 0)、SRP72(r=0.839 5)、MRPL1(r=0.829 3)关联较为密切(P＜0.000 1,FDR<0.000 1)；与GRSF1负相关的基因中，RAB11B(r=-0.727 4)、E4F1(r=-0.715 7)，KLHL36(r=-0.710 9)、CORO1B(r=-0.709 5)、WDR24(r=-0.704 0)关联较为密切(P＜0.000 1,FDR<0.000 1)。基因富集分析结果显示，GRSF1高表达样本富集于核糖体构成（P=0.000，FDR=0.000）、RNA催化活性（P=0.000，FDR=0.000）、氧化磷酸化（P=0.000，FDR=0.000）等基因集。结论:GRSF1高表达与结肠癌患者预后不良相关，可作为结肠癌患者判断预后、早期诊断的潜在生物标志物以及早期治疗的分子靶点。

Expression of GRSF1 in colon cancer tissue and its correlation with prognosis

Objective:To explore the correlation between G-rich RNA sequence binding factor 1(GRSF1) expression level and clinical pathological features as well as the prognosis of colon cancer.Methods:From comprehensive Gene Expression Omnibus(GEO) and The Cancer Genome Atlas (TCGA) data were download.Gene expression profiles and corresponding clinical data of 565 colon cancer tissues were collected.The expression of GRSF1 in colon cancer tissues and adjacent tissues was analyzed using Graphpad Prism 5,and divided into low-GRSF1 expression group (n=396) and high-GRSF1 expression group (n=169).The GRSF1 gene expression data and corresponding clinical information in the samples were analyzed using IBM SPSS Statistics 25.Comparison of measurement and enumeration data between groups was performed by t test and Chi-square test,respectively.Survival analysis was performed using log-rank test.Univariate and multivariate analyses of survival data were performed using Cox proportional hazards models.Gene Set enrichment analysis (GSEA) was used to predict GRSF1 related functions and gene pathways.Results:TCGA,CPTAC and GEO database analysis indicated that compared with normal tissues,the expression levels of GRSF1 in patients with colon cancer were significantly higher than those in adjacent tissues (P＜0.000 1).GRSF1 expression level was associated with T stage (χ2=5.797,P=0.016),M stage (χ2=7.365,P=0.000 7),N stage (χ2=6.664,P=0.010),and clinical TNM stage (χ2=4.264,P=0.039).The overall survival rate and relapse-free survival rate of the high GRSF1 expression group were significantly higher than those of the low GRSF1 expression group (P=0.031,P=0.039).Cox univariate analysis showed that the prognosis of colon cancer patients was correlated with GRSF1 expression level (P=0.001),T stage (P=0.002),M stage (P=0.000),N stage (P=0.000) and TNM stage (P=0.000).Cox multivariate analysis showed that the prognosis of colon cancer patients was correlated with GRSF1 expression level (P=0.033),T stage (P=0.012),M stage (P=0.000),and clinical TNM stage (P=0.023).Gene interaction analysis suggested that among the gene positively correlated with GRSF1,EIF4E (r=0.854 2),LARP1B(r=0.847 6),ABCE1(r=0.843 0),SRP72(r=0.839 5) and MRPL1(r=0.829 3) were closely correlated(P＜0.000 1 FDR<0.000 1).Among the gene negatively correlated with GRSF1,RAB11B(r=-0.727 4),E4F1(r=-0.715 7),KLHL36(r=-0.710 9),CORO1B(r=-0.709 5) and WDR24(r=-0.704 0) are closely correlated (P＜0.000 1 FDR<0.000 1).Gene enrichment analysis results showed that GRSF1 expression was highly enriched in structural constituent of ribosome (P=0.000,FDR=0.000),catalytic activity,acting on RNA(P=0.000,FDR=0.000),ribosome (P=0.000,FDR=0.000),oxidative phosphorylation (P=0.000,FDR=0.000),and other gene sets.Conclusion:High expression of GRSF1 is associated with poor prognosis in colon cancer patients,and can also be used as a potential bio-marker for prognosis,early diagnosis and potential molecular target for early treatment in colon cancer patients.