基于生物信息学探索肺鳞癌肿瘤微环境及寻找关键基因  

Exploration on the tumor microenvironment of lung squamous cell carcinoma and search of key genes based on bioinformatics

作　　者：朱妍 王玉秀 徐兴祥 许文景 葛辉 凡国华 杨俊俊 吴书清 闵凌峰 Zhu Yan;Wang Yuxiu;Xu Xingxiang;Xu Wenjing;Ge Hui;Fan Guohua;Yang Junjun;Wu Shuqing;Min Lingfeng(Department of Respiratory and Critical Care Medicine,Northern Jiangsu People′s Hospital,Clinical Medical College of Yangzhou University,Yangzhou 225001,China)

机构地区：[1]扬州大学临床医学院,苏北人民医院呼吸与危重症医学科,扬州225001

出　　处：《国际呼吸杂志》2022年第3期200-210,共11页International Journal of Respiration

基　　金：国家自然科学基金(81870033、81800049、81900085)。

摘　　要：目的探究肺鳞癌肿瘤微环境,并寻找影响肺鳞癌肿瘤微环境的关键基因。方法使用R语言ESTIMATE算法计算TCGA数据库中肺鳞癌样本的基质评分、免疫评分与综合评分;CIBERSORT算法计算肺鳞癌22种肿瘤浸润免疫细胞丰度。根据surv_cutpoint函数分别获得3种评分的最佳截断值并分为各自的高评分组与低评分组,进行3种评分的高评分组与低评分组5年限制平均生存时间分析及差异表达分析。将差异表达分析获取的基因定义为差异表达基因,再根据蛋白质相互作用网络和单因素Cox回归分析结果确定关键基因。最后,利用Timer数据库探索关键基因与肿瘤浸润免疫细胞之间的相互关系。结果肺鳞癌基质评分、免疫评分与综合评分均低于癌旁组织,差异均有统计学意义(P值均<0.001)。限制平均生存时间分析显示,随访5年,3种评分的高评分组均较低评分组预后差(P值均<0.05)。肺鳞癌组织中22种肿瘤浸润免疫细胞含量有18种与癌旁组织比较,差异均有统计学意义(P值均<0.05)。差异分析确定了影响肿瘤微环境的972个差异基因,将972个差异基因进行单因素Cox回归分析,确定了135个与生存相关的基因(P值均<0.05),与蛋白相互作用网络中前50个中枢基因进行交叉,获得3个与肿瘤微环境相关且影响肺鳞癌预后的关键基因AGTR2、CXCL2、ADORA1。最后Timer数据库探究显示关键基因表达量与肿瘤浸润免疫细胞具有相关性(P值均<0.05)。结论通过生物信息学探索了肺鳞癌肿瘤微环境,筛选出3个与肺鳞癌预后相关且影响肿瘤微环境的关键基因AGTR2、CXCL2、ADORA1,为肺鳞癌提供了潜在的治疗靶点。Objective To explore the tumor microenvironment of lung squamous cell carcinoma(LUSC)and to find key genes influencing the tumor microenvironment of LUSC.Methods ESTIMATE algorithm based on R Language was applied to calculate the stromal score,immune score and estimate score of LUSC samples in TCGA database,CIBERSORT algorithm was applied to calculate the abundance of 22 types of tumors infiltrating immune cells of LUSC samples in TCGA database.The best cutoff values of scores of three groups were acquired based on the function surv_cutpoint,three groups were divided into high scores group and low scores group.Restricted mean survival time and differential expression analysis on high scores group and low scores group of three scores were implemented.The genes acquired by differential expression analysis were identified as differentially expressed genes(DEGs).Key genes were determined based on protein-protein interaction(PPI)network and univariate Cox regression analysis.Finally,Timer database was applied to explore the relations between three key genes and tumor infiltrating immunocytes.Results The stromal score,immune score and estimate score of LUSC tumor tissues were lower than those of para-carcinoma tissue,with statistical significance difference(all P<0.001).Restricted mean survival time showed that the prognosis of high scores groups of three kinds of scores was poor through five years of follow-up(all P<0.05).Twenty two tumor infiltrating immune cells were found in LUSC tissues,and the difference was statistically significant(all P<0.05).Differential expression analysis identified 972 DEGs influencing the tumor microenvironment.Three key genes involving AGTR2,CXCL2 and ADORA1 relating to the tumor microenvironment and prognosis of LUSC were acquired through univariate Cox regression analysis and protein-protein interaction(PPI)network construction.The exploration on Timer database showed that the expression of key genes was correlated with tumor infiltrating immune cells.Conclusions The tumor microenviron

关 键 词：肺肿瘤 癌 鳞状细胞 肿瘤微环境 癌症基因组图谱 免疫细胞浸润 

分 类 号：Q811.4[生物学—生物工程] R734.2[医药卫生—肿瘤]