基于多组学分析胶质母细胞瘤免疫微环境关键基因的研究  

Research on key genes of glioblastoma immune microenvironment based on multi-omics analysis

作　　者：依日扎提·艾力 努尔斯曼古丽·买买提明 秦虎[1] 王增亮[1] 汪永新[1,3] Yirizhati Aili;Nuersimanguli Maimaitiming;Qin Hu;Wang Zengliang;Wang Yongxin(Department of Neurosurgery,the First Affiliated Hospital of Xinjiang Medical University,Urumqi 830054,China;Department of Hepatobiliary Surgery,Cancer Hospital,Chinese Academy of Medical Sciences,Beijing 100004,China;Key Laboratory of Precision Diagnosis and Clinical Transformation of Nervous System Tumors,Xinjiang Medical University,Urumqi 830054,China)

机构地区：[1]新疆医科大学第一附属医院神经外科中心,乌鲁木齐830054 [2]中国医学科学院肿瘤医院肝胆外科,北京100004 [3]新疆医科大学神经系统肿瘤精准诊疗与临床转化重点实验室,乌鲁木齐830054

出　　处：《中华神经外科杂志》2025年第2期187-194,共8页Chinese Journal of Neurosurgery

基　　金：新疆维吾尔自治区自然科学基金重点项目(2022D01D70);新疆维吾尔自治区自然科学基金青年项目(20231107414);2022年度"青年科研起航"专项基金(2022YFY-QKQN-61)。

摘　　要：目的探讨胶质母细胞瘤(GBM)免疫微环境中的关键基因,并分析其在GBM患者预后预测中的作用。方法获取基因表达综合数据库中的GBM单细胞测序数据,包含了8例样品(4例GBM组织,4例癌旁组织)共91791个单细胞测序结果,结合癌症基因组图谱、FinnGen R10-C3_GBM数据库进行差异基因及孟德尔随机化分析,筛选GBM免疫微环境关键基因。通过CIBERSORT算法研究关键基因免疫学特征,最终通过人类蛋白图库、单样本基因集富集分析(ssGSEA)评分(样本分为高、低ssGSEA评分组)及列线图预测模型分析关键基因在预测GBM预后中的价值。结果单细胞测序数据分析获得了21个细胞亚群。对GBM与癌旁组织的差异基因结合孟德尔随机化分析得到9个差异基因,其中6个关键基因在转录组学水平上表达差异具有统计学意义(均P<0.05),分别为RHO家族相互作用细胞极化调节器2(RIPOR2)、S100钙结合蛋白A6(S100A6)、热休克蛋白家族H成员1(HSPH1)、核受体亚家族4 A组成员1(NR4A1)、核仁蛋白10(NOP10)和细胞色素B-245β链(CYBB)基因。其中,CYBB、NOP10和S100A6基因高表达与M2型巨噬细胞浸润正相关,而与自然杀伤细胞活化及M0型巨噬细胞浸润负相关(均P<0.05);HSPH1及NR4A1基因高表达与M0型巨噬细胞浸润正相关,而与M2型巨噬细胞浸润负相关;RIPOR2基因高表达与中性粒细胞浸润正相关,而与M0型巨噬细胞浸润负相关(均P<0.05);CYBB基因与多种免疫抑制细胞的浸润相关。6个关键基因主要富集在血管生成、活性氧、转化生长因子-β及缺氧相关、胆固醇代谢、凋亡及排斥反应、补体及凝血级联等相关通路。进一步分析显示,CYBB、HSPH1、NOP10、NR4A1及S100A6基因均在GBM细胞中高表达;且与低ssGSEA评分组相比,高ssGSEA评分组的总体生存时间、无进展生存时间及疾病特异性生存时间均下降(均P<0.05)。基于6个关键基因建立的列线图预测模型分析显示,其对GObjectiveTo investigate the key genes in the immune microenvironment of glioblastoma(GBM)and to analyze their roles in predicting the prognosis of GBM patients.MethodsGBM single cell sequencing data in the Gene Expression Omnibus database were obtained,including a total of 91,791 single cell sequencing results from 8 samples(including 4 GBM tissues and 4 para-cancerous tissues).Differential gene and Mendel randomization analysis was performed in combination with The Cance Genorne Atlas and the FinnGen R10-C3_GBM database to screen key genes of GBM immune microenvironment.CIBERSORT algorithm was used to study the immunological characteristics of key genes,and finally the value of key genes in prognosis prediction of GBM was analyzed by human protein library,single sample gene set enrichment analysis(ssGSEA)scores(samples were divided into high and low ssGSEA groups)and the nomogram prediction model.ResultsThe single cell data were analyzed and 21 cell subpopulations were obtained.The differential genes of GBM and adjacent tissues combined with Mendelian randomization analysis obtained 9 differential genes,among which 6 key genes had statistically significant expression differences at the transcriptomic level(all P<0.05),as follows:RHO family interaction cell polarization regulator 2(RIPOR2),S100 calcium-binding protein A6(S100A6),member H of the heat shock protein family 1(HSPH1),member A of nuclear receptor subfamily 4(NR4A1),ribonucleoprotein 10(NOP10),and cytochrome B-245βchain(CYBB)gene.Among them,the high expression of CYBB,NOP10 and S100A6 genes was positively correlated with M2 macrophage infiltration,while it was negatively correlated with natural killer cell activation and M0 macrophage infiltration(all P<0.05).The high expression of HSPH1 and NR4A1 genes was positively correlated with M0 macrophage infiltration,while it was negatively correlated with M2 macrophage infiltration(all P<0.05).The high expression of RIPOR2 was positively correlated with neutrophil infiltration,and while it was negatively cor

关 键 词：胶质母细胞瘤 单细胞分析 孟德尔随机化分析 肿瘤微环境 免疫细胞 

分 类 号：R73[医药卫生—肿瘤]