Identification of m6A-related genes and m6A RNA methylation regulators in pancreatic cancer and their association with survival

Identification of m6A-related genes and m6A RNA methylation regulators in pancreatic cancer and their association with survival
复制标题

胰腺癌中 m6A 相关基因和 m6A RNA 甲基化调节因子的鉴定及其与生存的关系

DOI:
10.21037/atm.2020.03.98
复制
发表时间:
2020-03-01
影响因子:
--
通讯作者:
Hou, Baohua
Hou, Baohua
中科院分区:
医学4区
文献类型:
--
作者:
Geng, Yan;Guan, Renguo;Hou, Baohua

文献摘要

被引文献

相似文献

背景N6-甲基腺苷(m6 A)修饰在肿瘤发生和转移中具有重要的作用,它可以在多个水平上改变基因的表达甚至功能,包括RNA的剪接、稳定性、易位和翻译。本研究旨在对胰腺癌中m6 A RNA甲基化调控因子和m6 A相关基因及其与生存期的关系进行全面的研究。方法采用单变量考克斯回归分析、蛋白质-蛋白质相互作用分析、LASSO考克斯回归分析、风险预测模型、STRING、斯皮尔曼和一致性聚类分析等方法,对15个已报道的m6 A RNA甲基化调控因子和1 393个m6 Avar相关基因进行分析。结果发现283个m6 A RNA甲基化相关基因和4个m6 A RNA甲基化调控因子,包括RNA结合基序蛋白15(RBM 15)、甲基转移酶样蛋白14(胃L14)、脂肪量和肥胖相关蛋白(FTO)、α-酮戊二酸依赖性双加氧酶AlkB同源物5(ALKBH 5),在美国癌症联合委员会(AJCC)分期系统的不同阶段之间存在显著差异。蛋白质间相互作用分析表明,表皮生长因子受体(EGFR)、plectin-1(PLEC)、BLM RecQ样解旋酶(BLM)和波罗样激酶1(PLK 1)与其他基因关系密切,可视为网络中的枢纽基因。LASSO考克斯回归和风险预测模型结果显示,AJCC分期、T和N期、KRAS突变状态和x8 q23. 3 CNV片段突变在高危和低危亚组之间存在显著差异。术后1 ~ 5年的AUC均大于0.7,且逐年增加。最后,我们发现TCGA样本分为k=7的亚组后,KRAS突变状态和AJCC分期在这些组之间存在显著差异。此外,我们确定了四个m6 A RNA甲基化相关基因在这七个亚组中表达显著不同,包括VII型胶原α 1链(COL 7A 1),支链氨基酸转氨酶1(BCAT 1),锌指蛋白596(ZNF 596)和PLK 1。结论本研究系统分析了m6 A RNA甲基化相关基因的表达、蛋白质相互作用、潜在功能及预后价值,为进一步研究RNA m6 A甲基化及其相关基因在胰腺癌中的作用提供了重要线索。
Background N6-methyladenosine (m6A) modification holds an important position in tumorigenesis and metastasis because it can change gene expression and even function in multiple levels including RNA splicing, stability, translocation and translation. In present study, we aim to conducted comprehensive investigation on m6A RNA methylation regulators and m6A-related genes in pancreatic cancer and their association with survival time. Methods Based on Univariate Cox regression analysis, protein-protein interaction analysis, LASSO Cox regression, a risk prognostic model, STRING, Spearman and consensus clustering analysis, data from The Cancer Genome Atlas (TCGA) and the International Cancer Genome Consortium (ICGC) database was used to analyze 15 m6A RNA methylation regulators that were widely reported and 1,393 m6A-related genes in m6Avar. Results We found that 283 candidate m6A RNA methylation-related genes and 4 m6A RNA methylation regulatory factors, including RNA binding motif protein 15 (RBM15), methyltransferase like 14 (METTL14), fat mass and obesity-associated protein (FTO), and α‐ketoglutarate‐dependent dioxygenase AlkB homolog 5 (ALKBH5), differed significantly among different stages of the American Joint Committee on Cancer (AJCC) staging system. Protein-protein interaction analysis indicated epidermal growth factor receptor (EGFR), plectin-1 (PLEC), BLM RecQ like helicase (BLM), and polo like kinase 1 (PLK1) were closely related to other genes and could be considered as hub genes in the network. The results of LASSO Cox regression and the risk prognostic model indicated that AJCC stage, stage T and N, KRAS mutation status and x8q23.3 CNV fragment mutation differed significantly between the high-risk and the low-risk subgroups. The AUCs of 1 to 5 years after surgery were all more than 0.7 and increased year by year. Finally, we found KRAS mutation status and AJCC stage differed significantly among these groups after TCGA samples divided into subgroups with k=7. Moreover, we identified four m6A RNA methylation related genes expressed significantly differently among these seven subgroups, including collagen type VII alpha 1 chain (COL7A1), branched chain amino acid transaminase 1 (BCAT1), zinc finger protein 596 (ZNF596), and PLK1. Conclusions Our study systematically analyzed the m6A RNA methylation related genes, including expression, protein-protein interaction, potential function, and prognostic value and provides important clues to further research on the function of RNA m6A methylation and its related genes in pancreatic cancer.