南方医科大学学报 ›› 2022, Vol. 42 ›› Issue (12): 1822-1831.doi: 10.12122/j.issn.1673-4254.2022.12.10

• • 上一篇    下一篇

组蛋白赖氨酸去甲基化酶家族在膀胱癌中的表达模式及其潜在作用:基于多组学分析

付小聪,余光创,郭艳芳   

  1. 南方医科大学基础医学院生物信息学系,广东 广州 510515
  • 出版日期:2022-12-20 发布日期:2023-01-12

Expression pattern of the histone lysine demethylase family and its potential role in bladder cancer: a multi-omics analysis

FU Xiaocong, YU Guangchuang, GUO Yanfang   

  1. Department of Bioinformatics, School of Basic Medical Sciences, Southern Medical University, Guangzhou 510515, China
  • Online:2022-12-20 Published:2023-01-12

摘要: 目的 探讨 19 个组蛋白赖氨酸去甲基化酶(KDMs)在膀胱癌多组学中的表达模式与潜在作用。方法 本研究使用UALCAN和GSCALite分析来自TCGA的膀胱癌样本的KDMs的转录表达和甲基化水平、体细胞变异多组学高通量测序数据。使用Kaplan Meier-Plotter和Assistant for clinical bioinformatics探讨KDMs的表达对BLCA样本的预后影响。利用Timer和GSCALite分析KDMs在膀胱癌中的免疫浸润和药物敏感性。结果 分析结果首先揭示了KDMs在膀胱癌多组学中的表达特征,并不是所有KDMs都具有一致的表达模式。转录组数据分析显示KDM1A/1B/2B/4A/4B/5B/5C的表达上调(P<0.05),而KDM3B/6B/7C的表达下调。甲基化水平差异分析显示KDM1A/3B/4A/4B/4C/5A/5B/5C/7B的甲基化水平下调,而KDM7C甲基化水平上调(P<0.05)。相关性分析显示,14个KDMs家族成员的转录水平与甲基化水平呈负相关,其中KDM1A最显著。突变分析显示,KDM6A非同义突变频率最高,突变种类最多,且与其它KDMs 的非同义突变具有互补性。生存分析显示,KDM3A/4C/5D/6A/7B对BLCA患者总生存率具有保护性作用,而KDM3B/5B/5C对BLCA患者无复发生存率具有危险性影响。综合预后模型证实KDM4C/6A/7B具有膀胱癌预后生物标志物的潜在作用,其表达与BLCA患者免疫浸润呈正相关。药物敏感性分析显示,KDM2B/3B/4B/4C/5A与大多数抗癌药物呈负相关,而KDM2B/4B与6种抗癌药物呈正相关(P<0.05)。结论 本研究系统性地揭示了KDMs基因家族在膀胱癌中的高突变互补性、过表达与低甲基化负相关性的特征,并与膀胱癌预后、免疫浸润和药物敏感性密切相关。

关键词: 关键词:膀胱癌;组蛋白赖氨酸去甲基化酶;多组学

Abstract: Objective To investigate the expression patterns of 19 histone lysine demethylases (KDMs) and their role in bladder cancer. Methods In this study, UALCAN and GSCALite were used to analyze the transcriptional expression, methylation level and somatic variation of KDMs in bladder cancer samples from TCGA. Kaplan Meier- Plotter and Assistant for clinical bioinformatics were used to investigate the effect of KDMs expression on the prognosis of BLCA samples. The immune infiltration and drug sensitivity of KDMs in bladder cancer were analyzed by Timer and GSCALite. Results The KDMs did not show consistent expressions patterns in bladder cancer, where the expressions of KDM1A/1B/2B/4A/4B/5B/5C were significantly upregulated while those of KDM3B/6B/7C were significantly downregulated. Methylation data analysis showed that methylation levels of KDM1A/3B/4A/4B/4C/5A/5B/5C/7B were significantly downregulated and that of KDM7C was upregulated. The transcription levels of 14 KDMs had significant negative correlations with their methylation levels, and among them KDM1A showed the strongest correlation. Mutation analysis revealed that KDM6A had the highest frequency of nonsynonymous mutations with the largest variety, and these mutations were complementary to nonsynonymous mutations of the other KDMs. Survival analysis showed that KDM3A/4C/5D/6A/7B were protective for OS while KDM3B/5B/5C adversely affected RFS of BLCA patients. Further comprehensive prognostic modeling confirmed that KDM4C/6A/7B were potential prognostic biomarkers of bladder cancer, and their expressions were positively correlated with immune infiltration in BLCA patients. KDM2B/3B/4B/4C/5A were negatively correlated with the sensitivity to most anticancer drugs, while KDM2B/4B were positively correlated with the sensitivity to 4 anticancer drugs. Conclusion The expression patterns of the KDMs in bladder cancer highlight a high mutation complementarity and a negative correlation between over-expression and hypomethylation level closely related with the prognosis, immune infiltration and drug sensitivity.

Key words: bladder cancer; histone lysine demethylase; multi-omics