铁死亡相关基因影响肝细胞癌肿瘤微环境及预后风险研究

Effects of Ferroptosis-Related Genes on Microenvironment and Prognosis in Hepatocellular Carcinoma

目的:通过生物信息学方法,揭示肝细胞癌(hepatocellular carcinoma,HCC)的异质性,探讨铁死亡相关基因与HCC肿瘤微环境(tumor microenvironment,TME)的关系,并构建基于铁死亡相关基因的预后模型,为预防HCC患者复发提供新的思路。方法:基于TCGA数据库和NODE数据库,利用非负矩阵分解聚类分析鉴定分子亚型,通过ESTIMATE算法和CIBERSORT算法用于构建免疫评分,利用Cox回归进行拟合风险评估模型并计算风险得分,并基于风险得分及临床信息构建列线图,利用体外实验验证mRNA的表达水平。结果:HCC的异质性体现为两种不同铁死亡相关的免疫分子亚型,其免疫评分差异显著(P<0.001)。筛选出G6PD、KIF20A、RRM2和TXN四个最佳候选铁死亡相关基因,并进行拟合构建风险评估模型,列线图对HCC患者的复发具有良好的预测能力。体外实验提示四个最佳候选基因均在人肝癌细胞表达水平增高(P<0.05)。结论:HCC可分为两种分子亚型,分子亚型与HCC的复发有关,铁死亡在TME中可能起着关键作用。四个铁死亡相关基因可能是预防HCC复发的潜在因素,预后模型对HCC复发有一定的预测价值。

Objective:To explore the heterogeneity of hepatocellular carcinoma(HCC),and investigate the association between ferroptosis-related genes and tumor microenvironment(TME),then construct a prognosis prediction model based on these genes through bioinformatics methods.The model aims to offer new strategies for preventing recurrence in HCC patients.Methods:Based on TCGA and NODE databases,non-negative matrix factorization clustering was applied to identify molecular subtypes of ferroptosis-related genes.The ESTIMATE and CIBERSORT algorithms were employed to calculate immune scores,while Cox regression analysis was used to develop a risk assessment model and calculate risk scores.A nomogram incorporating risk scores and clinical features was established,and in vitro experiments were conducted to validate mRNA expression levels.Results:HCC was classified into two distinct ferroptosis-related immune molecular subtypes,showing significant difference in immune scores(P<0.001).G6PD,KIF20A,RRM2 and TXN were identified as the optimal candidate genes for the risk assessment model.The nomogram accurately predicts the likelihood of HCC recurrence.In vitro experiments revealed increased expression levels of these four genes in liver cancer cells(P<0.05).Conclusion:HCC can be classified into two molecular subtypes which are associated with HCC recurrence,suggesting a critical role of ferroptosis in the TME.The four identified genes may be potential factors for preventing HCC recurrence,and the prognosis prediction model shows significant potential for predicting recurrence of HCC.