Data_Sheet_1_A Novel Mitochondrial-Related Nuclear Gene Signature Predicts Overall Survival of Lung Adenocarcinoma Patients.PDF

Background: Lung cancer is the leading cause of cancer-related death worldwide, of which lung adenocarcinoma (LUAD) is one of the main histological subtypes. Mitochondria are vital for maintaining the physiological function, and their dysfunction has been found to be correlated with tumorigenesis and disease progression. Although, some mitochondrial-related genes have been found to correlate with the clinical outcomes of multiple tumors solely. The integrated relationship between nuclear mitochondrial genes (NMGs) and the prognosis of LUAD remains unclear.
Methods: The list of NMGs, gene expression data, and related clinical information of LUAD were downloaded from public databases. Bioinformatics methods were used and obtained 18 prognostic related NMGs to construct a risk signature.
Results: There were 18 NMGs (NDUFS2, ATP8A2, SCO1, COX14, COA6, RRM2B, TFAM, DARS2, GARS, YARS2, EFG1, GFM1, MRPL3, MRPL44, ISCU, CABC1, HSPD1, and ETHE1) identified by LASSO regression analysis. The mRNA expression of these 18 genes was positively correlated with their relative linear copy number alteration (CNA). Meanwhile, the established risk signature could effectively distinguish high- and low-risk patients, and its predictive capacity was validated in three independent gene expression omnibus (GEO) cohorts. Notably, a significantly lower prevalence of actionable EGFR alterations was presented in patients with high-risk NMGs signature but accompanied with a more inflame immune tumor microenvironment. Additionally, multicomponent Cox regression analysis showed that the model was stable when risk score, tumor stage, and lymph node stage were considered, and the 1-, 3-, and 5-year AUC were 0.74, 0.75, and 0.70, respectively.
Conclusion: Together, this study established a signature based on NMGs that is a prognostic biomarker for LUAD patients and has the potential to be widely applied in future clinical settings.

背景：肺癌是全球范围内癌症相关死亡的首要病因，其中肺腺癌（lung adenocarcinoma, LUAD）是主要的组织学亚型之一。线粒体对于维持机体生理功能至关重要，其功能失调已被证实与肿瘤发生及疾病进展密切相关。尽管已有研究发现部分线粒体相关基因仅与多种肿瘤的临床结局相关，但核编码线粒体基因（nuclear mitochondrial genes, NMGs）与肺腺癌预后之间的整体关联仍未明确。
方法：从公共数据库中下载肺腺癌的核编码线粒体基因列表、基因表达数据及相关临床信息。采用生物信息学方法筛选出18个与预后相关的核编码线粒体基因，并以此构建风险特征模型（risk signature）。
结果：通过LASSO回归分析筛选出18个核编码线粒体基因，分别为NDUFS2、ATP8A2、SCO1、COX14、COA6、RRM2B、TFAM、DARS2、GARS、YARS2、EFG1、GFM1、MRPL3、MRPL44、ISCU、CABC1、HSPD1及ETHE1。这18个基因的mRNA表达水平与其相对线性拷贝数变异（copy number alteration, CNA）呈正相关。与此同时，所构建的风险特征模型可有效区分高风险与低风险患者，其预测效能在3个独立的基因表达综合数据库（Gene Expression Omnibus, GEO）队列中得到了验证。值得注意的是，携带高风险核编码线粒体基因特征的患者，其可操作EGFR变异的检出率显著更低，同时其肿瘤免疫微环境的炎症浸润程度更高。此外，多因素Cox回归分析显示，在纳入风险评分、肿瘤分期及淋巴结分期后，该模型仍保持稳定；其1年、3年及5年的受试者工作特征曲线下面积（area under the curve, AUC）分别为0.74、0.75及0.70。
结论：本研究构建了基于核编码线粒体基因的特征模型，可作为肺腺癌患者的预后生物标志物，有望在未来的临床场景中得到广泛应用。