背景:葡萄膜黑色素瘤(UVM)是成人的恶性眼内肿瘤。靶向氧化磷酸化相关基因(OXPHOS)可能在抗肿瘤治疗中发挥作用。然而,UVM中氧化磷酸化的临床意义尚不清楚.
方法:从KEGG途径获得134个OXPHOS相关基因,TCGAUVM数据集包含80个样本,作为训练组,而GSE22138和GSE39717用作验证集。进行LASSO回归以鉴定OXPHOS相关的预后基因。从Cox多元回归分析获得的系数用于计算风险评分,这促进了预后模型的构建。Kaplan-Meier生存分析,使用时间“timeROC”包装进行logrank测试和ROC曲线。通过Cibersort工具分析低、高危人群的免疫细胞频率。通过“maftools”R软件包分析了特定的基因组改变。分析低危组和高危组之间的差异表达基因,并进行基因本体论(GO)和GSEA。最后,我们在体外通过基因沉默验证了CYC1在UVM中的功能。
结果:共鉴定出9个OXPHOS相关预后基因,包括NDUFB1、NDUFB8、ATP12A、NDUFA3、CYC1、COX6B1、ATP6V1G2、ATP4B和NDUFB4。基于9个OXPHOS相关预后基因构建UVM预后风险模型。高危组患者的预后较低危组差。此外,ROC曲线显示,预测UVM患者1~5年生存率的模型曲线下面积均大于0.88。在GSE22138和GSE39717数据集中的外部验证显示,这9个基因也可用于评估和预测患有UVM的患者的总体存活。风险评分水平与免疫细胞频率和特定基因组改变有关。低危组和高危组之间的DEGs富含肿瘤OXPHOS和免疫相关通路。体外实验,CYC1沉默显著抑制UVM细胞增殖和侵袭,诱导细胞凋亡。
结论:总而言之,基于UVM中氧化磷酸化相关基因的预后风险评分模型旨在增强对该疾病的认识.这种预后风险评分模型可能有助于寻找UVM患者的潜在治疗靶标。CYC1在UVM中起癌基因作用。
BACKGROUND: Uveal melanoma (UVM) is a malignant intraocular tumor in adults. Targeting genes related to oxidative phosphorylation (OXPHOS) may play a role in anti-tumor therapy. However, the clinical significance of oxidative phosphorylation in UVM is unclear.
METHODS: The 134 OXPHOS-related genes were obtained from the KEGG pathway, the TCGA UVM dataset contained 80 samples, served as the training set, while GSE22138 and GSE39717 was used as the validation set. LASSO regression was carried out to identify OXPHOS-related prognostic genes. The coefficients obtained from Cox multivariate regression analysis were used to calculate a risk score, which facilitated the construction of a prognostic model. Kaplan-Meier survival analysis, logrank test and ROC curve using the time \"timeROC\" package were conducted. The immune cell frequency in low- and high-risk group was analyzed through Cibersort tool. The specific genomic alterations were analyzed by \"maftools\" R package. The differential expressed genes between low- or high-risk group were analyzed and performed Gene Ontology (GO) and GSEA. Finally, we verified the function of CYC1 in UVM by gene silencing in vitro.
RESULTS: A total of 9 OXPHOS-related prognostic genes were identified, including NDUFB1, NDUFB8, ATP12A, NDUFA3, CYC1, COX6B1, ATP6V1G2, ATP4B and NDUFB4. The UVM prognostic risk model was constructed based on the 9 OXPHOS-related prognostic genes. The prognosis of patients in the high-risk group was poorer than low-risk group. Besides, the ROC curve demonstrated that the area under the curve of the model for predicting the 1 to 5-year survival rate of UVM patients were all more than 0.88. External validation in GSE22138 and GSE39717 dataset revealed that these 9 genes could also be utilized to evaluate and predict the overall survival of patients with UVM. The risk score levels related to immune cell frequency and specific genomic alterations. The DEGs between the low- and high- risk group were enriched in tumor OXPHOS and immune related pathway. In vitro experiments, CYC1 silencing significantly inhibited UVM cell proliferation and invasion, induced cell apoptosis.
CONCLUSIONS: In sum, a prognostic risk score model based on oxidative phosphorylation-related genes in UVM was developed to enhance understanding of the disease. This prognostic risk score model may help to find potential therapeutic targets for UVM patients. CYC1 acts as an oncogene role in UVM.