J Proteome Res. 2020 Jul 22.
Lei You,
Yingying Fan,
Xinyu Liu,
Shujuan Shao,
Lei Guo,
Hamada A A Noreldeen,
Zaifang Li,
Yang Ouyang,
Enyou Li,
Xue Pan,
Tianyang Liu,
Xin Tian,
Xiangnan Li,
Fei Ye,
Guowang Xu.
Unclarified molecular mechanism and lack of practical diagnosis biomarkers hinder the effective treatment of non-small-cell lung cancer. Herein, we performed liquid chromatography-mass spectrometry-based nontargeted metabolomics analysis in 131 patients with their lung tissue pairs to study the metabolic characteristics and disordered metabolic pathways in lung tumor. 339 metabolites were identified in metabolic profiling. And 241 differential metabolites were found between lung carcinoma tissues (LCTs) and paired distal noncancerous tissues, amino acids, purine metabolites, fatty acids, phospholipids and most of lysophospholipids significantly increased in LCTs, while 3-phosphoglyceric acid, phosphoenolpyruvate, 6-phosphogluconate and citrate decreased. Additionally, pathway enrichment analysis revealed that energy, purine, amino acid, lipid, and glutathione metabolism are markedly disturbed in LCa. Using binary logistic regression, we further defined candidate biomarkers for different subtypes of lung tumor. Xanthine and PC 35:2 were selected as combinational biomarkers for distinguishing benign from malignant lung tumors with 0.886 area under curve (AUC) value, and creatine, myoinositol and LPE 16:0 were defined as combinational biomarkers for discriminating adenocarcinoma from squamous cell lung carcinoma with 0.934 AUC value. Overall, metabolic characterization and pathway disturbance demonstrated apparent metabolic reprogramming in LCa. The defined candidate metabolite marker panels are useful for subtyping of lung tumors to assist clinical diagnosis.