Esophageal cancer and lung cancer are among the most common cancers worldwide with millions of new cases annually. Esophageal cancer patients at an advanced stage suffer from a poor five-year survival rate. However, only fewer than 30% of esophageal cancer cases were diagnosed at an early stage. For lung cancer, malignant pleural effusion (MPE) is an important hallmark for late-stage patients with metastasis. However, other causes of pleural effusions including tuberculosis bring difficulties in the diagnosis of MPE. It is necessary to develop novel diagnostic biomarkers and elucidate the pathological mechanism of esophageal cancer and lung cancer. Metabolic reprogramming is an emerging hallmark of cancer. It has been clear that metabolites play a critical role in cancer development and impose vulnerabilities that could be targeted for cancer therapy. The overall objective of this study is to comprehensively characterize the metabolic dysregulation in esophageal cancer and lung cancer for biomarker discovery and pathological elucidation, by using liquid chromatography--mass spectrometry (LC-MS)-based metabolomics and lipidomics. Paired tumors and normal adjacent tissues from esophageal squamous-cell carcinoma (ESCC) patients were first analyzed through global metabolomic and lipidomic profiling. Tumors were clearly separated from the normal tissues based on the partial least-square discriminant analysis (PLS-DA) model (R2Y >0.85 and Q2Y >0.79 in metabolomic profiling and R2Y >0.70 and Q2Y >0.67 in lipidomic profiling). A preliminary list of 41 polar metabolites and 65 lipids were identified to be significantly perturbed in tumor tissues. Kynurenine, spermidine, citicoline, as well as several glucosylceramides and phosphatidylcholines (PC) showed excellent predictive potential with area under curve (AUC) values better than 0.95 in receiver operating characteristic (ROC) models. Major elevated metabolic pathways were polyamine biosynthesis, glycerophospholipid metabolism, methionine mechanism, arginine and proline mechanism, and kynurenine metabolism, suggesting active amino acid biosynthesis and lipid biosynthesis in ESCC. The potential biomarkers and dysregulated pathways discovered above in ESCC tissue was further validated using targeted metabolomic, lipidomic and proteomic profiling. Polyamine biosynthesis was found to be activated in ESCC through the overexpression of tumor promoting ornithine decarboxylase and spermidine/spermine synthases. Upregulated levels of S-adenosylmethionine and DNA (cytosine-5)-methyltransferase 1 implied DNA hypermethylation in ESCC. Elevated purines in tumors were generated through the overexpression of methylenetetrahydrofolate dehydrogenases. Active phospholipid biosynthesis in tumors was promoted by overexpression of choline transporters and synthase of citicoline, which may accelerate the tumor growth. Dysregulation of coenzyme A species with different fatty acyl chains showed the same trend as of phospholipids, implying the specific activation of relevant acyltransferases in the phospholipid remodeling pathway. Moreover, essential amino acids exhibited a higher upregulation trends in patients with high-grade tumor or with cancer recurrence. Collectively, this study revealed the detailed metabolic dysregulations in ESCC tumor tissues, discovered potential metabolite biomarkers and identified therapeutic targets of ESCC. In order to explore the clinical application of the discovered biomarkers, metabolomic and lipidomic profiling was further performed on ESCC plasma samples. Eight metabolites were found to be simultaneously upregulated in ESCC tumors and plasma samples, indicating their potential as tumor-derived plasma biomarkers. Among them, a panel of five tumor-derived plasma biomarkers consisting of arginine, acetylspermidine, methylguanosine, dimethylguanosine and cystine showed good diagnostic potential in the cross validation. These biomarkers are related with polyamine biosynthesis and purine metabolism, which are critical to support tumor growth. For lung cancer, MPE from lung adenocarcinoma patients were investigated by LC-MS/MS-based metabolomic and lipidomic profiling. In PLS-DA models, the MPE samples were clearly separated from benign pleural effusion samples from pulmonary tuberculosis patients. A group of 17 polar metabolites and 45 lipids were identified to be significantly perturbed in MPE. For diagnostic purposes, ether lipid biomarkers, including PCs, lyso-PCs and phosphatidylethanolamines, showed an excellent predictive ability with the highest AUC value of 0.953 in ROC models. Furthermore, downregulated ether lipids and upregulated oxidized polyunsaturated fatty acids in MPE reflected the elevated oxidative stress and peroxisome disorder in lung cancer patients, which offers deeper understanding in lung cancer pathology.
|Date of Award||11 Aug 2020|
|Supervisor||Zongwei CAI (Supervisor)|
- Mass spectrometry
- Liquid chromatography