Tissue Metabolomic Fingerprinting Reveals Metabolic Disorders Associated with Human Gastric Cancer Morbidity.

Hu Song,Lei Wang,Huan-Liang Liu,Xiao-Bin Wu,Hua-She Wang,Zhong-Hui Liu,Yun Li,De-Chang Diao,Hong-Lei Chen,Jun-Sheng Peng
DOI: https://doi.org/10.3892/or.2011.1302
2011-01-01
Oncology Reports
Abstract:The principal way to improve the outcome of gastric cancer (GC) is to predict carcinogenesis and metastasis at an early stage. The aims of the present study were to test the hypothesis that distinct metabolic profiles are reflected in GC tissues and to further explore potential biomarkers for GC diagnosis. Gas chromatography/mass spectrometry (GC/MS) was utilized to analyze tissue metabolites from 30 GC patients. A diagnostic model for GC was constructed using orthogonal partial least squares discriminant analysis (OPLS-DA), and the metabolomic data were analyzed using the non-parametric Wilcoxon rank sum test to identify the metabolic tissue biomarkers for GC. Over 100 signals were routinely detected in one single total ion current (TIC) chromatogram, and the OPLS-DA model generated from the metabolic profile of the tissues adequately discriminated the GC tissues from the normal mucosae. Among the low-molecular-weight endogenous metabolites, a total of 41 compounds, such as amino acids, organic acids, carbohydrates, fatty acids and steroids, were detected, and 15 differential metabolites were identified with significant difference (p<0.05). A total of 20 variables were noted which contributed to a great extent in the discriminating OPLS-DA model (VIP value >1.0), among which 12 metabolites were identified using both VIP values (VIP >1) and the Wilcoxon test (p<0.05). In conclusion, the identification of the metabolites associated with GC morbidity potentially revealed perturbations of glycolysis, fatty acid β-oxidation, cholesterol and amino acid metabolism. These results suggest that tissue metabolic profiles have great potential in detecting GC and may aid in understanding its underlying mechanisms.
What problem does this paper attempt to address?