Amino Acid Metabolomics Using LC-MS/MS: Assessment of Cancer-Cell Resistance in a Simulated Tumor Microenvironment.

We performed a comprehensive quantification of 20 amino acids in RPMI 1640 medium-cultured human colorectal adenocarcinoma cells to evaluate the efficacy of 5-fluorouracil treatment under hypoxic and hypoglycemic conditions, which mimic the tumor microenvironment. In this study, we developed a simple and comprehensive analytical method by using LC-MS/MS connected to the Intrada amino acid column, which eluted amino acids within 9 min. The present method covered a linearity range of 3.6 - 1818 µM, except for Gly (227 - 1818 µM), Ala, Asp, His (7.1 - 1818 µM each), and Trp (3.6 - 909 µM). The limits of detection were in the range of 0.02 - 38.0 pmol per injection in a standard solution. Amino acid concentration data were analyzed using principal-component analysis to represent samples on two-dimensional graphs. Linear discriminant analysis was used to classify samples on the score plots. Using this approach, the effect of 5-fluorouracil treatment could be successfully discriminated at high discrimination rates. Moreover, several amino acids were extracted from corresponding loading plots as candidate markers for distinguishing the effects of the 5-fluorouracil treatment or tumor microenvironmental conditions. These results suggest that our proposed method might be a useful tool for evaluating the efficacy of anticancer drugs in the tumor microenvironment.

Assessment of the efficacy of anticancer drugs by amino acid metabolomics using fluorescence derivatization-HPLC.

Metabolomic studies conducted for evaluating cancer pathogenesis and progression by monitoring the amino acids metabolic balance hold great promise for assessing current and future anticancer treatments. We performed a comprehensive quantification of 21 amino acids concentrations in cultured human colorectal adenocarcinoma cells treated with the anticancer drugs 5-fluorouracil, irinotecan, and cisplatin. A precolumn fluorescence derivatization-HPLC method involving 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate was used. Amino acid concentration data were analyzed by principal-component analysis and partial least-squares multivariate statistical methods to represent samples on two-dimensional graphs. The hierarchical cluster analysis and linear discriminant analysis were used to classify the samples on the score plots. Unlike the cluster analysis approach, the linear discrimination analysis classification successfully distinguished anticancer drug-treated samples from the untreated controls. Moreover, three candidate amino acids (serine, aspartic acid, and methionine) were identified from the loading plots as potential biomarkers. Our proposed method might be able to evaluate the effectiveness of anticancer therapy even in small laboratories or medical institutions.