Potato metabolomics by GC–MS: what are the limiting factors?

Metabolic profiling methods are not ideally suited to the simultaneous analysis of all metabolite classes within a biological sample and must be optimized for maximum applicability. Several factors related to the optimization, validation and limitations of a GC–MS-based metabolic profiling method for potato were examined. A key step is conversion of reducing sugars to methyloximes, and optimum reaction conditions were 50 °C for 4 h. Shorter times or lower temperatures resulted in incomplete oximation whereas longer times and higher temperatures caused hydrolysis of sucrose, the major tuber dissacharide. Metabolite concentration gradients were observed in tuber sections. Glucose, fructose, alanine, methionine, threonine and tyrosine were more concentrated in the interior, whereas asparagine, putrescine, and caffeic and chlorogenic acids were higher in the skin and citrate was concentrated at the tuber’s bud end. These results impact upon choice of sampling strategy, consequently the use of freeze-dried (FD) material from a sampling protocol developed to avoid gradient-induced bias was examined. Using FD material, the method was highly linear and there was little qualitative or quantitative difference in the metabolite composition between fresh and FD material. The short- and long-term repeatability of the method was studied, and the use of reference materials to monitor and to improve data quality is discussed. Ascorbate is an important tuber metabolite that is readily measured by targeted approaches, but can be a problem in metabolic profiling. It was shown for standards and FD potato that ascorbate was largely degraded during oximation, although some survived in FD material.