Metabolite Analysis in Mammalian Cells: How to Generate Reliable Data Sets?

As a basis for the development of predictive mathematical models in systems biology and for a quantitative understanding of cellular metabolism, reliable experimental data sets of intracellular metabolites are indispensable. A prerequisite to obtain such data sets is the identification of a suitable sample preparation method. In this work, we optimized the extraction procedure for a wide range of intracellular metabolites from adherent mammalian cells in culture. In a screening of several commonly used extraction protocols using Madin-Darby canine kidney (MDCK) cells a MeOH/CHCl3 method appeared to be a promising candidate. In the following experiment, this method was optimized based on a central composite experimental design with four response variables: Nucleotide Amount, Energy Charge, Fructose-1,6-bP (F16bP) Amount, and Absorption at 280 nm. After fitting polynomial equations, each response was transformed to an individual desirability function and then combined to find an overall optimum for the extraction conditions. Using optimal settings, recoveries were determined and most metabolites were close to 100% with exception of nucleotide diphosphates with values significantly above 100% and nucleotide mono-and tri-phosphates below 100%. These conversions were suspected to be caused by a nucleotide kinase activity during extraction. Modification of the protocol with an earlier addition of CHCl3 led to recoveries for all metabolites of about 100%. The method was then applied for the analysis of intracellular metabolites during a starvation experiment and distinct changes could be seen for intracellular metabolite pools of MDCK cells.