Metabolite profiling of human plasma by different extraction methods through gas chromatography-mass spectrometry-An objective comparison

For the comprehensive metabolite profiling of human plasma, sample preparation is a crucial step. In this investigation, we have compared 10 different extraction techniques for metabolite profiling by GC-MS. Six one-dimensional (1D) and four two-dimensional (2D) extraction techniques involving solvent precipitation, molecular weight cut off tube (MWCOT) and solid phase extraction (SPE) by using silica, RP C18, cation and anion were investigated. Pooled samples of 50 Healthy Male Plasma (HMP), 50 Healthy Female Plasma (HFP) and 100 Healthy Pakistani Plasma (HPP) were subjected to these extraction methods for comparison purposes. Metabolites obtained were identified through NIST mass spectral (Wiley registry), METLIN and Fiehn RTL libraries. XCMS Software was used for the detection of metabolic features, retention time correction, alignment, annotation and statistical analysis in each method. 116-34 peaks were detected by various methods and approx 33% of the peaks were characterized in each method. Hierarchical clustering of the 10 extraction methods showed a low similarity index (50.1%) which indicated different chemical nature of metabolites, resulting from different methods. Venn diagram highlights the GC-MS peaks (33-77%) common in various methods. Metabolites which were different in male and female groups were detected using a threshold value of p <= 0.0001, q <= 0.001 and fold change >= 3 by employing Welch's t-test and identified through METLIN. Results indicated that 2D-C18 and 2D-silica offers a comprehensive metabolite profile in term of reproducibility, number of peaks and difference in metabolite pattern of male and female. (C) 2013 Elsevier B.V. All rights reserved.