Development of a widely-targeted metabolomics method based on gas chromatography-mass spectrometry

Abstract

Gas chromatography-mass spectrometry (GC-MS) detectors are widely used detection instruments owing to their distinct advantages over other analytical techniques, including lower sample consumption, higher sensitivity, faster analysis speed, and simultaneous separation and analysis. Metabolomics is an important component of system physiology that concerns systematic studies of the metabolite spectrum in one or more biological systems, such as cells, tissues, organs, body fluids, and organisms. Unfortunately, conventional GC-MS detectors also feature low scan rates, high ion loss rates, and a narrow concentration detection range, which limit their applications in the field of metabolomics. Therefore, establishing a GC-MS-based metabolomic analysis method with wide coverage is of great importance. In this research, a widely-targeted metabolomics method based on GC-MS is proposed. This method combines the universality of untargeted metabolomics with the accuracy of targeted metabolomics to realize the qualitative and semi-quantitative detection of numerous metabolites. It does not require a self-built database and exhibits high sensitivity, good repeatability, and strong support for a wide range of metabolic substances. The proposed method was used to establish the relationship between the retention time of straight-chain fatty acid methyl esters (FAMEs) and their retention index (RI) in the FiehnLib database based on the metabolite information stored in this database. We obtained a linear relationship that could be described by the equation y=40878x-47530, r2=0.9999. We then calculated the retention times of metabolites in the FiehnLib database under the experimental conditions based on their RI. In this way, the effects of significant variations in peak retention times owing to differences in the chromatographic column, temperature, carrier gas flow rate, and so on can be avoided. The retention time of a substance fluctuates within a certain threshold because of variations in instrument performance, matrix interference, and other factors. As such, the retention time threshold of the substance must be determined. In this paper, the retention time threshold was set to 0.15 min to avoid instrument fluctuations. The optimal scan interval was optimized to 0.20 s (possible values=0.10, 0.15, 0.20, 0.25, and 0.30 s) because longer sampling periods can lead to spectral data loss and reductions in the resolution of adjacent chromatographic peaks, whereas shorter sampling periods can result in deterioration of the signal-to-noise ratio of the collected signals. The metabolite quantification ions were optimized to avoid the interference of quantification ion peak accumulation in the case of similar peak times, and a selected ion monitoring (SIM) method table was constructed for 611 metabolites, covering 65% of the metabolic pathways in the KEGG (Kyoto Encyclopedia of Genes and Genomes). The developed method covered 39 pathways, including glycolysis, the tricarboxylic acid cycle, purine metabolism, pyrimidine metabolism, amino acid metabolism, and biosynthesis. Compared with the full-scan untargeted GC-MS method, the widely-targeted GC-MS method demonstrated a 20%-30% increase in the number of metabolites detected, as well as a 15%-20% increase in signal-to-noise ratio. The results of stability tests showed that 84% of the intraday relative standard deviations (RSDs) of metabolite retention times were less than 2% and 91% of that were less than 3%; moreover, 54% of the interday RSDs of metabolite retention times were less than 2% and 76% of that were less than 3%. The detection and analysis results of common biological samples confirmed that the proposed method greatly improved the quantity and signal-to-noise ratio of the detected metabolites and is applicable to substances that are thermally stable, volatile, or volatile after derivation and have relative molecular masses lower than 600. Thus, the widely-targeted GC-MS method can expand the application scope of GC-MS in metabolomics.