Systematic Evaluation of Liquid Chromatography (LC) Column Combinations for Application in Two-Dimensional LC Metabolomic Studies

In comparison to proteomics, the application of two-dimensional liquid chromatography (2D LC) in the field of metabolomics is still premature. One reason might be the elevated chemical complexity and the associated challenge of selecting proper separation conditions in each dimension. As orthogonality of dimensions is a major issue, the present study aimed for the identification of successful stationary phase combinations. To determine the degree of orthogonality, first, six different metrics, namely, Pearson's correlation coefficient (1 - vertical bar R vertical bar), the nearest-neighbor distances (H-NND), the asterisk equations (A(O)), and surface coverage by bins (SCG), convex hulls (SCCH), and alpha-convex hulls (SC alpha H), were critically assessed by 15 artificial 2D data sets, and a systematic parameter optimization of a-convex hulls was conducted. SG(G), SC alpha H with alpha = 0.1, and H-NND generated valid results with sensitivity toward space utilization and data distribution and, therefore, were applied to pairs of experimental retention time sets obtained for >350 metabolites, selected to represent the chemical space of human urine. Normalized retention data were obtained for 23 chromatographic setups, comprising reversed-phase (RP), hydrophilic interaction liquid chromatography (HILIC), and mixed-mode separation systems with an ion exchange (IEX) contribution. As expected, no single LC setting provided separation of all considered analytes, but while conventional RPxHILIC combinations appeared rather complementary than orthogonal, the incorporation of IEX properties into the RP dimension substantially increased the 2D potential. Eventually, one of the most promising column combinations was implemented for an offline 2D LC time-of-flight mass spectrometry analysis of a lyophilized urine sample. Targeted screening resulted in a total of 164 detected metabolites and confirmed the outstanding coverage of the 2D retention space.

Automated summary

Compared to proteomics, the use of two-dimensional liquid chromatography (2D LC) in metabolomics is still in its early stages. The study aimed to identify successful stationary phase combinations and optimized parameters to determine orthogonality. Through the analysis of 15 artificial 2D data sets, the study identified the most effective column combinations for metabolomics. The implementation of these column combinations in an offline 2D LC time-of-flight mass spectrometry analysis of a lyophilized urine sample resulted in the detection of 164 metabolites, showcasing the comprehensive coverage of the 2D retention space.