Isomer selectivity of one- and two-dimensional approaches of mixed-mode and hydrophilic interaction liquid chromatography coupled to tandem mass spectrometry for sugar phosphates of glycolysis and pentose phosphate pathways

In this study, the chromatographic behavior of mixed-mode and hydrophilic interaction liquid chromatography (HILIC) with the mixed-mode HILIC/strong anion-exchange (SAX) column HILICpak VT-50 2D and the two HILIC columns Atlantis Premier BEH Z-HILIC and Acquity Premier BEH Amide was assessed with regard to their separation capability of the metabolites from the glycolysis and pentose phosphate pathways. Chromatographic conditions were evaluated with the aim of achieving separation of the isomeric glycolytic phosphorylated carbohydrate metabolites free from isomeric interferences and thus allowing for selective targeted analysis by liquid chromatography with tandem mass spectrometry (MS/MS) using multiple reaction monitoring acquisition. The effects of pH values (8.0/9.0/10.0) of the ammonium bicarbonate buffer and gradient time were investigated during HILIC-MS/MS analysis, with the optimal conditions found at pH = 10.0. Separation of the pentose phosphate isomers (ribose 5- and 1-phosphate, xylulose 5-phosphate and ribulose 5-phosphate) was achieved on the mixed-mode HILIC/SAX (HILICpak VT-50 2D) column and HILIC BEH Amide column. Column performance was evaluated based on the direct comparison of chromatographic parameters, i.e. peak width at 50% and peak tailing factors of the individual metabolites. Parity plots were generated allowing a direct comparison between the normalized retention times and assessment of orthogonality of all 3 stationary phases evaluated. Separation of 7 biologically relevant hexose monophosphates metabolites turned out to be challenging by HILIC-MS/MS, with the BEH Amide providing the best individual results for such a separation. However, fructose 6-phosphate and glucose 1-phosphate co-eluted. Therefore, an on-line heart-cutting HILIC-Mixed Mode 2D-LC-QToF experiment was conducted, allowing the separation of this critical isomer pair. In this setup, the BEH Amide column in the 1D separated the majority of target metabolites, while a heart-cut of the peak from totally coeluted fructose 6-phosphate and glucose 1-phosphate was separated in the 2D with HILICpak VT50-2D column, thus allowing undisturbed determination of the glycolytic phosphorylated carbohydrate metabolites due to their chromatographic separation from hexose monophosphate metabolites. The assay specificity towards 7 common hexose monophosphates was characterized (glucose 1- and 6-phosphate, galactose 1- and 6-phosphate, fructose 6-phosphate, mannose 1- and 6-phosphate). The selectivity of some rare hexose monophosphates (allose 6-phosphate, tagatose 6-phosphate, sorbose 1-phosphate) was also tested.