Fundamental study of ion trapping and multiplexing using drift tube-ion mobility time-of-flight mass spectrometry for non-targeted metabolomics.

ABSTRACT

This study of ion accumulation/release behavior relevant to ion mobility-mass spectrometry (IM-MS) as employed for non-targeted metabolomics involves insight from theoretical studies, and controlled reference experiments involving measurement of low and high molecular mass metabolites in varying concentrations within a complex matrix (yeast extracts). Instrumental settings influencing ion trapping (accumulation time) and release conditions in standard and multiplexed operation have been examined, and translation of these insights to liquid chromatography (LC) in combination with drift tube IM-MS measurements has been made. The focus of the application is non-targeted metabolomics using carefully selected samples to allow quantitative interpretations to be made. Experimental investigation of the IM-MS ion utilization efficiency particularly focusing on the use of the Hadamard transform multiplexing with 4-bit pseudo-random pulsing sequence for assessment of low and high molecular mass metabolites is compared with theoretical modeling of gas-phase behavior of small and large molecules in the IM trapping funnel. Increasing the trapping time for small metabolites with standard IM-MS operation is demonstrated to have a deleterious effect on maintaining a quantitative representation of the metabolite abundance. The application of these insights to real-world non-targeted metabolomics assessment of intracellular extracts from biotechnologically relevant production processes is presented, and the results were compared to LC×IM-MS measurements of the same samples. Spiking of a uniformly 13C-labeled yeast extract (as a standard matrix) with varying amounts of natural metabolites is used to assess the linearity and sensitivity according to the instrument mode of operation (i.e., LC-MS, LC×IM-MS, and LC×[multiplexed]IM-MS). When comparing metabolite quantification using standard and multiplexed operation, sensitivity gain factors of 2-8 were obtained for metabolites with m/z below 250. Taken together, the simulation and experimental results of this study provide insight for optimizing measurement conditions for metabolomics and highlight the need for implementation of multiplexing strategies using short trapping times as relative quantification (e.g., in the context with non-targeted differential analysis) with sufficient sensitivity and working range is a requirement in this field of application.