Main content area

Proposing a validation scheme for 13C metabolite tracer studies in high-resolution mass spectrometry

Schwaiger-Haber, Michaela, Hermann, Gerrit, El Abiead, Yasin, Rampler, Evelyn, Wernisch, Stefanie, Sas, Kelli, Pennathur, Subramaniam, Koellensperger, Gunda
Analytical and bioanalytical chemistry 2019 v.411 no.14 pp. 3103-3113
Komagataella pastoris, anion exchange chromatography, carbon, experimental design, hydrophilic interaction chromatography, mass spectrometry, metabolic flux analysis, metabolites, quality control, reversed-phase liquid chromatography, selenium, stable isotopes
¹³C metabolite tracer and metabolic flux analyses require upfront experimental planning and validation tools. Here, we present a validation scheme including a comparison of different LC methods that allow for customization of analytical strategies for tracer studies with regard to the targeted metabolites. As the measurement of significant changes in labeling patterns depends on the spectral accuracy, we investigate this aspect comprehensively for high-resolution orbitrap mass spectrometry combined with reversed-phase chromatography, hydrophilic interaction liquid chromatography, or anion-exchange chromatography. Moreover, we propose a quality control protocol based on (1) a metabolite containing selenium to assess the instrument performance and on (2) in vivo synthesized isotopically enriched Pichia pastoris to validate the accuracy of carbon isotopologue distributions (CIDs), in this case considering each isotopologue of a targeted metabolite panel. Finally, validation involved a thorough assessment of procedural blanks and matrix interferences. We compared the analytical figures of merit regarding CID determination for over 40 metabolites between the three methods. Excellent precisions of less than 1% and trueness bias as small as 0.01–1% were found for the majority of compounds, whereas the CID determination of a small fraction was affected by contaminants. For most compounds, changes of labeling pattern as low as 1% could be measured. Graphical abstract