Comparative High-Speed Profiling of Carboxylic Acid Metabolite Levels by Differential Isotope-Coded MALDI Mass Spectrometry Albert Koulman, Daniel Petras, Vinod K. Narayana, Laura Wang, and Dietrich A. Volmer* Medical Research Council, Elsie Widdowson Laboratory, Cambridge, United Kingdom This present work describes the development of a novel high throughput comparative matrix-assisted laser de- sorption ionization (MALDI) mass spectrometry profiling technique for endogenous compounds using a new isotope- coded label for relative quantitation of carboxylic acids. The key new aspect of this technique was a differential label, 3-hydroxymethyl-1-ethylpyrididinium iodide (HMEP), designed specifically for simultaneous quantitative MALDI analysis of two physiological states. The HMEP-d 0 and HMEP-d 5 variants of the label were applied to profiling endogenous fatty acid levels during a fish oil interven- tion study, using the metabolite profile of every indi- vidual person in the study as their own personal analytical reference standard. Initially, analytical fig- ures of merit such as sensitivity, linear dynamic range, limit of quantitation, and precision were determined from the comparative quantitation experiments. Im- portantly, the permanently charged HMEP mass tag not only increased the ionization efficiency of the studied fatty acids but also ensured that the solution phase properties of the analytes became more similar, allowing the use of CHCA as a single MALDI matrix compound for the entire range of analytes. The label exhibited a further very unique feature; it provided complete suppression of MALDI matrix-related ions. The MALDI assay was able to generate the data much faster than conventional gas chromatography (GC) methods for fatty acids. It is shown in this study that analyzing a single sample took less than 10 s with analytical results of comparable quality to those ob- tained by GC. Profiles and fingerprints of metabolites in biological samples are important measures for describing physiological processes, signaling and metabolic pathways, regulatory events, or for discovering and measuring biomarkers in biological samples. 1 The emerging field of metabolomics exploits these comprehensive snapshots of biological systems by considering the entire comple- ment of low-molecular weight compounds contained in a biological sample (the “metabolome”) and correlating them to specific physiological states, biological pathways, to effects of pharmaco- logical modulations or to phenotypes. 2 The measurement of metabolite profiles and fingerprints is not trivial. The metabolites in biological samples exhibit a wide range of different chemical structures and they are present over extended dynamic ranges (at least 10 orders of magnitude 3 ) in the metabolome, making the comprehensive analytical determi- nation extremely challenging. 4 One of these challenges originates from the quantitative determination of endogenous molecules, many of which will be unknown at the discovery stage of metabolomics and for most of which no analytical reference standards exist. This ambiguity not only makes quantifications very difficult but also creates uncertainty over the range of metabolites covered during the analysis. 5 The analytical technique chosen will certainly not be amenable to analysis of all metabolites contained in the sample and many compounds will be present at levels below the detection limit. In addition, overlapping signals in nuclear magnetic resonance (NMR) and mass spectrometry (MS), and ion suppression from biological cocomponents in mass spectrometry experiments introduce additional limitations. 6 As a result, many metabolomics approaches (e.g., MS-based shotgun metabolomics), will introduce unwanted and often unknown biases to analysis. As a result, they often do not reach very deep into complex metabolomes such as plasma or urine, only observing the most abundant metabolites, thus requiring extensive fraction- ation steps for simplification. 7 There is an additional difficulty arising from the significant interpersonal differences 8 of the metabolite profiles, particularly in genetically diverse humans, complicating the interpretation of changes of metabolite profiles across individuals further. These interpersonal variations make the use of generic biological reference materials for the plasma or urine metabolome impos- sible. Ideally, if changes in the metabolome of an individual person are measured for specific metabolites in a longitudinal study, they should be determined in comparison to the same metabolome of the same individual as “internal standard”, before intervention started. Simple calibration routines using pooled plasma as often * To whom correspondence should be addressed. Phone: +44 (0)1223 43 7550. Fax +44 (0)1223 43 7515. E-mail: Dietrich.Volmer@mrc-hnr.cam.ac.uk. (1) Lindon, J. C.; Holmes, E.; Nicholson, J. K. Curr. Opin. Mol. Ther. 2004, 6, 265–272. (2) Fiehn, O. Plant Mol. Biol. 2002, 48, 155–171. (3) Human Metabolome Database, http://www.hmdb.ca. (4) Want, E. J.; Nordstro ¨m, A.; Morita, H.; Siuzdak, G. J. Proteome Res. 2007, 6, 459–468. (5) Want, E. J.; Cravatt, B. F.; Siuzdak, G. Chembiochem. 2005, 6, 1941–1951. (6) Werner, E.; Heilier, J. F.; Ducruix, C.; Ezan, E.; Junot, C.; Tabet, J. C. J. Chromatogr., B 2008, 871, 143–163. (7) Annesley, T. M. Clin. Chem. 2003, 49, 1041–1044. (8) German, J. B.; Watkins, S. M.; Fay, L. B. J. Am. Diet Assoc. 2005, 105, 1425–1432. Anal. Chem. 2009, 81, 7544–7551 10.1021/ac900562j CCC: $40.75  2009 American Chemical Society 7544 Analytical Chemistry, Vol. 81, No. 18, September 15, 2009 Published on Web 08/24/2009