Stable Isotope Assisted Assignment of Elemental Compositions for Metabolomics

Abstract

Assignment of individual compound identities within mixtures of thousands of metabolites in biological extracts is a major challenge for metabolomic technology. Mass spectrometry offers high sensitivity over a large dynamic range of abundances and molecular weights but is limited in its capacity to discriminate isobaric compounds. In this article, we have extended earlier studies using isotopic labeling for elemental composition elucidation (Rodgers, R. P.; Blumer, E. N.; Hendrickson, C. L.; Marshall, A. G. J. Am. Soc. Mass Spectrom. 2000, 11, 835-40) to limit the formulas consistent with any exact mass measurement by comparing observations of metabolites extracted from Arabidopsis thaliana plants grown with (I) (12)C and (14)N (natural abundance), (II) (12)C and (15)N, (III) (13)C and (14)N, or (IV) (13)C and (15)N. Unique elemental compositions were determined over a dramatically enhanced mass range by analyzing exact mass measurement data from the four extracts using two methods. In the first, metabolite masses were matched with a library of 11,000 compounds known to be present in living cells by using values calculated for each of the four isotopic conditions. In the second method, metabolite masses were searched against masses calculated for a constrained subset of possible atomic combinations in all four isotopic regimes. In both methods, the lists of elemental compositions from each labeling regime were compared to find common formulas with similar retention properties by HPLC in at least three of the four regimes. These results demonstrate that metabolic labeling can be used to provide additional constraints for higher confidence formula assignments over an extended mass range.

Mass spectrometry for the identification of the discriminating signals from metabolomics: Current status and future trends

“…The comparison of the monoisotopic masses from unlabeled and labeled biological extracts gives access to the number of both C and N atoms, decreasing the number of possible chemical formulas. A proof of concept has been performed on extracts from the model plant Arabidopsis thaliana: double isotopic labeling ( 13 C and 15 N) resulted in unique assignment of 87% of 5000 formulas versus 20% with a typical approach [71].…”

Metabolomics of Arabidopsis Thaliana

“…Shifts in m/z values in 13 C or 15 N enriched samples allows determination of numbers of C and N atoms in each ion species, thus narrowing down the possibilities for empirical formula. Applications of this have been described for sample introduction by direct infusion MS (Giavalisco et al, 2008) and also by LC-MS (Hegeman et al, 2007). The isotope enrichment method may not be amenable for larger seeded plants, even other models such as Medicago truncatula, whose seed is only 50 times bigger than Arabidopsis, because the contribution of 14 N and 12 C from the metabolites stored in the seed is too large to be overcome completely by replacement with isotope enriched medium after germination.…”

“…Of course, to solve this problem one could grow such plants through one complete generation to ensure that the seeds are heavily labelled, but the costs of performing such experiments are too high for most laboratories. Arabidopsis is easy to label since one can place about 100-300 dry seeds in 50 mL of media containing complete replacement of 14 N with 15 N nitrate and 12 C with 13 C 6 glucose (see Hegeman et al, 2007). It is possible to make Arabidopsis plants nearly completely replaced at all N and C atoms simultaneously in this procedure.…”

“…To add confidence to empirical formulae determined by mass spectroscopy, another method, which is only possible so far in plants, is to grow the plant to near-complete isotope replacement, with 99% 15 N or 13 C, and then determining whether the feature increases its mass exactly according to that predicted from the suspected formula (Hegeman et al, 2007). If the mass is known with sufficient precision for that feature, the list of candidates can be small, but additional experiments are clearly needed to be sure.…”

Nanoscale Ion-Pair Reversed-Phase HPLC−MS for Sensitive Metabolome Analysis

“…Potential metabolites were identified by stable 13 C isotope assignment. 27,28 For attribution of molecular formulas to detected mass peaks, a list of 14'943 molecular formulas was extracted from PubChem compound database (http://pubchem.ncbi.nlm.nih.gov/) from all compounds of category metabolic pathways composed of C and H, N, O, P, or S. The list was used to generate mass lists for assignment of molecular formulas. A molecular formula of the database was assigned to a mass peak when i. the differences between theoretical and measured m/z values of the monoisotopic peak M 0 , of the corresponding U- 13 C labeled peak M UL as well as the delta value M UL -M 0 were below 1 mmu, and ii.…”