<p>Annotation of untargeted
high-resolution full-scan LC-MS metabolomics data remains a difficult task. Existing
literature suggests that LC-MS peaks can be divided into multiple major
categories including “Background”, “Isotope”, “Adduct”, “Fragment” and
“Candidate metabolite”. Among these, adduct annotation is a particular
challenge, as the same mass difference between peaks can arise from adduct
formation, fragmentation, or different biological species. To address this,
here we describe a Buffer Modification Workflow (BMW), in which the same sample
is run by LC-MS in both liquid chromatography solvent with <sup>14</sup>NH<sub>3</sub>-acetate
buffer, and in solvent with the buffer modified with <sup>15</sup>NH<sub>3</sub>-formate.
Buffer switching results in characteristic mass and signal intensity changes
for adduct peaks, facilitating their annotation. In analyzing the candidate
metabolite peaks, we recognized that some paradoxically increased in intensity
over time between sample preparation and analysis. We show that such peaks are
formed by chemical reactions between known metabolites and the extraction
buffer and accordingly categorize these peaks as “Reaction Product”. Comparison using yeast extracts of BMW with a
stable isotope labeling-based workflow suggests that BMW captures > 90% of
candidate metabolites. This new workflow is well-suited to biological samples
that cannot be readily isotope labeled, such as mammalian tissues and tumors.
Application to mouse liver identified 26% of ~ 27,000 total peaks across
positive and negative mode as candidate metabolites, of which ~ 2600 showed HMDB
or KEGG database formula match.</p>