Item

Abstract

: Spatial metabolomics using mass spectrometry
imaging (MSI) is a powerful tool to map hundreds to thousands
of metabolites in biological systems. One major challenge in MSI is
the annotation of m/z values, which is substantially complicated by
background ions introduced throughout the chemicals and
equipment used during experimental procedures. Among many
factors, the formation of adducts with sodium or potassium ions, or
in case of matrix-assisted laser desorption ionization (MALDI)-
MSI, the presence of abundant matrix clusters strongly increases
total m/z peak counts. Currently, there is a limitation to identify
the chemistry of the many unknown peaks to interpret their
biological function. We took advantage of the co-localization of
adducts with their parent ions and the accuracy of high mass resolution to estimate adduct abundance in 20 datasets from different
vendors of mass spectrometers. Metabolites ranging from lipids to amines and amino acids form matrix adducts with the commonly
used 2,5-dihydroxybenzoic acid (DHB) matrix like [M + (DHB-H2O) + H]+ and [M + DHB + Na]+
. Current data analyses neglect
those matrix adducts and overestimate total metabolite numbers, thereby expanding the number of unidentified peaks. Our study
demonstrates that MALDI-MSI data are strongly influenced by adduct formation across different sample types and vendor platforms
and reveals a major influence of so far unrecognized metabolite−matrix adducts on total peak counts (up to one third). We
developed a software package, mass2adduct, for the community for an automated putative assignment and quantification of
metabolite−matrix adducts enabling users to ultimately focus on the biologically relevant portion of the MSI data.