Improving Metabolite Annotations in On-Tissue Chemical Derivatization Mass Spectrometry Imaging by Functional Group...

Matrix-assisted laser desorption ionization mass spectrometry imaging (MALDI-MSI) enables the direct visualization of metabolites from tissue sections with high spatial resolution. However, its application to untargeted spatial metabolomics is hindered by poor ionizing compounds and challenges in accurate metabolite annotation. On-tissue chemical derivatization (OTCD) is commonly employed to enhance the ionization of metabolites bearing specific functional groups, and platforms such as METASPACE facilitate high-throughput annotation of derivatized features. Nevertheless, distinguishing structural isomers for a large number of metabolites remains a major challenge, often resulting in incorrect annotations. To address this limitation, we developed an improved annotation workflow for OTCD-MALDI-MSI by integrating two filtering strategies. Functional group filtering leverages SMARTS-based substructure matching to retain only those metabolites that react with the applied OTCD reagent. In parallel, gas-phase hydrogen-deuterium exchange (HDX) in the MALDI source is used to determine the number of labile hydrogens for each feature, enabling the exclusion of annotations that are inconsistent with HDX behavior. We applied this workflow to MALDI-MSI of maize root sections using Girard's reagents T and P, along with the plant-specific COCONUT metabolite database. The combined filtering strategy reduced incorrect annotations by ∼67%, from ∼7.3 annotations per unique feature without filtering to ∼2.4 with filtering, substantially improving annotation accuracy and confidence. By coupling OTCD signal enhancement with structurally informed filtering, this workflow advances the utility of MALDI-MSI for untargeted spatial metabolomics, enabling more reliable and scalable metabolite profiling in complex biological tissues.