Untargeted Spatial Metabolomics and Spatial Proteomics on the Same Tissue Section

Abstract

An increasing number of spatial multiomic workflows have recently been developed. Some of these approaches have leveraged initial mass spectrometry imaging (MSI)-based spatial metabolomics to inform the region of interest (ROI) selection for downstream spatial proteomics. However, these workflows have been limited by varied substrate requirements between modalities or have required analyzing serial sections (i.e., one section per modality). To mitigate these issues, we present a new multiomic workflow that uses desorption electrospray ionization (DESI)-MSI to identify representative spatial metabolite patterns on-tissue prior to spatial proteomic analyses on the same tissue section. This workflow is demonstrated here with a model mammalian tissue (coronal rat brain section) mounted on a poly(ethylene naphthalate)-membrane slide. Initial DESI-MSI resulted in 160 annotations (SwissLipids) within the METASPACE platform (≤20% false discovery rate). A segmentation map from the annotated ion images informed the downstream ROI selection for spatial proteomics characterization from the same sample. The unspecific substrate requirements and minimal sample disruption inherent to DESI-MSI allowed for an optimized, downstream spatial proteomics assay, resulting in 3888 ± 240 to 4717 ± 48 proteins being confidently directed per ROI (200 μm × 200 μm). Finally, we demonstrate the integration of multiomic information, where we found ceramide localization to be correlated with SMPD3 abundance (ceramide synthesis protein), and we also utilized protein abundance to resolve metabolite isomeric ambiguity. Overall, the integration of DESI-MSI into the multiomic workflow allows for complementary spatial- and molecular-level information to be achieved from optimized implementations of each MS assay inherent to the workflow itself.