Correlative MS Imaging for cellular identification and analysis of in situ cryo-ET

Abstract

In situ cryo-electron tomography (cryo-ET) offers direct access to protein structures in their native state and environment. A central challenge for this technique is the identification and classification of imaged cells, as sample heterogeneity with respect to cell type and functional state can significantly impact the data. Currently, cells can be classified using subtomogram averaging or fluorescent labeling, which require sufficiently distinguishable macromolecules, genetic modifications, or the availability of suitable labels. In MALDI-MS Imaging (MSI) individual cells can be identified and classified reliably, making use of the distributions of phospholipid species, metabolites and small proteins with characteristic profiles for specific cell types. Here we present an integrated pipeline combining MSI and in situ cryo-ET for single-cell identification and classification. Our approach enables MSI data acquisition directly from EM grids and can be used on any eukaryotic single-cell cryo-ET sample without prior modification in less than 36 h. We apply this workflow to describe cell-to-cell heterogeneity in cultured HeLa cells, to distinguish between distantly related HeLa and Polytomella cells, and between primary cultured rat neurons and glial cells. Finally we present a data integration workflow combining cryo-ET, light microscopy and MSI aligned in the same coordinate system to facilitate efficient data visualization and interpretation. We anticipate that this approach will change the way that structural biologists can approach cellular heterogeneity in cryo-ET, as cells can be identified and classified reliably post acquisition.