What if it were possible to image the locations of multitudes of molecules in a tissue or even a single cell and to determine their chemical identities without using any fluorescent labels or antibodies? It's not a fanciful idea—a technique called mass spectrometry imaging (MSI) has the potential to do just this.

Mass spectrometry imaging maps the molecular composition of tissues and cells. Reproduced in part from Kompauer, M. et al. Nat. Methods 14, 1156–1158, 2017. Credit: Kim Caesar/Springer Nature

Mass spectrometry was first used decades ago to image elemental distribution in tissues and cells. In 1997, the biomolecule-friendly technique of matrix-assisted laser desorption–ionization (MALDI) was applied for the first time to localize proteins and peptides in tissue samples. This breakthrough work by Richard Caprioli and coworkers launched a new field of label-free molecular imaging using mass spectrometry.

In a typical implementation of MSI, a laser scans a tissue surface to ionize molecules from defined spots; a mass spectrometer records a full mass spectrum for each spot. Images representing the distributions of mass signals of interest over the tissue section are then reconstructed using computer programs. Molecules can be identified by interpreting their mass spectra. As long as the signal is sufficient, the distribution of just about any molecule—such as peptides, metabolites, and lipids—can be mapped using MSI. Besides numerous applications in basic research, MSI also has great potential for following how a drug distributes in tissue or for disease diagnosis.

Recent advances in proteomics have benefitted other applications that depend on mass spectrometry, including MSI. New ion sources and high-resolution, high-accuracy mass analyzers have led to substantial improvements in mass spectrometry sensitivity. Such developments have also spurred numerous new approaches for data analysis.

In 2017, advances in MSI made it possible to resolve subcellular molecular distributions (e.g., Nat. Methods 14, 90–69, 2017; Nat. Methods 14, 1175–1183, 2017). Methods were also developed to improve the imaging of molecules in 3D samples (Nat. Methods 14, 1175–1183, 2017) and to capture corresponding information about sample topography (Nat. Methods 14, 1156–1158, 2017). Statistical tools for assessing molecular identification by MSI also help advance research (e.g., Nat. Methods 14, 57–60, 2017).

There is still much room for improving sensitivity to detect low-abundance molecules, quantification, and molecular identification—particularly of metabolites—from mass spectra. Will the developments of the last year and those yet to come open up MSI technology for broader application? It is an area worth watching.