MALDI imaging of the cochlea

The problem

The cochlea is tiny, encased in bone and built from many distinct cell types, which makes it hard to section and image without losing the molecular information you came for. Bulk methods average that heterogeneity away.

What we built

A streamlined sample-preparation and processing workflow for MALDI mass spectrometry imaging (MALDI-MSI) of the mouse cochlea. Sample collection was cut to as few steps as possible so that flash-frozen neonatal mouse heads could be used directly, then sectioned sagittally on a cryostat.

Matrix (NEDC, N-(1-naphthyl) ethylenediamine dihydrochloride) was applied by sublimation with controlled recrystallization and compared head to head with automated spraying. Sublimation gave higher signal-to-noise for about two-thirds of assigned ions, especially lipids and fatty acids, and an effective resolution near 5 µm versus 10–15 µm for spraying; spraying still did better for some polar metabolites such as aspartate. Imaging ran on a timsTOF fleX, with features annotated by spectral libraries, collision cross section and MS/MS.

What it showed

Diverse metabolites and lipids mapped across the cochlear region with distinct spatial distributions. Unsupervised clustering recovered reproducible physiological regions, including the otic capsule and the spiral ducts. Some molecules sat in very specific places: a sulfatide concentrated in the organ of Corti, and ADP appeared higher in endolymph than in perilymph. The paper proposes using the method to compare genetic, noise-induced and drug-induced hearing loss, and to study other small, complex structures.

My role

First author. The method came out of UF’s Center for Advanced Spatial Biomolecule Research, where I ran MALDI timsTOF imaging day to day. The same NEDC sublimation and recrystallization approach is featured in HTX Imaging Application Note 69, where I am credited as a data contributor.