Herein, we assess the complementarity
and complexity of data that
can be detected within mammalian lipidome mass spectrometry imaging
(MSI) via matrix-assisted laser desorption ionization
(MALDI) and nanospray desorption electrospray ionization (nano-DESI).
We do so by employing 21 T Fourier transform ion cyclotron resonance
mass spectrometry (FTICR-MS) with absorption mode FT processing in
both cases, allowing unmatched mass resolving power per unit time
(≥613k at m/z 760, 1.536
s transients). While our results demonstrated that molecular coverage
and dynamic range capabilities were greater in MALDI analysis, nano-DESI
provided superior mass error, and all annotations for both modes had
sub-ppm error. Taken together, these experiments highlight the coverage
of 1676 lipids and serve as a functional guide for expected lipidome
complexity within nano-DESI-MSI and MALDI-MSI. To further assess the
lipidome complexity, mass splits (i.e., the difference
in mass between neighboring peaks) within single pixels were collated
across all pixels from each respective MSI experiment. The spatial
localization of these mass splits was powerful in informing whether
the observed mass splits were biological or artificial (e.g., matrix related). Mass splits down to 2.4 mDa were observed (i.e., sodium adduct ambiguity) in each experiment, and both
modalities highlighted comparable degrees of lipidome complexity.
Further, we highlight the persistence of certain mass splits (e.g., 8.9 mDa; double bond ambiguity) independent of ionization
biases. We also evaluate the need for ultrahigh mass resolving power
for mass splits ≤4.6 mDa (potassium adduct ambiguity) at m/z > 1000, which may only be resolved
by advanced FTICR-MS instrumentation.