Correlative Chemical Imaging and Spatial Chemometrics Delineate Alzheimer Plaque Heterogeneity at High Spatial Resolution

We present a novel, correlative chemical imaging strategy based on multimodal matrix-assisted laser desorption/ ionization (MALDI) mass spectrometry imaging (MSI), hyperspectral microscopy, and spatial chemometrics. Our workflow overcomes challenges associated with correlative MSI data acquisition and alignment by implementing 1 + 1-evolutionary image registration for precise geometric alignment of multimodal imaging data and their integration in a common, truly multimodal imaging data matrix with maintained MSI resolution (10 mu m). This enabled multivariate statistical modeling of multimodal imaging data using a novel multiblock orthogonal component analysis approach to identify covariations of biochemical signatures between and within imaging modalities at MSI pixel resolution. We demonstrate the method's potential through its application toward delineating chemical traits of Alzheimer's disease (AD) pathology. Here, trimodal MALDI MSI of transgenic AD mouse brain delineates beta-amyloid (A beta) plaque-associated co-localization of lipids and A beta peptides. Finally, we establish an improved image fusion approach for correlative MSI and functional fluorescence microscopy. This allowed for high spatial resolution (300 nm) prediction of correlative, multimodal MSI signatures toward distinct amyloid structures within single plaque features critically implicated in A beta pathogenicity.

Automated summary

We propose a novel correlative chemical imaging strategy that integrates multimodal matrix-assisted laser desorption/ionization (MALDI) mass spectrometry imaging (MSI), hyperspectral microscopy, and spatial chemometrics. Our approach overcomes challenges associated with data acquisition and alignment by using evolutionary image registration and multiblock orthogonal component analysis. This allows for the identification of covariations of biochemical signatures between and within imaging modalities. We demonstrate the potential of our method in delineating chemical traits of Alzheimer's disease pathology and improving image fusion for correlative MSI and fluorescence microscopy.