Background
The presence of amyloid pathology can have a profound effect on the surrounding cellular neighborhood. While this impact has been mainly investigated for amyloid plaques in the context of Alzheimer's disease (AD), other forms of amyloid deposits can also be found in the brain and in other organs. In the pancreas, amyloid deposits consist of islet amyloid polypeptide (IAPP) and are a hallmark of type 2 diabetes (T2D). Notably, T2D has been associated with an increased risk of developing AD, and as such T2D is a common comorbidity of AD. It has therefore been suggested that these diseases may share pathophysiological processes. To advance our understanding in this respect, we compared the cellular and transcriptomic responses related to the proximity of amyloid pathology across the AD brain and T2D pancreas.

Method
Xenium single-cell spatial transcriptomic profiling was applied to tissue sections from a human post-mortem AD brain (150,060 cells) and a T2D pancreas (256,907 cells). Spatial transcriptomics images were integrated with amyloid histopathology images to determine the proximity of individual cells to amyloid deposits. Together with cell type predictions, this enabled the investigation and cross-organ comparison of amyloid-associated changes in cell type composition and gene expression changes.

Result
With respect to cell type composition, in the brain a higher proportion of microglia could be observed close to amyloid pathology, while in the pancreas this was mirrored by a higher proportion of macrophages as well as a higher proportion of activated stellate cells. Cell type specific differential gene expression analysis based on amyloid proximity revealed many cell types with altered gene expression, including astrocytes, microglia, oligodendrocytes and endothelial cells in the brain and acinar, alpha and activated stellate cells in the pancreas. Comparison across organs revealed 16 shared genes differentially expressed with proximity to amyloid deposits, including CAV1, CXCR4, MS4A6A, SNCG, and SOX2.

Conclusion
Here we spatially investigate the impact of amyloid deposits on the cellular and transcriptomic microenvironment in the brain and pancreas. Our analysis revealed a common set of amyloid proximity related genes, providing insight into potentially shared pathological pathways underlying AD and T2D.

Gangliosides are acidic glycosphingolipids, containing ceramide moieties and oligosaccharide chains with one or more sialic acid residue(s) and are highly diverse isomeric structures with distinct biological roles. Matrix-assisted laser desorption/ionization imaging mass spectrometry (MALDI IMS) enables the untargeted spatial analysis of gangliosides, among other biomolecules, directly from tissue sections. Integrating trapped ion mobility spectrometry with MALDI IMS allows for the analysis of isomeric lipid structures in situ. Here, we demonstrate the gas-phase separation and identification of disialoganglioside isomers GD1a and GD1b that differ in the position of a sialic acid residue, in multiple samples, including a standard mixture of both isomers, a biological extract, and directly from thin tissue sections. The unique spatial distributions of GD1a/b (d36:1) and GD1a/b (d38:1) isomers were determined in rat hippocampus and spinal cord tissue sections, demonstrating the ability to structurally characterize and spatially map gangliosides based on both the carbohydrate chain and ceramide moieties.