Back

Glial subcellular specialisation resolved with high resolution spatial transcriptomics

Boulger, S. L.; Melgosa-Ecenarro, L.; Zielonka, M.; Pilch, K. S.; Wijesinghe, S. S.; Radulescu, C. I.; Matthews, P. M.; Barnes, S. J.; Mallach, A.

2026-07-17 neuroscience
10.64898/2026.07.12.737168 bioRxiv
Show abstract

Cells in the brain have complex structures with extended processes. This complex morphology supports diverse specialized functions in health and disease, and specifically, cell processes appear to be critical for cellular integration and signalling. Here, we developed a new spatial averaging framework to recover and interrogate molecular phenotypes of glial processes in spatial transcriptomics (ST) data. We characterised cell type specific signatures associated with processes of astrocytes and microglia in both mouse and human brain tissue. Astrocytic processes were enriched for transcripts related to neuronal support relative to their soma, while microglial processes preferentially expressed genes linked to specific microglial states. When investigated in tissue from brains with Alzheimers Disease (AD) pathology, we found that local amyloid-{beta} pathology was associated with subcellular differences in transcriptomes in both an amyloid-{beta} mouse model and human AD patient tissue. Specifically, astrocytic and microglial processes oriented toward amyloid-{beta} plaques exhibited distinct molecular changes in comparison to processes extending away into plaque free areas, suggesting polarized glial responses to pathology. Our work thus outlines a general method for the selective characterisation of transcriptomics of glial processes in mouse and human ST data and provides evidence for differential transcriptomic responses between the soma and processes of glia in health and disease. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=139 SRC="FIGDIR/small/737168v1_ufig1.gif" ALT="Figure 1"> View larger version (53K): org.highwire.dtl.DTLVardef@157455corg.highwire.dtl.DTLVardef@8ac0a2org.highwire.dtl.DTLVardef@16ccbf5org.highwire.dtl.DTLVardef@1c24cd0_HPS_FORMAT_FIGEXP M_FIG C_FIG

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.