Spatial Transcriptomics Reveals the Macroscale Signatures of Cortical Cell Types on Anatomic Magnetic Resonance Imaging
Abstract Establishing how intensity in magnetic resonance images (MRIs) covaries with molecular factors can help to derive quasi-mechanistic biological insights from neuroimaging. Gray/white matter contrast (GWC) of $$\:{T}_{1}$$ -weighted MRI is an abundantly available phenotype often used to investigate aging and cognition. Despite growing interest in investigating microstructure using GWC, its relationship to cell type-related tissue properties is understudied. We leverage spatially resolved post-mortem transcriptomic data from six donors to identify associations between cell type marker genes and GWC averaged across cognitively normal living adults ( $$\:N$$ = 577, mean age = 56.2 years). Astrocyte marker gene transcription is significantly associated with cortical variation in mean GWC, referred to as $$\:C$$ . Additionally, estimated relative astrocyte abundance correlates significantly with $$\:C$$ ( $$\:r$$ = 0.409, $$\:p$$ < 0.01). Gene transcription correlations with $$\:C$$ are enriched for oligodendrocyte marker genes only within the frontal lobe, which is known to age fastest. Transcriptions of genes related to iron trafficking and myelination help to identify potential explanations for associations of $$\:C$$ to astrocyte and oligodendrocyte distributions. A multivariate linear model of cell abundances explains 31% ( $$\:r$$ = 0.56) of variation in $$\:C$$ , of which $$\:\sim$$ 40% is attributable to neurons, ~33% to endothelial cells, and ~15.7% to microglia. Findings are replicated in two additional independent cohorts of cognitive normal adults ( $$\:N$$ = 577 in each), indicating that reported associations are not dependent on scanner parameters. Our findings empower interpretation of GWC and its use in studying aging and disease.
Authors
- Nahian F. Chowdhury (ORCID: https://orcid.org/0000-0003-3535-7280)
- Tamara Jafar (ORCID: https://orcid.org/0000-0002-4763-3409)
- Andrei Irimia
Institutions
- University of Southern California (US)
- King's College London (GB)
Publication Details
- Journal
- Neuroinformatics
- Published
- 2026-11-21
- DOI
- https://doi.org/10.1007/s12021-026-09828-8
- Primary Topic
- Single-cell and spatial transcriptomics
- Type
- article
- Field-Weighted Citation Impact
- 0.00