Back

High-Dimensional Single-Cell Analysis Reveals Coordinated Age-Dependent Neuroinflammatory Microglia-T cell Circuits in the Brain

Ali, M. A.; Siam, M. H. B.; Vardaman, D.; Bolding, C.; Brazell, J. N.; Whatley, A. D.; Risley, C. A.; Tidwell, H.; Hossain, S. N.; Samal, J.; Harms, A. S.; Tyrrell, D. J.

2025-12-13 immunology
10.64898/2025.12.10.693494 bioRxiv
Show abstract

Aging and cerebrovascular pathology drive neuroinflammation in vascular dementia (VaD) but immune mechanisms underlying this interplay remain unresolved. Leveraging multi-modal high-dimensional imaging, flow cytometry, and split pool ligation transcriptomic sequencing in a mouse model of VaD, we constructed a brain immune cell atlas spanning young and aged mice in health and disease. We profiled microglia, T cells, macrophages, neutrophils, and B cells and integrated transcriptomics, cell-cell communication, multiplex imaging, and comparative analysis with human microglia. We found striking depletion of Ccr7+ naive T cells and expansion of Gzmk+ cytotoxic Cd8+ effector memory T cells in the aging brain. At the same time, microglia shifted toward a pro-inflammatory state with enhanced activity of major histocompatibility class complex I (MHC-I) to T cell receptor and co-stimulation from CD86 to CD28. These shifts suggest enhanced neuroinflammatory polarization within the aged brain and in VaD. These signals were strongest from activated microglia to Gzmk+ Cd8+ TEM cells, indicating that age-related microglial polarization may sustain cytotoxic T cell activation in the aged brain. Our findings suggest pro-inflammatory microglia and Gzmk+ CD8+ TEM cells are central drivers of immune brain aging and highlights a therapeutic potential to disrupt age-related neuroinflammatory cascades in VaD.

Matching journals

The top 7 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.