Back

Single Cell Mapping Identifies CD14+ Macrophages as Central Orchestrators of CD8+ T Cell Driven Immune Niches in clinical Type 1 diabetes

Shivamadhu, M. C.; Zhang, X.; Yechoor, V. K.; Prentice, K.; Razani, B.; Wheeler, M. B.; Khan, M. S. R.

2026-08-12 pathology
10.64898/2026.08.06.743319 bioRxiv
Show abstract

Type 1 diabetes (T1D) is an autoimmune disease characterized by CD8 T cell-mediated destruction of pancreatic {beta} cells; however, the cellular interactions that organize immune activation within human islets remain poorly understood. Here, we integrated thirteen CD45 immune cell single-cell RNA sequencing datasets from human islets spanning non-diabetic donors, stage 3 T1D, and type 2 diabetes (T2D) to comprehensively define immune cell heterogeneity and decipher the intercellular communication networks that drive islet autoimmunity. We identified distinct macrophage states, including CD14 inflammatory macrophages, CD14/TREM2 macrophages, and quiescent-like macrophages, together with CD8 T cells and mast cells. Trajectory and communication analyses revealed CD14 macrophages as central immune hubs that coordinate antigen presentation, costimulatory signaling, and inflammatory chemokine production. Compared with non-diabetic and type 2 diabetic islets, T1D macrophages displayed a disease-specific inflammatory program characterized by enhanced TNF, IL18, CCL3, CCL4, CCL5, and ICOSLG expression, supporting CD8 T cell recruitment and activation. Spatial transcriptomic analysis of human T1D pancreas further demonstrated a {beta}-cell-macrophage-CD8 T cell inflammatory niche, where macrophage-derived CCL3/CCL4/CCL5 and CD8 T cell-expressed CCR5 suggest a chemokine-mediated mechanism of immune targeting. Together, these findings identify CD14 macrophages as key orchestrators of a feed-forward inflammatory circuit driving human islet autoimmunity.

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.