Back

Endothelial Caveolin-1 and CXCL10 promote transcellular migration of autoreactive T cells across the blood-brain barrier

Trevino, T. N.; Almousawi, A. A.; Ochoa-Raya, A.; Zemanski, K.; Oliveira, S. D.; Marottoli, F. M.; Tai, L. M.; Minshall, R. D.; Lutz, S. E.

2022-11-16 immunology
10.1101/2022.11.15.516689 bioRxiv
Show abstract

CXCL10 is an interferon-inducible chemokine that can recruit CXCR3+ leukocytes to the central nervous system, leading to neuroinflammation, demyelination, and neuronal losses. How CXCL10 promotes leukocyte extravasation and diapedesis across the blood-brain barrier - formed by brain endothelial cells - is poorly understood. Here, we report that CXCL10 mediates CD4+ T cell migration through the brain endothelial cell cytoplasm (transcellular), but not cell-cell junctions (paracellular), via the vesicular trafficking protein Caveolin-1. Caveolin-1 promotes CXCL10 aggregation into cytoplasmic stores in brain endothelial cells in vitro to provide the local, high concentration necessary for recruitment of CXCR3+ leukocytes. This process also requires LFA-1 activity. In the absence of Caveolin-1, endothelial CXCL10 is secreted, and the local signaling cues are lost. Consistent with our in vitro data, genetic ablation of Caveolin-1 in endothelial cells reduces the severity of active experimental autoimmune encephalomyelitis (EAE), a murine model for multiple sclerosis, by decreasing the infiltration of CXCR3+ T cells into the CNS. Moreover, loss of Caveolin-1 protects against the adoptive transfer of autoreactive T cells. Our findings establish a novel mechanism by which brain endothelial cells utilize Caveolin-1 dependent CXCL10 intracellular stores to license T cells for transcellular migration across the blood-brain barrier.

Matching journals

The top 4 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.