The accumulation of T cells in the rheumatoid synovium is thought to be antigen-driven. This implies a highly permissible trafficking of T cells to and from peripheral circulation through the synovial endothelium. We propose that among the factors critical to T cell infiltration of synovial tissues is thrombospondin-1 (TSP) which has been reported to deposit on vessel walls and subendothelium of synovial tissues. While TSP is known to be transiently released into the extracellular matrix by thrombin-activated platelets, its cellular source in the synovium is not known. However, in vitro studies show that endothelial cells can produce TSP, along with its receptor CD36, in response to a variety of cytokines, such IFNgamma, IL6, and TGFbeta. Inasmuch as these cytokines are abundantly expressed in the synovium, it is likely that TSP maybe produced by the endothelial cells of the tissue. Therefore, we propose to examine the patterns of TSP and CD36 expression in the synovium. We will determine whether or not there is a coordinate expression of this receptor-ligand pair on endothelial cells. Since TSP can also bind CD47, an adhesion molecule that is constitutively expressed on T cells, we will evaluate the hypothesis that a CD47-TSP-CD36 trimolecular complex promotes extravasation of T cells. Experiments have been designed to specifically evaluate the specificities of this trimolecular interaction as it relates to the transendothelial migration of T cells. We also will assess whether this complex synergizes with other adhesion molecules. Inasmuch as CD36 is also expressed on macrophages, which are in abundance in the tissue, we will also evaluate the role of the CD47-TSP-CD36 complex in T cell activation. This is based on our preliminary studies showing that crosslinking of CD47 by specific antibody or immobilized TSP is comitogenic with anti-CD3 for T cells in vitro. These results suggest a T cell costimulatory function for CD47 that is reminiscent of CD28- dependent costimulation. In the context of the synovium, we propose that in vivo, CD47-TSP-CD36 interaction plays an important role in the activation of the CD4+ CD28null T cells. These unusual cells have been shown previously to accumulate in high frequencies among patients with rheumatoid arthritis, and have been correlated with disease severity in some case. An analysis of CD47-dependent effector functions of CD4+CD28null T cells could provide valuable insights on the role of these unusual cells in the pathogenesis of rheumatoid synovitis.