Organ transplantation has become increasingly important as a treatment for many human diseases. Despite the dramatic improvements in immunosuppression in recent years, acute and chronic rejection remain significant problems. It has become increasingly evident that the presence of T cell memory correlates with both acute and chronic rejection episodes. Endothelial cells (EC) have been shown to preferentially activate memory T cells and, as the lining of every transplanted organ, they are in a unique position to provide signals to alloreactive memory T cells. EC activation of memory T cells depends in part on the costimulatory molecule LFA-3 in addition to other signals. The nature of these other signals is not well understood. In this dissertation I further explore the mechanisms by which EC activate memory T cells and investigate the role that these mechanisms play in a model of memory T cell mediated allograft rejection.

Several newly described costimulatory pathways, ICOS-ICOSL, 4-1BB-41BBL, and OX40-OX40L, have recently been recognized as important players in the generation and function of memory T cells. EC can express all three of the ligands from these pathways and also increase expression in response to the cytokine tumor necrosis factor (TNF) and in co-cultures with T cells. Furthermore, blockade of these pathways using monoclonal antibodies in co-cultures causes reductions in T cell production of IFN-γ and IL-2 as well as decreased T cell proliferation.

To examine the contribution of memory T cells to allograft rejection in vivo, I describe a chimeric model of human skin grafted onto SCID mice that are adoptively transferred with various subsets of human T cells. In this model, rejection of the human skin by the transferred T cells is mediated by memory T cells alone. Using this model, the effect of different costimulatory pathways on allograft rejection is tested. Interestingly, blockade of any of the three costimulatory pathways can diminish the T cell response to the allograft. The role of the B7-CD28 pathway is also examined and found to partially contribute to the memory T cell response.

Finally, using observations that memory T cells can be further divided into two functionally distinct subsets known as central and effector memory T cells, the response of these two subsets to EC is explored. Central memory T cells respond to EC by producing IL-2 while effector memory T cells generate IFN-γ. The basis for this difference may be due to the differential expression and sensitivity of the two subsets to costimulatoryn by CD27 and CD28. Further, in the chimeric human-SCID model effector memory T cells alone can mediate allograft rejection while central memory T cells cannot.

Our results strongly support the idea that the EC capacity to activate memory T cells and its subsets depends in part on the CD28-B7 pathway and memory T cell specific costimulatory molecules. Furthermore, targeting CD28 or these molecules in vivo can attenuate allograft rejection mediated by memory T cells.