Murine recipients of fully mismatched donor marrow are protected from lethal graft-versus-host disease by the in vivo administration of rapamycin but develop an autoimmune-like syndrome.

We investigated the ability of the macrolide antifungal agent rapamycin (RAPA) to inhibit murine graft-vs-host disease induced across the MHC barrier. An optimum dose (1.5 mg/kg) given for 14 days beginning on the day of transplant (and then three times weekly until 1 mo) effectively and significantly (p < 0.001) protected 80% of irradiated B10.BR recipients of C57Bl/6 bone marrow/spleen grafts for over 90 days, whereas 80% of control mice died by day 37. Using a congenic model in which a mixture of Ly5.1+ bone marrow (T cell-depleted) and Ly5.2+ spleen cells allowed us to distinguish mature and immature cells, we found that RAPA inhibits the splenic expansion of mature donor-derived T cells in B10.BR recipients after bone marrow transplantation. In addition, phenotyping studies revealed that RAPA causes a massive reduction of immature CD4+CD8+ T cells in the thymus, indicating that RAPA probably interferes with maturation of immature CD3-CD4-CD8- T cells to CD4+CD8+ T cells. There was also a predilection toward development or intrathymic retention of the more mature CD3+CD4-CD8+ or CD3+CD4+CD8- cells in the thymus of long term survivors. These same observations were made in different experiments with mice given syngeneic bone marrow transplantation and RAPA. However, RAPA administration was associated with the occurrence of an autoimmune-like syndrome, consisting of ulcerative dermatitis, hepatic bile duct proliferation, and nondestructive lymphoid peribronchiolar infiltration of the lung. RAPA interfered with the deletion of potentially self-reactive T cells that occurs in thymic development. The failure of clonal deletion was observed in allogeneic and syngeneic transplants given RAPA, although only the allografted mice experienced an autoimmune-like syndrome. Some, but not all, of the nondeleted V beta populations were functionally active. These new findings bear certain dissimilarities to the syndrome and lesions observed with cyclosporin A treatment, particularly in the observation of bile duct proliferation and ulcerative skin lesions. Nonetheless, because of the potent effect of RAPA in preventing lethal graft-vs-host induced across the MHC, further investigation of the immune consequences of this highly effective compound is warranted.