Introduction: Tolerance induction would avoid xenograft rejection and the need for toxic levels of non-specific immunosuppressive therapy. We have developed two complementary approaches to inducing human immune tolerance to the pig, namely mixed xenogeneic chimerism, which tolerizes T cells, B cells and NK cells, and porcine thymic transplantation, which induces T cell tolerance by combined clonal deletion and Treg function. However, in immunocompetent hosts, thymus transplantation requires recipient thymectomy. Furthermore, human T cells developing in porcine thymus grafts lack Tregs specific for human tissue-restricted antigens and have weak HLA-restricted antigen recognition. To circumvent these problems, we combined mixed xenogeneic chimerism with double thymus transplantation, modelling porcine thymic transplantation and mixed chimerism in a human without native thymectomy.

Methods: Thymectomized pig cytokine transgenic (PCT)-NSG mice received a thymocyte-depleted swine fetal thymus graft and a human fetal thymus graft under the capsule of each kidney. Human fetal liver CD34+ cells were given i.v. and half of the mice also received hCD47-transgenic pig bone marrow cells. Animals were sacrificed 19 weeks post-transplantation and thymocytes and splenocytes were analyzed and compared between groups. Splenic T cells, enriched by depletion of mouse cells and human B cells, were assessed for proliferative responses in CFSE-MLR.

Results: Overall reconstitution of human cells (T cells, B cells and monocytes) in peripheral blood was delayed and reduced in mixed chimeric mice. Pig chimerism consisting mainly of T cells was persistent in chimeric mice. Overall human cell reconstitution was also decreased in spleen and bone marrow in chimeric compared to non-chimeric mice. Pig chimerism was detected in both spleen and bone marrow of chimeric mice. Human T cells from both groups were hyporesponsive to autologous human stimulators while mounting robust responses to human allogeneic stimulators (Fig.1 and Fig.2 A), indicating tolerance to the human autologous antigens. Allogeneic control T cells showed a strong response to the autologous dendritic cells (Fig.2 B). While human T cells from chimeric mice showed specific hyporesponsiveness to the donor pig with robust third-party pig responses (Fig.2 A), human T cells from two of three non-chimeric mice proliferated robustly to the donor pig (Fig.1). These data suggest that specific human T cell tolerance to the donor pig xenoantigens was achieved in the presence of mixed chimerism and a human plus a porcine thymus.

Conclusion: Our data suggest that combining porcine thymus transplantation with mixed xenogeneic chimerism may achieve optimal tolerance while ensuring recognition of pathogens in the context of human and porcine MHC molecules without requiring thymectomy of the host.