Induction of hyperacute rejection of skin allografts by CD8+ lymphocytes

Background. Second-set rejection is generally regarded as a phenomenon mainly mediated by humoral cytotoxic antibodies, although a few discordant data have been presented. In the reported experiments, we have taken advantage of the absence of production of specific cytotoxic alloantibodies contrasting with the normal development of transplantation cellular immunity, in two murine models: chimeric mice and RAG mice. Methods. Chimeras (BALB/c-->CBA) were obtained by transplantation of 2x10(7) fetal liver cells from BALB/c (H-2d) mice to lethally irradiated CBA (H-2k) mice. After hyperimmunization with third-party C57/ BL6 (B6) (H-2b) skin transplants and with injections of 2x10(7) B6 spleen cells, antibody production, and skin graft survival were analyzed. To identify further the factors or cells responsible for accelerated rejection of B6 skin transplants in hyperimmunized chimeras, transfer experiments were carried out involving the injection of serum, whole spleen cells, spleen T cells, spleen CD8(+) T cells or spleen CD4(+) T cells from chimeras into BALB/c mice that had received 6 Gy irradiation. The recipient mice were then grafted with B6 skin. Similarly, the immunodeficient RAG mice were used to construct a model of recipient animals with anti-H-2d hyperimmunized B6 T cells in the total absence of antibody. Results. In chimeras, anti-B6 cytotoxic antibodies were not detectable in any of hyperimmunized chimeric mice, yet accelerated rejection of B6 skin transplant occurred: a graft survival of 8.6+/-0.5 days (d), comparable to 8.9+/-0.8 d survival in CBA control mice subjected to the same hyperimmunization procedure, and significantly shorter than that in non-hyperimmunized (BALB/c-->CBA) chimeras (11.6+/-0.5 d) or in non-hyperimmunized CBA control mice (12.1+/-0.6 d). High titers of anti-B6 cytotoxic antibodies were present in the serum of hyperimmunized CBA control mice. In transfer experiments, the graft survival was over 14 d in mice treated with irradiation alone, with irradiation + serum or with irradiation + CD4(+) T cells. It was significantly shorter in mice treated with irradiation + whole spleen cells, with irradiation + T cells or with irradiation + CD8(+) T cells (8.9+/-0.8 d). Similarly, in immunodeficient RAG mice, reconstitution of the T cell compartment with T cells from hyperimmunized B6 mice led to accelerated rejection of BALB/c skin allografts (11.4+/-1.1 d vs. 18.8+/-0.8 d when T cells were provided by nonimmunized mice). In a second transfer of cells from these reconstituted RAG: mice into naive RAG mice, CD8(+) T cells were shown to induce accelerated rejection of skin allografts (12.0+/-0.6 d) whereas CD4(+) T cells were much less efficient (16.5+/-0.1 d). Conclusion. These data indicate that T cells, and especially the CD8(+) subset, can be responsible for second-set rejection in the absence of anti-donor antibodies in chimeric and RAG mouse models. These sensitized CD8(+) T cells are also likely to play an important role in normal mice, in addition to that of cytotoxic antibodies.