Immunomodulation by intrathymic injection of donor leukocytes in rhesus monkeys. Transplantation 2001; 72: 1432.

COMMENTARY The strategy to achieve acquired central immunological tolerance by injection of foreign antigens into the thymus was popularized in the early 1990s. It was first demonstrated by Posselt et al. that an intrathymic allogeneic islet graft not only survived indefinitely under the cover of a brief course of immunosuppression, but also, lead to a state of immunological tolerance as indicated by acceptance of a second peripheral graft without additional immunosuppression (1). It was hypothesized that exposure of developing T cells to foreign antigen results in deletion of those clones reactive to the donor in a manner mimicking self-tolerance induction. The brief immunosuppression required at the time of intrathymic inoculation was postulated to be necessary to eliminate the already mature donor reactive T cells present in peripheral immune tissues. Many investigators have now demonstrated that this strategy is effective at attaining acceptance of a variety of allogeneic tissues (including heart, pancreas, liver, and islet) using small animal models (2). The broad success reported in rodent systems led to hope that this approach might ultimately be applied to achieve a long-lived state of transplantation tolerance in man. Further support for the general principal that exposure of developing T cells in the thymus to a foreign antigen can lead to immune tolerance is found in studies of autoimmunity, in which intrathymic inoculation with the relevant autoantigen apparently leads to eradication of T cells clones that would otherwise develop auto-aggressive behavior. This was first demonstrated by transplantation of pancreatic islets to the thymus of diabetes prone BB rats (3). This protocol efficiently prevented development of autoimmune diabetes despite the fact that the precise islet β cell autoantigen involved in triggering and targeting the diabetogenic autoimmune cascade has not been identified. The precise immunologic mechanisms that underlie these studies of autoimmunity and allogeneic tissue transplantation are poorly understood. A number of investigators have documented clonal deletion follows intrathymic antigen exposure however, this is rarely complete (2). More puzzling is evidence that intrathymic inoculation with isolated peptide or protein antigen can often induce a state of systemic tolerance without peripheral depletion of already mature T cells by immunosuppression. This suggests strongly that clonal deletion may not be the sole mechanism mediating acquired central tolerance after Pintrathymic antigen exposure. Induction of regulatory T cells by thymic inoculation has been proposed to explain these findings but clear evidence is lacking. Despite the widespread success in small animal models, translation of the general principals defined in these experiments to the clinical setting remains elusive. To date, the most encouraging work comes from experiments in large animal using pancreatic islets. Brayman et al. studying dogs (4), and Mullen et al. studying miniature swine (5), found that the thymus may be a site suitable for engraftment of allogeneic islet transplants. However, a state of immunological tolerance has not been demonstrated conclusively. The work by Jonker et al. in this issue is important in that it represents perhaps the most forthright attempt to evaluate the translational potential of an acquired central tolerance system. In this work, rhesus monkeys were inoculated i.v., intracutaneously, or intrathymically with peripheral blood lymphocytes from allogeneic donors. Donors and recipients were selected based on serological mismatching for at least a single class I and class II locus. To deplete peripheral T cells, ATG was administered to some groups. Three to four weeks after donor cell injection, animals were skin grafted and followed both for the survival of the graft and for donor responses in in vitro assays of immune reactivity. Intrathymic inoculation with allogeneic cells in recipients not treated with ATG did not lead to obvious prolongation and may have sensitized one recipient. Of note, the sensitized animal was found to have an “involuted” thymus suggesting technical difficulties associated with the injection process. In the group receiving both intrathymic allogeneic cells and ATG (there were only three animals), there may have been a slight prolongation of the survival in one graft, however, the results were not statistically or biologically significant. Cellular and humoral reactivity was evident after intracutaneous and i.v. cellular treatment. This was not evident, however, in animals receiving intrathymic cells and ATG. At face value, the work of Jonker seems to dispel the notion of intrathymic donor antigen treatment as suitable approach to achieve clinical transplantation tolerance. Perhaps the most incisive and pertinent question posed by this work, however, is not whether the protocol used would have succeeded if applied to humans but whether it would have succeeded if applied to mice. In the now vast experimental literature that has examined various protocols of intrathymic antigen injection induced transplantation tolerance in rodents, there is only a single report in which tolerance to skin grafts has been achieved (6). In this case, the donor/recipient mouse pair was selected based on a MHC class I disparity, whereas, the same protocol being ineffective for fully allogeneic donor recipient pairs. In contrast to these results, acceptance of heart, vascularized pancreas, islets, and liver grafts has been achieved almost routinely. Why tolerance to skin has proved so challenging remains enigmatic and highlights how little we know about the mechanisms mediating tolerance development. Although we have suggested that acquired central tolerance recapitulates the mechanisms of self-tolerance, this obviously over simplifies a complex process. The fact that allogeneic tolerance and skin graft acceptance consistently occurs with classical neonatal tolerance induction (7) but not tolerance induction after intrathymic injection of donor lymphoid cells indicates that our understanding of this topic is incomplete. Thus, the results of Jonker et al. although provocative in their lack of success even in the small number of animals studied, will require further experiments with other allogeneic tissue or organ grafts to assess definitively the parallel to small animal work. Moreover, although such studies may help to shed light on the human application of this approach, it seems likely that the ultimate success with tolerance induction protocols in the clinical setting will await a better mechanistic understanding of induced tolerance. This will almost certainly be first revealed by small animal experiments.