Dieter Kabelitz & Ottmar Janssen

Department of Immunology, Paul-Ehrlich-Institute, Langen, Germany

Received 1/24/97; Accepted 1/31/97 On-line 2/15/97

The elimination of potentially self-reactive thymocytes during intrathymic T-cell development involves apoptotic mechanisms. The above summarized observations indicating that antigen can also trigger AICD in mature peripheral T-lymphocytes suggest that programmed cell death also plays a role in the establishment of peripheral tolerance (12). Moreover, the intentional induction of AICD might form the basis for novel immunotherapeutical strategies.

4.1 Role of AICD in peripheral tolerance

While it is clear that programmed cell death is the major mechanism of the intrathymic elimination of self-reactive lymphocytes, it is less obvious that apoptotic deletion contributes to the establishment of tolerance to foreign antigens in the peripheral immune system. The establishment and maintenance of peripheral tolerance is a multi-step process that may involve functional "anergy" (146, 147), downmodulation of cell surface TCR expression (147), as well as physical elimination of antigen-reactive T-cells (93, 149, 150; see ref. 151 for review). In the setting of allograft transplantation, it is highly desirable to induce tolerance in the transplant recipient towards donor MHC and non-MHC antigens. Pearson and coworkers reported that brief treatment of C3H/He recipient mice with anti-CD4 mAb together with C57/BL10 donor cells induced specific tolerance of subsequent C57/BL10 cardiac allografts (152). It is possible (although not proven in this study) that CD4 cross-linking primed the recipient T-cells for AICD following TCR-mediated recognition of donor alloantigens, in accordance with the in vitro studies (86). In support of this possibility, it was reported by Markmann and coworkers that the clonal deletion of antigen-reactive T-cells following the intrathymic inoculation with Mls-disparate lymphoid cells required the additional application of anti-CD4 mAb (153). In experimental models of organ transplantation, the intrathymic injection of donor lymphoid cells has been successfully used to prolong cardiac and liver allograft survival, possibly through the induction of AICD in alloreactive T-cells (154, 155). Furthermore, some immunosuppressive drugs that are used in clinical transplantatiom medicine might in part exert their effect through the modulation of AICD. While cyclosporin A (CsA) and FK506 inhibit AICD triggered through CD3 or CD2 in vitro (156, 157), both drugs can actually enhance apoptosis of immature thymocytes in vivo (158, 159). Interestingly, FK506 but not CsA, also augmented superantigen-induced AICD of peripheral T-lymphocytes following the injection of SEB into mice (159). Such a mechanism might contribute to the clinical efficacy of FK506. Finally, it is noteworthy that the clinically used anti-CD3 mAb is a potent inducer of in vitro AICD in activated human T-lymphocytes (76). Therefore, AICD might contribute to the immunosuppressive effect of anti-CD3 treatment, in addition to other mechanisms such as modulation of cell surface CD3 expression.

Recent studies suggest that the Fas/Fas-L system also plays a role in the prevention of graft rejection. Bellgrau and coworkers reported that testicular allografts expressing Fas-L escaped rejection by inducing apoptosis in Fas-expressing recipient T-cells activated by graft antigens (160). It should be noted, however, that the expression of Fas-L might not always be beneficial for graft survival. In contrast to Bellgrau et al. (160), Yagita and coworkers observed a severe inflammatory rejection of Fas-L expressing baby hamster kidney (BHK) fibroblasts, apparently mediated by neutrophils (161). Nevertheless, it appears that the constitutive expression of Fas-L forms the basis for the maintenance of immunological tolerance in so-called "immune privileged sites" such as the anterior chamber of the eye (162).

4.2 Therapeutical perspectives in autoimmune diseases

The discovery that antigen can delete antigen-reactive mature T-lymphocytes through the induction of AICD has raised great interest in the potential clinical application of this approach in the treatment of certain autoimmune diseases. A widely studied model of T-cell mediated autoimmune disease is experimental autoimmune encephalomyelitis (EAE). The injection of the relevant autoantigen, myelin basic protein (MBP), or the transfer of MBP-specific T-cells, triggers EAE in susceptible rats and mice. In situ, apoptotic T-cells can be detected in the central nervous system of Lewis rats suffering from EAE (163). During the spontaneous recovery from EAE, antigen-specific down-regulation of MBP-reactive T-cells occurs, due to the selective apoptotic elimination of autoreactive T-lymphocytes from the target organ (164, 165). These results suggest the possibility that the application of specific (auto)antigen may be used to delete unwanted autoreactive T-cells by the induction of AICD. In fact, several experimental autoimmune diseases can be prevented by the injection of large doses of the relevant autoantigen (166, 167). Although it was not reported in these studies whether apoptosis was involved in the induction of tolerance, there are well-documented examples that high doses of autoantigen can delete autoreactive T-cells via AICD. Critchfield and coworkers induced EAE in mice by the adoptive transfer of MBP peptide-reactive T-cells. In contrast to mice injected with a control antigen, mice injected with the relevant MBP peptide had a dramatic deletion of T-cells expressing the MBP-reactive TCR, and did not develop disease (168). Peripheral deletion of antigen-reactive T-lymphocytes can also be achieved by other routes of antigen application. In the studies of Chen and coworkers, the oral application of high concentrations of ovalbumin triggered the deletion of antigen-reactive T-cells in the Peyer's patches of mice transgenic for the ovalbumin-specific TCR (169). Thus, the deletion of antigen-reactive T-cells via the induction of AICD is a promising new approach in the treatment of autoimmune and certain other diseases (170, 171). However, in contrast to experimental models, the relevant autoantigen in human disease is frequently unknown, and antigen-independent strategies of modulation of AICD and apoptosis need to be pursued.