Detection of donor-derived cells by polymerase chain reaction in neonatally tolerant mice. Microchimerism fails to predict tolerance.

Although it has long been appreciated that establishment of chimerism is important in the acquisition and maintenance of allograft tolerance, the importance of this relationship has been reemphasized recently. Using the exquisite sensitivity of the polymerase chain reaction we have studied qualitatively and quantitatively the presence of donor-derived chimeric cells in relation to the ability of neonatally injected mice to display skin graft tolerance or rejection. We have found that virtually all mice that receive neonatal injections of F1 hematopoietic cells acquire donor-derived chimerism that is detectable in blood, spleen, lymph nodes, and thymus. Surprisingly, neither the presence nor the quantity of chimeric cells predicts whether a particular neonatally injected mouse will accept or reject donor-specific skin allografts. Moreover, whether the test skin allograft is accepted (tolerance) or rejected by neonatally injected mice, chimerism typically remains detectable within recipient lymphoid tissues. In functionally tolerant mice, challenge with a test skin allograft actually leads to a remarkable expansion in donor-derived genetic sequences, implying that donor-derived cells have been induced by the graft to undergo proliferation. Since persistence of chimerism without proliferation after test grafting is characteristic of mice that fail to display tolerance, we believe that achievement of a threshold level of donor-derived alloantigen may be necessary to retain the tolerant state. We conclude that chimerism is essential for the induction of neonatally induced tolerance, and its expansion may play an important role in the maintenance of that tolerance, when challenged by an allograft.

Cellular mechanisms that maintain neonatally-induced tolerance of class II alloantigens. Evidence that factor-mediated suppression silences cytotoxic T cell activity.

Virtually all neonatal mice of the A strain background are rendered profoundly and permanently tolerant of test skin allografts if they receive an i.v. inoculation of semiallogeneic hematopoietic cells expressing class II disparate alloantigens. After neonatally injected mice reach immunologic maturity, their lymphoid organs have been found to contain 1) tolerogen-specific CD4+ T cells that proliferate and secrete IL-4 when stimulated in vitro with class II tolerogens and 2) tolerogen-specific CD8+ T cell that fail to differentiate into cytotoxic effector cells. In this study, experiments are described that investigate the possibility that tolerance is maintained by regulatory T cells of the Th2-type. When A.TH T cells were stimulated in vitro with A.TL alloantigens in the presence of lymphoid cells from tolerant mice, tolerogen-specific cytotoxic T cells responses were absent or greatly diminished. When regulatory cells from tolerant mice were fractionated and tested, the cell type responsible for suppression proved to be CD4+ class II tolerogen-specific and gamma irradiation sensitive. Moreover, suppression of tolerogen-specific cytotoxic T cell generation was achieved when regulatory cells and naive responder cells were separated by a transwell barrier and supernatants harvested from cultures in which tolerant cells were stimulated specifically with class II tolerogens also inhibited cytotoxic T cell generation. Thus, suppression appears to be mediated by a soluble factor(s) produced by regulatory T cells. We conclude that tolerance of class II alloantigens is maintained, at least in part, by regulatory cells of the Th2-type that secrete factors that suppress the generation of tolerogen-specific effector cells required for rejection of solid tissue allografts.

Cellular mechanisms that maintain neonatally-induced tolerance of class II alloantigens. Evidence that precursor cytotoxic T cells are present but silenced.

Clonal deletion of alloantigen-specific lymphocytes was the first explanation advanced to explain why neonatal mice that receive injections of alloantigenic hematopoietic cells mature into adults that accept donor-derived skin allografts indefinitely, i.e., they are immunologically tolerant. However, numerous other passive and active regulatory mechanisms have been invoked to explain neonatal transplantation tolerance. In A strain mice (A.TH, las, A.TL, lak) rendered tolerant of class II-only alloantigens, formal evidence exists demonstrating that tolerogen-reactive T cells are not eliminated. In fact, tolerogen-reactive T cells are present in peripheral lymphoid organs and can secrete lymphokines (IL-2/IL-4) when stimulated with tolerogen-bearing cells in vitro. Despite the presence of cytokines in these cultures, class II-specific T cells are usually not generated, raising the possibility that selective deletion of these cells may contribute to the tolerant state. To examine this issue, limiting dilution analysis was performed and revealed that tolerant mice possess significantly diminished precursor cytotoxic T cell frequencies. Virtually all cytotoxic T cells generated by normal A.TH mice in response to A.TL class II Ags are CD8+ cells. Moreover, the frequency of donor I-E reactive V beta 5 cells among CD4+ and CD8+ subpopulations among tolerant mice was comparable to naive mice. This suggests that the peripheral lymphoid organs of tolerant mice are functionally deleted of tolerogen-specific cytotoxic T cells and that tolerogen-specific CD8+ T cells are present in normal numbers of tolerant mice. Therefore, this circumstantial evidence implies that tolerogen-specific T cells have not been physically eliminated in class II tolerant mice. Instead, either tolerogen-specific CD8+ T cells have been rendered anergic or active suppression prevents their activation in vitro and presumably in vivo.

Ontogeny of tolerogen-responsive lymphocytes following neonatal inoculation of class II disparate semiallogeneic cells.

Spleens of adult mice of the A strain background that were rendered tolerant as neonates of class II alloantigens only (A.TH tolerant of A.TL, A.TL tolerant of A.TH) contain large numbers of tolerogen-responsive T cells, many of which secrete IL-4, but not IL-2. Since these spleens also contain suppressor cells that can adoptively transfer skin allograft acceptance in vivo and can prevent generation of class II-specific cytotoxic T cells in vitro, it is important to determine the origins during postnatal development of these cells. Class II disparate, semiallogeneic haematopoietic cells were injected into newborn A.TH and A.TL mice. Periodically thereafter (1 to 60 days post-injection, but prior to challenge with a tolerogen-bearing skin graft), thymocytes and splenocytes from these mice were examined in vitro for tolerogen-specific reactivity in mixed lymphocyte reactions during which proliferation and IL-2 and IL-4 production were assayed. Within 24 hours of neonatal injection, the thymus and spleens of injected mice were profoundly depleted of tolerogen-responsive T cells. However, there was no commensurate loss of I-E-related V beta 5+ cells in the thymus of A.TH mice that received neonatal inoculations of I-E-bearing A.TL cells. During the ensuing weeks, tolerogen-responsive proliferative and IL-4-secreting T cells were detected in thymus and spleen. However, not until after the mice were more than 60 days of age were tolerogen-responsive cells able to secrete IL-2. Since physical clonal deletion of tolerogen-related V beta 5+ cells is a characteristic of neither neonatal nor adult A.TH and A.TL mice that received injections of semiallogeneic cells at birth, and since tolerogen-responsive IL-4 producing cells exist in adult mice that have permanently accepted (A.TH x A.TL)F1 skin grafts, our results imply that the tolerogen-responsive T cells detected in adult tolerant mice are descendants of the novel IL-4-producing T cells that arise in the thymus almost immediately after the tolerance conferring inoculum of semiallogeneic cells. The possible mechanisms responsible for generation of IL-4-producing, tolerogen-responsive T cells and the role of these cells in maintenance of tolerance of class II alloantigens are discussed.

Influence of I-E expression on induction of neonatal transplantation tolerance.

Neonatal transplantation tolerance was one of the first experimental systems to reveal that tolerance could be achieved to non-self antigens in living animals. Functional and direct evidence (obtained by the use of monoclonal antibodies directed at T cell receptors specifically reactive with I-E molecules) confirm that tolerance is achieved, at least in part, via clonal elimination of developing thymocytes. In this report, we show that induction of tolerance of class I alloantigens in neonatal mice is governed by expression of I-E molecules. Neonatal I-E non-expressor mice proved to be highly resistant to the acquisition of class I tolerance if the donor inoculum expressed disparate class I antigens as well as I-E molecules. The spleens of the few class I-tolerant, I-E non-expressor mice that were generated were found to be depleted of I-E-reactive (RR315+) T cells, whereas no such depletion was observed in their neonatally injected, but non-tolerant littermates. By contrast we found no resistance to tolerance of I-A alloantigens when neonatal I-E non-expressor mice received injections of I-A-disparate, I-E-bearing donor cells. In these tolerant mice, splenic I-E-reactive T cells were readily detected in apparently normal amounts. These results indicate that lack of I-E expression in newborn mice confers resistance to tolerance induction to class I alloantigens, especially when the latter are expressed on donor cells that also display I-E molecules. The possible mechanisms operating to produce resistance to tolerance induction in neonatal mice are discussed, including the possibilities that (a) I-E may act as a restricting element during tolerance induction (an ontogenic process), and (b) the expression of I-E on H-2-disparate, I-E-expressing test skin allografts may provide a source of "help" for CD8+ cytotoxic T cell precursors, leading to graft rejection.

Ontogeny of cytotoxic T-cell repertoire modification.

The specificity of the residual anti-B10 cytotoxic T-cell response of B10. A mice rendered neonatally tolerant of B10 was compared to the anti-B10 response of adult and neonatal normal B10. A mice. The response of both spleen cells and thymus cells from adult and neonatal normal mice was biased toward Kb. This was in contrast to the response from tolerant spleen which was biased toward Db. The results suggest that the repertoire of normal mice is established neonatally and does not change radically without specific antigenic challenge. Furthermore the fact that the residual tolerogen specific cytotoxic T-cell precursors (pTc) in tolerant mice have a different repertoire to normal neonates makes it unlikely that they are remnants of a neonatal repertoire that developed prior to the full establishment of tolerance. Taking into account the present and previous results, the residual tolerogen-specific Tc in tolerant mice most likely represent a population of cells that has escaped tolerance induction due to their low avidity for antigen and provide the first evidence that avidity plays a role in tolerance among cytotoxic T cells.

Modification of the cytotoxic T cell repertoire in neonatal tolerance. Evidence for preferential survival of cells with low avidity for tolerogen.

Clonal deletion of developing lymphocytes with potential reactivity for self is thought to play a crucial role in the establishment of self tolerance. One prediction of the clonal deletion hypothesis is that cells bearing receptors with high affinity for self are more likely than cells with low affinity receptors to be deleted from the repertoire. Experimental models of B cell tolerance have provided evidence for the preferential survival of low affinity cells with specificity for tolerogen in tolerant animals, but no comparable evidence exists for T cells. To examine this issue in T cells, cytotoxic T cell lines specific for the Kb mutant class I H-2 molecule, bm1, were generated from C57BL/6 mice rendered neonatally tolerant of bm1 and compared with anti-bm1 lines generated from normal mice. Compared with normal lines, those from tolerant mice differed in five ways: 1) they grew more slowly; 2) they were less efficient at lysing bm1 targets; 3) they showed different patterns of lysis against a panel of third party targets; 4) their cytotoxic activity against bm1 could be increased in the presence of leukoagglutinin, whereas the activity of normal lines was not increased by leukoagglutinin; and 5) their cytotoxic activity was more susceptible to inhibition by anti-Lyt-2 antibody. Taken together, these results demonstrate that the repertoire of the remaining tolerogen-specific cytotoxic T cells in neonatally tolerant mice is different from the normal C57BL/6 anti-bm1 repertoire, and the results are consistent with the idea that the remaining tolerogen-specific cells are low avidity cells that have preferentially escaped the clonal deletion process.