TCR transgenic CD8+ T cells activated in the presence of TGFβ express FoxP3 and mediate linked suppression of primary immune responses and cardiac allograft rejection

Judith A. Kapp 0 1 Kazuhito Honjo 0 Linda M. Kapp 1 Xiao yan Xu 0 Alana Cozier 0 R. Pat Bucy 0 0 Department of Pathology , Spain Wallace Building, 619 South 19th Street , University of Alabama at Birmingham , Birmingham, AL 35233-7331 , USA 1 Department of Ophthalmology Although CD41CD251FoxP31 regulatory T cells play a role in allograft tolerance, the role of CD81 cells with immunosuppressive function is less clear. To address this issue, spleen cells from Rag-1deficient TCR transgenic (Tg) mice expressing a receptor for ovalbumin (OVA) in the context of MHC class I (OT1) were activated with OVA expressing antigen-presenting cell (APC) in the presence or absence of exogenous transforming growth factor b (TGFb). TGFb inhibited the expression of IFN-c, granzyme B and the lytic activity of the OT1 T cells while inducing FoxP3 expression in 5-15% of the cells. By contrast, FoxP3 expression was not detected in naive OT-1 T cells or OT-1 T cells activated without exogenous TGFb. TGFb-activated OT1 cells inhibited the activation of Kd-specific CD81 CTL responses by normal B6 T cells and the proliferation by Kd-specific CD41 TCR Tg T cells, but only if the OVA epitope was co-expressed by Kd1 APC. This antigen-specific inhibitory activity, referred to as linked suppression, was neither mediated by residual lytic activity within the activated OT1 T cells nor did it depend upon IL-10 or TGFb. Suppression correlated with inhibition of CD86 expression on CD11c1 APC. TGFb-activated OT1 T cells also delayed the rejection of heterotopic, vascularized cardiac allografts mediated by anti-Kd-specific CD41 TCR Tg T cells, but only if the cardiac allograft expressed both OVA and Kd as transgenes. Prolonged survival of allografts was associated with rapid migration of the FoxP31 OT1 T cells into the donor heart raising the possibility that suppression may be mediated within the allograft. These data show that TGFb-activated CD81 T cells mediate antigenspecific, APC-focused patterns of suppression in vitro and in vivo. Introduction Over 50 years ago, Medawar and colleagues (1) demonstrated that induction of antigen-specific tolerance in rodents could prolong allograft survival in the absence of pharmacological agents. Although multiple different mechanisms that could contribute to tolerance have been elucidated, the contribution of active suppression mediated by antigenspecific T cells to this process has had variable popularity. The seminal experiments of Gershon and Kondo (2) showing that antigen-specific tolerance could be adoptively transferred with lymphocytes from tolerant to naive mice, a phenomenon they labeled as infectious tolerance, raised the possibility that tolerance could be maintained by an active process. CD8+ T lymphocytes were shown to mediate antigen-specific tolerance and were referred to as suppressor T cells (Ts) (3, 4). Although the ability of antigen-specific CD8+ T cells to adoptively transfer non-responsiveness was confirmed by numerous investigators in a variety of systems including transplant rejection [reviewed in (5)], the mechanisms involved in suppression were not identified and interest in this pathway waned. The concept that T cells can mediate immunological tolerance through active suppression has again become widely accepted on the basis of the observations that a distinct subset of naturally occurring CD4+CD25+ T cells (6), referred to as regulatory T cells (Tregs), has the capacity to prevent autoimmune disease mediated by endogenous, self-reactive T cells (7, 8). The presence of CD25 on these resting T cells allowed them to be physically isolated and shown to have specific immunosuppressive activity. Molecular analysis of Tregs identified the unique expression of a transcription factor from the forkhead/winged helix family (FoxP3) (9, 10), which functions as a master switch driving differentiation of naive T cells into the Treg lineage (1015). Interest in the role of CD8+ Ts was renewed by the observation that delivery of antigens into the anterior chamber of the eye induced the activation of CD8+ T cells that inhibit responses of primed effector Tcells (16) and that the activation of the CD8+ Ts depended upon transforming growth factor b (TGFb), a normal constituent of the aqueous humor (17). Antigen-specific CD8+CD28 Ts have also been identified in the blood of rejection-free cardiac transplant recipients (18, 19). Moreover, FoxP3 is up-regulated in CD8+CD28 human (20) and rat T cells (21) that have suppressive activity. Our long-term goal is to determining the mechanisms by which CD8+ Ts can prolong allograft survival. However, the regulation of alloantigen-specific rejection is difficult to dissect because of the complexity of natural alloantigens, the large number of T cells expressing heterogeneous receptors that can recognize even limited alloantigenic differences between the host and the donor animals and multiple pathways of antigen presentation. Consequently, we have used a reductionist approach to develop a model in which the key elements of this complex in vivo response can be isolated. To generate CD8+ Ts, spleen cells from the ovalbumin (OVA)-specific Rag1 / TCR transgenic (Tg) mice expressing a receptor for OVA in the context of MHC class I (OT1) (22) were incubated with spleen cells from mice expressing a chicken OVA transgene (23) with or without exogenous TGFb. To restrict the specific antigenic epitopes that drive the T cell response, we have generated TCR Tg mice (TCR75) with specificity for a single peptide derived from the H-2Kd MHC class I molecule presented by the I-Ab class II molecule (K5d4 68/I-Ab) on the C57BL/6 (B6) background (24). Antigen-presenting cells (APCs) and hearts were obtained from Tg B6 mice, expressing the genomic sequence of H-2Kd (25), the chicken OVA gene or both genes. Here, we report that non-lytic CD8+ OT1 T cells activated in the presence of TGFb expressed FoxP3 and exhibited linked suppression in vitro and in vivo. Experimental animals B6 (H-2b) mice were purchased from the National Cancer Institute (Frederick Cancer Research and Development Center, Frederick, MD, USA). C57BL/6-Tg(TcraTcrb)1100Mjb (22) mice, also referred to as OT1, which recognize OVA257264 in the context of H2Kb were a generous gift of Michael Bevan (University of Washington, Seattle, WA, USA) and were bred and maintained in the animal facilities at the University of Alabama at Birmingham (UAB). The C57BL/6Tg(TcraTcrb)TCR75Rpb (TCR75) mice, which express TCRab specific for I-Ab/H-2K5d4 68 epitope, have been previously characterized (24). The OT1 and TCR75 Tg strains were crossed onto the B6.Rag1 / background to exclude expression of other TCR specificities. B6.Rag1 / .OT1 mice were also crossed to perforin, or pore-forming protein (pfp), knockout B6 Prf1tm1Sdz (26) obtained from Jackson Laboratory (Bar Harbor, ME, USA), to produce B6.Rag1 / .OT1.pfp / mice (referred to as OT1.pfp / ). B6.Rag1 / .OT1 mice were also crossed to (...truncated)