T cells have the capacity to eliminate tumors but the signaling pathways by which they do so are incompletely understood. T cell priming requires activation of the transcription factors AP-1, NFAT and NF-κB downstream of the TCR, but whether activation of T cell-NF-κB in vivo is required for tumor control has not been addressed. In humans and mice with progressively growing tumors, the activity of T cell-intrinsic NF-κB is often reduced. However, it is not clear if this is causal for an inability to reject transformed cells, or if it is a consequence of tumor growth. T cell-NF-κB is important for T cell survival and effector differentiation and plays an important role in enabling T cells to reject cardiac and islet allografts, suggesting the possibility that it may also be required for tumor elimination. In this study, we tested whether normal T cell-NF-κB activation is necessary for the rejection of tumors whose growth is normally controlled by the immune system.

The analysis of gene function in mature and activated natural killer cells has been hampered by the lack of model systems for Cre-mediated recombination in these cells. Here we have investigated the utility of GzmbCre for recombination of loxp sequences in these cells predicated on the observation that Gzmb mRNA is highly expressed in mature and activated natural killer cells. Using two different reporter strains we determined that gene function could be investigated in mature natural killer cells after GzmbCre mediated recombination in vitro in conditions that lead to natural killer cell activation such as in the cytokine combination of interleukin 2 and interleukin 12. We demonstrated the utility of this model by creating GzmbCre;Rosa26IKKbca mice in which Cre-mediated recombination resulted in expression of constitutively active IKKβ, which results in activation of the NFκB transcription factor. In vivo and in vitro activation of IKKβ in natural killer cells revealed that constitutive activation of this pathway leads to natural killer cell hyper-activation and altered morphology. As a caveat to the use of GzmbCre we found that this transgene can lead to recombination in all hematopoietic cells the extent of which varies with the particular loxp flanked allele under investigation. We conclude that GzmbCre can be used under some conditions to investigate gene function in mature and activated natural killer cells.

Despite numerous advances in the definition of a role for regulatory T cells (Tregs) in facilitating experimental transplantation tolerance, and ongoing clinical trials for Treg-based therapies, critical issues related to the optimum dosage, antigen-specificity, and Treg-friendly adjunct immunosuppressants remain incompletely resolved. In this study, we used a tractable approach of MHC tetramers and flow cytometry to define the fate of conventional (Tconvs) and Tregs CD4+ T cells that recognize donor 2W antigens presented by I-Ab on donor and recipient antigen-presenting cells (APCs) in a mouse cardiac allograft transplant model. Our study shows that these endogenous, donor-reactive Tregs comparably accumulate in the spleens of recipients undergoing acute rejection or exhibiting costimulation blockade-induced tolerance. Importantly, this expansion was not detected when analyzing bulk splenic Tregs. Systemically, the distinguishing feature between tolerance and rejection was the inhibition of donor-reactive conventional T cell (Tconv) expansion in tolerance, translating into increased percentages of splenic FoxP3+ Tregs within the 2W:I-Ab CD4+ T cell subset compared to rejection (~35 vs. <5% in tolerance vs. rejection). We further observed that continuous administration of rapamycin, cyclosporine A, or CTLA4-Ig did not facilitate donor-specific Treg expansion, while all three drugs inhibited Tconv expansion. Finally, donor-specific Tregs accumulated comparably in rejecting tolerant allografts, whereas tolerant grafts harbored <10% of the donor-specific Tconv numbers observed in rejecting allografts. Thus, ~80% of 2W:I-Ab CD4+ T cells in tolerant allografts expressed FoxP3+ compared to ≤10% in rejecting allografts. A similar, albeit lesser, enrichment was observed with bulk graft-infiltrating CD4+ cells, where ~30% were FoxP3+ in tolerant allografts, compared to ≤10% in rejecting allografts. Finally, we assessed that the phenotype of 2W:I-Ab Tregs and observed that the percentages of cells expressing neuropilin-1 and CD73 were significantly higher in tolerance compared to rejection, suggesting that these Tregs may be functionally distinct. Collectively, the analysis of donor-reactive, but not of bulk, Tconvs and Tregs reveal a systemic signature of tolerance that is stable and congruent with the signature within tolerant allografts. Our data also underscore the importance of limiting Tconv expansion for high donor-specific Tregs:Tconv ratios to be successfully attained in transplantation tolerance.

Clinical observations that kidney transplant recipients receiving belatacept who experienced T cell-mediated acute rejection can be successfully treated and subsequently maintained on belatacept-based immunosuppression suggest that belatacept is able to control memory T cells. We recently reported that treatment with CTLA4-Ig from day 6 posttransplantation successfully rescues allografts from acute rejection in a BALB/c to C57BL/6 heart transplant model, in part, by abolishing B cell germinal centers and reducing alloantibody titers. Here, we show that CTLA4-Ig is additionally able to inhibit established T cell responses independently of B cells. CTLA4-Ig inhibited the in vivo cytolytic activity of donor-specific CD8+ T cells, and the production of IFNγ by graft-infiltrating T cells. Delayed CTLA4-Ig treatment did not reduce the numbers of graft-infiltrating T cells nor prevented the accumulation of antigen-experienced donor-specific memory T cells in the spleen. Nevertheless, delayed CTLA4-Ig treatment successfully maintained long-term graft acceptance in the majority of recipients that had experienced a rejection crisis, and enabled the acceptance of secondary BALB/c heart grafts transplanted 30 days after the first transplantation. In summary, we conclude that delayed CTLA4-Ig treatment is able to partially halt ongoing T cell-mediated acute rejection. These findings extend the functional efficacy of CTLA4-Ig therapy to effector T cells and provide an explanation for why CTLA4-Ig-based immunosuppression in the clinic successfully maintains long-term graft survival after T cell-mediated rejection.

Dominant infectious tolerance explains how brief tolerance-inducing therapies result in lifelong tolerance to donor antigens and "linked" third-party antigens, while recipient sensitization and ensuing immunological memory prevent the successful induction of transplant tolerance. In this study, we juxtapose these 2 concepts to test whether mechanisms of dominant infectious tolerance can control a limited repertoire of memory T and B cells. We show that sensitization to a single donor antigen is sufficient to prevent stable transplant tolerance, rendering it unstable. Mechanistic studies revealed that recall antibody responses and memory CD8+ T cell expansion were initially controlled, but memory CD4+Foxp3- T cell (Tconv) responses were not. Remarkably, naive donor-specific Tconvs at tolerance induction also acquired a resistance to tolerance, proliferating and acquiring a phenotype similar to memory Tconvs. This phenomenon of "linked sensitization" underscores the challenges of reprogramming a primed immune response toward tolerance and identifies a potential therapeutic checkpoint for synergizing with costimulation blockade to achieve transplant tolerance in the clinic.