Addition of an IL-15 mutant/FCγ2A antagonist protein protects islet allografts from rejection overriding costimulation blockade

IL-15, a powerful chemoattractant and growth factor for antigen-activated T cells and NK cells, contributes to lymphocyte homeostasis and selectively induces the activation and proliferation of memory CD8+ T cells.1,2 Analysis of the phenotype of IL-15−/− mice3 and IL-15Rα−/− mice4 demonstrate that IL-15 is a critical factor for innate and adaptive immune responses. IL-15 is produced by the monocyte/macrophage lineage, by epithelial and endothelial cells but not by T-cells. The IL-15 receptor (IL-15R) expressed upon T cells is composed of three chains. Two are constitutively expressed and shared with the IL-2 receptor (IL-2R), the IL-2Rβ and IL-2Rγc chains, and a unique IL-15Rα chain is expressed upon activation of, but not during resting of mononuclear leukocytes.5,6 Because IL-15 and IL-2 share two receptor signaling components (IL-2/IL-15Rβγc), they display similar biological properties in vitro. Nevertheless, there are differential patterns of IL-15/IL-15Rα expression and in vivo studies have revealed major nonredundant roles for these two T-cell growth factors (TCGFs). In organ transplantation, T cells play a critical role in initiating the immune response leading ultimately to the effector mechanisms mediating allograft rejection. An analysis of biopsies taken from rejecting human renal allografts has revealed that intragraft gene expression of IL-15 is a far more consistent marker of rejection than IL-2 gene expression.7 Moreover, IL-2 expression is not a prerequisite for allograft rejection as IL-2 knockout (KO) mice promptly reject islet and cardiac allografts.7,8 Immunohistology in the IL-2 KO recipients reveals infiltration of rejecting grafts by CD4 and CD8 T cells as well as robust intragraft expression of the IL-4, IL-7, and IL-15 as well as granzyme B genes.8 Since costimulation blockade drastically decreases post-transplant expression of IL-2,9 the clonal proliferation of alloreactive T cells in hosts treated with costimulation blockade may be explained by the action of a non-T-cell-derived cytokine, such as IL-15. Although costimulatory blockade of CD28/CTLA4/B7 or CD40/CD40L interactions reduces immune activation and provides a powerful inhibition of alloimmune responses in both rodents and primates,9–11 in some circumstances it failed to induce permanent engraftment. Several reports suggest that CD8+ T cells are responsible for costimulation blockade-resistant rejection in various models, such as intestinal and skin allograft models.12,13 For example, adjunctive treatment with anti-CD8 mAb or other agents to deplete CD8+ T cells, are required to induce tolerance in these models.14 The resistance of CD8+ T cells to costimulation blockade may be explained by a poor effect of CTLA4Ig and αCD154 Abs to regulate activation and proliferation of CD8+ compared to CD4+ T cells.14,15 Recently, Smith et al16 have reported the prolonged survival of heart allografts in a treatment protocol consisting of soluble IL-15 receptor- α chain proteins and nondepleting anti-CD4 mAb. Thus the expression of IL-15 may be linked to CD8-dependent costimulation-resistant rejection. As costimulation blockade resistant rejection is mediated by CD8+ T cells and as IL-15 is essential for the activation and proliferation of CD8+ T cells, we hypothesized that targeting the IL-15/IL-15R system in conjunction with costimulation blockade would provide a new perspective for inducing allograft tolerance. The development of agents targeting the receptor and signaling elements of IL-15 may provide new perspectives for treatment of diseases associated with the IL-15/IL-15R pathway. The idea of creating an IL-15 antagonist was based on the homology of glutamine residues within the C-terminus of the four helix structures shared by IL-2 and IL-15, and on the report that a C-terminal glutamine in IL-2 is crucial for IL-2 binding to IL-2Rγ chain.17 In our laboratory a novel IL-15 mutant/Fcγ2a protein was genetically constructed, linked to murine Fcγ2a and expressed in mammalian cells. The IL-15 mutant was constructed by replacing the codons for the C-terminal glutamine amino acid residues with codons for aspartic acid (ie, Q101 and Q108).18 The mutant IL-15 moiety is a high-affinity IL-15R site-specific antagonist. The Fc portion of this fusion protein confers longevity and is designed to support cytocidal activities against IL-15R α+ cells insofar as the Fc-related sequences of mouse IgG2a isotype that support complement activation and activation of FcR+ phagocytes are present.18 Moreover, this IL-15 mutant/Fcγ2a protein competitively inhibits IL-15-triggered cell proliferation and does not activate the STAT-signaling pathway. Because the receptor site-specific antagonist IL-15 mutant/Fcγ2a protein had a prolonged half-life in vivo and the potential for destruction of IL-15R + leukocytes, the immunosuppressive effects of this agent were examined in mice with induced delayed-type hypersensitivity (DTH) and collagen-induced arthritis (CIA), a murine model for RA. In the DTH model, treatment with IL-15 mutant/Fcγ2a protein attenuated DTH responses and mononuclear leukocyte infiltration within the DTH sites.18 In the CIA model, treatment with IL-15 mutant/Fcγ2a protein markedly decreased the incidence and severity of arthritis in mice (S. Ferrari-Lacraz, inpreparation). As this unique IL-15 mutant/Fcγ2a protein displays promising effects on T-cell-derived pathogenesis, we wanted to determine its role and relative merits as a new therapeutic strategy in transplantation.