1) To elucidate the mechanisms by which a group of cytokines utilizing the common cytokine receptor g chain (gc) promote the differentiation and homeostasis of lymphoid cells ;
2) To dissect the potential roles played by distinct lymphoid subsets during the innate response to infectious pathogens ;
3) To study the development of Natural Killer (NK) cells using a novel alymphoid mouse model and to assess the potential of this model for the transplantation of human xenografts.
1) Cytokines have been assigned pleiotropic and redundant functions based on their ability to promote a variety of biological responses in lymphocytes in vitro. Still, our understanding of the essential roles played these different soluble factors in the differentiation of lymphocytes in vivo is only partly elucidated. Several cytokines binding to receptor complexes sharing the common cytokine receptor g chain (gc, a functional component of the receptors for interleukins (IL)-2, -4, -7, -9 and -15) have been shown to promote lymphoid proliferation in vitro and early lymphoid development in vivo. However, despite their well-established in vitro activities (IL-2, for example, was first identified as "T cell growth factor"), the relative importance for gc cytokines in naive and memory T cell homeostasis and for antigen-induced proliferation in vivo has not been clearly defined. In addition, in vivo evidence of gc cytokine redundancy has not been provided. Understanding when and how cytokines act on lymphocytes during immune responses is important for designing rational therapeutic treatment protocols to modify these responses. The ability to modulate immunity could have beneficial outcomes in the context of human diseases, including immune deficiency and autoimmunity.
2) Lymphocyte subpopulations have been identified based on their expression of different types of antigen receptors (B cells, ab T cells, gd T cells) as well as differentiation markers (NK and NK-T cells). These diverse lymphoid subsets have distinct roles during immune responses, and the co-operation between different subsets is required for establishing long-lasting immunity. While the cellular participants in adaptive immunity have been well characterized, less is known about the lymphoid cells which control innate immune responses. Since innate priming may determine the quality of adaptive responses, knowledge of the lymphocyte subsets involved in innate immunity could have an impact on diverse experimental therapies including vaccine development against infectious agents and tumors.
3) In vivo models provide the most biologically relevant conditions to study physiological processes. However, studies involving human tissues are limited due to the fact that xenogenic tissues are rapidly rejected following engraftment in mice. The development of NK-deficient mouse strains provides a) the means to study the genetic program which controls the development of NK cells, and b) new mouse models for establishing human xenografts. The ability to graft human tissues in mice or to develop antigen-specific B cell responses in mice using human lymphocytes has broad applications for the study of human disease processes, the development of novel gene transfer therapies and the generation of human monoclonal antibodies for adoptive immunotherapy.
1) Gene targeting in mice as well as the study of rare human genetic disorders have demonstrated that signaling through cytokine receptors sharing the common cytokine receptor g chain (gc) are important for normal lymphoid development. In humans, defects in gc cause a severe combined immunodeficiency disease (SCID), characterized by an absence of mature T and NK cells, while circulating B cells are detected. Mice made deficient in gc also lack gd T, NK-T and NK cells, but in contrast to gc- patients have peripheral ab T cells and a reduction in B cells. Our previous and current research efforts focused on defining the mechanisms of abnormal lymphopoiesis in gc- mice. We have demonstrated that a) the absence of gd T cells in gc- mice results from a survival defect; b) ab T cell development is " rescued " in gc- mice by an independent signaling pathway involving the pre-TCR; c) gc signals act in concert with growth factors and the pre-TCR to promote thymocyte development and d) the mature ab T cells that develop in gc- mice provoke autoimmune symptoms including secondary hematopoiesis, colitis and B cell loss. Together these results confirm the multiple roles played by gc cytokines for development and function of the lymphoid system. Because the lymphoid subsets involved in innate immunity (gd T, NK-T, NK cells) appeared more gc-dependent than those involved in adaptive immunity (ab T, B cells), we hypothesized that the homeostatic mechanisms which maintain these two " arms " of the immune system may be different. Projects to address the role of gc-dependent cytokines in peripheral lymphoid homeostasis using TCR transgenic mice and inducible gene targeting are described in this proposal.
Other important observations made with gc- mice and in mice made deficient in individual gc-dependent cytokines (e.g. : IL2, IL-4, IL-7-deficient mice) revised our understanding of cytokine redundancy. When immunological phenotypes were compared between these different mice, it became clear that individual gc-dependent cytokines played unique roles in lymphoid differentiation and that the absence of one gc-dependent cytokine could not be compensated by another. Thus the notion of cytokine redundancy in vivo does not appear to apply to cytokines using the gc chain.
2) Infection models in mice have provided an important means to dissect the cellular components involved in immune responses to pathogens. Infection with Listeria monocytogenes has been used as a prototypical model for responses to intracellular bacteria, which provoke a T-helper 1-dominated response. Studies in immunodeficient mice have shown that innate immunity to Listeria can proceed in the absence of T and B cells. A model has emerged where an IL-12/IFN-g positive feedback loop results in the activation of macrophage effectors by NK cells. Our own studies in gc- mice suggested that NK-independent pathways could prime innate effectors. The concept of T-cell dependent priming of innate immunity could have advantages for the host by protecting against unrelated infections. However, this alternative pathway of innate priming had not been demonstrated in normal animals. Previous infection studies using gene ablated animals have been useful in identifying those lymphoid components which were required for immune responses to pathogens, but did not report on the ability of individual subsets to respond or provide immunity. Using a new alymphoid mouse model (described below) for selective lymphoid reconstitution, the roles of individual lymphoid subsets in innate immunity can be deciphered. This approach, coupled with the analysis of cytokine roles in peripheral lymphoid homeostasis, may bring insights into the role played by unique lymphoid subsets in innate immunity and how cytokines may modulate them.
3) Few satisfactory mouse models exist for studies involving the transplantation of xenografts. Different drawbacks include innate mechanisms which result in the rejection of tissues, and shortened lifespans of the host due to spontaneous tumor development or susceptibility to infection. One mechanism involved in the rejection of xenografts involves natural killer (NK) cells. Because gc-deficient mice manifest an absolute deficiency in this lymphoid subset, they provide a means a) to examine the genetic pathways involved in the differentiation of NK cells and in the control of their effector functions and b) to establish and test NK-deficient mouse models for human tissue transplantation. Along these lines, we have established a novel alymphoid mouse strain (lacking all mature T, B and NK cells) which combines mutations in the Recombinase Activating Gene-2 and gc (RAG2/gc mice). RAG2/gc mice are viable, fertile and easily maintained under pathogen free conditions. We have previously shown that RAG2/gc mice can accept grafts of human peripheral blood cells and tumors. In our ongoing projects presented in this proposal, we demonstrate that RAG2/gc mice are also excellent hosts for studying NK cell development, the physiology of NK cells and for human solid tissue transplantation.
1. Roles for gc-dependent cytokines in the peripheral lymphoid homeostasis. (J. Di Santo)
Our previous studies demonstrated that the pre-TCR provided an essential gc-independent signal for ab T cell development. This observation, however, did not explain how ab T cells in the absence of gc could further develop, exit the thymus and survive in the periphery. During intrathymic selection, auto-reactive TCRs are deleted to maintain self-tolerance. However, the presence of activated, dysregulated T cells in the periphery of gc- mice suggested that the process of negative selection might be defective. Alternatively, the process of T cell activation in gc- mice might simply reflect antigen stimulation in the periphery.
We have begun to address these questions using TCR transgenic mice. Our previous results showed that positive and negative selection of CD8+ cells TCR transgenic thymocytes was normal in gc- mice. In contrast, no CD8+ T cells were detected in the peripheral lymphoid organs of these mice suggesting that CD8+ ab T cells also have a strict requirement for gc cytokines for their peripheral survival. However, since the majority of peripheral ab T cells in gc- mice are CD4+, the selection processes which operate for these cells may be different. TCR transgenic mice which select CD4+ ab T cells specific for male antigen (CD4-HY) have been recently generated. Preliminary results show that intrathymic selection of this TCR is also not affected in the absence of gc. Interestingly, the peripheral survival of CD4+ ab T cells in female gc- transgenic mice is markedly reduced. This contrasts with the situation found in non-transgenic gc- mice where a CD4+ ab T cell population persists and expands with age. Since female mice do not harbor male antigen, this suggests that the majority of peripheral T cells in gc- mice result from antigen stimulation. Consistent with this hypothesis, we observed that antigen-induced proliferation of gc- male-reactive cells was normal. This surprising result suggests that T cell proliferation is not driven by gc cytokines and call into question the role for IL-2, -4, -7, -9 and 15 in vivo. To identify the gc-dependent cytokines required for naive T cells survival, transfer studies will be made into mice deficient in either IL-2, -4, -7 or 15. These mutant mice have been obtained in order to generate a series of RAG2-deficient cytokine-deficient hosts, which will prove useful in further studies.
Another strategy to identify roles for gc-dependent cytokines makes use of an inducible gene targeting strategy to conditionally delete the gc locus in vivo. Mice have been generated carrying a " floxed " gc locus which can be deleted in a temporal fashion using the Cre recombinase under the MxCre promoter. In this way, gc-deficiency can be selective induced under defined conditions in the adult animal and during the course of an immune response.
We have begun to identify the role of different signaling pathways emanating from gc-containing receptors for their biological responses in vivo. In a collaboration with Dr. M. Goldsmith (Gladstone Institute, San Francisco), a novel gene bypass strategy was demonstrated in gc- and JAK3 mutant mice which are in deficient in the gc signaling pathway. A gc/erythropoietin receptor chimera which selectively recruited another signaling pathway (JAK2) could restore lymphoid function in these two mouse strains. This study demonstrated that JAK kinases appear non-specific in their roles and provide a new possibility for correction of genetic diseases characterized by defects in critical metabolic pathways. In collaboration with Pr. T. Taniguchi (Tokyo, Japan) we have studied the regions of the gc intracytoplasmic domain which are essential for gc signaling in vivo. Transgenic mice expressing mutant gc molecules have been generated and crossed to gc- mice. These projects will attempt to decipher the pathways implicated in the gc signaling.
2. Elucidating the genetic program which controls NK cell differentiation. (F. Colucci, C. Roth)
NK cells participate in both innate and adaptive immunity by their prompt secretion of cytokines including IFN-g which activate macrophages and by their ability to lyse virally infected cells and tumor cells without prior sensitization. Although these characteristics of NK cells are well documented, little is known about the genetic program which orchestrates NK development or about the signaling pathways which trigger NK effector functions.
RAG2/gc mice should prove useful in dissecting murine NK development and function by hematopoietic complementation. In order to validate our experimental approach, we have generated lymphoid chimeras using RAG2/gc mice by injection of either syngeneic bone marrow or fetal liver hematopoietic progenitors from normal or mutant mice. We could demonstrate long-term reconstitution of normally functioning T, B and NK cells in vivo (antigen-specific Ig responses, tumor rejection) and in vitro (proliferation, and NK cytotoxicity). Studies using fetal liver cells and/or ES cells mutated for either 1) proto-oncgenes (vav-1); 2) cytokine receptors or their signaling cascades (c-kit); 3) intracellular signaling molecules (syk); or 4) transcription factors (PU.1, myb) have been performed. By exploiting this unique system of NK complementation in RAG2/gc mice, we hope to gain a more complete understanding of the molecular signals required for NK differentiation in vivo. In addition, a system of in vitro NK cell differentiation from hematopoietic precursors has been established, which generates cells expressing inhibitory Ly49 receptors. These experiments demonstrate that Ly49 acquisition is a sequential process dependent on interactions with MHC molecules. Future studies using RAG2/gc mice expressing different MHC alleles will allow us to determine whether this process proceeds in a similar fashion in vivo.
In collaboration with Dr. B.A. Croy (Ontario, CA), RAG2/gc mice have been used as recipients to study the role of uterine NK cells. The function of these pregnancy-associated NK cells remains ill-defined, and they have been hypothesized to play an immunological role at fetal-maternal interface. We have observed that RAG2/gc mice have normal reproductive capacity (normal litter size) thereby ruling out an essential role for uterine NK cells in fetal development. In contrast, normal placental development and especially the development of the uterine vasculature, appears to require IFN-g secretion from uterine NK cells. These observations suggest that defects in uterine NK cell function may be involved in syndromes involving abnormal placentation or in pre-eclampsia.
3. Defining the minimum cellular components for immune responses against infections. (J. Di Santo)
Previous experiments using gene targeted mice lacking one lymphocyte subset have helped to identify the cells that are essential for resistance against pathogenic organisms. However, this approach cannot identify the minimal cellular requirements for an effective immune response. RAG2/gc mice may be suitable for reconstructing " custom-made " immune systems. Using this idea, we are developing an approach to dissect innate immunity by reconstituting RAG2/gc mice with selected lymphoid subsets. Subsequently, these mice will be challenged with Listeria monocytogenes, used as the model pathogen.
The initial experiments will be performed after selective reconstitution of RAG2/gc mice. This will be accomplished by transferring into RAG2/gc mice single purified lymphoid subsets (CD4+ or CD8+ ab T cells, gd T cells, NK-T cells, NK cells) or combinations of two or more subsets. Mice will subsequently be infected with Listeria (or in parallel studies, Shigella). The development of the infection will be monitored by animal survival, production of protective cytokines (IFN-g), quantification of bacterial load, and histopathology. Further reconstitution experiments using cytokine- or cytokine receptor-deficient cells will help to identify the necessary soluble interactions required for innate immunity in this system. Studies using a murine model of dysentery (Shigella) are being performed in collaboration with Dr. P. Sansonetti.
4. A novel mouse model for human tissue transplants (J. Di Santo)
RAG2/gc mice have been evaluated for their ability to accept human tumor grafts and normal human peripheral blood leukocytes (PBL). In comparison to the NOD/SCID strain (the most widely used model), RAG2/gc mice could be grafted to the same extent by human PBL and were clearly superior when tumors were utilized. Moreover, RAG2/gc mice could be grafted without prior conditioning (ie. irradiation) and no spontaneous tumor formation was observed (which is a common drawback to the NOD/SCID mouse model). Using RAG2/gc mice as a starting point, a series of novel immunodeficient mouse strains are being generated with additional defects in natural defense pathways. For example, we have already incorporated a mutation affecting the complement C5 component onto the RAG2/gc background (RAG2/gc/C5 mice). In a similar fashion, we have created RAG2/gc/NOD mice. These derivative RAG2/gc mouse strains may provide additional grafting potential for certain types of human tissues.
We are currently examining the utility of RAG2/gc mice and their derivatives as potential hosts for transplantation of normal and transformed human tissues. In collaboration with Dr. G. Butler-Browne (URA 2115, Paris) and Dr. T. Partridge (Hammersmith Hospital, London) we recently found that human skeletal muscle can be tranplanted into the muscles of RAG2/gc mice and persist for up to 3 months, in contrast to previous studies using other immunodeficient mouse strains where rejection was consistently observed. These new approach should provide the possibility to study human myodystrophies in vivo and to test new gene transfer therapies for these diseases.
We believe that the projects presented herein will enhance our knowledge of the mechanisms of lymphoid development, of innate and adaptive immune responses and will furnish new animal models to study human disease processes. These projects are fundamental in nature, but have important implications for clinical medicine.