We have developed a model for studying tolerance to persistent low dose antigen in vivo, which results in the generation of a large number of anergic (hyporesponsive) T cells. We call this state adaptive tolerance. We inject CD4+, antigen-specific T cells from a T cell receptor transgenic mouse on a Rag2-/- background (a monospecific T cell population) into a second transgenic mouse expressing the antigen in a lymphopenic background (no other T cells). Within 24 hours after transfer, the T cells are all activated by the antigen (as evidenced by an increase in size and expression of CD69), and proliferate extensively for several days, increasing in number about 100-fold. This expansion is followed by a deletional phase during which 50% of the cells disappear. Finally, the population reaches a steady state level in which the cells appear to be refractory to restimulation in vivo and in vitro. In this adaptive tolerant state, cytokine responses to high doses of antigen in vitro are inhibited 90%. However, in vivo bromodeoxyuridine labeling shows a slow T cell turnover of about 5% of the cells per day, and B cell help is still sufficient for the mice to eventually develop a mild form of chronic autoimmune arthritis or polychondritis depending on the antigen and the T cell specificity. The hyporesponsive state is reversible if the cells are transferred again into a second host not expressing the antigen. If the retransfer is into a host expressing the antigen, the T cells remain hyporesponsive and slowly decrease their IL-2 and IFN gamma production by another 6-10 fold over 3-4 weeks. This deeper state of anergy suggests that the tolerance process is adaptable to different levels. Finally, if the initial antigen-bearing host is T cell replete, i.e., contains a normal polyclonal T cell population, then the naive T cell response is blunted. The expansion phase is curtailed by 5 to 10 fold, the differentiation into effector T cells (producing IFN-gamma) does not occur very much, and the anergic T cells disappear with a half life of 9-12 days. As a consequence, the mice do not develop any autoimmune disease. Recently, the lab has focused on determining what causes the difference in the antigen-specific naive CD4+ T cell response when the cells are transferred into T cell-sufficient versus T cell-deficient antigen-bearing hosts. The difference must somehow be related to the presence of other T cells in the environment. The critical cells are peripheral alpha/beta T cells because adult thymectomy does not alter the full-host effects and introducing the TCR alpha KO mutation eliminates them. To identify the inhibitory T cell, we reconstituted T cell-deficient mice with normal CD4+/Foxp3+, CD4+/Foxp3-, or CD8+ T cells purified by flow cytometry. One to 7 days later, naive antigen-specific CD4+ T cells were injected and their expansion and longevity monitored. Both expanded polyclonal CD4+ T cell populations partially facilitated the loss of the antigen-specific T cells after the latter became anergic. This decrease from peak cell numbers occurred over a 3 week time period and was about 20 fold, compared to only a 2 fold loss in a host reconstituted with CD8+ T cells. Since these experiments suggested that CD4+Foxp3- T cells could mediate at least one of the effects of a full host, we considered the possibility that the differences in outcome in the T cell-sufficient host were caused by T cell competition for nutrients, growth factors, or activation niches. To test this type of model we transferred the original antigen-specific T cells into a second TCR transgenic, Rag2-/- host specific for another antigen, but also expressing the first antigen. These hosts contain around 5 to 10 million naive T cells filling the space in lymph nodes and spleen. Despite this, the original naive antigen-specific T cells adoptively transferred into each such hosts expanded fully (100-fold), differentiated into IFN-gamma producers, entered the adaptive tolerant state, and persisted for months. This result was observed with 5 different CD4+ TCR transgenics. One of them was specific for an antigen presented by the same MHC molecule as that used by the first antigen. Thus, non-specific competition for growth and survival factors such as IL-7 would seem not to be responsible for the inhibitory effects of the polyclonal T cell population. The next step was to clone the inhibitory CD4+ T cell population in order to further determine its phenotype and characterize its mechanism of inhibition. Polyclonal CD4+ T cells were plated at 100 cells per well from a V beta3 transgenic mouse crossed to a TCR alpha-deficient mouse. The cells (expressing only one alpha and beta chain each) were then stimulated with phorbol ester, ionomycin, and anti-CD28 monoclonal antibody, and after expansion to several million progeny, tested for their ability to inhibit the survival of naive B10.A TCR-5C.C7 T cells transferred into an antigen-expressing lymphopenic host. A few subpopulations were found to be greatly enriched for deletional activity, and this inhibition was specific as the cells did not inhibit the survival of a different TCR transgenic population recognizing another antigen. RNA was isolated from the deletor and non-deletor populations and their pool of TCR alpha chain genes separately sequenced. In general there were between 10 and 20 unique alpha chain sequences per expanded colony. Each alpha-chain gene from a deletor pool was introduced into a V beta3 Rag2-/- bone marrow culture using retroviral infection. These populations were then used to create radiation-induced bone marrow chimeras expressing the unique alpha/beta TCR. After several months the chimeras were challenged with naive B10.A TCR-5C.C7 T cells and examined 40 days later for survival of the 5C.C7 T cells. Only one set of chimeras expressing a unique Valpha2 receptor showed strong deletor activity. Experiments in vitro showed that this T cell receptor is not specific for the cytochrome c antigen recognized by the original naive 5C.C7 T cells. In addition, mixing the deltor population with naive 5C.C7 T cells did not prevent the latter from responding to cytochrome c in vitro. We next explored the possibility that the deletor T cell recognizes self peptides. Retroviral infected-bone marrow chimeras were set up using either the Valpha2 deletor receptor or a Valpha10 control non-deletor receptor in a Rag2-/- host not expressing the cytochrome antigen. When the latter mouse was injected with B10.A TCR-5C.C7 T cells, the cells underwent the process of homeostatic proliferation, thought to be induced by recognition of self peptide/MHC complexes under lymphopenic conditions. In contrast, the presence of T cells expressing the Valpha2 receptor almost completely prevented this proliferation, but not that of A1M-specific T cells. In addition, the survival of a labelled cohort of B10.A TCR-5C.C7 T cells in an intact B10.A host was diminished by injection of a million Valpha2 receptor-bearing T cells. These results suggest that the Valpha2 receptor has specificity for certain self peptides and can compete with 5C.C7 T cells to prevent their expansion and/or survival. In 2012 we identified the self peptide as the GP peptide, encoded by an endogenous retrovirus that had previously been shown to play a role in the thymic positive selection of the cytochrome-specific T cell. Overall are results suggest that the CD4 T cell repertoire is functionally organized into small colonies or microniches, within which homeostatic competition operates. Such organiztion can help avoid excessive loss of bystander specificities during multiple immune responses, preserving the valuable diversity of the T cell repertoire.
|Chappert, Pascal; Bouladoux, Nicolas; Naik, Shruti et al. (2013) Specific gut commensal flora locally alters T cell tuning to endogenous ligands. Immunity 38:1198-210|