Autoreactive T cells that are capable of inducing disease exist in normal adult animals, but are maintained in a dormant or inactive state due to the suppressive functions of CD4+CD25+Foxp3+ regulatory T cells (Treg). Our recent studies have focused on defining the mechanism of action of the CD4+CD25+Foxp3+ Treg cells in vitro and on an analysis of their potential dysfunction in autoimmune disease: (1) Recent reports have demonstrated that peripheral CD4+Foxp3- T cells can be induced in vitro to express Foxp3 in the presence of TGF-beta. The demonstration that TGFβ is a potent inducer of Foxp3 expression and Treg function in vitro and in vivo raised the possibility that one potential mechanism of action of Treg is to induce Treg de novo from nave T cell precursors in a TGFβ-dependent manner thus facilitating a type of infectious tolerance. Previous studies have demonstrated that the TGFβ propeptide, latency associated peptide (LAP), and thus presumably TGFβ is expressed on the surface of Treg. Cell surface LAP was only expressed by activated, but not resting, Foxp3+ T cells and not by activated Foxp3- T cells. Since activated Treg express functional LAP/TGF-beta complexes on their cell surface, we determined whether Treg would be able to confer infectious tolerance by inducing Foxp3 expression in CD4+Foxp3- T cells. Co-culture of pre-activated Foxp3+ Treg, but not pre-activated CD4+Foxp3-, T cells with nave CD4+Foxp3- T cells resulted in the induction of Foxp3 expression in 10-30% of the nave T cells. The addition of LAP completely inhibited the induction of Foxp3, demonstrating that Treg-dependent induction of Foxp3 is TGF-beta-dependent. Pre-activated Treg from TGF-beta deficient (-/-) mice or from mice with T selective defect in processing TGF-beta (furin-/- mice) were incapable of inducing Foxp3 in nave responders, thus demonstrating that the Treg were the source of the TGF-beta. The CD4+Foxp3+ T cells converted from Foxp3- cells were suppressive in vitro and could inhibit the development of colitis when co-transferred with CD4+Foxp3- T cells into RAG-/- mice. Co-transfer into normal mice of pre-activated antigen-specific Treg with nave antigen-specific responder T cells also resulted in induction of Foxp3 in the responders after immunization of the recipients with antigen. Treg-mediated Foxp3 induction appears to be tightly regulated and dependent on T cell activation. We propose that that in vivo simultaneous stimulation of both the Treg and the antigen-specific T cells likely occurs on the platform of the DC in the form of a 3-cell interaction. As a single DC can present more than one organ-derived antigen, it is possible that the Treg-TGFβ pathway of Foxp3 induction can result in expansion of Treg of broad antigen specificities and thus serve as an important adjuvant in the use of Treg for cellular biotherapy. (2) We are analyzing the role of the transcription factor Helios (Ikzf2) in Treg function. Helios is a transcription factor belonging to the zinc finger containing Ikaros family comprised of five transcription factors-- Ikaros, Helios, Aiolos, Eos and Pegasus. Ikaros is expressed in all immune cells. Ikaros has the capacity to dimerize, not only with other Ikaros isoforms, but also with Aiolos or Helios isoforms. Ikaros can act both as a repressor and an activator of gene transcription. Helios can dimerize with itself, as well as with other family members. In the mouse, Helios is detected in the earliest hematopoietic sites of the embryo, in adult hematopoietic stem cells and, with decreasing expression, in the maturing erythroid, macrophage, T and B lineage cells. The Helios gene (Ikzf2) encodes protein isoforms with zinc-finger domains exhibiting considerable homology to Ikaros family proteins. Abundant Helios expression was observed in the thymus with very little expression in bone marrow and brain, and no detectable expression in spleen, liver, kidney or muscle. Human and mouse Helios share 97% identity in their amino acid sequences. We screened a differentially expressed cDNA library from CD4+CD25+ T cells and found that helios was expressed at much higher levels when compared to CD4+CD25- T cells. Real-time PCR confirmed that helios mRNA is expressed at a rate 75-fold greater in CD4+CD25+ T cells than in CD4+CD25- T cells. Helios was cloned into a retroviral vector containing an IRES and GFP expression cassette. CD4+CD25- T cells were infected with the vector in an attempt to confer a regulatory phenotype. However, post-transduction, GFP+ cells could not be detected. The control vector and vectors containing other genes of interest were shown to be infectious and the isolated GFP+ cells expressed the proper genes. We therefore concluded that expression of helios in CD4+CD25- cells lead to cell death. Helios was then cloned into the hCD2 mini-locus expression vector to generate transgenic mice. One founder line was chosen for its high expression of helios mRNA in CD4+CD25+, CD4+CD25- and CD8+ T cells. The percentage and total numbers of these three cell populations are unchanged compared to wild-type (WT) mice. The proliferative capacity of CD4+CD25- and CD8+ cells from these mice is normal; the Treg phenotype of the cells is also intact as they are both anergic and suppressive. Moreover, CD4+CD25- and CD8+ cells do not appear to be suppressive despite the aberrant expression of helios. Strikingly, the helios transgenic mice have profound phenotypes when challenged in vivo. Helios Tg mice are completely resistant to the induction of EAE, while their WT counterparts succumb to an ascending paralysis. Infection of helios Tg mice with T. gondii is lethal while WT mice are able to clear the pathogen in an IFNγ-dependent manner. Finally, helios Tg mice infected with a chronic variant of LCMV have an extremely high virus titer in serum three months after infection, while WT mice have cleared the virus from their serum. (3) We have studied the mechanism of selection of Foxp3+ Treg in the thymus. The role of the TCR in the development of Foxp3+ Treg is highly controversial. Some models suggest that Foxp3+ Treg development requires a signal through a TCR with high avidity for a self-peptide presented in the thymus. The prediction of this model is that the Treg repertoire is restricted to self-reactive TCRs. Other investigators proposed that the increased percentages of Treg in these mice were most likely caused by a decreased susceptibility of Treg to deletion during the process of negative selection. We have isolated and cloned a TCR derived from a thymic-derived Treg and generated a transgenic mouse expressing this Treg TCR. We also bred this mouse onto a RAG-/- background to prevent endogenous TCR rearrangement and limit TCR expression to the Treg-derived TCR. Expression of this TCR as a transgene resulted in marked deletion of both CD4+Foxp3- and CD4+Foxp3+ thymocytes in the transgenic mice on a conventional background; almost complete deletion was seen in transgenic mice on the RAG-/- background. No evidence of preferential usage of the autoreactive TCR by Tregs was observed as a lower proportion (31%) of Foxp3+ cells expressed the the Treg TCR than Foxp3-CD4+ cells (53%). We do not believe this result is consistent with the self-ligand induced instructive model of Treg development. As thymic cellularity is normal, it is likely that the Treg-TCR is specific for a peptide presented by an APC residing in the thymic medulla.
Showing the most recent 10 out of 18 publications
