Year: 2007
Type: article
Abstract: The transcription factor Foxp3 is generally believed to control all aspects of regulatory CD4+ T (Treg) cell development. Two recent letters to Nature have reported that cells with a Treg cell surface phenotype can develop in the presence of reduced or even absent FoxP3 expression, but lack suppressive activity and may acquire a Th2-effector phenotype. Regulatory circuits are an essential part of the immune system, providing a balance between the need for potent responses against pathogens, protection of tissues against collateral damage and maintenance of self tolerance. T cells with suppressive functions were first described in 1971,1 but it was not until the discovery of reliable surface markers to identify a discrete subset of suppressor or regulatory cells responsible for preventing spontaneous autoimmune disease in mouse models2 that Treg cells assumed their rightful place in the pantheon of T cells. Identification of mutations in the transcription factor Foxp3 in immunoregulatory diseases of both mouse and man led to the elucidation of Foxp3 as crucial for the survival and function of Treg cells.3,4,5,6 The Foxp3 locus is on the X-chromosome, so that disease-causing mutations manifest as a sex-linked condition. Male mice with natural loss-of-function mutations (the ‘scurfy’ mutant)7 or genetically engineered inactivation8 of Foxp3 die at 3–4 weeks of age of lymphoproliferative disease with infiltration of secondary lymphoid organs, liver and skin, accompanied by anaemia and thrombocytopaenia. Mutations of the human Foxp3 gene manifest as a syndrome termed immune dysregulation, polyendocrinopathy, enteropathy and X-linked inheritance (IPEX),3,4,5 which shares many features with the murine scurfy disease.9 In heterozygous females, random X-chromosome inactivation gives rise to two populations of ‘Treg’ cells: bona fide Treg cells that transcribe the wild-type Foxp3 locus, and those that transcribe the mutant Foxp3 locus. In general, heterozygotes do not develop autoimmune disease, suggesting that immune homeostasis can be maintained by selecting fewer than the normal number of functional Treg cells. Consistent with this finding, reconstitution of scurfy or Foxp3 knockout male mice with wild-type Treg cells, either by means of bone marrow reconstitution or mature T-cell transfer, can markedly reduce T-cell proliferation8 and increase lifespan.10 In mice, Foxp3 is first expressed in a small subset of thymic CD4+ single positive cells that express a surface phenotype characteristic of peripheral Treg cells, namely CD25+CTLA-4+. During ontogeny, the appearance of the CD4+CD25+ subset in the thymus precedes Foxp3 expression by several days, and a subset of thymic CD4+CD25+ cells never acquires Foxp3 expression. In humans, on the other hand, a significant proportion of thymic but not peripheral Foxp3+ cells are negative for CD25,11 suggesting some inter-species differences in Foxp3 expression. Although it is believed that the majority of peripheral Treg cells are derived from the thymic Foxp3+ subset, de novo expression of FoxP3 and suppressive function can be induced in mature non-Treg CD4+ cells by a number of exogenous factors including TGF-β12,13 and vitamin D.14 Transfection of Foxp3 also confers regulatory function to naïve mouse8,15,16 and human Foxp3− CD4+ cells.17,18 These data have generally been interpreted as indicating that Foxp3 acts as a simple on–off switch between Treg and conventional T-cell development, a textbook example of a lineage-specifying master-regulatory transcription factor. However, there is mounting evidence that the biology of Foxp3 may not be quite so simple. Several laboratories have reported that expression of Foxp3 is induced, albeit transiently and at relatively low levels, by in vitro activation of human Foxp3-negative cells.19,20,21 Antigen-experienced CD4+ T cells sorted from human peripheral blood express at least 10-fold more Foxp3 mRNA than naïve cord blood CD4+ T cells, but do not suppress in vitro.22 Moreover, Foxp3-dependent anergy and suppressive activity can be dissociated in transfected cells.21 Two recent letters to Nature have shed light on some of the subtleties of Foxp3 function.23,24 Both studies used mouse models in which the endogenous Foxp3 gene locus was modified by inserting a nucleotide sequence encoding a fluorescent protein. Several groups had previously performed the same manipulation while preserving Foxp3 protein expression, so that normal Treg cells were fluorescent.25,26,27 In the models described here, Foxp3 expression was disrupted, to identify and study the fate of cells that underwent thymic selection into the Treg pathway but could not express Foxp3 normally (Figure 1). In the model of Wan and Flavell,23 a gene cassette co-expressing luciferase and enhanced green fluorescent protein (eGFP) was inserted into the 3′UTR of the endogenous Foxp3 locus. eGFP+ cells expressed 5- to 10-fold less Foxp3 protein than normal Treg cells, which the authors suggested may have been due to destabilising of the mRNA by AU-rich elements in the luciferase sequence. This unexpected finding allowed them to study the effect of reduced Foxp3 expression on Treg phenotype and function. Gavin et al.24 knocked GFP into a disrupted Foxp3 locus, so that GFP+ cells were unable to make any Foxp3 protein. In male mice in which all Treg cells transcribed the modified eGFP-encoding locus inserted by Wan and Flavell, lymphoproliferative disease developed with slower kinetics than in mice totally lacking Foxp3 expression (3 months compared with 3–4 weeks). The disease manifestations were similar, with lymphoproliferation accompanied by scaly skin and prominent blepharitis (inflammation of the eyelids characteristically caused by Th2 cells).28 Several aspects of normal Treg biology were preserved in these mice, despite the reduced Foxp3 expression. The number of Treg cells selected in the thymus, the rate of expansion in lymphopenic hosts and the anergic response to in vitro stimulation were all normal. However suppressive activity was lost both in vitro and in vivo. In addition, eGFP+ cells expressed reduced levels of CTLA-4, GITR and, crucially, CD25, so that they competed poorly with wild-type Treg cells after co-transfer to immunodeficient hosts and were underrepresented in female heterozygotes, presumably owing to reduced IL2 signalling. The most interesting abnormality in the eGFP+ cells was their spontaneous differentiation to a Th2 effector phenotype in vivo. Th2 skewing was preserved in vitro and was resistant to the inhibitory effect of IFNγ. It was not seen in female heterozygotes, in which only a very minor increase in cells making IL-2, IL-17 and IFNγ was seen. Th2 skewing in male mice was also seen in the model of Gavin et al., in which Foxp3 was completely absent from GFP+ cells. Thus spontaneous differentiation to a Th2 phenotype does not require Foxp3, but occurs in cells with a selected ‘Treg’ T cell receptor (TCR) repertoire in an environment lacking regulatory function provided in trans (Figure 1). The report of Gavin et al. concentrated on analysis of heterozygous females. Once again, fluorescent cells expressed less CD25, CD44, GITR, CTLA-4 and ICOS than wild-type Treg cells and were relatively underrepresented in heterozygotes. They maintained low expression of IL-7R, a characteristic feature of Treg cell, despite the absence of Foxp3.11,29,30,31 They were anergic to TCR stimulation but had lost suppressive activity in vitro and in vivo. Interestingly, when restrained in vivo by wild-type Treg cells, they spontaneously acquired the ability to produce IL10 and IL-17 but not IL-4, consistent with the data from Wan and Flavell. After transfer of sorted GFP+ cells into lymphopaenic hosts, the majority of cells no longer transcribed the Foxp3/GFP locus, suggesting that Foxp3 positively regulates its own transcription, a finding that was independently confirmed by acute inactivation of Foxp3 in mature Treg cells.32 Gavin et al. went on to study the transcriptome of the GFP+ cells from thymus and peripheral lymphoid organs of female heterozygotes. After performing cluster analysis based on relative expression in GFP+ cells compared with wild-type Treg cells and CD25+FoxP3-CD4+ cells, the largest cluster shared between thymus and periphery was those transcripts selectively upregulated in fully functional Treg cells. A smaller sized cluster showed increased expression in GFP+ cells versus wild-type FoxP3− cells, and a further increase in Treg cells versus GFP+ cells. The authors interpreted these data as indicating that Foxp3 not only initiates a transcription program that is relatively stable between thymus and periphery, but that it enhances the expression of genes already induced within cells destined to enter the Treg pathway. Proteins encoded by the upregulated genes were predominantly located on the cell surface or secreted. Of genes selectively downregulated in Treg cells, PDE3B (cyclic nucleotide phosphodiesterase 3B, cGMP-inhibited) was the most repressed. Transfection of PDE3B into Treg cells caused a decrease in expression of genes encoding mitochondrial and biosynthetic proteins and reversion to a surface phenotype very similar to that of GFP+ cells that could not make Foxp3 protein, suggesting that downregulation of PDE3B is an important consequence of Foxp3 function in cells selected into the Treg pathway in the thymus. In naïve Foxp3− cells, inhibition of PDE3B increased viability during expansion driven by TCR ligation and IL-2. Thus, the downregulation of PDE3B in Foxp3+ cells may be required to prevent apoptosis and sustain proliferation in response to the chronic TCR and IL-2 signals to which self-reactive cells constitutively expressing high-affinity IL-2 receptors would be exposed in vivo. The results presented in these two papers suggest that Foxp3 expression amplifies pre-existing features induced by thymic selection of self-reactive cells and modifies their biology so that they can survive the apoptotic effect of chronic TCR and IL-2 signalling in the periphery. The data also show that the allergic manifestations of Foxp3 deficiency in mice (skin inflammation and blepharitis) and man (eczema) can be attributed to Th2 differentiation of defective Treg cells, which then appear to drive further Th2 differentiation of conventional T cells via IL-4 secretion. In contrast, in diseases caused by regulatory deficiency owing to loss of Treg cells themselves rather than loss of Foxp3 expression, the pathology is more usually the consequence of a Th1-mediated autoimmune attack, caused by spontaneous effector differentiation of conventional T cells rather than Treg cells.33,34 Why cells with a Treg TCR repertoire should be biased towards Th2 differentiation is not entirely clear. There is some evidence that Th2 differentiation in the absence of overt cytokine or microbial signals reflects intermediate rather than high avidity TCR signals,35 and this model is consistent with thymic selection of T cells with intermediate affinity for self in the Treg lineage (Figure 1). However, Th2 effector differentiation was seen only in male mice, that is, in the absence of the regulatory function provided in trans in female heterozygotes. This result suggests that the TCR signals received by self-reactive Treg cells under normal circumstances are of even lower avidity than those that lead to Th2 differentiation.36 How Treg cells maintain immune homeostasis in vivo remains a crucial issue in Treg biology. Although suppression in vitro requires cell–cell contact, in vivo visualisation of Treg cells showed them contacting dendritic cells rather than effector T cells37 By comparing Treg cells with GFP+ cells with no suppressive function, Gavin et al. found a number of Foxp3-dependent effector molecules that could be involved in suppression, including a number of secreted factors, ectonucleotidases, proteins involved in reactive oxygen generation, wound repair and synaptic vesicles. Whether Treg cells use a range of different mechanisms, or whether we are yet to discover their crucial functional mediator(s) remains to be determined. A scheme to explain thymic selection of Treg cells and its consequences for peripheral regulation. The avidity of Treg cells for self antigen is believed to lie between the high avidity that triggers negative selection and the low avidity that leads to positive selection. When cells with appropriate TCRs are selected in the thymus but cannot express Foxp3 normally, their differentiation program is subverted, although not all aspects of Treg function are lost. Importantly, if they respond to self antigen in the periphery in individuals lacking a source of normal Treg cells, they differentiate into Th2 cells that cause the characteristic eczema of the IPEX syndrome.
Source: Immunology and Cell Biology
Author: Barbara Fazekas de St Groth
Institutions: The University of Sydney, Centenary Institute
Language: English
Topic: T-cell and B-cell Immunology
Subfield: Immunology
Field: Immunology and Microbiology
Domain: Life Sciences
Open Access status: bronze
OpenAlex ID: W2019038677
