Therapeutic potential of TGF-β-induced CD4+Foxp3+ regulatory T cells in autoimmune diseases

Abstract

Foxp3+ T regulatory cell (Treg) subsets play a crucial role in the maintenance of immune homeostasis against self-antigen. The lack or dysfunction of these cells is responsible for the pathogenesis and development of many autoimmune diseases. Therefore, manipulation of these cells may provide a novel therapeutic approach to treat autoimmune diseases. In this review, we provide current opinions concerning the classification, developmental and functional characterizations of Treg subsets. A particular emphasis will be focused on the therapeutic role of TGF-β-induced CD4⁺Foxp3⁺ cells (iTregs) in established autoimmune disease. Moreover, the similarity and disparity of iTregs and naturally occurring, thymus-derived CD4⁺CD25⁺Foxp3⁺ regulatory T cells (nTregs) have also be discussed. While the proinflammatory cytokine IL-6 can convert nTregs to IL-17-producing cells, iTregs induced by TGF-β are resistant to the effects of this cytokine. Understanding this difference may play a key role in determining how Tregs can be used in the treatment of established autoimmune diseases.

Introduction

One of the key features of the immune system is its ability to discriminate self antigen from non-self antigens, providing effective immune responses against microbial antigens but preventing deleterious autoimmune responses [1]. Although several mechanisms, such as clonal deletion (physiological deletion of self-reactive lymphocytes) and anergy (functional inactivation of self-reactive lymphocytes), have all been suggested as mechanisms responsible for the immune unresponsiveness to self antigens, there is also a substantial evidence that T cell–mediated active suppression of self-reactive T cells by suppressor or regulatory T (Treg) cells is another essential mechanism of self-tolerance in the periphery [1,2].

Gershon and Kondo first suggested thymocytes contained a population of suppressor T cells in the early 1970s [3]. Other data supporting the existence of a suppressor T cell population in normal thymocytes came from experiments in which neonatal thymectomy (NTx) of normal mice between day 2 and 4 after birth resulted in the development of autoimmune thyroiditis, gastritis, orchitis, prostatitis, and sialoadenitis [4]. Nonetheless, a resurgent interest in Treg cells was restored in 1995 when Sakaguchi’s group found a CD4⁺ subset that constitutively expresses CD25 (IL-2 receptor α-chain) in the thymus and displays the suppressive activity [5]. These thymus-derived, naturally-occurring CD4⁺CD25⁺ (nTreg) cells can migrate to periphery 3 days after birth and constitute 5–10% of peripheral CD4⁺ T cells in normal naïve mice. Removal of CD4⁺CD25⁺ T cells elicited many autoimmune diseases that are similar with that in NTx mice. Reconstitution with normal CD4⁺CD25⁺ T cells within a limited period after NTx prevented the development of autoimmunity [5].

The discovery of Foxp3, a major X-linked forkhead/winged helix transcription factor associated with nTreg development and function, provides an important molecular insight into this essential cell lineage [6–8]. Foxp3 is mainly expressed on endogenous CD4⁺CD25⁺ Treg cells but not on conventionally activated CD4⁺ cells in mouse [1]. A mutation of Foxp3 gene leads to total lack of the CD4⁺Foxp3⁺ Treg cells and development of Scurfy mice that display the most severe form of multi-organ inflammation [9]. The development of Foxp3^gfp “knock-in” mice that express fluorescent proteins under control of the Foxp3 promoter provides an important tool to investigate phenotype and functional characteristics of purified and viable Treg cells since Foxp3 staining cannot be used to isolate living Tregs [9].

CD4⁺CD25⁺ cells also exist in humans although only the CD4⁺CD25^bright cell population display the immune suppressive effect [10]. Recent evidence suggests that although Foxp3 expression in murine lymphocytes appears to be a definitive marker for Treg cells, this may not be the case for human Tregs. New data demonstrate that FOXP3 (the analog of murine Foxp3) may be upregulated in rapidly proliferating human T cells and might be viewed as an activation marker for human T cells [11–13]. Nonetheless, mutations of the human FOXP3 gene cause a disease called IPEX (immune dysregulation, polyendocrinopathy, enteropathy, X-linked syndrome), which is characterized by autoimmune disease in multiple endocrine organs [14–16].

Classification of Treg cells

It is now well recognized that the Treg cell family consists of multiple heterogeneous subsets that include CD4⁺, CD8⁺, CD4⁻CD8⁻ T cells, NK T cells and γδ T cells [17,18]. Among these Treg subsets, CD4⁺ Treg cells are better understood than other Treg cell subsets since they are key players in the Treg cell network and display many important functional characteristics. CD4⁺ Treg subsets can be further classified into two main populations, thymus-derived, natural CD4⁺CD25⁺Foxp3⁺ cells (nTregs) as described above and those that can be induced from CD25⁻ precursors in peripheral lymphoid organs. Induced Treg cells can be developed by the administration of self antigen, peptide or anti-CD3 antibody in vivo [19–22] or induced ex vivo with IL-10 (IL-10-producing type 1 regulatory T cells, Tr1 cells) [23] or with IL-2 and TGF-β (transforming growth factor β) (iTreg) in vitro [24–27]. Induced iTreg cells developed in vivo and Tr1 cells may not express Foxp3 and their role in the autoimmunity has been reviewed elsewhere [28, 29]. Conversely, iTregs induced ex vivo by IL-2 and TGF-β express Foxp3 and share some similar phenotypic and functional characteristics with nTregs although they may also exhibit substantial differences in the development and functionality. In this review, we will compare and contrast nTreg and iTreg cells and focus on the potential therapeutic role of iTreg cells in autoimmune diseases.

Similarities and disparities between nTreg and iTreg cells

Most CD4⁺ nTreg and iTreg cells express CD25, α chain of IL-2 receptor. iTreg cells can be induced from CD4⁺CD25⁻ precursors cells extrathymically [24]. When Foxp3 was identified as a specific hallmark for nTreg cells, studies immediately confirmed that the combination of IL-2 and TGF-β was able to induce CD25⁻Foxp3⁻ precursors to express Foxp3, and with this expression came suppressive activity [25–27]. Both Treg subsets express CTLA-4 (cytotoxic T lymphocyte antigen 4), GITR (gluccorticoid-induced tumor necrosis factor receptor), CCR4 (chemokine receptor) and CD62L (L-Selectin), and most are previously activated cells (i.e. CD45RB^low in the mouse and CD45RO in the human) [30].

Both Treg cells subsets produce little IL-2 and IFN-γ, but they may produce immunosuppressive cytokines such as active TGF-β and IL-10 [24,31]. Both of them also express membrane-bound TGF-β [32,33]. They display a weak proliferative capacity when stimulated with anti-CD3 but the addition of exogenous IL-2 can restore their proliferation [25–27]. Both Treg cell subsets suppress T cell immune response in vitro similarly. They suppress CD4+ effector cell activation, proliferation and cytokine production, as well as CD8+ effector cell activation, proliferation and cytotoxicity activity in vitro.

Whereas the adoptive transfer of nTreg cells can prevent the induction of autoimmune gastritis, colitis, EAE, diabetes, lupus and rescue the multiorgan autoimmune disease in Scurfy mice [7, 34–37], adoptive transfer of iTreg cells similarly suppressed lupus, diabetes, EAE and colitis [38–40]. Although the mechanism of action of Treg cells is still poorly understood, it is generally believed that cell-cell contact is required for the in vitro suppressive activity of both Treg cell subsets, and cytokines such as TGF-β and/or IL-10 are possibly required for the in vivo suppressive activity of nTreg cells [35, 41, 42]. We recently demonstrated that TGF-β and/or IL-10 are also required for the inhibitor effect of iTreg cells on mouse lupus-like syndromes (Zhou XH and Zheng SG, unpublished data). Both Treg cell subsets can also possess the ability to induce conventional CD4⁺CD25⁻ cells to become a new generation of iTreg cells in vitro and in vivo [26, 43–45]. Nonetheless, there are substantial differences in the development and functionality between nTreg and iTreg cells.

A distinction is often made between nTregs that develop in thymus and iTregs that develop extrathymically [46]. Although endogenous CD4⁺CD25⁺Foxp3⁺ cells in the periphery were viewed as nTregs [1], we believe these cells may be a mixture of nTregs and iTregs since infection, tumor, food and allergens can enlarge this cell population [30]. There are no specific phenotypic hallmarks that can distinguish iTregs from nTregs. By definition, CD4⁺CD25⁺Foxp3⁺ cells found in the thymus can be viewed as nTregs.

IL-2 and TGF-β play an essential role in the differentiation and development of iTreg cells. Other TGF-β superfamily members such as Activin A and BMP also have independent or synergistic role with TGF-β in iTreg induction [47–50], but they are redundant for the development of nTreg cells since CD4⁺CD25⁺ cells exist in the thymus of TGF-β/TGF-β receptor deficient mice, or IL-2 and IL-2 receptor deficient mice [51, 52]. However, these cytokines are essential for the maintenance of Foxp3 expression and for the survival of both subsets in vitro and in vivo [47, 52].

Additionally, whereas the CD28/B7 signal pathway is essential for the development of nTreg cells, this signal pathway is not needed for the development of iTregs although this signal pathway promotes the expansion of iTreg cells [53, 54]. In addition, although the lack of a functional CTLA-4/B7 signal pathway markedly interferes with the development of iTregs, CTLA-4 deficient mice have a normal frequency and function of nTreg cells, notwithstanding other reports that CTLA-4 is important for the suppressive activity of nTreg cells [35, 55]. Another member of the CD28 and CTLA-4 receptor family, inducible T cell co-stimulator (ICOS), also favors Treg development [56]. By contrast, Ox40, a member of the tumor necrosis factor (TNF)–TNF receptor (TNFR) superfamily, negatively regulates the development and function of both nTregs and iTregs [57, 58].

nTreg and iTreg cells also differ in their plasticity and stability in an inflammatory milieu. Recent studies revealed that the combination of IL-6 and TGF-β can induce mouse T cells to produce IL-17 [59, 60]. IL-23 produced by immunogenic DCs promotes the survival and expansion of these cells and enables these T cells to become into pro-inflammatory Th17 cells [61]. Although these cells seem to play an important role in immune defense against certain bacterial infections, expansion of these cells is also closely related to the pathogenesis and development of many autoimmune diseases such as multiple sclerosis, rheumatoid arthritis, psoriasis and possibly SLE. The TGF-β/TGF-β receptor signaling pathway is essential for the development of Th17 cells [62], probably via an upregulation of RORγt expression, a transcription factor needed for IL-17 [63]. Interestingly endogenous TGF-β produced by Foxp3⁺CD4⁺CD25⁺ nTreg cells enables IL-6 to convert some of these cells to Th17 thus losing their previous phenotypic characteristics and functional activities [33, 64]. nTreg cells can also be converted into Th2 cells even in a Th1-polarizing environment if Foxp3 expression is attenuated [65]. We also observed that nTregs can be converted into Th1 cells when they were strongly stimulated with anti-CD3 in the presence of IL-6 [62]. More recently, Tsuji M et al reported that the environment of the Peyer’s patches (PPs) down-regulates Foxp3 expression by nTregs and induce them to become CD4+ follicular B helper T(T_FH) which is required B cell function in the germinal centers [66].

Unlike nTreg cells, we recently found that iTregs were completely resistant to the IL-6-driven Th1 and Th17 cell conversion. This resistance cannot be explained by the lack of TGF-β production by iTregs. In fact, we observed that both nTregs and iTregs expressed and produced similar levels of membrane-bound TGF-β and active TGF-β. In addition, we also observed that iTregs are unable to convert into Th1/Th17 cells after exogenous TGF-β was added to the cultures [33]. Recently, we conducted an experiment in vivo to better establish the stability of both Tregs in live animals. When nTregs were adoptively transferred into mice with evident CIA, >50% of these cells in draining lymph nodes lost Foxp3 expression in one week after cell transfer. The majority of Foxp3⁻ cells began to express IL-17 and IFN-γ but not IL-4. Conversely, iTregs cells mostly maintained Foxp3 expression and did not convert into Th1/Th2/Th17 cells [67]. The differences of stability can contribute to their different functionality in autoimmune diseases.

Although both nTregs and iTregs are known to express Foxp3, recent studies have revealed that the Foxp3 gene locus and enhancer in nTregs and iTregs are structurally distinct. Floess et al reported that while complete demethylation of CpG motifs within the foxp3 locus in nTregs was observed, only partial demethylation of CpG motifs was found in iTregs. They also demonstrated that DNA methylation affects Treg stability and suppressive capacity in vitro [68]. More information came from LaI et al, who identified a unique and evolutionarily conserved CpG-rich island of the Foxp3 nonintronic upstream enhancer and discovered that a specific site within it was unmethylated in nTregs but not in iTregs [69]. It is still arguable, however, that methylation of the Foxp3 locus or enhancer affects the stability and functionality of iTregs. We and others have found that iTregs expressed stable Foxp3 if these cells were restimulated with IL-2 and/or TGF-β [47]. Recently, we observed that addition of all-trans retinoic acid (atRA) to TGF-β not only promoted Foxp3 expression, but also sustained Foxp3 expression and survival although atRA did not change the Foxp3 locus methylation [70]. In addition, these groups found that the addition of the DNMT (DNA methyltransferase) inhibitor 5-aza-2’-deoxycytidine (AzaC) can induce demethylation of the Foxp3 locus and enhancer, suggesting that epigenetic regulation can overcome the problem of stability and functionality in iTreg cells. It is noted iTreg cells can be induced in vivo and these iTreg cells displayed both stable Foxp3 expression and demethylated TSDR (Treg-specific demethylated region) [71]. It is unknown why iTregs generated in vitro and in vivo have a different methylation status in TSDR.

Therapeutic role of iTreg cells in autoimmune diseases

Adoptive transfer of nTregs or iTregs can rescue Scurfy mice and prevent many autoimmune diseases [7, 34–37], suggesting that manipulation of these cells may have the potential to treat or cure autoimmune diseases.

To be of use as therapeutic agents for inflammatory and autoimmune diseases, Treg cells must be able to inhibit ongoing T cell responses and reverse established pathology. Although one group showed the therapeutic ability of CD4⁺CD25⁺ nTreg cells to reverse established colitis [72], most studies revealed that the therapeutic effects of nTregs on established autoimmune diseases is fairly unsatisfactory. While adoptive transfer of nTregs can markedly prevent the development of CIA, injection of these cells to established CIA was almost unable to suppress the disease progression [73]. A study from Daikh’s lab observed that the effect of adoptive transfer of in vitro expanded CD4+CD25+CD62L+ cells from (NZBxNZW) F1 (BWF1) lupus-prone mice into BWF1 mice that had developed proteinuria is modest [37]. Additionally, nTregs also did not suppress Th17 cell-mediated autoimmune gastritis (AIG) [75].

There are several reasons may explain for this unsatisfactory effect of iTregs on established autoimmune diseases: Firstly, the low frequency and poor growth make sufficient numbers for adoptive immunotherapy a significant challenge. Although expansion in vitro can overcome this problem, it has been observed that repeated stimulation leads to the diminished Foxp3 expression and the suppressive activity [76]. Secondly, it is very likely that when antigen-specific effector T cells are induced, nTregs can not counteract the effector T cell response and function. For example, T cells from lupus mice and other autoimmune diseases may be resistant to treatment with nTregs although the functional properties of nTregs from these diseases are normal [77]. Additionally, when inflammation is established, a large amount of proinflammatory cytokines, such as IL-6, IL-1 and TNF-α is released. It has been reported that increased levels of IL-6 released by activated DCs abolished the suppressive ability of nTregs [78]. TNF-α possibly has a similar ability to affect the suppressive activity of nTregs [79]. Lastly, it is possible that there is a dynamic alteration of nTregs in an inflammatory milieu. It has been recently reported by Strober’s lab and our group that nTregs can be converted into IL-17-producing cells (Th17) when stimulated with IL-6 [33, 64], we have observed that these nTregs lost Foxp3 expression and suppressive activity when they had been stimulated with IL-6 [33]. As Th17 cells have been demonstrated to be involved in the pathogenesis of many autoimmune diseases, it, therefore, raises a concern that not only does nTreg therapy lack the therapeutic effect on the established autoimmune diseases, but may even worsen the autoimmune diseases since high levels of IL-6 and other pre-inflammatory cytokines have been identified in many autoimmune diseases [80].

Unlike nTregs, several studies have revealed that manipulation of iTregs may have a therapeutic potential on established autoimmune diseases. Adoptively transferred iTreg to EAE mice were active both prophylactically and after priming, suggesting iTreg are highly potent suppressors of autoimmune encephalomyelitis [40]. Similarly, iTregs not only controlled acute onset of T1D elicited by Th1 effectors in SCID recipients and protected NOD mice from spontaneous disease when administered during the preinsulitis phase, but also suppressed the disease progression during the established insulitis phase [39]. Aricha R et al also reported that administration of ex vivo-generated iTreg cells after the induction of experimental autoimmune myasthenia gravis can significantly ameliorate disease manifestations [81]. We have also demonstrated that a single injection of iTregs not only prevented other parental T cells from inducing lymphoid hyperplasia, B cell activation, and an immune complex glomerulonephritis, but also decreased the titer of anti-DNA, suppressed proteinuria, and doubled survival when cells are adoptively transferred to animals that had already developed anti-dsDNA Abs [38]. Others have confirmed this effect [82]. More interestingly, while both nTregs and iTregs can suppress AIG initiated by Th1 and Th2 cells, only Ag-specific iTregs but not nTregs suppressed Th17-mediated AIG [75]. Recently, we found that adoptive transfer of iTregs to established CIA resulted in superior therapeutic effects when compared to nTreg cells [67]. Taken together, these data indicate that iTregs may have an advantageous therapeutic effect on the established autoimmune diseases.

iTregs in mouse and human

While iTregs in mouse have been well studied, it is unclear if similar protocols can generate iTregs in humans. Unlike mouse CD4+ cells, stimulation with polyclonal mitogens without exogenous TGF-β induced transient FOXP3 expression, and endogenous TGF-β produced by CD4⁺ cells or contaminated in culture medium may account for the FOXP3 induction [83]. The addition of TGF-β to the cultures markedly increased FOXP3 expression after one week of culture, however, these cells were not anergic, produced high levels of IL-2, and experienced vigorous proliferation when restimulated with anti-CD3 antibody [83]. These cells were unable to suppress T cell proliferation in vitro and therefore one questions whether IL-2 and TGF-β has ability to induce iTreg cells in humans [83]. Furthermore, it must be remembered that unlike mouse cells, FOXP3 expression is not a specific hallmark for human Treg cells since it is also transiently expressed on activated proliferating human T cells [11–13].

As nTregs and iTregs differ in their principal antigen specificities and strength of T-cell receptor (TCR) stimulation needed for their generation [30], we believe that both TCR stimulation strength and other co-stimulatory factors may affect the differentiation of TGF-β-induced iTregs. In fact, we have previously reported that human CD4+ cells can be induced to become Treg with IL-2 and TGF-β. CD4+ cells stimulated with staphylococcal enterotoxin B and these cytokines become potent suppressors of T cell-dependent IgG production [24]. Additionally, naive CD4⁺ cells stimulated with alloantigens and IL-2 and TGF-β also become Treg that suppress the development of allospecific CTL activity [84].

It is likely that the FOXP3 locus remains methylated in human TGF-β-induced CD4⁺ cells affects their suppressive activity. LaI et at observed that addition of AzaC to TGF-β restored the ability of the latter to induce CD4+ iTregs in humans and they believed that AzaC induces demethylation of Foxp3 gene enhancer, then endows these human CD4+ with suppressive capacity [69]. However, our group recently found that addition of atRA to TGF-β also enabled human CD4⁺ cells to become suppressor cells (Lu L et al. Manuscript submitted). The addition of atRA did not significantly change methylation of FOXP3 locus but markedly promoted CD4⁺ cells to become mature CD4+ cells that expressed high levels of CD45RO and lost CD127 expression although they also increased FOXP3 expression. We therefore believe that after 1 wk these human CD4⁺ cells are only partially differentiated suppressor cells. We observed that upon repeated stimulation, the cells became anergic, and some then expressed membrane-associated TGF-β, and T cell proliferation was strongly suppressed (Wang J and Zheng SG. Unpublished data). Thus, we believe that polyclonal stimulation of human CD4+ cells with IL-2 and TGF-β induces both mouse and human CD4+CD25− cells to become Foxp3⁺CD4⁺CD25⁺ suppressor cells, but the human cells need more time or other cofactors to function in a manner similar to mouse suppressor cells.

Conclusion

We have summarized the similarities and differences of both nTregs and iTregs and highlighted the differences of stability and functionality of Treg cell subsets in the inflammatory environment. While nTregs can be converted into Th1/Th17 cells when stimulated with IL-6 and the therapeutic effects on established autoimmune diseases are unsatisfactory, iTregs are stable and seem to be effective in the treatment of established autoimmune diseases. Although there are different requirements for the induction of iTregs in humans compared to mouse, it is likely that we can expect the future development of human iTregs using TCR stimulation with IL-2, TGF-β and other additives such as AzaC or atRA. These encouraging findings underscore the need to explore the therapeutic potential of iTregs in human immune-mediated diseases. In this approach, iTreg cells are induced from naïve CD4⁺ precursor cells in a patient, enriched, expanded ex vivo, and reinfused. This approach is advantageous because the iTreg product can be analyzed phenotypically and functionally prior to infusion, providing another level of safety. Furthermore, cell dosage can be tightly controlled [85].